LIBRARY

Michigan State
University

 

 

 

PLACE IN RETURN BOX
to remove this checkout from your record.
TO AVOID FINES return on or before date due.

 

MTE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

11” mm“

 

THE AMIDINIUM-CARBOXYLATE SALT BRIDGE AND
ELECTRON TRANSFER REACTIONS
By
James Patrick Kirby

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY

Department of Chemistry
1 997

ABSTRACT

THE AMIDINIUM—CARBOXLYATE SALT BRIDGE AND ELECTRON TRANSFER
REACTIONS
By
James Patrick Kirby

Salt bridges occur in a variety of natural structural motifs including proteins,
protein-protein complexes, protein-substrate complexes, and protein-DNA or protein—RN A
assemblies. In addition to their role as natural structural building blocks, salt bridges also
control reaction chemistry by modulating redox activity at the protein active site. One way
this occurs is when salt bridges mediate redox processes, as in sulfite and nitrite reductases
(SiRs and MRS). Salt bridges are also proposed pathways for electron transfer in
cytochrome c oxidase. We are interested in examining the affect of proton motion or the
proton position within a salt bridge on fixed distance electron transfer (ET) rates. There are
an array of chromophore/quencher complexes where an electron donor is held at some
fixed distance relative to an electron acceptor. Upon excitation of either the donor or
acceptor, depending on the design of the particular molecule, it is possible to time the
electron as it makes its journey from one point to another, that is to measure the rate of
photoinduced ET at some fixed distance. This strategy has been employed to characterize
the distance dependence of ET reactions, and also assess the role of solvation in fixed
distance ET reactions. Alternatively, pulse radiolysis experiments have been employed to
measure the rates of electron transfer at fixed distances. From these classical experiments
the existence of the Marcus inverted region has been established experimentally, nearly 30

years after it was proposed by theory. In the inverted region for ET, increasing the

exergisity of an ET reaction results in a decrease in the reaction rate, in sharp contrast to the
prediction of classical chemical reaction theory.

Presented here is a series of donor/acceptor complexes assembled via the non—
covalent amidinium—carboxylate salt bridge. A series of donor/acceptor complexes have
been chemically modified with the amidine functional group. It is shown that the amidine
group is readily transformed into a 1:1 arnidinium-carboxylate salt bridge with concomitant
protonation by a carboxylic acid. This salt bridge structural motif parallels Asp-Arg salt
bridges occurring in proteins, yet has the advantage of having only one specific two
hydrogen bond binding mode, where the guanidine group of arginine has multiple binding
modes. The synthetic details of preparing and characterizing amidines and their salt bridge
complexes are described. The association constants for various amidinium—carboxylate salt
bridge complexes have been determined by either titration or dilution 1H NMR methods.
The synthesis and characterization of 1:1 amidinium-carboxylate salt bridge complexes and
kinetics measurements aimed at exploring proton-coupled electron transfer are presented for

inorganic, bioinorganic and organic complexes.

To my parents, Jeannine and James, grandparents, Katherine and James Wood, and my
brothers and sisters, Anita, Keith, Linda, Marcy, and Johnn.

iv

ACKNOWLEDGEMENTS

Traveling the long and winding road towards completing the research and writing
of this thesis I was fortunate to meet and work with many talented individuals. Much of
their insight made the research for this thesis possible. Dan Nocera saw me through the
highs and lows of graduate school, always teaching me to keep things in perspective, and
to look at the “big picture”. I am grateful to him for his guidance and support. Many thanks
go to my guidance committee, Professor C. K. Chang, who served as my second reader,
Gerry Babcock and Ned Jackson. John Miller provided me with a unique experience at
Argonne National Laboratory mid-way through my graduate career. These individuals have
acted as my mentors, and have demonstrated to me how to think in excruciating detail about
scientific problems.

I would like to thank my friends and group members, past and present, including
Mark Torgerson, Jim Roberts, Dimitri Popostakis, Rich Marassas, Dan Engebretson, Sara
Helvoight, Einhard Schmidt, Adrian Ponce, Ying Liang, Carolyn Hsu, Zoe Pikramenou,
Yongqi Deng, Dean Lantero, Doug Motry, Carl Iverson, Ann MacIntosh, Al Barney,
Linda Kraus and Eric Saari for providing me with many entertaining moments, and for
making my graduate experience a lot of fun. These individuals (and the ones I forgot to list)
have made my life a little easier over the last few years. Finally, I would to thank God and
the St. Johns student parish for providing me with shelter and a spiritual outlet during the
many good times here, but also for the few times when my troubles and worries tried to get

the best of me.

TABLE OF CONTENTS

LIST OF FIGURES .................................................................................. x

LIST OF TABLES ................................................................................. xiv

LIST OF ABBREVIATIONS ..................................................................... xv
CHAPTER 1

PERSPECTIVES ON SALT BRIDGES AND ELECTRON TRANSFER .................... 1

A. SALT BRIDGES .................................................................................. l

1. Introduction ..................................................................................... l

2. Examples of Salt Bridges ...................................................................... 3

a. Guanidine Based Salt Bridges .............................................................. 3

b. Arginine Based Salt Bridges. .............................................................. 5

c. Amidine Based Salt Bridges ................................................................ 7

B. ELECTRON TRANSFER ..................................................................... 14

1. Introduction ................................................................................... l4

2. Theory to Describe Electron Transfer Rates ............................................... 15

a. Classical Chemical Reaction Rate Theory ............................................... 15

b. Marcus Theory ............................................................................. 16

c. Semiclassical Marcus Theory ............................................................. 17

3. Observation of and Implications of the Marcus Inverted Region ....................... 17

a. Free Energy Effects on kEr ................................................................ 17

b. Effects of Temperature on kEr in the Marcus Inverted Region ........................ 18

vi

c. Structural and Solvent Perturbations and the Effects on kET .......................... 18

4. Photoinitiated ET in Covalent Dyad and Triad Systems ................................. 20

a. ET in Porphyrin-Quinone D/A Dyad .................................................... 20

b. ET in D-Chromophore-A Triads ......................................................... 20

C. PROTON-COUPLED ELECTRON TRANSFER .......................................... 23
1. Introduction ................................................................................... 23

2. Photoinitiated ET in Semi-Synthetic Protein Mediated D/A Dyads ..................... 23

a. Histidine-33 Modified Cytochrome c .................................................... 23

b. Dipolar Effects in D/A modified a-Helical Polypeptides ............................. 24

3. Biological Salt Bridge Proton-Coupled Electron Transfer in Sirohemereductase ..... 24

4. Photoinitiated ET in Hydrogen Bonded D/A Dyads ...................................... 26

a. Carboxylic Acid Dimer D/A Dyads ...................................................... 26

b. ET in a Guanine-Cytosine Base Paired D/A Dyad ..................................... 30

D. THESIS OBJECTIVES ........................................................................ 30
1. Association of Amidinium with Carboxylate in Solution and in the Solid State ...... 30

2. The Amidinium—Carboxylate Salt Bridge and Proton-Coupled Electron Transfer. . . . 31

CHAPTER 2

SOLUTION AND SOLID STATE AMIDINTUM-CARBOXYLATES ...................... 35

A. INTRODUCTION ............................................................ I .................. 35

B. SYNTHESIS .................................................................................... 38
1. Tetrabutylammonium Benzoate ............................................................. 38
2. Benzamidinium Chloride .................................................................... 38
3. Benzamidinium-Benzoate .................................................................... 4O
4. 3-Amidinium-benzoate ....................................................................... 40

C. EQUILIBRIUM MEASUREMENT VIA THE NMR TECHNIQUE .................... 41

D. EFFECTS ON SALT BRIDGE EQUILIBRIUM CONSTANT .......................... 47

vii

1. Salt Bridge Hammett Effects from para Substituted Bemzamidiniums ................. 47

2. Effect of Benzoic Acid Concentration on the Solution Salt Bridge Equilibrium ...... 50

3. Effect of Chloride Ion Concentration on the Salt Bridge Equilibrium .................. 52

4. Carboxylate Concentration Effects ......................................................... 52

E. SOLID STATE STRUCTURE OF 3-AMIDINIUM-BENZOATE ........................ 54

F. EXPERIMENTAL .............................................................................. 64
CHAPTER 3

RUTI-IENIUM(II)TRIS(2,2’-BIPYRIDII\1E) COMPLEXES TO PROBE PCET .......... 66

A. INTRODUCTION .............................................................................. 66

B. BPY COMPLEXES OF RU(II) TO MEDIATE OXIDATIVE PCET .................... 69

1. Results and Discussion ...................................................................... 69

a. Solution Association of D/A PCET Complexes ........................................ 69

b. Photophysical Behavior ................................................................... 71

c. Kinetic Isotope Effects on PCET ........................................................ 72

C. TMBPY COMPLEXES OF RU(II) TO MEDIATE OXIDATIVE PCET ................ 76

1. Results and Discussion ...................................................................... 77

a. Solution Salt Bridge Association ......................................................... 77

b. Photophysics ............................................................................... 79

c. PCET ........................................................................................ 87

2. Conclusions ................................................................................... 88

D. DECB AND BPY COMPLEXES OF RU(II) TO MEDIATE REDUCTIVE PCET . . . . 90

1. Results and Discussion ...................................................................... 91

a. Solution Association ............................................................... - ........ 91

b. Photophysical Behavior ................................................................... 91

2. Conclusions ................................................................................... 95
E. MATERIALS AND METHODS .............................................................. 96

viii

CHAPTER 4

FUNCTIONALIZATION OF PORPHYRIN WITH THE AMIDINE GROUP ........... 102
A. INTRODUCTION ............................................................................. 102
B. RESULTS AND DISCUSSION ............................................................. 102
1. Synthesis ..................................................................................... 102
2. Solution Salt Bridge Association .......................................................... 104
3. Photophysical Behavior ..................................................................... 108
C. MATERIALS AND METHODS ............................................................. 108
CHAPTER 5
PULSE RADIOLYSIS EXPERIMENTS TO PROBE PCET ................................ 111
A. INTRODUCTION ............................................................................. 1 l l
B. BACKGROUND .............................................................................. 1 l2
1. Measurement of Equilibria Between Radical Anions .................................... 112
2. Kinetic Measurements Employing the Pulsed Radiolysis Experiment ................ 114
3. Hydrogen Production in the Pulse Radiolysis Experiment ............................. 117
C. RESULTS AND DISCUSSION ............................................................. 119
1. Synthesis ..................................................................................... 119
2. Salt Bridge Mediated Electron Transfer ................................................... 122
D. MATERIALS AND METHODS ............................................................. 126
LIST OF REFERENCES ......................................................................... 129

LIST OF FIGURES

Figure 1. Salt bridges composed of (a) 2guanidine (b) arginine and (c) amidine. X may be
SO-R', C-R', P(O-R')2, SO", and PO2 “. ........................................................ 4

Figure 2. Solid state materials based on the pseudo planar hexagonal salt bridge
interactions of guanidinium (G) with organic sulfonate (S) as shown schematically in (a); A
three dimensional pillared bilayer structural motif results from the interaction of 4,4'-
biphenyldisulfonate with guanidinium (b) which has been shown to accommodate small
organic molecules such as styrene (c). ............................................................. 6

Figure 3. Biological examples of DNA-Protein recognition (a) and RNA-Protein
recognition (b) ......................................................................................... 8

Figure 4. Stick structure representation of the active site of two chorismate mutase
enzymes-inhibitor bound complexes from Bacillus subtilis (a) and Saccharomysce cerevisia
(b) ....................................................................................................... 9

Figure 5. Schematic representation of a salt bridge complex formed between an a-helical
polypeptide, Ac-Ala—Ala—Gln-Asp—Ala-Ala-Asp-Ala-Ala-Ala-Ala-Ala-Gln—Ala—Ala-Tyr-
CONHZ, and a bis-guanidinium receptor molecule in water ................................... 10

Figure 6. An amidinium complex to model the action of phosphodiesterase .............. 12
Figure 7. A self-replicating amidinium-carboxylate system to model natural replication. 13

Figure 8. Log of the rate constant for intrarnolecular ET reactions, km, versus reaction
exergonicity, AG°, observed in eight molecules (0) and fit to theory (solid line) as
discussed in the text. ............................................................................... 19

Figure 9. Photoinitiated electron transfer reactions yielding charge separation in a
porphyrin based dyad ZnP—NQ (a), a porphyrin based triad TMDAB—ZnP—NQ (b), and a
ruthenium based triad PZT—Ru(dmb)3—AQ (c) .................................................. 21

Figure 10. Hydrogen bonded protein mediated ET reactions in (a) a modified cytochrome
c with a surface AsRu(II)-His-33 group and (b) a functionalized Ot-helix in the peptide
sequence Boc-NH-Ala-Aib-Ala-(DMA or Pyr-Ala)-Ala-Aib-Ala-Ala-Aib-Ala-(Pyr or
DMA-Ala)-Aib-Ala-COzEt which sets up a helix dipole moment (“M“). .................. 25

Figure 11. A schematic representation of the active site of SiRs and MRS proteins. (a)
siro heme where a rectangle is used to represent the heme atomic framework; (b), substrate
(protonated sulfite) bound enzyme; (c) inhibitor (phosphate ion) bound enzyme ........... 27

Figure 12. Photoinitiated electron transfer in zinc porphyrin (ZnP) hydrogen bonded
complexes where the electron acceptor is a dinitro aromatic, ZnP—(COZH)2-DNB (a), an
iron(III) porphyrin, ZnP—(COZH)2-Fe(III)P (b) and benzoquinone, ZnP—GC—BQ (c).. . . 29

Figure 13. The amidinium-carboxylate salt bridge formed by the equilibrium between an
amidinium and a carboxylate group to yield a primary two-point hydrogen bond (a);
stacking of amidinium-carboxylate salt bridges through secondary one-point hydrogen
bonding (b); D/A amidinium-carboxylate salt bridge complexes with two distinct
orientations of the salt bridge for the study of fixed distance ET (c) .......................... 32

Figure 14. Synthetic reaction sequence employed for the preparation of (a)
tetrabutylammonium benzoate by treatment of benzoic acid with tetrabutylarnmonium
hydroxide; (b) para-substituted benzamidinium chloride salts by reaction of para substituted
benzonitriles with sodium methoxide followed by ammonium chloride; (c) benzamidinium-
benzoate salt bridges by precipitation from water; and (d) the preparation of 3-amidinium-
benzoate by reaction of 3—cyanobenzoic acid with sodium methoxide, tetrabutylarnmonium
hydroxide and crystallization from water ((1). ................................................... 39

Figure 15. 1H NMR spectrum of benzamidinium chloride at a concentrtion 5.2 mM in
DMSO-d6 (bottom). The IH resonances of the aromatic ring appear between 7.5 and 8.0
ppm. The amidinium protons appear as a broad singlet at 9.3 ppm, but separate and
sharpen to two singlets in the presence of 16.6 mM benzoic acid (second from the bottom).
Selected spectra are shown upon addition of the tetrabutylamonium benzoate at
concentrations 0.0, 0.0, 1.5, 2.3, 3.0, and 5.0 mM (bottom to top) ......................... 43

Figure 16. Plot of the change in chemical shift of the benzamidinium N-Hn proton versus
the concentration of benzoate from 0.0 to 25 mM. The concentration of benzamidinium
chloride is 5.5 mM, and the concentration of benzoic acid is 16.6 mM throughout all

concentrations of benzoate (as the tetrabutylamonium salt) .................................... 44

Figure 17. Job’s plot of the mole fraction of the benzamidinium-benzoate salt bridge
complex vs. the mole fraction of benzamidinium (6.3 mM) as the concentration of benzoate
is varied from 0.0 to 25 mM ....................................................................... 45

Figure 18. Hammett plot for K and para substituted benzamidiniums ..................... 49

Figure 19. The effect of benzoic acid concentration on K. The chloride concentration is
held constant at 0.005 M ........................................................................... 51

Figure 20. Affect of chloride ion concentration on the observed binding constant (K).
The concentration of benzoic acid is constant 0.050 M for each measurement of K ........ 53

Figure 21. Change in chemical shift of benzamidinium NHax (Q) and NHeq (A) with the
addition of benzoate 0 to 886 mM. ............................................................... 55

Figure 22. Proposed mechanism for stepwise association of three carboxylates to one
amidinium, yielding a net four hydrogen bonds to the amidinium in solution ............... 56

Figure 23. Structure of 3-aminidinium-benzoate; oxygen (red), nitrogen (blue), carbon
(light gray) and hydgrogen (dark gray). The aromtic hydrogens are omitted for clarity. . .59

Figure 24. The packing of the zig-zag tapes in the be plane of the crystal strucutre of l.
The amidinium-carboxylate salt brdgie is roatated 26.1° (plane defined by the oxygen and
nitrogen atoms of the salt bridge) out of the plane of the aromatic ring. ..................... 60

Figure 25. View of the interlayer one-point secondary hydrogen bonding network. Each
subunit of 1 is connected to four other distinct subunits of 1 via the secondary one-point

hydrogen bonding interaction (two above and two below with respect to the the layers of
two-point primary interactions displayed in Figure 24). ....................................... 61

Figure 26. Rotation of the crystal structure of 1 such that the ladder is viewed in the
plane of the paper. The ladder structure is formed from the hydrogen bonding between
protons external to the salt bridge of a given tape and the carboxylate oxygens of salt
bridges in neighboring interlayers. ............................................................... 62

Figure 27. Complexes to study PCET; 2 is based on [(bpy)2Ru"(Mebpy-amH“)]3+ and is
based on [(bpy)2Ru"(Mebpy- COOH)]2”. ....................................................... 70

Figure 28. 1H NMR spectrum of [(bpy)2Ru"(Mebpy-amH*)]3*. Selected spectra are
shown upon addition of the tetramethylammonium salt of 3,5-dinitrobenzoate at
concentrations 0, 0.44, 0.89, 2.2, and 3.1 mM in DMSO.d,3 (bottom to top). The IH
resonances of the bipyridines appear between 7.3 and 9.2 ppm, two broad singlets flanking
9.5 ppm signify the H resonances of the amidinium protons. For the spectrum at 0.44 M
carboxylate added, the two resonances are coincident .......................................... 73

Figure 29. Plot of the chemical shift of the anridinium protons hydrogen-bonded to
carboxylate versus concentration of carboxylate from 0.44 to 40 mM. ...................... 74

Figure 30. Job’s plot of the relative salt-bridge complex concentration vs. the mole
fraction of [(bpy)2Ru"(Mebpy-amH*)]3+ (3.8 leD as the concentration of DNBCOO‘ is
varied from 0.0 to 40.0 mM ....................................................................... 75

Figure 31. Complexes to study PCET; 4 is based on [(tmbpy)2Ru"(Mebpy-amH*)]3*, 5
is based on [(tmbpy)2Ru"(Mebpy'COO‘)]+ and 6 is based on [(tmbpy)2Ru"(Mebpy-
COOH)]2+. .......................................................................................... 78

Figure 32. Spectral changes of lH NMR spectrum of [(tmbpy)2Ru"(bpy-amH*)]3” with
added tetrabutylammonium 3,5-dinitrobenzoate. Selected spectra are shown upon addition
of 3,5-dinitrobenzoate at concentrations 0.0, 2.3, 3.7, 6.6, 10.3 and 13.9, mM in DMSO
d,5 (bottom to top). The 1H resonances of the bipyridines appear between 7.0 and 9.2 ppm,
two broad singlets flanking 9.5 ppm signify the 'H resonances of the axial (internal) and
equatorial (external) amidinium protons .......................................................... 80

Figure 33. Plot of the chemical shift of the resonances shown in Figure 32 for
[(tmbpy)2Ru"(Mebpy-amH*)]3” and additional resonances of the amidinium protons
hydrogen-bonded to carboxylate versus concentration of carboxylate from 0.0 to 80 mM.81

Figure 34. Job’s plot of the relative salt-bridge complex concentration vs. the mole
fraction of [(tmbpy)2Ru"(Mebpy—amH“)]3+ (6.3 mM) as the concentration of 3,5-
DNBCOO‘ is varied from 0.0 to 80 mM ......................................................... 82

Figure 35. 'H NMR spectrum of the tetraphenylborate salt of 3,5-DNBamH“ with
addition of the PF; salt of [(tmbpy)2Ru"(Mebpy-COO’)]* at concentrations 0.0, 1.8, 2.9
and 3.9 mM in DMSO-d6 (bottom to top). The 1H resonances of the axial (internal) and
equatorial (external) amidinium protons appear between 9.3 and 12.0 ppm ................. 83

Figure 36. Plot of the chemical shift of the resonances shown in Figure 35 and additional

resonances of the amidinium protons hydrogen-bonded to Ru" carboxylate complex from
0.0 to 10 mM ........................................................................................ 84

xii

Figure 37. Job’s splot of the relative salt-bridge complex concentration vs. the mole
fraction of 3, 5- DNBamH” (2. 9 mM) as the concentration of the [(tmbpy)2Ru"(Mebpy-
COO )] is varied from 0. 0 to 10 mM ............................................................ 85

Figure 38. Complenxes to study PCET; 7 is based on [(decb)2Ru" (Mebpy- -amH )]3", 8 is
based on [(bpy)2RuII (Mebpy- -amH" )]3" and 915 based on [(decb)2Ru "(Mebpy- COO' )] .92

Figure 39. Preparation of ocatethyl porphyrin with the amidine functional group: (i)
(EtO)2PCH2CN, NaNHz; (ii) MeAl(Cl)NH2, NaOH; (iii) 6M H2304, NaOH, Zn(OAc)2,
NaOH; (iv) 3,4—dinitrobenzoic acid. ............................................................ 103

Figure 40. Selected 1H NMR spectra of 12 in the presence of one equivalent of 3,4-
DNBCOOH for concentrations of 1.7, 2.6, 3.5, 4.3, 6.1, 13.0 and 26.0 mM in DMSO—d6
(bottom to top). The spectral range captures the amidinium and vinylic protons of the
interface and the meso protons of the octaethyl porphyrin ring. ............................. 105

Figure 41. Plot of the chemical shift of the amidinium protons hydrogen bonded to
carboxylate versus concentration of 12 and 3,4—DNBCOOH. .............................. 107

Figure 42. Electron flow in the pulse radiolysis experiment for mixed aromatic D/A
compounds. ........................................................................................ l 18

Figure 43. Synthesis of complexes for the study of electron transfer using the pulse
radiolysis technique, (i) instant ylide; (ii) perchloric acid, dichloromethane; (iii)
hydroxylamine hydrochloride, formic acid; and (iv) DBU, water. .......................... 121

Figure 44. Typical decay profiles on the nanosecond timescale for radical anions of

{FlCazAmCy} (top), {FlCazAmN a} (middle), and {CyCa:AmN a} (bottom) generated via
the pulse radiolysis experiment in NMP and monitored at 650 nm. ......................... 123

Figure 45. Observed ET rates in {FlCa: AmNa} measured at 1.1, 4.,4 7..6 and 18.1
mM. ................................................................................................. 125

xiii

LIST OF TABLES

Table 1. Hammett effects on the association constants of para-substituted benzamidiniums

(X) with benzoate ................................................................................... 48
Table 2. Primary hydrogen bond distances for l. ............................................ 63
Table 3. Secondary one-point hydrogen bond distances for l. ............................. 63

Table 4. Rates for unimolecular and bimolecular electron transfer for donor-acceptor
complexes with amidinium-carboxylate and dicarboxylic acid dimer bridges in
dichloromethane at 22 °C. ......................................................................... 89

xiv

ACN
Aib

AmzCa
AQ
Arg
Asp
BPh4
bpy
BQ

Ca

cyt c

decb
DMAB
DMSO

LIST OF ABBREVIATIONS

excited state

acceptor

acetyl

acetonitrile
a—aminoisobutyric acid
alanine
amidinium-carboxylate
anthraquinone

arginine

aspartate

tetraphenylborate
2,2’-bipyridine
benzoquinone

carboxylate

cytochrome c

donor

doublet

dichloromethane
4,4’-di(ethylcarboxy)-2,2-bipyridine
N ,N-dimethylaminobenzene

dimethylsulfoxide

XV

DNA
DNB

FCWD

GC
Gln
His
NBu4
NMR
NMP

NQ

PCET

ppm

THF
tmbpy
TMDAB

deoxyribonucleic acid
dinitrobenzene

electron transfer

ethyl

Franck-Condon weighted density of states
guanidinium

guanine-cytosine

glycine

histidine

tetrabutylammonium

nuclear magnetic resonance
N-methylpyrrolidone

l ,4-naphthaquinone

protein

pentet

proton-coupled electron transfer
parts per million

pyrene

phenothiazene

quartet

ribonucleic acid

sulphonate

singlet

triplet

tetrahydrofuran
4,4’,5,5’-tetramethyl-2,2-bipyridine
N,N ,N ’ N’ -tetramethyl-p-diaminobenzene

xvi

Tyr tyrosine

ZnP zinc(II)porphyrin

xvii

Chapter 1
PERSPECTIVES ON SALT BRIDGES AND ELECTRON TRANSFER

A. Salt Bridges

1. Introduction

Hydrogen bonds are ubiquitous in nature, biomirnetic and synthetic systems
forming the basis for biological and chemical suprarnolecular assemblies as described by
Lehn.I A particular type of hydrogen bonding interaction is the salt bridge, where the
components of the hydrogen bond consist of a positively charged proton donor and a
negatively charged proton acceptor. A complementary two point salt bridge hydrogen
bonding interaction occurs between 1,3 spaced charged pairs such as the one shown
schematically on the left side of Eq. 1.1. This and other salt bridge structural motifs are
present throughout biological systems and are involved in essential biological
transformations. In bioenergetics for instance, the coupling of proton motion to the motion
of the electron is evident where salt bridges are present. However, the role that salt bridges

play in such schemes has yet to be clearly defined.

\ \
\ \

‘N—H ------ o ‘N—H§+ ,o\8-
----- (:6) @:/X ""'<:@ ::":‘@;’X
N—H ------ O 'N—H5’+ ‘05-

I I
I I

’ ’ (1.1)

primary secondary

The stabilizing forces that result in the formation of a salt bridge are a combination
of hydrogen bonding interactions and electrostatic interactions, which stem from the
charged nature of the individual components of the salt bridge. First, there exists a strong
primary electrostatic interaction associated with the attraction between positively charged
adjacent N—H protons and negatively charged oxygen atoms. This is in part symbolized by
the dashed lines connecting H with O in the left side of Eq. 1.1. But there exists another
electrostatic force acting to stabilize the salt bridge, a secondary electrostatic interaction
(SEI).2 This SEI is due to the attraction between the diagonal arrangement of partially
positive charged N-H protons and the partially negative charged oxygen atoms. On the
right side of Eq. 1.1, this important contribution to the overall stabilization of the salt
bridge moiety is symbolized by diagonal dashed lines.

