EXPLORING DESIGN STRATEGIES TO TUNE THE ELECTRONIC STRUCTURE
AND ULTRAFAST DYNAMICS OF IRON(II) POLYPYRIDYL CHROMOPHORES
By
Lindsey Louise Jamula

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Chemistry - Doctor of Philosophy
2013

ABSTRACT
EXPLORING DESIGN STRATEGIES TO TUNE THE ELECTRONIC STRUCTURE
AND ULTRAFAST DYNAMICS OF IRON(II) POLYPYRIDYL CHROMOPHORES
By
Lindsey Louise Jamula
The overall goal of this research project is to develop low cost sensitizers for Dyesensitized solar cells (DSSCs) by moving to first row transition metal-based
chromophores, specifically iron(II) polypyridyl complexes. To be an effective sensitizer
the MLCT state of a chromophore must be sufficiently long lived for injection into the
semiconductor to occur. Unfortunately, iron(II) polypyridyl complexes intrinsically have
short lived MLCT states due to ultrafast deactivation to lower lying ligand field states.
The research has been developed to gain more understanding of the fundamental causes
underlying the ultrafast processes. This study is highly collaborative within our group
and the aim here has been along the synthetic front, providing the means through which
we may begin to tackle this complicated problem.
Iron(II) polypyridyl complexes commonly possess a strained coordination
environment through which orbital degeneracy is removed.

The dense ligand field

manifold that arises from the deviation from octahedral symmetry may be mediating the
charge transfer to ligand field excited state deactivation. We set out to develop a highly
symmetric iron(II) chromophore through which we may begin to investigate this theory.
Our first efforts yielded a complex with a nearly perfect octahedral coordination
environment that exhibits some exciting properties along with a significantly altered
electronic structure.

A series of symmetric iron(II) polypyridyl complexes were prepared to evaluate
the impact of symmetry and substituent effects on the electronic structure and the
synthesis and characterization will be presented. We also made the first step toward
implemeting such a chromophore in a DSSC by introducing anchoring groups to the
periphery of the ligands through which it can bind to a metal oxide semiconductor. We
have not yet developed a viable sensitizer for a TiO2, however we have uncovered some
promising possibilities and the future design strategies that are proposed that may bring
us one step closer.

Copyright by
LINDSEY LOUISE JAMULA
2013

For my wonderful husband Michael

v

ACKNOWLEDGEMENTS

First of all I would like to thank my advisor Jim McCusker for all of the guidance
and support he has provided throughout my graduate career. His enthusiasm for science
is contagious and to be included amongst the talented scientists that make up the
McCusker group has been a great pleasure. I would also like to thank my undergraduate
advisor, Professor Sheila R. Smith at the University of Michigan-Dearborn, for her
guidance and encouragement to pursue a graduate degree. She is an exceptional role
model and presented a research opportunity that sparked my interest in inorganic
chemistry and allowed me to develop the skills necessary to take on graduate research.
I would like to thank every member of the McCusker group, I am blessed to have
had the opportunity to work with all of you.

You have all contributed in some way to

my success which I am grateful for, and consider you all friends and not just colleagues.
Dong, thank you for sharing your synthetic expertise, your guidance has been invaluable.
Allison, without your hard work and passion toward our concerted efforts, my project
could not have gone anywhere. I am so very grateful for your friendship, support, and
your talents that aided in our mutual success.
I would like to thank my entire family for their unwavering support and
encouragement. I cannot put into words how important my family is to me and every bit
of my success stems from their belief in me. Mom and Dad, you instilled the confidence
in me to achieve anything I set out to do and I am so very grateful. Rob and Jessica,
thanks for always looking out for your little sister and despite the fact that I now have
more degrees than you, haha, I know my place in the hierarchy and will always look up

vi

to the both of you. Granny, you are my role model and I strive to be as amazing as you
are. Grandma and Grandpa, I could not have asked for better grandparents and I hope I
make you proud.
To all of my extended family, I feel blessed to have all of you in my life. We cannot
choose our relatives, however I somehow hit the jackpot with you and consider myself
lucky to be part of this family. Speaking of luck, I also hit the jackpot when I met my
husband. Mike, you have brought so much to my life, your love and support throughout
this endeavor was essential and I appreciate you so much. You and Tyler Anne have
brightened up my life enormously and I am so grateful for you both.
I want to acknowledge Brenda, my mentor and friend, for the great person you are
and imparting your work ethic. Jenny, my oldest friend, for always being there through
thick and thin. Lisa, my cohort in this prolonged adventure in graduate school, I could
not have hung in there without you. You have become my closest friend and I am so
appreciative for you. Rick, my friend and personal cupid, thank you for all of the fun
outings and changing my life by introducing me to my husband. Kayla, a great friend
who I can always count on. Drew, the most reliable person I have ever met, you are a
great friend. Kate, Eileen, Jennie, Larry, Dani, Troy, Joel, Adam, Scott, and Brenna for
all of the good times during, and after work that have made all these years in grad school
bearable.

vii

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................. xi
LIST OF FIGURES ......................................................................................................... xiii
LIST OF SCHEMES........................................................................................................ xix
Chapter 1. Introduction to the Design of Iron(II) Based Chromophores ............................1
1.1 Introduction ........................................................................................................1
1.2 Relaxation Dynamics of Iron(II) Complexes .....................................................6
1.3 Design Strategies Toward the Development of Iron(II)-Based Sensitizers .....10
1.4 Contents of Dissertation ...................................................................................13
REFERENCES ..................................................................................................................15
Chapter 2. Introducing a Highly Symmetric Coordination Environment to Iron(II) ........19
2.1 Introduction ......................................................................................................19
2.2 Experimental ....................................................................................................23
2.2.1 Synthesis ..........................................................................................23
2.2.2 Physical Measurements ....................................................................30
2.2.3 Theoretical Calculations ..................................................................31
2.3 Results and Discussion ...................................................................................32
2.3.1 Synthesis of 2,6-Di(2-carboxypyridyl)pyridine ...............................32
2.3.2 Coordination Chemistry of 2,6-Di(2-carboxypyridyl)pyridine ........34
2.3.3 Synthesis of Bis(2,6-di(2-carboxypyridyl)pyridine)iron(II) .............37
2.3.4 Properties of Bis(2,6-di(2-carboxypyridyl)pyridine)iron(II) ............40
2.3.5 Quest for a Control Molecule ...........................................................49
2.4 Concluding Comments.....................................................................................55
APPENDIX ........................................................................................................................57
REFERENCES ..................................................................................................................73
Chapter 3. Development of Symmetric Analogs to Investigate the Influence of the
Bridging Group ..................................................................................................................79
3.1 Introduction ......................................................................................................79
3.2 Experimental ....................................................................................................82
3.2.1 Synthesis ...........................................................................................82
3.2.2 Physical Measurements .....................................................................87
3.2.3 Theoretical Calculations ...................................................................87
3.3 Results and Discussion ....................................................................................88
3.3.1 Significance of the  Interaction in Achieving the Preferred
Geometry...........................................................................................88
viii

3.3.2 Synthesis and Coordination of 2,6-Di(2-(N-methylimine)pyridyl)pyridine .............................................................................................90
3.3.3 Synthesis and Coordination of 2,6-Di(2-vinylidenepyridyl)pyridine .............................................................................................96
3.3.3 Properties of Bis(2,6-di(2-vinylidenepyridyl)pyridine)iron(II) ........98
3.3.4 Continuing Quest for a Control Molecule ......................................103
3.4 Concluding Comments...................................................................................104
APPENDIX ......................................................................................................................106
REFERENCES ................................................................................................................123
Chapter 4. Tuning the Energetics of Excited States by Synthetic Modification of the
Ligand System ..............................................................................................126
4.1 Introduction ....................................................................................................126
4.2 Experimental ..................................................................................................128
4.2.1 Synthesis .........................................................................................128
4.2.2 Physical Measurements ...................................................................133
4.3 Results and Discussion ..................................................................................133
4.3.1 Design of the π-Extended Ligand System ......................................133
4.3.2 Synthesis and Coordination of 2,6-Di(2-carboxypyridyl)-4-ptolylpyridine ....................................................................................135
4.3.3 Preliminary Investigation into the Effects of Extending the
-System .........................................................................................140
4.3.4 Alternative Strategies to Tune the Energetics of Excited States.....143
4.4 Concluding Comments...................................................................................144
APPENDIX ......................................................................................................................146
REFERENCES ................................................................................................................158
Chapter 5. Functionalization of an Iron(II) Chromophore for Attachment to Metal
Oxide Substrates ..............................................................................................................161
5.1 Introduction ....................................................................................................161
5.2 Experimental ..................................................................................................164
5.2.1 Synthesis .........................................................................................164
5.2.2 Physical Measurements ...................................................................169
5.2.3 Theoretical Calculations .................................................................170
5.3 Results and Discussion ..................................................................................170
5.3.1 Installation of Anchoring Groups to 2,6-Di(2-carboxypyridyl)pyridine ...........................................................................................170
5.3.2 Preparation of [Fe(dcpap)2]2+ and [Fe(dcpep)2]2+ ..........................173
5.3.3 Alternative Anchoring Groups for Dye-Sensitization ....................182
5.4 Concluding Comments...................................................................................186
APPENDIX ......................................................................................................................187
ix

REFERENCES ................................................................................................................207
Chapter 6. Concluding Comments and Future Directions ..............................................211
6.1 Concluding Comments...................................................................................211
6.2 Future Directions ...........................................................................................212
APPENDIX ......................................................................................................................221
REFERENCES ................................................................................................................224

x

LIST OF TABLES

Table 2-1.

Crystallographic data for [Fe(dcpp)2](PF6)2 .............................................41

Table 2-2.

Selected bond lengths from the X-ray crystal structure of [Fe(dcpp)2](PF6)2
....................................................................................................................42

Table 2-3.

Selected bond angles from the X-ray crystal structure of
[Fe(dcpp)2](PF6)2 ......................................................................................42

Table 2-4.

Electrochemical potentials for the oxidation and first reduction of
[Fe(bpy)3](PF6)2, [Fe(terpy)2](PF6)2, and [Fe(dcpp)2](PF6)2, measured
with Ag/AgNO3 reference with 0.1 M TBAPF6 electrolyte in acetonitrile,
externally referenced to ferrocene (adapted from reference 2) ..................47

Table 2-5.

Geometric parameters for the X-ray crystal structure of
[Fe(dcpp)2](PF6)2 ......................................................................................62

Table 3-1.

Crystallographic data for [Fe(dvpp)2](PF6)2 .............................................99

Table 3-2.

Selected bond lengths from the X-ray crystal structure of [Fe(dvpp)2](PF6)2
..................................................................................................................100

Table 3-3.

Selected bond angles from the X-ray crystal structure of [Fe(dvpp)2](PF6)2
..................................................................................................................100

Table 3-4.

Electrochemical potentials for the oxidation and first reduction of
[Fe(bpy)3](PF6)2, [Fe(terpy)2](PF6)2, [Fe(dvpp)2](PF6)2, and [Fe(dcpp)2](PF6)2, measured with Ag/NO3 or Ag/AgCl reference with 0.1 M TBAPF6
electrolyte in acetonitrile, externally referenced to ferrocene .................103

Table 3-5.

Geometric parameters for the X-ray crystal structure of
[Fe(dvpp)2](PF6)2 ....................................................................................118

Table 4-1.

Electrochemical potentials for the oxidation and first reduction of
[Fe(dcpp)2](PF6)2 and [Fe(dcptp)2](PF6)2, measured with Ag/AgCl
reference with 0.1 M TBAPF6 electrolyte in acetonitrile, externally
referenced to ferrocene ............................................................................142

xi

Table 5-1.

Crystallographic data for [Fe(dcpep)2](PF6)2..........................................176

Table 5-2.

Selected bond lengths from the x-ray crystal structure of [Fe(dcpep)2](PF6)2
..................................................................................................................176

Table 5-3.

Selected bond angles from the x-ray crystal structure of [Fe(dcpep)2](PF6)2
..................................................................................................................177

Table 5-4.

Electrochemical potentials for the oxidation and first reduction of
[Fe(dcpp)2](PF6)2 and [Fe(dcpep)2](PF6)2, measured with Ag/AgCl
reference with 0.1 M TBAPF6 electrolyte in acetonitrile, externally
referenced to ferrocene. *These values are approximations as hydrolysis
may have occurred and the degree of protonation is unknown. ..............181

Table 5-5.

Geometric parameters for the X-ray crystal structure of [Fe(dcpep)2](PF6)2
..................................................................................................................204

Table 6-1.

Selected bond lengths from the x-ray crystal structure of the tri-nuclear
cryptand [Cu2Fe(5-Lehn-cryptand)]4+ ....................................................217

xii

LIST OF FIGURES

Figure 1-1.

Diagram of the Grätzel cell. Green arrows represent forward processes that
allow the cell to function regeneratively. The red dashed arrows represent
processes that hinder cell function. For interpretation of the references to
color in this and all other figures, the reader is referred to the electronic
version of this dissertation ...........................................................................2

Figure 1-2.

Structures of [Ru(dcbpy)2(NCS)2]2+ (N3) and [Ru(tcterpy)(NCS)3]2+
(black dye) ...................................................................................................4

Figure 1-3.

Representation of the excited states of the sensitizer and TiO2 ...................6

Figure 1-4.

a) Typical potential energy level diagram of Fe(II) based complexes
(adapted from reference 12 and references therein). b) Occupancy of the
ligand field split d orbitals in iron(II) complexes ........................................8

Figure 2-1.

Structures of 2,2’:6’,2”-terpyridine (terpy) and 2,6-di(2-carboxypyridyl)pyridine (dcpp) ...........................................................................................21

Figure 2-2.

Structures of tris(2,2’-bipyridine)iron(II) [Fe(bpy)3]2+, bis(2,2’:6’,2”terpyridine)iron(II)
[Fe(terpy)2]2+, and bis(2,6-di(2-carboxypyridyl)pyridine)iron(II) [Fe(dcpp)2]2+ ..................................................................22

Figure 2-3.

1H

Figure 2-4.

ORTEP Drawing of the cation of [Fe(dcpp)2]2+ obtained from singlecrystal X-ray structure determination. Atoms are represented as 50%
probability thermal ellipsoids. Hydrogen atoms and anions are omitted for
clarity .........................................................................................................41

Figure 2-5.

Structures of [Fe(bpy)3]2+, [Fe(terpy)2]2+, and [Fe(dcpp)2]2+ showing the
coordination environment with approximate bond angles .........................43

Figure 2-6.

Ground state electronic absorption spectrum of [Fe(bpy)3]2+ (black),
[Fe(terpy)2]2+ (red), and [Fe(dcpp)2]2+ (blue) in CH3CN (collected by
Allison Brown, adapted from reference 2) ................................................44

NMR of dcpp in (CD3)2CO (top) and CD3OD (bottom) .....................36

xiii

Figure 2-7.

Results from a TD-DFT calculation on [Fe(dcpp)2]2+ showing the
interligand π stacking interaction between the carbonyl of one ligand with
the pyridyl ring of the other. Also shown is the schematic of the structural
parameters determined from the crystal structure used to characterize the
l.p.—π contact ............................................................................................46

Figure 2-8.

Proposed molecular orbital diagrams of Fe(terpy)2]2+ (left) and
Fe(dcpp)2]2+ (right) showing effects of ligand orbital movement on ligand
field splitting ..............................................................................................48

Figure 2-9.

1H NMR spectrum of [Fe(dcpp) ](PF ) in (CD ) CO ............................58
2
6 2
3 2

Figure 2-10. 13C NMR spectrum of [Fe(dcpp)2](PF6)2 in (CD3)2CO ...........................59
Figure 2-11. ESI-MS [Fe(dcpp)2](PF6)2 ........................................................................60
Figure 2-12.

IR spectrum of [Fe(dcpp)2](PF6)2 ............................................................61

Figure 2-13. 1H NMR spectrum of [Cd(dmpp)2](PF6)2 in (CD3)2CO ..........................65
Figure 2-14. 13C NMR spectrum of [Cd(dmpp)2](PF6)2 in (CD3)2CO .........................66
Figure 2-15. ESI-MS of [Cd(dmpp)2](PF6)2 ..................................................................67
Figure 2-16. IR spectrum of [Cd(dmpp)2](PF6)2 ...........................................................68
Figure 2-17. 1H NMR spectrum of 2,6-di(2-ketalpyridyl)pyridine (dkpp) in (CD3)2CO.
Following recrystallization the spectrum is less resolved, therefore the
spectrum of the crude material is shown....................................................69
Figure 2-18. 13C NMR spectrum of 2,6-di(2-ketalpyridyl)pyridine (dkpp) in
(CD3)2CO ..................................................................................................70
Figure 2-19. ESI-MS of 2,6-di(2-ketalpyridyl)pyridine (dkpp) .....................................71
Figure 2-20. IR spectrum of 2,6-di(2-ketalpyridyl)pyridine (dkpp) ...............................72
Figure 3-1.

Structures of 2,6-di(2-(N-methylimine)pyridyl)pyridine (dmipp) and 2,6di(2-vinylidenepyridyl)pyridine (dvpp) .....................................................80

Figure 3-2.

Optimized Structures of [Fe(dcpp)2](PF6)2 (left), [Fe(dmipp)2](PF6)2
(center), and [Fe(dvpp)2](PF6)2 (right) ......................................................81
xiv

Figure 3-3.

Ground state absorption spectrum of [Fe(dmipp)2]2+ ...............................94

Figure 3-4.

Characteristic shift of the 6-position protons shown by comparison of the
1H NMR ((CD ) CO) of the dcpp ligand (top) and [Fe(dcpp) ]2+ complex
3 2
2
(bottom)......................................................................................................95

Figure 3-5.

1H

Figure 3-6.

ORTEP Drawing of the cation of [Fe(dvpp)2]2+ obtained from singlecrystal X-ray structure determination. Atoms are represented as 50%
probability thermal ellipsoids. Hydrogen atoms and anions are omitted for
clarity. Also shown is the schematic of the structural parameters determined
from the crystal structure used to characterize the π—π interaction .........99

Figure 3-7.

Ground state electronic absorption spectrum of [Fe(dvpp)2]2+ (red), and
[Fe(dcpp)2]2+ (blue) in CH3CN ...............................................................102

Figure 3-8.

1H

Figure 3-9.

13C

NMR ((CD3)2CO) of the [Fe(dmipp)2]2+ complex, note the chemical
shift of the 6-position protons far downfield .............................................96

NMR of [Fe(dmipp)2](PF6)2 in (CD3)2CO ........................................107
NMR of [Fe(dmipp)2](PF6)2 in (CD3)2CO ......................................108

Figure 3-10. ESI-MS of [Fe(dmipp)2](PF6)2 ...............................................................109
Figure 3-11. IR of [Fe(dmipp)2](PF6)2.........................................................................110
Figure 3-12. 1H NMR of 2,6-di(2-vinylidenepyridyl)pyridine (dvpp) in (CD3)2CO...111
Figure 3-13. 13C NMR of 2,6-di(2-vinylidenepyridyl)pyridine (dvpp) in (CD3)2CO .112
Figure 3-14. ESI-MS of 2,6-di(2-vinylidenepyridyl)pyridine dvpp .............................113
Figure 3-15. 1H NMR of [Fe(dvpp)2](PF6)2 in (CD3)2CO ..........................................114
Figure 3-16. 13C NMR of [Fe(dvpp)2](PF6)2 in (CD3)2CO.........................................115
Figure 3-17. ESI-MS of [Fe(dvpp)2](PF6)2 in (CD3)2CO............................................116
Figure 3-18. IR of [Fe(dvpp)2](PF6)2 ...........................................................................117
Figure 3-19. 1H NMR of [Zn(dvpp)2](PF6)2 in (CD3)2CO..........................................121
xv

Figure 3-20. 13C NMR of [Zn(dvpp)2](PF6)2 in (CD3)2CO ........................................122
Figure 4-1.

Structures of desired analogs of 2,6-di(2-carboxypyridyl)-4-p-tolylpyridine
..................................................................................................................139

Figure 4-2.

Ground state electronic absorption spectrum of [Fe(dcpp)2](PF6)2 (blue
dashed) and [Fe(dcptp)2](PF6)2 (red solid) in acetonitrile ......................141

Figure 4-3.

1H

Figure 4-4.

13C

Figure 4-5.

ESI-MS of 4-p-tolylpyridine-2,6-dicarboxylate diethyl ester ..................149

Figure 4-6.

IR of 4-p-tolylpyridine-2,6-dicarboxylate diethyl ester ...........................150

Figure 4-7.

1H

Figure 4-8.

13C

Figure 4-9.

ESI-MS of 2,6-di(2-carboxypyridyl)-4-p-tolylpyridine (dcptp) ..............153

NMR of 4-p-tolylpyridine-2,6-dicarboxylate diethyl ester in CDCl3 147

NMR of 4-p-tolylpyridine-2,6-dicarboxylate diethyl ester in
CDCl3.......................................................................................................148

NMR of 2,6-di(2-carboxypyridyl)-4-p-tolylpyridine (dcptp)
in
(CD3)2CO ................................................................................................151
NMR of 2,6-di(2-carboxypyridyl)-4-p-tolylpyridine (dcptp) in
CDCl3.......................................................................................................152

Figure 4-10. IR of 2,6-di(2-carboxypyridyl)-4-p-tolylpyridine (dcptp) .......................154
Figure 4-11. 1H NMR of [Fe(dcptp)2](PF6)2 in (CD3)2CO .........................................155
Figure 4-12. 13C NMR of [Fe(dcptp)2](PF6)2 in (CD3)2CO........................................156
Figure 4-13. ESI-MS of [Fe(dcptp)2](PF6)2 .................................................................157
Figure 5-1.

Depiction of the three possible binding modes of the carboxylate group to
the titanium dioxide semiconductor, along with the orbital diagram for
ester-type binding mode (adapted from references 2 and 14)..................162

Figure 5-2.

Optimized structure of bis(2,6-di(2-carboxy(4-pyridylcarboxylicacid))pyridine)iron(II), [Fe(dcpap)2](PF6)2, shown in two orientations for
clarity. ......................................................................................................163

Figure 5-3.

ORTEP Drawing of the cation of [Fe(dcpep)2]2+ obtained from singlecrystal X-ray structure determination. Atoms are represented as 50%
xvi

probability thermal ellipsoids. Hydrogen atoms and anions are omitted for
clarity .......................................................................................................175
Figure 5-4.

Ground state absorption spectra of [Fe(dcpep)2](PF6)2 in acetonitrile (left)
and [Fe(dcpap)2](PF6)2 in methanol (right). The normalized absorption
spectra of two crops of [Fe(dcpep)2](PF6)2 clearly show that the spectrum
changes as the ester groups on the complex undergo hydrolysis (—
[Fe(dcpep)2](PF6)2) (- - - hydrolyzed [Fe(dcpep)2](PF6)2).....................179

Figure 5-5.

Band structure of TiO2 compared to the energy level diagram of
[Fe(dcpep)2](PF6)2 showing that the MLCT excited states are lower that the
conduction band of TiO2..........................................................................182

Figure 5-6.

Functional groups that have been investigated for anchoring to a metal oxide
substrate ...................................................................................................183

Figure 5-7.

1H

Figure 5-8.

13C

Figure 5-9.

ESI-MS of 2,6-di(2-carboxy(4-methylpyridyl))pyridine (dcmpp) ..........190

NMR of 2,6-di(2-carboxy(4-methylpyridyl))pyridine (dcmpp) in
(CD3)2CO ................................................................................................188
NMR 2,6-di(2-carboxy(4-methylpyridyl))pyridine (dcmpp) in
(CD3)2CO ................................................................................................189

Figure 5-10. IR of 2,6-di(2-carboxy(4-methylpyridyl))pyridine (dcmpp) ...................191
Figure 5-11. 1H NMR of 2,6-di(2-carboxy(4-pyridylcarboxylicacid))pyridine (dcpap) in
(CD3)2SO .................................................................................................192
Figure 5-12. 13C NMR of 2,6-di(2-carboxy(4-pyridylcarboxylicacid))pyridine (dcpap)
in (CD3)2SO .............................................................................................193
Figure 5-13. ESI-MS
of
2,6-di(2-carboxy(4-pyridylcarboxylicacid))pyridine
(dcpap) .....................................................................................................194
Figure 5-14. IR of 2,6-di(2-carboxy(4-pyridylcarboxylicacid))pyridine (dcpap) ........195
Figure 5-15. 1H NMR of [Fe(dcpap)2](PF6)2 in CD3OD.............................................196
Figure 5-16. 13C NMR of [Fe(dcpap)2](PF6)2 in CD3OD ...........................................197
Figure 5-17. ESI-MS of [Fe(dcpap)2](PF6)2 ................................................................198

xvii

Figure 5-18. IR spectrum of [Fe(dcpap)2](PF6)2..........................................................199
Figure 5-19. 1H NMR of 2,6-di(2-carboxy(4-pyridylcarboxylate ethyl ester))pyridine
(dcpep) in CDCl3 .....................................................................................200
Figure 5-20. 13C NMR of 2,6-di(2-carboxy(4-pyridylcarboxylate ethyl ester))pyridine
(dcpep) in CDCl3 .....................................................................................201
Figure 5-21. ESI-MS of 2,6-di(2-carboxy(4-pyridylcarboxylate ethyl ester))pyridine
(dcpep) .....................................................................................................202
Figure 5-22. 1H NMR of [Fe(dcpep)2](PF6)2 in (CD3)2CO ........................................203
Figure 6-1.

Cryptand ligand structures .......................................................................213

Figure 6-2.

ORTEP Drawing of the cation of [Cu2Fe(5-Lehn-cryptand)]4+ obtained
from single-crystal x-ray structure determination (in progress). Hydrogen
atoms and anions are omitted for clarity ..................................................217

Figure 6-3.

ESI-MS of [Cu2Fe(5-Lehn-cryptand)]4+. [C48H48Cu2F24FeN14P4]4+ m/z
=
251.1;
[C48H48Cu2F24FeN14P4](PF6)2}2+
m/z
=
647.1;
+ m/z = 1439.1 ...................................222
[C48H48Cu2F24FeN14P4] (PF6)3}

Figure 6-4.

Cyclic voltammogram of [Cu2Fe(5-Lehn-cryptand)]4+ measured with a
Ag/AgCl reference with 0.1 M TBAPF6 electrolyte in acetonitrile (50
mV/s), referenced to ferrocene. The presence of the feature at -0.49 V is
attributed to stripping/deposition of copper, therefore the complex degrades
under the present conditions and must be recollected .............................223

xviii

LIST OF SCHEMES

Scheme 2-1. Stepwise synthesis of 2,6-di(2-methylenepyridyl)pyridine (dmpp) and 2,6di(2-carboxypyridyl)pyridine (dcpp) .........................................................33
Scheme 2-2. Alternate synthesis of 2,6-di(2-carboxypyridyl)pyridine (dcpp) ...............34
Scheme 2-3. Equilibrium of dcpp in water or methanol .................................................35
Scheme 2-4. Synthesis of Ru(dcpp)2](NO3)2 (adapted from reference 3) .....................38
Scheme 2-5. Synthesis of [Fe(dcpp)2](PF6)2 ..................................................................39
Scheme 2-6. Synthesis of [Cd(dcpp)2]2+ ........................................................................52
Scheme 2-7. Synthesis of [Cd(dmpp)2](PF6)2................................................................53
Scheme 2-8. Synthesis of [Cd(dcpp)2](PF6)2 via oxidation of [Cd(dmpp)2](PF6)2 ......54
Scheme 2-9. Synthesis of 2,6-di(2-ketalpyridyl)pyridine (dkpp) ...................................55
Scheme 3-1. Attempted synthesis of [Fe(dmpp)2](PF6)2 ...............................................88
Scheme 3-2. General synthetic scheme for the preparation of 2,6-di(2-(N-methylimine)pyridyl)pyridine (dmipp) .................................................................91
Scheme 3-3. Attempted preparation of the desired [Fe(dmipp)2]2+ resulting in the
alternate coordination mode .......................................................................93
Scheme 3-4. Synthesis
of
2,6-di(2-vinylidenepyridyl)pyridine
(dvpp)
and
2+ .............................................................................................97
[Fe(dvpp)2]
Scheme 4-1. Retrosynthetic analysis of substituted 2,6-di(2-carboxypyridyl)pyridine ....................................................................................................136
Scheme 4-2. Preparation of 4-p-tolylpyridine-2,6-dicarboxylate diethyl ester ............136
Scheme 4-3. Preparation of 2,6-di(2-carboxypyridyl)-4-p-tolylpyridine (dcptp) .........138
Scheme 4-4. Preparation of [Fe(dcptp)2](PF6)2 ...........................................................140
Scheme 5-1. Preparation of 2,6-di(2-carboxy(4-methylpyridyl))pyridine (dcmpp) .....171
xix

Scheme 5-2. Preparation
of
2,6-di(2-carboxy(4-pyridylcarboxylicacid))pyridine
(dcpap) .....................................................................................................172
Scheme 5-3. Preparation of 2,6-di(2-carboxy(4-pyridylcarboxylate ethyl ester))pyridine
(dcpep) .....................................................................................................173
Scheme 5-4. Preparation of [Fe(dcpep)2]2+..................................................................174
Scheme 5-5. Preparation of [Fe(dcpap)2]2+..................................................................177
Scheme 5-6. Proposed installation of acetylacetonate (acac) anchoring group ............184
Scheme 5-7. Proposed installation of hydroxamic acid anchoring group ....................185
Scheme 6-1. Preparation of the iron(II) dicopper(I) Lehn cryptand trinuclear complex
[Cu2Fe(5-Lehn-cryptand)]4+ ...................................................................216

xx

Chapter 1. Introduction to the Design of Iron(II)-Based Chromophores

1.1 Introduction
While global energy consumption is ever increasing, the world is failing to put the
global energy system on a more sustainable path.

1

Society relies heavily on fossil fuels

for our energy needs. 40% of energy consumption in the U.S. goes toward electric power
production and with the current infrastructure in place and wide accessibility of coal,
transitioning away from the current system will not be timely or easy.2 The electric
power sector also happens to be the leading source of CO2 emissions in this country.3
With increasing concerns over global climate change we must move toward renewable,
carbon-neutral sources of energy. The sun is the most scalable among the available
sources, providing enough energy in one day to supply the world’s energy needs for one
year.4 Current photovoltaic technology is expensive and cannot compete cost effectively
with fossil fuels, therefore solar energy lies at the bottom of the list of renewable energy
sources in use today.
A cheaper alternative to the commercially available silicon based devices was
developed in 1991 with the introduction of a nanocrystalline titanium dioxide (TiO2)
based dye-sensitized solar cell (DSSC) or Grätzel cell.5

TiO2 is a wide band gap

semiconductor which cannot absorb or make use of visible light and is therefore
sensitized with a chromophore chemically adsorbed to the surface. The TiO2 based
DSSC exhibited a 7% light harvesting efficiency. Dye-sensitization was not a new
concept; the significant advance was the use of nanoparticles of TiO2 which greatly
increased the surface area of the semiconductor. The efficiency of single junction DSSCs
1

improved to approximately 11% where it has held steady for nearly twenty years, despite
much efforts in research. Regardless of the low efficiency, the practicality of these
devices lies in the prospect to utilize inexpensive materials for components of the cell.

Conducting
glass

CB
Counter
electrode

S*/S+

eR/RSo/S+

VB
TiO2

LOAD

Figure 1-1. Diagram of the Grätzel cell. Green arrows represent forward processes that
allow the cell to function regeneratively. The red dashed arrows represent processes that
hinder cell function. For interpretation of the references to color in this and all other
figures, the reader is referred to the electronic version of this dissertation.
The Grätzel cell is depicted in Figure 1-1. The chromophore is chemically bound
to the TiO2 semiconductor which is coated on an indium or fluorine doped tin oxide
(FTO) substrate. Upon absorption of a photon, the sensitizer is promoted to an excited
state that can inject an electron into the conduction band of TiO2 causing the charge
separation. Ideally, electrons will move through the nanocrystalline TiO2 to the back
contact and flow through the cell to the external load. Residual current will flow to the
counter electrode then to the redox mediator in solution where it can re-reduce the
2

oxidized sensitizer.

Unwanted processes also occur in the cell, such as charge

recombination with the oxidized sensitizer or the redox couple.
A considerable amount of the research toward DSSC optimization has involved
the sensitizer. The sensitizer must have high molar absorptivity in the solar spectrum and
be able to bind strongly to the semiconductor.

It must be able to undergo charge

separation, with an excited state above the conduction band in order to inject an electron,
and a low energy highest occupied molecular orbital (HOMO) to be reduced by the redox
couple. The metal-to-ligand charge transfer (MLCT) excited states of d6 coordination
compounds, specifically ruthenium(II) polypyridyl chromophores, make them efficient
for solar harvesting and sensitization of wide-bandgap semiconductors.6 The “N3” and
“black dye,” shown in Figure 1-2, demonstrate good performance in DSSCs and highlight
the charge separation capability of the dye. The polypyridyl ligands have low lying π*
molecular orbitals capable of accepting an electron upon excitation of the complex and
the thiocyanates stabilize the charge on the ruthenium. The carboxylic acid group can
form ester linkages to the TiO2 resulting in strong binding with good electronic
communication between the dye and semiconductor.

3

Figure 1-2. Structures of [Ru(dcbpy)2(NCS)2]2+ (N3) and [Ru(tcterpy)(NCS)3]2+ (black
dye).
The latest developments in structural design of ruthenium photosensitizers were

recently reviewed.7 The strategies employed are too numerous to describe here, but
overall, variations on the standard N3 chromophore have led to much insight on
achieving enhanced light absorption, extending absorption to the near-IR, enhancing
stability, and the understanding of charge recombination. Despite the vast understanding
gained, the best performing ruthenium chromophores to date are similar to the “Grätzel
type” sensitizer.8 Though these ruthenium chromophores have ideal properties for the
sensitization of TiO2, they are a costly component of the cell. Various alternatives have
been investigated, including organic dyes, porphyrins, and phthalocyanines.9 In 1998,
Ferrere and Gregg reported the observation of a photocurrent from an iron(II) based
chromophore, sparking our interest in an iron(II) sensitizer as an ideal substitute due to
the abundance and low cost of first row metals.10 The feasibility of utilizing such a
chromophore is hindered by the intrinsic properties of first row transition metals.

4

Gaining a better understanding of the ultrafast relaxation dynamics of iron(II) complexes
is critical to developing a viable chromophore for DSSCs.

Moving away from

ruthenium(II) based dyes may open the door to the replacement of another costly
component of cell, the platinum counter electrode, by utilizing an alternate redox couple.
We know from ligand field theory that the d orbitals in a transition metal ion are
degenerate. In the presence of an octahedral field, the d orbitals split into non-bonding
(t2g) and anti-bonding (eg*) sets of orbitals. In second and third row transition metal
complexes this octahedral splitting is large resulting in high energy anti-bonding orbitals
regardless of the identity of the ligands. In first row transition metal complexes the
octahedral splitting is not as great resulting in lower energy ligand field states that are
more accessible, therefore the charge separation dynamics are much different.
In order to achieve a photocurrent in a DSSC, an excited state of a chromophore
must lie above the conduction band of TiO2 and must be sufficiently long lived for an
electron to be injected into the semiconductor prior to relaxation to a lower lying state. In
these ruthenium based chromophores, absorption of a photon results in a MLCT excited
state; an electron is promoted from a metal based d orbital into a π* orbital of the ligand.
As shown in Figure 1-3, the MLCT states are the lowest lying excited states, positioned
above the conduction band of TiO2, and are sufficiently long lived (ns) for injection to
occur. In iron(II) complexes, photoexcitation results in promotion of an electron into an
MLCT state followed by ultrafast relaxation into lower lying ligand field states and it is
unclear as to whether efficient injection can occur.

5

LF
1MLCT

1MLCT
CB

3MLCT

3MLCT

LF
TiO2

Fe(II)

Ru(II)
VB

Figure 1-3. Representation of the excited states of the sensitizer and TiO2.

