photophysical studies of chromium sensitizers …
TRANSCRIPT
PHOTOPHYSICAL STUDIES OF CHROMIUM SENSITIZERS DESIGNED FOR EXCITED
STATE HOLE TRANSFER TO SEMICONDUCTORS AND SEQUENTIAL
HOLE/ELECTRON TRANSFERS FROM PHOTOEXCITED CADMIUM SULFIDE
NANORODS TO MONONUCLEAR RUTHENIUM WATER-OXIDATION CATALYSTS
by
HUAN-WEI TSENG
B.S., National Chung-Cheng University, Taiwan, 2003
M.S., University of Houston, Houston, Texas, 2008
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirement for the degree of
Doctor of Philosophy
Department of Chemistry and Biochemistry
2013
This thesis entitled:
Photophysical Studies of Chromium Sensitizers Designed for Excited State Hole Transfer to
Semiconductors and Sequential Hole/Electron Transfers from Photoexcited Cadmium Sulfide
Nanorods to Mononuclear Ruthenium Water-Oxidation Catalysts
written by Huan-Wei Tseng
has been approved for the Department of Chemistry and Biochemistry
Associate Professor Niels H. Damrauer
Professor Cortlandt G. Pierpont
Date________________
The final copy of this thesis has been examined by the signatories, and we
find that both the content and the form meet acceptable presentation standards
of scholarly work in the above mentioned discipline.
iii
Tseng, Huan-Wei (Ph.D., Chemistry)
Photophysical Studies of Chromium Sensitizers Designed for Excited State Hole Transfer to
Semiconductors and Sequential Hole/Electron Transfers from Photoexcited Cadmium Sulfide
Nanorods to Mononuclear Ruthenium Water-Oxidation Catalysts
Dissertation directed by Associate Professor Niels H. Damrauer
This dissertation describes three research projects related to solar cells and solar water
splitting with a goal of utilizing solar energy, a renewable energy source. The first project is
focused on photophysical studies of four newly-synthesized Cr(III) tris-bipyridyl complexes
featuring the 4-dmcbpy (dimethyl 2,2’-bipyridine-4,4’-dicarboxylate) ligand. Static and time-
resolved emission results suggest that the complexes store ∼1.7 eV of energy for multiple
microseconds. Using cyclic voltammetry, it is found that the inclusion of 4-dmcbpy shifts the
E1/2 of CrIII/II
by +0.2 V from the homoleptic parent complexes without 4-dmcbpy. All four
complexes have excited state potentials of CrIII*/II
between +1.8 and +2.0 V vs. NHE, placing
them among the most powerful photooxidants reported and making them candidates for hole-
injection sensitizers.
The second project continues with Cr(III) complexes, but using iminopyridine Schiff base
ligands. Two complexes feature hexadentate ligands and the other two are their tris-bidentate
analogues. One of each pair contains methyl ester groups for attachment to semiconductors.
Cyclic voltammograms show that the hexadentate and tris-bidentate analogues have almost
identical reduction potentials, but the addition of ester substituents shifts the reduction potentials
by +0.2 V. The absorption spectra of the hexadentate complexes show improved absorption of
iv
visible light compared to the tris-bidentate analogues. For freshly prepared sample solutions in
CH3CN, time-resolved emission and transient absorption measurements for the Cr(III) tris-
bidentate ester complex show a doublet excited state with a 17-19 μs lifetime at room
temperature, while no emission or transient absorption signals from the doublet states are
observed for the hexadentate analogue under the same conditions. The dramatic difference is due
to the presence of a nonligated bridgehead nitrogen atom.
The third project features charge transfer interactions between a photoexcited cadmium
sulfide nanorod and [Ru(diethyl 2,2’-bipyridine-4,4’-dicarboxylate)(2,2’:6’,2”-terpyridine)Cl]+, a
mononuclear water-oxidation catalyst. Upon photoexcitation, hole transfer from the cadmium
sulfide nanorod oxidizes the catalyst (Ru2+→Ru
3+) on a 100 ps to 1 ns timescale. This is followed
by electron transfer (10-100 ns) from the nanorod to reduce the Ru3+
center. The relatively slow
electron transfer dynamics may provide opportunities for the accumulation of multiple holes at
the catalyst, which is required for water oxidation.
v
ACKNOWLEDGMENTS
First of all, I would like to thank my advisor, Professor Niels Damrauer, for giving me
the chance to learn nanosecond and femtosecond laser spectroscopy and to work on several
interesting research topics. I really appreciate his support and guidance.
It would not have been possible to get all the projects done without the contributions
from our collaborators. I would like to thank Professor Matt Shores and Ashley McDaniel for the
synthesis of the chromium complexes, Professor Anthony Rappé for the computations on the
chromium complexes, and Professor Gordana Dukovic and Molly Wilker for the synthesis and
spectroscopic studies on the cadmium sulfide nanorods. I would also like to thank my labmates:
Heather Meylemans, Erik Grumstrup, Mirvat Abdelhaq, Joshua Hewitt, Paul Vallett, Karen
Spettel, Jamie Snyder, Jasper Cook, Sean Ryan, Jessica Ramirez, Dr. Bijan Paul, Sam Shepard,
Thomas Carey, and Anna Curtis for their help in the lab.
Many thanks to my friends: Sophie Mok, Sandy Kao, Roxanne Liu, Shelly Wang, Xinyi
Gao, Kaiyun Chen, Kent Lin, Paula Ko, Cindy Hsu, Hsiang-Yu Chen, Wan-Ching Hsu, Jack Lin,
Joanna Chen, Huai-Chun Chen, Kai-Chun Chen, Chih-Hsuan Chang, Huei-Jiun Li, Chun-Wei
Shih, Julia Cheng, Anling Cheng, Jenar Wang, Paul Carr, Jane Tseng, Lindsay Huang, Katherine
Chou, Michelle Tai, Tai-Dan Hsu, Pin-Chin Maness, Fanny Liu, Nan Hu, Lynn Junglim Lee,
Younghee Lee, Patrick Konold, and Jean Chen for making my life beautiful in Colorado.
At the end, I would like to dedicate my dissertation to my family in Taiwan and Toronto
for their love and support.
vi
CONTENTS
CHAPTER 1. General Introduction Page
1.1 Energy Sources, Global Warming, and Renewable Energy 1
1.2 Dye-Sensitized Solar Cells 5
1.3 Solar Water-Splitting Systems 13
1.4 References 17
CHAPTER 2. Solution Phase Characterization of Strongly Photooxidizing
Heteroleptic Cr(III) Tris-Dipyridyl Complexes: Toward Efficient
Hole Transfer Sensitizers
2.1 Introduction 20
2.2 Experimental Section 23
2.3 Results and Discussion 34
2.4 Conclusions 58
2.5 Acknowledgment on Participation 59
2.6 References 59
CHAPTER 3. Photophysical Investigations of Cr(III) Hexadentate Iminopyridine
Complexes and Their Tris-Bidentate Analogues
3.1 Introduction 64
3.2 Experimental Section 67
3.3 Results and Discussion 77
3.4 Conclusions 103
3.5 Acknowledgment on Participation 103
3.6 References 104
vii
CHAPTER 4. Sequential Hole/Electron Transfer from Photoexcited Cadmium
Sulfide Nanorods to Mononuclear Ruthenium Water-Oxidation
Catalysts
4.1 Introduction 107
4.2 Experimental Section 109
4.3 Results and Discussion 115
4.4 Conclusions 137
4.5 Acknowledgment on Participation 138
4.6 References 138
viii
FIGURES
Page
Figure 1.1 U.S. energy consumption in 2010. 2
Figure 1.2 Atmospheric concentrations of carbon dioxide over the last 10000
years, and since 1750 (inset).