Due to the ionic nature of the salt bridge, there is a pronounced electrostatic
interaction with the medium surrounding the salt bridge. In solution, the strength of the salt
bridge bond is inversely proportional to the dielectric constant (as) of the surrounding
solvent molecules as described by Coulomb energy, Eq. 1.2, where ABC is the electrostatic
energy stabilizing the salt bridge within a solvent of some as, ql and q2 are the respective
charges of the salt bridge proton donor and proton acceptor, and r12 is the distance
separating qI and q2. Therefore, in solvents with low 85's, the salt bridge interactions are
expected to be strong; conversely in solvents with high 65's, the salt bridge interactions are

expected to be weak.

(91 ‘92)

AB :
C (‘12 '83) (1'2)

Within this context, salt bridges located in the interior of proteins display unique
behavior, inasmuch as that the as for the interior of proteins is dependent on the structure of

the local or surrounding protein side chain groups.3 The magnitude of the electrostatic

attraction in the salt bridge within the interior of a protein is effectively on a sliding scale,
because ABC is a function of the es of the surrounding protein side chain groups versus the
bulk dielectric of the solvent, water. This characteristic of salt bridges in biological systems

may be especially important in bioenergetic cycles based on charge transport.

2. Examples of Salt Bridges

Examples of salt bridge interactions are found in diverse areas or research ranging
from crystal engineering to biochemistry. A class of salt bridges important to this thesis are
ones formed from anion complexation with protonated positively charge guanidine,
ariginine and amidine groups, depicted schematically in Figure l. The molecular basis for
each of the three groups listed here determines the overall specificity and pattern of salt
bridge hydrogen bonding. For the guanidinium ion there exists a three fold symmetry axis,
and salt bridge interactions can occur through one, two or three sets of two-point hydrogen
bonds as shown in Figure l (a). Close inspection of the structure of the protein side chain
group, arginine, shows that it is in fact structurally rooted to guanidine group through a
covalent link from one of the three guanidinium nitrogen atoms to the polypeptide
backbone. For the arginine group there exists lower symmetry giving rise to salt bridge
hydrogen bonding interactions though one or two sets of two-point hydrogen bonds as
shown in Figure 1 (b). The diverse structural motifs germane to the arginine group are
extensively employed in an array of biological functions in nature. Arginine based salt
bridge hydrogen bonds are essential in protein recognition and activation of DNA, RNA as
well as in catalysis in biological systems. Finally, the amidinium group displays only one

mode of specific two point hydrogen bonding as shown by Figure l (c).

a. Guanidine Based Salt Bridges
A series of crystalline salt bridge complexes structures have been investigated by
Etter and Ward based on the interaction of guanidinium with a variety of different sulfonate

groups.4 One of the most notable systems is the recently reported guanidinium-sulfonate

leQ 1-
o' O
H H
\ \ \
H N-H H N—H

/ ’ {H (9 lN—H
I” IH I” IH
0;,6 0;,6
x: x=o
(a)

o o
H\ N—H H\ N—H ------- o\
IN--- IN--- Gilx
H (+3 N—H H (9 N—H ------- o
, H H
o'\.‘@ ,.
X‘—=o’
(b)
H
\
N—H ------- o\
R «ea 9:}
N—H ------- o
/
H
(C)

H\ N—H ------- o\
/N. - - 9 I X
”H @H/N—H ------- o

O
\ .’ N—H ....... O\
/N.—.:<, 9 :/X
{H (9 lN—H ------- o
I” IH
ore ,’
x=o’

Figure 1. Salt bridges composed of (a) guanidine (b) arginine and (c) amidine. X may be

SO-R’, C-R', P(O-R')2, so': and Po, -.

salt sheet structure shown in Figure 2 (a). The basic building block shown there forms
from the packing of three guanidinium groups with three sulfonate groups yielding a
pseudo hexagonal packing motif. Represented in Figure 2 (b) is formation of pillared
bilayer sheets, upon co-crystallization of the two equivalents of guanidinium with one
equivalent of biphenyl-4,4'-disulfonate. In this structure, layered sheets pack together in
the crystal lattice. The biphenyl groups extend out from the sulphonates to form pillars
between the two dimensional sheets of salt bridge hydrogen bonds. This pillaring effect of
the biphenyl spacers creates large channels between the salt bridge layers. These open
channels have been structurally characterized by x-ray crystallography with the channels
occupied by methanol or host molecules such as styrene, as represented in Figure 2 (c).
These remarkable systems represent a pivotal achievement in the field of crystal engineering

inasmuch as they point the way towards the design of host lattices.

b. Arginine Based Salt Bridges.

Arginine salt bridges are prevalent in nature. They are especially prominent in zinc
finger protein-DNA complexes. Crystallographic studies on this class of proteins reveals
the formation of protein-DNA assemblies, which are stabilized by buttressed salt-bridge
interactions between aspartate (Asp) and arginine (Arg) within the protein main chain
forming two hydrogen bonds to a guanine-cytosine (GC) base pair in the major groove of
DNA, Figure 3 (a).5 A similar structural motif has been revealed for a class of proteins,
which interact in a likewise fashion forming protein-RNA complexes as revealed by
solution NMR structural studies, Figure 3 (b).6 RNA stem loops form a similar recognition
complex with arginamide, which exhibits buttressed phosphodiester salt bridge interaction
in place of the carboxylate interaction the DNA complex. These studies reveal that salt-
bridges are the key to the assembly of protein/DNA or RNA complexes.

Arginine salt-bridges also figure prominently in biological transformations. The

conversion of chorismate ion to prephenate ion, Eq. 1.3, is a rare example in nature of the

 

 

i e :
“D’s"O‘x .
‘H\ ,H" H‘R “H\ /H“
‘ H :N G) H"”’O\\‘H :N ® H' a
\N” ‘N/ \N,’ ‘N/ |
' ' I I S
'7' H '7' H ‘G/ \ ,
(a) E 5 5 e 5 = I I
’,('),§C:’}é~\ ”,é.?s.;é“ /s\G/s\
li\ “H\ /H” ”x '
x01? 16 “.03
\H\ ”s' ,H"
'7‘ "i‘
*l 'f'
061' O I ' l
{.32. ~c/S\ s’
‘8’ \ _____,, /
O I H‘ L/ \G/ \
H\ @ ’N-H / \Gfi :
(b) 1 2 N-.’. __> ~é/ \
O H ’N—H / /G/S\c’
\S— H ’ \ / ‘
x \ /G \G/5
,G" »S\ /
0973,}0 ’G/sx
OO

 

 

(c) O =Q

 

 

 

 

Figure 2. Solid state materials based on the pseudo planar hexagonal salt bridge
interactions of guanidinium (G) with organic sulfonate (S) as shown schematically in (a); A
three dimensional pillared bilayer structural motif results from the interaction of 4,4'-
biphenyldisulfonate with guanidinium (b) which has been shown to accommodate small
organic molecules such as styrene (c).

Claisen rearrangement (pericylic reaction) and is present in the biochemical synthetic
pathway of the essential amino acid phenyl alanine. Chorismate mutase from Escherichia
coli accelerates the transformation by more than 6 orders of magnitude. The x-ray crystal
structure of two chorismate mutase proteins have been solved with an inhibitor bound to

the active site, Figure 4.7'8

(1.3)

   

chorismate ion prephenate ion

A synthetic peptide sequence containing a pair of aspartates (Asp) spaced at an
interval of i + 3 along the peptide main chain is shown to recognize a synthetic bis-
guanidinium receptor in water.9 The proposed solution structure involves two salt bridges
between the bis-guanidinium (arginine like) receptor and the two carboxylates from the i +

3 spaced Asp peptides along the (rt-helix, as shown schematically in Figure 5.

c. Amidine Based Salt Bridges

A model system for the action of phosphodiesterase is presented in Figure 6.'0 Here
the primary hydrogen bonding interactions formed by the amidinium group to a
phosphodiester salt bridge is employed to withdraw electron density from the
phosphodiester, which then undergoes an intrarnolecular trans-phosphodiesterification
reaction. This reaction pathway is at the heart of the action for phosphodiesteraces, which
are involved in the transformation of phosphodiesters such as those located in the backbone

linkages of DNA and RNA. This enzyme catalyses the overall transformation by more than

(b)
GS ‘N—H----o
’
'7'
’ taro
7P.
6" 0

Figure 3. Biological examples of DNA-Protein recognition (a) and RNA-Protein

recognition (b).

(a)

(b)

 

Figure 4. Stick structure representation of the active site of two chorismate mutase
enzymes-inhibitor bound complexes from Bacillus subtilis (a) and Saccharomysce cerevisia

(b).

10

 

Figure 5. Schematic representation of a salt bridge complex formed between an (Jr-helical
polypeptide, Ac-Ala-Ala-Gln-Asp-Ala-Ala—Asp—Ala-Ala—Ala-Ala-Ala-Gln-Ala-Ala—Tyr—
CONHZ, and a bis-guanidinium receptor molecule in water.

ll

twelve orders of magnitude. Salt bridge interactions are important in the active site of this
enzyme. Systems aimed at modeling the active site of the protein contain amidinium linked
to an ethanol pendant group of an aromatic. This protein mimic reacts efficiently with
model phosphodiesters and displays Michaelis-Menten type reaction kinetics, which is
consistent with an enzyme substrate complex from which an irreversible reaction occurs.
The crystal structure of the product of this reaction shows strong hydrogen bonding
between the phosphodiester group of the product and the amidinium group of the template.
The kinetics were assessed against control molecules and Michaelis—Menten parameters kcm
and Km could be determined and a 9000—fold rate enhancement was calculated. This model
system is an elegant example of how salt bridge interactions can be employed in a synthetic
setting to influence reaction chemistry.

Amidinium—carboxylate salt bridges are the key recognition component for a self-
replicating chemical system, Figure 7.” In this pseudo-living system, a complementary
pair of reactive molecules were prepared where one molecule contains the carboxylate and
aldehyde groups while a second molecule contains the amidinium and amine groups. In the
first step of the sequence, a molecule of each are brought into contact by two
complementary salt bridge interactions. In the reaction scheme described in Figure 8,
Kiedrowski shows how a self-replicating system reacts to form templates of the original
molecule and water. The next step allows for the reaction of the localized aldehyde group
with the amine group to yield an imine linkage and water. At this stage the molecule is
transformed into a template dimer, which is then free to exchange into free template as
governed by the equilibrium characteristic of the salt bridge. The net concentration of the
template doubles and each can then go on to form two more copies of the template, in a
numerical progression. In a very simplistic sense, the autocatalysis described is analogous
to one circuit of the polymerase chain reaction (PCR) technique employed by molecular

biologist for gene amplification, which is a special case of autocatalysis.

12

 

 

 

Figure 6. An amidinium complex to model the action of phosphodiesterase.

l3

 

Figure 7. A self-replicating amidinium-carboxylate system to model natural replication.

14

B. Electron Transfer

1. Introduction

Electron transfer (ET) reactions have been characterized for a variety of model
systems though few have involved salt bridge interactions. More generally an electron
donors (D) and electron acceptors (A) are separated at some fixed distance defined by a
rigid spacer group. Systems of this type have been investigated as models of the early
events in photosynthesis, where natural D/A moieties are held at fixed distances and
orientations as governed by the supporting protein matrix. Experiments to measure the rates
for ET reactions in D/A model systems invoke the use of time resolved spectroscopy.
Absorption of photons in the form of visible pulsed laser light can yield excited states (D*
or A*) in molecular ensembles, which can subsequently undergo charge separation ET
reactions via a reductive mechanism as described by Eq. 1.4 or by an oxidative mechanism
as described by Eq. 1.5, relative to the excited state. Whether charge separation proceeds
through the reductive or oxidative pathway depends on thermodynamics of the particular
D/A system to model ET reaction rates. The kinetics for the subsequent ET events can be
monitored by transient absorption spectroscopy techniques or by times resolved emission

lifetime measurements for which specific examples will be described in a later section of

this chapter.
D*—spacer—A ——-) D"“—spacer—A“1 (1 .4)
D—spacer—A* ———> D’*‘—spacer— "‘ (1.5)

In simple dyad D/A systems, the light absorbing chromophore is either the ultimate
D or A. This basic design has been elaborated in recent years to include triads, tetrads and
pentads. In the simplest extension to triads, the reductive and oxidative components are

constructed about the light-absorbing chromophore. The ET events originate from the

15

chromophore excited state (chromophore*) for which charge separation events involving

both oxidation and reduction are summarized by Eq. 1.6.

D’“-chromophore"‘-A

/ \

D—ch romophore*—A D°“‘—chromophore—A"1 ( l .6)

\ D-chromophore”"—A"1 /

Photons are not a prerequisite for triggering ET reactions. Irradiation with high
energy pulsed electrons can produce reactive charged radical anions (D“') or radical
cations, depending on the chemical conditions of the pulse radiolysis experiment. As
depicted in Eq. 1.7, measurement of ET rates in systems of this type have been studied and

have provided a wealth of insight into the physical description and understanding of ET.

D“‘—spacer—A ————> D—spacer— "‘ (1.7)

2. Theory to Describe Electron Transfer Rates

a. Classical Chemical Reaction Rate Theory
The foundations of ET theory are rooted in the Fermi equation or golden rule
equation, Eq. 1.8, which states that the reaction rate for electron transfer, km.” is a product

of an electronic term and the nuclear coordinates. For instance, with respect to fixed

21: 2
kET =7|HDA| FCWD (1.8)

distance ET in the pulse radiolysis experiment, the electronic term is quantified by the
electronic coupling matrix element, HDA, and is obtained by solving the time dependent

Schrodinger equation for the of mixing of the two wave functions at the intersection of the

l6

reactant and product potential energy surfaces corresponding to the molecules depicted on
the left side of Eq. 1.7, and the right side of Eq. 1.7. This intersection point, or crossing
point, is where the nuclear coordinates are presumed to be frozen in some transition state
geometry, according to the Born-Oppenheimer approximation. Restated, in the transition
state, at the curve-crossing, the nuclear coordinates of the reactant surface are equal to the
nuclear coordinates on the product surface. The nuclear coordinates of the system form the
Franck-Condon-weighted density of states (FCW D). This theory is operative in the non-
adiabatic limit where the electronic coupling matrix element, HDA, is much smaller than the
thermal energy, kBT (HDA << kBT, where kB is Boltzmann's constant, and T is the absolute

temperature of the system).‘2

b. Marcus Theory
The classical Marcus expression describing kEr is obtained when the FCWD term in

Eq. 1.17 is substituted with the expression shown in Eq. 1.9 for parabolic reactant and

{0.5 + AG°)2}
(4nxskBT)

 

FCWD: (41c).,1<,,'r)"/2 exp — (1.9)

product potential energy surface. The displacement of the wells or potential energy surfaces
describing the nuclear coordinates of the product surface and reactant surface is its as

described by Marcus in Eq. 1.10. The definition of As as put forth by Marcus is the energy
__ 2 —1 -1 -1 —1 —1
xs-e [(2%) Ha) 4RD.) ](es ) (1.10)

required to change the solvation geometry in the solvent shell surrounding the system, in
going from the reactant potential energy surface to the product potential energy surface.
This energy is essentially a Coulomb energy for an ellipsoid as calculated by approximating

the size of the solvent shell of system by a radius for the donor and acceptor portions to the

17

system or molecule as to and rA which are separated by some distance RDA and the dielectric

of the solvent and the optical dielectric constant 80p.

c. Semiclassical Marcus Theory

The semiclassical Marcus expression as put in its final form by Jortner, Eq. 1.11,
includes the electron-phonon coupling term S which is the ratio of the inner sphere
reorganizational energy, Av, and a high frequency molecular vibrational mode, hv, Eq.
1.12. The summation of this quantum mechanical vibrational progression is employed in
describing kEr and is obtained by substituting the FCWD of Eq. 1.8, with the expressions
shown in Eq. 1.11 and 1.12.

FCWD = (41tlskBT)—V2 z exp(—S) S— exp[-— {

 

(its + AG° + (ohv)2}
(0! (1.11)

(41tASkBT)

(0:0

Szlv/hv (1.12)

3. Observation of and Implications of the Marcus Inverted Region

a. Free Energy Efiects on kg,

The pulse radiolysis technique is a powerful experimental tool for studying fixed
distance intrarnolecular ET reactions. In the pulse radiolysis experiment short pulses of
high energy electrons generate negative ions, which are strong reductants, such as the
biphenyl radical anion, which can go on to reduce a neighboring electron acceptor group.
The reaction proceeds with a unimolecular rate constant for the ET event, kn, as shown
schematically in Eq. 1.13. Applying this technique to the study of ET reactions, Miller and
Closs have provided the first unequivocal experimental evidence for the existence of the
Marcus inverted region, where it is demonstrated that increasing the ET reaction driving

force (—AG°) beyond the solvent reorganization energy (Its) results in a decrease in km, as

18

described by Eq. 1.11.‘3 A plot of the observed ET rate constant, kET, versus —AG° in the
shown in Figure 8. Also shown is the fit to the theory shown above, Eq. 1.11. The
agreement is excellent between the fit and the experimentally observed km, with the use of
four adjustable parameters in the fit, AG°, 15, AV and hv. This result is in stark contrast to
classical theories developed to describe reaction chemistry and bears out Marcus’ elegant
prediction of the inverted region more than 30 years prior to the experiments of Miller and

Closs.

ker

0

e—1
g , . __.. a , . (1.13)
A “ya A A y‘“ AM
b. Efiects of Temperature on kg in the Marcus Inverted Region
Semiclassical Marcus expressions are revealed in temperature dependence studies
on kEr in the normal and inverted regions.” Temperature studies on normal region ET are
predicted by the classical Marcus expression Eq. 1.9 due to the fact that quantum
mechanical effects in the normal region are minimal; restated, in the normal region most ET
occurs primarily to the lowest vibrational level or when w is 0, Eq. 1.11 reduces to Eq. 9.
However, little or no temperature dependence on k5.r in the inverted region. This confirms

the validity of the quantum mechanical treatment of the FCWD described in Eq. 1.11.

c. Structural and Solvent Perturbations and the Eflects on kn.

Other studies aimed at emphasizing the importance of the semi-classical Marcus
expression have shown the effects on kEr by perturbing its by altering the molecular
structure of the D portion of the dyad system. In a molecule where a rigid fluorene D group
is substituted for the biphenyl D group, Its is changes by 0.15 eV. This translates to an

increase in kEr in the normal region, where A is 2-naphthalene, by factor of 8 and a

19

 

1010

l llLlLl

109

l I l l 1 1 1 ll
Inhgmx
O

1 11111:]
O

.—L
O
\l
O
0’0
x
'11

r r lllLll
I.
O

8

 

 

 

—AG °/eV

Figure 8. Log of the rate constant for intrarnolecular ET reactions, km, versus reaction
exergonicity, AG°, observed in eight molecules (0) and fit to theory (solid line) as
discussed in the text.

20

decrease in kEr in the inverted, where A is benzoquinone region by a factor of 2.5, Eq.

1.14.'5

 

4. Photoinitiated ET in Covalent Dyad and Triad Systems

a. ET in Porphyrin-Quinone D/A Dyad

Wasliewski has prepared and studied a wide range of molecules based on reducing
excited state of metalloporphyrins such as the complex shown in Figure 9 (a).”’ The ZnP*
in this particular system is in proximity to an electron deficient electron acceptor,
naphthaquinone (NQ). Photoexcitation produces an ET reaction in 25 picoseconds,
yielding a charge separated state represented schematically by Eq 1.16. The charge
separated species decays in 370 picoseconds back to the ground state via a thermal ET
reaction. The progress of the reaction is monitored by time resolved transient absorption
spectroscopy incorporating an optical delay line, which allows for time increments on the

order of a few picoseconds to be assessed in kinetic measurements.

ZnP*—NO ———> ZnP‘“—NQ"’ (1.16)

b. ET in D-Chromophore-A Triads

To extend the lifetime of this charge separated state, as is the case in photosynthesis
in biology, a sequence of oxidation and reductive steps are employed to extend the overall
charge separation out over an even greater distance. This process is summarized in Eq 1. 17
for the triad complex shown Figure 9 (b).'7 This is essentially a two step ET process. The

process combines ET to the NQ in one direction and a hole transfer (a form of ET) in the

21

(a)

 

TM DAB-Zn P—NQ

I-Z

. 060

(C)

 

PZT-Ru(dmb)3—AQ

Figure 9. Photoinitiated electron transfer reactions yielding charge separation in a
porphyrin based dyad ZnP—NQ (a), a porphyrin based triad TMDAB—ZnP—NQ (b), and a
ruthenium based triad PZT—Ru(dmb)3—AQ (c).

22

opposite direction to the tetramethyl-p-diaminobenzene (TMDAB moiety) to yield an overall
charge separated state on the right side of Eq. 1.17. These consecutive ET events are
complete in under 500 picoseconds. The resulting charge separated species decays with a
time constant of 4 milliseconds, long enough to record its EPR signal, which is remarkably
similar to that measured in protein systems. It is an impressive result that adding one more
ET event into the scheme extends the overall lifetime of the charge separated species by

greater than a factor of a million.

TMDAB—ZnP*-NQ ———> TMDAB’”—ZnP—NQ"1 ( l . l7)

Meyer and coworkers have also prepared D-chromophore-A complexes of the type
shown in Figure 9 (c) as photophysical models or mimics of the light harvesting and charge
separation events in photosynthesis.'8 The model incorporates [Ru(dmb)3]2* (dmb is 4,4’-
dimethyl-2,2’-bipyridine) as a synthetic support from which two distinct types of redox
active organic moieties are attached. Each dmb ligand is functionalized, one with PZT, and
two with anthraquinone. Absorption of visible light by [Ru(dmb)3]2+ yields an excited
state, [Ru(dmb)3*]2", which simultaneously reduces one of the anthraquinones to its radical
anion (km = 580 nm, 8 = 4000 M“I cm“), and oxidizes the PZT to its radical cation (km:
510 nm), as monitored by time resolved transient absorption spectroscopy, Eq. 1.18. The
lifetime of the charge separated species is 150 nanoseconds, where systems functionalized
with only a PZT or only an anthraquinone live for a few nanoseconds. This remarkable
system demonstrates how light can be efficiently converted into stored energy through
carefully tuning intrarnolecular oxidation and reduction reactions with the excited state of a

molecule.

PZT-Ru(dmb)3*-AQ —> PZT'*‘-Ru(dmb)3-AQ“ (1.13)

23

C. Proton-Coupled Electron Transfer

1. Introduction

Protons can significantly affect the rates of biological electron transfer. Many
proteins and enzymes derive their function by mediating the rates of electron transfer by a
proton. Nowhere is this better demonstrated than in the active site of cytochrome c oxidase.
Although the distances separating the two heme centers from the binuclear Cu A centers are
similar, electron transfer to heme a is 102 - 10" faster than that to heme a3.l9 Proton transfer
accompanying heme reduction is believed to be the origin of the slow electron transfer to
heme a3 and accordingly the controlling factor for directional electron transport via heme
a.20 Such proton-coupled electron transfer (PCET) events continue to emerge in the

structure/function relations of a variety of other proteins and enzymes including

21-23 24.25 26.27
1,

Photosystem I nitrogenase, non-heme iron containing proteins, multicopper

8 and reductasesz9‘3' to name a few. Nevertheless, despite the importance of

oxidases,2
PCET in bioenergetic conversion processes of this diverse biological machinery, the

mechanistic details of how the electron couples to the proton remain largely undefined.
2. Photoinitiated ET in Semi-Synthetic Protein Mediated D/A Dyads

a. Histidine-33 Modified Cytochrome c

It was recognized long ago that ET reactions in semi-synthetic proteins could be
applied to study the effect of the coupling pathway within the protein. From these studies
Gray began to define the role of the protein structure in ET events in natural systems. Their
results stem from synthetically tailoring a class of cytrochrome c proteins with the redox
active pendent groups.32 The overall electron transfer reactions studied was an ET reaction
form a surface modified histidine group to the iron(III) heme irnbedded within the protein
interior. The net ET reaction is given below, Eq. 1.22. The reaction is initiated by the
reductive quenching of the surface AsRu(III)His-33 by a [Ru(bpy)3*]2+ present in solution.

Alternatively the reaction can be initiated using the pulse radiolysis technique, yielding

24

essentially the same unimolecular rate constant. The reaction rate for the ET reaction listed
in Eq. 1.22 is 50 milliseconds, as measured by laser flash photolysis, or by the pulse
radiolysis technique. The results of their experiments show that electrons can efficiently
tunnel through covalent bonding and hydrogen bonding present within the folded protein

structure, Figure 10 (a).

ASRu(ll)-(His-33)-Cyt C—Fe(lll)P ——) AsRu(lll)-(His-33)-Cyt C—Fe(ll)P (1.22)

b. Dipolar Efi‘ects in D/A modified a-Helical Polypeptides

Model helical peptide systems have been developed by Fox and coworkers to
examine the effects of helical dipole moments (11%) on the rate of ET, Figure 10 (b).33 ET
rates where measured to the excited state electron acceptor (A*) of pyrene (Pyr*), which is
reduced by a pendent electron donor (D), N ,N-dimethylarninobenzene (DMA), by the
mechanism outlined in Eq. 1.23. Owing to the design of the system, the ET could be
initiated along a pathway coincident and aginst the pennenant dipole of the helix. Measured
rates in a variety of solvents reveal km/km to be 27 in THF, showing that an intervening

dipole moment can significantly affect ET rates.