The ligand field states in iron(II) complexes are metal based and thus have
unfavorable electronic overlap with TiO2, therefore injection from an upper lying excited
state prior to any relaxation would be required. Lian and coworkers found evidence for
electron injection from vibrationally hot excited states in ruthenium, iron, and rhenium
complexes.11 The focus of our research is to study the deactivation from the initially
formed excited state in iron(II) complexes with the overall intention of slowing down the
relaxation process to promote injection.
1.2 Relaxation Dynamics of Iron(II) Complexes
As mentioned above, we know from ligand field theory that the d orbitals in a
transition metal ion are degenerate. In the presence of an octahedral field, the d orbitals
6

split into non-bonding (t2g) and anti-bonding (eg*) sets of orbitals. For d4-d7 first row
transition metal ions, oxidation state of the metal and the ligand field strength will
determine the ground state. In the presence of a strong ligand field, octahedral splitting is
large and the low energy conformation is achieved by pairing spins, while in the presence
of a weak field ligand, the high spin state is lower in energy. Iron(II) polypyridyls, such
as tris(2,2’-bipyridine)iron(II) and bis(2,2’:6’,2”-terpyridine)-iron(II) complexes are
approximated as possessing octahedral symmetry and exist in the low spin 1A1 state,
where there are six electrons in the t2g nonbonding orbitals. In the 5T2 high spin state,
there are four electrons in the t2g nonbonding orbitals and two in the eg* σ-antibonding
orbitals. The formation of the 5T2 state is accompanied by a 0.2 Å elongation of the FeN bond. The occupancy of the 1A1 and 5T2 states are shown in Figure 1-4, along with a
typical energy level diagram of octahedral iron(II) complexes. The diagram is general to
the family of iron(II)-based systems with the ordering of the ligand field states as
calculated from the Tanabe-Sugano diagram for d6 octahedral complexes and is not
quantitative with absolute energies.

7

a)

1

MLCT

1

b)
T2

5T2 High Spin

1

T1

E
n
e
r
g
y

eg

3

MLCT
t2g

3

T2

3

T1
5

T2

1

A1

1A1 Low Spin
eg
t2g

0

0.1

0.2

∆R Fe-N (Å)

Figure 1-4. a) Typical potential energy level diagram of Fe(II) based complexes
(adapted from reference 12 and references therein).12 b) Occupancy of the ligand field
split d orbitals in iron(II) complexes.
Iron(II)

polypyridyls

have

been

widely

investigated

chemically

and

photophysically. It is well established that upon photoexcitation, the low spin complex
undergoes a metal to ligand charge transfer, followed by ultrafast deactivation to a long
lived ligand field state. It has been shown that this long lived ligand field state is the 5T2
state and ultrafast electronic absorption measurements reveal sub-picosecond dynamics
associated with the formation of the 5T2 state.13

Further work using stimulated

resonance Raman and femtosecond transient absorption experimentally verified that the

8

5T
2

state is established in 200 fs.14 The mechanism for the rapid conversion (S = 2)

was thought to proceed via 1A1 → 1MLCT → 5T2. It has also been determined that rapid
deactivation from the charge-transfer manifold occurs in less than 100 fs, which appears
to be characteristic of the whole class of low spin iron(II) polypyridyl complexes.15 Due
to the short time scale, the mechanism mediating the deactivation to the transient 5T2
state is still unclear. Much more is known about the conversion from the long lived
excited state to ground state since the ligand field excited state lasts for hundreds of
picoseconds to hundreds of nanoseconds, depending on the ligands. Also, the dynamics
of iron(II) spin-crossover complexes have received a lot of attention in recent years, with
much interest stemming from potential applications in molecular switches and display
devices.16–18
Ultrafast optical studies by Chergui and coworkers have provided important
mechanistic information on the excited state evolution of tris(2,2’-bipyridine)iron(II).12,19,20 The observation of two distinct emission features, along with other
evidence, allowed them to conclude that the first event is ultrafast intersystem-crossing
(ISC) from the 1MLCT to the 3MLCT (<30 fs) with strong mixing between the 1MLCT
and 3MLCT states, followed by ISC to the 5T2 state. A computational study by de Graaf
et al. provided additional evidence for their findings from the analysis of the energy
dependence of electronic states on the changes in geometry associated with the LS-HS
spin state conversion.21 Consequently, the currently recognized mechanism of excited
state evolution for iron(II) polypyridyls is 1A1 → 1MLCT → 3MLCT → 5T2.

9

1.3 Design Strategies Toward the Development of Iron(II)-Based Sensitizers
As mentioned above, a photocurrent has been observed from and iron(II)-based
chromophore,

specifically

[Fe(dcpby)2(CN)2]

(Fe(2,2’-bipyridine-4,4’-dicarboxylic

acid)2(CN)2).10 As with other iron(II) polypyridyls, upon photoexcitation to a charge
transfer state it undergoes ultrafast relaxation (<100 fs) to the long lived ligand field state
that is localized on the metal. Electron injection is not possible from this ligand field
state as there is unfavorable overlap and it lies below the conduction band of TiO2.11
Injection had to be occurring from an initially formed MCLT state and the quantum yield
for injection was determined to be ~12% following excitation into the low energy charge
transfer band.22 We can think about this injection efficiency in terms of a competition of
rates: the rate of injection versus rate of relaxation. Therefore, there are two approaches
we could take toward developing a viable iron(II) sensitizer: 1) speed up the electron
injection process, or 2) slow down the relaxation process.
The first approach is quite a tall order, as it seems that conditions are just right to
promote injection. In Fe(dcpby)2(CN)2, the carboxylate anchoring groups are electron
withdrawing and are located in the 4-position of the dye, which is along the transition
dipole which should mediate the electron transfer process. Furthermore, carboxylate
groups provide good electronic communication with the TiO2 since there is orbital
overlap between the extended π system of the carboxylate and the d orbitals of titanium.
Recent theoretical work by Jakubikova and coworkers has shown that injection from this
specific molecule will vary depending on the binding mode, but ranges between 0.3-7.3
ps through a carboxylate linker.23 It is possible that alternate anchoring groups could
decrease this rate, however we believed we had better chances taking on the second
10

approach of slowing down the relaxation rate, which has taken us down two distinct
paths.
Our initial approach to investigate the mechanism of the 1MLCT → 5T2
deactivation was inspired by what is known for the 1A1 ↔ 5T2 spin state interconversion.
It is well known that population of eg* σ-antibonding orbitals results in an increase in
metal – ligand bond length when a complex converts from the 1A1 to the 5T2 state. This
observation leads to the possibility that spin state interconversion is coupled to the
vibrational modes of the complex. The hypothesis originated from the work by Purcell
and Vanquickenborne who established a connection between enantiomerization of d6
complexes and spin state interconversion.24,25 McCusker et al. proposed that torsional
modes play an essential role in the kinetics of spin state interconversion between 5T2 ↔
1A
1

states and found evidence that suggests there is a correlation.26 In attempts to

determine a deactivating coordinate of the MLCT excited state, we set out to investigate
the extent to which a torsional coordinate might modulate MLCT to ligand field kinetics
by introducing steric bulk to the terminal rings of 2,2’:6’,2”-terpyridine (terpy) and
preparing the bis-tridentate iron(II) complexes.27

The terpy series yielded a lot of

information on the correlation between charge transfer deactivation and ground state
recovery, but did not significantly slow the charge transfer deactivation, therefore we set
out on an alternative strategy.28
In general, iron(II) polypyridyls are approximated to possess octahedral
symmetry, while in reality the ligand environments are commonly strained, reducing the
coordination symmetry.

Research from our collaborative efforts with Huse and

coworkers has shown that varying chemical composition and ligand–field symmetry
11

influence the valence charge density of ground and excited states.29–31 Symmetry effects
could also influence the ultrafast dynamics of the charge transfer deactivation. Deviation
from octahedral symmetry will remove the orbital degeneracy and increase the number of
states shown in the energy diagram in Figure 1-4, and the increased density of states may
be mediating the ultrafast relaxation. In an effort to minimize the number of these state
that we believe to be mediating the deactivation, the goal of the present investigation was
to introduce a more symmetrical coordination environment to an iron(II) center.
Most of the work presented herein stems from the second strategy, on moving
toward a more symmetrical environment we stumbled on a molecule with some exciting
properties. The symmetric coordination environment did not yield the anticipated effect
of slowing down the charge transfer deactivation process, however it appeared to effect
significant changes to the electronic structure. This observation led us down a new path
of examining ways to tune the electronic structure even further through synthetic
modification of the ligand system. There is a chance that this new system has presented
us with the opportunity to make a first row transition metal complex behave more like a
second row metal complex, which would have major implications for use in DSSCs. We
have also taken the next step toward implementation of an iron(II)-based chromophore
into a DSSC by installing anchoring groups to this new system.
Though much of our efforts have been focused on this new, ever evolving
strategy, we have not abandoned our original approach of investigating vibrational
modes. Following up on the terpy series, which involved one torsional coordinate, we
were interested in locking the complex in a rigid ligand framework to minimize all
movement: the torsional modes as well as the known Fe-N bond elongation that

12

accompanies the charge transfer to ligand field transition. In the last chapter I will
introduce our current work along this front based on Lehn-type cryptand ligand
systems.32,33
1.4 Contents of Dissertation
The focus of the research described in this dissertation involves the design,
synthesis, and characterization of a series of highly symmetrical iron(II) polypyridyl
complexes for the analysis of the influence of coordination environment and substituent
effects on the electronic structure. Chapter 2 presents the preparation of a new iron(II)
complex possessing nearly perfect octahedral symmetry. The interesting coordination
chemistry of the highly versatile carbonyl bridged ligand is highlighted and discussed in
detail. The new complex possesses some very exciting properties which are examined
through a comparison with two pseudo-octahedral complexes.

Design and synthetic

efforts toward the preparation of an isostructural control molecule are also discussed.
Chapter 3 introduces the design of two additional symmetric iron(II) complexes.
We set out to replace the carbonyl bridges with alternate functional groups in order to
investigate whether the interesting properties discussed in Chapter 2 arise from the
symmetrical geometry or from inductive effects from the bridging groups of the ligand.
The development of the synthesis of the ligands and corresponding iron(II) complexes is
discussed.
Chapter 4 lays the groundwork for an investigation into determining ways to
effectively tune the energetics of the excited states of these new symmetric iron(II)
complexes. The design of a series of molecules in which substituents are introduced to
the periphery of the ligands in an effort to modulate the physical properties will be
13

presented. The first of this new series has been successfully prepared and will be
discussed.
Chapter 5 moves the discussion away from the fundamentals from the previous
chapters and moves toward the overall application of the new iron(II) chromophores.
This section details the installation of anchoring groups through which these new
chromophores may bind to a semiconductor. As a starting point we have successfully
introduced the most commonly employed carboxylate anchoring group to our new ligand
system and the synthesis and characterization will be discussed. A survey of other
functional groups that may be utilized, along with proposed synthetic methods toward
their development will be presented.
Chapter 6 introduces a new system that we have just begun to investigate. As a
follow up to the study of the correlation of torsional modes with ultrafast deactivation
that was described in my Masters Thesis, we set out to design iron(II) polypyridyl
complexes with fixed cage ligand structures.27 A Lehn-type cryptand was employed to
prepare a tri-nuclear assembly which may possess interesting redox properties and the
synthesis and preliminary characterization will be presented.

14

REFERENCES

15

REFERENCES

(1)

International Energy Agency World Energy Outlook, Executive Summary; 2012.

(2)

U.S. Energy Information Administration Annual Energy Review; 2012.

(3)

Environmental Protection Agency Inventory of U.S. Greenhouse Gas Emissions
and Sinks: 1990-2011; 2013.

(4)

Report of the Basic Energy Sciences Workshop on Solar Energy Utilization;
Department of Energy: Washington DC, 2005.

(5)

O’Regan, B.; Grätzel, M. Nature 1991, 353, 737–740.

(6)

Ardo, S.; Meyer, G. J. Chemical Society Reviews 2009, 38, 115–64.

(7)

Yin, J.-F.; Velayudham, M.; Bhattacharya, D.; Lin, H.-C.; Lu, K.-L. Coordination
Chemistry Reviews 2012, 256, 3008–3035.

(8)

Nazeeruddin, M. K.; Kay, A.; Miiller, E.; Liska, P.; Vlachopoulos, N.; Gratzel, M.;
Lausanne, C.-; April, R. 1993, 115, 6382–6390.

(9)

Clifford, J. N.; Martínez-Ferrero, E.; Viterisi, A.; Palomares, E. Chemical Society
Reviews 2011, 40, 1635–46.

(10)

Ferrere, S.; Gregg, B. A. Journal of the American Chemical Society 1998, 120,
843–844.

(11)

Asbury, J. B.; Wang, Y.-Q.; Hao, E.; Ghosh, H. N.; Lian, T. Research on
Chemical Intermediates 2001, 27, 393–406.

(12)

Cannizzo, a.; Milne, C. J.; Consani, C.; Gawelda, W.; Bressler, C.; van Mourik, F.;
Chergui, M. Coordination Chemistry Reviews 2010, 254, 2677–2686.

(13)

McCusker, J. K.; Walda, K. N.; Dunn, R. C.; Simon, J. D.; Madge, D.;
Hendrickson, D. N. J. Am. Chem. Soc. 1993, 115, 298.

(14)

Smeigh, A. L.; Creelman, M.; Mathies, R. A.; McCusker, J. K. Journal of the
American Chemical Society 2008, 130, 14105–7.

(15)

Juban, E. a.; Smeigh, A. L.; Monat, J. E.; McCusker, J. K. Coordination Chemistry
Reviews 2006, 250, 1783–1791.

16

(16)

Hauser, A.; Enachescu, C.; Daku, M. L.; Vargas, A.; Amstutz, N. Coordination
Chemistry Reviews 2006, 250, 1642.

(17)

Gütlich, P.; Koningsbruggen, P. J. Van; Renz, F. Structure and Bonding 2004,
107, 27–75.

(18)

Halcrow, M. A. Polyhedron 2007, 26, 3523–3576.

(19)

Gawelda, W.; Cannizzo, A.; Pham, V.-T.; Mourik, F. van; Bressler, C.; Chergui,
M. J. Am. Chem. Soc. 2007, 129, 8199.

(20)

Bressler, C.; Milne, C.; Pham, V.-T.; Elnahhas, a; van der Veen, R. M.; Gawelda,
W.; Johnson, S.; Beaud, P.; Grolimund, D.; Kaiser, M.; Borca, C. N.; Ingold, G.;
Abela, R.; Chergui, M. Science (New York, N.Y.) 2009, 323, 489–92.

(21)

De Graaf, C.; Sousa, C. Chemistry - A European Journal 2010, 16, 4550–4556.

(22)

Ferrere, S. Chemistry of Materials 2000, 12, 1083–1089.

(23)

Bowman, D. N.; Blew, J. H.; Tsuchiya, T.; Jakubikova, E. Inorganic Chemistry
[Online Early Access] DOI:10.1021/ic4007839 Publication Date (Web): July 9,
2013.

(24)

Purcell, K. F. Journal of the American Chemical Society 1979, 101, 5147–5152.

(25)

Vanquickenborne, L. G.; Pierloot, K. Inorganic Chemistry 1981, 20, 3673–3677.

(26)

McCusker, J. K.; Rheingold, A. L.; Hendrickson, D. N. Inorganic Chemistry 1996,
35, 2100–2112.

(27)

Jamula, L. L. Masters Thesis, Michigan State University, 2010.

(28)

Brown, A. M. PhD Dissertation, Michigan State University, 2011.

(29)

Huse, N.; Kim, T. K.; Jamula, L.; McCusker, J. K.; de Groot, F. M. F.; Schoenlein,
R. W. Journal of the American Chemical Society 2010, 132, 6809–16.

(30)

Huse, N.; Cho, H.; Hong, K.; Jamula, L.; Groot, F. M. F. De; Kim, T. K.;
Mccusker, J. K.; Schoenlein, R. W. Journal of Physical Chemistry Letters 2011, 2,
880–884.

(31)

Cho, H.; Strader, M. L.; Hong, K.; Jamula, L.; Gullikson, E. M.; Kim, T. K.; de
Groot, F. M. F.; McCusker, J. K.; Schoenlein, R. W.; Huse, N. Faraday
Discussions 2012, 157, 463.

17

(32)

Rodriguez-Ubis, J.-C.; Alpha, B.; Plancherel, D.; Lehn, J.-M. Helv. Chim. Acta
1984, 67, 2264.

(33)

De Mendoza, J.; Mesa, E.; Rodríguez-Ubis, J.-C.; Vázquez, P.; Vögtle, F.;
Windscheif, P.-M.; Rissanen, K.; Lehn, J.-M.; Lilienbaum, D.; Ziessel, R.
Angewandte Chemie International Edition in English 1991, 30, 1331–1333.

18

Chapter 2. Introducing a Highly Symmetric Coordination Environment to Iron(II)
2.1 Introduction
Ruthenium(II) polypyridyl complexes are commonly employed as chromophores
in DSSCs.

Though these ruthenium chromophores have ideal properties for the

sensitization of TiO2, they are a costly component of the cell.

An iron(II) based

chromophore is an ideal alternative due to the abundance and low cost of first row metals.
The efficacy of an iron(II) chromophore is hindered by the ultrafast relaxation to ligand
field states, which is in competition with injection into the conduction band of the
semiconductor. In order to overcome this challenge, we are investigating the mechanism
of the ultrafast deactivation with the ultimate goal of slowing down this process.
Our initial approach involved the introduction of steric bulk to the terminal rings
of 2,2’:6’,2”-terpyridine (terpy) and preparation of the bis-tridentate iron(II) complexes to
investigate a torsional deactivation pathway.1 The steric tuning had a minimal effect on
the relaxation rate leading us to pursue a new strategy.2 The deviation from a truly
octahedral coordination environment of the planar terpy ligands may be facilitating the
deactivation to the lower lying ligand field states.

Tris(2,2’-bipyridine)iron(II)

([Fe(bpy)3]2+) and bis(2,2’:6’,2”-terpyridine)iron(II) ([Fe(terpy)2]2+) are approximated to
possess octahedral symmetry, but the strained ligands reduce the symmetry to D3 and C2,
respectively.

We propose that the reduced symmetry of these complexes may be

mediating the charge transfer deactivation. Therefore, the goal of the present research is
the design and preparation of highly symmetric iron(II) chromophores. In designing such
a system, it is our intent to remove ligand restraints in order to achieve the desired

19

symmetry while still following the criteria necessary to develop a viable sensitizer for
DSSCs.
To attain the desired symmetry the most obvious first choice would be
Hexapyridineiron(II) ([Fe(py)6]2+), which is known and was structurally characterized
nearly fifty years ago.3

From the X-ray crystal structure investigation it becomes

immediately apparent that this complex is not a viable option. The complex achieves the
desired octahedral symmetry, however the Fe-N bond lengths are greater than 2.2 Å
signifying a high spin ground state. Furthermore, the complex is off-white and is not
stable in solution, losing a pyridine in exchange for a solvent molecule in acetonitrile.4
Overall, high spin iron(II) complexes lack the stability and molar absorptivity necessary
to meet the criteria for a sensitizer.
The inspiration for the molecule in this investigation arose from research efforts
toward improving the photophysical properties of bis-tridentate ruthenium(II) complexes.
[Ru(terpy)2]2+ has structural advantages over [Ru(bpy)3]2+ but suffers from poor excited
state properties. The 3MLCT of [Ru(terpy)2]2+ is relatively short lived (250 ps) and the
complex has uncharacteristically weak emission at room temperature.5 These anomalous
features are attributed to the accessibility of the metal centered states. Unfavorable bite
angles of the mer coordinated terpy leads to a weak ligand field which stabilizes the
metal centered states, making them thermally accessible. There is much interest in the
pursuit of a more octahedral coordination environment by expanding the bite angles to
achieve a stronger ligand field and destabilize the metal centered states, rendering them
less accessible.6

Although our motivations for moving toward a more symmetrical

20

environment differ, ours being to collapse the ligand field manifold into more discrete
states, the expansion of the bite angles of tridentate ligands is an ideal path to pursue.
Recent work by Schramm et al. introduced an octahedral ruthenium(II)
polypyridyl complex with unprecedented photophysical properties.7 The expansion of
the coordination environment of a terpy-like ligand resulted in improved bite angles and
the formation of a complex that has a significantly extended excited state lifetime, a
higher room temperature quantum yield, and remains functional in aerated conditions.
The synthesis of the ruthenium(II) complex is complicated by the nature of the ligand.
Despite some synthetic hurdles that the preparation of the analogous iron(II) complex
would potentially entail, it seemed an ideal choice for this investigation.
The ligand of interest in this study, 2,6-di(2-carboxypyridyl)pyridine (dcpp), was
first introduced by Abarca et al. in 1998, while the coordination chemistry was not
investigated until 2005.8,9 The ligand is essentially an extended version of the widely
studied and commercially available di(2-pyridyl)ketone (dpk), which has been known
and well described in the literature since it was first prepared in the early 1950s.10
Unlike the rigid terpy ligand, the expanded dcpp has the flexibility to coordinate in a
more symmetrical environment (Figure 2-1).

Figure 2-1. Structures of 2,2’:6’,2”-terpyridine (terpy) and 2,6-di(2-carboxypyridyl)pyridine (dcpp).
21

The dcpp ligand is highly sensitive to reaction conditions, therefore it rarely
remains intact upon coordination to a metal center. In most cases the highly reactive
carbonyl bridge is transformed in solution resulting in a plethora of binding possibilities.
The versatility of the ligand makes it attractive to a wide range of research areas,
including the preparation of polynuclear clusters, single molecule magnets, and
coordination polymers.9
While the reactivity of dcpp was of great concern at the outset of this study, the
target octahedral iron complex, bis(2,6-di(2-carboxypyridyl)pyridine)iron(II) has been
successfully prepared and thoroughly investigated. The consequences of moving toward
more symmetrical coordination environment are examined through a comparison of the
new iron(II) complex with the well-known pseudo-octahedral tris(2,2’-bipyridine)iron(II)
and bis(2,2’:6’,2”-terpyridine)iron(II) complexes (Figure 2-2).

The synthesis and

characterization will be fully described here, while further discussion on the photophysics
and implications were examined by Allison Brown in her PhD Dissertation.2

Figure 2-2. Structures of tris(2,2’-bipyridine)iron(II) [Fe(bpy)3]2+, bis(2,2’:6’,2”terpyridine)iron(II) [Fe(terpy)2]2+, and bis(2,6-di(2-carboxypyridyl)pyridine)iron(II)
[Fe(dcpp)2]2+.
22

In an effort to gain more insight on the properties of the new iron(II) complex, it
is necessary to prepare a control molecule. An isostructural d10 molecule would aid in
distinguishing properties that may be attributed to the complex from ligand based effects.
Analysis of the ligand alone would be inadequate due to the free lone pairs on the
nitrogen atoms as well as the vastly different arrangement that the dcpp assumes on upon
coordination. Attempts to isolate either a zinc(II), gallium(III), or cadmium(II) analog
have been unsuccessful thus far. Unlike with the complexation of the iron(II) compound,
the reactivity of the dcpp appears to be detrimental to the formation of these complexes.
Many different routes employing a variety of reagents and conditions were pursued and
will be discussed.
2.2 Experimental
2.2.1 Synthesis
General. All chemicals were of reagent grade, purchased from Sigma-Aldrich,
Alfa Aesar, Acros Organics, or Strem Chemicals and used as received unless otherwise
noted.

Solvents were purchased from Sigma-Aldrich, Jade Scientific, Spectrum,

Mallinckrodt, EMD Chemical, or CCI and were purified using standard purification
techniques.11 All air-sensitive reactions were carried out under inert atmosphere by
standard Schlenk techniques utilizing thoroughly deoxygenated solvents that were
degassed by the freeze—pump—thaw method. 1H NMR and 13C NMR were recorded
with Varian UnityPlus-300 MHz, Varian UnityPlus-500 MHz, and Agilent DDR2 500
MHz spectrometers. Ground state absorption spectra were obtained on a Varian Cary 50
spectrophotometer.

IR spectra were obtained on a Mattson Galaxy 5000 FTIR.

23

Elemental analysis was obtained through the Analytical facilities at Michigan State
University and Columbia Analytical Services (Tucson, AZ). Electrospray mass spectra
(ESI-MS) were obtained from the staff of the MSU Mass Spectrometry Facility. The
characterization data of previously unknown compounds are included in the appendix at
the end of this chapter.
2,6-Di(2-methylenepyridyl)pyridine (dmpp). The compound was prepared
according to the published procedure from 2-picoline (14.9 g, 160 mmol) and 2,6difluoropyridine (2.30 g, 20 mmol).12

The crude product was purified by vacuum

distillation yielding a yellow oil. Yield: 4.96 g (95%). The oil does not crystallize upon
standing as described in the literature until a KOH pellet is added to the product to dry it
promoting crystallization. mp 46oC. 1H NMR (500MHz, CDCl3) δ (ppm): 8.53 (ddd,
2H, 6-H of the pyridyl arms (py-a)), 7.56 (td, 2H, 4py-a), 7.50 (t, 1H, 4-H of the bridging
pyridyl (py-b)), 7.23 (dt, 2H, 3py-a), 7.09 (dtd, 2H, 5py-a), 7.05 (d, 2H, 3/5py-b), 4.33 (s,
4H).
2,6-Pyridinedicarboxylate diethyl ester. The compound was prepared from 2,6pyridinedicarboxylic acid (12.0 g, 71.8 mmol) according to the literature procedure.13
The crude product is obtained as a colorless oil which does not readily crystallize. The
product is precipitated by dissolution of the oil in minimal ethanol and flash freezing by
submersion in liquid nitrogen.

The solid must be filtered quickly to avoid

melting/redissolution. Once in the solid form the compound may be recrystallized by
dissolution in ethanol at room temperature and placing in the freezer. The hygroscopic
solid must be stored in a vacuum desiccator to maintain integrity. Yield: 11.2 g (70%).
1H

NMR (CDCl3), δ (ppm): 8.31 (d, 2H); 8.04 (t, 1H); 4.5 (q, 4H); 1.47 (t, 6H).
24

2,6-Di(2-carboxypyridyl)pyridine (dcpp). This compound was prepared via
two different routes, by modification of previously published procedures.
Method A.14 To a 100 mL round bottom flask equipped with a reflux condenser were
added dmpp (0.510 g, 1.95 mmol), excess selenium dioxide (1.35 g, 12.2 mmol), and
glacial acetic acid (50 mL) and the solution was refluxed for 8 hours. The reaction
mixture was filtered hot through celite to remove selenium and the solvent evaporated in
vacuo. The brown residue was dissolved in CHCl3 and washed with saturated NaHCO3
(aq). The product was purified by column chromatography on silica with 2% EtOH in
CHCl3 and recrystallized by dissolving by sonication in EtOAc at room temperature and
adding hexanes till cloudy (~1:1) and cooling to 0oC. Yield 0.176 g (30%).
Method B.15 To a 250 mL round bottom air-free flask were added 2-bromopyridine (3.12
g, 19.7 mmol) and dry THF (~80 mL) via cannula. Under nitrogen, the solution was
cooled to -78oC in a dry ice/acetone bath. 1.6 M n-BuLi solution in hexanes (12.3 mL,
19.7 mmol) was added dropwise over 30 minutes and the solution stirred for an
additional 30 minutes.

To a separate air free flask was added diethyl 2,6-

pyridinedicarboxylate (2.01 g, 8.96 mmol) and dry THF (~20 mL) and the solution
transferred dropwise via 22 gauge cannula to the reaction mixture over 40 minutes. The
solution was stirred for an additional 2 hours at -78oC before quenching with methanol
(20 mL) and allowing to warm to room temperature and stir overnight. 10% HCl (40
mL) was added and the organic layer removed. The acidic aqueous layer was washed
with CH2Cl2 (1 x 50 mL) and then basified with 5 M NaOH. The basic aqueous layer
was then extracted with CH2Cl2 (3 x 100 mL). The organic layer dried with MgSO4,
filtered, and evaporated. The crude product is fairly clean and purification is not
25

necessary, but may be purified as in Method A. Yield: 1.80 g (70%). mp 135.0-135.5. 1H
NMR ((CD3)2CO, 500 MHz) δ (ppm): 8.69 (ddd, 2H, 6py-a), 8.28 (t, 1H, 4py-b), 8.24 (d,
2H, 3/5py-b), 8.07 (dt, 2H, 3py-a), 7.95 (td, 2H, 4py-a), 7.57 (dtd, 2H, 5py-a).
[Fe(dcpp)2](PF6)2.

To an air free flask dcpp (0.585 g, 2.02 mmol) and

deoxygenated 4:3 MeOH/H2O (50 mL) via cannula were added. The solution was gently
warmed (40°C) under nitrogen to promote dissolution. A separate air free flask was
charged with 1 equivalent of FeCl2·2H2O (0.146 g, 0.897 mmol) and 4:3 MeOH/H2O (20
mL).

The Fe(II) solution was transferred via cannula to the ligand solution which

immediately turned pale green then gradually darkened to deep blue. The reaction was
allowed to stir for 18 hours at room temperature. A separate air free flask was charged
with 4 equivalents of NH4PF6 (0.620 g, 3.80 mmol) and 4:3 MeOH/H2O (15 mL) which
was then transferred via cannula to the reaction mixture. The solution was concentrated
by evaporation under a stream of N2 yielding dark blue precipitate which was filtered and
rinsed with H2O. The crude product was dissolved in acetonitrile and the solution was
washed with hexanes to remove grease. The acetonitrile solution was concentrated then
purified by passage through a basic alumina column. The product was recrystallized by
diethyl ether vapor diffusion into a 1:1 acetone/acetonitrile solution of the complex.
Yield: 0.451 g (54%).

1H

NMR ((CD3)2CO, 500 MHz): δ (ppm) 8.75 (t, 2H, J = 7.6

Hz, 4py-b), 8.49 (d, 4H, J = 7.6 Hz, 3/5py-b), 8.34 (td, 4H, J = 7.8, 1.2 Hz, 4py-a), 8.19
(dd, 4H, J = 7.8, 1.7 Hz, 3py-a), 8.05 (dd, 4H, J = 5.9, 1.2 Hz, 6py-a), 7.49 (td, 4H, J =
5.9, 1.7 Hz, 5py-a). 13C NMR ((CD3)2CO, 500 MHz): δ (ppm) 180.69, 160.92, 159.21,
159.08, 144.60, 140.93, 131.20, 128.96, 128.05. TOF-MS [ESI, m/z (rel int)]: 317.0

26

(100) [C34H22N6O4Fe]2+, 779.1 (20) {[C34H22N6O4Fe](PF6)}+. IR (KBr, cm-1): 1699.2
s, 1683.4 s, 1596.1 m, 1434.7 w, 1328.4 m, 1242.6 w, 986.2 w, 915.9 w, 836.5 br, 760.6
m, 667.9 w.

Elemental Analysis for C34H22N6O4FeP2F12·CH3OH, Calculated:

C,

43.96; H, 2.74; N, 8.78; Fe, 5.84. Found: C, 43.83; H, 2.73; N, 8.79; Fe, 5.80. UV-Vis
(CH3CN) λ(ε(M-1cm-1)): 354 nm (11000), 445 nm (2500), 513 nm (4600), 606 nm
(6200).
[Cd(dmpp)2](PF6)2. To a 50 mL Erlenmeyer flask were added dmpp (0.761 g,
2.91 mmol), Cd(CH3CO2)2 (0.217 g, 0.941 mmol), and 1:1 EtOH/H2O (20 mL) and the
solution was allowed to stir overnight in the presence of air. 10 equivalents of NH4PF6 in
H2O was added and a beige precipitate formed.

The product was filtered then

recrystallized by ether diffusion into an acetone solution yielding white crystalline
needles. Yield: 0.666 g (72%). 1H NMR ((CD3)2CO, 500 MHz): δ (ppm) 8.23 (t, 2H, J =
7.8 Hz, 4py-b), 8.12 (td, 4H, J = 7.8, 1.7 Hz, 4py-a), 7.93 (dd, 8H, J = 7.8, 1.7 Hz, 3py-a,
3/5py-b), 7.70 (dq, 4H, J = 5.3, 1.0 Hz, 6py-a), 7.35 (td, 4H, J = 5.3, 1.0 Hz, 5py-a), 4.61
(d, 4H, J = 14.5 Hz), 4.43 (d, 4H, J = 14.5 Hz). 13C NMR ((CD3)2CO, 500 MHz): δ
157.84, 157.66, 148.86, 142.36, 141.37, 126.39, 125.15, 124.60, 43.37. TOF-MS [ESI,
m/z (rel int)]: 318.0 (25) [C34H22N6Cd]2+. IR (KBr, cm-1): 1604.1 m, 1575.1 m, 1482.8
m, 1440.3 m, 1111.7 w, 1016.4 m, 842.2 br, 779.4 w.

Elemental Analysis for

C34H22N6CdP2F12, Calculated: C, 44.54; H, 2.42; N, 9.17. Found: C, 44.01; H, 3.07; N,
9.06.
[Cd(dcpp)2]2+. Many routes have been employed to prepare this compound
successfully; however, it could not be isolated in pure form.
27

Method A. [Cd(dcpp)2](PF6)2 was prepared by modification of a literature procedure.16
To a round bottom flask dcpp (0.101 g, 0.349 mmol), Cd(NO3)2·4H2O (0.050 g, 0.162
mmol), and dry acetonitrile (30 mL) were added and the solution stirred for 6 hours. The
solution was evaporated and the remaining solid dissolved in H2O/MeOH and 4
equivalents of NH4PF6 were added yielding a sticky white solid. The solid was dissolved
in acetonitrile and precipitated with ether, but instead of yielding a solid, a coagulated
gel-like suspension forms. The gel is filtered yielding a flaky white solid that contains
the desired product. TOF-MS [ESI, m/z (rel int)]: 346.0 (5) [C34H22N6O4Cd]2+, 837.0
(5) {[C34H22N6O4Cd](PF6)}+.
Method B. [Cd(dcpp)2](PF6)2. To a round bottom flask dcpp (0.120 g, 0.415 mmol),
anhydrous Cd(CH3CO2)2 (0.033 g, 0.143 mmol), and dry ethanol (20 mL) were added
and the solution was warmed at 50oC overnight. 20 equivalents of NH4PF6 in 4 mL of
ethanol were added and the solution concentrated yielding a thick gel-like solid as in
Method A. The gel is filtered yielding a flaky white solid that contains the desired
product. TOF-MS [ESI, m/z (rel int)]:

346.1 (10) [C34H22N6O4Cd]2+, 837.0 (5)

{[C34H22N6O4Cd](PF6)}+.
Method C. [Cd(dcpp)2](ClO4)2. A 25 mL Erlenmeyer flask was charged with dcpp
(0.0987 g, 0.341 mmol) and ethanol (3 mL). Cd(ClO4)2·xH2O (0.0369 g, 0.119 mmol) in
1 mL of ethanol was added dropwise with stirring. During addition solution becomes
cloudy then coagulates forming a gel as in Method A. A flaky white solid is obtained
that contains the desired product. TOF-MS [ESI, m/z (rel int)]:
[C34H22N6O4Cd]2+, 791.0 (20) {[C34H22N6O4Cd](ClO4)}+.
28

346.0 (15)

Method D.

[Cd(dcpp)2](PF6)2.

A 125 mL Erlenmeyer flask was charged with

[Cd(dmpp)2](PF6)2, DMF (50 mL), and activated carbon. The reaction mixture was
heated to 100oC and bubbled with air for 5 hours. The carbon was filtered and water was
added to the filtrate to precipitate the product which was found to be a mixture of
[Cd(dcpp)2](PF6)2 and dcpp.