3
Figure 1.3 Observed changes in (a) global average surface temperature, (b)
global average sea level from tide gauge (blue) and satellite (red)
data and (c) North Hemisphere snow cover for March-April.
3
Figure 1.4 A schematic representation of the dye-sensitized solar cell
prototype.
5
Figure 1.5 Chemical structure of the N3 dye and a photo of a nanocrystalling
TiO2 film on a piece of conducting glass sensitized by the N3 dye.
7
Figure 1.6 Chemical structure of the black dye and a photo of a nanocrystalling
TiO2 film on a piece of conducting glass sensitized by the black
dye.
7
Figure 1.7 Chemical structure of the CYC-B11 dye. 8
Figure 1.8 Chemical structures of the YD2-o-C8 dye and a co-sensitizer Y123
dye.
9
Figure 1.9 A schematic representation of the energy levels of the dye-
sensitized NiO solar cell.
10
Figure 1.10 Chemical structures of erythrosin B and TPPC. 11
Figure 1.11 A graphical representation of the photoelectrochemical water-
splitting cell by Fujishima and Honda.
13
Figure 1.12 A schematic representation of the components in a solar water-
splitting system.
15
Figure 1.13 Band gaps and band-edge positions for selected semiconductors and
the redox electrochemical potentials for water splitting.
16
Figure 2.1 X-ray crystal structure of the [Cr(phen)2(4-dmcbpy)]3+
complex
cation.
36
ix
Figure 2.2 Full crystal structure for the complex
[Cr(phen)2(4-dmcbpy)](OTf)3∙1.3CH3CN
36
Figure 2.3 Electronic absorption spectra for Cr(III) dipyridyl complexes in 1 M
HCl(aq) at room temperature.
39
Figure 2.4 Electronic absorption spectra for Cr(III) complexes containing
phenanthroline-based ligands.
40
Figure 2.5 Emission spectra for Cr(III) polypyridyl complexes in deoxygenated
1 M HCl(aq) following excitation at 320 nm.
43
Figure 2.6 Temperature dependence of the observed rate constant for
[Cr(bpy)3]3+
and [Cr(4-dmcbpy)3]3+
in degassed 1 M HCl(aq)
48
Figure 2.7 Temperature dependence of the observed rate constant (kobs = 1/τobs)
for the Cr(III) polypyridyl complexes in degassed 1 M HCl(aq).
50
Figure 2.8 Comparison of cyclic voltammograms for the 4-dmcbpy containing
complexes in 0.1 M TBAPF6 acetonitrile solution.
51
Figure 2.9 Transient absorption spectra on a μs timescale for Cr(III)
polypyridyl complexes in 1 M HCl(aq).
57
Figure 3.1 Structures of the complex cations [Cr(pod-ester)]3+
and
[Cr(impy-ester)3]3+
.
79
Figure 3.2 The full complex cation, and crystallographically independent anion
and solvent molecules in the structure of
[Cr(pod-ester)](BF4)3·CH3CN
80
Figure 3.3 Alternate view of Cr(III) complex cation in the structure of
[Cr(pod-ester)](BF4)3·CH3CN, highlighting the octahedral
coordination of the Cr center.
80
Figure 3.4 The full structure of [Cr(impy-ester)3](BF4)3 including disordered
components
83
Figure 3.5 Comparison of cyclic voltammograms for all four complexes in
0.1 M TBAPF6 acetonitrile solution.
84
Figure 3.6 Electronic absorption spectra collected in acetonitrile for Cr(III)
complexes in the UV range (main) and the visible range (inset).
87
x
Figure 3.7 Electronic absorption spectra of the free ligands pod and pod-ester
in acetonitrile and impy and impy-ester in pentane (~10 μM to 1
mM).
87
Figure 3.8 Emission spectra for freshly prepared solutions of
[Cr(pod-ester)](BF4)3 and [Cr(impy-ester)3](BF4)3 in deoxygenated
acetonitrile at room temperature following 355 nm excitation.
91
Figure 3.9 2E excitation spectrum for a fresh prepared solution of
[Cr(impy-ester)3](BF4)3 in deoxygenated acetonitrile at room
temperature.
92
Figure 3.10 Emission spectra for [Cr(pod-ester)](BF4)3 and
[Cr(impy-ester)3](BF4)3 in acetonitrile right after dissolution and at
24 hours.
95
Figure 3.11 Computed excitation energies for selected Cr(III) complexes. 98
Figure 3.12 Natural transition orbitals (NTOs) for the lowest (a) doublet and (b)
quartet transitions for [Cr(pod-ester)]3+
(left) and [Cr(impy-ester)3]3+
99
Figure 3.13 Calculated doublet and quartet excited state absorption spectra of
[Cr(bpy)3]3+
and transient absorption spectrum of [Cr(bpy)3]3+
.
100
Figure 4.1 Chemical structure of the water-oxidation complex
[Ru(deeb)(tpy)Cl](PF6).
108
Figure 4.2 TEM image of CdS NRs. 116
Figure 4.3 UV-vis absorption spectra of the CdS NRs, [Ru(deeb)(tpy)Cl]+, and
their mixture.
117
Figure 4.4 Energy level diagram depicting the band edges of CdS NRs and the
redox potentials of [Ru(deeb)(tpy)Cl]+.
119
Figure 4.5 PL spectra of CdS NRs with increasing amounts of
[Ru(deeb)(tpy)Cl]+ and constant CdS NR concentration.
120
Figure 4.6 PL spectra for CdS NR and CdS NR with free ligands tpy, deeb, or
with [Ru(deeb)(tpy)Cl]+.
120
xi
Figure 4.7 Quenching of CdS PL as a function of concentration of
[Ru(deeb)(tpy)Cl]+. In (a) and (b), band gap and trap emission are
fit with the Stern-Volmer model for dynamic collisional quenching,
whereas in (c) and (d) the same data is fit with a linear form of the
Langmuir adsorption isotherm.
122
Figure 4.8 Fraction of PL quenched for band gap (475 nm) and trap (700 nm)
emission as a function of [Ru(deeb)(tpy)Cl]+ concentration and the
[Ru(deeb)(tpy)Cl]+ : CdS NR ratio.