DMA—or-helix—Pyr* —————> DMA"‘—or-helix—Pyr"1 (1.23)

3. Biological Salt Bridge Proton-Coupled Electron Transfer in Sirohemereductase

An important example of PCET in nature is the action of a certain iron-sulfur
protein Sirohemereductase. Iron-sulfur proteins are well known for their ability to act as
electron carriers to supply the electrons required in biological transformations. However,
they also catalyze small molecule activation. Sirohemereductases, which encompass sulfite
reductase (SiRs) and nitrite reductase (N iRs), catalyze the conversion of sulfite to sulfide
(Eq. 1.24). The overall transformation requires the delivery of 6 protons and 6 electrons to
the substrate bound protein active site. In the duration of the overall reaction, no

intermediate oxidation products can detected free in solution. This is an indication that all

 

(a

v

A5Ru(ll)-(His-33)—Cyt C—Fe(|ll)P

v v

_ 0. w
DMA _ (RP-Nb

DMA Pyr*
kET— d kET+
O ”helix
Pyr* DMA

(b) DMA—a-helix—Pyr*

Figure 10. Hydrogen bonded protein mediated ET reactions in (a) a modified cytochrome
c with a surface AsRu(II)-His-33 group and (b) a functionalized a—helix in the peptide
sequence Boc-NH-Ala-Aib-Ala-(DMA or Pyr-Ala)-Ala—Aib-Ala—Ala—Aib-Ala-(Pyr or
DMA-Ala)-Aib-Ala-C02Et which sets up a helix dipole moment (tum).

26

intermediate reduction and protonation states of the substrate remain bound to the protein
active site, implying that the flow of protons and electrons into the protein active site is

somehow coupled.

8032' + 6H” + 69‘ ————> 82' + BHZO (1.24)

Displayed in Figure 11 is a schematic representation of SiRs, taken from a protein
crystal structure solved at a resolution 1.9 A.” The key features of the structure are
highlighted and clearly display the presence of an extensive array of salt bridge hydrogen
bonds which act to perform multiple tasks in the overall transformation, including anion
recognition, siroheme cofactor stabilization, and to create an access channel to shuttle

protons into the active site, while allowing water molecules to exit the active site.

4. Photoinitiated ET in Hydrogen Bonded D/A Dyads

Recent examples of synthetic dyad molecules to study ET reactions have been
extended to include naturally relevant hydrogen bonds. The basis for such systems is that
hydrogen bonds can be employed to mediate charge separation reactions as the spacer
group or linkage in fixed distance D/A ET. These systems emphasize that the spacer group
separating D and A need not be a covalent group but can be some non-covalent hydrogen
bond. Presented in Figure 10 are systems which fall into this category. In these D/A model
systems, the spacer group is some type of hydrogen bond. These studies are peaking
interests in a new wave of the ever changing field of ET research. The role of the proton
within the hydrogen bond is being defined in the context of bioenergetics and its role in

proton—coupled electron transfer (PCET) is being revealed.

a. Carboxylic Acid Dimer D/A Dyads
Nocera and coworkers have led the way in investigating electron transfer through

hydrogen bonds and have developed a research program that delves into establishing the

27

 

 

 

(a) = / Fe(lll) \
H ’1
l
y“
=. (D H H
H/“m. -~..I .......... H—N,
" (9N. 3:-..NH
0%Siogw ______
(b) l‘_\
/ Fe(||l)
/S——Fe/S
Fe—l-S/l
I/Fe-i78
S——Fe
H ’1
(:5 1’6) III 6:2-..H—N, g9
M. Gig/o G)
H ‘-. / \O@--.H.N \
\
H

Hm

Figure 11. A schematic representation of the active site of SiRs and MRS proteins. (a)
siro heme where a rectangle is used to represent the heme atomic framework; (b), substrate
(protonated sulfite) bound enzyme; (c) inhibitor (phosphate ion) bound enzyme.

28

underpinnings of PCET.35 The initial studies focused on ET reactions in the midst of a
carboxylic acid dimer. One such system is displayed in Figure 12 (a). In their system a
zinc porphyrin (ZnP) is funtionalized with an acetic acid group which forms a heteromeric
dimer with 3,4-dinitrobenzoic acid. In solvents of low dielectric constant, the carboxylic
acid dimer forms spontaneously. In dichloromethane, the association constant for the
dimerization is on the order of 500 M". Also formed in solution are nearly a statistical
distribution of homomeric dimers, which are ET inactive by design, and are not considered
in this discussion. Pulsed laser initiated kinetics measurements where performed to assess
the dynamics for ET in this system. The lifetimes for the charge separation and charge
recombination were assessed by time resolve transient absorption techniques. Fixed
distance ET reactions occur form the excited state of the zinc porphyrin to yield charge
separated state. For the 'H isomer charge separation occurs in 20 picoseconds, while
charge recombination (not shown) occurs in 100 picoseconds. Similar kinetics studies have
been performed where deuterium (2H) is exchanged into the carboxylic acid dimer. An
interesting kinetic isotope effect was observed (k,,/kD = 1.6) for the forward charge
separation step (Eq. 1.19) and similarly a kinetic isotope effect was observed for the charge

recombination step (kH/kD = 1.7).

ZnP*—(COZ"2H)2-DNB ——> ZnP'+‘-(CO,‘-"’H),—DNB"1 (1.19)

Therien and coworkers have assessed photoinitiated ET reaction in other carboxylic
acid dimers.36 In Figure 12 (b) is displayed a diporphyrin molecule where a benzoic acid
dimer substituted in the para position separates a zinc porphyrin by iron porphyrin. The
photoinitiated charge separation reaction, Eq. 1.20, is slower than the previous reaction
with a time constant of 114 picoseconds. This is in part due to the larger distance separating
D and A. Yet again this system demonstrates that coupling through hydrogen bonds is

feasible to enable efficient electron transfer.

ZnP*—(CO,H),_—Fe(lll)P ———> ZnP’*‘-(COZH)2—Fe(ll)P (1.20)

29

mg:

(b) ZnP-(p-PhCOzH)2-Fe(III)P

 

wary;

(c) Zn P—GC—BQ

Figure 12. Photoinitiated electron transfer in zinc porphyrin (ZnP) hydrogen bonded
complexes where the electron acceptor is a dinitro aromatic, ZnP-(COZH)2—DNB (a), an
iron(III) porphyrin, ZnP-(COZH)2—Fe(III)P (b) and benzoquinone, ZnP-GC—BQ (c).

30

b. ET in a Guanine-Cytosine Base Paired D/A Dyad

Sessler and coworkers have created an interesting ET assembly based on the
hydrogen bonding interactions taken from the natural guanine and cytosine base pairsf"7
Their system is an example of specific hydrogen bonding and molecular recognition.
Imparted into their system is the specificity of the guanine-cytosine (GC) base pairing
interactions, Figure 12 (c), where the carboxylic acid dimers interactions are nonspecific
and in solution there exists competing equilibria yielding dimers of D and dimers of A. The
ET reaction takes place between the excited state of the porphyrin and the benzoquinone
moiety , Eq. 1.21, with a time constant of 740 picoseconds. The longer time scale for the
reaction is in part due to the longer distance created by the GC spacer group. ET is again

efficient, and the presence of hydrogen bonds does not greatly purturb ET reaction rates.

ZnP*—GC—BQ ————> ZnP““—GC—BQ“’1 (1.21)

D. Thesis Objectives

1. Association of Amidinium with Carboxylate in Solution and in the Solid State

The salt bridge is presented here as a versatile building block in natural and non-
natural molecular architectures. In particular, as depicted in Figure 1, the amidinium-
carboxylate salt bridge stands out for its specificity in that it possess only one specific two
point hydrogen bonding mode relative to salt bridges based on guanidine and arginine.
Chapter 2 presents a study on the factors controlling the amidinium-carboxylate salt bridge
formation. Shown in Figure 13 (a) is the equilibrium expression for the formation of an
amidinium-carboxylate salt bridge in solution. In the solid state there are other points of
contact for maximizing all possible hydrogen bonding by forming additional one point
secondary hydrogen bonds, Figure 13 (b), not to be confused with secondary electrostatic

interactions. The factors governing hydrogen bond formation in the amidinium-carboxylate

31

salt bridge will be described. These include electronic, concentration, ionic strength,
counter ion, secondary hydrogen bonding and finally pH effects on the observed
equilibrium constant, K. In doing so, experimental challenges include the preparation and
isolation of the individual components of the salt bridge, namely the challenge of
synthesizing the individual molecules with either the amidinium or carboxylate functional

groups.

2. The Amidinium-Carboxylate Salt Bridge and Proton—Coupled Electron Transfer

With the amidinium-carboxylate salt bridge moiety understood, we will study
PCET by tethering a D/A system via the salt bridge, Figure 13 (c). A series of studies
aimed at exploring the coupling of the motion of an electron to the motion of the proton
through the synthesis of a series of novel compounds where electron donors and acceptors
are tethered together through the dynamic interactions of the amidinium functional group to
the carboxylate functional group. These compounds are presented as one possible way in
which the proton motion couples to electron motion through a highly polar salt bridge
created by the intrinsic electrical fields germane to groups mentioned above. In these
systems the positive charge of an amidinium group is the result of
protonation. Likewise the charge of the carboxylate group is the result of deprotonation.
How an electron transfer reaction proceeds in the presence of such a salt bridge is revealed
by electron transfer measurements. PCET reactions are initiated by either photoexcitation or
pulse radiolysis.

Chapters 3, 4 and 5 present measurements of ET rates in D/A systems modified
with the amidinium-carboxylate salt bridge, Figure 13 (c). Orientation of the salt bridge is
an important issue with regard to PCET as the proton is postulated to set up a potential
energy gradient that the electron must interact with on its journey from D to A, and also will

exhibit Franck-Condom barriers and electronic coupling barriers that are dependent on the

32

N _ ......
/
H H
H
\ .’
N—H ------ o
<' 6) G:>-—§
N—H ------ é
/ ‘\\
(b) ‘.
H
x /
9 ...... _
§-—<:e (+3".
{3 ------ H—N
x \
I ,H
H H
\ /
N— ------ O o ------ H—N
(C) D ‘: C9 6} A D (G) (9} A
— ------ o ------ -N
/ \
H H

Figure 13. The amidinium-carboxylate salt bridge formed by the equilibrium between an
amidinium and a carboxylate group to yield a primary two-point hydrogen bond (a);
stacking of amidinium-carboxylate salt bridges through secondary one-point hydrogen

bonding (b); D/A amidinium-carboxylate salt bridge complexes with two distinct
onentations of the salt bridge for the study of fixed distance ET (c)

33

position of the proton within the salt bridge interface. Represented in Figure 13 (c) are the
two distinct orientations of the amidinium-carboxylate salt bridge with respect to D and A.
Chapter 3 presents the results of photoinduced electron transfer measurements on
donor—(salt bridge)—acceptor complexes based on [Ru(bpy)3*]2". Laser flash excitation
prompts an electron to transfer from the reducing excited state of a derivatized Ru(H)
bipyridine donor complex to a dinitrobenzene acceptor juxtaposed by the salt bridge
interface in an oxidative electron transfer reaction, Eq. 1.25. The rate of electron transfer
through the D—(amidinium-carboxylate)—A salt bridge is compared to electron transfer in

the switched salt bidge complex, D—(carboxylate—amidinium)—A.

D*-salt bridge—A ———> D“‘-salt bridge— "‘ (1.25)

Laser flash excitation also prompts an electron to transfer to the excited state of
other derivatized Ru(II) bipyridine complexes. In this system the excited state of a
derivatized Ru(H) bipyridine complex acts as an electron acceptor to a dialkyarninoaromatic
donor compounds juxtaposed by the salt bridge interface in an reductive electron transfer

reaction, Eq. 1.26.

A*—salt bridge—D —-> A“‘—sa|t bridge—D’+1 (1.26)

Chapter 4 describes the preparation of porphyrins modified the amidine functional
group which undergo a reaction with carboxylic acids to form an amidinium—carboxylate
salt bridge. The synthesis of porphyrin amidines permits donor-acceptor complexes to be
juxtaposed by a zwitterionic hydrogen-bonded interface, formed upon the molecular
recognition of a carboxylic acid by the porphyrin amidine. It is demonstrated that the salt
bridge not only pre-organizes the donor-acceptor pair, but it also mediates electron transfer

from the photoexcitable Zn(II) porphyrin to a nitroaromatic acceptor, Eq. 1.25.

34

Chapter 5 presents a PCET system as an extension of the Closs and Miller systems,
and utilizes the pulse radiolysis technique to study ET reactions in radical anions of

aromatic groups linked by the amidinium-carboxylate salt bridge, Eq. 1.27.

D"‘-salt bridge—A ————> D—salt bridge—A"1 (1.27)

Chapter 2
SOLUTION AND SOLID STATE AMIDINIUM-CARBOXYLATES

A. Introduction

Innumerable types of hydrogen bonding motifs are continually being revealed in
nature where three dimensional structures are constructed from an assembly of hydrogen
bonds and covalent bonds. One particular type of hydrogen bond is the salt bridge. In
nature salt bridges act both as active catalysts and as passive structural motifs. More than
100 years ago Pinner isolated benzamidinium-benzoate salts in his pursuit of a synthetic
methodology toward the amidinium functional group.38

Since Pinner’s pioneering work, a series of benzamidinium-benzoates have been
prepared by Krechl and coworkers, Eq 2.1, where the para substituent, X, is H, NHZ‘ t-
butyl, CGHS, Br, and N02.” They found that the NMR chemical shifts in solution could be

correlated to electronic contributions from either the benzamidinium or the benzoate

H
\

H\
N-H ----- o N-H ----- o
N-H------o N—H ------ o
H/

/
H

groups across the salt bridge in DMSO-d6 solution. Yet in DZO there is no influence on the
NMR chemical shifts. From their results it was concluded that the salt bridge fromed by the
1:] interaction of benzamidinium with the benzoate group is formed readily in DMSO-d6

solution but not in DZO. The factors that govern the equilibrium to form a salt bridge

35

36

between the benzamidinium group and benzoate group in either solution or in the solid state
have only begun to be uncovered.

Before initiating a study of the factors involved in the formation of the
benzamidinium-benzoate salt bridge, it is necessary to isolate pure salts of both the
benzamidinium group and the benzoate groups. General methodologies are described here
that allow for the convenient isolation and purification of large quantities of benzamidinium
chloride salts and tetrabutylammonium benzoate salts. With these compounds in hand, it is
possible to assess the equilbria in solution through the formation of a strong two point
hydrogen bonding interaction.

In solution, the amidinium group of a para substituted benzamidinium (Am)
spontaneously interacts with the carboxylate group of benzoate (Ca) to form the
benzamidinium-benzoate salt bridge (AmCa) as described by the equilibrium depicted in
Eq. 2.2. By carefully changing the relative concentration of amidinium and carboxylate, the

equilibrium constant, K, describing the equilibrium between Am and Ca forming the AmCa

Heq K Heq
8‘11»: 9 N-Hu‘"?
N-H... 0 N~Hgm0
I /
H,q Ha,
Am Ca AmCa

salt bridge may be assessed. Factors that influence the equilibrium to form the AmCa salt
bridge include the electronic structure of the benzamidinium portion of the salt bridge
system, the solution pH, and the ionic strength. At high ratios of Ca to Am, secondary
hydrogen bonding interactions are observed. For the equilibrium displayed in Eq. 2.2,
where X is hydrogen, methyl, phenyl or trifluoromethyl group an interesting substituent

effect on the observed K is observed and may be rationalized in terms of a standard

37

Hammett equation relating changes in K to the electronic structure of the para
benzamidinium substituent.

Solution studies of the salt bridge may be extended to the solid state. The molecule
3-amdininium-benzoate, shown in Eq. 2.3, combines both the carboxylate and amidinium

groups into one molecule and is synthesised and structurally characterized by

   

9 ”\N.’H : ”\NfH.
9’. L9 -———> . .3 '19 3'
O" ‘~N,H I -. .' ‘T/ (2.3)
1‘. i 1,.
solution solid state

crystallographic techniques for the first time. Its solid state structure provides a link to the
observed solution primary two point and secondary one point hydrogen bonding modes in
the benzamidinium-benzoate system. The data presented here, from solution to the solid
state, shows conclusively both the weaker secondary hydrogen bonding interactions and
the strong primary two point hydrogen bonding interactions in both solution and the solid
state. These results advance our understanding of the solution behavior of the arnidinium-
carboxylate salt-bridge as well as the factors that contribute to stabilizing this interesting
structural motif in solution. The link between the solution and solid state salt bridge
structures are clear in terms of a propensity of the system to achieve a closed set of
hydrogen bonding interactions. In this closed set, the carboxylate group yearns to donate 4
lone pairs of electrons to 4 hydrogen bonding protons donated from the nitrogens of the
amidinium group. Similarly, the amidinium group yearns to donate all 4 of its protons to 4
lone pairs of electrons. The formation of this closed set of hydrogen bonds in 3-amidinium—
benzoate is represented in Eq. 2.3. In the solid state, there appears to exist a propensity to

achieve closure inasmuch as one lone pair of electrons is donated to each proton.

38

B. Synthesis

1. Tetrabutylammonium Benzoate

The carboxylate form of acids are conventionally obtained by neutralization of the
acid proton with strong base. Often sodium carboxylate salts of common carboxylic acids
are commercially available. The drawback of sodium carboxylate salts is their poor
solubility in organic solvents, thus they do not pose as viable compounds for solutions
studies of salt bridges involving the introduction of carboxylates into aprotic organic
solvents. However the tetrabutylammonium ion is exceedingly soluble in organic solvents,
as it is commonly employed in phase transfer catalysts. For our studies we wish to examine
salt bridge formation in non-aqueous aprotic solvents; therefore, we require the carboxylate
form of carboxylic acid as its tetrabutylammonium salt.

Treatment of benzoic acid with an equivalent of tetrabutylammonium hydroxide in
methanol yields an equivalent each of water and the tetrabutylammonium benzoate, as
summarized by the reaction shown in Figure 14 (a). Removal of the methanol under
reduced pressure yields a tarry residue, which is dissolved in freshly distilled benzene and
treated with basic alumina, as described by Kemp.40 Filtration to remove the basic alumina
followed by freeze drying the solution, yields tetrabutylammonium benzoate in its

anhydrous cyrstalline form.

2. Benzamidinium Chloride

Classically, rather harsh and cumbersome conditions are required for the
preparation of amidinium chloride salts from nitriles. Dry hydrochloric acid consumes a
nitrile in dry alcohol to yield the corresponding imidate ester as a hydrochloride salt.
Addition of dry ammonia to the imidate ester hydrochloride salt affords the amidinium
chloride salt as developed by Pinner at the end of last century. Conversely, as shown in

Figure 14 (b), base catalyzed addition of alcohol to para-substituted benzonitriles generates

39

(a) "7H3 _’ WCEV W

(b) X—QCEN —> x‘©—%:—H —‘> x©——<;>-H cue

/
H 3C H,

H
1.-.. J
(C) X “2C9 09 WNW Q

H
H

\

N-H ----- o
N—H ------ o
x

1 \ ’H
(d) c”/~ 91 9 N
O Q. :9
CH3 I
H

Figure 14. Synthetic reaction sequence employed for the preparation of (a)
tetrabutylammonium benzoate by treatment of benzoic acid with tetrabutylammonium
hydroxide; (b) para-substituted benzamidinium chloride salts by reaction of para substituted
benzonitriles with sodium methoxide followed by ammonium chloride; (0) benzamidinium-
benzoate salt bridges by precipitation from water; and (d) the preparation of 3-amidinium-
benzoate by reaction of 3-cyanobenzoic acid with sodium methoxide, tetrabutylammonium
hydroxide and crystallization from water (d).

40

free base imidate esters (bracketed in Figure 14 (a)). Addition of ammonium chloride to the

free base imidate ester yields para-substituted benzamidinium chloride salts.

3. Benzamidinium-Benzoate

With both tetrabutylammonium benzoate and para-substituted benzamidinium
chloride salts in hand it is possible to obtain benzamidinium-benzoate salt bridges free of
extraneous counter ions by spontaneous assembly in aqueous solution, Figure 14 (c).
When an aqueous solution of an amidinium chloride salt is combined with an equivalent of
a tetrabutylammonium carboxylate salt, a precipitate forms. The 1H and 13C NMR as well
as mass spectroscopy confirm the presence of both the amidinium and carboxylate portions
of the salt bridge. Integration of the peaks in the 1H NMR spectrum reveal the
stoichiometry of the amidinium to carboxylate is 1:1. This reaction sequence is similar to a
procedure developed where sodium benzoate salts are used instead of tetrabutylammonium

benzoate salts.

4. 3-Amidinium-benzoate

An interesting molecule can be prepared that contains both the amidinium and
carboxylate functional groups in one molecule. This molecule 3-arnidinium-benzoate is
prepared from the corresponding 3-cyanobenzoic acid as described in Figure 14 (d). This
synthesis essentially combines both steps shown in Figure 14 (a), (b) and (c) into one. An
excess of sodium methoxide performs dual duty, first to deprotonate the benzoic acid to
yield the 3-cyanobenzoate and then to catalyze the transformation of the nitrile group to the
free base imidate ester intermediate, shown as a bracketed intermediate in Figure 14 ((1).
Addition of an excess of ammonium chloride yields an intractable white solid, which
crystallizes from hot water to yield large prismatic crystals suitable for x-ray

crystallography.

41

C. Equilibrium Measurement via the NMR Technique

Displayed in Eq. 2.2 is the equilibrium for a para substituted benzarnidiniun, Am
with benzoate, Ca to form the benzamidinium-benzoate salt bridge, AmCa, in solution. The
equilibrium constant, K, for the equilibrium depicted in Eq. 2.2 is defined in Eq. 2.4 as a
function of the equilibrium concentrations of Am, Ca, and AmCa, denoted as [Am], [Ca],
and [AmCa] respectively. Simple equilibria of this type have an exact mathematical solution

derived from solving a quadratic formula relating the change in initial non-equilibrium

_ [AmCa]

— [Am][Ca] (24)

concentrations for Am and Ca, designated by [Am]0 and [Ca]0 respectively, upon achieving
some equilibrium concentration of the AmCa salt bridge complex, [AmCa], as govemed by
K. The solution to this quadratic formula can be derived to define [AmCa] as expressed in
Eq. 2.5 in terms of the initial non-equilibrium concentration values of [Am]0 and [Ca]o, as
determined by preparing standard stock solutions of either benzamidinium chloride or

tetrabutylammonium benzoate employing the usual analytical volumetric techniques and K.

 

[AmCa] = g[% + [Am]o + [Ca]0 - J (112 + [Am]o + [Ca]0 )2 _ 4[Am]o[Ca]o:l (2.5)

The 'H NMR spectrum provides chemical shift information for freely solvated Am
and the salt bridge complex AmCa. The chemical shifts for the N-Hu, the protons involved
in the primary two point salt bridge hydrogen bonding motif are designated 8A,“ and 6mm,
respectively, and are separated in the lH NMR spectrum by some difference in chemical

shift, A5 as defined by the expression shown in Eq. 2.6. Next the observed

max ’

carboxylate-concentration-dependent chemical shift, 8 for N-H”, is defined as the

obs’

weighted average of 8A",Cal for the AmCa salt bridge form of N-Hu, and 8A,“ for Am, the

non-salt bridge forms of N-Hax. This is due to the exchange rate of the salt bridge being

42

faster then the time scale for the lH NMR experiment. Finally, upon substitution of Eq 2.5
into Eq 2.7, Eq. 2.8 is obtained, where 501,. is defined in terms of one dependent variable,
K. The chemical shift dependence of N-Hax is evident experimentally as shown in Figure
15 where displayed are lH NMR spectra for benzamidinium chloride and benzarninium

chloride with benzoic acid and varying amounts of tetrabutylammonium benzoate.

 

Asmax = 5AmCa - 8Am (2‘6)
AmCa]
8 = 5 [ -A5 2.7
obs Am + [Amlo max ( )

 

 

2

50,, = 5,," + £230 [% + [Am]o + [Ca]o — J(% + [Am]o + [6.110) — 4[Am]o[Ca]oJ (2.8)

In the 1H NMR spectrum of benzamidinium chloride both N-Hu and N-ch appear
as a single broad peak centered at 9.3 ppm, the bottom spectrum in Figure 15. Integration
of the area of this peak corresponds to the expected 4 protons. Addition of benzoic acid to a
solution of benzaminidinium chloride transforms the broad peak into two resolvable peaks
corresponding to N-Hax and N-Heq, second spectrum from the bottom of Figure 15, where
the integral region of each peak is equivalent to two protons each. A small effect is
observed on the 50b. for N-Hax and N-Heq, but they essentially remain centered at 9.3 ppm.
The change in the appearence of the line shapes for these two peaks in the presence of
benzoic acid reflects the change in the relaxation dynamics of the nuclear spin of the
protons corresponding to N-Hu and N-ch. Effects such as this are well known for altering
the media surronding a spin system such as by changing the acidity of the solution. The
chemcial exchange of the N-H protons with the acidic proton of the carboxylic acid group
may also contribute to the appearence of the peaks corresponding to N-Hu and N —Heq in the
IH NMR spectrum.

43

 

 

 

 

I I I I I
12.0 11.0 10.0 9.0 8.0 7.0

PM“

Figure 15. 'H NMR spectrum of benzamidinium chloride at a concentrtion 5.2 mM in
DMSO-d6 (bottom). The 1H resonances of the aromatic ring appear between 7.5 and 8.0
ppm. The amidinium protons appear as a broad singlet at 9.3 ppm, but separate and
sharpen to two singlets in the presence of 16.6 mM benzoic acid (second from the bottom).
Selected spectra are shown upon addition of the tetrabutylamonium benzoate at
concentrations 0.0, 0.0, 1.5, 2.3, 3.0, and 5.0 mM (bottom to top).

3.5

 

3.0 —

2.5 —

2.0 —

1.5—

A5obs I ppm

1.0~

0.5 —

 

 

Mod
0.0 5.0 10.0 15.0 20.0 25.0

 

Benzoate / mM

Figure 16. Plot of the change in chemical shift of the benzamidinium N -Hax proton versus
the concentration of benzoate from 0.0 to 25 mM. The concentration of benzamidinium
chloride is 5.5 mM, and the concentration of benzoic acid is 16.6 mM throughout all
concentrations of benzoate (as the tetrabutylamonium salt).

45

 

 

 

 

0.4
0.3 7
(U
E
< 0.2 ~
X
0.1 ‘
00 I I I Q
0 0.25 0.5 0.75 1

Figure 17. Job’s plot of the mole fraction of the benzamidinium-benzoate salt bridge
complex vs. the mole fraction of benzamidinium (6.3 mM) as the concentration of benzoate
is varied from 0.0 to 25 mM

46

Addition of tetrabutylammonium benzoate to the solution containing benzamindium
chloride and benzoic acid has large effects on 501,. for both N-Ham and N-H“l of the
amidinium group of benazminidium chloride. The large downfield changes in 80,,s for N-

H over 3.0 ppm is a direct indication of AmCa salt bridge formation as shown in Eq.