TOF-MS [ESI, m/z (rel int)]:

346.1 (30)

[C34H22N6O4Cd]2+.
To date, all attempts to purify: column chromatography on silica, reverse phase silica,
neutral alumina, basic alumina, and size-exclusion chromatography have been
unsuccessful, therefore the pure product is yet to be isolated.
2,6-Di(2-ketalpyridyl)pyridine (dkpp). The protection of the carbonyls of dcpp
was carried out by modification of a published procedure.17 To a 100 mL round bottom
flask equipped with a Dean-Stark trap was added dcpp (0.774 g, 2.68 mmol), freshly
distilled ethylene glycol (2 mL), dry toluene (50 mL), and 3 drops of H2SO4. The
reaction mixture was refluxed utilizing an oil bath for 6 days. Upon heating black residue
forms in the flask, sticking to the sides. Heating to 120oC for 5 days resulted in no water
collection in trap so warmed to 125oC on the last day, resulting in the removal of water.
The toluene layer is poured into an Erlenmeyer flask and placed in the freezer to
precipitate the product. Yield: 0.423 g, (42%). The black residue from reaction flask was
dissolved in CHCl3 and washed with 5% Na2CO3, and found to contain a mixture of
unreacted dcpp, partially reacted monoketal, and the diketal product. TLC on silica (in
CH3CN or CH2Cl2) shows clear separation of the three compounds, but with the desired
product remaining at the baseline no attempts were made to isolate it by column

29

chromatography. mp 177-178oC. 1H NMR ((CD3)2CO, 500 MHz): δ 8.37 (d, 2H, J =
4.8 Hz, 6py-a), 7.79 (t, 1H, J = 7.7 Hz, 4py-b), 7.68 (td, 2H, J = 7.7 Hz, 4py-a), 7.60 (d,
2H, J = 7.9 Hz, 3/5py-b), 7.58 (d, 2H, J = 7.9 Hz, 3py-a), 7.20 (dtd, 2H, J = 4.8 Hz, 5pya), 4.03 (dm, 8H, J = 4.4 Hz). 13C NMR ((CD3)2CO, 500 MHz): δ 148.32, 136.12,
135.80, 135.57, 122.53, 121.15, 119.98, 109.99, 65.26. TOF-MS [ESI, m/z (rel int)]:
378.1 (100) [C21H19N3O4]H+. IR (KBr, cm-1): 3046.6 w, 2950.2 w, 2886.5 m, 1588.8
m, 1573.4 m, 1466.5 m, 1442.4 m, 1279.2 w, 1215.4 m, 1133.1 s, 1031.5 s, 992.2 s,
949.6 m, 834.5 m, 797.1 m, 762.5 m, 669.2 m.
2.2.2 Physical Measurements
X-ray Structure Determination. Single-crystal X-ray diffraction data for
[Fe(dcpp)2](PF6)2 was acquired at the X-ray Facility of Michigan State University. Xray quality single crystals were obtained by slow diffusion of diethyl ether into an
acetonitrile solution of the compound. A black block crystal with dimensions 0.15 × 0.15
× 0.08 mm was mounted on a Nylon loop using very small amount of paratone oil. Data
were collected using a Bruker CCD (charge coupled device) based diffractometer
equipped with an Oxford Cryostream low-temperature apparatus operating at 173 K.
Data were measured using omega and phi scans of 0.5° per frame for 30 s. The total
number of images was based on results from the program COSMO18 where redundancy
was expected to be 4.0 and completeness of 100% out to 0.83 Å. Cell parameters were
retrieved using APEX II software19 and refined using SAINT on all observed reflections.
Data reduction was performed using the SAINT software20 which corrects for Lp.
Scaling and absorption corrections were applied using SADABS21 multi-scan technique,
30

supplied by George Sheldrick. The structures are solved by the direct method using the
SHELXS-97 program and refined by least squares method on F2, SHELXL-97, which are
incorporated in SHELXTL-PC V6.10.22

Anisotropic thermal parameters were refined

for all non-hydrogen atoms. Hydrogen atoms were localized in their calculation positions
and refined by using the riding model.
Cyclic Voltammetry. Electrochemical measurements were carried out using a
CH Instruments CH620D electrochemical analyzer to determine the E1/2 for ligand
reductions and metal oxidations of each complex. Solutions of the compounds were
prepared in distilled CH3CN containing NBu4PF6 (ca. 0.1 M) as the supporting
electrolyte.

A standard three-electrode setup was used with a platinum working

electrode, carbon rod counter electrode, and a Ag/AgCl or Ag/AgNO3 electrode as the
reference. All measurements were made inside an inert atmosphere glovebox. Data was
acquired at a scan rate of 100 mV s-1. After data collection ferrocene was added to
subsequent scans for reference.
2.2.3 Theoretical Calculations
Theoretical calculations were performed using Gaussian 03.23

The initia4

geometry of the [Fe(dcpp)2]2+ molecule was obtained from the crystal structure data and
subsequently optimized in two steps. The first optimization was performed using the
density functional B3LYP with the 3-21G** basis set, followed by a second optimization
using the B3LYP functional with the 6-31G* basis set. Optimizations were repeated
using the polarizable continuum model (PCM) to account for solvent effects. Frequency
calculations at the B3LYP/6-31G* level of theory were performed on the final optimized

31

structures to ensure that these geometries corresponded to global minima.

Time-

dependent DFT was used to calculate energies of excited states and obtain models of the
molecular orbitals.
2.3 Results and Discussion
2.3.1 Synthesis of 2,6-Di(2-carboxypyridyl)pyridine
2,6-Di(2-carboxypyridyl)pyridine, dcpp, may be prepared by a variety of
methods. It was first synthesized by Abarca et al. from a general strategy they developed
to yield a series of potential helicating ligands from triazolopyridines.8 Due to the wide
interest in the fascinating coordination possibilities of this ligand, alternate routes have
been developed.7,15,24–26
A clear starting point toward the development of the synthesis of [Fe(dcpp)2]2+
was to look to the route of the preparation of the ruthenium analog by Schramm et al. 7
Due to the nature of the ligand (vide infra), the complex could not be prepared by simply
coordinating the ligand to the metal.

Instead, the target molecule was obtained by

simultaneous coordination/ligand oxidation from 2,6-di(2-methylenepyridyl)pyridine,
dmpp. Therefore, it was an obvious choice to prepare the dmpp first, from which the
dcpp may be prepared by a simple oxidation procedure.
A one step procedure for the preparation of dmpp, or tripyridinedimethane, was
developed by Vedernikov et al.27 The route involves lithiating 2,6-lutidine, creating a
bis(nucleophile) to react with pyridine. The simple one step procedure yields impure
crude products requiring tedious fractional distillation and results in only moderate
yields. An improved method was developed by Dyker and Muth in 2004.12 This
32

procedure involves the lithiation of 2-picoline, to be used as the nucleophile to react with
2,6-difluoropyridine. It is crucial to use at least a four-fold excess of the nucleophile,
since the product has two acidic methylene bridges that each lead to consumption of one
equivalent of the lithiated picoline. This optimized reaction was chosen and provided a
95% yield of the dmpp, as shown in Scheme 2-1. Dmpp is subsequently oxidized with
selenium dioxide to dcpp.

Scheme 2-1. Stepwise synthesis of 2,6-di(2-methylenepyridyl)pyridine (dmpp) and 2,6di(2-carboxypyridyl)pyridine (dcpp).
The selenium dioxide oxidation, or Riley oxidation, is commonly employed in our
lab for the oxidation of the active methyl groups on substituted bipyridyls. The method
may be utilized to oxidize methyl groups to alcohols, and with excess oxidant will
convert them to ketones.

Though this route is convenient, selenium dioxide is not

pleasant to work with; it readily sublimes, has a harsh, unpleasant odor, can burn the nose
and throat upon exposure, and even result in garlic breath. Furthermore, much difficulty
arises in removing the malodorous selenium byproducts completely while maintaining a
decent yield. Overall, this method stinks, literally, and another route was pursued.

33

An alternate method for the preparation of dcpp was adapted from work by
Goldsmith and Stack.15

This convenient method involves the lithiation of 2-

bromopyridine followed by nucleophilic attack of a diethyl ester of 2,6pyridinedicarboxylate, as shown in Scheme 2-2.

The reaction is relatively clean,

requiring little effort in purification, and the product is obtained in 70% yield. This route
is also preferential in that, by slight modifications to the starting materials, it allows for
ease of functionalization to install anchoring groups and other structural modifications of
the molecule, as will be discussed in later chapters of this dissertation.

Scheme 2-2. Alternate synthesis of 2,6-di(2-carboxypyridyl)pyridine (dcpp).
2.3.2 Coordination Chemistry of 2,6-Di(2-carboxypyridyl)pyridine
The coordination chemistry of dcpp has been a hot topic in recent years. From the
time the ligand properties were first investigated in 2005 to 2009 there were already 14
crystallographically established coordination modes of dcpp.9,28 By introducing the
carbonyl bridges, not only are there two new possible coordinating sites, the oxygen
atoms, but the carbon of the carbonyl is highly susceptible to transformation resulting in
various forms of the ligand.
One of the first things we learn in introductory organic chemistry is about the
reactivity of carbonyl compounds, with organic texts referring to the addition of a
nucleophile to a carbonyl group as the simplest of all organic reactions.29 The concept is
34

introduced with the example of an aldehyde or ketone dissolved in water or alcohol, in
which water acts a nucleophile and attacks the electrophilic carbonyl forming the hydrate,
or gem-diol, and alcohol will produce a hemiacetal or hemiketal. As shown in Scheme 23, an equilibrium exists between a ketone and its hydrated form that usually lies far
toward the ketone, however, special factors can shift the equilibrium more toward the
hydrated form. In the case of dcpp, the electron withdrawing 2-pyridyl groups increase
the electrophilicity of the carbonyl carbon which may shift this equilibrium more toward
the hydrated form.

Scheme 2-3. Equilibrium of dcpp in water or methanol.
The hydrated or hemiketal forms of dcpp are not stable in the solid form and have
not been isolated. In solution, evidence for the formation of the hemiketal can be seen in
the 1H NMR spectrum (Figure 2-3). Dcpp remains intact in deuterated acetone (shown),
acetonitrile, or chloroform, but evidence of hemiketal formation appears in deuterated
methanol. Quantitative NMR was not performed, but from integration of a standard 1H
NMR spectrum it appears that the equilibrium lies toward the ketone with a 4:1 ratio of
ketone to hemiketal.

35

Figure 2-3. 1H NMR of dcpp in (CD3)2CO (top) and CD3OD (bottom).
Coordination of the pyridyl nitrogen atoms to a metal ion may shift the
equilibrium further toward the hydrated form. Donation of the nitrogen lone pair to the
metal will increase polarization of the carbon-carbon bond between the pyridyl carbon
and carbonyl carbon, making it more electrophilic.

Also, a mechanism for metal

promoted hydrolysis/solvolysis has been proposed based on a crystal structure study on
Cu(II) dcpp complexes.24 The planar dcpp wraps around the metal in a helical fashion
and the nucleophilic solvent (H2O or CH3OH) is captured by the metal through a weak
interaction forming a distorted quasi-octahedral structure.

This coordination

environment, together with the steric effect of the helical structure, dictates the

36

nucleophilic attack by the solvent. This mechanism has been termed metal-promoted,
helical chelation-induced hydrolysis and asymmetric solvolysis.
Due to the reactivity of dcpp, small variations in stoichiometry, reaction
conditions, and choice of solvent lead to different structures. It is unusual that the ligand
remains intact upon coordination, and even more so in the presence of water or
methanol.9 This seems to be the case with the coordination chemistry of iron(III) and
dcpp, as well as with the chemistry of iron(II) and di(2-pyridyl)ketone (dpk).30–33 Since
the motivation behind much of the research is the preparation of polynuclear clusters,
single molecule magnets, and coordination polymers, in which hydrolysis of the ligand is
preferred, it appears that not many attempts have been made to coordinate the ligand
intact.
2.3.3 Synthesis of Bis(2,6-di(2-carboxypyridyl)pyridine)iron(II)
According to the work by Schramm et al. bis(2,6-di(2-carboxypyridyl)pyridine)ruthenium(II) could not be prepared via coordination of the ligand to the metal.7
The authors claim that neither working in the exclusion of air, nor higher temperatures,
nor microwave heating could produce the target molecule. Though not explicitly stated,
it is likely that the reactive carbonyl is problematic. The target molecule was obtained by
simultaneous coordination/ligand oxidation from 2,6-di(2-methylenepyridyl)pyridine,
dmpp, shown in Scheme 2-4. This simultaneous coordination/ligand oxidation has been
found to occur in other Ru(II) polypyridyls with activated positions on the ligand back
bone.34

37

Scheme 2-4. Synthesis of Ru(dcpp)2](NO3)2 (adapted from reference 3).
The iron(II) chemistry likely could not proceed by the same method, since Fe(II)
in solution is highly susceptible to oxidation to Fe(III). Nevertheless, the method was
attempted, to no avail. Identical conditions were utilized, as well as a few variations in
solvent choice, temperatures, order of addition, and attempts to initially exclude oxygen.
In all cases, there was no evidence of the desired product by ESI-MS. Also, the samples
were paramagnetic by NMR, likely from the formation of Fe(III) compounds.
The standard route employed in our lab for the preparation of Fe(II) polypyridyl
complexes is the addition of one equivalent of FeCl2·2H2O to two equivalents of ligand
in methanol/water.

The reasoning for the choice of solvent mixture is as follows:

methanol for solubilizing the ligand, water for solubilizing the iron(II) reagent, and ease
of isolation since the PF6- salt of the complex may be easily metathesized from this
solvent mixture, which almost always is obtained in pure form upon filtration of the
reaction mixture. Ignoring all precedence in the literature for this particular ligand
system (vide supra), the standard route was attempted.

38

[Fe(dcpp)2](PF6)2 was

successfully prepared and the optimized reaction conditions produced a 54% yield of
pure product (Scheme 2-5).

Scheme 2-5. Synthesis of [Fe(dcpp)2](PF6)2.
Unlike with the preparation of most iron(II) polypyridyl complexes, which
produce high yields of pure products right out of the gate, the formation of [Fe(dcpp)2]2+
is accompanied by formation of a byproduct. The reaction mixture initially turns green
then darkens as the desired complex forms, and longer reaction times increase the yield
of the desired product. Varying the ligand to metal ratio (i.e., 4:1, 3:1, 2:1) does not have
a significant effect on product/byproduct formation, and it was determined that optimal
yield is obtained from a slight excess of ligand (2.25:1 ratio). Presumably, the byproduct
arises from hydrate or hemiketal formation, therefore an alternate solvent system may
prevent the formation of the byproduct. However, since the byproduct and excess ligand
may be easily removed by column chromatography, an alternate solvent system was not
employed. The blue colored [Fe(dcpp)2]2+ may be purified by the same method as the
Ru(II) analog, by silica gel chromatography with 20:1 acetonitrile/KNO3 (aq). The
excess ligand is eluted first followed by [Fe(dcpp)2]2+, while the byproduct remains on
39

the column. A better method for purification was found to be chromatography on basic
alumina with acetonitrile. The alumina column runs much more quickly and retains the
ligand and byproduct, with only the pure product being eluted. The product may then be
recrystallized by diethyl ether vapor diffusion into a 1:1 acetone/acetonitrile.
The formation of this complex in water/methanol is a curiosity and requires
further investigation. Since the focus of the current research is on the properties of the
resulting complex, the reaction was not further analyzed, but an observation about the
synthesis should be noted. Attempts to isolate [Fe(dcpp)2]Cl2 from the reaction were
unsuccessful, yielding only the byproduct and unreacted ligand, metathesis to the PF6salt was a must.

I am uncertain of the reason behind this, is the counter ion that

important? Is the formation affected by the change in pH upon adding an excess of the
acidic NH4PF6 reagent? Alternate solvent systems, pH adjustments, and alternative
counterions should be investigated.
2.3.4 Properties of Bis(2,6-di(2-carboxypyridyl)pyridine)iron(II)
Single crystals suitable for X-ray crystallography were obtained and the crystal
structure solved by Dr. Dong Guo. As was the intended goal, an octahedral coordination
environment is achieved with [Fe(dcpp)2]2+ and the crystal structure as shown in Figure
2-4. The crystallographic data is listed in Table 2-1 and selected geometric parameters
are listed in Tables 2-2 and 2-3.

40

Figure 2-4. ORTEP Drawing of the cation of [Fe(dcpp)2]2+ obtained from single-crystal
X-ray structure determination. Atoms are represented as 50% probability thermal
ellipsoids. Hydrogen atoms and anions are omitted for clarity.
Table 2-1. Crystallographic data for [Fe(dcpp)2](PF6)2.
[Fe(dcpp)2](PF6)2 a
Empirical formula
Formula weight
Temperature/K
Crystal system
Space group
a/Å
b/Å
c/Å
Volume/Å3
Z

C34H22F12FeN6O4P2 ρcalcmg/mm3
924.37
173(2)
Orthorhombic
Pbcn
14.5603 (3)
15.4333 (3)
15.0431 (3)
3380.39 (12)
4

/mm-1
F(000)
2θmax
Reflections collected
Independent reflections
Reflections with I > 2(I)
R(int)
R1 b
wR2 c
Goodness-of-fit on F2

a Obtained with graphite monochromated Mo Kα radiation (λ = 0.71073
 Fc  Fo . c wR2 = {  [w(Fo2  Fc2)2/  [w(Fo2)2]}1/2.

41

1.816
0.66
1856
56.6°
21855
4199
3524
0.072
0.047
0.133
1.05

Å) b R1 =  Fo 

Table 2-2. Selected bond lengths from the X-ray crystal structure of [Fe(dcpp)2](PF6)2.
Selected Bond Lengths (Å)
Fe1—N2i
1.974 (2)
Fe1—N2
1.974 (2)
Fe1—N1
1.985 (2)
Fe1—N1i
1.985 (2)
Fe1—N3i
1.989 (2)
Fe1—N3
1.989 (2)
Symmetry code (i) −x, y, −z+1/2
Table 2-3. Selected bond angles from the X-ray crystal structure of [Fe(dcpp)2](PF6)2.
N2i—Fe1—N2
N2i—Fe1—N1
N2—Fe1—N1
N2i—Fe1—N1i
N2—Fe1—N1i
N1—Fe1—N1i
N2i—Fe1—N3i
N2—Fe1—N3i

Selected Bond Angles (°)
178.96 (8)
N1—Fe1—N3i
89.07 (7)
90.48 (7)
N1i—Fe1—N3i
177.63 (6)
88.79 (6)
N2i—Fe1—N3
91.86 (6)
88.79 (6)
N2—Fe1—N3
88.87 (6)
90.48 (7)
N1—Fe1—N3
177.63 (6)
90.61 (9)
N1i—Fe1—N3
89.07 (7)
88.92 (9)
N3i—Fe1—N3
91.35 (9)
91.86 (6)
Symmetry code (i) −x, y, −z+1/2

The flexibility of the dcpp ligand has allowed for improved bite angles around the
metal center, coordinating in a helical fashion. A much more symmetrical environment is
achieved as compared to the coordination environment of the rigid bpy and terpy ligands
as shown in Figure 2-5.

42

Figure 2-5. Structures of [Fe(bpy)3]2+, [Fe(terpy)2]2+, and [Fe(dcpp)2]2+ showing the
coordination environment with approximate bond angles.
An important characteristic that a DSSC chromophore must possess is
absorptivity in the visible region. Chromophores containing conjugated ligand systems,
like bpy and terpy fit this criteria. The dcpp ligand is not conjugated in the normal sense
of the word, but is cross-conjugated. A cross-conjugated compound is one in which there
are three unsaturated groups, where two of them are conjugated to a third but are not
conjugated to each other.35 In the case of dcpp, the pyridyl rings are not conjugated to
each other, but are each conjugated to the carbonyl groups. We were uncertain of the
implications of this since cross-conjugated systems are considered to be much less able to
promote electron delocalization than true conjugation, until recently that is. Recent
theoretical investigations have found that the connectivity pattern in cross-conjugation
will lead to enhanced electron delocalization effects for electronically excited states.36
From the optical absorption spectrum in Figure 2-6, we can see that the lack of a
truly conjugated system is not detrimental to the molar absorptivity, and the oscillator
strength was determined to be comparable for all three systems.2 The sharp feature in the
[Fe(terpy)2]2+ spectrum, that is not seen in the other two, arises from the delocalization
43

over the three rings, signifying that delocalization is in fact not extended across the entire
dcpp ligand in the new complex.

The red-shifted absorption maximum of the

[Fe(dcpp)2]2+ is ideal for a sensitizer, as it covers more of the solar spectrum.

[Fe(bpy)3]

2+
2+

[Fe(terpy)2]

2+

[Fe(dcpp)2]

Figure 2-6. Ground state electronic absorption spectrum of [Fe(bpy)3]2+ (black),
[Fe(terpy)2]2+ (red), and [Fe(dcpp)2]2+ (blue) in CH3CN (collected by Allison Brown,
adapted from reference 2).
When the dcpp ligand twists from its planar arrangement upon coordination the
orbital overlap across the conjugated branches within one ligand will be disturbed.
However, medium - stacking intramolecular interactions are observed between each
ketone plane of one dcpp ligand and a peripheral pyridine ring of another dcpp ligand.
The shortest inter-planar atom…atom separations and dihedral angles are 2.87 Å (19.3°)
and 2.84 Å (18.8°) for the stacked pairs, respectively, with the center-to-center
separations of these stacked pairs of 3.26 Å and 3.40 Å. The cooperative effects of inter44

ligand interactions might be a reason for the formation of the tight coordination of Fe(II)
core.
Carbonyl lone pair (l.p.)—π interactions have recently been recognized as
significant supramolecular bonding contacts by the scientific community.37–41

The

importance of these l.p.—π interactions were first recognized in biological
macromolecules nearly twenty years ago, but this interaction was seemingly overlooked
in small synthetic molecules as they have been rarely reported.42 Mooibroek et al.
performed a detailed analysis of the Cambridge Structure Database and found that these
carbonyl l.p.— interactions are not uncommon with nearly 6000 occurances.39 The
authors note that the majority of these contacts with benzene rings are weak and the
incorporation of nitrogen into the ring strengthens this interaction. Electron deficiency of
the pyridine ring clearly induces the formation of stronger l.p.—π interaction.
In the case of the [Fe(dcpp)2]2+, the π-acidic nature of the pyridine rings and
metal coordination induced polarization both favor the carbonyl l.p.—π interaction. This
is quite noteworthy since the metal coordination induced polarization is also known to
facilitate nucleophilic addition at the carbonyl carbon in other complexes with the dcpp
ligand.

The formation of this particular molecule is likely aided by this l.p.—π

interaction, stabilizing the carbonyl in addition to stabilizing the entire molecule.
From the crystal structure data of [Fe(dcpp)2]2+ we determined the distance (d)
between the centroid of the terminal pyridine rings to the oxygen of the carbonyl bridges
on the opposing ligand to be 3.351-3.365 Å. The carbonyl stacks with the pyridyl ring in
a nearly planar arrangement with 18.8-19.3o angle between planes. Preliminary DFT

45

calculations suggest the significant interligand π interactions, may facilitate
delocalization over the entire molecule as shown in Figure 2-7.

d (centroid to O) = 3.351 – 3.365 Å

o

ω (angle between planes) = 18.8-19.3

Figure 2-7. Results from a TD-DFT calculation on [Fe(dcpp)2]2+ showing the
interligand π stacking interaction between the carbonyl of one ligand with the pyridyl
ring of the other. Also shown is the schematic of the structural parameters determined
from the crystal structure used to characterize the l.p.—π contact.
The electrochemistry, photoluminescence investigation, and proposed effects of
symmetry on ligand field strength are described in detail elsewhere, but in the interest of
being thorough will also be mentioned here.2 The electrochemistry of [Fe(dcpp)2]2+ was
found to differ greatly from [Fe(bpy)2]2+ and [Fe(terpy)2]2+, with the oxidation and first
reduction potentials being much more positive (Table 2-4). The electrochemical behavior
is attributed to the electron withdrawing carbonyls along with better metal ligand overlap.

46

Table 2-4. Electrochemical potentials for the oxidation and first reduction of
[Fe(bpy)3](PF6)2, [Fe(terpy)2](PF6)2, and [Fe(dcpp)2](PF6)2, measured with Ag/AgNO3
reference with 0.1 M TBAPF6 electrolyte in acetonitrile, externally referenced to
ferrocene (adapted from reference 2).
Compound
[Fe(bpy)3]2+
[Fe(terpy)2]2+
[Fe(dcpp)2]2+

E 1 / 2 [ox]
0.665 V
0.715 V
1.295 V

E 1 / 2 [red]
-1.775 V
-1.675 V
-0.965 V

There are four distinct reductions which are facilitated by the electron
withdrawing carbonyls. The significant shift of the oxidation potential is consistent with
the inductive influence of ligand substituents on the Fe(II)/Fe(III) redox couple, with
more electron withdrawing groups leading to a more positive redox couple.43 This
inductive effect appears to be more significant as compared to substituted terpy ligands,
likely due to enhanced metal ligand overlap and stronger π accepting capacity of the
ligand, leading to larger ligand field splitting and stabilization of the t2g set.
A stronger ligand field may be achieved by improving metal ligand orbital
overlap by expanding the coordination environment. The lower energy of the ligandbased π*-orbitals leads to a better energetic match with the π-symmetry d-orbitals of the
metal center, stabilizing the t2g set. This allows us to propose the molecular orbital
diagrams in Figure 2-8, depicting the effect that the structural differences have on the
ligand field splitting.

47

Fe2+

Fe2+

terpy

dcpp




d

3d





Figure 2-8.
Proposed molecular orbital diagrams of Fe(terpy)2]2+ (left) and
Fe(dcpp)2]2+ (right) showing effects of ligand orbital movement on ligand field splitting.
We are currently investigating whether we have increased the ligand field strength
sufficiently to result in an inversion of the lowest energy excited states from the 5T2 to
3T .
1

Preliminary photoluminescence experiments have provided evidence for weak

emission, suggesting that we may be accessing a triplet ligand field state.

Weak

emission was detected in a 9:2 butyronitrile/propionitrile glass at 80 K at 800 nm, which
is at the limit of our detector so further evidence was required.2 A solid-state emission
experiment was performed by Christian Reber at the Université de Montréal who also
found evidence of photoluminescence, with the presence of weak emission at low
temperature centered around 700 nm.

As for other possible sources of the

photoluminescence, we cannot rule out the 3MLCT state, an emissive impurity, or a
ligand based effect. In order to narrow down the source of the emission, it is crucial to
develop a control molecule.
48

2.3.5 Quest for a Control Molecule
In the review of the coordination chemistry of di(2-pyridyl)ketone (dkp) and 2,6di(2-carboxypyridyl)pyridine (dcpp) the authors state “at the outset of our efforts we were
aware that such ligands could not be incorporated in metal complexes in, e.g. aqueous or
alcoholic reaction media.”9 I seem to have stumbled upon the exception to the rule with
the coordination of iron(II) and was not optimistic to find a second when setting out to
prepare a control analog.
The first choice for a control molecule is the zinc(II) analog. Being a first row
transition metal as well as having a divalent charge would make it the most isostructural
with the Fe(II) analog, while at the same time provide insight from the absence of d-d
transitions. Coordination chemistry of Zn(II) with dcpp is unknown in the literature, but
Zn(II) with dpk has been widely investigated in the context of polynuclear cluster
chemistry.9,44–46 Dpk is highly sensitive to the reaction media in reactions with zinc,
and although rare, can remain intact in a reaction.44 In most cases the carbonyl is
hydrated, even by trace amounts of water in the solvent or in the starting materials.45
Many routes were attempted employing various Zn(II) reagents, solvents, and
reaction conditions. Most of the zinc(II) reagents are hydrates, so even in the situations
where dry solvent is used, the reaction fails to yield any of the desired product (by ESIMS). The instance in the literature where the ligand remained intact was accomplished by
adding ZnCl2 to dkp in acetonitrile and layering in hexane/ether and allowing to sit. This
method was attempted with ZnCl2 and dcpp in the inert atmosphere dry box, and even
under these dry, mild conditions there was no evidence of the desired product.

49

Coordination of 2,6-di(2-methylenepyridyl)pyridine, dmpp, to Zn(II) is possible but
attempts to oxidize the ligand after coordination are too harsh and strip the ligand from
the metal. After many fruitless attempts I changed direction toward the preparation of a
gallium(III) analog.
Gallium is not technically a transition metal, but a post-transition metal, or poor
metal, in the p-block of the periodic table with a common oxidation state of +3 which is
d10. A difference in oxidation state is not ideal, but due to the difficulty encountered with
the Zn(II) chemistry, it was worth pursuing. The coordination chemistry of Ga(III) is
rather unexplored, but medicinal applications have sparked more interest in the field.
The survey of the literature by Bandoli et al. found only 17 homoleptic bis-tridentate
Ga(III) compounds, with only 6 of these being meridionally bound.47 Ga(III) not only
binds to nitrogen, but readily coordinates to oxygen, also the tendency of Ga to form
hydroxo-bridged dimers raised many concerns.

Despite the apprehension, the wide

availability and ease of preparation of anhydrous reagents made it worth a shot.
To avoid any of the hydration problems with the reactive carbonyl of dcpp and the
propensity of the formation of hydroxo-bridges dimers of gallium, every effort was made
to exclude water. Reactions utilizing the commercially available anhydrous gallium(III)
isopropoxide, gallium(III)chloride, and gallium(III) iodide were attempted with no
success. In exploring the literature, I stumbled upon an interesting gallium reagent that
was used to prepare [Ga(bpy)3]3+.48 The researchers found that the reaction of “GaI”
with one equivalent of bpy in toluene unexpectedly led to the high yield formation of
[Ga(bpy)3]I3. They found that presence of water led to the formation of gallium dimers,

50

but since water may easily be excluded from the reaction in toluene, this route was
pursued.
“GaI” is a very interesting, easily prepared reagent, that was recently reviewed
highlighting its versatility in organic synthesis and gallium cluster formation.49 GaI
appeared in the literature in 1955 and identity of its formulation has been redetermined
numerous times.50–53 The currently accepted formulation is a mixture of sub-iodides and
predominately exists as the mixed valence salt [(Ga+)2][Ga2I62-]. The reagent is simply
prepared by sonication of elemental gallium and iodine in toluene at 30oC. “GaI” is used
in disproportionation reactions in which gallium(III) complexes form, accompanied by
the deposition of elemental gallium which may be easily removed from the reaction. 48,54
Unfortunately, attempts to prepare [Ga(dcpp)2]3+ employing “GaI” have failed.
One last ditch effort with the gallium chemistry was based off of the work by
Manessi et al.55 They prepared the only known homoleptic, bis-tridentate gallium(III)
complex in which the ligand coordinates strictly through nitrogens.

They utilized

(Et4N)[GaCl4] which is simply prepared from Et4NCl and GaCl3 in ethanol to synthesize
the target molecule.

An analogous method was attempted for the preparation

[Ga(dcpp)2]3+, but not surprisingly, has also failed to yield any of the desired product.
Looking back at the hydration behavior of dkp it is important to note that metals
in high oxidation states facilitate the hydration while low oxidation states hinder it.56
This may be why Ga(III) has been difficult. A second point to note is that the hydration
behavior is attributed to steric strain imposed on the carbonyl carbon upon complexation.
The small size of Zn(II) and Ga(III) may be contributing to the transformation of the
ketone in dcpp. Investigating other options for a control molecule, we thought about
51

cadmium(II). We were hesitant to move to second row transition metals, but since the
larger size may alleviate some of the strain it was a viable option.
Finally, the quest for a control molecule began to yield some positive results,
[Cd(dcpp)2]2+ has been successfully prepared under a variety of conditions (Scheme 26).

Reactions

employing

Cd(NO3)2·4H2O,

anhydrous

Cd(CH3CO2)2

and

Cd(ClO4)2·xH2O in either acetonitrile or ethanol have produced the target molecule. As
was the case with the iron(II) analog, the reaction is plagued by byproduct formation.
Cd(II) ions are used to prepare metal organic coordination polymers and from the ESIMS it appears that polymerization may be occurring.57,58 To date, all attempts to purify
the complex have been unsuccessful. Avoiding hydrated Cd(II) reagents and alcohol or
aqueous solvents would be preferable, but we encounter solubility issues with anhydrous
reagents in anhydrous solvents, and attempts to prepare the complex under these
conditions have not produced results.

Scheme 2-6. Synthesis of [Cd(dcpp)2]2+.

52

An alternate course of action toward the preparation of [Cd(dcpp)2]2+ was to
synthesize it by the route of Schramm et al. by simultaneous coordination/ligand
oxidation (vide supra).

Air must not be an efficient oxidant in this case since

[Cd(dmpp)2]2+ forms with no evidence of [Cd(dcpp)2]2+.

The synthesis of

[Cd(dmpp)2](PF6)2 was adjusted to allow for cleaner isolation and was prepared in 72%
yield according to Scheme 2-7. Our current efforts are now directed at oxidizing the
ligand after complexation.

Scheme 2-7. Synthesis of [Cd(dmpp)2](PF6)2.
A variety of oxidation procedures have been attempted with little success. The
selenium dioxide catalyzed procedure that was used to prepare the dcpp ligand was found
to be too harsh as it strips the ligand from the metal center. Mild sodium periodate
mediated procedures have also been attempted, but led to the same result.59,60 Air would
be the ideal oxidant, but prolonged heating in DMF while being bubbled with air resulted
in no change, signifying the need for a catalyst. Auto-oxidation of carbon acids can be
achieved utilizing potassium t-butoxide as a catalyst, but again this would likely wreak
havoc on the metal complex.61 Looking into other options for a catalyst, we thought
53

about activated carbon. Activated carbon is commonly employed as a catalyst support,
but it actually possesses catalytic activity on its own due to the surface oxides and
structure. I attempted the same procedure of heating in DMF while bubbling with air, but
this time in the presence of activated carbon, and achieved the best result to date (Scheme
2-8). This seems to be the most promising route as the major products are the desired
complex and dcpp, without evidence of the high mass byproducts that were formed in the
previous methods. Efforts are ongoing with this reaction to find the right temperature
conditions to limit the degradation of the metal complex.

Scheme 2-8. Synthesis of [Cd(dcpp)2](PF6)2 via oxidation of [Cd(dmpp)2](PF6)2.
A new approach with the cadmium(II) chemistry that we are just beginning to
look at involves protection of the carbonyl. Protection of the carbonyl groups of dcpp is
readily achieved by the traditional route of acid catalyzed ketalization as shown in
Scheme 2-8.17 Much higher yields are common with this chemistry and will likely be
achieved in due course, but to date this reaction was performed only once, suffered from
incomplete conversion, and yet to be optimized.

54

Scheme 2-9. Synthesis of 2,6-di(2-ketalpyridyl)pyridine (dkpp).
Initially there was much reluctance toward pursuing protection chemistry since
deprotection commonly employs harsh hydrolytic conditions.17,62 Also, sterics of a
protected ligand may present an issue to complex formation, but researchers have found
that the ketal protected dkp can coordinate to Cu(II), Co(II), Ni(II), and Pd(II), so
hopefully this will not be an issue.63,64 If the protected dkpp ligand can even bind, it is
crucial that the complex remain intact upon deprotection so mild conditions are
necessary. Mild deprotection methods have recently been developed that are much more
promising for this chemistry.65–68 To date, the coordination of dkpp to Cd(II) is yet to be
attempted, but if successful, a promising deprotection method developed by Sun et al.
will be pursued.65 Deprotection is achieved under extremely mild conditions by catalytic
iodine in acetone that can provide high yields in just minutes under neutral conditions.65

2.4 Concluding comments
A highly symmetric iron(II) polypyridyl complex with some very exciting
properties has been successfully prepared. Unexpectedly, the coordination of the highly
reactive of 2,6-di(2-carboxypyridyl)pyridine ligand to iron(II) proceeds smoothly in
aqueous solution, against all precedence in the literature. The resulting complex provides
55

a good starting point toward the investigation of the effect of geometry on the ultrafast
dynamics of iron(II) polypyridyls. Though we have not achieved the overall goal of
slowing down the charge transfer deactivation to the ligand field state,2 we may have
developed an iron(II) complex with a strong enough ligand field to achieve inversion of
the lowest energy excited state. The pursuit for a control molecule is ongoing and
necessary to gain more information on this system.