124
Figure 4.9 TA spectra of (a) CdS NRs, (b) CdS with [Ru(deeb)(tpy)Cl]+, and
(c) CdS NRs, CdS, and [Ru(deeb)(tpy)Cl]+ in MeOH taken 2 ps
after excitation with a 400 nm pump.
126
Figure 4.10 TA decay kinetics at 470 nm for CdS NRs in the presence and
absence of [Ru(deeb)(tpy)Cl]+.
127
Figure 4.11 The same TA decay kinetics data as in Figure 4.10 plotted using a
linear time axis.
127
Figure 4.12 Proposed charge transfer steps between photoexcited CdS NR and
[Ru(deeb)(tpy)Cl]+.
129
Figure 4.13 TA decay kinetics at 470 nm for CdS NRs alone, and in the
presence of the following: [Ru(deeb)(tpy)Cl]+, the hole scavenger
ascorbate, the electron acceptor methylene blue, and both
[Ru(deeb)(tpy)Cl]+ and Asc.
130
Figure 4.14 Steady-state absorption spectra of a sample containing CdS NRs,
[Ru(deeb)(tpy)Cl]+, and ascorbate taken before and after TA data
collection.
131
Figure 4.15 Transient absorption kinetics of CdS NRs with fixed concentration
of CdS NRs and varying CdS NR : [Ru(deeb)(tpy)Cl]+ ratios.
132
Figure 4.16 (a) The onset of ET as a function of [Ru(deeb)(tpy)Cl]+ : CdS NR
ratio. (b) QEET and the excited electron lifetimes plotted versus both
concentration of [Ru(deeb)(tpy)Cl]+ and [Ru(deeb)(tpy)Cl]
+ : CdS
NR ratio.
133
xii
SCHEMES
Page
Scheme 2.1 Target Structures of Cr(III) Complexes 22
Scheme 2.2 Preparative Routes for the [Cr(NN)2(4-dmcbpy)](OTf)3 Complexes 35
Scheme 2.3 Preparative Routes for [Cr(4-dmcbpy)3](BF4)3 35
Scheme 3.1 Podand Tripodal and Bidentate Iminopyridine Ligands Used in This
Study
65
Scheme 3.2 Preparation of the Cr(III) Complexes 77
xiii
TABLES
Page
Table 2.1 Crystallographic Data for [Cr(phen)2(4-dmcbpy)](OTf)3 ∙1.3CH3CN 27
Table 2.2 Selected Bond Lengths and Angles for
[Cr(phen)2(4-dmcbpy)](OTf)3∙1.3CH3CN
37
Table 2.3 Room Temperature Electronic Absorptions for Cr(III) Tris-Dipyridyl
Complexes
42
Table 2.4 Photophysical Data for Cr(III) Polypyridyl Complexes in 1 M HCl(aq) 45
Table 2.5 Ground and Excited State Reduction Potentials for Cr(III) Polypyridyl
Complexes
52
Table 3.1 Crystallographic Data for [Cr(pod-ester)](BF4)3 ∙CH3CN and
[Cr(impy-ester)3](BF4)3
78
Table 3.2 Average Cr−N Distances (Å) for [Cr(pod)](ClO4)3,
[Cr(pod-ester)](BF4)3·CH3CN, and [Cr(impy-ester)3](BF4)3
81
Table 3.3 Ground State Reduction Potentials for Cr(III) Complexes 85
Table 3.4 Computed Excitation Energies (EEs) for Selected Complexes (eV) 98
Table 4.1 Results of analysis of TA dynamics displayed in Figure 4.15. 132
1
CHAPTER 1
General Introduction
1.1 Energy Sources, Global Warming, and Renewable Energy
The activities of human civilization rely on the utilization of energy. In old times people
made fire to generate heat and light. After the Industrial Revolution, wood- or coal-burning
steam engines were used to convert heat into kinetic energy needed to drive trains. Power plants
today use coal or oil to generate electricity. Bulbs use electricity to light up the night. Cars and
airplanes consume gasoline. Computers and cell phones use electricity. Cooking and heating at
home takes natural gas. The usage of energy in different forms and their inter-conversion adds
convenience to our life. However, the rapid growth of energy demand has become one of the
biggest concerns about the continuation of our comfortable modern lifestyle.
For the United States, the annual energy consumption has tripled from 1950 to 2010.1 In
2010, more than 90% of the energy consumed was from non-renewable sources (including
petroleum, natural gas, coal, and nuclear) and less than 10% was from renewable sources
(including hydropower, biomass, geothermal, solar, and wind).2 The percentages of each energy
source are shown in detail in Figure 1.1. Petroleum, natural gas, and coal are depletable energy
sources since their reserves are limited and thus new energy sources are needed. In addition to
delivering the energy we need, burning fossil fuels also generates byproducts including the
greenhouse gas carbon dioxide.
2
Figure 1.1 U.S. energy consumption in 2010. Note that hydropower is considered as a
conventional energy source and is accounted separately from other renewable sources. Figure
taken from Reference 2. Copyright 2011 U.S. DOE.
Since the Industrial Revolution, the global increase of carbon dioxide concentration in the
atmosphere is mainly due to fossil fuel use and land use change.3 The increase of carbon dioxide
concentration has a great impact on climate change and global warming. According to the
Intergovernmental Panel on Climate Change,3 “Warming of the climate system is unequivocal,
as is now evident from observations of increases in global average air and ocean temperatures,
widespread melting of snow and ice, and rising global average sea level.” In Figure 1.2, we can
see the fast increase of carbon dioxide concentration in the 20th
century. The consequences
include global increases in temperature, sea level rise, and a decrease in land snow coverage
which can be seen in Figure 1.3. With their non-renewable nature and huge effect on our
environment, fossil fuels need to be replaced to ensure the future sustainability of human
development.
Other than fossil fuels, nuclear is one of the major energy sources and accounted for
8.6% of U.S. energy consumption in 2010 (Figure 1.1). However, the Three Mile Island accident
in 1979 and the Chernobyl disaster in 1986 have clearly revealed the seriousness of nuclear
events.4-5
More recently, the Fukushima Daiichi nuclear disaster in 2011 has changed the
direction of energy policy towards renewable energy in Japan and many other countries.6-7
3
Figure 1.2 Atmospheric concentrations of carbon dioxide over the last 10000 years, and since
1750 (inset). Measurements are shown from ice cores (symbols with different colors for different
studies) and atmospheric samples (red lines). Figure taken from Reference 3. Copyright 2007
IPCC.
Figure 1.3 Observed changes in (a) global average surface temperature, (b) global average sea
level from tide gauge (blue) and satellite (red) data and (c) North Hemisphere snow cover for
March-April. All changes are relative to corresponding averages for the period 1961-1990.
Figure taken from Reference 3. Copyright 2007 IPCC.
4
Renewable energy resources, which are associated with low amounts of carbon dioxide
emission, have received attention with the increase of awareness that carbon dioxide emissions
lead to global warming.8 Although renewable energy is currently only a small portion of the total
energy consumption in the U.S., some findings by the Department of Energy show that between
2000 and 2010 the installed renewable energy capacity more than quadrupled in the U.S. and
globally.2 This is encouraging because people have started to pay attention to the balance
between economic development and the protection of our environment, and governments have
begun taking measures to ensure us a sustainable Earth.