2.2. Likewise the N -Heq experiences a slight upfield shift of around 0.5 ppm. The chemical
shift information, in conjuction with the concentration information, is used in Eq. 2.8 to fit
the data to K using a non-linear least squares method. Plotted in Figure 16 is 50,,s as a
fucntion of the benzoate concentration. A fit to Eq. 2.8 is shown in Figure 16 for which a
K of 2500 M "' is obtained. The fit gives a good K for the entire concentration range
studied for this systems. The stoichiometry of the AmCa salt bridge is confirmed by a Job
plot Figure 17. In the Job plot the method of continuos variation is employed to reveal that
the mole fraction of the AmCa has is optimum when the mole fraction of Am is 0.5. That is
the concentration ratio of Am and Ca that yields the optimum mole fraction AmCa in
solution.

There are limitations to this technique, primarily concentration at which one works.
Also for the titration procedure a great limitation is the presence of spectator ions. For each
amidinium group there will be a chloride ion in solution. Likewise for each carboxylate
group there will be a tetrabutylarmnonium ion present in solution. The affects of these
groups is accounted for by a separate ion dependence on the association constant study.
The observed binding constant is also dependent on the concentration of acid present in
solution. This can be accounted for by measuring K at different concentrations of benzoic
acid while at a constant concentration of benazamidinium chloride. Conversely the affects
of chloride ion on K can be assessed by measuring K at different concentrations of

benzarnidinim chloride while maintaining constant concentration of benzoic acid. The

results of these experiments are presented in the proceeding sections of this chapter.

47

D. Effects on Salt Bridge Equilibrium Constant

1. Salt Bridge Hammett Effects from para Substituted Bemzamidiniums

Recently it was shown by Wilcox that solution hydrogen bonding interactions
between N-arylureas, with substituted in the para—position, and sulphonates display a
strong correlation between the observed association and the o” Hammett parameter.
Furthermore, calculated surface electric potentials are shown to effectively predict the
observed association energy within the same series of compounds. Thus the correlation
between the electrostatic nature of hydrogen bonding and observed association energies is
established. These results reinforce the power of the Hammett parameter in predicting
electronic effects on chemical equilibria. We thought that the electrostatic nature of the
amdinium-carboxylate salt bridge was a logical next step to pursue the electrostatic effects
on solution hydrogen bond formation.

Presented here is a similar Hammett study on K, Eq. 2.02, as a function of the
electron donating and electron withdrawing groups of para substituted benzamidinium
group. Listed in Table 1 are the substituents used in this study along with the
corresponding Hammett 6pm parameters and the observed K for each substituent. The
results of this study reveal that the electronic structure of the molecule is important in
controlling the primary two point hydrogen bond formed in the amidinium-carboxylate salt
bridge. These effects are due to polarization of the para substituted benzamidinium group
due to the electrostatic nature of the salt bridge hydrogen bond. Changes in the electronic
structure are manifested in changes in polarity of the hydrogen bonding group, in this case
the Hammett parameters are adjusted for modifying the electronic structure about the
amidinium group. However the a benzoate group can be nearly equally perturbed as shown

by results aimed at exploring the simple chemical shift data.

48

Table l. Hammett effects on

the association constants of para—substituted
benzamidiniums (X) with benzoate.

 

 

X om“ Kb
CH3 on 1490
H 0.00 2504
C6H5 0.05 2846
CF, 0.53 13253

 

aValues for (Spam are from March, J. Advanced Organic Chemistry, 4th Edition J. Wiley

and Sons: New York, 1992, p. 280. bIn DMSO-d6 at 25.0 °C. the error in values of K
are :I: 15%.

49

1.0

 

    

 

|09(KIK,.) = 96
p = 1.40

 

 

 

 

 

 

-{12 (10 (L2 CL4 (16

para

Figure 18. Hammett plot for K and para substituted benzamidiniums.

50

The effects can be explained in terms of simple electron donor/acceptor interactions.
The donation of electron density into the amidinium group results in stronger N-H bonds
but decreases the relative charge on the H. This weakening of the charge results in a direct
weakening of the electrostatic interaction of the salt bridge. Conversely, removing electron
density by the inductive action of the trifluoromethyl groups weakens the N-H bonds, and
makes the charge on the H higher thereby directly increasing the electrostatic attraction to
the carboxylate moiety, yielding an overall increase in observed K.

These results are summarized in Figure 18 where the slope of the plot of the log of
K versus 6p yields p. The value of 1.4 for p, obtained in the fit, reflects the strong
electrostatic nature of the hydrogen bonding interation within the amidnium-carboxylate salt

bridge.

2. Effect of Benzoic Acid Concentration on the Solution Salt Bridge Equilibrium

The effects on the binding constant are not limited to the electronic effects. Due to
the polar nature of the salt bridge other factors are involved in controlling the binding such
as the addition of other hydrogen bonding groups. Especially important is the effect of
adding excess benzoic acid to solutions due to the fact that the protons in the salt bridge are
more easily observed when the conjugate acid is present at some concentration.

Figure 19 is the result of three separate titration experiments where the
concentration of chloride and amidinium is constant at 5.2 mM and the concentration of
benzoic acid is constant at 10, 50 and 100 mM. The fit of the data is arbitrary and is in
place only as a visual aid. The results show that there is a sizable decrease in the observed
K with increasing benzoic acid concentration. This is due to several factors. First the
increase in the polarity of the solution will change the dielectric properties of the solution,
directly decreasing the electrostatic interaction. The carboxylic acid is arnphoteric in that it

can compete for hydrogen bonding to the amidinium group through one and two point

51

 

2500 ~ 0

2000 —

1500 -

mm1

1000 —

500 4

 

 

O ' I . I ' I ' I ' I . I
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14

 

Benzoic acid I M

Figure 19. The effect of benzoic acid concentration on K. The chloride concentration is
held constant at 0.005 M.

52

hydrogen bonds. Finally by increasing the pH of the solution will also affect the hydrogen

bonding interactions.

3. Effect of Chloride Ion Concentration on the Salt Bridge Equilibrium

The ionic strength of a solution affects equilibria present in that solution. The
interaction energy of the amidinium-carboxylate salt bridge, determined by Coulombs’s
law, is directly effected by the ionic strength of the solution as described by the Debye-
Huckle equation. Therefore, the association constant for the formation of salt bridges
should also depend on the ionic strength of the solution. This quantitiy can be
systematically varied with the addition of extraneous counter ions by changing the initial
concentration of the amidinium chloride in a particaular titration experiment, or by the
addition of extraneous ions in the form of tetrabutylammonium chloride.

As shown in Figure 20, there are large effects on the observed binding constant
with ions present. These effects suggest that the lone pair electrons and protons that interact
with the solvent give a direct pathway to tuning the electronic interactions in the salt bridge
by the solvent. Added chloride ions, which can weakly donate 4 lone pair electrons and has
a net negative charge, compete with carboxylate to interact with the positively charged
amidinium group. This interaction need not be bonding. Furthermore the chloride can
perturb the system by simple non-specific ion-ion interactions, thereby scrambling the
electrostatic interactions between net ion partners. These non-specific ionic interactions

effectively lower the observed binding constant for the observed salt bridge interaction.

4. Carboxylate Concentration Effects

The preceding discussion has focused upon the interesting concentration dependent
chemical shifts at low concentrations, and the effects on the equilibrium constant for the
formation of the amidinium-carboxylate salt bridge. However, when the concentrations of
benzoate is raised far above and beyond the concentrations used in assessing K in the

primary hydrogen bonding mode expressed by Eq. 2.2, it is evident that secondary

53

 

1 200

1000 7

800 4

600 —

KIM"1

400 —

   

tetrabu lammonium
200 — c loride ——>

 

 

 

O T"l"'l"'l"'TII'TIITTT"
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14

Chloride I M

Figure 20. Affect of chloride ion concentration on the observed binding constant (K).
The concentration of benzoic acid is constant 0.050 M for each measurement of K.

54

hydrogen bonding equilibria are present, though with much lower relative equilibrium
constants relative to the primary two point hydrogen bonding salt bridge interaction.

Figure 21 displays the concentration dependence of the amidinium protons at 63.0
mM over a concentration range of carboxylate from 0.0 to 864.0 mM with the carboxylic
acid concentration constant at 226.0 mM. The early concentration dependence fits to the
equilibrium expression shown in Eq. 2.2, where a K of 125 M‘l is obtained. This primary
hydrogen bonding mode fits well to the equation below 5 equivalents, but not at higher
concentrations. This suggests that secondary equilibria are also present. In particular, the
concentration dependence of the N Heq becomes obvious when the concentration of
carboxylate approaches 10 equivalents or more. This suggest an association constant
weaker then the primary mode yet observable nonetheless.

A model to explain the behavior of N ch at concentrations of benzoate near
saturation is outlined in Figure 22. First, the primary two point hydrogen bonds form, then
in a stepwise manner, a second and a third benzoate associate with the two NHeq’s to form
a total of four hydrogen bonds to the benzamidinium group through the two additional
weaker one point hydrogen bonds. This creates a net closed set of hydrogen bonds to the

amidinium group as put forth by Etter and Ward.

E. Solid State Structure of 3-Amidinium-benzoate

Evidence for the weaker secondary hydrogen bonds in the amidinium-carboxylate
salt bridge observed in the benazamindinium-benzoate salt bridge in the presence of
exceedingly high concentrations of benzoate is revealed in the examination of the solid
state structure of 3-amidinium-benzoate. The construction of solids by self-assembly
requires that the noncovalent interactions of structural subunits, referred to as
suprarnolecular synthons,"l be controlled in an exact and predictable manner. Hydrogen-

bonding, dipole, hydrophobic, aromatic -stacking and van der Waals forces are some of

55

 

3.5

2.5‘ e

2.0“

1.57

Afiobs / ppm

 

 

 

.0.5 I ' ' I ' ' ' I ‘ ' ' I ' ' ' I
0.0 0.2 0.4 0.6 0.8 1.0
Benzoate / M

Figure 21. Change in chemical shift of benzamidinium NH” (O) and NH“. (A) with the
addition of benzoate 0 to 886 mM.

56

 

 

e>—I,
H (5 H\
N—H , N-H ------ C?
,5, l__ 4., 9%
Iii—H /N-H ------ o
H H
1 _
O 7 2.
a. Q
.’ e‘>——§
O "
Q o Q~.
d / " H
o ------ H—N o _____ _N/
—-——> (’ 9 ® W ____> ,1 \\
<——— , .’ +___ < 9 G) ,’
o ------ H—N ‘~ .
\H o ------ H—N\
.H
9'
efi>—1’
6

Figure 22. Proposed mechanism for stepwise association of three carboxylates to one
amidinium, yielding a net four hydrogen bonds to the amidinium in solution.

57

the noncovalent intermolecular interactions that have been exploited to self-assemble
organic molecules into solids.42 Though less common, electrostatic forces are emerging as
a particularly important noncovalent interaction in assembling subunits in extended
arrays.43 The nondirectionality of the electrostatic interaction, the inability to de-solvate
charged species from polar environments, and the need for counterions to maintain charge
neutralization in the crystalline environment is overcome when the synthon is a salt bridge.
The pioneering work of Etter and Ward44 and others ”’46 has shown that salt bridges can be
used to assemble tapes and layers. Yet the ability to set the dimensionality between tapes or
layers can be difficult. We wondered whether the control offered by a salt bridge in
establishing the three-dimensional structure of these biological systems would be useful for
the design of the extended crystalline arrays. Our approach centered on the amidinium-
carboxylate salt bridge as a synthon for self—assembly. We now report the three
dimensional structure of 3—amidinium benzoate (1), Figure 23, and show that Whiteside’s
recent prediction47 is correct; namely, favorable secondary electrostatic interactions of
synthons with small substituents are important in establishing the three-dimensional motif
of crystalline solids.

Figure 24 shows the bc plane of the crystal structure of 15""49 This view highlights
the primary interaction of the two-point hydrogen bond of the amidinium-carboxylate salt
bridge. Two favorable secondary electrostatic interactions” confer significant stability on
the interface“ as indicated by the relatively short N to O distances of 2.766(1) and
2.790(2) A and by the ability to self assemble the interface from water. A zig—zag tape
pattern results from the catenation of the salt bridges along chains. The tapes align in a
head-to-head arrangement along the c-axis of the unit cell and pack together at a van der
Waals contact distance to create the sheets of the be plane. The repeat distance between
positional atoms of neighboring tapes is 6.93 A thereby establishing the b-dimension of the

unit cell. Whereas the phenyl rings lie in the be plane, a rotation of the amidinium group by

58

30.0(1)° from the plane of the aromatic ring is complemented by a 23.3(l)° counter-rotation
of the carboxylate group. This 7° difference in rotation of the two groups results in a slight
twist of the amidinium-carboxylate interface (dihedral angle of 8.41° for the planes defined
by the CN2 and CO2 atoms of the salt bridge) with respect to the bc plane.

The canting of the salt bridges permits the formation of secondary hydrogen
bonding interaction, which extends the dimensionality of the structure beyond the sheets of
zig-zag tapes, shown in Figure 24. The amidinium protons outside the salt bridge form
hydrogen bonds to carboxylate oxygens of tapes from adjacent sheets above and below the
bc plane. This secondary interaction generates ladders as shown in Figure 25, where the
crystal structure has been rotated such that one of the ladders is projected in the plane of the
paper. The hydrogen-bonding arrangement of the ladder necessarily enforces the
amidinium—carboxylate salt bridges in neighboring sheets to arrange such that their dipoles
are opposed, thereby setting the head-to-tail orientation of tapes along the ladder. This
orientation has the added benefit of satisfying the dominant dipole interaction within the 3-
D structure. The dipoles of salt bridges above and below a plane in the ladder structure are
at a distance of 3 A whereas the dipoles within the bc plane are at a distance of 7 A.
Consequently, the maximum stabilization of the dipoles arising from the Coulombic
potential is derived from interlayer salt bridges in the head-to-tail arrangement.

The crystal structure of 1 in the ac plane (Figure 24) shows that ladders intersect the
bc plane at an oblique angle to interconnect the intralayer planes of zig-zag tapes within
sheets to form a layered structure with a d-spacing of 3.16 A. The ladders are parallel to
each other but alternately tilt in and out of the ac plane owing to the rotation of the
amidinium and carboxylate groups with respect to the aromatic rings (vide supra). Though
the one-point hydrogen bond is slightly weaker than the two point interaction, as evidenced
by the longer N to o distances of 2.878(5) A and 2.904(6) A, it composes the ladders that

radiate from the aromatic subunits to provide the scaffolding to support the layer structure,

59

 

Figure 23. Structure of 3-aminidinium-benzoate; oxygen (red), nitrogen (blue), carbon
(light gray) and hydgrogen (dark gray). The aromtic hydrogens are omitted for clarity.

60

 

Figure 24. The packing of the zig-zag tapes in the bc plane of the crystal strucutre of l.
The amidinium-carboxylate salt brdgie is roatated 26.1° (plane defined by the oxygen and
nitrogen atoms of the salt bridge) out of the plane of the aromatic ring.

61

 

Figure 25. View of the interlayer one-point secondary hydrogen bonding network. Each
subunit of l is connected to four other distinct subunits of 1 via the secondary one-point
hydrogen bonding interaction (two above and two below with respect to the the layers of
two-point primary interactions displayed in Figure 24).

62

 

Figure 26. Rotation of the crystal structure of 1 such that the ladder is viewed in the
plane of the paper. The ladder structure is formed from the hydrogen bonding between
protons external to the salt bridge of a given tape and the carboxylate oxygens of salt
bridges in neighboring interlayers.

63

Table 2. Primary hydrogen bond distances for 1.

 

 

Atoml Atom2 Distance
0(1) N (2) 2.766(1)“
0( 1) H(2A) 1.779b
0(2) N(1) 2.790(2)a
0(2) H(lA) 1.825”

 

3Numbers in parentheses are estimated standard deviations in the least significant digits.
aDetermined from difference maps used to calculate the position of the hydrogen.

Table 3. Secondary one-point hydrogen bond distances for l.

 

 

Atoml Atom2 Distancea
0(1) N(2) 2.878(5)“
0(1) H(2B) 2.011b
0(2) N(1) 2.904(6)“
0(2) H(lB) 2.104b

 

“Numbers in parentheses are estimated standard deviations in the least significant digits.
a‘Deterrnined from difference maps used to calculate the position of the hydrogen.

64

which is further stabilized by the van der Waals interactions of the aromatic rings from
neighboring sheets.

The most successful approaches to crystal engineering take advantage of the highly
directional and relatively strong nature of the hydrogen bond. In the crystal structure of l ,
the full complement of four protons and four lone pairs of the amidinium-carboxylate salt
bridge creates closure of the hydrogen bonding network. Two protons internal to the salt
bridge establish a primary interaction with two lone pairs of an individual carboxylate to
comprise zig-zag tapes that are analogous to that of isophthalic acid.52 Here the two point
intermolecular hydrogen bond is established with a carboxylic acid dimer. However, in this
structure, there is no secondary interaction to connect the layers owing to the absence of
protons external to the hydrogen bonding interface. For the amidinium-carboxylate salt
bridge of l, the two external protons of the salt bridge are complemented by two lone pairs
of oxygens from adjacent interlayer carboxylates to form the common ladder structure
characteristic of benzamidinium-carboxylate bridges.53 The one-point hydrogen bond of the
salt bridge is strong with respect to other non-covalent secondary interactions used for
traditional crystal engineering and it is the important interaction that establishes the three-
dimensional motif of the crystal. The results reported here thus further establish the

importance of salt bridges as useful synthons in crystal engineering.

F. Experimental

Synthesis of 3-amidinium-benzoate The intermediate imidate ester
carboxylate is prepared in situ by adding an excess of freshly generated sodium methoxide
to a methanol solution of 3-cyanobenzoic acid, under an argon atmosphere. An excess of
ammonium chloride is added after 24 h of stirring, at ambient temperature. The product
was filtered and washed several times with water, methanol and ether. The proton NMR
spectrum (D20, 300 MHz) shows the appropriate peaks for the unsymetn'cally substituted
meta aromatic ring: 57.67 (1H, t), 7.89 (1H, d), 8.18 (doublet, 1H) and 8.20 ppm (1H,

65

s). Calcd. for: C8H8N202: C, 58.48%; H, 4.87%; N, 17.05%. Found: C, 57.55%; H,
4.88%; N, 16.18%.

Chapter 3
Ruthenium(ll)tris(2,2’-bipyridine) Complexes to Probe PCET

A. Introduction

Pioneering work of Meyer and coworkers showed that photoexcited
ruthenium(II)tris-2,2’-bipyridine ([Ru(bpy)3*]*2) effeciently undergoes ET reactions in
solution via oxidative or reductive quenching. [Ru(bpy)3"‘]“2 in the presence of
nitroaromoatic groups (ArNOz) undergoes an oxidative ET reaction as shown in Eq. 3.1.
In the presence of dialkylaminoaromatics (ArNRz), [Ru(bpy)_.,*]+2 undregoes a reductive ET
reaction as described by Eq. 3.2. Based on these systems, salt bridge complexes for the

study of proton-coupled electron transfer (PCET) have been developed.

[Ru(bpy)3*]+2 + ArNO2 ——> [F1u(bpy)3]‘“3 + ArNOz"1 (3.1)

[Ru(bpy)3*]*2 + ArNR2 —-> [Ru(bpy)3]’*‘ + ArNth'” (3.2)

Three bipyridine ligands for this study are displayed in 3.3. These bidentate ligands
are distinguished for their electron donating and withdrawing capabilities. The bpy ligand
acts a benchmark with no electron donating or withdrawing functional groups attached to it,
while the trnbpy ligand, a bipyridine appended with four methy groups, is relatively
electron rich. The decb ligand has two ester functional groups which act to withdraw
electron density from its 1: system rendering it relatively electron deficient. These ligands
are ancillary to the salt bridge and can be used ot establish the proper electronics for PCET

studies.

66

67

   

/\
/ O
N\ |

(3.3)

N/
\ I O

O\/

bpy tmbpy decb

The salt bridge is established with the ligands displayed in 3.4. The Mebpy-amH+
and Mebpy-COO' ligands can establish the formation of the amidinium-carboxylate salt
bridge moiety; the known Mebpy-COOH bipyridine ligand funcitonalized with the
carboxylic acid group is capable of forming symmetrical carboxylic acid dimers and also is

the synthetic precursor to Mebpy-COO‘.

   

/
N\ I
(3.4)
/
I o
\
O\
H
Mebpy-amH” Mebpy-COO‘ Mebpy-COOH

The 4’-methyl-2,2’-bipyridine (Mebpy) ligand with amidine or carboxylate
functionalities is derived from the aldehyde, which may be converted to the nitrile
following Olah’s procedure54 in good yields. Conversion of the aldehyde to the carboxylic
acid is accomplished under standard oxidation conditions.73 The amidine group is afforded
by adapting Garigipati’s method” of reacting nitrile with Weinreb’s amide transfer reagent,
methyl aluminum(III) chloroarnide, In our hands, the amidinium chloride is directly

obtained from this conversion, Eq. 3.5.

68

'- ‘1

 

 

[CH3
Cl—Al H\
NHz N H20 N-H
<3'C2N All ——> H ———> ((9 (2| 9 (3.5)
0" ‘CHs NH2 N—H
_ J H,

Alternatively, for the polypyridine modification, we have obtained 4-methyl-4'-
amidinium-2,2'-bipyridine (Mebpy-amH*) chloride in higher yields by the base-catalyzed
reaction of the nitrile with methoxide to afford the imidate ester, which smoothly reacts

with ammonium chloride56 as shown in Eq. 3.6.

 

H
N-H NH4CI ‘N-H
(VCEN ————> [ V ———> H19 Cle (3.6)
,0 ,-H
H3C H

The oxidative quenchers for this study are based on the electron acceptror 1,3-

dinitrobenzene and are listed in Eq. 3 .7 .
H\ /H H\
o” 9)
T .
H
N02 N02 N02
3,5-DNB-amH” 3,5-DNBCOO‘ 3,5-DNBCOOH

The reductive quenchers for this study are based on the electron donor on N ,N-

dimethylbenzene, and are listed in Eq. 3.8.

H
H\N/ 9
H o; C?)
\ I O (3.8)
H t I
CH3 CH3

4-DMAB-amH+ 4-DMAB-COO‘

69

B. BPY Complexes of Ru(II) to Mediate Oxidative PCET

Reported here is the rate of proton-coupled electron transfer (PCET) through the
amidinium-carboxylate salt bridge. The donor is electronically excited Ru(bpy)32+ (bpy =
bipyridine), the acceptor is 3,5-dinitrobenzene (3,5-DNB), and a salt bridge is formed from
the association of an amidine to a carboxylic acid. The influence of proton motion on the
rate of electron transfer through the salt bridge is revealed by a deuterium isotope effect.
PCET in 2 is compared to 3 where a symmetric —(COOH)2— interface has been introduced

between the donor/ acceptor pair, Figure 27.
1. Results and Discussion

a. Solution Association of D/A PCET Complexes

The amidinium-carboxylate salt bridge is formed directly upon mixing the
appropriate carboxylic acid with free base amidine. Two favorable secondary electrostatic
interaction'8 within the amidinium-carboxylate interface are manifested in high association
constants even when the solvent is polar. Figure 28 shows the 'H NMR spectrum of
(bpy)zRu"(Mebpy-amH)3+ in DMSO-d6 at 19.5 °C. The ascending traces highlight the
changes in the chemical shifts of the amidinium protons, which appear as broad singlets at
~9.5 ppm, upon titration with the tetramethylammonium salt of the carboxylate. A
concentration-dependent downfield shift of >2.0 ppm is observed for the protons involved
in hydrogen bonding to the carboxylate whereas the chemical shift of the adjacent protons,
which are not bound to the carboxylate, varies marginally from 9.55 to 9.65 ppm over all
concentrations.

A least squares fit of a plot of the chemical shift of the hydrogen-bonded amidinium

protons vs the carboxylate concentration at 19.5 °C yields K = 1136 :t 93 M", Figure

855°C

29.'9 Unfortunately, the solubility of 2 in CHzClz, the solvent in which electron transfer

kinetics were determined, is too low to permit K to be reliably ascertained by NMR

3580C

titration experiments. Nonetheless we observe the association constant for 2 to increase

70

(bpy)2R"U<

 

 

N02

N02

Figure 27. Complexes to study PCET; 2 is based on [(bpy)2Ru"(Mebpy-amH+)]3+ and is
based on [(bpy)2Ru"(Mebpy- COOH)]2*.

71

with decreasing solvent polarity (K = 2432 M’1 in CH3CN at 19.5 °C). On this basis,

the DMSO and CH3CN association constants represent a lower limit for the formation of 2
in our electron transfer experiments. A Job’s plot, Figure 30, confirms the 1:1
stoichiometry of the amidinium-carboxylate salt bridge complex. For the case of 3, Kama,
measured by techniques previously described by us for the symmetric —(COOH)2—
interface,8 is 702 M" in CHZClZ, which is comparable to our previous measurements of a
dicarboxylic interface bridging a dinitrobenzoic acid acceptor to a porphyrin donor (K =

85 SOC

552 M").

b. Photophysical Behavior

The luminescence of [(bpy)2Ru"(bpy-am)]2" in CHZCl2 is quenched upon the
addition of 3,5-dinitrobenzoic acid (3,5-DNBCOOH). This result is consistent with an
electron transfer quenching mechanism, which is well-established for the reaction between
electronically excited Ru"tris(polypyridyl) complexes and nitroaromatic acceptors.” In the
absence of 3,5-DNBCOOH, the decay of the (bpy)2Ru"(bpy-amH)3+ excited state is
monoexponential with a lifetime of 1300 ns, which decreases upon complexation to
aliphatic carboxylates or benzoate (1:0 = 850 ns). However, when 3,5-DNBCOOH is
present, a biexponential decay of the luminescence is observed whereupon one lifetime
component is dependent on the concentration of acceptor and the other is not over a 3,5-
DNBCOOH concentration range of 0.1 to 5 mM ([aipy)2Ru"(Mebpy-am)2*] = 0.1 mM).