56

APPENDIX

57

APPENDIX

Figure 2-9. 1H NMR spectrum of [Fe(dcpp)2](PF6)2 in (CD3)2CO.

58

Figure 2-10. 13C NMR spectrum of [Fe(dcpp)2](PF6)2 in (CD3)2CO.

59

TOF MS ES+

100
317.0

%
84.1

102.1
143.1
164.1

634.1 653.1

338.4

779.1

0
50

250

450
m/z

Figure 2-11. ESI-MS [Fe(dcpp)2](PF6)2.

60

650

850

%
75
T
r 70
a 65
n
s 60
m
i 55
t
t 50
a 45
n
c 40
e

3439.3

1596.3 1328.4

986.1
761.3

1699.5
1683.4

846.6

4000

3500

3000

2500

2000
Wavenumbers

Figure 2-12. IR spectrum of [Fe(dcpp)2](PF6)2

61

1500

1000

Table 2-5. Geometric parameters for the X-ray crystal structure of [Fe(dcpp)2](PF6)2.

Fe1—N2
Fe1—N2i
Fe1—N1i
Fe1—N1
Fe1—N3
Fe1—N3i
O1—C6
O2—C12
N1—C1
N1—C5
C1—C2
C1—H1A
C2—C3
C2—H2A
C3—C4
C3—H3A
C4—C5
C4—H4A
C5—C6
N2—C11
N2—C7
C6—C7
C7—C8
C8—C9
N2—Fe1—N2i
N2—Fe1—N1i
N2i—Fe1—N1i
N2—Fe1—N1
N2i—Fe1—N1
N1i—Fe1—N1
N2—Fe1—N3
N2i—Fe1—N3
N1i—Fe1—N3
N1—Fe1—N3
N2—Fe1—N3i
N2i—Fe1—N3i
N1i—Fe1—N3i
N1—Fe1—N3i
N3—Fe1—N3i
C1—N1—C5
C1—N1—Fe1
C5—N1—Fe1

Bond Lengths (Å)
1.9737 (17)
C8—H8A
1.9738 (17)
C9—C10
1.9844 (15)
C9—H9A
1.9845 (15)
C10—C11
1.9893 (15)
C10—H10A
1.9893 (15)
C11—C12
1.213 (2)
C12—C13
1.211 (2)
N3—C17
1.346 (3)
N3—C13
1.358 (2)
C13—C14
1.385 (3)
C14—C15
0.9300
C14—H14A
1.375 (3)
C15—C16
0.9300
C15—H15A
1.386 (3)
C16—C17
0.9300
C16—H16A
1.387 (3)
C17—H17A
0.9300
P1—F1
1.499 (3)
P1—F2
1.351 (2)
P1—F5
1.359 (2)
P1—F3
1.503 (3)
P1—F6
1.387 (3)
P1—F4
1.385 (3)
Bond Angles (o)
178.96 (8)
C10—C9—C8
90.48 (7)
C10—C9—H9A
88.79 (6)
C8—C9—H9A
88.79 (6)
C9—C10—C11
90.48 (7)
C9—C10—H10A
90.61 (9)
C11—C10—H10A
88.87 (6)
N2—C11—C10
91.86 (6)
N2—C11—C12
89.07 (7)
C10—C11—C12
177.63 (6)
O2—C12—C13
91.86 (6)
O2—C12—C11
88.87 (6)
C13—C12—C11
177.63 (6)
C17—N3—C13
89.07 (7)
C17—N3—Fe1
91.35 (9)
C13—N3—Fe1
117.25 (16)
N3—C13—C14
121.21 (13)
N3—C13—C12
120.94 (13)
C14—C13—C12
62

0.9300
1.384 (3)
0.9300
1.388 (3)
0.9300
1.502 (3)
1.493 (3)
1.353 (2)
1.359 (2)
1.389 (3)
1.389 (3)
0.9300
1.385 (3)
0.9300
1.380 (3)
0.9300
0.9300
1.5783 (14)
1.5910 (16)
1.5960 (14)
1.6043 (16)
1.6152 (14)
1.6156 (14)
118.4 (2)
120.8
120.8
119.12 (18)
120.4
120.4
123.31 (17)
119.35 (16)
117.13 (16)
122.36 (17)
121.24 (17)
115.58 (15)
117.05 (15)
122.23 (12)
120.63 (12)
123.05 (18)
118.69 (15)
117.92 (16)

Table 2-5 (cont’d)
N1—C1—C2
N1—C1—H1A
C2—C1—H1A
C3—C2—C1
C3—C2—H2A
C1—C2—H2A
C2—C3—C4
C2—C3—H3A
C4—C3—H3A
C3—C4—C5
C3—C4—H4A
C5—C4—H4A
N1—C5—C4
N1—C5—C6
C4—C5—C6
C11—N2—C7
C11—N2—Fe1
C7—N2—Fe1
O1—C6—C5
O1—C6—C7
C5—C6—C7
N2—C7—C8
N2—C7—C6
C8—C7—C6
C9—C8—C7
C9—C8—H8A
C7—C8—H8A
N1—C1—C2
N1—C1—H1A
C2—C1—H1A
C3—C2—C1
C3—C2—H2A
C1—C2—H2A
C2—C3—C4
C2—C3—H3A
C4—C3—H3A
C3—C4—C5
C3—C4—H4A
C5—C4—H4A
N1—C5—C4
N1—C5—C6
C4—C5—C6
C11—N2—C7
C11—N2—Fe1

122.75 (18)
118.6
118.6
119.43 (19)
120.3
120.3
118.72 (19)
120.6
120.6
118.95 (19)
120.5
120.5
122.54 (18)
119.61 (16)
117.68 (17)
116.81 (17)
121.34 (12)
121.65 (12)
120.26 (17)
119.44 (18)
119.75 (16)
122.65 (17)
120.20 (17)
117.13 (17)
119.58 (18)
120.2
120.2
122.6 (3)
118.7
118.7
119.3 (3)
120.3
120.3
119.1 (3)
120.5
120.5
118.7 (3)
120.6
120.6
122.7 (2)
119.7 (2)
117.5 (2)
116.7 (2)
121.29 (17)

C15—C14—C13
C15—C14—H14A
C13—C14—H14A
C16—C15—C14
C16—C15—H15A
C14—C15—H15A
C17—C16—C15
C17—C16—H16A
C15—C16—H16A
N3—C17—C16
N3—C17—H17A
C16—C17—H17A
F1—P1—F2
F1—P1—F5
F2—P1—F5
F1—P1—F3
F2—P1—F3
F5—P1—F3
F1—P1—F6
F2—P1—F6
F5—P1—F6
F3—P1—F6
F1—P1—F4
F2—P1—F4
F5—P1—F4
F3—P1—F4
F6—P1—F4
C13—C14—C15
C13—C14—H14A
C15—C14—H14A
C16—C15—C14
C16—C15—H15A
C14—C15—H15A
C15—C16—C17
C15—C16—H16A
C17—C16—H16A
N3—C17—C16
N3—C17—H17A
C16—C17—H17A
O6A—P1—O5A
O6A—P1—O2A
O5A—P1—O2A
O6A—P1—O4A
O5A—P1—O4A
63

118.65 (18)
120.7
120.7
118.68 (18)
120.7
120.7
119.53 (19)
120.2
120.2
122.83 (17)
118.6
118.6
91.46 (9)
91.42 (8)
90.78 (9)
90.33 (9)
178.01 (9)
90.02 (8)
179.03 (9)
89.01 (8)
89.42 (7)
89.18 (8)
90.54 (8)
89.67 (9)
177.98 (8)
89.47 (9)
88.62 (7)
118.6 (2)
120.7
120.7
119.0 (3)
120.5
120.5
119.5 (3)
120.2
120.2
122.7 (2)
118.6
118.6
91.35 (13)
91.40 (11)
90.54 (12)
90.39 (13)
178.14 (13)

Table 2-5 (cont’d)
C7—N2—Fe1
121.81 (17)
O1—C6—C5
120.0 (2)
O1—C6—C7
119.9 (2)
C5—C6—C7
119.5 (2)
N2—C7—C8
122.8 (2)
N2—C7—C6
120.3 (2)
C8—C7—C6
116.8 (2)
C9—C8—C7
119.6 (3)
C9—C8—H8A
120.2
C7—C8—H8A
120.2
Symmetry codes: (i) −x, y, −z+1/2.

O2A—P1—O4A
O6A—P1—O3A
O5A—P1—O3A
O2A—P1—O3A
O4A—P1—O3A
O6A—P1—O1A
O5A—P1—O1A
O2A—P1—O1A
O4A—P1—O1A
O3A—P1—O1A

64

90.09 (12)
90.56 (12)
89.83 (13)
178.00 (11)
89.49 (12)
179.04 (13)
89.22 (12)
89.36 (10)
89.03 (12)
88.68 (10)

Figure 2-13. 1H NMR spectrum of [Cd(dmpp)2](PF6)2 in (CD3)2CO.

65

Figure 2-14. 13C NMR spectrum of [Cd(dmpp)2](PF6)2 in (CD3)2CO.

66

100

262.1

TOF MS ES+

│◄x30

635.1

%

318.1
649.1
131.6

420.0

0
100

300

500

700
m/z

Figure 2-15. ESI-MS of [Cd(dmpp)2](PF6)2.

67

900

%
80
T
70
r
a
n 60
s
m50
i
t 40
t
a 30
n
c 20
e
4000

1440.3
1575.1

1016.4

779.4

1604.1

842.2
3500

3000

2500

2000

1500

Wavenumbers
Figure 2-16. IR spectrum of [Cd(dmpp)2](PF6)2.

68

1000

Figure 2-17. 1H NMR spectrum of 2,6-di(2-ketalpyridyl)pyridine (dkpp) in (CD3)2CO.
Following recrystallization the spectrum is less resolved, therefore the spectrum of the
crude material is shown.

69

Figure 2-18. 13C NMR spectrum of 2,6-di(2-ketalpyridyl)pyridine (dkpp) in (CD3)2CO.

70

100

TOF MS ES+

378.1

%
400.1

290.1
189.6
0

200

334.1
400

600
m/z

Figure 2-19. ESI-MS of 2,6-di(2-ketalpyridyl)pyridine (dkpp).

71

800

%
50
T
r 45
a
n 40
s
35
m
i
t 30
t 25
a
n 20
c
e 15
4000

3046.6
2886.5

762.5
834.5

1442.4
1588.8

1215.4
992.2
1031.5
1133.1

3500

3000

2500

2000

1500

Wavenumbers
Figure 2-20. IR spectrum of 2,6-di(2-ketalpyridyl)pyridine (dkpp).

72

1000

REFERENCES

73

REFERENCES

(1)

Jamula, L. L. Masters Thesis, Michigan State University, 2010.

(2)

Brown, A. M. PhD Dissertation, Michigan State University, 2011.

(3)

Doedens, R.; Dahl, L. Journal of the American Chemical Society 1966, 88, 4847–
4855.

(4)

McGhee, L.; Siddique, R. M.; Winfield, J. M. Journal of the Chemical Society,
Dalton Transactions 1988, 1309.

(5)

Sauvage, J.; Collin, J.; Chambron, J.; Guillerez, S. Chemical Reviews 1994, 94,
993–1019.

(6)

Hammarström, L.; Johansson, O. Coordination Chemistry Reviews 2010, 254,
2546–2559.

(7)

Schramm, F.; Meded, V.; Fliegl, H.; Fink, K.; Fuhr, O.; Qu, Z.; Klopper, W.; Finn,
S.; Keyes, T. E.; Ruben, M. Inorganic Chemistry 2009, 48, 5677–84.

(8)

Abarca, B.; Ballesteros, R.; Elmasnaouy, M. Tetrahedron 1998, 54, 15287–15292.

(9)

Stamatatos, T. C.; Efthymiou, C. G.; Stoumpos, C. C.; Perlepes, S. P. European
Journal of Inorganic Chemistry 2009, 3361–3391.

(10)

De Jonge, A. P.; den Hertog, H. J.; Wibaut, J. P. Recueil des Travaux Chimiques
des Pays-Bas et de la Belgique 1951, 70, 989–996.

(11)

Perrin, D. D.; Armarego, W. L. F. Purification of Laboratory Chemicals, 3rd ed.;
Pergamon Press: New York, 1988.

(12)

Dyker, G.; Muth, O. European Journal of Organic Chemistry 2004, 4319–4322.

(13)

Soutif, J.-C. J.-C.; Doan, N. T. H. H.; Montembault, V. European Polymer Journal
2006, 42, 126–132.

(14)

Newkome, G. R.; Joo, Y. J.; Evans, D. W.; Fronczek, F. R.; Baker, G. R. The
Journal of Organic Chemistry 1990, 55, 5714–5719.

(15)

Goldsmith, C. R.; Stack, T. D. P. Inorganic Chemistry 2006, 45, 6048–55.

(16)

Yan, C.; Chen, Q.; Chen, L.; Feng, R.; Shan, X.; Jiang, F.; Hong, M. Australian
Journal of Chemistry 2011, 64, 104.
74

(17)

Newkome, G. R.; Taylor, H. C. R.; Fronczek, F. R.; Delord, T. J. The Journal of
Organic Chemistry 1984, 49, 2961–2971.

(18)

COSMO V1.58, Software for the CCD Detector Systems for Determining Data
Collection Parameters 2008.

(19)

APEX2 V2008.5-0, Software for the CCD Detector System 2008.

(20)

SAINT V 7.34 Software for the Integration of CCD Detector System 2008.

(21)

SADABS V2.008/2 Program for absorption corrections using Bruker-AXS CCD
based on the method of Robert; Blessing, R. H. Acta Cryst. 1995, A51, 33–38.

(22)

Sheldrick, G. M. Acta Cryst. 2008, A64, 112–122.

(23)

Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; J. A. Montgomery, J.; Vreven, T.; Kudin, K. N.; Burant, J. C.;
Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.;
Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.;
Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A.
J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth,
G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels,
A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Ortiz, J. V; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.;
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Challacombe, M.; Gill, P. M. W.; Chen, B. J. W.; Wong, M. W.; Gonzalez, C.;
Pople, J. A. Gaussian 03, Revision D.01 2004.

(24)

Chen, X.-D.; Mak, T. C. W. Inorganica Chimica Acta 2005, 358, 1107–1112.

(25)

Abarca, B.; Ballesteros, R.; Chadlaoui, M. ARKIVOC 2008, 2008, 73–83.

(26)

Goldsmith, C. R.; Jonas, R. T.; Stack, T. D. P. Journal of the American Chemical
Society 2002, 124, 83–96.

(27)

Vedernikov, A. N.; Huffman, J. C.; Caulton, K. G. Inorganic Chemistry 2002, 41,
6244–6248.

(28)

Chen, X.-D.; Mak, T. C. W. Journal of Molecular Structure 2005, 748, 183–188.

(29)

Clayden, J.; Greeves, N.; Warren, S.; Peter Wothers Organic Chemistry; Oxford
University Press: New York, 2001.

75

(30)

Chen, X.-M. X.-D.; Du, M.; He, F.; Mak, T. C. W. Polyhedron 2005, 24, 1047–
1053.

(31)

Georgopoulou, A. N.; Sanakis, Y.; Raptopoulou, C. P.; Psycharis, V.; Boudalis, A.
K. Polyhedron 2009, 28, 3251–3256.

(32)

Boudalis, A. K.; Donnadieu, B.; Nastopoulos, V.; Clemente-Juan, J. M.; Mari, A.;
Sanakis, Y.; Tuchagues, J.-P. J.-P.; Perlepes, S. P. Angewandte Chemie
(International ed. in English) 2004, 43, 2266–70.

(33)

Boudalis, A. K.; Sanakis, Y.; Clemente-Juan, J. M.; Donnadieu, B.; Nastopoulos,
V.; Mari, A.; Coppel, Y.; Tuchagues, J.-P.; Perlepes, S. P. Chemistry - A European
Journal 2008, 14, 2514–26.

(34)

Rau, S.; Schwalbe, M.; Losse, S.; Görls, H.; McAlister, C.; MacDonnell, F. M.;
Vos, J. G. European Journal of Inorganic Chemistry 2008, 2008, 1031–1034.

(35)

Phelan, N. F.; Orchin, M. Journal of Chemical Education 1968, 45, 633–637.

(36)

Limacher, P. a.; Lüthi, H. P. Wiley Interdisciplinary Reviews: Computational
Molecular Science 2011, 1, 477–486.

(37)

Egli, M.; Sarkhel, S. Accounts of Chemical Research 2007, 40, 197–205.

(38)

Wan, C.-Q.; Chen, X.-D.; Mak, T. C. W. CrystEngComm 2008, 10, 475.

(39)

Mooibroek, T. J.; Gamez, P.; Reedijk, J. CrystEngComm 2008, 10, 1501.

(40)

Kumar Seth, S.; Dey, B.; Kar, T.; Mukhopadhyay, S. Journal of Molecular
Structure 2010, 973, 81–88.

(41)

Quiñonero, D.; Estarellas, C.; Frontera, A.; Deyà, P. M. Chemical Physics Letters
2011, 508, 144–148.

(42)

Egli, M.; Gessner, R. V. Proc. Natl. Acad. Sci. U. S. A. 1995, 92, 180–184.

(43)

Chambers, J.; Eaves, B.; Parker, D.; Claxton, R.; Ray, P. S.; Slattery, S. J.
Inorganica Chimica Acta 2006, 359, 2400–2406.

(44)

Katsoulakou, E.; Lalioti, N.; Raptopoulou, C. P.; Terzis, A.; Manessi-Zoupa, E.;
Perlepes, S. P. Inorganic Chemistry Communications 2002, 5, 719–723.

(45)

Katsoulakou, E.; Bekiari, V.; Raptopoulou, C. P.; Terzis, A.; Lianos, P.; ManessiZoupa, E.; Perlepes, S. P. Spectrochimica Acta. Part A, Molecular and
Biomolecular Spectroscopy 2005, 61, 1627–38.

76

(46)

Lalioti, N.; Raptopoulou, C. P.; Terzis, A.; Aliev, A. E.; Gerothanassis, I. P.;
Manessi-Zoupa, E.; Perlepes, S. P. Angewandte Chemie 2001, 113, 3311–3314.

(47)

Bandoli, G.; Dolmella, a; Tisato, F.; Porchia, M.; Refosco, F. Coordination
Chemistry Reviews 2009, 253, 56–77.

(48)

Baker, R. J.; Jones, C.; Kloth, M.; Mills, D. P. New Journal of Chemistry 2004, 28,
207.

(49)

Baker, R. J.; Jones, C. Dalton Transactions 2005, 1341–8.

(50)

Corbett, J. D.; McMullan, R. K. Journal of the American Chemical Society 1955,
77.

(51)

Wilkinson, M., Worrall, I. J., . Journal of Organometallic Chemistry 1975, 93, 39.

(52)

Gerlach, C.; Hönle, W.; Simon, A. Z. Anorg. Allg. Chem. 1982, 7.

(53)

Green, M. L. H.; Mountford, P.; Smout, G. J.; Speel, S. R. Polyhedron 1990, 9,
2763–2765.

(54)

Jurca, T.; Dawson, K.; Mallov, I.; Burchell, T.; Yap, G. P. a; Richeson, D. S.
Dalton Transactions 2010, 39, 1266–72.

(55)

Manessi, A.; Papaefstathiou, G. S.; Raptopoulou, C. P.; Terzis, A.; Zafiropoulos,
T. F. Journal of Inorganic Biochemistry 2004, 98, 2052–62.

(56)

Knight, J. C.; Amoroso, A. J.; Edwards, P. G.; Prabaharan, R.; Singh, N. Dalton
Transactions 2010, 39, 8925–36.

(57)

Gou, L.; Zhang, B.; Hu, H.-M.; Chen, X.-L.; Wang, B.-C.; Wu, Q.-R.; Qin, T.;
Tang, Z.-X. Journal of Molecular Structure 2008, 889, 244–250.

(58)

Wu, Q.-R.; Wang, J.-J.; Hu, H.-M.; Wang, B.-C.; Wu, X.-L.; Fu, F.; Li, D.-S.;
Yang, M.-L.; Xue, G.-L. Inorganic Chemistry Communications 2010, 13, 715–
719.

(59)

Shaikh, T. M. A.; Emmanuvel, L.; Sudalai, A. The Journal of Organic Chemistry
2006, 71, 5043–6.

(60)

Tabatabaeian, K.; Mamaghani, M.; Mahmoodi, N. O.; Khorshidi, a. Catalysis
Communications 2008, 9, 416–420.

(61)

Neumann, R.; Sasson, Y. The Journal of Organic Chemistry 1984, 49, 1282–1284.

77

(62)

Newkome, G. R.; Taylor, H. C. R. The Journal of Organic Chemistry 1979, 44,
1362–1363.

(63)

Newkome, G. R.; Gupta, V. K.; Taylor, H. C. R.; Fronczek, F. R. Organometallics
1984, 3, 1549–1554.

(64)

Newkome, G. R.; Taylor, H. C. R.; Fronczek, F. R.; Gupta, V. K. Inorganic
Chemistry 1986, 25, 1149–1154.

(65)

Sun, J.; Dong, Y.; Cao, L.; Wang, X.; Wang, S.; Hu, Y. The Journal of Organic
Chemistry 2004, 69, 8932–4.

(66)

Das, S.; Borah, R.; Devi, R.; Thakur, A. Synlett 2008, 2741–2762.

(67)

Kumar, D.; Kumar, R.; Chakraborti, A. Synthesis 2008, 1249–1256.

(68)

Olson, M. E.; Carolan, J. P.; Chiodo, M. V.; Lazzara, P. R.; Mohan, R. S.
Tetrahedron Letters 2010, 51, 3969–3971.

78

Chapter 3. Development of Symmetric Analogs to Investigate the Influence of the
Bridging Group

3.1 Introduction
The development of the highly symmetric [Fe(dcpp)2]2+ provided a nice starting
point to our investigation.

Introducing a nearly perfect octahedral coordination

environment appears to give rise to fascinating properties. The drastic difference in
physical properties and seemingly significant increase in ligand field strength as
compared to other known iron(II) polypyridyls led us to design analogs to uncover more
information.
The motivation behind the design of the next set of molecules is two-fold. First,
the development of a control molecule is crucial, as an isostructural d10 molecule would
aid in distinguishing properties that may be attributed to the complex from ligand based
effects. The highly reactive carbonyl has proven to be problematic in the coordination
chemistry of Zn(II), Ga(III), and Cd(II), it is therefore in our best interest to develop a
ligand system in which we replace the carbonyl groups with a less reactive bridging
group to allow for more accessibility to a control molecule.

Second, can the

characteristics be attributed strictly to the enhanced geometry or are they more a result of
the identity of the ligand? The electron withdrawing carbonyl groups in the ligand
seemingly have a substantial impact on the properties of the complex so we set out to
prepare ligands with alternate bridging groups to investigate this further.
As described in Chapter 1, an important characteristic that a DSSC sensitizer must
possess is absorptivity in the visible region. Chromophores containing conjugated ligand
systems

fit

this

criteria, and we found
79

in

the investigation of 2,6-di(2-

carboxypyridyl)pyridine (dcpp) that a cross-conjugated ligand fits as well. In an effort to
investigate the importance of the carbonyl bridge, while still retaining the optical
properties, it was pertinent to maintain the cross-conjugation. Also, the  interaction of
the carbonyl groups with the facing pyridyl rings appears to be essential to achieving the
overall geometry of the complex, therefore it was in our best interest to preserve this
aspect of the structure.
We set out to prepare analogs that we presumed would be isostructural with
[Fe(dcpp)2]2+ and allow us to differentiate between the effects of the ligand arrangement
and substituent effects. Two ideal choices for this investigation involve replacing the
carbonyl bridging groups with imine and vinylidene functional groups, shown in Figure
3-1. By systematically moving away from the more electronegative oxygen to nitrogen
then to carbon should allow us to distinguish between the effects of inductive
stabilization of the π* orbitals of the ligand and interligand interactions.

Figure 3-1. Structures of 2,6-di(2-(N-methylimine)pyridyl)pyridine (dmipp) and 2,6di(2-vinylidenepyridyl)pyridine (dvpp).
The simplest imine that may be introduced is a primary ketamine (>C=N-H), but
an imine bearing an alkyl group (>C=N-CH3), or Schiff base, would be more stable and
is more synthetically accessible and was therefore chosen. As for the carbon analog,
vinylidene (>C=CH2), or 1,1-ethenediyl, was chosen for its simplicity and is also
80

synthetically accessible. The ligands 2,6-di(2-(N-methylimine)pyridyl)pyridine (dmipp)
and 2,6-di(2-vinylidene-pyridyl)pyridine (dvpp) were prepared, and as with dcpp, have
the flexibility to coordinate in a symmetrical environment.
Concern was raised at the outset of this study that the geometric differences in
these functional groups may interfere with the ability to achieve an isostructural analog.
Using the X-ray crystal structure of [Fe(dcpp)2]2+ as a starting point, geometry
optimizations of the desired complexes were obtained and indicate that the differences
should not hinder coordination and will likely yield isostructural molecules as shown in
Figure 3-2.

Figure 3-2. Optimized Structures of [Fe(dcpp)2](PF6)2 (left), [Fe(dmipp)2](PF6)2
(center), and [Fe(dvpp)2](PF6)2 (right).
Additional apprehension at the beginning of this investigation involves the Schiff
base analog. First, imines are usually unstable and are easily hydrolyzed to the carbonyl.
Furthermore, with the installation of two additional nitrogens we are introducing potential
coordination sites. Unfortunately, attempts to prepare [Fe(dmipp)2]2+ in the desired
coordination environment have failed, instead the dmipp ligand coordinates through the

81

central pyridyl ring and the imine nitrogens. The synthetic efforts toward the preparation
of this analog will be discussed, along with future strategies to overcome the problem.
Fortunately, the dvpp analog presented no such issue and was successfully prepared, and
the synthesis and characterization will be discussed. The properties were examined and a
comparison with [Fe(dcpp)2]2+ will be presented.
The absence of the highly reactive carbonyl in the dvpp ligand renders it less
susceptible to hydrolysis. This allows for more ease in the preparation of a control
molecule and the [Zn(dvpp)2]2+ complex has been prepared. This analog may be useful
in our continuing efforts in the quest for a control molecule for the [Fe(dcpp)2]2+.
3.2 Experimental
3.2.1 Synthesis
General. All chemicals were of reagent grade, purchased from Sigma-Aldrich,
Alfa Aesar, Acros Organics, or Strem Chemicals and used as received unless otherwise
noted.

Solvents were purchased from Sigma-Aldrich, Jade Scientific, Spectrum,

Mallinckrodt, EMD Chemical, or CCI and were purified using standard purification
techniques.1

All air-sensitive reactions were carried out under inert atmosphere by

standard Schlenk techniques utilizing thoroughly deoxygenated solvents that were
degassed by the freeze—pump—thaw method. 1H NMR and 13C NMR were recorded
with Varian UnityPlus-300 MHz, Varian UnityPlus-500 MHz, and Agilent DDR2 500
MHz spectrometers. Ground state absorption spectra were obtained on a Varian Cary 50
spectrophotometer.

IR spectra were obtained on a Mattson Galaxy 5000 FTIR.

Elemental analysis was obtained through the Analytical facilities at Michigan State
82

University. Electrospray mass spectra (ESI-MS) were obtained from the staff of the
MSU Mass Spectrometry Facility. The characterization data of previously unknown
compounds are included in the appendix at the end of this chapter.
2,6-Di(2-(N-methylimine)pyridyl)pyridine (dmipp).

This compound was

prepared by modification of previously published procedures.2,3

A 200 mL round

bottom flask equipped with a Dean-Stark trap was charged with dcpp (0.165 g, 0.569
mmol), methylamine in methanol (2 mL, 2.0 M, 4.00 mmol), dry methanol (10 mL), and
dry benzene (30 mL). The solution was refluxed for 72 hours under N2, and the solvent
subsequently evaporated to yield a dark green oil. The oil appears to be a mixture of
three products (by TLC and NMR), but attempts to purify resulted in hydrolysis,
therefore the crude oil product is used as-is in the subsequent reaction.
[Fe(dmipp)2](PF6)2. The crude dmipp (0.180 g, 0.569 mmol) from the previous
reaction is dissolved in dry MeOH (~10 mL). The brown solution is transferred to a
second air-free flask containing (NH4)2Fe(SO4)2·6H2O (0.0555 g, 0.142 mmol) which
turns dark blue immediately. The reaction was stirred for 30 min at rt and NH4PF6
(0.100 g, 0.617 mmol) was added. Diethyl ether (10 mL) was added and the solution
placed in the freezer. After 18 hours no precipitate had formed so the solution was
concentrated and water added and the blue/purple product filtered. Crude Yield: 0.1047 g
(76%). The crude product contains at least five compounds that may be separated by
column chromatography on neutral alumina with acetonitrile. The first band to elute was
intense purple and determined to be [Fe(dmipp)2](PF6)2 coordinated through the imine
groups and central pyridine ring. Subsequent fractions did not correspond to the correct
mass (according to ESI-MS) and were increasingly paramagnetic by NMR. There was no
83

evidence of the product in the desired coordination mode.

Characterization of

[Fe(dmipp)2](PF6)2 coordinated through the imine groups: 1H NMR ((CD3)2CO, 500
MHz): δ (ppm) 8.93 (ddd, 4H, J = 4.8, 1.6, 1.0 Hz, 6py-a (pyridyl arm)), 8.55 (t, 2H, J =
8.0 Hz, 4py-b (bridging pyridyl)), 8.40 (d, 4H, J = 8.0 Hz, 3py-b), 8.18 (td, 4H, J = 7.8,
1.7 Hz, 4py-a), 7.95 (dt, 4H, J = 7.8, 1.0 Hz, 6py-a), 7.75 (dtd, 4H, J = 4.8, 1.0 Hz, 5pya), 2.75 (s, 12H, N-CH3). 13C NMR (CD3)2CO, 500 MHz): δ (ppm) 175.0, 160.4, 150.9,
147.8, 137.8, 137.1, 128.4, 126.3, 125.7, 40.8. TOF-MS [ESI, m/z (rel int)]: 343.1 (100)
[C38H34N10Fe]2+, 831.2 (15) {[C38H34N10Fe](PF6)}+. IR (KBr, cm-1):

1588.3 m,

1585.0 m, 1465.1 w, 1384.1 m, 1336.0 m, 1152.5 w, 1118.4 w, 1071.8 w, 840.4 br, 757.9
m.
2,6-Di(2-vinylidenepyridyl)pyridine (dvpp). The target molecule was prepared
analogously to 2,2’-vinylidenedipyridine according to the procedure by Summers et al.4
from dcpp (1.999 g, 6.91 mmol).

The brown crude oil was purified by silica gel

chromatography with acetonitrile as the eluent. The bright yellow band is collected (Rf
= 0.32), which immediately begins to darken after leaving the column and upon
evaporation yields the dark red oil product. Yield 0.688 g (35%). The pure product
should be stored in the dark and used fairly quickly (within days), as it polymerizes to a
tacky, insoluble orange solid. 1H NMR ((CD3)2CO, 500 MHz): δ 8.61 (ddd, 2H, J = 4.8,
1.8, 1.0 Hz, 6py-a), 7.81 (t, 1H, J = 7.8 Hz, 4py-b), 7.77 (td, 2H, J = 7.7, 1.8 Hz, 4py-a),
7.50 (dt, 2H, J = 7.9, 1.0 Hz, 3py-a), 7.44 (d, 2H, J = 7.8 Hz, 3/5py-b), 7.31 (td, 2H, J =
4.8, 1.0 Hz, 5py-a), 6.10 (d, 2H, J = 1.8 Hz), 6.05 (d, 2H, J = 1.8 Hz). 13C NMR
((CD3)2CO, 500 MHz): δ 157.60, 157.03, 148.99, 148.87, 136.84, 136.14, 123.04,

84

122.47, 121.80, 119.39. TOF-MS [ESI, m/z (rel int)]: 286.2 (20) [C19H15N3]H+, 571.4
(100) 2[C19H15N3]H+.
[Fe(dvpp)2](PF6)2.

To an air-free flask dvpp (0.163 g, 0.570 mmol) and

deoxygenated 1:1 MeOH/H2O (30 mL) via cannula were added. The solution was gently
warmed (40°C) under nitrogen to promote dissolution of the oil. A separate air-free flask
was charged with (NH4)2Fe(SO4)2·6H2O (0.0690 g, 0.176 mmol) and 1:1 MeOH/H2O
(20 mL). The Fe(II) solution was transferred via cannula to the ligand solution which
immediately turned red, darkening quickly to deep brown. The reaction was allowed to
stir for 2 hours at room temperature. A separate air-free flask was charged with 20
equivalents of NH4PF6 (0.605 g, 3.73 mmol) and 1:1 MeOH/H2O (10 mL) which was
then transferred via cannula to the reaction mixture. The solution was concentrated by
evaporation under a stream of N2 yielding brown precipitate which was filtered and
rinsed with H2O. The crude product was dissolved in acetonitrile and the solution was
washed with hexanes to remove grease. The acetonitrile solution was concentrated then
purified by passage through a basic alumina column. Two distinct bands form, a red
band followed by green band.

The product (red band) eluted first and may be

recrystallized by slow evaporation of methanol or by diethyl ether vapor diffusion into a
1:1 acetone/methanol solution of the complex.

Yield: 0.0822 g (51%). 1H NMR

((CD3)2CO, 500 MHz): δ 8.31 (t, 2H, J = 7.9 Hz, 4py-b), 8.03 (d, 4H, J = 7.9 Hz, 3/5pyb), 7.99 (td, 4H, J = 7.7, 1.4 Hz, 4py-a), 7.81 (d, 4H, J = 7.7, 0.7 Hz, 3py-a), 7.30 (d, 4H,
J = 5.9, 0.7 Hz, 6py-a), 7.05 (td, 4H, J = 5.9, 1.4 Hz, 5py-a), 5.69 (s, 4H), 5.44 (s, 4H).
13C

NMR ((CD3)2CO, 500 MHz): δ 162.60, 162.32, 158.14, 144.18, 139.67, 138.87,

85

125.57, 124.75, 124.11, 123.16.

TOF-MS [ESI, m/z (rel int)]:

313.1 (100)

[C38H30N6Fe]2+, 771.1 (1) {[C38H30N6Fe](PF6)}+. IR (KBr, cm-1): 3100.0 w, 1594.8
m, 1566.5 m, 1478.8 s, 1436.6 m, 1384.8 m, 1287.1 w, 1164.4 m, 1091.3 w, 1021.0 w,
960.5 m, 842.3 br, 759.9 s. Elemental Analysis for C38H30N6FeP2F12, Calculated: C,
49.80; H, 3.30; N, 9.17. Found: C, 49.42; H, 3.29; N, 8.99. UV-Vis (CH3CN) λ(ε(M1cm-1)):

281 nm (33000), 384 nm (7700), 445 nm (6800), 491 nm (7000).

[Zn(dvpp)2](PF6)2. The compound was prepared by modification of a literature
procedure.5 To a 50 mL Erlenmeyer flask was added dvpp (0.0389 g, 0.136 mmol) and
CHCl3 (6 mL). To a separate flask was added Zn(CH3CO2)2·2H2O (0.0123 g, 0.0560
mmol) and MeOH (5 mL). The Zn solution was transferred to the ligand solution by
pipet and allowed to stir for 30 min. NH4PF6 (0.0600 g, 0.368 mmol) was dissolved in
methanol and added to the reaction mixture. Diethyl ether was added to precipitate the
peach product.

The product was recrystallized by slow evaporation of a methanol

solution yielding colorless plate crystals. Yield: 0.0341 g (66%). 1H NMR ((CD3)2CO,
500 MHz): δ 8.36 (t, 2H, J = 7.9 Hz, 4py-b), 8.15 (td, 4H, J = 7.8, 1.5 Hz, 4py-a), 7.95
(d, 4H, J = 7.9 Hz, 3/5py-b), 7.93 (dd, 4H, J = 5.4, 0.7 Hz, 6py-a), 7.86 (d, 4H, J = 7.9
Hz, 3py-a), 7.49 (td, 4H, J = 6.5, 0.7 Hz, 5py-a), 5.66 (s, 4H), 5.57 (s, 4H). 13C NMR
((CD3)2CO, 500 MHz): δ 157.30, 157.24, 148.43, 142.67, 142.31, 141.26, 126.92,
125.77, 125.07, 124.66. TOF-MS [ESI, m/z (rel int)]: 317.1 (2) [C38H30N6Zn]2+.