The research activities described in this dissertation are related to renewable energy,
especially the utilization of solar energy. Fundamental research was conducted using new
molecules and materials with a goal of converting sunlight into electricity or fuels. The devices
which generate electricity from sunlight using sensitizers and semiconductors are called ‘Dye-
Sensitized Solar Cells’ and the concepts will be described in Section 1.2. For fuels, there are
systems that can absorb sunlight to split water and then form molecular oxygen and hydrogen.
By doing that, solar energy is converted and stored in the form of chemical bonds for future use.
Such solar water-splitting systems will be discussed in Section 1.3.
5
1.2 Dye-Sensitized Solar Cells
The prototype of the dye-sensitized solar cell (DSSC) was introduced by Michael Grӓtzel
and Brian O’Regan in 1991.9 A schematic representation of the DSSC components can be seen
in Figure 1.4.10
Figure 1.4 A schematic representation of the dye-sensitized solar cell prototype. In this sample
module, the semiconductor is TiO2, the dye is cis-Ru(NCS)2(2,2’-bipyridine-4,4’-dicarboxylate)2,
and the electrolyte is a 3I−/I3
− couple. Figure taken from Reference 10. Copyright 2005 American
Chemical Society.
In this system, the porous semiconductor film is composed of TiO2 nanocrystals (gray
dots in the figure). TiO2 does not absorb visible light, therefore dye molecules (red dots in the
figure) with a high visible-absorbance are needed to harvest sunlight and the metal complex cis-
Ru(NCS)2(2,2’-bipyridine-4,4’-dicarboxylate)2 is used for this purpose. The carboxylate groups
on the metal complex are responsible for the attachment to the semiconductor surface and the
porosity of the semiconductor film provides a high surface area to maximize the amount of dye
6
molecules attached. After the absorption of a photon, the sensitizer (S) is promoted from the
electronic ground state to the excited state. The sensitizer in the excite state (denoted by S*) is
capable of injecting an electron to the conduction band of TiO2 forming an oxidized species (S+).
The electron in the conduction band would then diffuse towards the electrode, travel through an
external circuit to provide electrical power, and arrive at the counter electrode on the other side
of the solar cell. The redox mediator (I3−/3I
− couple) shuffles the electron from the counter
electrode to the oxidized sensitizer (S+) to regenerate the sensitizer (S) and therefore completes
the circuit.
In the past two decades, hundreds of dyes have been used as sensitizers.11
The first dye
with a high overall light-to-electricity energy conversion efficiency of 7-10% was cis-
Ru(NCS)2(2,2’-bipyridine-4,4’-dicarboxylic acid)2, also known as the “N3 dye”.9,12
The
chemical structure of the N3 dye and a photo of a TiO2 film sensitized with N3 dye are shown in
Figure 1.5. Since 1991, the efficiency of the DSSC with this red N3 dye had been unmatched by
any other dyes until the discovery of the “black dye” (Figure 1.6) in 2001 which allowed cells to
be built with an efficiency of 10.4%.13
7
Figure 1.5 Chemical structure of the N3 dye, cis-Ru(NCS)2(2,2’-bipyridine-4,4’-dicarboxylic
acid)2, and a photo of a nanocrystalling TiO2 film on a piece of conducting glass sensitized by
the N3 dye. The thickness of the TiO2 film is about 5 μm. Picture taken from Reference 10.
Copyright 2005 American Chemical Society.
Figure 1.6 Chemical structure of the black dye, Ru(NCS)3(2,2’:6’,2”-terpyridine-4,4’,4”-
tricarboxylic acid), and a photo of a nanocrystalling TiO2 film on a piece of conducting glass
sensitized by the black dye. The thickness of the TiO2 film is about 5 μm. Picture taken from
Reference 10. Copyright 2005 American Chemical Society.
In 2009, a ruthenium complex coded “CYC-B11” (Figure 1.7) possessing a similar ligand
set to the N3 dye (two bidentate ligands and two SCN− monodentate ligands) was synthesized
with two hexylthio-bithiophene substituents on one of the bipyridine ligands.14
The introduction
of the hexylthio-bithiophene substituents on the ancillary ligand increased the molar absorptivity
8
of the complex and the light-harvesting ability and efficiency of the DSSC using the CYC-B11
dye were both improved. The DSSC device showed an overall light-to-electricity energy
conversion efficiency of 11.5%.
Figure 1.7 Chemical structure of the CYC-B11 dye. Figure taken from Reference 14. Copyright
2009 American Chemical Society.
The efficiencies of the solar-cell devices have a dependence on the nature of the redox
shuttle.11
Instead of ruthenium polypyridyl complexes as mentioned above, a zinc porphyrin
complex (YD2-o-C8 dye in Figure 1.8) was revealed in 2011 with overall light-to-electricity
energy conversion efficiencies of 7.6% using an I3−/3I
− redox shuttle and 11.9% using a
CoIII/II
tris(bipyridyl) redox shuttle.15
During the sensitizer-regeneration process the energy loss in
terms of voltage is high when using I3−/3I
−. One alternative is the Co
III/IItris(bipyridyl) redox
shuttle with a relatively higher reduction potential. However, the kinetics associated with one-
electron, outer-sphere reduction of the CoIII
species makes it facile for the oxidized partner of the
redox shuttle to intercept the injected electron in TiO2 which ends up lowering the efficiencies of
DSSCs. The design of the YD2-o-C8 dye includes long-chain alkyloxy groups on the dye
molecule which are hypothesized to hinder the unwanted charge recombination process, thus
serving to increase the overall efficiency. The efficiency of the DSSC (using CoIII/II
tris(bipyridyl)
9
redox shuttle) could be further enhanced to 12.3% with the help of a co-sensitizer, Y123 (Figure
1.8). Ligand designs with the long-chain alkyloxy groups can also be applied to other dye
molecules, such as the ruthenium complexes mentioned above. The dyes with long-chain
alkyloxy groups can be used in conjunction with the CoIII/II
tris(bipyridyl) redox shuttle to
increase the overall efficiencies for solar-cell devices.
Figure 1.8 Chemical structures of the YD2-o-C8 dye and a co-sensitizer Y123 dye.
All the examples discussed above were based on the process of sensitizers absorbing
photons, followed by electron injection to the TiO2 semiconductor. Instead of electron injection,
we wondered if it would be possible to use sensitizers to inject holes into semiconductors and to
make solar cells with photocurrents running in the opposite direction to the electron-injection
DSSCs. We also wondered if a DSSC could couple both the electron- and hole-injection
functions to construct a tandem device. This tandem device would combine the voltages and
currents generated by each electrode, which is expected to increase the overall efficiency of the
DSSC. After looking into the literature, we found that both hole-injection solar cells16-26
and
10
tandem devices27-30
had been made and tested. However, the scope of fundamental studies is
vastly smaller compared to electron-injection DSSCs.