The origin of the concentration-dependent decay is easily understood. The Stem-
Volmer plot of the concentration-dependent lifetime is linear over the entire quencher
concentration range and the intercept is unity. The bimolecular rate constant of 2 .4 x 109 M“
's‘l is in accordance with that measured by Meyer et al. for Ru(bpy)32+ and 3,5-
dinitrobenzoic acid (kEr = 1.6x 109 M"s").15 More significantly, the bimolecular kinetics
of 2 is similar to that for the reaction of the (bpy)2Ru”(Mebpy-arnH)3+ with the ester, 3,5-
DNBCOOEt (kET = 1.9x109 M"s"), which is unable to associate with the amidine

72

(association of the amidinium with the nitro functionality is not likely,20 which we also
confirmed by NMR). The concentration independent lifetime decay component is
attributable to electron transfer for the associated pair shown 2. An intramolecular rate
constant of 4.3(9) x 106 s‘1 is determined from the concentration—independent lifetime decay
component.

The reaction of (bpy),Ru"(Mebpy-coon)2+ with 3,5-DNBCOOH in CH,C12
exhibits similar characteristics to the reaction of the quencher with (bpy)2Ru"(Mebpy-am)2*.
Lifetime decays are biexponential, also exhibiting concentration-independent and dependent
components. The bimolecular rate constant derived for the latter is 1.4x 109 M"s“, which
is comparable to the reaction rate constant of (bpy)2Ru"(Mebpy-COOH)2+ with the ester
3,5-DNBCOOEt (kET = 1.1 x 109 M"s"). Despite a 0.07 V smaller driving force (AG(2) =
—0.21 V, AG(3) = —0.14 V),2| the intramolecular rate constant of 8.0(4)x 106 s" for 3 is

nearly a factor of two greater than that observed for 2.

c. Kinetic Isotope Efi‘ects on PCET

Does the attenuated PCET process in 2 suggest that the salt bridge is affecting the
rate of the ET? To more directly address this issue, we exchanged the 1HS of the amidine
and carboxylic acid by 2H and measured the electron transfer kinetics of the associated pair.
The intrarnolecular rate constant for deuterated 2 was 3.2x 10‘5 s", yielding kH/kD =
1.34(3). This isotope effect is in line with that measured for 3 (kn/kD = 1.5) and it is
consistent with our previous measurement of electron transfer for a porphyrin donor—
(COOH)2—acceptor complex (kH/kD = 1.6 and 1.7). As recently discussed, a deuterium
isotope effect in these hydrogen bonded systems arises from the modulation of the
electronic coupling matrix element by proton motion within the interface,”22 thereby
providing a mechanism for the asymmetric salt bridge to engender a PCET reaction.

The effect of proton motion on the electron transfer rate may be manifested in ways

other than the electronic coupling. Unlike the —(COOH)2— interface, significant charge

73

 

 

of r
' 5
.‘ t l
: i i
u l '
\ . - _ l
I ' l ! l
. l l; |.
l . '1
t l
. l
l
I

[M l .9 .’ '1‘) l s:
‘M/ \" ‘le ‘VW-JJ V “J; W , .

 

 

llllllIllllllllllllllllllllllllllllIlll

11.0 10.0 9.0 8.0 7.0
ppm

 

Figure 28. 1H NMR spectrum of [(bpy)2Ru"(Mebpy-arnH*)]3+. Selected spectra are
shown upon addition of the tetramethylammonium salt of 3,5-dinitrobenzoate at
concentrations 0, 0.44, 0.89, 2.2, and 3.1 mM in DMSO—d6 (bottom to top). The 'H
resonances of the bipyridines appear between 7.3 and 9.2 ppm, two broad singlets flanking
9.5 ppm signify the H resonances of the amidinium protons. For the spectrum at 0.44 M
carboxylate added, the two resonances are coincident.

 

74

 

 

 

 

 

 

2.5
- J.)
2.0
E
E
\ 1.5
E
.C
a)
E
.2
E 1.0
(D
.C
0
<1
0.5
0.0 b I L l J l l l
0.0 10.0 20.0 30.0 40.0

[3,5-DNBenzoate] / mM

Figure 29. Plot of the chemical shift of the amidinium protons hydrogen-bonded to
carboxylate versus concentration of carboxylate from 0.44 to 40 mM.

75

 

0.32

0.21 -

0.11 -

x [(bpy)2Fiu"AmH+ : ‘OOCDNB]

 

 

0.00 t t I
0.00 0.25 0.50 0.75 1 .00

x [(bpy)2RU"AmH+]

 

Figure 30. Job’s plot of the relative salt-bridge complex concentration vs. the mole
fraction of [(bpy)2Ru"(Mebpy-amH“)]3+ (3.8 mM) as the concentration of DNBCOO’ is
varied from 0.0 to 40.0 mM.

76

rearrangement occurs upon proton motion in the salt bridge, and this charge redistribution
will couple to solvent (i.e. to give rise to additional Franck-Condon factors arising from
proton motion).'°" Moreover, the electron is transferred through the permanent electrostatic
field of the salt bridge in 2 thereby modifying the energetics of the overall reaction. These
issues may be addressed by comparing the rates of electron transfer for a donor—
(amidinium—carboxylate)—acceptor complex to its congener where the interface is switched
(i.e. donor—(carboxylate-amidinium)—acceptor). For the systems here, such a comparative
study is obscured by the possibility of transferring an electron from the ancillary bipyridine
ligand in addition to the transfer of an electron from the derivatized bipyridine ligand
because the two metal-to-ligand charge transfer states are close in energy.”24 We therefore
have gone on to assess the issue of PCET rates for switched interface systems with
complexes featuring a single bipyridine ligand, and with complexes in which the energy of
the metal-to-ligand charge transfer state for the ancillary ligands are energetically far

removed from the derivatized amidinium or carboxylate bipyridine ligand.

C. TMBPY Complexes of Ru(II) to Mediate Oxidative PCET

A direct experimental measure of the affect of a salt bridge on electron transfer is to
undertake a comparative kinetics study of a donor—(amidinium-carboxylate)—acceptor
suprarnolecule and its switched interface donor—(carboxylate-amidinium)—acceptor
congener. We now report such a study for the supramolecular series of complexes where
the donor is [(tmbpy)2Ru"(Mebpy-amH*)]3+ or [(tmbpy)2Ru"(Mebpy-COO“)]+ (tmbpy =
3,3’,4,4’-tetramethyl-2,2’-bipyridine, Mebpy-amH+ = 4-methyl-4’-amidinium-2,2’-
bipyridine, Mebpy-COO' = 4-methyl-4’-carboxylate-2,2’-bipyridine) and the acceptor is
the complementary carboxylate- or amidinium-modified 3,5-dinitrobenzene (4 and 5,
respectively in Figure 31). To further extend the study, we have also prepared the same
donor/acceptor pair bridged by the symmetric —(COOH)2— interface (6, also shown in
Figure 31). The metal complexes are afforded by standard reactions of the modified Mebpy

77

ligands with the cis-dichloride [(tmbpy)2RuCl,'].57 Electron transfer measurements of this
series of complexes reveal that the salt bridge can extraordinarily influence the rates of
electron transfer thereby allowing us to uncover new contributing effects to the mechanism

of biologically relevant PCET.
I. Results and Discussion

a. Solution Salt Bridge Association.

The Ru(H) carboxylate and amidinium complexes readily form the salt bridge with
the corresponding acceptor. Complexes 4 and 5 have been thoroughly characterized by
NMR. Figure 32 shows the lH NMR spectral changes resulting for the association of
[(tmbpy)2Ru"(Mebpy-arnI-I’)]3+ to 3,5-dinitrobenzoate (3,5-DNBCOO‘) in DMSO-d6. As

8 a signature of the salt

observed previously for guanidinium-carboxylate salt bridges,5
bridge is the concentration-dependent downfield shift of amidinium protons involved in
hydrogen bonding (NHax) and an insensitivity of the chemical shift for the amidinium
protons external to the salt bridge (N ch). The chemical shift of the NH” protons vary by
2.4 ppm upon their hydrogen bonding association to the carboxylate, whereas the chemical
shift of the Nch protons varies by <0.3 ppm over the same concentration range. A
nonlinear least-squares fit of chemical shift of the hydrogen-bonded amidinium protons vs.
the carboxylate concentration (Figure 33), as described by Wilcox,59 yields an association

constant (K ) of 386 M". A 1:1 stoichiometry of the supramolecule complexes is

established by the Job’s plot 6° of the 1H NMR titration data reproduced in Figure 34,
which shows that the optimal formation of 4 occurs at equimolar concentrations of the
amidinium and carboxylate (i.e., a 0.5 mole fraction). This is definitive evidence for the 1:1
stoichiometry of the amidinium-carboxylate salt bridge.

Similar behavior is observed for 5 (Figures 35 - 37) The NHeq resonance (9.53
ppm) of 3,5-dinitrobenzamidinium (3,5-DNBamH") exhibits an insignificant shift upon the

addition of [(tmbpy)2Ru"(Mebpy-C007]+ (A ppm < 0.2 over 10 mM), while NHax varies

'78

 

 

N02
N02
/N
11
(tmbpylzfitk
N
N02
N02
55
I/N
(tmbpmmk
N
N02
N02

 

Figure 31. Complexes to study PCET; 4 is based on [(tmbpy)2Ru"(Mebpy-amH’)]3’, 5
is based on [(tmbpy)2Ru"(Mebpy-COO‘)]” and 6 is based on [(tmbpy)2Ru"(Mebpy-

(3()()ll)]2+.

79

by 2.5 ppm over the 10 mM concentration range, Figure 35. The larger shift of NHe,q
resonances over the smaller concentration range as compared to 4 is indicative of the higher
association constant of 2297 M" for 5 in DMSO.

The greater association constant of 5 is consistent with the electron withdrawing
nitro groups conferring a decreased basicity of the carboxylate group on 3,5-DNB,
resulting in a weaker hydrogen bonding interaction for 4. In less polar solvents, the
association constant greatly increases. The low solubility of 4 and 5 in CH2C12 (the solvent
in which electron transfer kinetics were determined, vide infra) precludes reliable

determination of Kassoc by N MR titration experiments.

b. Photophysics

The design of the excited state structure of the Ru(II) polypyridyl complex is crucial
to a proper kinetics study of the electron transfer reactions of 4-6. The electron transfer
reaction to the 3,5-dinitrobenzene acceptor of 4-6 is initiated by laser excitation of the
metal-to-ligand charge transfer (MLCT) transition of the Ru(II) polypyridyl complex. Two
potential reaction pathways arise for the oxidative quenching reaction, depending on the
energetics of the MLCT excited state. For the case where the lowest energy MLCT excited
state is localized on the salt bridge modified Mebpy ligand, photoexcitation will place the
transferring electron directly into the PCET pathway. Alternatively, a lowest energy MLCT
excited state involving the ancillary polypyridine ligand will remove the excited electron
from the PCET reaction pathway. Indeed, we have confronted this issue in our earlier
studies of the oxidative quenching of [(bpy)2Ru"(Mebpy-X)]M (X = COO‘, n = 1, X =
amH’, n = 3). In the absence of tetramethyl substitution of the bipyridine rings, the MLCT
excited states involving the ancillary bpy and the Mebpy-amH+ and Mebpy-COO‘ ligands
are close in energy thereby obscuring a comparative PCET study owing to the presence of
the two competing reaction pathways.

The electronic spectra of [(tmbpy)2Ru"(Mebpy-amH’)]"+, [(tmbpy)2Ru"(Mebpy-

8O

 

NHax NHeq

 

 

 

 

l l l I

12.0 11.0 10.0 9.0 8.0 7.0
ppm

 

Figure 32. Spectral changes of 1H NMR spectrum of [(tmbpy)2Ru"(bpy-amH”)]3+ with
added tetrabutylammonium 3,5-dinitrobenzoate. Selected spectra are shown upon addition
of 3, 5- dinitrobenzoate at concentrations 0. 0, 2. 3, 3. 7, 6. 6, 10. 3 and 13. 9, mM 1n DMSO-
d6 (bottom to top). The 'H resonances of the bipyridines appear between 7. 0 and 9. 2 ppm,
two broad singlets flanking 9. 5 ppm signify the 'H resonances of the axial (internal) and
equatorial (external) amidinium protons.

 

81

 

A Chemical Shift (N—H) / ppm

 

 

 

0.0 l l l l l l l
0.0 10 20 30 40 50 60 70 80

[3,5-DNBenzoate] / mM

Figure 33. Plot of the chemical shift of the resonances shown in Figure 32 for
[(tmbpy)2Ru"(Mebpy-amH*)]3+ and additional resonances of the amidinium protons
hydrogen-bonded to carboxylate versus concentration of carboxylate from 0.0 to 80 mM.

 

82

 

 

 

 

0.8
O
0.7 — O .
O
E‘
z 0.6 — ' '
8
O O
o
E. 0.5 —
3': ' '
E
:2 o
g 0.4 —
N O
E
.D O
E 0.3 — 0
:3 .
O O
0.2 — .
0.1 l l l I l l l
0 0.25 0.5 0.75 1

x [(tmbpy)2Ru"AmH+]

Figure 34. Job’s plot of the relative salt-bridge complex concentration vs. the mole
fraction of [(tmbpy)2Ru"(Mebpy-amH’)]3+ (6.3 mM) as the concentration of 3,5-
DNBCOO' is varied from 0.0 to 80 mM

83

 

 

 

I’n‘
I, \\ M/ \‘
I \_ t
/’ V / \\
‘1' \\
/ \‘V ‘1.
IX
\\ ”/‘~\
\‘ / \"
‘\ ‘.
\M I/ fix
1 J ~\
\
I
*I. \\.-‘
p, I
‘1‘ / \
\ / .
\I‘J/ “\
/‘ ‘\
/.
M/ \
-_—’/ V \V. 1' 1w

 

 

 

12.0 11.0 10.0 9.0
ppm

Figure 35. 'H NMR spectrum of the tetraphenylborate salt of 3,5-DNBamH” with
addition of the PF; salt of [(tmbpy)zRu"(Mebpy-COO‘)]+ at concentrations 0.0, 1.8, 2.9
and 3.9 mM in DMSO-d6 (bottom to top). The 1H resonances of the axial (internal) and
equatorial (external) amidinium protons appear between 9.3 and 12.0 ppm.

 

84

 

 

 

 

2.5 ”
2.0 ‘
E
o.
o.
f
2' 1.5 ”
SE
.:
(D
R
.9
0E) 1.0 ‘
.c
0
<1
0.5 P
00 l I l l
0 2 4 6 8 10

[(tmbpy)2Ru"COO-] / mM

Figure 36. Plot of the chemical shift of the resonances shown in Figure 35 and additional

resonances of the amidinium protons hydrogen-bonded to Ru" carboxylate complex from
0.0 to 10 mM.

85

 

 

 

 

1.0
O
. o

0.8 — .
E
Z O
D
3E: o
it 0.6 "
6 o
O
9 o
‘3
i“ 0.4 _
>.
o.
.o O
.E.
2'

0.2 ‘

00 l l I Q

0 0.25 0.5 0.75 1
x[+HAmDNB]

Figure 37 . Job’s plot of the relative salt-bridge complex concentration vs. the mole
fraction of 3,5-DNBamH+ (2.9 mM) as the concentration of the [(tmbpy)2Ru"(Mebpy-
COO’)]" is varied from 0.0 to 10 mM.

 

86

coo-)1; and [(tmbpy)2Ru"(Mebpy-COOH)]2", are typical of Ru(II) polypyridyl
complexes. Absorption profiles are dominated by a high energy 1t1t* intraligand absorption
band and a split lower energy dn-rt* MLCT absorption band. Nonaqueous solutions of the
complexes exhibit a strong red. Luminescence decay exhibits monoexponential behavior
and a fit of luminescence data yield lifetimes of 860, 1030 and 770 ms for
[(trnbpy)2Ru"(Mebpy-amH“)l’*. [(tnrlblw)2Ru"(Mebpy-(300)]+ and [(nnbpy)2Ru"(Mebpy-
COOH)]2’, respectively.

The relative energies of the relevant MLCI‘ excited states for systems 1-3 may be
evaluated by analyzing the emission profiles of the various homoleptic complexes as

follows,74

AG“o = Eo + x (3,9)
x = (AV0,1,2)2(16kBT ln2)‘I (3,10)

where the excited state energy (AGcf) is related to the energy of the luminescence
maximum (E0) and the reorganization energy (x) containing solvent and low frequency
modes, which are treated classically and related to the full width at half-maximum, AVOJ , 2
for a single vibronic component. The relation of the spectroscopic energy to a free energy
arises from negligible pressure-volume work (and hence the energy is primary enthalpic)‘SI
and small electronic entropic contributions accompanying excited state production.‘52
Measured values of E0 = 14,843 cm‘1 and A VH2 = 2476 cm"l for Ru(II) Iris-(Mebpy-
amH*) and E0 = 15,359 cm" and AV“2 = 1826 cm" for Ru(II) tris-(Mebpy-COO’)
complex yield excited state energies (E04,) of 2.17 and 2.08 eV, respectively, as compared
to that reported for Ru(tmbpy),2+ (E0.o = 2.56 eV“). The destabilization of the dn: -—>
1t*(tmbpy) MLCT by nearly 0.4 CV with regard to the MLCT excited states of Mebpy-

amH" and Mebpy-COO‘ ensures that photoexcitation of 4-6 cleanly promotes the

87

transferring electron on to the Mebpy ligand, from where it can smoothly advance to the

dinitrobenzoic acceptor.

c. PCET

Consistent with these excited state energetics, the Ru(II) excited state undergoes
electron transfer reactions with unbound and bound acceptor. The luminescence from the
[(tmbpy)2Ru"(Mebpy-COO‘)]" bound to benzamidinium and [(tmbpy)2Ru"(Mebpy-
arnidirlium)]3+ bound to benzoate is long-lived with observed excited state lifetimes of 1200
and 470 ns, respectively. Owing to the large association constants of 4 and 5 in CHZClz,
these lifetimes are independent of benzamidinium and benzoate at greater concentrations
than that of the metal complex. Conversely, the presence of the 3,5-DNB electron accepting
group leads to efficient quenching of the Ru(II) excited state luminescence for 4-6. For
each system, biphasic decay kinetics are observed for the quencher concentration range
investigated. One component of the emission decay is clearly dependent upon the
concentration of quencher whereas the other component remains concentration-
independent. The concentration-dependent lifetimes obey typical linear Stem-Volmer
quenching kinetics and the rate constants for the bimolecular reactions of the respective
constituents of supramolecule assemblies are listed in Table 4. The bimolecular reaction at
low concentrations likely involves complexed quencher and free Ru(II) complex as a result
of the high association constants of the supramolecule assemblies 4 and 5. The bimolecular
rate constants for 4-6 accord well with the kinetics of equi-exergonic bimolecular reactions
between Ru"tris(polypyridyl) and nitroaromatic quenchers.63 Of greater relevance to PCET
are the attendant concentration-independent rate constants, which we attribute to the
unimolecular electron transfer of the associated complexes.64 The striking result of these
data listed in Table 1 is that the electron transfer rate constant through 4 is considerably

slower than for its switched interface congener 5. The differences in the unimolecular rate

 

88

constants for 4 and 5 reveal that the rate of electron transfer depends significantly on the
salt bridge and its orientation with respect to the electron transfer pathway.

The slower rate of 4 may have several origins. In 4, the permanent dipole (5’8‘) of
the salt bridge is in the direction of electron transfer whereas in 5 electron transfer opposes
the dipole. Internal electrostatic fields affect the rates of electron transfer by altering the
driving force of reaction relative to the isolated constituents."5*’7 By using typical bond
distances of [Ru(bpy)3]2", DNB and the amidinium—carboxylate salt bridge, we calculate
that a 0.37 V (less favorable for 4) field-induced difference between the driving force for
intrarnolecular electron transfer (AEC = ezler) . 68 This thermodynamic attenuation of the
electron transfer rate constant for 4 may be further augmented by a greater reorganization
energy associated with the salt bridge. In 4, accompanying proton transfer (from the Run
amidinium donor to the carboxylate acceptor) can stabilize the charge of the electron as it
develops on the acceptor. In this case, since the proton charge is strongly coupled to the
solvent dipoles, as with the electron, charge shift within the salt bridge may be
accompanied by significant solvent polarization thereby giving rise to additional Franck-
Condon factors. This is not the case for 5. Here, the proton is already residing on the
acceptor, and hence it is likely to remain upon the arrival of the electron. Finally,
differences in H-bonding strengths of the asymmetric interfaces may be manifested in
differences in electronic coupling efficiencies."9 The electron withdrawing nitro groups will
stabilize the negative charge on the carboxylate, resulting in a weaker hydrogen bond in 6

and hence correspondingly weaker electronic coupling pathway.

2. Conclusions

Our results show that intervening salt bridges can profoundly mediate the rates of
intrarnolecular electron transfer. Unlike a symmetric interface, the salt bridge can
significantly affect the rate of electron transfer from contributions of the electrostatic

potential, Franck-Condom factors, and electronic coupling arising from the asymmetric

 

89

Table 4. Rates for unimolecular and bimolecular electron transfer for donor-acceptor
complexes with amidinium-carboxylate and dicarboxylic acid dimer bridges in
dichloromethane at 22 °C.

 

 

PCET Complex AG° / eV"l kEr/ 109 M" s" kPCEr/ 106 s"
4 —0.14 1.2 8.4
5 —0.34 3.3 310
6 —0.23 3.2 43

 

“Excited state redox potentials of the [(tmbpy)2Ru"(Mebpy-amH’)]3+,
[(tmbpy)2Ru"(Mebpy-COO’)]" and [(tmbpy)2Ru"(Mebpy-COOH)]2+ donors were
determined from the simple thermodynamic relation: E,,2(*Ru"”) = AG,”o - E,,2(Ru"”)
where E,,2(Ru"") and E,,2(*Ru"”) are the ground state and excited state Ru"ll reduction
potentials, respectively. The excited state free energies (AG °) of 2.23, 2.11 and 2.22 eV
were determined from Eqs. (3.9) and (3.10) and the Rum’fisreduction potentials of 1.05,
1.02 and 1.05 eV were measured by cycilc voltammetry (CHzCl2 containing 0.1 M
tetrabutylammonium hexafluorophosphate as supporting electrolyte), respectively.
Reduction potentials of 3,5-dinitrobenzoate and 3,5-dinitrobenzamidine were measured to
be —1.04 and —0.85 V vs SCE, respectively. Protonation of the acceptors was estimated to
facilitate reduction by 0.1 V.

90

charge distribution. Such effects will be present in any biological system where the
developing charge resulting from proton motion coupled to the electron transfer pathway is
not compensated (e.g. proton pumps, oxygen activation). Current studies are underway to
begin disentangling the contributions of the different PCET mechanisms for electron

transfer through asymmetric interfaces.

D. DECB and BPY Complexes of Ru(II) to Mediate Reductive PCET

With an electronic framework established for the PCET reactions of Ru(H)
polypyridyl complexes established, we wondered whether the influence of the salt bridge
on electron transfer rates was specific to an oxidative quenching pathway. To address this
issue, we undertook electron transfer reactions of Ru(II) polypyridyls juxtaposed to donors
via the amidinium-carboxylate salt bridge. In the case of the reductive quenching pathway,
the PCET reaction effectively amounts to hole transfer through the amidinium-carboxylate
salt bridge (as opposed to electron transfer for the oxidative pathway). Accordingly, a
meaningful PCET study demands that the excited electron be removed from the PCET
pathway by MLCT to the ancillary bipyridyl ligand. We now report that modification of the
ancillary bpy ligands by diethylcarboxy (decb) permits PCET reactions of Ru(II)
polypyridyl excited states to be investigated by reductive quenching, Figure 38.
Supramolecule assemblies 7 and 8 have been prepared in which an electron may transfer
from N,N’-dimethylaniline to the electronically excited Ru(II) polypyridyl acceptor through
the intervening amidinium-carboxylate salt bridge interface. Electron transfer into the Ru(H)
metal center is fast when the Ru(II) polypyridyl complex is attached to the arnidinium-side
of the salt bridge. This behavior is opposite to that previously observed by us for the
oxidative quenching pathway where electron transfer is fast when the Ru(H) polypyridyl
complex is attached to the carboxylate-side of the salt bridge. These differences in the
oxidative and reductive quenching reactions may be understood by considering the

orientation of the salt bridge relative to the direction of electron transport. Complex 9 was

91

also studied however it does not appear to undergo an ET reaction due to thermodynamic

constaints.
1. Results and Discussion

a. Solution Association

Amidinium—carboxylate salt bridge complexes 7 and 8 readily form in CH3CN
solutions. The exceptionally high association of the salt bridge agrees well with
Jorgensen’s classification for two favorable secondary interactions composing the
hydrogen bonded interface”, bolstered by the stabilization imparted by the molecular
recognition of the negatively charged carboxylate by the positively charged amidinium. The
significantly higher Kassoc of 7 is consistent with an increased basicity of the carboxylate

group conferred by the electron donating dimethylamino group, resulting in a stronger

hydrogen bonding interaction than that observed for switched interface system 9.

b. Photophysical Behavior

The long-lived luminescence from the Ru(II) polypyridyl complexes is preserved
upon formation of the salt bridge in the absence of an electron donating group. Addition of
benzoate to acetonitrile solutions of [(bpy)2Ru"(Mebpy-amH*)]3’ and [(decb)2Ru"(Mebpy-
amH*)]3+ lead to a modest attenuation in the observed excited state lifetimes (‘c = 975 and
1330 ns, respectively), which remain relatively independent of benzoate concentrations.
Similarly, long-lived excited state lifetimes of 900 and 1130 ns are observed for
[(bpy)2Ru"(Mebpy-COO“)? and [(decb)2Ru"(Mebpy-COO')]" associated to
benzamidinium. The Ru(II) excited state luminescence of [(bpy)2Ru"(Mebpy-amH“)]3+ and
[(decb)2Ru"(Mebpy-amH")]3+ is dramatically quenched when phenyl of benzoate is
substituted with the dimethylaniline. This result is consistent with an electron transfer
quenching mechanism, which is a well-established reaction between electronically excited

Ru(II) tris(polypyridyl) complexes and dialkylylamine aromatic electron donors.

 

92

(decb)2Fii|l<

 

/ |
N
u/ \
(bpy)2Ru\ H
N/ |
\ I 'N\
((9 H‘~\
bk 9
H/ Hs“ 6’)
‘0 CH
I 3
'1'
CH3
8

 

Figure 38. Complexes to study PCET; 7 is based on [(decb)2Ru"(Mebpy-amH*)]3’, 8 is
based on [(bpy)2Ru"(Mebpy-amH*)]3+ and 9 is based on [(decb)2Ru"(Mebpy- COO‘ )]".