86

3.2.2 Physical Measurements
X-ray Structure Determination. Single-crystal X-ray diffraction data was
acquired at the X-ray facility of Michigan State University by Dr. Richard J. Staples.
Data were collected as described in Chapter 2 of this dissertation.
Cyclic Voltammetry. Electrochemical measurements were carried out using a
CH Instruments CH620D electrochemical analyzer to determine the E1/2 for ligand
reductions and metal oxidations of each complex. Solutions of the compounds were
prepared in distilled CH3CN containing NBu4PF6 (ca. 0.1 M) as the supporting
electrolyte.

A standard three-electrode setup was used with a platinum working

electrode, carbon rod counter electrode, and Ag/AgCl or Ag/AgNO3 electrode as the
reference. All measurements were made inside an inert atmosphere glovebox. Data was
acquired at a scan rate of 50 mV s-1. After data collection ferrocene was added to
subsequent scans for reference.
3.2.3 Theoretical Calculations
Geometry Optimizations. The initial geometries of the molecules were
generated using SPARTAN or GaussView, or from crystal structure coordinates when
available, and subsequently optimized in two steps.6,7 The first optimization was
performed using the density functional B3LYP with the 3-21G** basis set, followed by a
second optimization using the B3LYP functional with the 6-31G* basis set. Frequency
calculations at the B3LYP/6-31G* level of theory were performed on the final optimized
structures to ensure that these geometries corresponded to global minima.

87

3.3 Results and Discussion
3.3.1 Significance of the  Interaction in Achieving the Preferred Geometry
Our pursuit to prepare [Fe(dcpp)2]2+ by the method of simultaneous ligand
oxidation/coordination, that was described in Chapter 2, was complicated by the
oxidation of iron(II) from the presence of oxygen. Therefore, efforts were made to
exclude

oxygen

and

obtain

the

methylene-bridged

coordination

compound

[Fe(dmpp)2]2+, as shown in Scheme 3-1. After many attempts there was never any
evidence of the desired complex by ESI-MS. By 1H NMR the various samples are
paramagnetic leading to the presumption that this complex may be similar to the
hexakispyridyl complex [Fe(py)6]2+. The absence of the carbonyl bridge results in a
weaker ligand field that cannot stabilize the low spin state, therefore the Fe(dmpp)2]2+ is
likely high spin and unstable in solution.

Scheme 3-1. Attempted synthesis of [Fe(dmpp)2](PF6)2.
As discussed in Chapter 2, the  interaction of the carbonyl oxygen with the
facing pyridyl rings appears to be essential to achieving the overall geometry of the
88

[Fe(dcpp)2]2+ complex. Introducing the imine and vinylidene bridges should allow for a
similar π interaction. In the detailed analysis of the Cambridge Structure Database that
was described in Chapter 2, the authors found that carbonyl lone pair (l.p.)—
interactions are quite common with nearly 6000 occurrences.8 They also found many
cases of imine groups with nitrogen l.p.— interactions, with over 1000 hits, but on
average, this interaction is notably weaker than the carbonyl interaction.

No l.p.-π

interaction is possible with the vinylidene bridge, but the more commonly known π-π
stacking interaction may be present.
We normally think of π-stacking as the non-covalent interaction occurring
between aromatic moieties, with an example of this in the stacking of benzene rings. The
presumption that aromaticity is a requirement for π-stacking has recently been in
question. Computational studies have suggested that stronger interactions may be
achieved between benzene and a non-aromatic moiety than between two benzene rings,
with π localization stabilizing the stacking interaction.9 This suggests that the blanket
phrase describing π-interactions as “aromatic interactions” is a misleading term and
should be more generally defined. Furthermore, a recent review by Martinez and Iverson
brings the use of the terms “-stacking” and “ interactions” into question, finding that
they are also overused in the literature and do not accurately describe the forces that drive
non-covalent association between molecules.10
With the recent debate over terminology and continually evolving research on
noncovalent interactions I am hesitant to suggest what interaction could exist.9–12 I can
only presume that the installation of a vinylidene bridging group may introduce sufficient

89

stabilization to allow for complex formation to occur, whether it be a result of the crossconjugation in the ligand or through a so-called “ interaction.”
3.3.2 Synthesis and Coordination of 2,6-Di(2-(N-methylimine)pyridyl)pyridine
Imines are known to be unstable and are easily hydrolyzed to the carbonyl. In
general, imines that are formed from ketones and primary amines are only stable enough
to isolate if the carbon or nitrogen bears an aromatic substituent. The target molecule
dmipp bears the pyridyl substituents on the imine carbons which we hoped would be
sufficient to stabilize the imine. Furthermore, if coordination with an iron(II) center
could occur in the desired conformation, the l.p.-π interaction with the nitrogen of the
imine and the facing pyridyl rings could further stabilize the imines.
Dmipp has been successfully prepared under various conditions according to the
general method shown in Scheme 3-2. In all cases, dry solvent must be used and water
should be removed from the reaction as it is formed. For imine formation in general, an
acid catalyst is added to speed up the reaction, with optimal pH 4-6. Below pH 4 too
much of the amine is protonated for the reaction to proceed and above pH 6 the proton
concentration is too low to allow for the protonation of the OH leaving group in the
dehydration step. The acid catalyst is not necessarily required and with more rigorous
heating and longer reaction time the reaction may still proceed.

90

Scheme 3-2. General synthetic scheme for the preparation of 2,6-di(2-(N-methylimine)pyridyl)pyridine (dmipp).
Initially, a route described by Yorke et al. was adapted for the preparation of
dmipp.13 Amberlyst-15 was employed as the acid catalyst in the reaction, which was
refluxed for 72 hours utilizing a Dean-Stark trap for water removal. There is evidence of
product formation, but it could not be isolated cleanly so alternate routes were pursued.
The acid catalyst was not required in this alternative procedure, which was also refluxed
for 72 hours with a Dean-Stark trap, with no obvious difference in conversion.2,3 As
with the previous method, product purification was problematic.
A third procedure was pursued in which the compound may be prepared via vapor
diffusion. The method by Thienthong et al. was described as a simplified procedure by
which imines may be prepared quantitatively, requiring little or no work-up or
purification.14 The reaction is carried out in an H-tube in which one arm is charged with
a volatile amine while the other arm contains the ketone and complete conversion can be
achieved in 24-48 hours by simple vapor diffusion. The authors investigated many
ketones and found that the reaction is assisted by a hydrogen bonding interaction with the
oxygen of the carbonyl and may proceed neat if a hydroxyl group is present in the
starting ketone. If no OH group is present, a small amount of dry methanol may be added
to the ketone and the reaction proceeds quantitatively. This method was attempted
91

utilizing a “cow” type distillation receiver. One arm was charged with 20 mL of 2.0 M
CH3NH2 in methanol and a second arm was charged with dcpp, 1 mL of dry methanol,
and a stir bar. The apparatus was sealed and the ketone suspension was stirred for 24
hours with no noticeable change, after another 48 hours the ketone had completely
solubilized and turned yellow, signifying that conversion to the imine had occurred. As
with the two previous procedures the oil product was found to be a mixture, with some of
the starting dcpp remaining as well as some partially converted product in with the
desired product. Allowing the reaction to proceed longer may produce better yield, but to
date this was not attempted.
By all routes, the preparation of the imine yields a mixture. The imine product is
susceptible to hydrolysis and therefore any attempt to purify by recrystallization or
column chromatography has proven to be difficult, resulting in reversion to dcpp.
Coordination of the dmipp ligand to iron(II) could stabilize the imine groups, possibly
allowing for cleaner separation of the compounds with less chance of hydrolysis,
therefore, the crude oil product was used without purification in the coordination
reaction. As anticipated, the reaction yielded a mixture of products due to the presence of
partially and unconverted dcpp in the ligand mixture, but could be easily separated by
column chromatography on neutral alumina. As predicted, coordination of the dmipp
ligand to iron(II) stabilized the imine groups, unfortunately it was achieved by
coordination of the imine nitrogens instead of the desired pyridyl coordination as shown
in Scheme 3-3.

92

Scheme 3-3. Attempted preparation of the desired [Fe(dmipp)2]2+ resulting in the
alternate coordination mode.
I was unable to obtain crystals suitable for X-ray crystal structure determination to
verify the alternate coordination mode, but from ground state electronic absorption and
1H

NMR data it became clear that the desired complex was not achieved.

[Fe(dmipp)2]2+ was not previously known, but iron(II) complexes possessing similar
ligand systems that coordinate through imine groups are known and have been well
characterized.15–18 The ground state absorption spectrum for Fe(dmipp)2]2+ in
acetonitrile is nearly identical to those published for similar compounds (Figure 3-3).
Toma and coworkers performed calculations to identify the features in the absorption
spectrum: the peaks at 600 nm and 485 nm are metal to ligand charge transfer (MLCT)

93

bands arising from dxy → π* and dxz, dyz → π*, respectively, and the weak band at 700
nm is attributed to a low energy ligand field transition.16,17 The consistency of features
in the [Fe(dmipp)2]2+ absorption spectrum with the literature reports provides strong
evidence of the coordination mode.

Absorbance

1.0

0.8

0.5

0.3

0.0
300

400

500
600
Wavelength (nm)

700

800

Figure 3-3. Ground state absorption spectrum of [Fe(dmipp)2]2+.
A second, more crucial piece of evidence is obtained from the 1H NMR spectrum
of the complex. Upon coordination of polypyridyl ligands to a metal center there is a
drastic shift of the 6-position protons, which is well known in terpyridine chemistry.19,20
The 6-position protons on the terminal rings of polypyridyl are the least shielded protons
in the free ligand and therefore their signal appears far downfield. A free ligand is
oriented in an anti arrangement to avoid strain between the lone pairs on the pyridyl
nitrogens, and upon coordination the pyridyls reorient into a syn arrangement causing the
94

observed upfield shift of the 6-position protons. These protons are influenced by the
metal-ligand bond as well as the close proximity to the adjacent ligand. We observe a
shift of the other protons as well, but it is usually not as drastic as with the 6-position.
This characteristic shift can be seen in the 1H NMR spectra of dcpp and [Fe(dcpp)2]2+,
shown in Figure 3-4. Since a clean 1H NMR spectrum of the dmipp ligand has not been
obtained, I cannot make the same comparison with the [Fe(dmipp)2]2+ spectrum;
however the 6-position protons in the complex appear far downfield signifying that they
are not affected by coordination to the metal (Figure 3-5) providing more evidence for the
alternate coordination mode.

Figure 3-4. Characteristic shift of the 6-position protons shown by comparison of the 1H
NMR ((CD3)2CO) of the dcpp ligand (top) and [Fe(dcpp)2]2+ complex (bottom).

95

Figure 3-5. 1H NMR ((CD3)2CO) of the [Fe(dmipp)2]2+ complex, note the chemical
shift of the 6-position protons far downfield.
Since the imine complex does not possess the desired conformation it is not useful
for the current investigation. We are currently pursuing the installation of a bulkier R
group to the imine nitrogen in an effort to prevent binding at these sites. First on deck is
the reaction of dcpp with t-butylamine.

t-Butylamine is much less volatile than

methylamine which may lead to more difficulty in obtaining the product and more
rigorous conditions are likely necessary. Another possibility that has been considered is
attempting the imine formation with the ligand already bound to the Fe(II) center. The
chances of this route producing results is highly unlikely, since the [Fe(dcpp)2]2+ appears
to be extremely stable. Furthermore, the carbonyls are stabilized by the l.p.-π interaction
with the pyridyl rings likely rendering them unreactive.
3.3.3 Synthesis and Coordination of 2,6-Di(2-vinylidenepyridyl)pyridine
2,6-Di(2-vinylidenepyridyl)pyridine (dvpp) was not previously known, however
is readily synthetically accessible since olefination reactions are quite commonplace with

96

a wide variety of options available.

With 2,6-di(2-carboxypyridyl)pyridine (dcpp)

already in hand it was fairly easy to develop the chemistry for the dvpp ligand. The
Wittig reaction was employed and dvpp was prepared analogously to 2,2’vinyldipyridine.4,21 The product was purified by column chromatography and elutes as a
yellow oil that quickly darkens to red upon standing. The product was used immediately
in the subsequent coordination reaction, as shown in Scheme 3-4.

Scheme 3-4. Synthesis of 2,6-di(2-vinylidenepyridyl)pyridine (dvpp) and [Fe(dvpp)2]2+.
The dvpp product contained a small amount of partially converted ligand in which
one carbonyl remains intact while the other was converted to the olefin. The presence of
this impurity became apparent in the subsequent formation of the iron(II) complex. It is
possible that this is a byproduct of the reaction that is carried through the silica column
purification, despite much care being taken in isolation and clear separation by TLC. It is
also possible that it may be occurring during the purification, as it is known that olefins

97

can undergo oxidative cleavage on silica in the presence of light and air.22,23

When

repeating this reaction, much more care was taken to exclude light during the workup and
purification. However, if the impurity is carried through to the coordination reaction it
may be easily removed during purification of the Fe(II) complex.
The product of the coordination reaction was obtained as a brown solid which was
easily separated into two distinct products by column chromatography on basic alumina.
The first band is red and is determined by ESI-MS to be the pure intended product. The
second band is dark green and according to ESI-MS is two mass units too heavy
signifying that it contains one carbonyl. Also, there is a carbonyl stretch in the IR
spectrum of this green byproduct.
3.3.3 Properties of Bis(2,6-di(2-vinylidenepyridyl)pyridine)iron(II)
Single crystals of [Fe(dvpp)2](PF6)2 suitable for X-ray crystallography were
obtained by slow evaporation of a methanol solution of the complex.

The crystal

structure of [Fe(dvpp)2]2+ was solved and found to possess an octahedral coordination
environment isostructural with the [Fe(dcpp)2]2+, shown in Figure 3-6.

The

crystallographic data is listed in Table 3-1 and selected geometric parameters are listed in
Tables 3-2 and 3-3. With an isostructural analog in hand we can now investigate the
source of the seemingly unique properties of the [Fe(dcpp)2]2+ complex.

98

d (centroid to CH2) = 3.23 – 3.38 Å o
ω (angle between planes) = 16.8-19.1
Figure 3-6. ORTEP drawing of the cation of [Fe(dvpp)2]2+ obtained from single-crystal
X-ray structure determination. Atoms are represented as 50% probability thermal
ellipsoids. Hydrogen atoms and anions are omitted for clarity. Also shown is the schematic
of the structural parameters determined from the crystal structure used to characterize the π—π
interaction.

Table 3-1. Crystallographic data for [Fe(dvpp)2](PF6)2.
[Fe(dvpp)2](PF6)2 a
Empirical formula
Formula weight
Temperature/K
Crystal system
Space group
a/Å
b/Å

C38H30F12FeN6P2 ρcalcmg/mm3
916.47
/mm-1
173.01
F(000)
Orthorhombic
2θ range
Pbca
Reflections collected
13.5610 (3)
Independent reflections
22.4737 (5)
R(int)

c/Å

25.5597 (5)

Final R indexes [I>=2σ (I)]

Volume/Å3

7789.7 (3)

Final R indexes [all data]

Z

8

Goodness-of-fit on F2

a Obtained with graphite monochromated Cu Kα radiation (λ = 1.54178
Fc  Fo . c wR2 = {  [w(Fo2  Fc2)2 /  [w(Fo2)2]}1/2.

99

1.563
4.746
3712
65.1° to 134.7°
19560
5468
0.1978
b = 0.0884
R1
wR2 c = 0.1911
R1 = 0.1659
wR2 = 0.2363
1.017

Å) b R1 =  Fo  

Table 3-2. Selected bond lengths from the X-ray crystal structure of [Fe(dvpp)2](PF6)2.
Selected Bond Lengths (Å)
Fe1—N1
1.988
Fe1—N2
1.989
Fe1—N3
1.998
Fe1—N4
1.986
Fe1—N5
1.990
Fe1—N6
1.996

(6)
(6)
(6)
(6)
(6)
(5)

Table 3-3. Selected bond angles from the X-ray crystal structure of [Fe(dvpp)2](PF6)2.

N2—Fe1—N5
N2—Fe1—N4
N5—Fe1—N4
N2—Fe1—N1
N5—Fe1—N1
N4—Fe1—N1
N2—Fe1—N3
N5—Fe1—N3

Bond Angles (o)
179.5 (3)
N4—Fe1—N3
89.4 (2)
N1—Fe1—N3
90.5 (2)
N2—Fe1—N6
89.0 (2)
N5—Fe1—N6
90.3 (2)
N4—Fe1—N6
87.9 (2)
N1—Fe1—N6
90.2 (2)
N3—Fe1—N6
90.4 (2)

92.8 (2)
179.0 (2)
90.4 (2)
89.7 (2)
179.5 (2)
92.5 (2)
86.6 (2)

In order to compare the two analogs and determine the influence of the bridging
groups we first need to look to what is known about substituent inductive effects. The
effects of electron withdrawing and donating substituents in the 4-position of terpyridine
on photophysical and electrochemical properties of ruthenium(II) complexes have been
investigated.24 Introducing any substituent to the 4-position of terpyridine in a Ru(II)
complex results in a red shift of the MLCT. Electron withdrawing substituents are
known to stabilize both the π* (LUMO) and t2g orbitals (HOMO), while electron
donating substituents destabilize them. The substituent effects of 4-substituted terpy
ligands on iron(II) redox properties have also been investigated.25 The research shows
that there is a linear increase in the Fe(II/III) couple as a function of increasing electron
withdrawing nature of the substituents. This inductive influence is greater with Fe(II)
100

than for Co(II) and is attributed to the stronger iron(II)-terpy binding strength. In our
case, the influence may be even more pronounced due to enhanced metal-ligand overlap
from the octahedral environment.
In the optical absorption spectrum, shown in Figure 3-7, we can see that the dvpp
analog possesses molar absorptivity in the visible region, as is key for use in dye
sensitization, but is significantly blue-shifted in comparison to the dcpp analog. Since the
metal-ligand overlap is comparable in the two systems it appears that the substituent
effect of the bridging groups has a significant impact on the optical properties.
Consistent with the what is known in the literature, the electron withdrawing carbonyl
groups lower the energy of the π* (LUMO) and t2g orbitals (HOMO), while the electron
donating vinylidene groups destabilize them. From the lower energy MLCT it appears
that the stabilization of the t2g is more pronounced in the dcpp analog compared to the
destabilization of the t2g in the dvpp, which we attribute to the better energy match of the
π orbitals of the metal and ligand that was proposed in Chapter 2 (Figure 2-8).

101

M o l a r A b s o r p t i v i t y (M-1 cm-1)

Wavelength (nm)

400

12000

500

10000

600

700 800

[Fe(dvpp)2]

2+

[Fe(dcpp)2]

8000

2+

6000
4000
2000
0
30

28

26

24

22

20

18

16

14

12

Energy x103 (cm-1)
Figure 3-7: Ground state electronic absorption spectrum of [Fe(dvpp)2]2+ (red), and
[Fe(dcpp)2]2+ (blue) in CH3CN.
The electrochemical potentials of the Fe(dvpp)2]2+ were measured by cyclic
voltammetry and the first oxidation and reduction potentials are listed in Table 3-4. The
more negative reduction potential is attributed to the electron donating vinylidene bridges
making the ligand harder to reduce. In comparison to the [Fe(dcpp)2]2+ complex, the less
positive oxidation potential is consistent with the inductive influence of ligand
substituents on the Fe(II/III) redox couple. The redox potentials are more in line with
[Fe(bpy)3]2+ and it appears that any stabilization gained by moving toward a more
symmetric geometry is countered by the inductive effect of the electron donating
vinylidene bridges.
102

Table 3-4. Electrochemical potentials for the oxidation and first reduction of
[Fe(bpy)3](PF6)2, [Fe(terpy)2](PF6)2, [Fe(dvpp)2](PF6)2, and [Fe(dcpp)2](PF6)2,
measured with Ag/NO3 or Ag/AgCl reference with 0.1 M TBAPF6 electrolyte in
acetonitrile, externally referenced to ferrocene.
Compound
[Fe(bpy)3]2+
[Fe(terpy)2]2+
[Fe(dvpp)2]2+
[Fe(dcpp)2]2+

E 1 / 2 [ox]
0.665 V
0.715 V
0.640 V
1.295 V

E 1 / 2 [red]
-1.775 V
-1.675 V
-1.840 V
-0.965 V

The isostructural analog [Fe(dvpp)2]2+ has provided some insight into the unique
characteristics of [Fe(dcpp)2]2+ and it is clear that the electron withdrawing carbonyl
groups have a significant impact on the properties. The photophysical investigation into
the ultrafast dynamics of this analog is currently underway. It seems that we have two
compounds at far ends of the spectrum and it would be ideal to obtain an analog
somewhere in the middle, so efforts will continue toward developing an imine analog.

3.3.4 Continuing Quest for a Control Molecule
A motive behind the design of analogs of [Fe(dcpp)2]2+ was to remove the
reactive carbonyl bridges in order to have easier access to a control molecule. The
[Fe(dvpp)2]2+ does not possess the exciting properties that the dcpp analog has, but a
control molecule could still prove to be useful. The zinc(II) analog may be easily
prepared by a simple procedure of adding a Zn(II) salt to the dvpp ligand and stirring for
30 minutes. The product is precipitated from the reaction with ether yielding fairly clean
product which can simply be recrystallized by ether diffusion.
103

With [Zn(dvpp)2]2+ in hand, our efforts are now turning toward back toward
[Zn(dcpp)2]2+ by oxidizing the olefin groups after coordination. As discussed in Chapter
2, after many efforts to oxidize the methylene bridges in [Cd(dmpp)2]2+ we have yet to
find the right conditions, most being too harsh and stripping the ligand from the metal
center. We may have better luck finding mild conditions by which we may oxidize the
vinylidene bridges. A promising possibility is the use of photochemistry, as it has been
shown that olefins can undergo oxidative cleavage on silica gel in the presence of light
and air.22,23
The ESI-MS data on this molecule is worrisome; though the product is present, it
is drastically overshadowed by many other peaks, possibly signifying instability.
Hopefully, the [Zn(dvpp)2]2+ is sufficiently stable to handle oxidative conditions and
remain intact. If it proves to be unstable, it should be just as simple to prepare the Cd(II)
analog which might be more stable toward oxidative conditions. Efforts are continuing
toward a control molecule for [Fe(dcpp)2]2+.
3.4 Concluding Comments
Our efforts to develop a symmetric analog was achieved with the design and
preparation of [Fe(dvpp)2]2+.

This molecule provided a nice contrast to the

[Fe(dcpp)2]2+ and allowed us to investigate the inductive effects of the bridging group of
the ligand. With only two analogs in hand we can only postulate, nevertheless, it appears
that the unique characteristics of the [Fe(dcpp)2]2+ can be attributed to the inductive
effects from the electron withdrawing carbonyl groups. However, these inductive effects

104

are presumably heightened by the enhanced geometry. Work is ongoing toward the
preparation of the imine analog to gain more information.
The other motive for this work was to develop an analog that would allow for
more access to a control molecule.

This also came to fruition, but since the

[Fe(dvpp)2]2+ analog does not possess the same exciting characteristics, the
[Zn(dvpp)2]2+ is not very useful to us as-is. We are currently in pursuit of methodology
by which we may oxidize the olefin groups to carbonyls to achieve the desired
[Zn(dcpp)2]2+.

105

APPENDIX

106

APPENDIX

Figure 3-8. 1H NMR of [Fe(dmipp)2](PF6)2 in (CD3)2CO.

107

Figure 3-9. 13C NMR of [Fe(dmipp)2](PF6)2 in (CD3)2CO.

108

TOF MS ES+

100
343.1

%

336.1
242.3

831.2
449.7

329.1

671.2

717.2
991.2

0
200

400

600
m/z

Figure 3-10. ESI-MS of [Fe(dmipp)2](PF6)2

109

800

1000

%

85

T 80
r
a 75
n
s 70
m
i 65
t
t 60
a
n 55
c 50
e
4000

1336.0 1071.8

1588.3

757.9

1384.1

840.4
3500

3000

2500

2000

1500

Wavenumbers
Figure 3-11. IR of [Fe(dmipp)2](PF6)2

110

1000

Figure 3-12. 1H NMR of 2,6-di(2-vinylidenepyridyl)pyridine (dvpp) in (CD3)2CO.

111

Figure 3-13. 13C NMR of 2,6-di(2-vinylidenepyridyl)pyridine (dvpp) in (CD3)2CO.

112

TOF MS ES+
100

571.4

%
428.8
286.2

588.4

856.6

714.5

0
200

300

400

500
m/z

600

700

Figure 3-14. ESI-MS of 2,6-di(2-vinylidenepyridyl)pyridine dvpp.

113

800

900

Figure 3-15. 1H NMR of [Fe(dvpp)2](PF6)2 in (CD3)2CO.

114

Figure 3-16. 13C NMR of [Fe(dvpp)2](PF6)2 in (CD3)2CO.

115

100
313.1

TOF MS ES+

│◄x30

%
771.1
360.1
645.2

286.1
0
200

400

600
m/z

Figure 3-17. ESI-MS of [Fe(dvpp)2](PF6)2.

116

800

1000

%

85

T 80
r
a 75
n
s 70
m
i 65
t
t 60
a
n 55
c 50
e
4000

3100.0
1164.4
1594.8

1436.6

960.5

1478.8
759.9
842.3
3500

3000

2500

2000

1500

Wavenumbers
Figure 3-18. IR of [Fe(dvpp)2](PF6)2.

117

1000

Table 3-5. Geometric parameters for the X-ray crystal structure of [Fe(dvpp)2](PF6)2.

Fe1
Fe1
Fe1
Fe1
Fe1
Fe1
N1
N1
N2
N2
N3
N3
N4
N4
N5
N5
N6
N6
C1
C2
C3
C4
C5
C6
C6
C8
C9
C10
C11
C12
C13
C13
C15
N1
N1
N1
N1
N2
N2
N2
N4

N1
N2
N3
N4
N5
N6
C1
C5
C8
C12
C15
C19
C20
C24
C27
C31
C34
C38
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
Fe1
Fe1
Fe1
Fe1
Fe1
Fe1
Fe1
Fe1

1.988(6)
1.989(6)
1.998(6)
1.986(6)
1.990(6)
1.996(5)
1.363(10)
1.345(10)
1.326(11)
1.366(11)
1.348(10)
1.364(10)
1.361(11)
1.354(11)
1.376(10)
1.338(10)
1.358(11)
1.360(11)
1.365(12)
1.389(13)
1.371(13)
1.389(11)
1.498(11)
1.361(14)
1.464(13)
1.408(13)
1.37(2)
1.35(2)
1.395(13)
1.494(14)
1.319(13)
1.480(11)
1.366(12)
N2
N3
N5
N6
N3
N5
N6
N1

89.0(2)
179.0(2)
90.3(2)
92.5(2)
90.2(2)
179.3(3)
90.4(2)
87.9(2)

Bond Lengths (Å)
C16
C17
C17
C18
C18
C19
C20
C21
C21
C22
C22
C23
C23
C24
C24
C25
C25
C26
C25
C27
C27
C28
C28
C29
C29
C30
C30
C31
C31
C32
C32
C33
C32
C34
C34
C35
C35
C36
C36
C37
C37
C38
P1
F1
P1
F2
P1
F3
P1
F4
P1
F5
P1
F6
P2
F7
P2
F8
P2
F9
P2
F10
P2
F11
P2
F12
Bond Angles (o)
C18 C19
N4 C20
C20 C21
C21 C22
C22 C23
N4 C24
N4 C24
C23 C24
118

1.387(14)
1.365(14)
1.359(12)
1.377(12)
1.385(18)
1.395(18)
1.402(12)
1.498(13)
1.316(14)
1.472(12)
1.378(12)
1.357(16)
1.373(16)
1.396(12)
1.503(13)
1.319(13)
1.481(13)
1.395(13)
1.367(17)
1.371(16)
1.370(12)
1.616(16)
1.552(14)
1.517(12)
1.578(11)
1.585(12)
1.594(14)
1.539(12)
1.553(10)
1.542(9)
1.628(10)
1.623(8)
1.587(9)
N3
C21
C22
C23
C24
C23
C25
C25

122.6(6)
123.6(9)
118.0(9)
120.4(7)
117.8(10)
122.5(8)
118.4(6)
119.0(8)

Table 3.5 (cont’d).
N4
N4
N4
N4
N5
N5
N6
C1
C5
C5
C8
C8
C12
C15
C15
C19
C20
C24
C24
C27
C31
C31
C34
C34
C38
N1
C1
C4
C3
N1
N1
C4
C7
C7
C8
N2
N2
C9
C10
C11
C10
N2
N2
C11

Fe1
Fe1
Fe1
Fe1
Fe1
Fe1
Fe1
N1
N1
N1
N2
N2
N2
N3
N3
N3
N4
N4
N4
N5
N5
N5
N6
N6
N6
C1
C2
C3
C4
C5
C5
C5
C6
C6
C6
C8
C8
C8
C9
C10
C11
C12
C12
C12

N2
N3
N5
N6
N3
N6
N3
Fe1
Fe1
C1
Fe1
C12
Fe1
Fe1
C19
Fe1
Fe1
Fe1
C20
Fe1
Fe1
C27
Fe1
C38
Fe1
C2
C3
C2
C5
C4
C6
C6
C5
C8
C5
C6
C9
C6
C8
C9
C12
C11
C13
C13

89.4(2)
92.8(2)
90.5(2)
179.5(2)
90.4(2)
89.7(2)
86.8(2)
120.9(5)
122.4(4)
116.5(6)
121.8(5)
118.4(7)
119.7(6)
121.9(5)
117.5(6)
120.5(5)
121.1(5)
121.0(5)
117.5(6)
120.1(5)
121.4(5)
118.6(6)
121.4(5)
117.8(6)
120.5(5)
123.9(7)
118.7(7)
118.4(7)
120.0(7)
122.2(7)
117.0(6)
120.7(7)
118.5(8)
123.0(8)
118.3(7)
119.7(6)
121.6(8)
118.7(8)
119.4(10)
119.5(8)
119.7(9)
121.4(9)
119.9(7)
118.7(7)

C26
C26
C27
N5
N5
C28
C29
C28
C29
N5
N5
C30
C33
C33
C34
N6
N6
C35
C36
C35
C38
N6
F2
F2
F2
F2
F3
F3
F3
F3
F3
F4
F4
F4
F5
F5
F6
F7
F7
F7
F7
F7
F8
F8
119

C25
C25
C25
C27
C27
C27
C28
C29
C30
C31
C31
C31
C32
C32
C32
C34
C34
C34
C35
C36
C37
C38
P1
P1
P1
P1
P1
P1
P1
P1
P1
P1
P1
P1
P1
P1
P1
P2
P2
P2
P2
P2
P2
P2

C24
C27
C24
C25
C28
C25
C27
C30
C31
C30
C32
C32
C31
C34
C31
C32
C35
C32
C34
C37
C36
C37
F1
F4
F5
F6
F1
F2
F4
F5
F6
F1
F5
F6
F1
F6
F1
F8
F9
F10
F11
F12
F10
F11

120.4(8)
121.1(8)
118.2(7)
119.7(6)
120.7(7)
119.7(7)
120.6(8)
119.3(7)
119.2(8)
121.7(8)
119.2(7)
119.2(7)
119.4(8)
121.8(8)
118.6(6)
117.9(6)
120.6(8)
121.5(8)
120.5(8)
118.7(7)
119.6(8)
122.7(8)
92.9(15)
92.9(13)
175.4(17)
89.7(14)
88.7(11)
95.5(16)
97.2(12)
88.9(11)
174.3(11)
171.3(9)
87.7(6)
85.0(8)
86.0(8)
85.9(8)
88.6(9)
98.3(11)
90.7(10)
173.5(13)
89.1(9)
92.3(11)
87.4(9)
172.4(10)

Table 3.5 (cont’d).
C14
C14
C15
N3
N3
C16
C15
C18
C19

C13
C13
C13
C15
C15
C15
C16
C17
C18

C12
C15
C12
C13
C16
C13
C17
C16
C17

121.3(8)
120.0(9)
118.5(6)
117.9(6)
121.8(7)
120.2(6)
120.0(7)
118.4(8)
119.7(8)

F8
F9
F9
F9
F9
F11
F12
F12

120

P2
P2
P2
P2
P2
P2
P2
P2

F12
F8
F10
F11
F12
F10
F10
F11

92.3(8)
86.3(9)
86.5(7)
95.5(8)
176.8(8)
85.3(5)
90.5(8)
85.5(5)

Figure 3-19. 1H NMR of [Zn(dvpp)2](PF6)2 in (CD3)2CO.

121

Figure 3-20. 13C NMR of [Zn(dvpp)2](PF6)2 in (CD3)2CO.

122

REFERENCES

123

REFERENCES

(1)

Perrin, D. D.; Armarego, W. L. F. Purification of Laboratory Chemicals, 3rd ed.;
Pergamon Press: New York, 1988.

(2)

Flores-Chávez, B.; Martínez-Ortega, B. A.; Alvarado-Rodríguez, J. G.; AndradeLópez, N. Journal of Chemical Crystallography 2005, 35, 451–456.

(3)

Andrade-López, N.; Hanna, T. a.; Alvarado-Rodríguez, J. G.; Luqueño-Reyes, A.;
Martínez-Ortega, B. a.; Mendoza-Espinosa, D. Polyhedron 2010, 29, 2304–2310.

(4)

Summers, G. .; Ndawuni, M. .; Summers, C. . Polymer 2001, 42, 397–402.

(5)

Hofmeier, H.; Hoogenboom, R.; Wouters, M. E. L.; Schubert, U. S. Journal of the
American Chemical Society 2005, 127, 2913–21.

(6)

SPARTAN 5.0 ed.; Wavefunction Inc.: Irvine, CA, 1997.

(7)

Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; J. A. Montgomery, J.; Vreven, T.; Kudin, K. N.; Burant, J. C.;
Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.;
Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.;
Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A.
J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth,
G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels,
A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Ortiz, J. V; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.;
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Challacombe, M.; Gill, P. M. W.; Chen, B. J. W.; Wong, M. W.; Gonzalez, C.;
Pople, J. A. Gaussian 03, Revision D.01 2004.

(8)

Mooibroek, T. J.; Gamez, P.; Reedijk, J. CrystEngComm 2008, 10, 1501.

(9)

Bloom, J. W. G.; Wheeler, S. E. Angewandte Chemie (International ed. in English)
2011, 50, 7847–9.

(10)

Martinez, C. R.; Iverson, B. L. Chemical Science 2012, 3, 2191.

(11)

Raju, R. K.; Bloom, J. W. G.; An, Y.; Wheeler, S. E. Chemphyschem : A European
Journal of Chemical Physics and Physical Chemistry 2011, 12, 3116–30.

124

(12)

Quiñonero, D.; Estarellas, C.; Frontera, A.; Deyà, P. M. Chemical Physics Letters
2011, 508, 144–148.

(13)

Yorke, J.; Dent, C.; Decken, A.; Xia, A. Inorganic Chemistry Communications
2010, 13, 54–57.

(14)

Thienthong, N.; Bergman, Y. E.; Perlmutter, P. Synthetic Communications 2009,
39, 2683–2693.

(15)

Clark, R. J. H.; Turtle, P. C.; Strommen, D. P.; Streusand, B.; Kincaid, J.;
Nakamoto, K. Inorganic Chemistry 1977, 16, 84–89.

(16)

Kuwabara, I. H.; Comninos, F. C. M.; Pardini, V. L.; Viertler, H.; Toma, H. E.
Electrochimica Acta 1994, 39, 2401–2406.

(17)

Toma, H. E.; Kuwabara, I. H.; Faria, D. L. A. de J. Braz. Chem. Soc. 1996, 7, 391–
394.