The first hole-injection DSSC was revealed in 1999 using either erythrosin B or
tetrakis(4-carboxyphenyl)porphyrin (TPPC) to sensitize NiO, a p-type semiconductor.16
A
schematic representation of the solar cell is in Figure 1.9 and the chemical structures of
erythrosin B and TPPC are shown in Figure 1.10.
Figure 1.9 A schematic representation of the energy levels of the dye-sensitized NiO solar cell.
The dye (D) in this picture is TPPC. Figure taken from Reference 16. Copyright 1999 American
Chemical Society.
In Figure 1.9, the nanostructured NiO film used is about 1 μm thick and is highly porous
with particle sizes of 10-20 nm based on the scanning electron microscopy images of the NiO
11
film. The redox couple used in the electrolyte is 3I−/I3
−. Upon photoexcitation, the electron
transfers from the excited dye to I3− in the electrolyte (process “a” indicated in Figure 1.9) and a
hole transfers from the HOMO of the oxidized dye to the valence band (VB) of NiO (process “b”
indicated in Figure 1.9). The hole produced in the VB of NiO diffuses to the conducting
electrode, then travels through the external load to the counter electrode, regenerates I3− from 3I
−
and thus completes the circuit.
Figure 1.10 Chemical structures of erythrosin B and TPPC.
The overall light-to-electricity energy conversion efficiencies for these systems were
found to be 0.0076% using erythrosin B as the sensitizer and 0.0033% using TPPC. There are
two reasons for the low overall efficiencies:26
(1) The redox potential of the I3−/3I
− couple is too
positive and is very close to the VB edge of the NiO semiconductor. This gives a huge energy
loss by way of process “a” indicated in Figure 1.9, thus making the maximum possible output
voltage (the potential difference between the I3−/3I
− couple and the upper edge of the NiO CB) to
be small (~ 0.1 V). (2) Due to the low mobility, the slow diffusion of holes in NiO makes it
12
easier for them to be intercepted by the redox couple in the electrolyte. This serves to decrease
output current.
In 2010, a hole-injection DSSC device was reported with 0.15% overall light-to-
electricity energy conversion efficiency.25
The device was based on a better quality NiO film
with more densely packed nanoparticles and a newly-designed organic sensitizer with a donor-
acceptor character. Although the efficiency was still low, this improvement by two orders of
magnitude was encouraging.
The first tandem DSSC was built in 2000 by combining the two working electrodes in
both electron-injection and hole-injection DSSCs.27
The photoanode used was an N3-sensitized
TiO2 film as discussed previously. The photocathode used was the erythrosin B-coated NiO film.
The electrolyte was I3−/3I
−. This tandem DSSC gave a photovoltage very close to the two
individual electrodes combined which was expected, but an overall light-to-electricity energy
conversion efficiency of 0.39% which is between the efficiencies of its component hole-injection
or electron-injection DSSCs by themselves. The possible reason for the low overall efficiency is
that the two electrodes gave different photocurrents, but the overall photocurrent of the tandem
DSSC was limited by the lower-photocurrent electrode.
Previous research has primarily focused on electron-injection solar cells rather than hole-
injection cells providing us with a relatively unexplored field of research and an opportunity to
investigate and contribute to improvements of cell efficiencies. In searching for potential hole-
injection sensitizers, we noted work on Cr(III) polypyridyl complexes done by Serpone and
Hoffman in the 1970’s to 1980’s.31-37
They showed that Cr(III) polypyridyl complexes are strong
excited-state photooxidants. For our purposes, the strongly photooxidizing abilities may meet our
13
needs in identifying sensitizers to construct hole-injection solar cells. The synthesis and
photophysical studies on a series of heteroleptic Cr(III) tris-dipyridyl complexes will be
described in Chapter 2. The studies on two Cr(III) complexes with hexadentate ligands and their
tris-bidentate analogues will be presented in Chapter 3.
1.3 Solar Water-Splitting Systems
The first device for the overall electrochemical photolysis of water was reported in 1972
by Akira Fujishima and Kenichi Honda using an n-type semiconductor TiO2 photoelectrode and
a platinum counter electrode.38
A graphical representation is shown in Figure 1.11. When the
TiO2 electrode was irradiated by photons from a xenon lamp, electrons flowed from the TiO2
electrode through the external circuit to the platinum electrode. With the help from external bias,
a reduction reaction (H2 evolution) occurred at the platinum electrode while an oxidation
reaction (O2 evolution) occurred at the TiO2 electrode.
Figure 1.11 A graphical representation of the photoelectrochemical water-splitting cell by
Fujishima and Honda. Figure taken from Reference 40. Copyright 2009 The Royal Society of
Chemistry.
14
There are two major disadvantages for this TiO2-based water-splitting system:39-40
(i) UV
light is needed for the system because TiO2 is a wide band-gap (> 3.0 eV) semiconductor, which
does not absorb visible light. In order to better utilize solar energy, materials that absorb visible
light are needed. (ii) Since the TiO2 conduction band edge is very close to the potential for H+-
reduction, the reducing power of the electrons in the TiO2 conduction band is insufficient for the
H2-evoling half-reaction. Materials with correct band-edge alignment are needed, i.e. the bottom
of the conduction band has to be more negative than the H+/H2 redox potential (0 V vs. NHE at
pH 0) and the top of the valence band needs to be more positive than the O2/H2O redox potential
(1.23 V vs. NHE at pH 0). Despite significant efforts in the search for new semiconductor
materials, relatively few systems have been found to split water under visible light and these
function with low overall quantum efficiencies.40-41
The general working principle for a semiconductor water-splitting system involves the
following critical processes: (i) light absorption and electron-hole pair formation, (ii) charge
separation, and (iii) redox reactions on the semiconductor surface.39
Bulk semiconductors
generally suffer from poor quantum conversion efficiencies in photocatalysis because the charge
carriers recombine significantly faster than they can be separated and utilized in a surface
reaction. This is partly due to the small overall surface area of bulk-sized material.42
With recent
developments in nanomaterials, colloidal semiconductor nanocrystals provide new opportunities
in the study of photocatalysis and solar fuel generation. Nanomaterials have the following
advantages over bulk materials: (i) large surface areas to facilitate heterogeneous reactions on the
semiconductor surface, (ii) small particle sizes, which increase electron-hole pair separation
efficiencies by shortening the electron- and hole-diffusion lengths for participating the redox
reactions on the surface, (iii) tunable optical properties to facilitate light absorption, and (iv)
15
tunable surface properties to provide the possibility of interacting with redox catalysts for more
efficient solar fuel generation.39,43
The design criteria discussed above for a solar water-splitting
system can be summarized in Figure 1.12.
Figure 1.12 A schematic representation of the components in a solar water-splitting system. The
circle represents a semiconductor nanocrystal. After the absorption of a photon, an electron-hole
pair is formed with an electron in the conduction band (CB) and a hole in the valence band (VB).