93

Biexponential decay kinetics are observed where one lifetime component of the
emission decay is dependent on the concentration of donor and the other is not over a
DMA-COO‘ concentration range of 0.062 to 1.24 mM ([(bpy)2Ru"(Mebpy-am)2*] = 0.051
mM). The concentration-dependent lifetimes obey typical linear Stem-Volmer quenching
kinetics and the bimolecular rate constants of 6.4X108 M"s" for [(bpy)2Ru“(Mebpy-
amH*)]3" and of 3.5x108 M"s" for [(decb)2Ru"(Mebpy-amH")]3+ accord well with the
kinetics of equi-exergonic bimolecular reactions between Ru(bpy)32+ and dialkylanilines.
Analyses of the concentration-independent lifetime decay curves yield intramolecular rate
constants of 1.9x 107 s" and 1.7x 109 s" for 8 and 7, respectively. The independence of
the rate constant on quencher concentration is consistent with an electron transfer
quenching reaction occurring within the associated pair.

When the interface is carboxylate and amidinium moieties are interchanged on the
Ru(II) acceptor and DMA donor, as in 9, no electron transfer quenching of the Ru(II)
luminescence is observed. The lifetime of [(bpy)2Ru"(Mebpy-COO’)]* and
[(decb)2Ru"(Mebpy-COO’)]+ is invariant with the addition of DMA-amH+ over a quencher
concentration range of 1.0 - 140 x 10'5 M.

The significant differences in the electron transfer reactivity between the 7 ,8 and 9
systems are reflected in the electron transfer driving forces. For N,N’-
dimethylaminobenzarnidinium, a reversible half-wave potential at 1.10 V vs. SCE was
observed in the cyclic voltammogram. Conversely, N ,N ’-dimethylarninobenzoate ('I'BA"
salt) was electrochemically irreversible as evidenced by the presence of only an anodic
wave in the cyclic voltammetric scan. Accordingly, the potential for this compound was
estimated from a Marcus curve by using methods previously described by Schuster and co-
workers and gives a redox potential of 0.70 V vs. SCE.

Consistent with the equilibrium measurements, the Ru(II) excited state undergoes
electron transfer with unbound and bound donor when the reaction is exergonic. The

luminescence from [(bpy) Ru"(Mebpy-amH")]3+ and [(decb) Ru“(Mebpy-amH‘*)]3+ is
2 2

94

efficiently quenched by DMA-COO”. That the bimolecular rate for the reaction of the bpy
complex, 8, is greater than that of the decb complex, 7, is sensible in light of the greater
driving force of the bimolecular electron transfer reaction of the former (AAG = 0.13 V).
Along these lines, the inversion of the relative ordering of the rate constants, and their large
difference (8 is ~102 less than that of its decb congener 7), for the salt-bridge assembled
acceptor-donor pair appears peculiar. We believe that the origins of this behavior arise from
disparate MLCT energetics and consequently electronic structure of the Ru(II) polypyridyl
acceptor complex.

The electron transfer reactions of 7 and 8 is established by photogenerating a hole
at the Ru(II) center by MLCT excitation. The energy ordering decb ~ Mebpy-amH’ << bpy
follows a simple trend where the energy of the MLCT excited state increases as the electron
withdrawing ability of the substituents on the polypyridyl rings decreases. In the case of 8,
simple Boltzmann considerations suggest that MLCT excitation places an electron on the
Mebpy-amH“ and thus the electron transfer pathway contains a photoreduced Mebpy
ligand. For 7, the decb and Mebpy-amH+ MLCT excited states are nearly equi-energetic
and therefore the photoexcited electron is partly removed from the electron transfer
pathway. In an electronic sense, the transferring electron traverses a less obstructed
pathway from DMA to the photogenerated hole on the metal.

The electronics of 7 and 8 are straightforward with regard to photoinitiating the
electron transfer reaction. The energetics for the dtt —> 1t*(polypyridine) MLCT excited
state are decb < bpy < Mebpy-COO‘. Photoexcitation of 9 cleanly establishes a hole at the
metal ion without any contribution of a photoreduced ligand along the electron transfer
pathway. Nevertheless, electron transfer is not observed for the salt-bridge complex.
Moreover, the corresponding bimolecular electron transfer reaction of 9 is negligible.
Inspection of the free energy involved in ET in 9 reveal that the electron transfer reactions

for this systems is endergonic.

 

95

The reactivity difference between the Ru(II) polypyridyl amidinium systems (7 and
8) and carboxylate congener (9) is primarily a result in a shift of the DMA donor potential
upon the presence of amidinium at the para position of dimethylaniline. DMA-amH" is 0.40
V more difficult to oxidize than DMA-COO‘, which is a slightly better reductant than DMA
(13,,2 = 0.81 V vs. SCE). The increased difficulty in oxidizing DMA-amH+ is sufficiently

destabilizing to drive the electron transfer reaction of 9 energetically uphill.

2. Conclusions

The overall electron transfer kinetics of 7 and 8 follows the emerging trend that
electron transport through salt bridges is fast for the D—(carboxylate-amidinium)—A
complexes. In this orientation, electron transfer is in the direction of the permanent dipole
of the salt bridge and the internal electric field contributes favorably to the driving force of
reaction relative to the isolated constituents."72 The thermodynamics may be further
augmented by favorable kinetics associated with a D—(carboxylate-amidinium)—A
orientation. Charge redistribution within the salt-bridge may be strongly coupled to the
solvent dipoles, thereby giving rise to Franck-Condon factors in excess of that expected for
simple electron transfer. In the case of the D—(carboxylate-amidinium)—A complex, the
proton already resides on the acceptor and hence it is likely to remain upon the arrival of the
electron. Franck-Condon factors arising from proton motion within the salt bridge are
therefore minimized. On the basis of the thermodynamic and kinetics considerations, we
can account for the differences in the oxidative and reductive quenching pathways. For the
oxidative quenching reaction, electron transfer should be fast out of the electronically
excited Ru(II) complex. Conversely, fast electron transfer into the metal center should be
observed for a reductive quenching pathway. In both cases, the orientation of the salt
bridge with regard to the direction of electron transfer is the same and favorable with regard
to electron transfer. Unfortunately, as was previously provided by a comparative kinetics

studies of switched interface systems for the oxidative quenching pathway, a more explicit

96

analysis of the effect of the salt bridge on electron transfer pathway is obviated by the large

difference in the electron transfer driving forces of 7 and 9.

E. Materials and Methods

Synthesis of 4-methyl-4’-amidinium-2,2’-bipyridine. The ligand
synthesis began with the preparation of 4-methyl-4’-carbonitrile-2,2’-bipyridine.
Hydroxylamine hydrochloride (1.8 g, 26 mmol) was added to a 40 ml formic acid solution
of 4-methyl-4’-carboxaldehyde-2,2’-bipyridine73 (4.0 g, 20 mmol). The solution was
refluxed for 24 h under an argon atmosphere. After cooling to room temperature, the
solution was poured into 200 ml of ice. Neutralization with an aqueous sodium carbonate
solution produced a white precipitate, which was extracted into dichloromethane (3x50
ml). The organic extracts were combined, dried over magnesium sulfate, and the solvent
was removed under reduced pressure to give the product, 4-methyl-4’-carbonitrile-2,2’-
bipyridine. Yield: 3.15 g, 80%; MS: 195.1 m/z; mp. 134-137 °C; 1H NMR (CD3CN): 5
8.79 (1H, d), 8.60 (1H, s), 8.50 (1H, d), 8.21 (1H, s), 7.62 (1H, d), 7.25 (1H, d), 2.12
ppm (3H, s);'3C NMR(CD3CN):8158.16, 154.76, 151.16, 150.23, 149.82, 126.68,
126.15, 123.54, 122.58, 122.03, 117.76, 21.22 ppm.

Conversion to the amidinium was achieved by charging a solution of freshly
prepared sodium methoxide in methanol (20 ml, 0.04 M) with 4-methyl-4’-carbonitrile-
2,2 ’-bipyridine (1.50 g, 7.7 mmol) under argon. The resulting solution was stirred at room
temperature for 24 h, ammonium chloride (0.43 g, 8.0 mmol) was added in one portion,
and stirring was continued for an additional 24 h under argon. Large colorless crystals of
(4-methyl-4’-amidinium-2,2’-bipyridine)Cl (Yield: 0.70 g, 71%) were collected and
dissolved in a minimal amount of water. The conversion of the Cl‘ salt to the PF; salt was
quantitative upon the addition of a saturated aqueous solution of ammonium
hexafluorophosphate to an aqueous solution of (4-methyl-4’-amidinium-2,2’-

bipyridine)Cl. MS: 212.1 m/z; lH NMR (CD3CN): 8 8.87 (1H, d), 8.65 (1H, s), 8.52

97

(1H, d), 8.27 (1H, s), 8.02 (4H, s), 7.64 (1H, d), 7.39 (1H, d), 2.41 ppm (3H, s); 13C
NMR (CD3CN): 5 166.79, 158.25, 154.82, 151.51, 150.21, 137.12, 126.82, 122.93,
122.44, 119.84, 21.26 ppm.

Synthesis of [(tmbpy)2Ru(Mebpy-amH+)][(PF6)3]. The ligand 4wmethyl-
4’-amidinium—2,2’-bipylidine chloride (0.09g, 0.36 mmol) was reacted with
(tmbpy)2Ru"C1274 (0.30 g, 0.50 rmnol) in refluxing 95% ethanol (20 ml) for 4 h. After
removing the solvent under reduced pressure, the resulting solid was dissolved in a
minimal amount of water and filtered. Addition of an aqueous saturated solution of
NH 4PF6 yielded [(trnbpy)2Ru"(Mebpy-amidinium)]3+ as its PF; salt, which was washed
with water (5x 15 ml), dried in air, and washed with anhydrous ethyl ether (5 x15 ml).
Yield: 0.17g, 40%; ES/MS [M]2*: 369.6 m/z; 'H NMR (DMSO-dé): 5 9.60 (2H, s), 9.37
(2H, s), 9.03 (1H, s), 8.71 (1H, s), 8.59 (4H, s), 7.97 (1H, d), 7.70 (1H, d), 7.59 (1H,
d), 7.42 (1H, d), 7.29 (1H, s), 7.28 (1H, s), 7.26 (1H, s), 7.17 (1H, s), 2.55 (3H, s),
2.41 (12H, 5), 2.08 ppm (12H, 5).

Synthesis of [(tmbpy)2Ru(Mebpy-COOH)][(PF6)2] and
[(tmbpy)2Ru(Mebpy-COO‘)][(PF6)]. A 95% ethanol (20 ml) solution containing 4-
methyl-4’-carboxylic acid-2,2’-bipyidine73 (0.05g, 0.23 mmol) and (tmbpy)2Ru"Cl2 (0.24
g, 0.39 mmol) was refluxed for 24 h. The solution was cooled and the solvent was
removed under reduced pressure. The remaining solid was dissolved in a minimal amount
of water and filtered. The orange PF; salt of [(tmbpy)zRu"(Mebpy-COOH)]2+ was
obtained upon the addition of an aqueous saturated solution of NH 4PF; The product was
further purified by charging a column of neutral alumina, activity 1, with a solution of the
complex salt and eluting with dichloromethane and ethanol 6:1 (v/v). Yield: 0.12g, 52%;
ES/MS [M]2’: 370.0 m/z; 1H NMR (CD3CN): 5 8.89 (1H, s), 8.45, (1H, s), 8.27, (4H,
m), 7.72 (1H, d), 7.60 (1H, d) 7.46 (1H, d), 7.36 (1H, s), 7.33 (1H, s), 7.32 (1H, s),
7.27 (1H, s), 7.15 (1H, d), 2.50 (3H, s), 2.42 (6H, s), 2.41 (6H, s), 2.07 (6H, s), 2.06

ppm (6H, s).

98

The carboxylate form of the complex was obtained by dissolving the chloride salt of
[(tmbpy)zRu"(Mebpy-COOH)]2+ in a minimal amount of water. The pH was adjusted to
~10 with sodium hydroxide and addition of saturated solution of sodium
hexafluorophosphate, added dropwise, yielded an orange precipitate. The precipitate was
filtered onto a fine flit and washed with water. Because the complex is slightly soluble in
aqueous solution, a minimal amount of water should be used for washing (3 x 5 ml). The
gummy, air-dried solid was washed with copious amounts of anhydrous ethyl ether to
remove any vestiges of water, thereby yielding a microcrystalline solid. IH NMR (DMSO~
d6): 5 8.85 (1H, s), 8.73 (1H, s), 8.57 (2H, s), 8.55 (2H, s), 7.66 (1H, d), 7.60 (1H, d),
7.45 ( 1H, d), 7.29 (5H, m), 2.50 (3H, s), 2.41 (9H, s), 2.39 (3H, s), 2.07 (6H, s), 2.06
ppm (12H, 8).

Synthesis of [NBu4][3,S-dinitrobenzoate]. Twenty three ml of a 1.0 M
tetrabutylammonium hydroxide methanol solution was added via burette to a flask
containing 3,5-dinitrobenzoic acid (5.0 g, 23 mmol). The mixture was stirred until the acid
was completely dissolved. A viscous oil remained upon removal of the methanol under
reduced pressure. The oil was dissolved in freshly distilled benzene and 10 g of basic
alumina was added to the solution, which was stirred for 10 min and filtered. The filtrate
was freeze-dried to yield the product as a fluffy white solid. Yield: 8.5g, 81%; MS: 212.1
m/z; 1H NMR (DMSO-d6): 5 8.90 (2H, s), 8.75 (1H, s), 3.20, (8H, s), 1.57 (8H, p)
1.28, (8H, s), 0.89 ppm (12H, t); l3C NMR (DMSO-d6): 8 162.97, 147.46, 146.00,
128.25, 117.91, 57.551, 23.11, 19.22, 13.42 ppm.

Synthesis of [3,5-dinitrobenzamidinium][BPh4]. The chloride salt of 3,5-
dinitrobenzamidinium (0.5 g, 2.03 mmol), which has previously been prepared by
Creary,75 was dissolved in a minimal amount water and residual solid was removed by
filtering through a fine flit. An aqueous solution of sodium tetraphenylborate (1.0 g, 2.92
mrnol) was added to the filtrate to afford a lemon yellow precipitate. This solid was filtered,

washed with water (5 x 15 ml), dried in air, and washed with anhydrous ethyl ether (5 x 15

99

ml). The metathesis reaction proceeded with nearly quantitative conversion. MS: 210.0
m/z; 'H NMR (DMSO- 6): 5 9.59 (4H, s), 9.15, (1H, s), 9.10, (2H, s), 7.24 (8H, s),
6.97 (8H, t), 6.83 ppm (4H, t); 13C NMR: 5 164.44, 163.79, 163.13, 162.48, 162.21,
147.87, 135.64, 131.071, 129.31, 125.37, 123.03, 121.60 ppm.

Synthesis of [Ru(decb)2C12]. An absolute ethanol (50 ml) solution containing
2,2’-bipyridine-4,4’-dicarboxylic acid (1.0 g, 4.1 rnrnol), RuCl3 °(HZO)3 (0.5 g, 1.9
mmol), LiCl (1.0 g, 23.6 mmol) and concentrated hydrochloric acid (1 ml) was refluxed
for 24 h. The reaction mixture was allowed to cool to room temperature and poured into ice
water (250 ml). A black microcrystalline precipitate forms immediately. The solid was
collected on a fine frit, washed with water, dried in air, and washed with anhydrous ether.
The product was obtained by column chromatographic separation on silica gel with CH2C12
and ethanol 50:1 v/v as the eluant. Yield: 0.70 g, 71%; ES IMS [M]+: 772.4 m/z; 'H NMR
(CD3CN): 5 10.22 (1H, d), 8.95 (1H, s), 8.77 (1H, s), 8.18 (1H, d), 7.77 (1H, d), 7.43
(1H, d), 4.51 (2H, q), 4.35 (2H, q), 1.47 (3H, t), 1.34 ppm (3H, t); 13C NMR (CDCl,): 5
164.00, 163.49, 160.22, 157.89, 154.95, 152.43, 136.24, 134.94, 124.89, 124.40,
121.76, 121.50, 62.51, 62.34, 14.24, 14.11 ppm.

Synthesis of [Ru(decb)2(Mebpy-COOH)][(PF6)2]. The ligand 4-methyl-
4’-carboxylic acid-2,2’-bipyidine (0.020 g, 0.093 mmol)76 was combined with
Ru(decb)2Cl2 (0.060 g, 0.078 mmol) in ethylene glycol (2 ml) and the mixture was heated
at 120 °C for 20 min. The ethylene glycol was removed by liquid-liquid extraction with
ether (2 l) to yield a tar-like orange residue, which was dissolved in a minimal amount of
water. Addition of ammonium hexafluorophosphate to the aqueous solution yielded an
orange precipitate. The solid was purified by neutral alumina, activity 1, and eluted with
dichloromethane and ethanol 6:1 (v/v). Yield: 32 mg, 34%; ES/MS [M]2’: 458.0 m/z; 'H
NMR (CD3CN): 5 9.03 (4H, s), 9.01 ( 1H, s), 8.60 (1H, s), 7.85 (9H, m), 7.73 (1H, d),
7.47 (1H, d), 7.26 (1H, d), 4.44 (8H. q). 1.39 (12H, t).

100

Synthesis of [Ru(decb)2(Mebpy-amH+)][(PF6)3]. The metal complex was
prepared by a chloride substitution reaction with [4—methyl-4’-amidinium-2,2’-
bipyridine][Cl]. A freshly distilled ethanol solution containing Ru(decb)2Cl2 (0.084 g, 0.1 1
mmol) and [4-methyl-4’-amidinium—2,2’-bipyridine][Cl] (0.048 g, 0.19 mmol) was
refluxed under argon for 18 days. A solution color change from deep purple to dark orange
was observed over the course of the reaction. The reacted solution was added to a short
silica gel column, and eluted with dichloromethane and ethanol 10:1 (v/v). Several bands
separated from the product, which remained immobilized on the silica gel. After all mobile
impurities were eluted from the column, the silica gel containing the product was collected
from the top of the column. A solution of ammonium hexafluorophosphate (0.5 g) in
dichloromethane and ethanol 10:1 (v/v) released the product from the silica gel. The silica
gel was removed by filtration. The mother liquor was taken to dryness, and water was
added to remove the excess ammonium hexafluorophosphate. The resulting solid was
collected on a fine frit, dried in air, and washed with dry ether to afford [Ru(decb)2(Mebpy-
amH*)][(PF6)3]. Yield: 0.056 g, 38%; ES-MS [M]2*: 456.9; 'H NMR (CD3CN): 5 9.08
(4H, s), 8.67 (1H, s), 8.46 (1H, s), 8.06 (4H, s), 7.90 (9H, m), 7.58 (1H, d), 7.50 (1H,
d), 7.33 (1H, d), 4.44 (8H, q), 2.62 (3H, s), 1.42 ppm (12H, t).

Synthesis of [NBu4][4-(N,N’-Dimethylamino)benzoate] . An equivalent
of a 1.0 M tetrabutylammonium hydroxide (31.5 ml) in methanol was added to N,N’-
dimethylarninobenzoic acid (5.20 g, 31.5 mmol). The mixture was stirred until the acid
was completely dissolved. A viscous oil remained upon removal of the methanol under
reduced pressure. The oil was dissolved in freshly distilled benzene and 10 g of basic
alumina was added to the solution, which was stirred for 10 min and filtered. The filtrate
was freeze-dried to yield the product as a fluffy white solid. Yield: 11.95 g, 93%; ES/MS
[M+2H]”: 166.2 m/z; 1H NMR (CD3CN): 5 7.76 (2H, d), 7.60 (2H, d), 3.09 (8H, m),
2.88 (6H, s), 1.56 (8H, m), 1.31 (8H, m), 0.91 ppm (12H, t); 13C NMR (CD3CN): 5
170.69, 151.88, 131.46, 131.25, 111.70, 59.12, 40.83, 24.32, 20.28, 13.84 ppm.

101

Synthesis of [4-(N,N-Dimethylamino)benzamidinium][PF6]. In a dry
Schlenk tube, 4-(N,N’-dimethylamino)benzonitrile (1.68 g, 11.5 mmol) was added via
cannula to a 1.0 M solution of methylchloroaluminumamide (35 ml). The mixture was
heated to 80 °C for 3 days. The reaction mixture was quenched with a chloroform slurry of
silica and the resulting mixture was stirred until gas evolution ceased. This suspension was
poured into a fine frit, and thoroughly washed with a chloroform/methanol solution.
Removal of solvent yielded the chloride salt of the product (1.48 g, 64%). The
hexafluorophosphate salt was obtained by dissolving this chloride salt in a minimal amount
of water, followed by the addition of a saturated solution of ammonium
hexafluorophosphate. The metathesis reaction proceeded with nearly quantitative
conversion. ES/MS [M]*: 164.4 m/z; 1H NMR (CD3CN): 5 7.59 (2H, d), 7.30 (2H, s),
7.06 (2H, s), 6.78 (2H, d), 3.05 (6H, s); 13C NMR (CD3CN): 5 166.167, 155.70,
130.58, 112.16, 111.78, 40.22 ppm.

Electrospray mass spectrometric (ES /MS) analyses were obtained with a Finnegan
mat (San Jose, CA) quadrupole mass spectrometer using a CH3CN mobile phase. A
CH3CN solution of the sample was infused directly into the vaporization nozzle of the
electrospray ion source at a flow rate of 3 ml min". Nitrogen was used as the nebulizing
gas at a pressure of 35 PSI. NMR spectra were recorded on a Varian VXR 300 and
samples were thermostatted at 298 (0.2) K.

Association constants for 4 and 5 were determined by measuring the lH NMR
chemical shift of amidinium protons by titrating DMSO-d6 solutions of the appropriate
amidinium compound with varying amounts of the complementary carboxylate compound.
Thus for 4, the hexafluorophosphate salt of [(tmbpy)2Ru"(Mebpy-amH")]3+ (6. 3 mM)
was titrated with tetrabutylammonium 3,5-dinitrobenzoate whereas 3,5-
dinitrobenzarnidinium tetraphenylborate (2.9 mM) was titrated with [(tmbpy)2Ru"(Mebpy-
COO“)]+ (PF; salt).

Chapter 4
F UNCTIONALIZATION OF PORPHYRIN WITH THE AMIDINE GROUP

A. Introduction

Due to the relavence of porphyrins in biological ET we became interested in
developing a facile route to amidine-modified porphyrins. Although a common synthetic
approach to amidines is the conversion of a nitrile to its imidate ester followed by
aminolysis, this method can be limited by the insolubility of porphyrin nitriles in methanol.
Therefore we have adapted Garigipati’s strategy77 to prepare porphyrin amidines for the
first time by using a metal-mediated amide transfer to a nitrile. We also show that an
amidinium-carboxylate salt bridge is readily formed from the porphyrin amidine donor and
carboxylic acid acceptor, thereby pemlitting us to measure the rate of a PCET reaction
through the salt-bridge. Zinc porphyrins have been popular donors in photoinitiated
electron transfer schemes because they can readily donate an electron from their singlet

excited state to a variety of acceptors.78

B. Results and Discussion

1. Synthesis

The synthesis of the Zn(II) porphyrin vinyl amidine employed for PCET
measurements is displayed in Figure 39. The amidine group is afforded by reaction of
Weinreb’s amide transfer reagent,79 methyl aluminum(III) chloroamide, with the

acrylonitrile, which in turn was obtained from the corresponding aldehyde. We have found

102

103

(i) x
CHO —>

(ii) ("R 12 R

NH

(iii) (iv)

13

   
  

Hoq
\N_ax---0 N02

/NH2 _._..

(G) 9’ N02

"CH= CH- CN

NH

CH=CH4

NH2

Figure 39. Preparation of ocatethyl porphyrin with the amidine functional group: (i)
(EtO)2PCH2 CN, NaNHz; (ii) MeAl(Cl)NH2, NaOH; (iii) 6M H 2804, NaOH, Zn(OAc)2,

NaOH; (iv) 3 ,4—dinitrobenzoic acid.

104

that amide transfer to the nitrile proceeds efficiently only with Ni(II) in the porphyrin ring;
no appreciable reaction is observed for free base or zinc porphyrins. The starting material
Ni‘lOEP aldehyde, 10,80'8' was converted to the corresponding acrylonitrile by using a
modified Wittig reaction employing diethyl cyanomethylphosphonate. Amidine 3 was
obtained by reaction of the nitrile 1 1 with Wienrebs reagent, methyl aluminum(III)
chloroamide, in toluene. This reaction mixture is stirred at 80 °C for 2 days under an argon
atmosphere and then quenched upon addition to a chloroform/ silica gel slurry. The
hydrochloride salt of 12 is released by thoroughly chloroform/methanol (20:1 v/v), and it
was isolated as a crystalline solid by removing the solvent mixture under reduced pressure.
Addition of 5 equivalents of NaOH to a methanol solution of the hydrochloride salt of l 2,
and trituration with water, yielded 12 as a pink solid. This product was washed with
water, and dried by rinsing with copious amounts of anhydrous ethyl ether. Removal of
Ni(II) and insertion of Zn(II) was accomplished with standard procedures to yield 1 3

quantitatively.82

2. Solution Salt Bridge Association

The binding constants of 12 and 13 to carboxylates were nearly identical. Figure
40 shows the major 'H NMR spectral changes resulting for the association of 12 to the
electron acceptor, 3,4-dinitrobenzoic acid (3,4—DNBCOOH) in DMSO—d6. The positions
of the nearly coincident singlets for the porphyrin meso protons at 9.6 ppm and the doublet
of the vinylic proton at 10.3 ppm are concentration invariant. Most striking is the behavior
of the amidinium protons, which appear as broad singlets. A concentration—dependent
downfield shift of >1.0 ppm is observed for the protons involved in hydrogen bonding to
the rinsing the carboxylate (sometimes called NHaXS") whereas the chemical shift of the
adjacent protons (Nch), which are not bound to the carboxylate, remains virtually

unchanged at 9.25 ppm over all concentrations.

 

105

 

. “-- / ‘ ‘~ ‘
M, MMM \W'Aeh ' _ mun/r w "W

n ‘ . . I‘
I L ' '
Mil \' I “y' ‘\ ‘ MIP/
ymwnfimw. .4 ‘W‘nflab )M“—fik—-W

n
7
o
l
.