(18)

Scott, J.; Gambarotta, S.; Korobkov, I.; Budzelaar, P. H. M. Journal of the
American Chemical Society 2005, 127, 13019–29.

(19)

Schubert, U. S.; Hofmeire, H.; Newkome, G. R. Modern Terpyridine Chemistry;
Wiley-VCH, 2006.

(20)

Jamula, L. L. Masters Thesis, Michigan State University, 2010.

(21)

Eckhard, I.; Summers, L. Australian Journal of Chemistry 1974, 27, 2511.

(22)

Shim, S. C.; Aronovitch, C.; Mazur, Y. The Journal of Organic Chemistry 1985,
50, 149–150.

(23)

Sigman, M. E.; Barbas, J.; Corbett, S.; Chen, Y.; Ivanov, I.; Dabestani, R. Journal
of Photochemistry and Photobiology A: Chemistry 2001, 138, 269–274.

(24)

Maestri, M.; Armaroli, N.; Balzani, V.; Constable, E. C.; Thompson, A. M. W. C.
Inorganic Chemistry 1995, 34, 2759–2767.

(25)

Chambers, J.; Eaves, B.; Parker, D.; Claxton, R.; Ray, P. S.; Slattery, S. J.
Inorganica Chimica Acta 2006, 359, 2400–2406.

125

Chapter 4. Tuning the Energetics of Excited States by Synthetic Modification of the
Ligand System

4.1 Introduction
Our intentions for moving toward a more symmetric coordination environment
were to collapse the ligand field manifold into discrete states to gain more understanding
into the charge transfer to ligand field deactivation pathway. We achieved the desired
geometry with the first molecule that we pursued and it was found to possess some exciting
properties. The observation of weak emission from this fascinating new molecule has led
us in a new direction. We believe that we may have introduced a strong enough ligand
field to iron(II) to achieve inversion of the lowest energy excited states. If we have in fact
been able to tune the energies of the lower excited states why couldn’t we do the same to
the higher energy levels and achieve inversion of a charge transfer excited state with a
ligand field state.
We propose that the emission arises from the 3T1 state (t2g5 eg1) which will possess
a different geometry than the 1A1 (t2g6) ground and 1MLCT excited states. The 1A1 to
1MLCT

absorption maximum is at 605 nm while the emission feature is centered around

700 nm, which corresponds to an energy separation of ~2200 cm -1. Considering the
displacement of the potential energy surfaces from the difference in geometry, the zero
point energy difference is likely even smaller. We propose that by making synthetic
modifications to the ligand it may be possible to achieve inversion of the 3T1 state with
3MLCT

state. As described in the introductory chapter, a major shortfall of iron(II)

polypyridyls for use in DSSCs is that upon excitation to the MLCT the molecules undergo
ultrafast deactivation to lower lying ligand field states. If we can prepare an iron(II)
126

complex with a charge transfer state as the lowest energy excited state it would open the
door to great possibility for replacing Ru(II) sensitizers in DSSCs.
In order to achieve inversion of the excited states, we have to determine ways to
stabilize the MLCT and destabilize the ligand field states. Tuning the energy of the 3MLCT
has been widely investigated for Ru(II) polypyridyls so we looked to this research for
inspiration.1–6 The introduction of electron withdrawing groups is known to lower the
energy of the MLCT state, which is apparent in our new [Fe(dcpp)2]2+ molecule. The
symmetry and inductive effects of the carbonyl groups in [Fe(dcpp)2]2+ have led to
stabilization of the metal t2g and ligand * orbitals as is evident in the electrochemistry
(Chapter 2).

From this trend, one possibility is that we could add more electron

withdrawing substituents to lower the energy even further. Yet one must proceed with
caution as too many electron withdrawing substituents may lower the basicity of the
nitrogens and cause difficulty with the formation of the complex.
Among many other possibilities, one stands out as a promising strategy: the
introduction of phenyl substituents to increase delocalization.

Introducing a phenyl

substituent, specifically p-tolyl, to terpy in Ru(II) complexes has been shown to stabilize
the 3MLCT by 1100 cm-1 relative to the 3MLCT of the unsubstituted analog.7 As for
achieving the destabilization of the ligand field states, the introduction of a tolyl substituent
could lead to an inductive increase in the basicity of the polypyridine, increasing the donor strength which may have the effect of increasing the ligand field splitting and
destabilizing the ligand field excited states.
We set out to prepare a series of molecules by introducing phenyl substituents to
the 2,6-di(2-carboxypyridyl)pyridine (dcpp) ligand. The first molecule in the series has
127

been successfully prepared and will be described. Additional design strategies to achieve
the excited state inversion will also be proposed.
4.2 Experimental
4.2.1 Synthesis
General. All chemicals were of reagent grade, purchased from Sigma-Aldrich,
Alfa Aesar, Acros Organics, Strem Chemicals, or TCI Chemical and used as received
unless otherwise noted. Solvents were purchased from Sigma-Aldrich, Jade Scientific,
Spectrum, Mallinckrodt, EMD Chemical, or CCI and were purified using standard
purification techniques.8

All air-sensitive reactions were carried out under inert

atmosphere by standard Schlenk techniques utilizing thoroughly deoxygenated solvents
that were degassed by the freeze—pump—thaw method. 1H NMR and 13C NMR were
recorded with Varian UnityPlus-300 MHz, Varian UnityPlus-500 MHz, and Agilent DDR2
500 MHz spectrometers. Ground state absorption spectra were obtained on a Varian Cary
50 spectrophotometer.

IR spectra were obtained on a Mattson Galaxy 5000 FTIR.

Elemental analysis was obtained through the Analytical facilities at Michigan State
University. Electrospray mass spectra (ESI-MS) were obtained from the staff of the MSU
Mass Spectrometry Facility. The characterization data of previously unknown compounds
are included in the appendix at the end of this chapter.
p-Tolylboronic acid. The compound was prepared according to the literature
procedure from p-bromotoluene (7.19 g, 42.0 mmol) and purified by recrystallization from
dissolution in hot toluene and adding hexanes and allowing it to slowly cool.9 Yield: 5.71
g (57%). 1H NMR (500MHz, CDCl3) δ (ppm): 8.14 (d, 2H), 7.33 (d, 2H), 2,46 (s, 3H).
128

4-Hydroxypyridine-2,6-dicarboxylate diethyl ester.

The compound was

prepared from chelidamic acid (9.80 g, 3.98 mmol) according to a patent procedure.10 The
product was obtained as a tacky white sold which is recrystallized by dissolving in hot
ethanol and adding hexanes till cloudy and placing in freezer. Yield: 7.07 g (61%). 1H
NMR (500MHz, (CD3)2CO) δ (ppm): 7.59 (s, 2H), 4.39 (q, 4H), 1.37 (t, 6H).
4-(Trifluoromethylsulfonyloxy)pyridine-2,6-dicarboxylate diethyl ester. The
compound was prepared by modification of previously published procedures.10,11 A 250
mL air-free round bottom flask equipped with an addition funnel was charged with 4hydroxypyridine-2,6-dicarboxylate diethyl ester (4.14 g, 17.3 mmol) and dry CH2Cl2 (40
mL) and cooled to -78oC in a dry ice/acetone bath. Once cool, dry pyridine was slowly
added via syringe (2.5 mL, 2.45 g, 30.9 mmol). To the addition funnel was added CH2Cl2
(25 mL) and trifluoromethanesulfonic anhydride (3.71 mL, 6.22 g, 22.1 mmol), which is
added dropwise to the reaction mixture over 15 minutes. The reaction was allowed to warm
to room temperature over 2 hours and water was added to quench the reaction. The layers
were separated and the water layer extracted with EtOAc (3 x 150 mL). The organic layers
were combined, dried with MgSO4, filtered and evaporated. The product may be used
without purification or may be purified by recrystallization from 3:1 EtOAc/Hexanes.
Yield: 6.06 g (94%). 1H NMR (500MHz, (CDCl3) δ (ppm): 8.17 (s, 2H), 4.52 (q, 4H),
1.47 (t, 6H).
4-p-Tolylpyridine-2,6-dicarboxylate diethyl ester. The compound was prepared
by modification of previously published procedures.10,12 To a 250 mL round bottom flask
was added 4-(trifluoromethylsulfonyloxy)pyridine-2,6-dicarboxylate diethyl ester (1.54 g,
4.14 mmol), p-tolylboronic acid (0.648 g, 4.77 mmol), Pd(PPh3)4 (0.380 g, 0.329 mmol, 8
129

mol%), N,N-diisopropylethylamine (DIPEA, ‘Hünig’s Base’) (1.6 mL, 9.19 mmol) and dry
dimethylformamide (~60 mL) and the reaction was stirred at 90-100oC overnight in the
dark. The solvent was removed by vacuum distillation. The remaining brown/green solid
was taken up in EtOAc (100 mL) and washed with H2O (2 x 100 mL) and once with
saturated NaCl (aq), dried with MgSO4, filtered and evaporated. The crude reaction
product may be purified by column chromatography on silica gel with 2% EtOAc in
CH2Cl2. The product glows on silica under longwave UV so the column may be monitored
by UV lamp. The unreacted triflate and boronic acid elute first followed by the product.
The product may be recrystallized by dissolving in ethanol and adding hexanes and cooling
to 0oC. Yield: 0.565 g (44%). 1H NMR ((CDCl3, 500 MHz): δ 8.51 (s, 2H), 7.68 (d, 2H,
J = 8.0 Hz), 7.36 (d, 2H, J = 8.0 Hz), 4.53 (q, 4H, J = 7.1 Hz), 2.45 (s, 3H), 1.49 (t, 6H, J
= 7.1 Hz). 13C NMR ((CDCl3, 500 MHz): δ 194.57, 164.94, 158.04, 150.86, 149.16,
130.11, 126.99, 125.23, 62.43, 21.32, 14.26. TOF-MS [ESI, m/z (rel int)]: 314.1 (100)
[C18H19NO4]H+, 336.1 (20) [C18H19NO4]Na+, 649.3 (15) 2[C18H19NO4]Na+. IR (KBr,
cm-1): 2976.7 w, 1717.0 s, 1600.7 m, 1377.9 m, 1345.7 s, 1257.4 s, 1140.5 m, 1069.7 m,
1023.5 m, 909.5 w, 830.3 m, 781.0 m, 745.6 w.
2,6-Di(2-carboxypyridyl)-4-p-tolylpyridine

(dcptp).

The

compound

was

prepared analogously to dcpp by modification of the published procedure.13 To a 250 mL
round bottom air-free flask was added 2-bromopyridine (0.563 g, 3.57 mmol) and dry THF
(~50 mL) via cannula. Under nitrogen, the solution was cooled to -78oC in a dry
ice/acetone bath. 1.6 M n-BuLi solution in hexanes (2.2 mL, 3.52 mmol) was added
dropwise over 15 minutes and the solution stirred for an additional 30 minutes. To a

130

separate air free flask was added 4-p-tolylpyridine-2,6-dicarboxylate diethyl ester (0.505
g, 1.61 mmol) and dry THF (~20 mL) and the solution transferred dropwise via 22 gauge
cannula to the reaction mixture over 30 minutes. The solution was stirred for an additional
30 minutes at -78oC before quenching with methanol (4 mL) and allowing to warm to room
temperature and stir overnight. To promote the separation of layers, hexanes (10 mL) was
added followed by the addition of 10% HCl (20 mL) and the organic layer removed. The
acidic aqueous layer was washed with CH2Cl2 (1 x 25 mL) and then basified with 5 M
NaOH. The basic aqueous layer was then extracted with CH2Cl2 (3 x 50 mL). The organic
layer dried with MgSO4, filtered, and evaporated. The product was purified by column
chromatography on silica with 2% EtOH in CHCl3 and recrystallized by dissolution in
minimal CH2Cl2 at room temperature and adding equal parts EtOAc and hexanes and
allowing to stand. Yield 0.292 g (48%). mp 173-174oC. 1H NMR ((CD3)2CO, 500 MHz):
δ 8.72 (ddd, 2H, J = 4.7, 1.5, 1.0 Hz, 6py-a (pyridyl arm)), 8.47 (s, 2H, 3/5py-b (bridging
pyridyl)), 8.13 (dt, 2H, J = 7.8, 1.0 Hz, 3py-a), 7.99 (td, 2H, J = 7.8, 1.7 Hz, 4py-a), 7.90
(d, 2H, J = 8.0, 1.7 Hz, Ph), 7.61 (dtd, 2H, J = 7.8, 4.7, 1.0 Hz, 5py-a), 7.44 (d, 2H, J = 8.0,
Ph), 2.45 (s, 3H). 13C NMR ((CDCl3, 500 MHz): δ 192.04, 153.89, 153.69, 150.54,
149.25, 140.22, 136.37, 133.08, 130.09, 127.11, 126.35, 126.19, 124.83, 21.33. TOF-MS
[ESI, m/z (rel int)]: 380.1 (100) [C24H17N3O2]H+, 402.1 (5) [C24H17N3O2]Na+, 781.3 (5)
2[C24H17N3O2]Na+. IR (KBr, cm-1): 3057.6 w, 1684.6 s, 1590.8 m, 1354.7 s, 1302.7 m,
1237.2 m, 1151.3 w, 1095.7 w, 1048.9 w, 1019.3 m, 991.1 m, 958.9 m, 903.8 w, 821.2 m,
764.7 m, 683.0 m. Elemental Analysis for C24H17N3O2, Calculated: C, 75.98; H, 4.52;
N, 11.07. Found: C, 75.22; H, 4.36; N, 11.00.
131

[Fe(dcptp)2](PF6)2.

To an air-free flask dcptp (0.115 g, 0.302 mmol) and

deoxygenated EtOH (50 mL) via cannula were added. The solution was warmed (50°C)
under nitrogen to promote dissolution.

A separate air-free flask was charged with

(NH4)2Fe(SO4)2·6H2O (0.0608 g, 0.155 mmol). The ligand solution was transferred via
cannula to the flask containing the iron(II) salt. The reaction was warmed to 70oC and
allowed to stir for 18 hours. After ~2 hours of warming the solution began to darken to an
olive green color and gradually darkened to intense blue and allowed to heat overnight. A
separate air-free flask was charged with 8 equivalents of NH4PF6 (0.202 g, 1.24 mmol)
and H2O (4 mL) which was then added via syringe to the reaction mixture. The solution
was concentrated by evaporation under a stream of N2 and water was added yielding dark
blue precipitate which was filtered and rinsed with H2O. The crude product was dissolved
in acetonitrile and the solution was washed with hexanes to remove grease. The acetonitrile
solution was concentrated then purified by passage through a basic alumina column.
Yield: 0.0381 g (23%). 1H NMR ((CD3)2CO, 500 MHz): δ (ppm) 8.75 (s, 4H, 3/5py-b),
8.37 (t, 4H, J = 7.8, 1.1 Hz, 4py-a), 8.20 (d, 4H, J = 7.8, 1.0 Hz, 3py-a), 8.16 (d, 4H, J =
5.8, 1.0 Hz, 5py-a), 8.06 (d, 4H, J = 8.3 Hz, Ph), 7.50 (t, 4H, J = 7.8, 5.8, 1.0 Hz, 5py-a),
7.49 (d, 4H, J = 8.3 Hz, Ph), 2.46 (s, 6H). 13C NMR ((CD3)2CO, 500 MHz): δ (ppm)
180.91, 160.94, 159.41, 159.31, 142.32, 141.89, 140.92, 131.06, 130.49, 129.01, 128.00,
127.48, 127.27, 20.48. TOF-MS [ESI, m/z (rel int)]: 407.0 (15) [C48H34N6O4Fe]2+, 959.1
(5) {[C48H34N6O4Fe](PF6)}+. UV-Vis (CH3CN) λ(ε(M-1cm-1)): 339 nm (21600), 381
nm (15500), 445 nm (2400), 515 nm (4000), 626 nm (8400).

132

4.2.2 Physical Measurements
Cyclic Voltammetry. Electrochemical measurements were carried out using a CH
Instruments CH620D electrochemical analyzer to determine the E1/2 for ligand reductions
and metal oxidations of each complex. Solutions of the compounds were prepared in
distilled CH3CN containing NBu4PF6 (ca. 0.1 M) as the supporting electrolyte.

A

standard three-electrode setup was used with a platinum working electrode, carbon rod
counter electrode, and a Ag/AgCl electrode as the reference. All measurements were made
inside an inert atmosphere glovebox. Data was acquired at a scan rate of 50 mV s-1. After
data collection, ferrocene was added to subsequent scans for reference.
4.3 Results and Discussion
4.3.1 Design of the π-Extended Ligand System
The goal of the present investigation is to achieve stabilization of a charge transfer
state and destabilization of the ligand field excited states of the symmetric iron(II)
polypyridyl complex [Fe(dcpp)2]2+. Inspired by the work of Damrauer and coworkers
(vide supra), we set out to introduce p-tolyl substituents to the pyridyl rings of 2,6-di(2carboxypyridyl)pyridine (dcpp).7
It is well known that extending the conjugation of polypyridyl ligand can
effectively lower the energy of the charge transfer state. A straightforward example of this
of this is evident in the comparison of bpy and terpy. Delocalization over three ring in
terpy compared to just two rings in bpy leads to a lower energy charge transfer. Introducing
phenyl substituents allows for further delocalization and effectively lowers the LUMO or
π* orbitals of the ligand.

The choice of the p-tolyl substituent as opposed to an
133

unsubstituted phenyl is two-fold. First, the electron donating methyl group on the phenyl
will make the functional group slightly electron donating which may have the effect of
increasing the -donor strength of the ligand, possibly destabilizing ligand field excited
states. Second, the methyl group will allow for further functionalization, i.e. installation
of an anchoring group which is essential for application as a sensitizer in DSSCs.
It has been shown that a single substituent on the central ring of terpy can have a
significant impact on the properties, therefore we chose to first introduce a single
substituent to the central ring of dcpp.4,14 Further synthetic elaboration will be pursued in
due course with the installation of substituents to the terminal pyridyl rings. Since the dcpp
ligand system differs from other polypyridyls in that it is cross-conjugated, it will be
interesting to see the effects of extending the conjugation of a single ring versus all rings.
If the p-tolyl substituent on the central pyridyl ring can produce the desired effect of
stabilization of the charge transfer state, can it then be further stabilized by the introduction
of p-tolyl groups on the terminal rings? Or do the presence of the cross-conjugated bridges
inhibit any cumulative stabilization?
As mentioned above, an added benefit of the tolyl group is the possible
enhancement of the -donor strength and destabilization of the ligand field states. We
propose another analog by which we can target this separately, by introducing electron
donating alkyl groups to the 5-position of the terminal pyridyl rings. If we could achieve
stabilization of the MLCT through the p-tolyl substituent on the central ring and
destabilization of the ligand field states through enhanced -donation from the terminal
rings, this would be ideal.

134

Taking a step back, and looking at the big picture, the preparation of heteroleptic
Ru(II) complexes is well known and provides much opportunity to tune the photophysical
properties. For instance, heteroleptic complexes carrying an electron withdrawing group
and an electron donating group on the two separate ligands always show lower emission
energies than the parent homoleptic complexes because the * orbital of the accepting
ligand is stabilized while the donating ligand destabilizes the metal centered t2g orbitals.4
Iron(II) complexes are much more labile, therefore heteroleptic compounds are extremely
difficult to obtain. Using the cross-conjugation to our advantage, if we can in fact achieve
the desired result of tuning the excited states by inducing different effects from varying the
substituents on the terminal rings versus the central ring, we are essentially working with
a heteroleptic-like iron(II) complex without the synthetic complications. We are just
beginning to investigate the possibilities with this ligand system and only time will tell, but
on top of the unique properties of [Fe(dcpp)2]2+ that were discussed Chapter 2, the
synthetic possibilities with this type of ligand could prove to be very exciting.
4.3.2 Synthesis and Coordination of 2,6-Di(2-carboxypyridyl)-4-p-tolylpyridine
The transformation of the bridging groups of 2,6-di(2-carboxypyridyl)pyridine
(dcpp), described in the previous chapter, was easily achieved from a one-step reaction
directly from dcpp. The installation of functional groups to the pyridyl rings is not as
straightforward, but from a retrosynthetic analysis of the preparation of dcpp it becomes
clear that we may introduce functional groups to the terminal rings and/or the central ring
independently prior to the final reaction step (Scheme 4-1). The accessibility of various
analogs through this route makes it highly desirable.

135

Scheme 4-1. Retrosynthetic analysis of substituted 2,6-di(2-carboxypyridyl)pyridine.
All of the desired analogs in this investigation possess the p-tolyl substituent on the
central ring, therefore we pursued this starting material first. 4-p-Tolylpyridine-2,6dicarboxylate diethyl ester was not previously described in the literature, but similar
structures exist and based on literature methods the target compound can be readily attained
through Suzuki coupling as shown in Scheme 4-2.10,12,15–17

Scheme 4-2. Preparation of 4-p-tolylpyridine-2,6-dicarboxylate diethyl ester.

136

The first step of the synthesis required the esterification of chelidamic acid. This
reagent is commercially available and was initially purchased at over $40/gram. The
reagent was used sparingly, due to the high expense, and the ester was prepared via the
standard sulfuric acid catalyzed Fischer esterification in ethanol. At this small scale the
yield of this reaction was very poor so the alternate esterification through the acyl chloride
was attempted, resulting in better yield. To have any hope of obtaining the final target
molecule, the high expense, scale, and yield of the first reaction left much to be desired and
this whole reaction scheme was almost abandoned. Luckily, TCI Chemical sells the same
reagent at higher purity for $6/gram, and at larger scale the yield was much improved.
Since organotriflates may be readily obtained from phenols, and the boronic acid
reagent is easily prepared from 4-bromotoluene, the Suzuki coupling route was an obvious
choice. The coupling reaction requires extremely dry conditions and despite my best
efforts to exclude water, the reaction suffered from lower than expected yields, achieving
only 44%. The reaction was undertaken during humid spring and summer months and
since most of the reagents and solvent are hygroscopic the reaction seemed to be adversely
affected. With the central pyridyl ring in hand, the first ligand for this study was prepared
analogously to dcpp in 48% yield, as show in Scheme 4-3.

137

Scheme 4-3. Preparation of 2,6-di(2-carboxypyridyl)-4-p-tolylpyridine (dcptp).
The reaction has only been completed once, therefore the 48% yield is acceptable
for a first attempt. It is truly a shame that we have only just begun to embark on this
investigation, because there is no doubt that, with just a little more time, the two other
desired analogs would be in hand (Figure 4-1). The procedure would be the same with the
only difference being the 2-bromopyridine starting material. 2-bromo-5-methylpyridine is
commercially available, while the 2-bromo-4-p-tolylpyridine may be prepared by a simple
palladium catalyzed coupling reaction of 2-bromo-4-iodopyridine with a Grignard reagent
of toluene.18

138

Figure 4-1. Structures of desired analogs of 2,6-di(2-carboxypyridyl)-4-p-tolylpyridine.
With dcptp we may begin to look at the effect of extending the -system of the dcpp
ligand. [Fe(dcptp)2]2+ has been successfully prepared, albeit by a slightly different
procedure that the other iron(II) complexes described in this dissertation. The standard
procedure of adding an iron(II) salt to a slight excess of ligand in methanol and water does
not work for this system. The p-tolyl substituent leads to solubility issues in this solvent
mixture and despite longer reaction times and prolonged heating, there was no color change
or other evidence of a reaction taking place and only unreacted ligand was recovered from
the reaction. Instead the product was obtained by reaction in ethanol by heating to just
under the boiling point for 24 hours. Again, unreacted ligand was recovered from the
reaction, but the product was present in 23% yield, shown in Scheme 4-4. Longer reaction
times or bringing to reflux may prove to increase the yield, but to date the reaction has yet
to be optimized.

139

Scheme 4-4. Preparation of [Fe(dcptp)2](PF6)2.

4.3.3 Preliminary Investigation into the Effects of Extending the -System
With the successful preparation of [Fe(dcptp)2]2+ we can begin to investigate the
effect of extending the -system. Since only one analog has been obtained we can only
compare the properties to [Fe(dcpp)2]2+ leaving many questions unanswered, however it
is a good starting point. From the optical absorption spectrum, shown in Figure 4-2, we
can see that the introduction of a p-tolyl group to the 4-position of the central pyridyl ring
leads to a red-shifted absorption maximum and an increase in the MLCT oscillator strength,
consistent with what is known for p-tolyl substituents on Ru(II) terpy chromophores.4,7

140

M o l a r A b s o r p t i v i t y (M -1c m -1)

25000

[Fe(dcptp)2]2+
[Fe(dcpp)2]2+

20000
15000
10000
5000
0

400

600

800

Wavelength (nm)
Figure 4-2. Ground state electronic absorption spectrum of [Fe(dcpp)2](PF6)2 (blue
dashed) and [Fe(dcptp)2](PF6)2 (red solid) in acetonitrile.
The electrochemical potentials of [Fe(dcptp)2]2+ were measured by cyclic
voltammetry and as with [Fe(dcpp)2]2+ there are four distinct ligand reductions attributed
to the presence of the four electron withdrawing carbonyl bridges. The oxidation and first
reduction potentials are listed in Table 4-1, along with the potentials of [Fe(dcpp)2]2+ for
comparison. The p-tolyl substituent induces a negative 95 mV shift in the oxidation
potential signifying destabilization of the metal t2g orbitals. The first reduction potential
is essentially unchanged as the negative 10 mV shift of the first reduction potential is within

141

error of the experiment, but could indicate a slight destabilization of the LUMO π* level
of the ligand.
Table 4-1. Electrochemical potentials for the oxidation and first reduction of
[Fe(dcpp)2](PF6)2 and [Fe(dcptp)2](PF6)2, measured with Ag/AgCl reference with 0.1 M
TBAPF6 electrolyte in acetonitrile, externally referenced to ferrocene.
Compound
[Fe(dcpp)2]2+
[Fe(dcptp)2]2+

E1 / 2 [ox]
1.295 V
1.200 V

E1 / 2 [red]
-0.965 V
-0.975 V

The electrochemistry is consistent with the work by Maestri et al. who found that
electron donating substituents on Ru(II)terpy complexes destabilize both the HOMO and
LUMO, and the HOMO t2g levels are destabilized more than the LUMO * levels.4 We
tend to think of p-tolyl substituent as an electron acceptor because of the lower energy
MLCT that arises. However, the electron-rich ring acts as a donating group as it leads to
an increase in the basicity of the nitrogens which makes the removal of one electron from
the metal t2g orbital easier, as the resulting Fe(III) can interact more strongly with the
donating ligands.4,7 The overall effect is a smaller HOMO-LUMO gap and lower energy
MLCT.
From this data it appears that we are on the right track as the introduction of single
tolyl substituent to the central ring of the ligands has led to a lower energy MLCT and
slight destabilization of the metal t2g orbitals. It will be interesting to see the effect of
introducing the additional substituents to the ligands that were shown in Figure 4-1.

142

4.3.4 Alternative Strategies to Tune the Energetics of Excited States
As mentioned above, one of the strategies to lower the energy of the MLCT state
is through the introduction of electron withdrawing groups to the periphery of the ligands.
Initially, there was hesitation along this front as too many electron withdrawing
substituents may lower the basicity of the nitrogens and cause difficulty with the formation
of the complex. However, in our efforts toward the overall goal of developing a viable
sensitizer, we set out to install substituents through which we may achieve attachment to
metal oxide substrates. As will be discussed in the next chapter, we have successfully
installed electron withdrawing carboxylate groups to the terminal rings of the dcpp ligand
and prepared the iron(II) complexes, however there are complications with characterization
of these analogs and we have yet to determine the effect of these substituents. Though we
cannot use these specific analogs in this present investigation, the apprehension toward this
strategy is lessened and this may be a feasible approach to lower the MLCT state.
Another approach that is gaining a lot of attention in recent years for Ru(II) dyes is
cyclometallated complexes in which a pyridyl ring is replaced with a phenyl ring.
Substituting a nitrogen with an anionic carbon, or strong -donor, results in a considerable
decrease in the 3MLCT energy.2 The metal based HOMO is significantly destabilized
while the LUMO * is destabilized by a smaller extent resulting in the lower MLCT. These
cyclometallating ligands allow for manipulation of both the ground and excited states
through installation of substituents on the phenyl ring.19 The Ru(II) cyclometallated
complexes are all heteroleptic in which only one pyridyl group has been replaced with a

143

phenyl, therefore, much difficulty would likely arise in introducing this type of ligand to
an iron(II) center, as iron(II) complexes are much more labile.
A more synthetically accessible strategy to introduce an anionic carbon would be
through the use of N-heterocyclic carbenes, in which the chemistry with iron(II) is
known.20 An exciting new report was recently published in which an iron(II) N-hetercyclic
carbene complex achieves a prolonged 3MLCT lifetime.21 The anionic carbene ligands
show greater-donor character compared to terpy, significantly destabilizing the 5T2
ligand field state, however, the LUMO is also destabilized resulting in higher energy charge
transfer states. If we may combine the strategies mentioned above to stabilize the charge
transfer state with the enhanced -donation from an anionic carbene ligand to destabilize
the ligand field state we may be able to achieve inversion of the excited states of an iron(II)
chromophore.
4.4 Concluding Comments
Efforts to modulate the energetics of the excited states of a symmetric iron(II)
complex through synthetic means has been initiated with the preparation of an analog of
[Fe(dcpp)2]2+ possessing an extended -system.

The 2,6-di(2-carboxypyridyl)-4-p-

tolylpyridine (dcptp) ligand was readily obtained through adaptation from known methods
and achieved through an analogous procedure for the preparation of the unsubstituted dcpp.
The coordination chemistry of dcptp by our standard procedure was complicated by the
solubility characteristics of the ligand, however, [Fe(dcptp)2]2+ was successfully obtained
by implementing more rigorous conditions. The increased delocalization that is achieved
through extending the conjugation of the central ring of dcpp has resulted in stabilization

144

of the MLCT and destabilization of the metal centered t2g orbitals as is evident in the
optical absorption and electrochemical data. This investigation is in its infancy and
additional characterization is necessary to determine the impact of these changes on the
excited state dynamics.

Further synthetic elaboration, involving the introduction of

substituents to the terminal rings of dcpp has been proposed, and should provide us more
information on the viability of the current strategy to effectively tune the excited state
energetics of iron(II) chromophores.

145

APPENDIX

146

APPENDIX

Figure 4-3. 1H NMR of 4-p-tolylpyridine-2,6-dicarboxylate diethyl ester in CDCl3.

147

Figure 4-4. 13C NMR of 4-p-tolylpyridine-2,6-dicarboxylate diethyl ester in CDCl3.

148

100

TOF MS ES+
314.1

%

336.1
649.3
286.1
0

200

852.2
400

600
m/z

Figure 4-5. ESI-MS of 4-p-tolylpyridine-2,6-dicarboxylate diethyl ester.

149

800

%
T
80
r
a
n 70
s
m 60
i
t
50
t
a
n 40
c
e
4000

2976.7
781.0
1023.5

1600.7
1377.9

1140.5

1345.7
1257.4
1717.0
3500

3000

2500

2000

1500

Wavenumbers
Figure 4-6. IR of 4-p-tolylpyridine-2,6-dicarboxylate diethyl ester.

150

1000

Figure 4-7. 1H NMR of 2,6-di(2-carboxypyridyl)-4-p-tolylpyridine (dcptp) in (CD3)2CO.

151

Figure 4-8. 13C NMR of 2,6-di(2-carboxypyridyl)-4-p-tolylpyridine (dcptp) in CDCl3.

152

100

TOF MS ES+
380.1

%

781.3

402.1

0
350

450

550

650

m/z
Figure 4-9. ESI-MS of 2,6-di(2-carboxypyridyl)-4-p-tolylpyridine (dcptp).

153

750

%
T 60
r
a
n 50
s
m
i 40
t
t
30
a
n
c
20
e
4000

3057.6
1428.4

1237.2

991.1

1590.8

1684.6

3500

3000

2500

2000

1354.7

1500

Wavenumbers
Figure 4-10. IR of 2,6-di(2-carboxypyridyl)-4-p-tolylpyridine (dcptp).

154

821.2
764.7

1000

Figure 4-11. 1H NMR of [Fe(dcptp)2](PF6)2 in (CD3)2CO.

155

Figure 4-12. 13C NMR of [Fe(dcptp)2](PF6)2 in (CD3)2CO.

156

100
380.1

TOF MS ES+

│◄x30

%
781.3

618.7

833.2

596.2

974.8

407.1
0

300

400

500

600

700
m/z

Figure 4-13. ESI-MS of [Fe(dcptp)2](PF6)2.

157

800

900

1000

1100

1200

REFERENCES

158

REFERENCES

(1)

Juris, A.; Balzani, V.; Barigelletti, F.; Campagna, S.; Belser, P.; Von Zelewsky, A.
Coordination Chemistry Reviews 1988, 84, 85–277.

(2)

Hammarström, L.; Johansson, O. Coordination Chemistry Reviews 2010, 254,
2546–2559.

(3)

Abrahamsson, M.; Lundqvist, M. J.; Wolpher, H.; Johansson, O.; Eriksson, L.;
Bergquist, J.; Rasmussen, T.; Becker, H.-C.; Hammarström, L.; Norrby, P.-O.;
Akermark, B.; Persson, P. Inorganic Chemistry 2008, 47, 3540–8.

(4)

Maestri, M.; Armaroli, N.; Balzani, V.; Constable, E. C.; Thompson, A. M. W. C.
Inorganic Chemistry 1995, 34, 2759–2767.

(5)

Damrauer, N. H.; Boussie, T. R.; Devenney, M.; McCusker, J. K. Journal of the
American Chemical Society 1997, 119, 8253–8268.

(6)

Constable, E. C.; Housecroft, C. E.; Thompson, A. C.; Passaniti, P.; Silvi, S.;
Maestri, M.; Credi, A. Inorganica Chimica Acta 2007, 360, 1102–1110.

(7)

Hewitt, J. T.; Vallett, P. J.; Damrauer, N. H. The Journal of Physical Chemistry A
2012, 116, 11536–47.

(8)

Perrin, D. D.; Armarego, W. L. F. Purification of Laboratory Chemicals, 3rd ed.;
Pergamon Press: New York, 1988.

(9)

White, J. R.; Price, G. J.; Schiffers, S.; Raithby, P. R.; Plucinski, P. K.; Frost, C. G.
Tetrahedron Letters 2010, 51, 3913–3917.

(10)

Belanger, D. B.; Siddiqui, M. A.; Curran, P. J.; Hamann, B.; Zhao, L.; Reddy, P.
A. P.; Tadikonda, P. K.; Shipps Jr., G. W.; Mansoor, U. F. Novel JNK Inhibitors.
WO 2008/082487 A2, 2008.

(11)

Wu, W.; Kong, H.-K.; Li, H.; Ho, Y.-M.; Gao, Y.; Hao, J.; Murphy, M. B.; Lam,
M. H.-W.; Wong, K.-L.; Lee, C.-S. European Journal of Organic Chemistry 2011,
2011, 5054–5060.

(12)

Pouessel, J.; Abada, S.; Le Bris, N.; Elhabiri, M.; Charbonnière, L. J.; Tripier, R.
Dalton transactions (Cambridge, England : 2003) 2013, 42, 4859–72.

(13)

Goldsmith, C. R.; Stack, T. D. P. Inorganic Chemistry 2006, 45, 6048–55.

(14)

Chambers, J.; Eaves, B.; Parker, D.; Claxton, R.; Ray, P. S.; Slattery, S. J.
Inorganica Chimica Acta 2006, 359, 2400–2406.
159

(15)

Xu, M.; Chen, C. Z.; Wan, P. Journal of Photochemistry and Photobiology A:
Chemistry 2008, 198, 26–33.

(16)

Ghoshal, A.; Sarkar, A.; Manickam, G.; Kumaran, R.; Jayashankaran, J. Synlett
2010, 2010, 1459–1462.

(17)

Ohe, T.; Miyaura, N.; Suzuki, A. The Journal of Organic Chemistry 1993, 58,
2201–2208.