The electron in the CB can migrate to the reduction catalyst (Catred) to reduce protons and the
hole in the VB can travel to the oxidation catalyst (Catox) to oxidize water. The reactions are:
4H+ + 4e
− → 2H2 and 2H2O + 4h
+ → O2 + 4H
+.
In searching for semiconductors suitable for solar water splitting, CdS catches attention
by having the correct band-edge alignment for water splitting and visible absorption with a 2.4
eV band gap which corresponds to a transition at 517 nm (Figure 1.13). However,
photocorrosion has always been an issue for metal sulfides to be used in catalysis.40
Proceeded
by photoexcitation, the positively charged hole in the VB tends to oxidize the S2−
in the CdS,
rather than the O2−
from water. The disfavored chemical reaction is: CdS + 2h+ → Cd
2+ + S.
Although photocorrosion is an issue on the oxidation side, CdS suspensions can serve as an H+-
16
reduction photocatalyst in the presence of hole scavengers.44-45
With the help of co-catalysts such
as Pt islands or hydrogenase on the surface, CdS nanocrystals have shown photocatalytic activity
for H+-reduction in the presence of hole scavengers.
46-48
Figure 1.13 Band gaps and band-edge positions for selected semiconductors and the redox
electrochemical potentials for water splitting. Picture taken from Reference 40. Copyright 2009
The Royal Society of Chemistry.
Based on the discussions above, significant progress has been made on the H+-reduction
reactions using CdS nanocrystals with co-catalysts. However, to our knowledge, CdS
nanocrystals have not been coupled with catalysts for photocatalytic water oxidation. We are
interested in constructing a heterostructure involving CdS nanocrystals and ruthenium water-
oxidation catalysts. We will study the interactions between them with a focus on excited state
charge transfer. By understanding the properties in this model system, we hope to make
17
contributions towards the design of an efficient solar water-splitting nano-heterostructure. The
detailed work on this project will be described in Chapter 4.
1.4 References
(1) Annual Energy Review 2011; Energy Information Administration, Office of Energy
Statistics, U.S. Department of Energy, Washington, DC 20585, 2012.
(2) 2010 Renewable Energy Data Book; U.S. Department of Energy, 2011.
(3) IPCC Summary for Policymakers; Cambridge University Press, Cambridge, United
Kingdom and New York, NY, USA., 2007.
(4) Walker, J. S. Three Mile Island : a nuclear crisis in historical perspective; University of
California Press: Berkeley, 2004.
(5) Ablokov, A. V.; Nesterenko, V. i. B.; Nesterenko, A. V.; Sherman, J. D. Chernobyl :
Consequences of the Catastrophe for People and the Environment; Published by
Blackwell Pub. on behalf of the New York Academy of Sciences: Boston, Mass., 2009.
(6) Watanabe, C. Japan Spurs Solar, Wind Energy With Subsidies for Renewables 26 August
2011, Bloomberg.
(7) Wittneben, B. B. F. Environ. Sci. Policy 2012, 15, 1-3.
(8) Boyle, G. Renewable Energy - Power for a Sustainable Future; 2 ed.; Oxford University
Press Inc.: New York, 2004.
(9) Oregan, B.; Gratzel, M. Nature 1991, 353, 737-740.
(10) Gratzel, M. Inorg. Chem. 2005, 44, 6841-6851.
(11) Hagfeldt, A.; Boschloo, G.; Sun, L. C.; Kloo, L.; Pettersson, H. Chem. Rev. 2010, 110,
6595-6663.
(12) Nazeeruddin, M. K.; Kay, A.; Rodicio, I.; Humphrybaker, R.; Muller, E.; Liska, P.;
Vlachopoulos, N.; Gratzel, M. J. Am. Chem. Soc. 1993, 115, 6382-6390.
(13) Nazeeruddin, M. K.; Pechy, P.; Renouard, T.; Zakeeruddin, S. M.; Humphry-Baker, R.;
Comte, P.; Liska, P.; Cevey, L.; Costa, E.; Shklover, V.; et al. J. Am. Chem. Soc. 2001,
123, 1613-1624.
(14) Chen, C. Y.; Wang, M. K.; Li, J. Y.; Pootrakulchote, N.; Alibabaei, L.; Ngoc-le, C. H.;
Decoppet, J. D.; Tsai, J. H.; Gratzel, C.; Wu, C. G.; et al. ACS Nano 2009, 3, 3103-3109.
(15) Yella, A.; Lee, H. W.; Tsao, H. N.; Yi, C. Y.; Chandiran, A. K.; Nazeeruddin, M. K.;
Diau, E. W. G.; Yeh, C. Y.; Zakeeruddin, S. M.; Gratzel, M. Science 2011, 334, 629-634.
(16) He, J. J.; Lindstrom, H.; Hagfeldt, A.; Lindquist, S. E. J. Phys. Chem. B 1999, 103, 8940-
8943.
(17) Morandeira, A.; Boschloo, G.; Hagfeldt, A.; Hammarstrom, L. J. Phys. Chem. B 2005,
109, 19403-19410.
18
(18) Borgstrom, M.; Blart, E.; Boschloo, G.; Mukhtar, E.; Hagfeldt, A.; Hammarstrom, L.;
Odobel, F. J. Phys. Chem. B 2005, 109, 22928-22934.
(19) Morandeira, A.; Fortage, J.; Edvinsson, T.; Le Pleux, L.; Blart, E.; Boschloo, G.;
Hagfeldt, A.; Hanmiarstrom, L.; Dobel, F. J. Phys. Chem. C 2008, 112, 1721-1728.
(20) Morandeira, A.; Boschloo, G.; Hagfeldt, A.; Hammarstrom, L. J. Phys. Chem. C 2008,
112, 9530-9537.
(21) Mori, S.; Fukuda, S.; Sumikura, S.; Takeda, Y.; Tamaki, Y.; Suzuki, E.; Abe, T. J. Phys.
Chem. C 2008, 112, 16134-16139.
(22) Qin, P.; Zhu, H. J.; Edvinsson, T.; Boschloo, G.; Hagfeldt, A.; Sun, L. C. J. Am. Chem.
Soc. 2008, 130, 17629-17629.
(23) Gibson, E. A.; Smeigh, A. L.; Le Pleux, L.; Fortage, J.; Boschloo, G.; Blart, E.; Pellegrin,
Y.; Odobel, F.; Hagfeldt, A.; Hammarstrom, L. Angew. Chem. Int. Ed. 2009, 48, 4402-
4405.
(24) Qin, P.; Linder, M.; Brinck, T.; Boschloo, G.; Hagfeldt, A.; Sun, L. C. Adv. Mater. 2009,
21, 2993-2996.
(25) Li, L.; Gibson, E. A.; Qin, P.; Boschloo, G.; Gorlov, M.; Hagfeldt, A.; Sun, L. C. Adv.
Mater. 2010, 22, 1759-1762.
(26) Odobel, F.; Le Pleux, L.; Pellegrin, Y.; Blart, E. Acc. Chem. Res. 2010, 43, 1063-1071.