1
1

I . I " - '
. , ~ 1 J -.
, q I" \\ ‘ - 1-
N 1. , I _ .' o,‘ \
l~..~w~fi-w~m-mfi»«~amps-11W?“ ‘9"Ir-NthwrhManx-«wrkvjvww 1 ’-‘ ”W- -.¢w-i-*r,~.~'~’t"‘w ”'1

I

4” ‘\ N." I

. . I \ I” ‘\
. - ‘J - ,- ~. .I ‘ . - w” ~
nWWMwWJ~m‘4-‘fi ~f#-’ “M, .‘N‘WW \f -_WW' ' “'1
l
'1' ‘. _' . .
' ' ' j 5* «49/ /’ "-
. ‘ I '- o. .4 . . so
,4. .0 . »_..- "W «v I ‘.’ _ | . M ' ”K
lm‘",~hdfi «ea—A - *Wwww. W "MAW va-rq—w' 1 \ w. ,1

: '. .
0. .\ \ I ‘ . l \
. ‘ ‘ .. .I“ I .-l' "
WW4"? MW «enviirwiunrir-‘V '1 “‘4

I
A. . d ‘ f
“‘ W'MR‘thfi-‘W "WW" wry-W" '

Mownqnv .t mammal -.-q~*.1-r.Io~WHM‘*‘-" "ff"

I
I I I

If, \M M, 5“...»

7 I ‘ "My-h M", a.
WMVMM‘nVVfiM‘wfiW 3" “H'WMMI.¥WW m“ I -n...b. “- "

 

 

llllllllllLLllLLLbllllllllllllll

10.75 10.50 10.25 10.00 9.75 9.50 9.25
ppm

 

Figure 40. Selected 'H NMR spectra of 12 in the presence of one equivalent of 3,4—
DNBCOOH for concentrations of 1.7, 2.6, 3.5, 4.3, 6.1, 13.0 and 26.0 mM in DMSO—d6
(bottom to top). The spectral range captures the amidinium and vinylic protons of the
interface and the meso protons of the octaethyl porphyrin ring.

106

Figure 41 plots the chemical shift of N H” versus concentration and the non-linear

least squares fit of these data34 to Eq. 4.1 yields a binding constant (K ) of 3500 M" in

DMSO—d6 at 23.5 °C. Titration of the chloride salt of 12 with the tetramethylammonium
salt of the carboxylate yields similar overall behavior for the chemical shifts of the
amidinium protons, but the overall binding constant is reduced owing to the presence of the
increased ionic strength resulting from the presence of an external salt. As expected on the
basis of substituent effects in the association of zwitterionic hydrogen-bonded interfaces,85
the binding constant increases for benzoate owing to the removal of the electron
withdrawing NO2 groups from the acceptor.

The observed binding constant for the amidinium-carboxylate salt bridge is
appreciable. In the context of Jorgensen’s classification of hydrogen bonded interfaces,86
the association of 12 and 13 to carboxylates is driven by two favorable secondary
electrostatic interactions. However, the favorable electrostatic charge within the salt bridge
also plays a significant role in its formation. Ureas bound to carboxylate also exhibit two
favorable secondary electrostatic interactions, but the association constants of these systems
in DMSO are reduced to ~150 M".87 The thirty-fold increase in binding of amidinium to
carboxylate most probably reflects the stabilization imparted by the molecular recognition of

the negatively charged carboxylate by the positively charged amidinium, as opposed to the

neutral urea.

 

 

 

 

2
1 1 1
8°“ - 5N" + Asmax[1+ 2[AmCa]oKalssoc _ \/[2[AmCa]oKassoc] + [[AmCa]oK..m0c JJ (4-1)

107

10.8

 

—-L

.0

o:
l

—L

.0

is
I

10.0 —

Chemical Shift / ppm
3
N
I

9.8 L—

 

 

 

9.6 l L l 1 l 1 l L I l I l 1 1

Figure 41. Plot of the chemical shift of the amidinium protons hydrogen bonded to
carboxylate versus concentration of 12 and 3,4—DNBCOOH.

108

3. Photophysical Behavior
The singlet excited state of 13 luminesces in CH2C12 at 25 °C thereby permitting
the excited state reactivity of studies.88 The luminescence of 13 when complexed to benzoic

acid decays monoexponentially with a lifetime of 434 picoseconds 0s = 580 nm, ltd“ =

760 nm). When benzoic acid is replaced by the well-known electron acceptor 3,4—
DNBCOOH, the luminescence maximum is unchanged, however, the luminescence
intensity is quenched and the excited state lifetime is diminished (1: = 327 picoseconds).
This observation is consistent with electron transfer from the singlet excited state of 4 to
3,4-DNBCOOH mediated by the amidinium-carboxylate salt bridge. A unimolecular rate
constant for photoinitiated PCET in 5 is 7.5 ><108 s" as calculated from the emission
lifetime decay kinetics. The results reported herein show that a salt bridge can not only pre-

organize a donor and an acceptor but that it can also mediate the electron transfer chemistry

between the donor-acceptor pair.

C. Materials and Methods

Synthesis of Ni"OEP-acrylonitrile (11). A solution of diethyl
cyanomethylphosphonate (0. 2 4 g, 1.36 mmol) in freshly distilled THF (5 mL) was added
dropwise through a cannula to sodium amide (0.053 g, 1.36 mmol) under an argon
atmosphere. The resulting solution immediately turned a deep red color, and was stirred for
an additional 4 hours, afterwhich a solution of Ni"OEP aldehyde, 10, (0.20 g, 0.32 mrnol)
in THF (10 mL) was added through a cannula. This reaction mixture was stirred for 24
hours under an argon atmosphere at 60 °C. The solution was then diluted with
dichloromethane (100 mL) and washed with water (3 x 100 mL). The organic solvents
where removed under reduced pressure, and the resulting solid was purified by silica gel

column chromatography, eluting with chloroform/hexane (1:1 v/v) to yield 11 (0.11 g,

 

109

53%): MS: m/z 641.4; 1H NMR (CDC13): 5 9.76 (1H, d), 9.45 (3H, s), 4.61 (1H, d),
3.81 (16H, m), 1.76 ppm (24H, m).

Synthesis of Ni"OEP amidine (12). A solution of methyl aluminum(III)
chloroamide in toluene (1.0 M, 5 mL, 5 mrnol) was added to 11 (0.25 g, 0.39 mrnol)
under an argon atmosphere. This reaction mixture was stirred at 80 °C for 2 days under an
argon atmosphere and then quenched upon addition to a chloroform/ silica gel slurry (20
mL). The hydrochloride salt of 12 was collected by thoroughly rinsing the silica gel slurry
with chloroform/methanol (20:1 v/v, 50 mL), and it was isolated as a crystalline solid by
removing the solvent mixture under reduced pressure. Addition of 5 equivalents of NaOH
to a methanol solution of the salt, and trituration with water, yielded 12 as a pink solid.
This product was washed with water, and dried by rinsing with copious amounts of
anhydrous ether; to yield 12 (0.23 g, 90%): MS: m/z 658.3; ‘H NMR (methanol-d4): 5
10.09 (1H, d), 9.40 (1H, s), 9.38 (2H, s), 5.29 (1H, d), 3.73 (16H, m), 1.65 ppm (24H,
m).

Synthesis of Zn"OEP-vinyI-amidine (13). A sample of 12 was placed onto
a fine frit, and concentrated sulfuric acid was added until 12 passed through the frit.
Crushed ice was added to the mixture, then sodium hydroxide was added until the solution
becomes basic. The free base amidine, free base porphyrin was extract into
dichloromethane. A saturated solution of zinc acetate is added, and resulting solution is
stirred for 30 minutes. Removal of Ni(II) and insertion of Zn(II) was accomplished with
standard procedures to yield 13 quantitatively.” ES—MS: [M]1+ 666.4; 1H NMR (13 with
benzoic acid in CDzClz with a trace of methanol-d4): 5 10.52 (1H, d), 10.12 (2H, 5), 10.02
(1H, s), 8.02 (2H, d), 7.45 (3H, m), 6.01 (1H, d), 4.05 (16H, m), 1.92 ppm (24H, m).

Synthesis of [Zn"OEP-vinyl-amH*][3,4-dinitrobenzoate] (14). An
equivalent each of 13 and 3,4—dinitrobenzoic acid are placed in dichloromethane. Hexane

is added until a precipitate forms. ES—MS: [M]1+ 877.8; lH NMR (DMF-d7): 5 10.97 (1H,

110

d), 10.96 (2H, 3), 10.18 (2H, 5), 10.15 (1H, s), 9.56 (2H, s), 8.68 (1H, s), 8.54 (1H,
d), 8.39 (1H, d), 6.40 (1H, d), 4.10 (16H, m), 1.95 ppm (24H, m).

 

Chapter 5
PULSE RADIOLYSIS EXPERIMENTS TO PROBE PCET

A. Introduction

The pules radiolysis experiment complement photochemical methods to study ET.
Pulse radiolysis experiments aimed at assessing the effects of amide hydrogen bonds on ET
in D/A modified polypeptides accentuate the complexity of the amide hydrogen bond.89 The
observed kinetic behavior with respect to pulse radiolysis initiated ET reactions in D/A
modified polypeptides showed multiple decay pathways. The kinetic behavior was best
explained by the presence of multiple hydrogen bonded configurations of the polypeptide
giving rise to statistical distribution hydrogen bonded structural motifs separating D and A,
for a given D/A modified polypeptide sequence. Subsequent photoinitiated ET studies on
D/A modified polypeptides in conjunction with circular dichroism experimental results
showed conclusively the relation between conformation and folding and the observed ET
rates. The effects of the orientation of the amide hydrogen bond in the polypeptide
backbone, in particular when the polypeptide backbone is folded into an Ot-helix is evident
in a photoinitiated ET rate measurements bonds, where the rate for ET correlates to the
direction of the helix dipole created by the amide hydrogen bonds forming an a—helix. 9°

One resolution to the problem of multiple amide hydrogen bonding modes has been
to make the switch to a different naturally relevant hydrogen bonding motif. To this end the
hydrogen bonding interaction of the amidinium-carboxylate salt bridge has been
successfully employed to structurally juxtapose D/A assemblies for ET measurements. For

the amidinium—carboxylate salt bridge shows greater structural specificity in that there exists

111

 

112

only one major structural conformation in a two—point primary hydrogen bonding motif.
Indeed, mono-phasic kinetic behavior is observed in D/salt-bridge/A system, where D is a
spyrofluorene group, the hydrogen bond is the amidinium-carboxylate salt-bridge, and A is
a 2-methylnaphthalene. For the first time the pulse radiolysis technique has been applied to
assess the rate for an intramolecular ET reaction in a D-salt bridge-A system, as depicted

schematically in Eq. 5.1.

D“‘-{salt bridge}-A ——> D—{salt bridge}-A"1 (5.1)

B. Background

1. Measurement of Equilibria Between Radical Anions
Absorption spectroscopy can be applied to measure equilibria in solutions of two

different radical anions of aromatic organic molecules D and A, as shown in Eq. 5.2.”

D"1 + A : D + A"1 (5.2)

The equilibrium represented in Eq 5.2 consists of four components at some
equilibrium concentration of D, A and their corresponding radical anions D"1 and A". The
equilibrium constant (Km) for Eq. 5.2 is defined by Eq. 5.3, and consists of the ratios of

the equilibrium concentrations of the different radical anions and neutral molecules.

K =10 llA"‘l 53
°" 010”] "’

 

The application of the absorption technique to measure Keq requires that for a given
detection wavelength one of the radical anions should have an appreciably greater extinction
coefficient than the other, for greater sensitivity, and that the neutral molecules should have

extinction coefficients near zero at a given wavelength. If these conditions are met, then Eq.

113

5.3 can be rewritten in terms of the absorbance intensity at a given wavelength for D"1 (ID~
1), A"l (IA-u) and an equilibrium mixture of the two, leg. This expression assumes that D

has a greater extinction coefficient than A and that [A] >> [A'"] and that [D] >> [D’"].

_(|Dk1- leq) [A]
(leg—IA... ) [o] (5.4)

 

eq

With Keq in hand it is possible to calculate the relative free energy change (AG°) for
the equilibrium shown in Eq. 5.2 by simply substituting the calculated Keq into Eq. 5.5.
For equilibria measurements employing the absorption spectroscopy technique, the radical
anions of interest can be generated in solution by treatment with of known concentrations
of aromatics with alkali metal. For the radical anions in the presence of alkali metals there

needs to be some corrections made for ion pairing.

AG° = —RTln(Keq) (5.5)

Values of Keq and the corresponding AG° determined spectrophotometrically for
one such equilibria are displayed in Eq. 5.6. This technique is useful for measuring ch out
to values of about 150, as dictated by how reliably the ratio of [D] to [A] can be accurately
adjusted using standard volumetric techniques.

Using this technique, the equilibria for such systems allows for the determination of
the relative AG° between a series radical anion (D'"/A)/(D/A’") pairs including pyrene and
anthracene. In fact, this technique is generally applicable to many organic radical anions for
which the electronic absorption spectra of their corresponding radical ions are known.92 In
one system the radical anion of pyrene (Pyr) acts as a D and anthracene (Ant) is the A, as
determined spectrophotometrically by comparing mixtures of solutions of their

corresponding radical anions (Pyr'", km = 492 nm, e = 105 M"cm"; Ant’", Am,“ = 370

 

114

nm, e = 5x104 M"cm"). Similar techniques are employed using the pulse radiolysis
technique; yet in the pulse radiolysis technique there is no alkali metal to compete in side
pairing reactions. The spectrophotometric technique and the pulse radiolysis technique give
similar results for the biphenyl (D)/naphthalene (A) system with their corresponding
radical anions (D’"/A)/(D/A'") for which AG° is determined to be —43 mV.

Pry"1 + Ant ———’ Pyr + Ant'“1 (K =74; AG° =—114 mV) (5.6)
(————- eq

2. Kinetic Measurements Employing the Pulsed Radiolysis Experiment

The time resolution for kinetic measurements in the pulse radiolysis experiment
stems from the ability to generated very short temporal pulses of electrons, 30 picoseconds,
from the 20 MeV linear accelerator (Linac) at Argonne National Laboratories. When high
energy pulses of electrons pass through media such as organic solvents, the solvent
becomes ionized. This is not due to depositing electrons into the solvent from the Linac
pulse, but rather due to electrons being ejected from the solvent molecules as the Linac
pulse passes through the solvent. The ejected electrons rapidly become solvated or are

solvated electrons represented as (e’") and solvent radical cation hole, [solvent‘*'] (Eq.

sol

5.7).

solvent —> (e"‘) + [solvent‘”] (5.7)

sol

The formation of solvated electron can be followed by transient absorption
spectroscopy, for the (e’")so,’s absorbs light throughout the visible spectrum and into the
infrared spectral region. In highly pure neat, oxygen free N-methylpyrrolidone (N MP), the
NMP solvated electron, (e'")NMP, has a lifetime of about 6 microseconds. For the
experiments described here (e'")so,’s are produced at a concentration of about 10
micromolar. However, operation of the Linac allows for adjustment of the size of the Linac
pulse giving flexibility to the concentration of (e'")so,. Stated simply, the concentration of

(e'")sol is dose dependent.

115

Once (e“')sol is produced in solution it can react through a variety of channels. Most
importantly for the work described here is that it reacts as a powerful reductant and can
attach itself to organic aromatic molecules to yield organic aromatic radical anions, Eq. 5 .8.
The rate constant for this reaction approaches the rate of diffusion in solution (10'0 M"s").
Restated, whenever (e‘")sol comes in contact with an organic aromatic group in solution it
attaches itself to the organic group producing an organic aromatic radical anions. The
formation of aromatic radical anions is conveniently followed by transient absorption
spectroscopy, as different aromatic organic radical anions are readily distinguished by their
absorption spectra. The rate for the electron attachment step as illustrated in Eq. 5.8 typical
aromatic around 2x10'° s"M" as determined by transient absorption spectroscopy.

(9“)

+ aromatic ——> aromatic"1 (5.8)

sol

When a solution contains an equimolar mixture of two different aromatic molecules,
a D and an A, the solvated electron, at 10 micromolar, will react without discrimination to
produce a non-equilibrium mixture of D"l and A'" at a concentration of approximately 5
micromolar each on the time scale of the Linac as shown in Eq. 5.9. These radical anion
species are generated on roughly the diffusional time scale, the time scale of the pulse
radiolysis technique (10l0 s"), which sets the time scale for which kinetic information can
be obtained about ensuing reactions between organic aromatic molecules and radical

anions.

(e"‘)so, (1011M) + D + A ——> D"1 (511M) + A"1 (5 1.1M) (5.9)

The radical anions generated using the pulse radiolysis technique can go on to react
or exchange electrons with other organic aromatic molecules present in solution as the
solution as a hole comes to equilibrium. That is, the radical species in solution achieve a
steady state. This process reveals kinetics information corresponding to electron transfer

form D"l to A as shown in Eq. 5.10. Again, this process is also followed by transient

116

absorption spectroscopy, as D‘" and A’" are easily distinguishable by their corresponding
absorption spectra. Changes in time in the color of the solution indicate changes in
electronic structure due to ET. For experiments described here the growth and decay of
colored aromatic radical anion species forms the basis by which ET between different

aromatic species is investigated.

D"1 -I- A ———> D -I- A"1 (5.10)

The above discussion covers the approach used to study intermolecular ET in
solution or frozen glasses containing a mixture of D and A molecules. When within a single
molecule there exists the presence of both D and A it is possible to assess the rate for

intrarnolecular ET, Eq. 5.1 1.

D"‘—spacer—A ———-9 D—spacer—A“1 (5. l 1)

Shown in Figure 5.1 is a scheme describing the cascade of events leading from
irradiation a solution of a D—spacer-A molecule to radical anions, which can then go on to
participate in intra- or intennolecular ET. The ensuing ET reactions are again predicated on
the fact that there exists some driving force for ET and that the initial radical anions are
formed in non—equilibrium mixture. For a 10 micromolar yield of solvated electron in the
presence of D—spacer—A, the initial electron attachment step yields 5 micromolar D’"-

spacer-A and 5 micromolar D—spacer—A’", Eq. 5.12.

(9“)50, (10 1M) + D—spacer—A ————>
(5.12)
D“‘—spacer—A (5 uM) + D—spacer—A“1 (5 uM)
As the D—spacer—A system comes to equilibrium via electron flow, there are two
major pathways that electron flow may take place between radical anions and aromatic

groups, by either intermolecular ET as is the case for non covalent D/A systems, or by

intrarnolecular ET. For the case of purely intermolecular ET, the observed rate constant,

117

kobs, has a linear dependence in the [A] for the rate of disappearance of D’" or the
appearance of A'", where the intercept is zero for the dependence of kobs on the
concentration of A. However in the case of the D—spacer—A systems, kobs shows quite
different behavior. A plot of the dependence of the concentration on kobs displays a non-
zero intercept as the concentration of D—spacer—A approaches zero. This is because in D—
spacer—A systems, kobs is the sum of inter- and intrarnolecular ET. In Eq. 5.13 is the rate
law for electron flow in D—spacer—A systems. In fact, to detennine the rate for
intrarnolecular ET, kmm it is necessary experimental requires one to vary the concentration
of D—spacer—A to determine the intercept thereby determining the rate for intrarnolecular

electron transfer.

k obs = k intra +[ID—$433991“Al'k inter (5- 13)

3. Hydrogen Production in the Pulse Radiolysis Experiment

A few words are needed to address the issue of hydrogen atom production in the
pulse radiolysis experiment. Indeed the solvated electron can react with protons to generate
the hydrogen atom. In water, under acidic conditions (pH 1), (e’")so,, the hydrated
electron, reacts with hydronium ion in a diffusion controlled reaction to produce the
hydrogen atom. For the experiments presented here however, this reaction is hardly worth
mentioning. This due to the competition of electron attachment reactions as described in
earlier sections. If the concentration of (e’")sol is 10 micromolar, then it follows that any
proton source or strong acid from the solvent hole is limited to a concentration of 10
micromolar (pH 6). Recalling that the concentration of aromatics used for ET are typically
10 millimolar, 1000 times more than the highest possible concentration of acid, combined
with the fact that the electron attachment reaction is also diffusion controlled, one can

estimate that the lowest total concentration of aromatic radical anions produced will be 9.99

118

solvent

D-A

Argonne Linac Pulse
(20 MeV, 30 picoseconds)

 

 

(9‘1 )sol

solvated electron

[solvent]'~~1

 

 

 

 

D~1—.A 7 e a-»

Figure 42. Electron flow in the pulse radiolysis experiment for mixed aromatic D/A
compounds.

119

micromolar (99.9%) and that the highest possible concentration of the hydrogen atom
produced will be 0.01 micromolar (00.1%). Thus, reaction of (e'")sol to yield the hydrogen
atom is negligible. Neglected in this arguement is the fact that the amidinium protons may
act as a weak acid and protonate the solvated electron. If this is the case hydrogen atom
production should not compete well with electron attachment. Experiments in fact show

this as presented in the following section.

C. Results and Discussion

1. Synthesis

The first synthetic challenge in preparing a D—salt bridge—A assembly for pulse
radiolysis was to choose appropriate groups to form a rigid structural framework from
which either a Ca functional group or an Am groups might be attached. The D/A organic
aromatic groups for the systems in this study are fluorene (F1) and naphthalene (Na), which
act as D and A, respectively as their absorption spectra corresponding to radical anions are
known as well as the free energy change for of 62 mV. In this first study, a spyrofluorene
molecule is functionalized with Ca to yield the D portion, FlCa of a D/A system, while
modification of 2—methylnaphthalene with Am yields the acceptor portion AmN a. The two
halves of the D/A system yield a D—salt bridge—A molecule {FlCazAmNa} for which
intrarnolecular salt bridge mediated electron transfer is assessed for the first time.

Displayed in Figure 43 are the synthetic steps for a series of compound for the
measure of ET by the pulse radiolysis technique. The synthesis of the spyroketone was
reported by Closs in his studies on triplet transfer.93 In the work here, it was found that the
yield, isolation and purification of the product 1-spyrofluorenecyclohexyl-4-ketone greatly
improves with the substitution of trifluoroacetic acid for sulfuric acid in the Lewis acid
catalysis intrarnolecular Friedel-Craft spyrocyclization step. The spyroketone is smoothly

converted to the enol ether using a modified Wittig reaction. Hydrolysis of the enol ether

 

120

yields fluorospyroaldehyde. The aldehyde is converted to the nitrile, which is in turn
converted to the carboxylic acid by acid hydrolysis. The salt bridge complexes
{FlCazAmCy}, {FlCazAmNa}, and {CyCa:AmNa} where prepared by simple precipitation
from aqueous solution. The ketone is converted to an enol ether through the action of the
instant ylide.94 The first challenge was to find the appropriate groups that contain a rigid
carbon frame work from which either a carboxylate functional group or an amidinium
functional group might be attached The fluorene functional group was chosen as the donor
portion of the system described here. Conversion of the enol ether to the aldehyde proceeds
efficiently with the action of acid catalysis in dichloromethane with perchloric acid.
Protonation, and loss of methanol followed by rearrangement leads to the aldehyde group.
Reaction of the aldehyde group with hydroxylamine hydrochloride in wet refluxing formic
acid gives the carboxylic acid group. The reaction most likely proceeds through a nitrile
intermediate, which is hydrolyzes under acidic conditions in the presence of water to the
corresponding amide and then to the corresponding carboxylic acid.

The formation of the salt bridge complexes {FlCazAmCy}, {FlCazAmNa}, and
{CyCa:AmNa} is quite straight forward. As described in previous section, the
corresponding carboxylic acids are dissolved in a minimal amount of water by titrating in
one equivalent of the strong base, DBU. Addition of a separate aqueous solution of the
amidinium chloride salt produces an immediate precipitation. Upon isolation and
characterization by mass spec and NMR it is shown that the precipitate is the 1:1
amidinium—carboxylate salt bridge, which can be isolated by simple filtration.

This method of preparing salt bridge complexes is applied here to a series of
complexes in an attempt to form aromatic complexes linked by a the non-covalent salt
bridge to probe intrarnolecular electron transfer. There are two classes of complexes that

were prepared in this study by this method.

121

©‘i (i)

 

 

 

{FICazAmNa}
o0 t
H o Cle HéN> (iv) (:9 O----H- 6N3
O-H H-NFEQ 0""H "ii—m
H
{FlCazAmCHy}

   

{CyCa:AmNa}

Figure 43. Synthesis of complexes for the study of electron transfer using the pulse
radiolysis technique, (1) instant ylide; (ii) perchloric acid, dichloromethane; (iii)
hydroxylamine hydrochloride, formic acid; and (iv) DBU, water.

122

2. Salt Bridge Mediated Electron Transfer

The system reported in this study display electron transfer through similar to fixed
distance electron transfer through rigid covalent spacers, but incorporates a salt bridge,
Figure 44. The donor for this study was the spyrofluorene molecule, used by Closs in his
studies of triplet transfer reactions. The spyro fluorene groups has several advantages, it is
very rigid, as seen by NMR and with respect to electron transfer, has a fewer degrees of
freedom, thereby decreasing the inner reorganizational energy. The acceptor in this case
was a 2-naphthylamidinium, which was reported by Garigapati.

In the pulse radiolysis experiments there are observed absorption at a given
wavelenght clearly inidicative of radical species that correspond to the radical anion of the
fluorene syro group. From previous studies is known that the potential for the spyro group
is around -2.6 volts. The absorption spectrum of the fluorene radical anion has been
reported by Shida and coworkers.

Some decay is observed for the transient at 680 nm in a 7.2 mM solution of
{FlCazAmCy} by pulse radiolysis experiment. Addition of {CyAmzCl} to 6.8 mM appears
to reduce the yield of the transient, but does not substantially increase the decay of the
radical anions. However addition of phenanthrene at 4.8 mM causes a normal decay of the
transient as would be expected for intrermolecular electron transfer from the {FICa:AmCy}
anion to biphenyl/phenanthrene at a potential of around ( 124 mV) or a driving force of This
gave a normal bimolecular ET quenching rate constant of 4.2x109 5" M". These results
suggest the the {FICa:AmCy} radical anion is fairly stable towards protonation by free
amidinium groups and that it is a potent electron donor and may be rewritten as {Fl‘
'Ca:AmCy} suggesting that the initial electron attachment step yields a fluorene localized

radical anion as suggested in Eq. 5.14.