(18)

LI, X.-H.; LIU, Z.-Q.; LI, F.-Y.; DUAN, X.-F.; HUANG, C.-H. Chinese Journal
of Chemistry 2007, 25, 186–189.

(19)

Bomben, P. G.; Robson, K. C. D.; Koivisto, B. D.; Berlinguette, C. P.
Coordination Chemistry Reviews 2012, 256, 1438–1450.

(20)

Hashimoto, T.; Urban, S.; Hoshino, R.; Ohki, Y.; Tatsumi, K.; Glorius, F.
Organometallics 2012, 31, 4474–4479.

(21)

Liu, Y.; Harlang, T.; Canton, S. E.; Chábera, P.; Suárez-Alcántara, K.; Fleckhaus,
A.; Vithanage, D. a.; Göransson, E.; Corani, A.; Lomoth, R.; Sundström, V.;
Wärnmark, K. Chemical Communications 2013, 1, 6412–6414.

160

Chapter 5. Functionalization of an Iron(II) Chromophore for Attachment to Metal
Oxide Substrates

5.1 Introduction
The molecules presented in this dissertation have provided us with the opportunity
to investigate the photophysical properties of symmetric iron(II) polypridyls. Though we
have yet to achieve a prolonged charge transfer state lifetime, we are pursuing another
aspect of the research toward the overall goal of implementing an iron(II) chromophore
into a DSSC. The complexes presented thus far are not suitable for dye sensitization as
they do not possess the means to bond to the semiconductor surface. To achieve efficient
injection we must first install anchoring groups.
Dye sensitization may be achieved by physisorption or chemisorption, but early
work by Gratzel and coworkers showed the incident photon to current efficiency is
improved 30-fold from the introduction of anchoring groups providing the ability to bind
to the semiconductor.1 The most widely employed anchor is the carboxylate functional
group, as it has been shown to bind to TiO2 with good stability and provide strong
electronic communication.2,3 The relative ease of installation of carboxylic acid groups
make it a clear front runner for this investigation. However, research into alternative
anchoring groups has been a hot topic in recent years and alternate possibilities will be
pursued in due course.4–12
Inorganic chromophores, specifically d6 coordination compounds, are ideal
sensitizers as they can undergo a metal to ligand charge transfer (MLCT) upon absorption
of a photon creating a charge separated state. The presence of carboxylate groups on the
periphery of the ligand allow for attachment to the semiconductor. Carboxylate groups are
161

ideal anchors for sensitization of TiO2 as there is optimal overlap of the d-orbital of
titanium with the extended  system of the carboxylate group, as shown in Figure 5-1.
Also shown is the multiple binding modes that are possible with carboxylate groups.
Researchers have found that initial binding involves one carboxylate linkage followed by
stronger bidentate or bridging modes due the rotational freedom of the molecule allowing
for immediate capture of another carboxylate group by a neighboring Ti atom.3,13

Figure 5-1. Depiction of the three possible binding modes of the carboxylate group to the
titanium dioxide semiconductor, along with the orbital diagram for ester-type binding
mode (adapted from references 2 and 14).2,14
The bpy and terpy ligands that are commonly employed bear multiple carboxylate
groups and their planar structure allows for binding at multiple sites. Research has also
shown that higher conversion efficiencies are obtained when there are numerous anchoring
groups as in the “N3” and “black dye” that were described in Chapter 1. The 2,6-(2162

carboxypyridyl)pyridine (dcpp) ligand, on the other hand, is not planar since upon
coordination to the metal center it takes on a helical arrangement. Though it is synthetically
feasible to introduce carboxylate groups to all three rings of dcpp, the helical conformation
of the dye molecule may hinder binding through multiple sites. If we functionalize the
central ring, binding is possible through only one site, on the other hand if we functionalize
the terminal rings, binding may be possible through both ligands to a TiO2 nanoparticle.
To determine if the functionalization of the terminal rings of dcpp would be sterically
feasible a geometry optimization was performed.

Figure 5-2. Optimized structure of bis(2,6-di(2-carboxy(4-pyridylcarboxylicacid))pyridine)iron(II), [Fe(dcpap)2](PF6)2, shown in two orientations for clarity.
The optimized structure, shown in Figure 5-2, shows that the carboxylate groups
should not interfere with the carbonyl bridge lone pair- interaction or the coordination
environment of the dcpp ligand to the iron(II) center, therefore we set out to prepare an
analog of dcpp that would allow for anchoring to a semiconductor surface. The carboxylic
acid functionalized ligand and corresponding iron(II) complex have been successfully
prepared and will be discussed. Carboxylic acids are known to be insoluble in most
163

solvents, therefore the ethyl ester analog of the ligand and complex have also been
prepared. Future work toward the installation of alternative anchoring groups will also be
presented.
5.2 Experimental
5.2.1 Synthesis
General. All chemicals were of reagent grade, purchased from Sigma-Aldrich,
Alfa Aesar, Acros Organics, or Strem Chemicals and used as received unless otherwise
noted.

Solvents were purchased from Sigma-Aldrich, Jade Scientific, Spectrum,

Mallinckrodt, EMD Chemical, or CCI and were purified using standard purification
techniques.15 All air-sensitive reactions were carried out under inert atmosphere by
standard Schlenk techniques utilizing thoroughly deoxygenated solvents that were
degassed by the freeze—pump—thaw method. 1H NMR and 13C NMR were recorded with
Varian UnityPlus-300 MHz, Varian UnityPlus-500 MHz, and Agilent DDR2 500 MHz
spectrometers. Ground state absorption spectra were obtained on a Varian Cary 50
spectrophotometer. IR spectra were obtained on a Mattson Galaxy 5000 FTIR. Elemental
analysis was obtained through the Analytical facilities at Michigan State University.
Electrospray mass spectra (ESI-MS) were obtained from the staff of the MSU Mass
Spectrometry Facility. The characterization data of previously unknown compounds are
included in the appendix at the end of this chapter.
2,6-Di(2-carboxy(4-methylpyridyl))pyridine (dcmpp).

The compound was

prepared analogously to 2,6-di(2-carboxypyridyl)pyridine (dcpp) by modification of the
published procedure.16 To a 250 mL round bottom air-free flask was added 2-bromo-4-

164

methylpyridine (4.87 g, 28.3 mmol) and dry THF (~120 mL) via cannula. Under nitrogen,
the solution was cooled to -78oC in a dry ice/acetone bath. 1.6 M n-BuLi solution in
hexanes (17.7 mL, 28.3 mmol) was added dropwise over 20 minutes and the solution stirred
for an additional 30 minutes.

To a separate air-free flask was added diethyl 2,6-

pyridinedicarboxylate (3.00 g, 13.5 mmol) and dry THF (~20 mL) and the solution
transferred dropwise via 22 gauge cannula to the reaction mixture over 30 minutes. The
solution was stirred for an additional 30 minutes at -78oC before quenching with methanol
(30 mL) and allowed to warm to room temp and stir overnight. 10% HCl (60 mL) was
added and the organic layer removed. The acidic aqueous layer was washed with CH2Cl2
(1 x 100 mL) and then basified with 5 M NaOH. The basic aqueous layer was then
extracted with CH2Cl2 (3 x 100 mL). The organic layer dried with MgSO4, filtered, and
evaporated. The product was purified by column chromatography on silica with 2% EtOH
in CHCl3 and recrystallized by dissolution in EtOAc at room temp and adding equal parts
hexanes and cooling to 0oC. Yield 2.56 g (60%). mp 174-175oC. 1H NMR ((CD3)2CO,
500 MHz): δ 8.50 (d, 2H, J = 4.9 Hz, 6py-a (pyridyl arm)), 8.24 (t, 1H, J = 7.8 Hz, 4py-b
(bridging pyridyl)), 8.17 (d, 2H, J = 7.8 Hz, 3/5py-b), 7.87 (s, 2H, J = 0.7 Hz, 3py-a), 7.40
(d, 2H, J = 4.9, 0.7 Hz, 5py-a), 2.42 (s, 6H). 13C NMR ((CD3)2CO 500 MHz): δ 192.84,
154.41, 154.15, 148.76, 147.93, 137.42, 127.18, 126.20, 125.19, 20.04. TOF-MS [ESI,
m/z (rel int)]: 318.1 (100) [C19H15N3O2]H+, 340.1 (20) [C19H15N3O2]Na+, 657.2 (20)
2[C19H15N3O2]Na+. IR (KBr, cm-1): 3045.2 w, 1682.9 s, 1596.9 s, 1571.1 w, 1478.5 w,
1408.6 w, 1288.4 m, 1200.4 m, 1170.6 m, 1050.5 m, 994.5 w, 855.2 m, 773.4 m, 668.7 m.

165

Elemental Analysis for C19H15N3O2, Calculated: C, 71.91; H, 4.76; N, 13.24. Found: C,
71.42; H, 4.71; N, 13.08.
2,6-Di(2-carboxy(4-pyridylcarboxylicacid))pyridine (dcpap). The compound
was prepared by modification of a literature procedure.17 In a 100 mL beaker dcmpp (1.01
g, 3.17 mmol) was dissolved in conc. H2SO4 (50 mL) and cooled to 0oC in a NaCl ice bath.
CrO3 (1.89 g, 18.9 mmol) was finely ground and added in small portions over 1 hour. The
dark green solution was warmed to 70-80oC for 2 hours then allowed to stir at room
temperature overnight. The solution was poured over 300 mL of ice and the pH raised by
slow addition of 5 M NaOH (~100-150 mL) until the product precipitates (pH ~2-3).
Filtering the pasty white solid may be time consuming and can clog a fritted filter. To
overcome this issue the suspension was first allow settle and most of the supernatant
decanted off. Using a fine fritted filter the product was gravity filtered, once a substantial
layer of product covers the frit a vacuum was applied by aspirator and the product may
then be easily filtered without clogging the frit. The product was washed with water until
the rinse was no longer green. The product was recrystallized by suspending in H2O (~25
mL) and 1 M NaOH added until fully dissolved. 10% HCl was added to precipitate the
product which was filtered as before and dried in a vacuum dessicator. Yield: 1.04 g (87%).
1H NMR ((CD ) SO, 500 MHz): δ (ppm) 8.82 (d,
3 2

2H, J = 4.9, 0.7 Hz, 6py-a), 8.29 (t, 1H,

J = 6.8 Hz, 4py-b), 8.29 (s, 2H, 3py-a), 8.24 (d, 2H, J = 6.8, 1.6 Hz, 3/5py-b), 7.97 (dd,
2H, J = 4.9, 1.6 Hz, 5py-a). 13C NMR ((CD3)2SO, 500 MHz): δ (ppm) 192.20, 165.87,
154.77, 153.48, 150.63, 139.42, 138.82, 127.64, 125.99, 123.90. TOF-MS [ESI, m/z (rel
int)]: 376.1 (100) {[C19H11N3O6]-H+}-. IR (KBr, cm-1): 3444.0 br, 3090.4 w, 1712.6 s,

166

1686.7 s, 1606.0 w, 1465.2 m, 1327.7 s, 1295.7 s, 1266.3 s, 1237.7 s, 1186.5 m, 1087.3 w,
1013.4 m, 963.6 m, 885.7 w, 827.5 w, 760.8 m, 707.6 w, 672.2 m.
[Fe(dcpap)2](PF6)2.

An air-free flask was charged with dcpap (0.118 g, 0.313

mmol), H2O (20 mL), and 1 M NaOH (0.62 mL, 0.620 mmol). The solution was bubble
degassed with nitrogen for 20 minutes. A separate air-free flask was charged with of
(NH4)2Fe(SO4)2·6H2O (0.0587 g, 0.150 mmol) and H2O (8 mL). The Fe(II) solution was
transferred via cannula to the ligand solution which immediately turned brown then deep
purple. The reaction was allowed to stir for 1 hour at room temperature. 0.1 M HPF6 (12
mL, 1.20 mmol) was added and the solution cooled in the refrigerator (~2-8oC). The dark
purple precipitate was filtered and rinsed with H2O. The crude product was dissolved in
methanol and filtered to remove unreacted ligand. Diethyl ether was added to the filtrate
to precipitate the pure product. Yield: 0.0345 g (21%). 1H NMR (CD3OD, 500 MHz): δ
(ppm) 8.69 (t, 2H, J = 7.8 Hz, 4py-b), 8.52 (s, 4H, J = 1.6 Hz, 3py-a), 8.48 (d, 4H, J = 7.8,
Hz, 3/5py-b), 7.92 (d, 4H, J = 5.9 Hz, 6py-a), 7.80 (dd, 4H, J = 5.9, 1.6 Hz, 5py-a). 13C
NMR (CD3OD, 500 MHz): δ (ppm) 180.13, 161.86, 160.06, 158.09, 142.74, 141.93,
131.29, 127.31, 126.40, 126.38.

TOF-MS [ESI, m/z (rel int)]:

405.0 (55)

[C38H22N6O12Fe]2+, 955.0 (2) {[C38H22N6O12Fe](PF6)}+. IR (KBr, cm-1): 3427.1 br,
3091.8 w, 1725.1 s, 1696,1 s, 1611.5 w, 1383.9 w, 1297.3 w, 1237.0 m, 1084.1 w, 1015.1
w, 842.1 s, 765.6 m, 669.0 m. Elemental Analysis for C38H22N6O12FeP2F12·5H2O,
Calculated: C, 38.34; H, 2.71; N, 7.06. Found: C, 38.53; H, 2.70; N, 7.02.
2,6-Di(2-carboxy(4-pyridylcarboxylate ethyl ester))pyridine (dcpep). Dcpap
(1.02 g, 2.71 mmol) and dry EtOH (20 mL) were added to a 100 mL round bottom flask.

167

The suspension was cooled in an ice bath and SOCl2 (2 mL, 27.5 mmol) was added
dropwise and the reaction stirred at room temperature overnight under nitrogen. The
reaction mixture was vacuum distilled to remove solvent. The residue was dissolved in
CHCl3 and washed with 1 M NaOH then water. The organic layer was dried with Na2SO4,
filtered, and evaporated leaving a faint yellow oil. The oil may be purified by dissolving
in acetone/hexanes and cooling to 0oC.

The precipitated product was filtered but

immediately reverts to a colorless oil in the filter. Yield: 1.14 g (97%). 1H NMR (CDCl3,
500 MHz): δ (ppm) 8.83 (d, 2H, J = 4.9 Hz, 6py-a), 8.55 (s, 2H, 3py-a), 8.34 (d, 2H, J =
7.8 Hz, 3/5py-b), 8.14 (t, 1H, J = 7.8 Hz, 4py-b), 7.95 (dd, 2H, J = 4.9, 1.4 Hz, 5py-a), 4.41
(q, 4H, J = 7.1 Hz), 1.40 (t, 6H, J = 7.1 Hz). 13C NMR (CDCl3, 500 MHz): δ (ppm) 191.33,
164.34, 154.75, 153.13, 149.86, 138.38, 137.88, 127.45, 125.22, 124.50, 62.07, 14.16.
TOF-MS [ESI, m/z (rel int)]: 434.1 (100) [C23H19N3O6]H+.
[Fe(dcpep)2](PF6)2. To an air-free flask dcpep (0.311 g, 0.718 mmol) and
deoxygenated MeOH (30 mL via cannula) were added. The solution was gently warmed
(40°C) under nitrogen to promote dissolution of the oil. A separate air-free flask was
charged with (NH4)2Fe(SO4)2·6H2O (0.121 g, 0.310 mmol). The ligand solution was
transferred via cannula to the flask containing the iron(II) salt. The reaction was allowed
to stir for 18 hours at room temperature. The solution turned pale brown but never darkens.
A separate air-free flask was charged with NH4PF6 (0.375 g, 2.30 mmol) and MeOH (10
mL) which was transferred via cannula to the reaction mixture. The pale brown solution
was concentrated by evaporation under a stream of N2 resulting in a dark colored film
coating the flask. The crude oily product was taken up with water and filtered. The product

168

was dissolved in acetonitrile and purified by passage through a neutral alumina column.
The product was recrystallized by diethyl ether vapor diffusion into an acetonitrile solution
of the complex. Yield: 0.0230 g (6%).

1H

NMR ((CD3)2CO, 500 MHz): δ (ppm) 8.87

(t, 2H, J = 7.8 Hz, 4py-b), 8.60 (d, 4H, J = 7.8 Hz, 3/5py-b), 8.51 (s, 4H, J = 1.7 Hz, 3pya), 8.32 (d, 4H, J = 5.9 Hz, 6py-a), 7.87 (dd, 4H, J = 5.9, 1.7 Hz, 5py-a), 4.50 (q, 8H, J =
7.1 Hz), 1.40 (t, 12H, J = 7.1 Hz).
5.2.2 Physical Measurements
X-ray Structure Determination. Single-crystal X-ray diffraction data was
acquired at the X-ray facility of Michigan State University. X-ray quality single crystals
were obtained by slow diffusion of diethyl ether into an acetonitrile solution of the
compound. Data were collected as described in Chapter 2 of this dissertation.
Cyclic Voltammetry. Electrochemical measurements were carried out using a CH
Instruments CH620D electrochemical analyzer to determine the E1/2 for ligand reductions
and metal oxidations of each complex. Solutions of the compounds were prepared in
distilled CH3CN containing NBu4PF6 (ca. 0.1 M) as the supporting electrolyte.

A

standard three-electrode setup was used with a platinum working electrode, carbon rod
counter electrode, and a Ag/AgCl or Ag/AgNO3 electrode as the reference. All
measurements were made inside an inert atmosphere glovebox. Data was acquired at a
scan rate of 50 mV s-1. After data collection, ferrocene was added to subsequent scans for
reference.

169

5.2.3 Theoretical Calculations
Geometry Optimizations. The initial geometries of the molecules were generated
using SPARTAN or GaussView, or from crystal structure coordinates when available, and
subsequently optimized in two steps.18,19 The first optimization was performed using the
density functional B3LYP with the 3-21G** basis set, followed by a second optimization
using the B3LYP functional with the 6-31G* basis set. Frequency calculations at the
B3LYP/6-31G* level of theory were performed on the final optimized structures to ensure
that these geometries corresponded to global minima.
5.3 Results and Discussion
5.3.1 Installation of Anchoring Groups to 2,6-Di(2-carboxypyridyl)pyridine
Carboxylic acid groups may be readily introduced to aromatic methyl groups
through various means, therefore we first need to introduce methyl substituents. From the
retrosynthetic analysis of dcpp, that was described in the previous chapter, we may
independently introduce functional groups to the terminal rings and/or central ring of dcpp
prior to the final reaction step. Therefore, we may readily adapt the method by which we
had previously prepared dcpp and the p-tolyl analog to introduce methyl groups to the
rings. 2-Bromo-4-methylpyridine is commercially available, and quite expensive, but it
may be readily obtained in high yield via bromination of 2-amino-4-methylpyridine.20
With the starting materials readily accessible 2,6-di(2-carboxy(4-methylpyridyl))pyridine
(dcmpp) was conveniently prepared via the standard route shown in Scheme 5-1.

170

Scheme 5-1. Preparation of 2,6-di(2-carboxy(4-methylpyridyl))pyridine (dcmpp).
The standard method we use in our lab for the introduction of carboxylic acid
functional groups to bpy ligands is the chromic acid oxidation, or variation of Jones
oxidation.17 The reaction requires harsh conditions and toxic reagents, but is highly
efficient for the transformation of aromatic methyl groups. The method involves the
preparation of chromic acid in situ from chromium trioxide and concentrated sulfuric acid.
The route is convenient for bpy reagents as the carboxylic acid product is easily
precipitated from the reaction with water. We set out to prepare 2,6-di(2-carboxy(4pyridylcarboxylicacid))pyridine (dcpap) by the same method and found that the carboxylic
acid product from the reaction of dcmpp does not precipitate when water is added, and the
carbonyl bridges are likely hydrated under these conditions. Attempts were made to extract
the product into various solvents to no avail. I attempted to neutralize the reaction mixture
in an effort to make extraction easier, but as soon as the pH was raised slightly the product
readily precipitated from the reaction and was isolated in 87% yield (Scheme 5-2).

171

Scheme 5-2. Preparation of 2,6-di(2-carboxy(4-pyridylcarboxylicacid))pyridine (dcpap).
The diethyl ester of the ligand was prepared to allow for easier characterization. It
was first prepared by Fischer esterification in low yield. As with the chelidamic acid
described in Chapter 4, the yield was much improved by utilizing thionyl chloride in
ethanol and the product obtained in 97% yield, shown in Scheme 5-3. The workup proved
to be slightly difficult in that the product may not be obtained as a solid. Various attempts
were made to crystallize the product and it would precipitate in cold solution, but as soon
as it was filtered it reverts to oil, seemingly severely hygroscopic. It is funny that I set out
to make 2,6-di(2-carboxy(4-pyridylcarboxylate ethyl ester))pyridine (dcpep) to allow for
easier characterization and ultimately the carboxylic acid analog has caused no difficulty
while the ester has caused numerous issues during the subsequent coordination reaction
and characterization.

172

Scheme 5-3. Preparation of 2,6-di(2-carboxy(4-pyridylcarboxylate ethyl ester))pyridine
(dcpep).
5.3.2 Preparation of [Fe(dcpap)2]2+ and [Fe(dcpep)2]2+
All attempts to crystalize the dcpep ligand have failed, therefore the oil product was
dried in a vacuum dessicator and used as an oil in the coordination reaction. A variation
of the standard coordination reaction was employed in which water was excluded, shown
in Scheme 5-4. The dcpep oil was dissolved by gently warming in oxygen-free methanol
and transferred to the iron(II) salt. The reaction does not appear to proceed as there is no
darkening observed, even with prolonged stirring or heating. During the preparation of
[Fe(dcpp)2]2+ the addition of NH4PF6 was found to drive the reaction forward, so NH4PF6
was added in hopes of completing the reaction with the esterified ligand. Again no color
change was observed, but upon evaporation of the reaction mixture a dark oily film coated
the flask. The brown oily residue was passed through an alumina column with acetonitrile
and a small amount of blue/purple product was obtained in 6% yield.

173

Scheme 5-4. Preparation of [Fe(dcpep)2]2+.
A few potential problems during the preparation of [Fe(dcpep)2]2+ that may be
leading to the poor yield are as follows. Ferrous ammonium sulfate is acidic, which is
usually beneficial in that when it is dissolved in solution the acidic conditions prevent
immediate oxidation of iron(II) to iron(III). As an alternative FeCl2·2H2O was also
employed, but did not produce better results. Though dry methanol was used in the
reaction, the dcpep is hygroscopic and likely introduced water to the reaction and esters
can readily undergo hydrolysis under acidic conditions. The NH4PF6 is also acidic which
may be further detrimental to the ester functional groups. This could explain one aspect of
the low yield, but could not be the reason that the ligand does not appear to be coordinating
as is evident from the lack of color change. The ligand may not be coordinating well due
to the presence of the highly electron withdrawing substituents which may induce the
174

following; 1) lowering the basicity of the pyridyl nitrogens making them less willing to
donate the lone pair electrons to form the M-L bond, and 2) induce polarization making the
carbonyl bridge carbon even more electropositive and susceptible to attack by the solvent,
as was discussed in Chapter 2. Luckily, enough pure product was obtained to grow a few
crystals and obtain the x-ray crystal structure shown in Figure 5-3. The crystallographic
data is listed in table 5-1. The ethyl ester groups on the terminal pyridyl rings have not
prevented the [Fe(dcpep)2]2+ from achieving the octahedral geometry as is evident in the
selected geometric parameters listed in Tables 5-2 and 5-3.

Figure 5-3: ORTEP Drawing of the cation of [Fe(dcpep)2]2+ obtained from single-crystal
X-ray structure determination. Atoms are represented as 50% probability thermal
ellipsoids. Hydrogen atoms and anions are omitted for clarity.

175

Table 5-1. Crystallographic data for [Fe(dcpep)2](PF6)2.
[Fe(dcpep)2](PF6)2 a
Empirical formula
Formula weight
Temperature/K
Crystal system
Space group
a/Å
b/Å

C48H41F12FeN7O12P2 ρcalcmg/mm3
1253.67
173.15
Monoclinic
P2 1/c
15.948 (2)
12.8826 (16)

c/Å

25.922 (3)

Volume/Å3

5104.6 (11)

Z

4

1.631
0.473
/mm-1
F(000)
2552
2θ range
3.56° to 50.78°
Reflections collected
40776
Independent reflections
9374
R(int)
0.1129
R b = 0.0811
Final R indexes [I>=2σ (I)] 1 c
wR2 = 0.1854
R1 = 0.1434
Final R indexes [all data]
wR2 = 0.2136
2
1.064
Goodness-of-fit on F

a Obtained with graphite monochromated Mo Kα radiation (λ = 0.71073 Å) b R =  F  
1
o
Fc  Fo . c wR2 = {  [w(Fo2  Fc2)2 /  [w(Fo2)2]}1/2.

Table 5-2. Selected bond lengths from the x-ray crystal structure of [Fe(dcpep)2](PF6)2.
Selected Bond Lengths (Å)
Fe1—N1
1.980(5)
Fe1—N2
1.970(5)
Fe1—N3
1.990(5)
Fe1—N4
1.967(5)
Fe1—N5
1.957(5)
Fe1—N6
1.988(5)

176

Table 5-3. Selected bond angles from the x-ray crystal structure of [Fe(dcpep)2](PF6)2.

N1—Fe1—N3
N1—Fe1—N6
N2—Fe1—N1
N2—Fe1—N3
N2—Fe1—N6
N4—Fe1—N1
N4—Fe1—N2
N4—Fe1—N3

Bond Angles (o)
179.8(2)
N4—Fe1—N6
88.6(2)
N5—Fe1—N1
90.0(2)
N5—Fe1—N2
90.2(2)
N5—Fe1—N3
89.9(2)
N5—Fe1—N4
90.3(2)
N5—Fe1—N6
89.7(2)
N6—Fe1—N3
89.7(2)

178.9(2)
91.1(2)
178.7(2)
88.7(2)
90.8(2)
89.6(2)
91.4(2)

In addition to the crystal structure, there was enough pure product to obtain a 1H
NMR spectrum, but the rest of the characterization proved to be problematic as the ester
groups hydrolyze. To obtain a better overall picture, the iron(II) complex of the carboxylic
acid functionalized ligand, [Fe(dcpap)2](PF6)2, was prepared.

The complex was

synthesized by the standard method with slight modification, as shown in Scheme 5-5.

Scheme 5-5. Preparation of [Fe(dcpap)2]2+.
177

The dcpap ligand was suspended in water and the two acid groups deprotonated by
the addition of two equivalents of NaOH in order to solubilize the ligand and the solution
was degassed by bubbling with N2. A solution of ferrous ammonium sulfate was added
and the reaction mixture immediately darkened to a deep purple signifying formation of
the iron(II) complex. HPF6 was added to protonate the acid groups and obtain the desired
complex. The product was obtained in 21% yield after 1 hour, but it is possible that if
allowed to react longer the yield may be improved. For the first time, as opposed to all of
the other complexes described in this dissertation, the complex was isolated by filtration
with no evidence of byproducts. There was only evidence of the desired complex and
unreacted ligand which was easily removed by dissolving the complex in methanol and
filtering the insoluble ligand. The carboxylic acid analog proved to be more stable than
the ethyl ester analog and the solubility of the complex in methanol allowed for thorough
characterization.
Measuring

the

molar

absorptivity

of

either

[Fe(dcpep)2](PF6)2

or

[Fe(dcpap)2](PF6)2 is complicated by the hygroscopic nature of the complexes. For the
ethyl ester analog various attempts were made to obtain an extinction coefficient but a
consistent ground state absorption spectrum could not be obtained as the ester groups
hydrolyze over time. The acid analog is more stable but gains weight in air as it takes on
water, therefore an accurate extinction coefficient could not be determined. The ground
state absorption spectra of the two complexes are shown in Figure 5-4. The spectrum of
[Fe(dcpep)2](PF6)2 shows the normalized absorption of two samples where it is evident
that the spectrum changes as hydrolysis of the ester groups occurs over time. The
hydrolysis is apparent by eye as there is a color change from blue/purple to more red/purple.
178

The blue solid line was [Fe(dcpep)2](PF6)2 taken as it was freshly obtained from the
column following the reaction. The purple dashed line is an older sample that has clearly
changed as a result of a different degree of protonation.

The spectrum of

[Fe(dcpap)2](PF6)2 is shown on the right for comparison. A true comparison cannot be
made as this spectrum was taken in methanol, but we can clearly see that the features are

1.2

1.2

Absorbance

Normalized Absorbance

comparable to the hydrolyzed ester sample.

0.8
0.4
0.0

0.8
0.4
0.0

400
600
800
Wavelength (nm)

400
600
800
Wavelength (nm)

Figure 5-4. Ground state absorption spectra of [Fe(dcpep)2](PF6)2 in acetonitrile (left)
and [Fe(dcpap)2](PF6)2 in methanol (right). The normalized absorption spectra of two
crops of [Fe(dcpep)2](PF6)2 clearly show that the spectrum changes as the ester groups on
the complex undergo hydrolysis (— [Fe(dcpep)2](PF6)2) ([Fe(dcpep)2](PF6)2).

- -

hydrolyzed

A comparison of the optical properties of these analogs to the other complexes in
this dissertation could provide useful information for our investigation of inductive effects.
Unfortunately the integrity of [Fe(dcpep)2](PF6)2 is in question, and [Fe(dcpap)2](PF6)2 is
insoluble in acetonitrile, consequently we cannot draw an accurate comparison.

179

Though the spectral features show a strong dependence on the degree of
protonation, electrochemistry, may provide some insight. Sepehrifard et al. recently
investigated a series of Ru(II)-terpy dyes in which carboxylic acid and ethyl ester analogs
were prepared, and from the electrochemical data it appears that the difference in anchoring
group, ester versus acid, has a minimal impact on the oxidation and reduction potential,
showing only small variation of between 20-100 mV.21 On the other hand, Ferrere has
found a strong dependence of the electrochemical oxidation potential of [Fe(4,4dicarboxylate-2,2’-bipyridine)2(CN)2] on the degree of protonation, with the acid being
300 mV more positive than the tetrabutylammonium salt.22
The electrochemical potentials of [Fe(dcpep)2](PF6)2 were collected by cyclic
voltammetry, and though the degree of protonation of the carboxylate groups are in
question which increases the error of the experiment, we can use these values as
approximations. The oxidation and first reduction potentials are listed in Table 5-4. It
appears as though the inductive effect of the electron withdrawing carboxylate groups is
shifting the oxidation potential and first reduction potential more positive compared to the
unsubstituted dcpp. These values signify that there is stabilization of the HOMO t2g levels
of the metal and LUMO π* levels of the ligand. Another notable feature of the
electrochemistry of [Fe(dcpep)2](PF6)2 is the presence of eight reversible reductions. The
ability of the ligands to undergo four reductions each is attributed to the presence of the
electron withdrawing functional groups: the four carbonyl bridges and four carboxylate
groups.

180

Table 5-4. Electrochemical potentials for the oxidation and first reduction of
[Fe(dcpp)2](PF6)2 and [Fe(dcpep)2](PF6)2, measured with Ag/AgCl reference with 0.1 M
TBAPF6 electrolyte in acetonitrile, externally referenced to ferrocene. *These values are
approximations as hydrolysis may have occurred and the degree of protonation is unknown.
Compound
[Fe(dcpp)2]2+
[Fe(dcpep)2]2+

E1 / 2 [ox]
1.295 V
1.580 V*

E1 / 2 [red]
-0.965 V
-0.780 V*

Though the ester groups in [Fe(dcpep)2](PF6)2 are susceptible to hydrolysis, the
core structure seems to be quite stable, which is evident from the positive oxidation
potential. The [Fe(dcpap)2](PF6)2 analog appears to be stable as well as there is no change
after days of standing in methanol solution in the presence of air. This is promising for
application as a sensitizer in DSSCs, since dying the semiconductor requires soaking in an
alcohol solution for at least 24 hours, also since DSSCs are commonly solution state cells
in which there is an acetonitrile electrolyte solution. As for sensitization of TiO2, the eight
electron withdrawing groups have stabilized the orbitals to the extent that even the charge
transfer excited states lies below the conduction band of TiO2, making it unlikely that this
complex could be a viable sensitizer (Figure 5-5). If we are able to effectively tune the
excited states by the strategies described in Chapter 4, it may still be possible to develop a
viable sensitizer.

181

CB

-

-0.5
-0.23
0.15

E (V vs. NHE)

TiO2
2.22

+

[Fe(dcpep)2]2+
VB

Figure 5-5: Band structure of TiO2 compared to the energy level diagram of
[Fe(dcpep)2](PF6)2 showing that the MLCT excited states are lower that the conduction
band of TiO2.
5.3.3 Alternative Anchoring Groups for Dye-Sensitization
As mentioned in the introduction, the most widely employed anchor is the
carboxylate functional group, as it has been shown to bind to TiO2 with good stability and
provide strong electronic communication.

However, the ester linkages are only stable

under anhydrous conditions and in the presence of water are susceptible to hydrolysis
causing detachment of the dye from the oxide surface. Though the cells we are interested
in utilize dry acetonitrile, long term stability may be improved by utilizing less reactive
anchoring groups. The complications encountered during characterization of the ester
analog described in the previous section, make it clear that the instability could present an
issue.

182

We are looking into other options for anchoring groups and have found many
possibilities in the literature which include catecholates,8 boronates,9 phosphonates,10
malonates,23 and isophthalates24 (Figure 5-6). Two anchoring groups that stand out for
their high stability on attachment to TiO2 are acetylacetonate (acac) and hydroxamate.4–6
In addition to their stability, the relative ease of installation of these groups make them a
desirable alternative to the carboxylate group.

Figure 5-6: Functional groups that have been investigated for anchoring to a metal oxide
substrate.
As with the carboxylic acid groups, the introduction of the anchoring groups
requires a methyl group. The installation of methyl substitutents to the terminal rings has
already been achieved, however we are also interested in functionalizing the central ring.
The central pyridyl ring starting material is known in the literature and may be obtained
through various procedures and once in hand, the standard method used to prepare all of
the dcpp analogs described in this dissertation may be utilized.25,26
183

The p-tolyl

stubstituent, on the dcptp ligand described in Chapter 4, will also allow for installation of
anchoring groups, which will be attempted in due course.
The chemistry for the installation of acac should be fairly straightforward, as shown
in Scheme 5-6. For simplicity, the reaction scheme shows the dcpp analog with only the
central ring bearing a methyl substituent, but the procedure should be applicable to
functionalization of all methyl or p-tolyl substituents on the various dcpp analogs. The
proposed procedure is based on the known method of functionalization of 4-methylpyridine
via low temperature coupling with acetyl chloride.27,28 The acac group is known to be
unstable and may undergo reverse Claisen condensation until bound by the TiO2 substrate,
therefore it should be freshly prepared and utilized quickly in order to maintain integrity.
The hydroxamic acid group may be a better option for stability reasons.

Scheme 5-6. Proposed installation of acetylacetonate (acac) anchoring group.
Hydroxamic acids possess a wide range of biological activities with antibacterial,
antifungal, anti-inflammatory, and anti-asthmatic properties, therefore several methods
have been developed for their preparation and they are well documented in the literature.
Among the many methods, one stands out as a highly convenient synthesis involving the
transformation of a carboxylic acid to a hydroxamic acid and is shown in Scheme 5-7.29
184

Scheme 5-7. Proposed installation of hydroxamic acid anchoring group.
This route is ideal, as the chemistry for the installation of carboxylic acid groups
has already been worked out and the hydroxamic acid may be obtained in one step, under
mild conditions in high yields. The method employs cyanuric chloride as a coupling agent,
NMM base, and DMAP as a catalyst and complete conversion to the hydroxamic acid may
be achieved by simply stirring at room temperature for up to 12 hours. Triazine byproducts
may be simply removed by aqueous workup and the product obtained in pure form and
high yield.
Experiment and theory both point to hydroxamate and acac anchoring groups
having good binding and injection characteristics, even better than carboxylate groups.4–
6,11

With the culmination of the proposed syntheses we will be one step closer toward the

development of a viable iron(II) sensitizer.