(27) He, J. J.; Lindstrom, H.; Hagfeldt, A.; Lindquist, S. E. Sol. Energ. Mat. Sol. Cells 2000,
62, 265-273.
(28) Nakasa, A.; Usami, H.; Sumikura, S.; Hasegawa, S.; Koyama, T.; Suzuki, E. Chem. Lett.
2005, 34, 500-501.
(29) Nattestad, A.; Ferguson, M.; Kerr, R.; Cheng, Y. B.; Bach, U. Nanotechnology 2008, 19,
295304.
(30) Nattestad, A.; Mozer, A. J.; Fischer, M. K. R.; Cheng, Y. B.; Mishra, A.; Bauerle, P.;
Bach, U. Nature Mater. 2010, 9, 31-35.
(31) Serpone, N.; Jamieson, M. A.; Henry, M. S.; Hoffman, M. Z.; Bolletta, F.; Maestri, M. J.
Am. Chem. Soc. 1979, 101, 2907-2916.
(32) Serpone, N.; Jamieson, M. A.; Emmi, S. S.; Fuochi, P. G.; Mulazzani, Q. G.; Hoffman, M.
Z. J. Am. Chem. Soc. 1981, 103, 1091-1098.
(33) Serpone, N.; Jamieson, M. A.; Sriram, R.; Hoffman, M. Z. Inorg. Chem. 1981, 20, 3983-
3988.
(34) Jamieson, M. A.; Serpone, N.; Hoffman, M. Z. Coord. Chem. Rev. 1981, 39, 121-179.
(35) Hoffman, M. Z.; Serpone, N. Isr. J. Chem. 1982, 22, 91-97.
(36) Bolletta, F.; Maestri, M.; Moggi, L.; Jamieson, M. A.; Serpone, N.; Henry, M. S.;
Hoffman, M. Z. Inorg. Chem. 1983, 22, 2502-2509.
(37) Serpone, N.; Hoffman, M. Z. J. Chem. Educ. 1983, 60, 853-860.
(38) Fujishima, A.; Honda, K. Nature 1972, 238, 37-38.
(39) Basic Research Needs for Solar Energy Utilization - Report on the Basic Energy Sciences
Workshop on Solar Energy Utilization; Argonne National Laboratory (U.S. DOE), 2005.
19
(40) Kudo, A.; Miseki, Y. Chem. Soc. Rev. 2009, 38, 253-278.
(41) Chen, X. B.; Shen, S. H.; Guo, L. J.; Mao, S. S. Chem. Rev. 2010, 110, 6503-6570.
(42) Grӓtzel, M. Energy Resources through Photochemistry and Catalysis; Academic Press:
New York, N.Y., 1983.
(43) Wilker, M. B.; Schnitzenbaumer, K. J.; Dukovic, G. Isr. J. Chem. 2012, 52, 1002-1015.
(44) Buhler, N.; Meier, K.; Reber, J. F. J. Phys. Chem. 1984, 88, 3261-3268.
(45) Matsumura, M.; Furukawa, S.; Saho, Y.; Tsubomura, H. J. Phys. Chem. 1985, 89, 1327-
1329.
(46) Vaneski, A.; Susha, A. S.; Rodriguez-Fernandez, J.; Berr, M.; Jackel, F.; Feldmann, J.;
Rogach, A. L. Adv. Funct. Mater. 2011, 21, 1547-1556.
(47) Amirav, L.; Alivisatos, A. P. J. Phys. Chem. Lett. 2010, 1, 1051-1054.
(48) Brown, K. A.; Wilker, M. B.; Boehm, M.; Dukovic, G.; King, P. W. J. Am. Chem. Soc.
2012, 134, 5627-5636.
20
CHAPTER 2
Solution Phase Characterization of Strongly Photooxidizing Heteroleptic
Cr(III) Tris-Dipyridyl Complexes: Toward Efficient Hole Transfer Sensitizers
2.1 Introduction
A large body of research has clarified the physical and synthetic prerequisites for
achieving efficient light-to-electrical energy conversion in dye-sensitized solar cells (DSSCs)
wherein excited states of inorganic chromophores can inject electrons into wide band gap
semiconductors.1-7
Early experimental successes, promising economic factors, and the sheer
magnitude of the scientific issues involved have meant that other paradigms for dye-sensitization
of charge transport remain relatively unexplored. One such opportunity involves photoinduced
interfacial hole transfer. Optimization of this paradigm would expose numerous opportunities in
solar energy conversion, including initiation of catalytic oxidative reactions critical for water
splitting,8-17
photocathodic solar cells (where current runs in the direction opposite to Grӓtzel
cells),18-21
and tandem photovoltaic cells,22-24
where both electrodes are photoactive. The
theoretical efficiency limit for third generation25
photovoltaic tandem cells, where each of the
chromophores absorbs a different portion of the solar spectrum, is ∼45%.23,26
This compares
favorably to the maximum∼30% efficiency achievable in Grätzel-type cells operating with one
active electrode.
Despite these promises, relatively little is known about the physio-chemical factors that
must be controlled if photoinduced hole injection processes are to be exploited for solar energy
conversion. To our knowledge, there are only a few reports in the literature where this initial
21
photophysical mechanism drives a photocathodic current in a DSSC device.21,27-30
There are only
three systems reported where hole injection is time-resolved and shown to be ultrafast18,31-32
and
only three disclosures where hole transfer participates in a dye-sensitized heterojunction solar
cell.33-35
Finally, only in three reports has hole injection functioned as one-half of a tandem
photovoltaic cell.23,36-37
The latter of these is the current efficiency record holder for p-type
DSSCs (0.20% overall efficiency). Clearly, whereas the paucity of results alludes to the
significant challenges involved in this area, it also offers the freedom to explore new materials
and methods for controlling energetics and carrier-transfer rates.
Searching for molecular sensitizers capable of initiating excited-state oxidation of wide
band gap semiconductors, we note tris-dipyridyl complexes of Cr(III) as one promising class of
compounds. Serpone and Hoffman studied homoleptic analogues for solar energy conversion
purposes about 25 years ago.38-44
Parent complexes such as [Cr(bpy)3]3+
or [Cr(phen)3]3+
have
excited state redox potentials sufficient to oxidize water to dioxygen if 4e− oxidation could be
achieved. They also have long excited state lifetimes, which should promote hole injection into
an attached semiconductor surface. Although they absorb visible light ∼50 times more weakly
than [Ru(bpy)3]2+
(at 450 nm),38,45
chromium is several orders of magnitude more abundant than
ruthenium,46
and ligand modifications can improve absorption properties (vide infra).
Heteroleptic polypyridyl complexes of Cr(III) represent potentially functional model
systems, which to our knowledge have not been studied as components of hybrid materials.