(e“‘)sol + {FICa:AmCy} ———> {F|"‘Ca:AmCy} (5.14)

123

 

.0
N
I

{FICa:AmCy};-1

   
 
 
   

{FICa:AmNa}"1

.0
—L
1

 

 

 

 

Absorbance (650 nm)

L{\.C\3yCa:AmNa}"‘

{

 

i

 

 

 

I I

0 39 78 117 156

Time I nanoseconds

Figure 44. Typical decay profiles on the nanosecond timescale for radical anions of
{FlCazAmCy} (top), {FlCazAmNa} (middle), and {CyCa:AmNa} (bottom) generated via
the pulse radiolysis experiment in N MP and monitored at 650 nm.

124

In a control experiment it is shown that the the radical anion {Fl'"Ca:AmCy}
donates its electron to phenanthrene (Phen), as displayed in eq 5.15. The free energy
change for this process is has been estimated to be about 100 mV. The reaction is followed
by both the decay of transient {Fl’"Ca:AmCy} species appearing at 650 nm, or by the
growth of Phen’" with a characteristic absorbance at 1100 nm. The rate for the ET reaction

has a bimolecular rate constant of ki = 4.2 x 109 M"s", Eq. 5.9. This value in line with

met

the results of other bimolecular ET reactions. This control experiment helps to establish the

reducing power of {Fl'"Ca:AmCy}.

{Fl“‘Ca:AmCy} + Phen -———> {FICa:AmCy} + Phen"1 (5.15)

In the pulse radiolysis experiments involving {CyCa:AmNa} the concentration
range studied is from 0.21 to 10.77 mM. For each concentration studied the rate for
electron attachment from the solvated electron could be assessed in a global fit from the
decay data. From these data an electron attachment rate of km”: 2.0 x 1010 for the reaction
shown in equation 5.16. This rate is assessed by observing the decay of the solvated
electron which absorbs in the near infra red where the absorption due to radical anions in
negligible. From these data it appears that the electron attaches to the AmNa portion of the

salt bridge complex, and is nearly identical to electron attachment to naphthalene.

(e"‘)so,+ {CyCa:AmNa} ——> {CyCa:AmNa“‘} (5.16)

And in the mixed aromatic salt bridge system what appears to be normal electron
transfer is observed, Eq. 5.17. The ET rates, kobs follow a linear dependence on the
concenctration of {FICa:AmNa}, Figure 45, from which a fit to the rate law given in Eq.
5.13 yields the intra- and intermolecular ET rates (Km: 3(1) x 107 s"; k 2 x 109 M"s’

I).

inter

{FI"‘Ca:AmNa} ——> {F lCa:AmNa"‘} (5.17)

125

 

 

 

 

8—
.1
'1'
(D
E2 67
\U)
Q -1
O
x
4...
2 l I I l I I l I T I I I I I I T l l I
0 5 10 15 20

[{FICa:AmNa}] / mM

Figure 45. Observed ET rates in {FICa:AmNa} measured at 1.1, 4.4, 7.6. and 18.1
mM.

126

D. Materials and Methods

Preparation of N-Methylpyrrolidone (NMP). Reagent grade NMP (500
mL) was placed in a single neck round bottom flask, and calcium hydride (50 g) was
added. The solution was distilled at 30 mmHg. A 300 ml fraction was collected boiling at
40 °C, at 30mmHg. This was combined with sodium and terphenylene, and again purified
in a second distillation step.

Synthesis of Fluorene-9-spyro-4-cyclohexanone. A slurry of dry
trifluoroacetic acid and the protected 2-biphenyl-1-hydroxylcyclohexan-4-one are allowed
the to stir at room temperature for 3 hours. Upon addition of the a Sm the solution turns a
deep blue color, after 15 minutes or so the solution becomes a deep red color, with no
further color change after which the reaction is quenched by careful addition to iced water.
The ice water solution is neutralized with sodium carbonate, 200 mL of DCM is added and
the extracts are washed with water (3 x 100 mL). The organic extracts are dried with
anhydrous magnesium sulfate, filtered, and the solvent is removed under reduced pressure,
yielding a colorless oil. Addition of a minimal amount of hexane initiates a dramatic
crystallization of the product, which is isolated by filtration in about 60% yield.

Synthesis of Fluorene-9-spyro-4-carboxaldehyde. A solution of instant
ylide (5.0 g) was generated in freshly distilled toluene under an inert atmosphere. After
refluxing for 1 h, the solution was cooled to room temperature, and the ketone (1.0 g), in
20 ml of dry toluene was added through a canula. This solution was refluxed overnight.
The next day it was taken to dryness, and passed through a short column of silica gel to
remove phosphorus. The product was collected, the solvents were removed by rotary
evaporation, and the product was taken into dichloromethane and washed with water. The
product was dried with magnesium sulfate, and the solvent was removed by rotary
evaporation. The product crystallizes with the addition of hexanes. The enol ether was not

isolated. The product was hydrolyzed by refluxing in 50 ml of a solution of THF/HzO/HCl

127

in a ratio of 5.0:l.0:0.5 for 1hr. The hydrolysis was followed by TLC, 2:1
dichloromethane/hexanes. And the product aldehyde has an Rf of 0.6, 70%. The product
gave satisfactory characterized 'H NMR (CDC3): 5 9.92 (1H, s), 7.75 (2H, t), 7.67 (1H,
d), 7.51 (1H, d), 7.31 (4H, m), 2.64 (1H, s), 2.05 (5H, m), 1.73 (2H, m), 1.55 ppm
(1H, 8).

Synthesis of F luorene-9-spyro-4-carboxylic acid. The products from
above where combined with an excess of Hydroxylarninehydrochloride and refluxed in
formic acid overnight. The product was quenched with 10% sodium hydroxide, washed
dichloromethane acidified with hydrochloric acid, which produced a white precipitate. This
precipitate was extracted with dichloromethane. Removal of solvents under reduced
pressure yielded the acid. MS: m/z 278.2; 1H NMR (CDC13): 5 7.83 (1H, d), 7.72 (2H,
m), 7.33 (5H, m), 2.72 (1H, m), 2.23 (4H, m), 2.01 (2H, m), 1.66 ppm (2H, d); 13C
NMR (CDC13): 5 182.17, 152.71, 151.41, 140.01, 139.33, 127.29, 127.17, 126.75,
125.58, 122.89, 120.07, 119.74, 49.27, 42.09, 34.33, 25.02 ppm .

Synthesis of Cyclohexylamidinuium chloride {CyAm:Cl}. This
compound was prepared by the litrature procedures employing sodium methoxide in
methanol. MS: m/z 126.2;'H NMR (DMSO-d6): 5 8.98 (2H, s), 8.82 (2H, s), 2.43 (1H,
t), 1.64 (7H, m), 1.22 ppm (3H, m).

Synthesis of [Cyclohexylamidinuium][fluorene-9-spyro-4-
carboxylate] ({FICa:AmCy}). The acid (0.062 g, 0.22 mmol) above was combined
with DBU (0.030 g, 0.20 mmol) in water. This solution was added to an aqueous solution
of cyclohexylamidinium chloride. A white precipitate formed immediately. The solution
was mixed thoroughly, then filtered. The filtrate was then washed with a minimal amount
of water, air dried then washed with an excess of anhydrous ether, to afford the product
(0.069 g, 0.17 mmol) in 85% yield. 'H NMR (DMSO-d6): 5 11.60 (2H, br), 8.25 (2H,
br), 7.82 (3H, m), 7.53 (1H, d), 7.32 (4H, m), 2.38 (2H, m), 1.90 (11H, m), 1.32 ppm
(7H, m).

128

Synthesis of [2-Naphthylacetamidinium][fluorene-9-spyro-4-
carboxylate] ({FICa:AmNa}). The same procedure was used as for {FICa:AmCy}
described above. The product {F lCa:AmNa} was obtained in 93% yield. MS: m/z 278.2
(CaFl), 183.1 (NaAm); 'H NMR (DMSO-d6): 5 10.45 (br), 7.60 (7H, m), 7.53 (4H, m),
7.31 (4H, m), 3.84 (2H, s), 2.26 (1H, t), 1.95 (6H, m), 1.38 ppm (2H, d).

Synthesis of 2-Naphthylacetamidinuium cyclohexylcarboxylate
{CyCa:AmNa}. The same procedure used for {F lCa:AmCy} described above is used
here. The product {CyCa:AmNa} was obtained in 20% yield. 'H NMR (DMSO-d6): 5
10.25 (br), 7.90 (4H, m), 7.51 (3H, m), 3.84 (2H, s), 1.60 (6H, m), 1.18 ppm (5H, m).

List of References

Lehn, J. -M. Supramolecular Chemisty, VCH, Weinheim, 1995.

(a) Pranata, J.; Wierschke, S. G.; Jorgensen, W. L. J. Am. Chem. Soc. 1991,
113, 2810. (b) Jorgensen, W. L.; Pranata, J. J. Am. Chem. Soc. 1990, 112,
2008.

(a) Honig, B.; Nicholls, A. Science 1995, 268, 1144; (b) Gunner, M. R.;
Nicholls, A.; Honig, B. J. Phys. Chem. 1996, 100, 4277.

(a) Russell V. A.; Etter, M. C.; Ward, M. E. J. Am. Chem. Soc. 1994, 116,
1941. (b) Russell V. A.; Etter, M. C.; Ward, M. E. Chem. Mater. 1994, 6, 1206;
(c) Russell V. A.; Ward, M. E. Chem. Mater. 1996, 8, 1654; (a) Russell, V. A.;
Evans, C. C.; Li, W.; Ward, M. D. Science 1997, 276, 579.

Pavletich, N. P.; Pabo, C. O. Science 1991, 252, 809.

(a) Puglisi, J. D.; Chen, L.; Frankel, A. D.; Williamson, J. R. Proc. Natl. Acad.
Sci. USA 1993, 90, 3680-3684; (b) Puglisi, J. D.; Tan, R.; Calnan, B. J.;
Frankel, A. D.; Williamson, J. R. Science 1992, 257, 76.

(a) Ganem, B Angew. Chem. Int. Ed. Engl. 1996, 35, 937; (b) Lee, A. Y.;
Karplus, P. A.; Ganem, B.; Clardy, J. J. Am. Chem. Soc. 1995, 117, 3627.

(a) Chook, Y. M.; Ke, H.; Lipscomb, W. N. Proc. Natl. Acad. Sci. USA 1993,
90, 8600; (b) Chook, Y. M.; Gray, J. V.; Ke, H.; Lipscomb, W. N. J. Mol. Biol.

129

10.

ll.

12.

13.

14.
15.

16.
17.

18.

19.

20.
21.

130

1994, 240, 476.

Albert, J. S.; Goodman, M. 8.; Hamilton, A. D. J. Am. Chem. Soc. 1995, 117,
1143.

Gébel, M. W.; Bats, J. W.; Diiner, G. Angew. Chem. Int. Ed. Engl. 1992, 31,
207.

Terfort, A.; Kiedrowski, G. von Angew. Chem. Int. Ed. Engl. 1992, 35, 654.
(a) Closs, G. L.; Calcaterra, L. T.; Green, N. J.; Penfield, K. W.; Miller, J. R. J.
Phys. Chem. 1986, 90, 3673; (b) Closs, G. L.; Miller, J. R. Science 1988, 240,
440.

Miller, J. R.; Beitz, J. V.; Huddleston, R. K. J. Am. Chem. Soc. 1984, 106,
5057.

Liang, N.; Miller, J. R.; Closs, G. L. J. Am. Chem. Soc. 1990, 112, 5354.
Miller, J. R.; Paulson, B. P.; Bal. R.; Closs, G. L. J. Phys. Chem. 1995, 99,
6923.

Wasielewski, M. R.; Niemczyk, M. P. J. Am. Chem. Soc. 1984, 106, 5043.

(a) Wasielewski, M. R.; Gaines 111, G. L.; O’Neil, M. P.; Svec, W. A.;
Niemczyk, M. P. J. Am. Chem. Soc. 1990, 112, 4559; (b) Wasielewski, M. R.;
Gaines 111, G. L.; O’Neil, M. P.; Svec, W. A.; Niemczyk, M. P. Mol. Cryst. Liq.
Cryst. 1991, 194, 201.

Opperman, K. A.; Mecklenburg, S. L.; Meyer, T. J. Inorg. Chem. 1994, 33,
5295.

Ramirez, B. E.; Malmstrfim, B. G.; Winkler, J. R.; Gray, H. B. Proc. Natl. Acad.
Sci. USA 1995, 92, 11949.

Brzezinski, P. Biochemistry 1996, 35, 5611.

(a) Graige, M. S.; Paddock, M. L.; Bruce, J. M.; Feher, G.; Okamura, M. Y. J.

22.

23.

24.
25.

26.

27.

28.

29.

30.

31.
32.

131

Am. Chem. Soc. 1996, 118, 9005. (b) Okamura, M. Y.; Feher, G. Annu. Rev.
Biochem. 1992, 61, 861.

Takahashi, H; Maroti, P.; Wraight, C. In Electron and Proton Transfer in
Chemistry and Biology; Muller, A., Ratajczaks, H., Junge, W., Diemann, E., Eds.;
Elsevier: Amsterdam, 1992; p 219.

Babcock, G. T.; Barry, B. A.; Debus, R. J.; Hoganson, C. W.; Atarnian, M.;
McIntosh, L.; Sithole, 1.; Yocum, C. F. Biochemistry 1989, 28, 9557.

Burgess, B. K.; Lowe, D. J. Chem. Rev. 1996, 96, 2983.

(a) Pilato, R. S.; Stiefel, E. I. In Bioinorganic Catalysis; Reedijk, 1., Ed.; Dekker:
New York, 1993, p 131. (b) Stiefel, E. I. In Molybdenum Enzymes, Cofactors, and
Model Systems; ACS Symposium Series 535; Stiefel, E. I., Coucouvanis, D.,
Newton, W. E., Eds.; American Chemical Society: Washington, DC, 1993, p 1. (c)
Stiefel, E. 1. Proc. Natl. Acad. Sci. USA 1973, 70, 988.

Feig, A. L.; Lippard, S. J. Chem. Rev. 1994, 94, 759.

Wallar, B. J.; Lipscomb, J. D. Chem. Rev. 1996, 96, 2625.

Solomon, E. 1.; Sundaram, U. M.; Machonkin, T. E. Chem. Rev. 1996, 96,
2563.

(a) Uhlin, U.; Eklund, H. Nature 1994, 370, 533. (b) Nordlund, P.; Eklund, H. J
Mol. Biol. 1993, 232, 123.

Ekberg, M.; Sahlin, M.; Eriksson, M.; Sjoberg, B.-M. J. Biol. Chem. 1996, 271,
20655.

Averill, B. A. Chem. Rev. 1996, 96, 2951.

(a) Winkler, J. R.; Nocera, D. G.; Yocom, K. M.; Bordignon, E.; Gray, H. B. J.
Am. Chem. Soc. 1982, 104, 5798; (b) Isied, S.S.; Worosila, G.; Atherton, S. J.
J. Am. Chem. Soc. 1982, 104, 7659.

33.

34.
35.

36.
37.

38
39

40.
41.
42.

43.

45.

46.

132

(a) Galoppini, E.; Fox, M. A. J. Am. Chem. Soc. 1996, 118, 2299; (b) Fox, M.
A.; Galoppini, E. J. Am. Chem. Soc. 1997, 119, 5277.

Crane, B. R.; Siegel, L. M.; Getzoff, E. D. Science 1995, 270, 59.

Turro, C.; Chang, C. K.; Leroi, G. E.; Cukier, R. 1.; Nocera, D. G. J. Am.
Chem. Soc. 1992, 114, 4013.

Rege, P. J. F. de; Williams, S. A.; Therien, M. J. Science 1995, 269, 1409.
Sessler, J. L.; Wang, B.; Harriman, A. J. Am. Chem. Soc. 1993, 115, 10418.
Pinner, A. Die Imidoc'ither und ihre Derivate. Oppenheim, Berlin 1892.

Krechl, J.; Smrckova, S.; Kuthan, J. Collect. Czech. Chem. Commun. 1990, 55,
460.

Kemp, D. 8.; Cox, D. D.; Paul, K. G. J. Am. Chem. Soc. 1975, 97, 7312.
Desiraju, G. R. Angew. Chem, Int. Ed. Engl. 1995, 34, 2311.

(a) Whitesides, G. M.; Mathias, J. P.; Seto, C. T. Science 1991, 254, 1312. (b)
MacDonald, J. C.; Whitesides, G. M. Chem. Rev. 1994, 94, 2383.

(a) Fagan, P. J.; Ward, M. O.; Calabrese, J. C. J. Am. Chem. Soc. 1989, 111,
1698; (b) Ward, M. O.; Fagan, P. J.; Calabrese, J. C.; Johnson, D. C. J. Am.
Chem. Soc. 1989, 111, 1719.

(a) Russell V. A.; Etter, M. C.; Ward, M. E. J. Am. Chem. Soc. 1994, 116,
1941. (b) Russell V. A.; Etter, M. C.; Ward, M. E. Chem. Mater. 1994, 6, 1206.
(c) Russell V. A.; Ward, M. E. Chem. Mater. 1996, 8, 1654.

Hosseini, M. W.; Ruppert, R.; Schaeffer, P.; Clan, A. De C.; Kyrisakas, N.;
Fischer, J. J. Chem. Soc., Chem. Commun. 1994, 2135.

(a) Aakerby, C. B.; Hughes, D. P.; Nieuwenhuyzen, M. J. Am. Chem. Soc.
1996, 118, 10134. (b) Aakeréy, C. B.; Hitchcock, P. B.; Moyle, B. D.; Seddon,
K. R. J. Chem. Soc., Chem. Commun. 1992, 553.

47.

48.

49.

50.

51.

52.
53.

133

Schwiebert, K. B; Chin, D. N.; MacDonald, J. C.; Whitesides, G. M. J. Am.
Chem. Soc. 1996, 118, 4018.

The benzoate of the previously known 3—amidinium chloride benzoic acid
(Wagner, G.; Vieweg, H.; Kuehmstedt, H. Phannazie, 1973, 28, 288) was
prepared in one-step synthesis as described in supporting information.

A Rigaku diffractometer with Mo Kat radiation (3. = 0.71073 A) was used to collect
data to a maximum of 20 = 50° from crystals of 0.65 x 0.40 x 0.20 mm dimensions
at 293 K. A total of 1397 independent reflections were collected, of which 1288
reflections were considered observed (1 > 20(1)) after Lp and absorption correction
(correction was based on w scans of a few suitable reflections with x values close
to 90°). Crystallographic data were as follows: CgHgNzoz, MW = 164.16,
monoclinic, space group P21/n (14), a = 7.253(2) A, b = 6.934(3) A, c =
14.602(2) A, 13 = 93.64(2)°, v = 733.0(4) A3, 2 = 4, Pcalc = 1.488 g cm'3. The
structure was solved by using direct methods (SHELXS) and refined by using full-
matrix least-squares procedures based on F2 (SHELXL-93) with R1 = 0.0398 and
wR2 = 0.1242.

(3) Pranata, J.; Wierschke, S. G.; Jorgensen, W. L. J. Am. Chem. Soc. 1991,
113, 2810. (b) Jorgensen, W. L.; Pranata, J. J. Am. Chem. Soc. 1990, 112,
2008.

(a) Kirby, J. P.; Dantzig, N. A. van; Chang, C. K.; Nocera, D. G. Tetrahedron
Lett. 1995, 36, 3477. (b) Roberts, J. A.; Kirby, J. P.; Nocera, D. G. J. Am.
Chem. Soc. 1995, 117, 8051.

Derissen, J. L. Acta Cryst. 1974, B30, 2764.

(a) Kratochvr’l, B.; Ondraé ek, J.; Maly , K.; Csordas, L. Collect. Czech. Chem.
Commun. 1988, 53, 294. (b) Kratochvr’l, B.; Ondraé ek, J.; Krechl, J.; Ha‘s ek, J.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

134

Acta Cryst. 1987, C43, 2182.

Olah, G. A.; Keumi, T. Synthesis 1979, 112.

Garigipati, R. A. Tetrahedron Lett. 1990, 31, 1969.

Schaefer, F. C.; Peters, G. A. J. Org. Chem. 1961, 26, 412.

Sullivan, B. P.; Salmon, D. J .; Meyer, T. J. Inorg. Chem. 1978, 12, 3334.

Miiller, G.; Riede, J.; Schmidtchen, P. Angew. Chem., Int. Ed. Engl. 1988, 27,
1516.

Wilcox, C. S. In Frontiers in Supramolecular Organic Chemistry and
Photochemistry; Schneider, H.-J., Durr, H., Eds.; VCH: Weinheim, 1991; p 123.
Albert, J. S.; Goodman, M. S.; Hamilton, A. D. J. Am. Chem. Soc. 1995, 117,
1143.

Hupp, J. T.; Neyhart, G. A.; Meyer, T. J.; Kober, E. M. J. Phys. Chem. 1992, 96,
10820

Balzani, V.; Scandola, F . Supramolecular Photochemistry; Ellis Horwood: New
York, 1991; Ch. 2.7.2.

Bock, C. R; Connor, J. A.; Gutierrez, A. R.; Meyer, T. J.; Whitten, D. G.;
Sullivan, B. P.; Nagle, J. K. J. Am. Chem. Soc. 1979, 101, 4815.

Assuming a diffiision-controlled on rate of 109 M‘1 s'l for these systems, from
the measured equilibrium constants for 1 and 2, off rates of 2 x 103 M and 102 s-
', respectively, are expected for these systems. Of course, the pertinent
equilibrium is with electronically excited Ru(ll) complex and not the ground state
complex, as measured by absorption spectroscopy. Notwithstanding, the largest
ApKa changes for Ru(II) polypyridyl complexes between ground and excited
states are 3; for instance, bpy’s directly appended with carboxylic acids exhibit

ApKas of 2.5. Hence the observed intrarnolecular electron transfer decay rates well

 

65.
66.
67.
68.

69.
70.

71.
72.
73.

74.
75.
76.

77.
78.

135

exceed equilibration rates, and intrarnolecular electron transfer occurs for a static
complex. This is not the case for 3, however, where the off rates are expected to
be commensurate with the electron transfer kinetics. For this system, the
interpretation of the observed PCET rate constant reported in Table 1 is
complicated by a dynamic equilibrium between acceptor and donor.

Galoppini, E.; Fox, M. A. J. Am. Chem. Soc. 1996, 118, 2299.

Stanley, R. J.; King, B.; Boxer, S. G. J. Phys. Chem. 1996, 100, 12052.

Heller, B. A.; Holten, D.; Kirmaier, C. Science 1995, 269, 940.

Scherer, T.; van Stokkum, I. M. H.; Brouwer, A. M.; Verhoeven, J. W. J. Phys.
Chem. 1994, 98, 10539.

Gray, H. B.; Winkler, J. R. Annu. Rev. Biochem. 1996, 65, 537.

(a) J. Pranata, S. G. Wierschke and W. L. Jorgensen, J. Am. Chem. Soc.,
113 (1991) 2810. b) W. L. Jorgensen and J. Pranata, J. Am. Chem. Soc., 112
(1990) 2008.

E. Galoppini and M. A. Fox, J. Am. Chem. Soc., 118 (1996) 2299.

B. A. Heller, D. Holten and C. Kirmaier, Science, 269 (1995) 940.

Peek, B. M.; Ross, G. T.; Edwards, S. W.; Meyer, G. J.; Meyer, T. J.; Erickson,
B. W. Int. J. Peptide Protein Res. 1991, 38, 114.

Opperman, K. A.; Mecklenburg, S. L.; Meyer. T. J. Inorg. Chem. 1994, 33, 5295.
Creary, X. J. Org. Chem. 1993, 58, 7700.

B. M. Peek, G. T. Ross, 8. W. Edwards, G. J. Meyer, T. J. Meyer and B. W.
Erickson, Int. J. Peptide Protein Res., 38 (1991) 114.

Garigipati, R. A. Tetrahedron Lett. 1990, 31 , 1969.

Kalyanasundaram, K. Photochemistry of Polypyridine and Porphyrin Complexes;
Academic Press: London, 1992.

79.
80.
81.

82.

83.

84.

85.

86.
87.

88.
89.
90.
91.
92.
93.
94.

136

Levin, J. 1.; Turos, E.; Weinreb, S. M. Synth. Comm. 1982, 12, 989.

Morgan, A. R.; Tertel, N. C. J. Org. Chem. 1986, 51, 1347.

Grigg, R.; Shelton, G.; Sweeney, A.; Johnson, A. W. J. Chem. Soc., Perkin
Trans. 1 1972, 1789.

Falk, J. E. Porphyrins and Metalloporphyrins; Elsevier: New York, 1964, pg.
129.

Gébel, M. W.; Bats, J. W.; Diimer, G. Angew. Chem. Int. Ed. Engl. 1992, 31,
207.

Wilcox, C. S. In Frontiers in Supramolecular Organic Chemistry and
Photochemistry; Schneider, H.-J. Durr, H. Eds.; VCH: Weinheim, 1991; pp.
123-143.

Wilcox, C. S.; Kim, E.-i.; Romano, D.; Kuo, L. H.; Burt, A. L.; Curran, D. P.
Tetrahedron 1995, 51 , 621.

Kelly, T. R.; Kim, M. H. J. Am. Chem. Soc. 1994, 116, 7072.

(a) Jorgensen, W. L.; Pranata, J. J. Am. Chem. Soc. 1990, 112, 2008. (b)
Pranata, J.; Wierschke, S. G.; Jorgensen, W. L. J. Am. Chem. Soc. 1991, 113,
2810.

Zaleski, J. M.; Chang, C. K.; Nocera, D. G. J. Phys. Chem. 1993, 97, 13206.
Meier, M. S.; Fox, M. A.; Miller, J. R. J. Org. Chem. 1991, 56, 5380.

Refer to the discusion on this topic presented in Chapter 1 of this theis.

Slates, R. V.; Szwarc, M. J. Phys. Chem. 1965, 69, 4124.

Shida, T. Electronic Absorpiton Spectra of Radical Ions, Elsevier, 1988.

Sigman, M. E; Closs, G. L. J. Phys. Chem. 1991, 95, 5012.

Petrzilka, M.; Gerrnann, A. Mol. Cryst. Liq. Cryst. 1985, 131, 327.