185

5.4 Concluding Comments
We have successfully prepared an analog of [Fe(dcpp)2]2+ that possesses the
functionality required to bind to a semiconductor. The ester functionalized [Fe(dcpep)2]2+
is susceptible to hydrolysis, therefore the characterization was complicated; however, the
acid functionalized [Fe(dcpap)2]2+ was also prepared and proved to be much more stable,
and stability is key for the eventual use as a sensitizer in DSSCs. Synthetic efforts toward
the development of alternative anchoring groups have been proposed and will also be
pursued. With the ability to bind to the semiconductor we may begin a new branch of this
research and investigate this new class of iron(II) chromophores on TiO2 films.

186

APPENDIX

187

APPENDIX

Figure 5-7. 1H NMR of 2,6-di(2-carboxy(4-methylpyridyl))pyridine (dcmpp) in
(CD3)2CO.

188

Figure 5-8. 13C NMR 2,6-di(2-carboxy(4-methylpyridyl))pyridine (dcmpp) in
(CD3)2CO.

189

100

TOF MS ES+
318.1

%

657.2

340.1
257.1

596.2

0
200

300

400

500

600

m/z
Figure 5-9. ESI-MS of 2,6-di(2-carboxy(4-methylpyridyl))pyridine (dcmpp).

190

700

%

80

T
r 70
a
n 60
s
m 50
i
t 40
t
30
a
n
20
c
e 4000

3045.2
1408.6
1571.1
855.2
1050.5 773.4
1200.4
1596.9 1288.4

668.7

1682.9
3500

3000

2500

2000

1500

1000

Wavenumbers
Figure 5-10. IR of 2,6-di(2-carboxy(4-methylpyridyl))pyridine (dcmpp).

191

Figure 5-11. 1H NMR of 2,6-di(2-carboxy(4-pyridylcarboxylicacid))pyridine (dcpap) in
(CD3)2SO.

192

Figure 5-12. 13C NMR of 2,6-di(2-carboxy(4-pyridylcarboxylicacid))pyridine (dcpap) in
(CD3)2SO.

193

100

TOF MS ES376.1

%

187.5
332.1
408.1
474.0

753.1
1167.6

0
100

300

500

700

900

1100

m/z
Figure 5-13. ESI-MS of 2,6-di(2-carboxy(4-pyridylcarboxylicacid))pyridine (dcpap).
.

194

%
80
T
r
a 75
n
s 70
m
i 65
t
t
a 60
n
c 55
e 4000

3090.4
3444.0
1465.2
1712.5

1686.7

3500

3000

2500

2000

996.9
1186.5
760.8
1237.7
672.2
1327.7
1295.7

1500

1000

Wavenumbers
Figure 5-14. IR of 2,6-di(2-carboxy(4-pyridylcarboxylicacid))pyridine (dcpap).

195

Figure 5-15. 1H NMR of [Fe(dcpap)2](PF6)2 in CD3OD.

196

Figure 5-16. 13C NMR of [Fe(dcpap)2](PF6)2 in CD3OD.

197

100

TOF MS ES+

378.1

405.0
%

809.1

582.3
224.1

434.1
1240.0

0
200

400

600

800
m/z

Figure 5-17. ESI-MS of [Fe(dcpap)2](PF6)2.

198

1000

1200

%
T
r 70
a
n
s
m 65
i
t
t 60
a
n
c
55
e
4000

1611.5

3091.8

669.0
1237.0

3427.1
1725.1
1696.1

3500

3000

2500

2000

1500

Wavenumbers
Figure 5-18. IR spectrum of [Fe(dcpap)2](PF6)2.

199

842.1

1000

Figure 5-19. 1H NMR of 2,6-di(2-carboxy(4-pyridylcarboxylate ethyl ester))pyridine
(dcpep) in CDCl3.

200

Figure 5-20. 13C NMR of 2,6-di(2-carboxy(4-pyridylcarboxylate ethyl ester))pyridine
(dcpep) in CDCl3.

201

100

TOF MS ES+
434.1

530.2

%

456.1
462.2

536.2
604.3
889.2

406.1

1037.4

0
300

400

500

600

700
m/z

800

900

1000

Figure 5-21. ESI-MS of 2,6-di(2-carboxy(4-pyridylcarboxylate ethyl ester))pyridine
(dcpep).

202

Figure 5-22. 1H NMR of [Fe(dcpep)2](PF6)2 in (CD3)2CO.

203

Table 5-5. Geometric parameters for the X-ray crystal structure of [Fe(dcpep)2](PF6)2.

Fe1
Fe1
Fe1
Fe1
Fe1
Fe1
O1
O1
O1
O2
O3
O4
O5
O5
O6
O7
O7
O7
O8
O9
O10
O11
O11
O12
N1
N1
N2
N2
N3
N3
N4
N4
N5
N5
N6
N6
C1
C1

N1
N2
N3
N4
N5
N6
C6
C7A
C7B
C6
C9
C15
C21
C22
C21
C29
C30
C30A
C29
C32
C38
C44
C45
C44
C1
C5
C10
C14
C16
C20
C24
C28
C33
C37
C39
C43
C2
C9

Bond Lengths (Å)
1.980(5)
C11
C12
1.970(5)
C12
C13
1.990(5)
C13
C14
1.967(5)
C14
C15
1.957(5)
C15
C16
1.988(5)
C16
C17
1.305(8)
C17
C18
1.53(9)
C18
C19
1.42(9)
C18
C21
1.207(8)
C19
C20
1.215(7)
C22
C23
1.208(7)
C24
C25
1.321(8)
C24
C32
1.460(7)
C25
C26
1.192(8)
C26
C27
1.340(9)
C26
C29
1.449(11)
C27
C28
1.66(3)
C30
C31
1.201(8)
C30A C31A
1.213(7)
C32
C33
1.209(8)
C33
C34
1.325(8)
C34
C35
1.455(7)
C35
C36
1.199(8)
C36
C37
1.352(8)
C37
C38
1.352(8)
C38
C39
1.352(8)
C39
C40
1.360(8)
C40
C41
1.359(8)
C41
C42
1.354(8)
C41
C44
1.355(8)
C42
C43
1.349(8)
C45
C46
1.370(8)
P1
F1
1.356(8)
P1
F2
1.349(8)
P1
F3
1.357(8)
P1
F4
1.380(8)
P1
F5
1.493(8)
P1
F6
204

1.373(9)
1.369(9)
1.378(9)
1.495(9)
1.492(8)
1.374(8)
1.387(9)
1.384(9)
1.499(9)
1.371(9)
1.490(10)
1.365(8)
1.493(9)
1.381(9)
1.397(9)
1.491(9)
1.366(9)
1.49(2)
1.49(5)
1.505(9)
1.364(8)
1.389(9)
1.386(9)
1.383(9)
1.507(9)
1.482(9)
1.386(9)
1.389(9)
1.384(9)
1.501(9)
1.384(9)
1.475(10)
1.513(6)
1.501(6)
1.568(6)
1.507(6)
1.553(6)
1.585(7)

Table 5-5 (cont’d).
C2
C3
C3
C4
C7A
C7B
C9
C10
N1
N1
N2
N2
N2
N4
N4
N4
N4
N5
N5
N5
N5
N5
N6
C6
C6
C7B
C21
C29
C29
C30
C44
C1
C5
C5
C10
C10
C14
C16
C20
C20
C24
C28

C3
C4
C6
C5
C8A
C8B
C10
C11
Fe1
Fe1
Fe1
Fe1
Fe1
Fe1
Fe1
Fe1
Fe1
Fe1
Fe1
Fe1
Fe1
Fe1
Fe1
O1
O1
O1
O5
O7
O7
O7
O11
N1
N1
N1
N2
N2
N2
N3
N3
N3
N4
N4

1.380(9)
1.372(9)
1.493(9)
1.379(9)
1.57(10)
1.49(7)
1.511(8)
1.383(9)
N3
N6
N1
N3
N6
N1
N2
N3
N6
N1
N2
N3
N4
N6
N3
C7A
C7B
C7A
C22
C30
C30A
C30A
C45
Fe1
Fe1
C1
Fe1
C14
Fe1
Fe1
Fe1
C16
Fe1
Fe1

P2
P2
P2
P2
P2
P2
N7
C47
Bond Angles (o)
179.8(2)
N3
88.6(2)
O5
90.0(2)
O6
90.2(2)
O6
89.9(2)
O5
90.3(2)
N4
89.7(2)
N4
89.7(2)
C25
178.9(2)
C24
91.1(2)
C25
178.7(2)
C25
88.7(2)
C27
90.8(2)
C28
89.6(2)
N4
91.4(2)
O7
114(3)
O8
121(4)
O8
16(5)
O7
115.2(5)
C31A
123.1(7)
O9
98.3(11)
O9
30.9(9)
C24
117.0(5)
N5
122.6(4)
C34
120.9(4)
C34
116.5(5)
C33
122.3(4)
C36
116.5(5)
C37
121.1(4)
N5
120.6(4)
N5
122.4(4)
C36
116.5(5)
O10
121.4(4)
O10
122.5(4)
C39
205

F7
F8
F9
F10
F11
F12
C48
C48
C20
C21
C21
C21
C22
C24
C24
C24
C25
C26
C26
C26
C27
C28
C29
C29
C29
C30
C30A
C32
C32
C32
C33
C33
C33
C34
C35
C36
C37
C37
C37
C38
C38
C38

1.582(5)
1.587(5)
1.583(5)
1.578(5)
1.602(5)
1.606(5)
1.130(12)
1.460(15)
C19
C18
O5
C18
C23
C25
C32
C32
C26
C27
C29
C29
C26
C27
C26
O7
C26
C31
O7
C24
C33
C33
C32
N5
C32
C35
C34
C35
C36
C38
C38
C37
C39
C37

123.0(6)
111.5(6)
125.7(6)
122.8(6)
106.8(6)
123.5(6)
118.9(5)
117.3(5)
119.5(6)
118.2(6)
122.1(6)
119.6(6)
118.4(6)
124.4(6)
110.2(6)
126.4(6)
123.4(7)
106.9(11)
102(3)
121.2(6)
119.0(6)
119.2(5)
118.8(5)
123.4(6)
117.7(6)
120.0(6)
117.7(6)
119.7(6)
123.2(6)
120.1(5)
116.6(6)
119.4(6)
120.7(6)
119.4(6)

Table 5-5 (cont’d).
C28
C33
C37
C37
C39
C39
C43
N1
N1
C2
C3
C2
C4
C4
C3
N1
O1
O2
O2
O1
O1
O3
O3
C1
N2
N2
C11
C12
C13
C12
N2
N2
C13
O4
O4
C16
N3
N3
C17
C16
C17
C19
C19
C20

N4
N5
N5
N5
N6
N6
N6
C1
C1
C1
C2
C3
C3
C3
C4
C5
C6
C6
C6
C7A
C7B
C9
C9
C9
C10
C10
C10
C11
C12
C13
C14
C14
C14
C15
C15
C15
C16
C16
C16
C17
C18
C18
C18
C19

C24
Fe1
Fe1
C33
Fe1
C43
Fe1
C2
C9
C9
C1
C6
C2
C6
C5
C4
C3
O1
C3
C8A
C8B
C1
C10
C10
C9
C11
C9
C10
C11
C14
C13
C15
C15
C14
C16
C14
C15
C17
C15
C18
C21
C17
C21
C18

115.9(5)
122.0(4)
121.9(4)
116.0(5)
121.6(4)
118.2(5)
119.9(4)
123.0(6)
118.9(5)
118.0(5)
119.4(6)
122.4(6)
118.4(6)
119.2(6)
119.5(6)
123.2(6)
112.8(6)
125.4(6)
121.7(7)
104(6)
106(6)
120.9(6)
119.7(6)
118.7(5)
119.3(5)
123.4(6)
117.3(6)
118.7(6)
119.1(6)
119.8(6)
122.5(6)
119.7(5)
117.7(6)
119.6(6)
121.0(6)
118.9(5)
118.6(5)
123.3(6)
117.9(6)
119.2(6)
117.9(6)
118.0(6)
124.1(6)
119.9(6)

N6
N6
C40
C39
C40
C42
C42
C41
N6
O11
O12
O12
O11
F1
F1
F1
F2
F2
F2
F2
F2
F3
F4
F4
F4
F4
F5
F5
F7
F7
F7
F7
F8
F8
F9
F9
F9
F10
F10
F10
F10
F10
F11
N7
206

C39
C39
C39
C40
C41
C41
C41
C42
C43
C44
C44
C44
C45
P1
P1
P1
P1
P1
P1
P1
P1
P1
P1
P1
P1
P1
P1
P1
P2
P2
P2
P2
P2
P2
P2
P2
P2
P2
P2
P2
P2
P2
P2
C48

C38
C40
C38
C41
C44
C40
C44
C43
C42
C41
O11
C41
C46
F3
F5
F6
F1
F3
F4
F5
F6
F6
F1
F3
F5
F6
F3
F6
F8
F9
F11
F12
F11
F12
F8
F11
F12
F7
F8
F9
F11
F12
F12
C47

119.7(5)
122.1(6)
117.9(6)
119.5(6)
120.0(6)
118.4(6)
121.6(6)
119.5(6)
122.1(6)
111.3(6)
125.9(6)
122.8(6)
110.8(6)
88.8(5)
86.0(5)
91.1(6)
91.6(5)
92.5(4)
97.5(5)
91.6(4)
176.1(5)
90.3(4)
170.3(5)
87.8(5)
96.7(5)
79.9(5)
173.5(5)
85.8(4)
90.4(3)
92.5(3)
89.7(3)
178.9(3)
89.4(3)
89.6(3)
90.3(3)
177.8(3)
88.6(3)
91.5(3)
177.7(3)
90.9(3)
89.3(3)
88.5(3)
89.2(3)
178.3(12)

REFERENCES

207

REFERENCES

(1)

Desilvestro, J.; Graetzel, M.; Kavan, L.; Moser, J.; Augustynski, J. Journal of the
American Chemical Society 1985, 107, 2988–2990.

(2)

Ardo, S.; Meyer, G. J. Chemical Society Reviews 2009, 38, 115–64.

(3)

Nazeeruddin, M. K.; Grätzel, M. Struct. Bond. 2007, 123, 113.

(4)

McNamara, W. R.; Milot, R. L.; Song, H.; Snoeberger III, R. C.; Batista, V. S.;
Schmuttenmaer, C. a.; Brudvig, G. W.; Crabtree, R. H. Energy & Environmental
Science 2010, 3, 917.

(5)

McNamara, W. R.; Snoeberger, R. C.; Li, G.; Schleicher, J. M.; Cady, C. W.;
Poyatos, M.; Schmuttenmaer, C. a; Crabtree, R. H.; Brudvig, G. W.; Batista, V. S.
Journal of the American Chemical Society 2008, 130, 14329–38.

(6)

McNamara, W. R.; Snoeberger III, R. C.; Li, G.; Richter, C.; Allen, L. J.; Milot, R.
L.; Schmuttenmaer, C. a.; Crabtree, R. H.; Brudvig, G. W.; Batista, V. S. Energy
& Environmental Science 2009, 2, 1173.

(7)

Xiao, D.; Martini, L. a; Snoeberger, R. C.; Crabtree, R. H.; Batista, V. S. Journal
of the American Chemical Society 2011, 133, 9014–22.

(8)

Abuabara, S. G.; Cady, C. W.; Baxter, J. B.; Schmuttenmaer, C. A.; Crabtree, R.
H.; Brudvig, G. W.; Batista, V. S. Journal of Physical Chemistry C 2007, 111,
11982–11990.

(9)

Altobello, S.; Bignozzi, C. .; Caramori, S.; Larramona, G.; Quici, S.; Marzanni, G.;
Lakhmiri, R. Journal of Photochemistry and Photobiology A: Chemistry 2004,
166, 91–98.

(10)

Zakeeruddin, S. M.; Nazeeruddin, M. K.; Pechy, P.; Rotzinger, F. P.; HumphryBaker, R.; Kalyanasundaram, K.; Grätzel, M.; Shklover, V.; Haibach, T. Inorganic
Chemistry 1997, 36, 5937–5946.

(11)

Ambrosio, F.; Martsinovich, N.; Troisi, A. The Journal of Physical Chemistry
Letters 2012, 3, 1531–1535.

(12)

Pal, S. K.; Sundström, V.; Galoppini, E.; Persson, P. Dalton Transactions
(Cambridge, England : 2003) 2009, 10021–31.

(13)

Thavasi, V.; Renugopalakrishnan, V.; Jose, R.; Ramakrishna, S. Materials Science
and Engineering: R: Reports 2009, 63, 81–99.
208

(14)

Wang, D.; Mendelsohn, R.; Galoppini, E.; Hoertz, P. G.; Carlisle, R. A.; Meyer, G.
J. The Journal of Physical Chemistry B 2004, 108, 16642–16653.

(15)

Perrin, D. D.; Armarego, W. L. F. Purification of Laboratory Chemicals, 3rd ed.;
Pergamon Press: New York, 1988.

(16)

Goldsmith, C. R.; Stack, T. D. P. Inorganic Chemistry 2006, 45, 6048–55.

(17)

Garelli, N.; Vierling, P. The Journal of Organic Chemistry 1992, 57, 3046–3051.

(18)

SPARTAN 5.0 ed.; Wavefunction Inc.: Irvine, CA, 1997.

(19)

Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; J. A. Montgomery, J.; Vreven, T.; Kudin, K. N.; Burant, J. C.;
Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.;
Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.;
Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A.
J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth,
G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels,
A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Ortiz, J. V; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.;
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Challacombe, M.; Gill, P. M. W.; Chen, B. J. W.; Wong, M. W.; Gonzalez, C.;
Pople, J. A. Gaussian 03, Revision D.01 2004.

(20)

Schubert, U. S.; Eschbaumer, C.; Hochwimmer, G. Synthesis 1999, 779–782.

(21)

Sepehrifard, A.; Chen, S.; Stublla, A.; Potvin, P. G.; Morin, S. Electrochimica
Acta 2013, 87, 236–244.

(22)

Ferrere, S. Inorganica Chimica Acta 2002, 329, 79–92.

(23)

He, J.; Lindström, H.; Hagfeldt, A.; Lindquist, S. Solar Energy Materials and
Solar Cells 2000, 62, 265–273.

(24)

Xia, H.-L.; Ardo, S.; Narducci Sarjeant, A. a; Huang, S.; Meyer, G. J. Langmuir :
The ACS Journal of Surfaces and Colloids 2009, 25, 13641–52.

(25)

Chavarot, M.; Socquet, S.; Kotera, M.; Lhomme, J. Tetrahedron 1997, 53, 13749–
13756.

(26)

Yu, X.; Yang, T.; Wang, S.; Xu, H.; Gong, H. Organic Letters 2011, 13, 2138–41.

209

(27)

Mackay, L. G.; Anderson, H. L.; Sanders, J. K. M. Journal of the Chemical
Society, Perkin Transactions 1 1995, 2269.

(28)

Olivier, J.-H.; Harrowfield, J.; Ziessel, R. Chemical Communications (Cambridge,
England) 2011, 47, 11176–88.

(29)

Giacomelli, G.; Porcheddu, A.; Salaris, M. Organic Letters 2003, 5, 2715–7.

210

Chapter 6. Concluding Comments and Future Directions
6.1 Concluding Comments
The overall goal of this research project is the development of low cost sensitizers
for DSSCs by moving to first row transition metal-based chromophores, specifically
iron(II) polypyridyl complexes. To be an effective sensitizer the MLCT state of a
chromophore must be sufficiently long lived for injection into the semiconductor to occur.
Unfortunately, iron(II) polypyridyl complexes intrinsically have short lived MLCT states
due to ultrafast deactivation to lower lying ligand field states. The research has been
developed to gain more understanding of the fundamental causes underlying the ultrafast
processes.
Iron(II) polypyridyl complexes commonly possess a strained coordination
environment through which orbital degeneracy is removed.

The dense ligand field

manifold that arises from the deviation from octahedral symmetry may be mediating the
charge transfer to ligand field excited state deactivation. To investigate this theory, we set
out to develop a highly symmetric iron(II) chromophore. The first efforts yielded a
complex with a nearly perfect octahedral coordination environment that exhibits some
exciting properties. Though the charge transfer state lifetime has not been extended, it may
have introduced a strong enough ligand field to achieve inversion of the lowest energy
excited state.

The study was then redirected toward unraveling the sources of the

seemingly unique characteristics. The development of a control molecule is crucial, but
despite our best efforts, has yet to be attained and still ongoing.
An isostructural iron(II) analog was developed to investigate the impact of the
carbonyl bridging groups within the ligands. The desired symmetry was achieved with a
211

vinylidene analog, but the characteristics were vastly different, therefore it appears that the
properties are driven by inductive effects from the bridging groups. Since the inductive
effects are seemingly enhanced by the symmetrical environment we propose to take
advantage of that to further modulate the energetics of the excited states. We have just
begun to investigate the possibilities and are setting out to systematically introduce
substituents to the periphery of the ligands in an effort to induce stabilization of the charge
transfer state and destabilization of the ligand field states. With the first analog, in which
an extended -system has been introduced, it appears that we are on the right track. If
further functionalization has the desired effect, an iron(II) complex with a redox active
3MLCT

excited state as the lowest energy excited state could be realized, which would

have major impact toward application in DSSCs. Though a viable sensitizer has yet to be
developed, the first step has been taken toward implementing the new symmetric iron(II)
chromophores in a DSSCs by installing carboxylate anchoring groups. Work is ongoing
along this front and methods have been proposed to introduce alternative anchoring groups.
The synthetic efforts presented here provide a nice starting point to investigate
fundamental aspects of iron(II) chromophores. Though the goal of this research is highly
focused, it has presented the opportunity to sample from all areas of chemistry to develop
the synthetic methods. There is much work left to be done, as was proposed throughout
each chapter, but the molecules that have been realized have laid the foundation for further
analysis.
6.2 Future Directions
The research presented in this dissertation evolved from the observation of the
exciting properties of the symmetric iron(II) chromophore that was developed; however,
212

work has been ongoing from the original investigation of the role that torsional modes may
play in the ultrafast dynamics. As mentioned in Chapter 1, and discussed in detail in my
Masters Thesis, we set out to investigate the extent to which a torsional coordinate might
modulate MLCT to ligand field kinetics by introducing steric bulk to the terminal rings of
terpy and preparing the bis-tridentate iron(II) complexes.1 The terpy series yielded a lot of
information on the correlation between charge transfer deactivation and ground state
recovery, but did not slow the charge transfer deactivation.2 The next approach involves
fixed ligand cage structures, introducing a rigid environment that would restrict motion
along any torsional coordinate. We set out to prepare the two cryptand ligands that were
developed by Lehn and coworkers, shown in Figure 6-1.3,4

Figure 6-1. Cryptand ligand structures.

The structure of [bpy.bpy.bpy] that is linked through the 6-position of the bpy rings
may make an ideal ligand to introduce the desired rigidity, but does raise a point of concern.
It is well established that substitution at the ortho position of polypyridines leads to the
stabilization of the high spin form when bound to iron(II).5 The iron-nitrogen bond lengths
in high spin iron(II) complexes are longer than in low spin complexes due to the population

213

of the antibonding orbitals. If the cage structure is able to bind to iron(II), it is our hope
that the compound is rigid enough to force the low spin ground state upon binding. If this
is the case, then the rigidity of the molecule may destabilize the high spin excited state,
thereby interfering with the rate of relaxation from the charge transfer to the ligand field
states. The ligand synthesis is quite time consuming, but progress has been made; however
efforts were turned toward the analog that is linked through the 5-position as its prospects
may be more exciting, and is much more synthetically accessible.
The Lehn cryptand is readily obtained through well established chemistry. The
method involves a straightforward Schiff base condensation of two equivalents of tris(2aminoethyl)amine (tren) and three equivalents of 5,5-diformyl-2,2’-bipyridine (dfbpy),
which may be prepared through many known methods from 5,5’-dimethyl-2,2’-bipyridine
(dmb).6–8 While 4,4’-dfbpy may be simply prepared by direct oxidation from 4,4’-dmb
with selenium dioxide, the 5-position is much less reactive and another route is necessary.9
5,5’-dmb may be oxidized to the carboxylic acid via Jones oxidation and converted to the
ethyl ester through Fisher esterification.10,11 The ester may be readily reduced to the
hydroxymethyl derivative with sodium borohydride, which can then be converted to dfbpy
through Swern oxidation.6,12
The tren capped 5-position bpy cage ligand was successfully prepared from the
method described by Lehn.4

The reaction was stirred at room temperature, in dry

acetonitrile for four days and the reaction progress monitored by eye since the orange
Schiff base product precipitates from the reaction as it forms. The ligand is simply filtered
recrystallized from methanol. The imine functional groups appear to be unstable toward

214

hydrolysis during the subsequent reaction to prepare the iron(II) complex, which has been
unsuccessful. All is not lost yet as there is an alternate plan for this structure.
The cryptand was originally designed for its complexation ability to bind to metals
through the tren caps and bpy ligands and trinuclear-silver(I) and binuclear-copper(I)
complexes are known.4 The Cu(I) ions showed selectivity to the four coordinate tren
position and left the bpy ligands free. As iron(II) has a high propensity for six coordination
we should be able to achieve a mixed metal trinuclear complex with Fe(II) selectively
coordinating to the bpys and Cu(I) coordinating to the imine groups.

This could be quite

interesting as it would contain an iron(II) chromophore in the center flanked by two Cu(I)
species. We intend to explore this molecule for potentially interesting redox properties, as
the presence of the Cu(I) centers could provide a remote site for electron transfer
regeneration, specifically we are looking into the possibility of reductive quenching of the
MLCT state of Fe(II) by Cu(I). This would be highly desirable for application in DSSCs
as upon photoexcitation, a bpy radical is created by the MLCT from Fe(II) to bpy leaving
Fe(III), if Cu(I) could effectively reduce Fe(III) prior to relaxation it could drive electron
transfer into the semiconductor.
The tri-nuclear complex was readily obtained, as shown in Scheme 6-1, through a
simple one pot procedure of dissolving the ligand in dichloromethane and adding
acetonitrile solutions of Fe(II) and Cu(I) salts and excess NH4PF6. The complexation
occurred quickly, as it turned blue from the addition of Fe(II) and turned red as Cu(I) was
added and it began to precipitate immediately upon the addition of NH4PF6, yielding a
dark purple/maroon solid. It is interesting to note that the coordination is highly selective
as the reaction was carried out in an inert atmosphere dry box creating much difficulty in

215

weighing out the desired quantities to ensure one equivalent of Fe(II) and two equivalents
of Cu(I). Due to the small scale it was nearly impossible to obtain the desired amounts and
excess Fe(II) and Cu(I) were added, however this did not appear to affect the outcome of
the reaction.

Scheme 6-1. Preparation of the iron(II) dicopper(I) Lehn cryptand trinuclear complex
[Cu2Fe(5-Lehn-cryptand)]4+.

The characterization of this complex is just beginning and the reaction conditions
yet to be optimized, as it was only prepared once as a test reaction. It is present by ESIMS, however the purity is difficult to ascertain at this point. Luckily, crystals suitable for
X-ray crystallography were grown by ether diffusion into an acetonitrile solution of the
compound, and the structure is in the process of being solved. Disorder from solvent
molecules has complicated the process, however it is clear that selective coordination of
the Fe(II) and Cu(I) in the desired positions has been achieved, as shown in Figure 6-2.
Another critical piece of evidence ascertained from the crystal structure is that a low spin

216

iron(II) center has been achieved, which is evident in the Fe—N bond lengths, listed in
Table 6-1.

Figure 6-2. ORTEP Drawing of the cation of [Cu2Fe(5-Lehn-cryptand)]4+ obtained from
single-crystal x-ray structure determination (in progress). Hydrogen atoms and anions are
omitted for clarity.

Table 6-1. Selected bond lengths from the x-ray crystal structure of the tri-nuclear
cryptand [Cu2Fe(5-Lehn-cryptand)]4+.
Selected Bond Lengths (Å)
Fe1—N1
1.975(7)
Fe1—N2
1.978(7)
Fe1—N3
1.977(7)
Fe1—N4
1.974(6)
Fe1—N12
1.986(7)
Fe1—N13
1.980(7)

Whether this cryptand can be employed as a redox relay type system will depend
on numerous factors which will be assessed in due course, with electrochemical and
spectroelectrochemical measurements, fluorescence upconversion measurements, time217

resolved spectroscopy, etc. Analysis of the electrochemistry of the intricate tri-nuclear
cryptand has just begun, however it is proving to be highly complicated and without further
characterization we cannot draw any conclusions just yet.

In mononuclear iron(II)

polypyridyl complexes the electrochemistry is fairly straightforward in that there is one
oxidation that may be attributed to the oxidation of metal center, and multiple reductions
that are attributed to the ligand. In multinuclear systems with similar cryptand ligands the
electrochemistry is much more complicated, and a few details about the known systems
will be outlined here in order to guide future efforts with this complex system.13–15
From the standard redox potentials of iron and copper, where Fe(III) + e- → Fe(II)
is 0.77 V and Cu(II) + e- →Cu(I) is 0.159 V vs. SHE, it is apparent that copper is more
easily oxidized than iron.16 However, the ligand environment and presence of other metal
centers will influence the redox chemistry. A comparison of a known di-copper cryptand
with analogous hetero-nuclear Cu(I) M(II) cryptands (M(II) = Fe(II), Ni(II), Co(II) Mn(II))
has provided evidence of an electrostatic influence on the oxidation potential of Cu(I).15
There is a correlation of the oxidation potential of Cu(I) with proximity to a divalent cation,
becoming harder to oxidize the closer it is to the divalent cation. The influence of the Cu(I)
center on the oxidation potential of the other metal is unclear as the researchers were unable
to distinguish the metal oxidation from a feature attributed to the oxidation of the ligand.
This raises concern with this system since the cryptand ligand is similar, and the possibility
of ligand oxidation will complicate the identification of features in the electrochemistry.
A second complication can arise from the relative ease of reduction of Cu(I) and
Cu(II), as it has been noted in various sources that sweeping negative can result in
stripping/deposition of copper on the electrode anywhere past around -0.9 to -1.1 V vs.

218

ferrocene (Fc).13,15

Evidence of the copper deposition can be observed with the

appearance of a wave at -0.49 V and the oxidation of the copper can then result in the
presence of [Cu(CH3CN)4]+ giving rise to an oxidation feature at +0.55 V vs. Fc, therefore
a small scan window must be utilized to prevent degradation of the complex.14 As it stands
we cannot gain any conclusive information from the electrochemistry that was taken as the
scan window was too large; the electrochemical data must be recollected and a bulk
electrolysis experiment should also be done.
The development of analogs of the cryptand system may provide useful information
to unravel the complicated electrochemistry. To date, the attempted isolation of the
cryptand with iron(II) alone has failed, however the di-copper(I) analog of this cryptand
could be useful. It is quite possible that instead of clarifying it could complicate the picture
even further as the environment would be vastly different; nevertheless it should provide
evidence involving the electrostatic influence of neighboring metal centers. Overall, this
is the first system of this type that we have set out to investigate and it appears to be quite
complex, and without further characterization it is unclear where this may lead. Keeping
in line with the running theme of this entire research project, we set out on the path toward
the investigation of ultrafast dynamics of iron(II)-polypyridyls and it appears to be ever
evolving in a new direction.
In closing, there are so many avenues that one may take from here, from all the
syntheses proposed throughout this dissertation to this new path presented here. As
outlined in the introduction, there are two approaches we could take toward developing a
viable iron(II) sensitizer: 1) speed up the electron injection process, or 2) slow down the
relaxation process. This research has been developed to tackle the second approach and

219

has diverged along two distinct paths, in which there is much left to be done along both
fronts. The investigation of torsional modes that was introduced in my Masters Thesis can
be pursued further with the development of rigid 6-position Lehn cryptand structure shown
in Figure 6-1. The syntheses proposed throughout chapters 2-4 can be carried out to further
investigate means to tune the electronic structure of iron(II) polypyridyl chromophores
through enhanced symmetry and inductive effects. The approach of speeding up the
electron transfer process has yet to be targeted, however the installation of alternative
anchoring groups that was proposed in Chapter 5 may be taken on as a first step. Finally,
the investigation into the trinuclear cryptand system presented here is just getting underway
and may present a nice starting point to a whole new area of research. Overall, the
prospects are wide open.

220

APPENDIX

221

APPENDIX

100

TOF MS ES+

│◄x30

251.1

%
235.1
1129.2

341.1
474.1

647.1

883.3

1029.2

1233.2

1439.1

0
200

400

600

800

1000

1200

1400

m/z
Figure 6-3. ESI-MS of [Cu2Fe(5-Lehn-cryptand)]4+. [C48H48Cu2F24FeN14P4]4+ m/z =
251.1; [C48H48Cu2F24FeN14P4](PF6)2}2+ m/z = 647.1; [C48H48Cu2F24FeN14P4]
(PF6)3}+ m/z = 1439.1.

222

C u r r e n t (A)

2.0E-6

0.0

-2.0E-6

2

1

0

-1

-2

-3

Potential (V)
Figure 6-4. Cyclic voltammogram of [Cu2Fe(5-Lehn-cryptand)]4+ measured with a
Ag/AgCl reference with 0.1 M TBAPF6 electrolyte in acetonitrile (50 mV/s), referenced
to ferrocene. The presence of the feature at -0.49 V is attributed to stripping/deposition of
copper, therefore the complex degrades under the present conditions and must be
recollected.

223

REFERENCES

224

REFERENCES

(1)

Jamula, L. L. Masters Thesis, Michigan State University, 2010.

(2)

Brown, A. M. PhD Dissertation, Michigan State University, 2011.

(3)

Rodriguez-Ubis, J.-C.; Alpha, B.; Plancherel, D.; Lehn, J.-M. Helv. Chim. Acta
1984, 67, 2264.

(4)

De Mendoza, J.; Mesa, E.; Rodríguez-Ubis, J.-C.; Vázquez, P.; Vögtle, F.;
Windscheif, P.-M.; Rissanen, K.; Lehn, J.-M.; Lilienbaum, D.; Ziessel, R.
Angewandte Chemie International Edition in English 1991, 30, 1331–1333.

(5)

Halcrow, M. a. Polyhedron 2007, 26, 3523–3576.

(6)

El-ghayoury, A.; Ziessel, R. The Journal of Organic Chemistry 2000, 65, 7757–
7763.

(7)

Narayan, K. S.; Geetha, K. V.; Nakmanovich, G.; Ehrenfreund, E.; Eichen, Y. The
Journal of Physical Chemistry B 2001, 105, 7671–7677.

(8)

Hodacová, J.; Budesínský, M. Organic Letters 2007, 9, 5641–3.

(9)

Sasaki, I.; Daran, J. C.; Balavoine, G. G. A. Synthesis 1999, 815–820.

(10)

Szeto, K. C.; Kongshaug, K. O.; Jakobsen, S.; Tilset, M.; Lillerud, K. P. Dalton
Transactions 2008, 2054–60.

(11)

Yu, S. C.; Hou, S.; Chan, W. K. Macromolecules 2000, 33, 3259–3273.

(12)

Telfer, S. G.; Bernardinelli, G.; Williams, A. F. Dalton Transactions 2003, 435–
440.

(13)

Brooker, S.; Ewing, J. D.; Ronson, T. K.; Harding, C. J.; Nelson, J.; Speed, D. J.
Inorganic Chemistry 2003, 42, 2764–73.

(14)

Coyle, J.; Downard, A. J.; Nelson, J.; McKee, V.; Harding, C. J.; Herbst-Irmer, R.
Dalton Transactions 2004, 2357–63.

(15)

Ronson, T. K.; Nelson, J.; Jameson, G. B.; Jeffery, J. C.; Brooker, S. European
Journal of Inorganic Chemistry 2004, 2570–2584.

(16)

Douglas, B.; McDaniel, D.; Alexander, J. Concepts and Models of Inorganic
Chemistry; Third Edit.; John Wiley & Sons, Inc.: New York, 1994.
225