Dipyridyl ligands with carboxylate functional groups located at the 4 and 4’ positions can serve
to anchor the sensitizer to metal oxide surfaces, as has been demonstrated extensively in Ru(II)-
containing analogues.1-7
As discussed in this chapter, the electronic properties of the Cr(III)
center can be tuned by judicious choice of the ancillary dipyridyl-type ligands (NN). Although
22
structurally homologous with Ru(II) complexes, the synthesis of heteroleptic Cr(III) dipyridyl
complexes is not straightforward, as efforts to activate the inert metal center often result in ligand
scrambling.47
Nevertheless, a recently disclosed methodology employing [(NN)2Cr(OTf)2]+
complexes as synthons47-49
shows the way to a new class of molecular species with potential for
efficient hole injection into semiconductor substrates. Herein, the preparations as well as
electrochemical and photophysical investigations of a family of structurally related heteroleptic
Cr(III) dipyridyl complexes (Scheme 2.1) are presented in this chapter. The solution phase
investigation of these compounds demonstrates their ability to act as strong photooxidants, and
the electronic flexibility afforded by ligand substitution allows us to explore fundamental
structure/function relationships in our search for efficient hole transfer to semiconductor
substrates.
Scheme 2.1 Target Structures of Cr(III) Complexes
23
2.2 Experimental Section
2.2.1 Preparation of Compounds
Unless otherwise noted, the syntheses of heteroleptic tris-dipyridyl Cr(III) complexes
were performed in air with atmospheric moisture excluded by use of a CaCO3-filled drying tube.
For synthetic routes employing Cr(II) starting materials and for the preparation of
[Cr(NN)2(OTf)2]OTf (OTf = trifluoromethanesulfonate), compound manipulations were
performed either inside a dinitrogen-filled glovebox (MBRAUN Labmaster 130) or via Schlenk
techniques on an inert gas (N2) manifold. The commercially obtained ligand 4,4’-dimethyl-2,2’-
bipyridine (Me2bpy) was recrystallized from ethyl acetate before use. The ligand dimethyl 2,2’-
bipyridine-4,4’-dicarboxylate (4-dmcbpy) was synthesized according to the literature.50
The
preparations of [Cr(phen)2(OTf)2](OTf), [Cr(bpy)2(OTf)2](OTf), and [Cr(CH3CN)4(BF4)2] have
been described elsewhere.51-52
The homoleptic complexes [Cr(NN)3](OTf)3, where (NN) is phen,
Ph2phen, or Me2bpy, were prepared by refluxing [Cr(NN)2(OTf)2]OTf in CH2Cl2,with 5 equiv of
the same (NN) ligand for 16 h and collecting the precipitated yellow solids by filtration. The
complex [Cr(bpy)3](BF4)3 was prepared analogously to [Cr(4-dmcbpy)3](BF4)3, using bpy in
place of 4-dmcbpy. Electronic absorption spectra,38
electrospray ionization mass spectrometry
(ESI-MS), and clean electrochemical traces confirmed the identity and purity of the previously
reported homoleptic complexes. Pentane was distilled over sodium metal and subjected to three
freeze-pump-thaw cycles. Other solvents were sparged with dinitrogen, passed over alumina, and
degassed prior to use. All other reagents were obtained from commercial sources and were used
without further purification.
24
[Cr(phen)2(4-dmcbpy)](OTf)3. Solid 4-dmcbpy (0.71 g, 2.62 mmol) was added to a
solution of [Cr(phen)2(OTf)2]OTf (1.50 g, 1.75 mmol) in 125 mL of dichloromethane and heated
to reflux. Over 5 days, a bright yellow precipitate formed. The solid was isolated by filtration,
washed with dichloromethane (3 x 30 mL), and dried in vacuo to afford 1.86 g (94%) of product.
IR (KBr pellet): νC=O 1728 cm-1
. μeff (295 K): 3.90 μB. ES+MS (CH3CN): m/z 228.27 ([M −
3OTf]3+
), 981.67 ([M − OTf]+). Anal. Calcd. for C41H28N6CrF9O13S3: C, 43.51; H, 2.49; N, 7.42.
Found: C, 43.23; H, 2.33; N, 7.27. Crystals suitable for X-ray analysis were obtained by slow
diffusion of diethyl ether into an acetonitrile solution of the compound.
[Cr(Ph2phen)2Cl2]Cl. Solid anhydrous CrCl3 (0.10 g, 0.60 mmol) was added to a
suspension of Ph2phen (0.40 g, 1.20 mmol) in 35 mL of absolute ethanol. A trace amount (<2 mg)
of zinc dust was added, and the mixture was heated to reflux for 1 h, resulting in a green-brown
mixture. The mixture was filtered, and an olive green solid was isolated from the filtrate by
rotary evaporation to afford 0.49 g (98%) of product. IR (KBr pellet): νC=N 1620 cm-1
. ES+MS
(CH3CN): m/z 788.27 ([M−Cl]+). The compound was used in the next synthetic step without
further purification or characterization.
[Cr(Ph2phen)2(OTf)2]OTf. Under a dinitrogen atmosphere, trifluoromethanesulfonic
acid (2 mL, 22.60 mmol) was slowly added to solid [Cr(Ph2phen)2Cl2]Cl (0.40 g, 0.49 mmol) to
give a red-orange solution. Dinitrogen was bubbled through the stirring solution for 24 h, after
which the solution was cooled in an ice bath and 250 mL of diethyl ether was added. After
standing 4 h, a beige-peach colored solid precipitated from solution. The solid was isolated by
vacuum filtration and rinsed with diethyl ether (3 x 30 mL) to afford 0.47 g (83%) of product. IR
(KBr pellet): νC=N 1625 cm-1
. ES+MS (CH3CN): m/z 1014.20 ([M − OTf]
+). The compound was
used in the next synthetic step without further purification or characterization.
25
[Cr(Ph2phen)2(4-dmcbpy)](OTf)3. Solid 4-dmcbpy (0.140 g, 0.515 mmol) was added to
a solution of [Cr(Ph2phen)2(OTf)2]OTf (0.209 g, 0.17 mmol) in 30 mLof dichloromethane and
heated to reflux. Over 14 days, a yellow precipitate formed. The solid was isolated by filtration,
washed with dichloromethane (3 x 10 mL) and collected to afford 0.07 g (27%) of product. IR
(KBr pellet): νC=O 1734 cm-1
. μeff (295K): 3.36 μB. ES+MS (CH3CN): m/z 329.80 ([M − 3OTf]
3+),
1285.60 ([M − OTf]+). Anal. Calcd. for C65H44N6CrF9O13S3: C, 54.36; H, 3.09; N, 5.85. Found:
C, 54.12; H, 3.05; N, 5.75.
[Cr(Me2bpy)2Cl2]Cl. Solid anhydrous CrCl3 (0.43 g, 2.71 mmol) was added to a solution
of Me2bpy (1.00 g, 5.43 mmol) in 60 mL of absolute ethanol. A trace amount (<4mg) of zinc
dust was added, and the mixture heated to reflux for 1 h, resulting in a green brown solution. A
gray-green solid precipitated from the reaction mixture upon cooling. It was collected by
filtration, washed with cold absolute ethanol (3 x 10 mL), and dried in vacuo to afford 1.12 g
(78%) of product. IR (KBr pellet): νC=N 1616 cm-1
. ES+MS (CH3CN): m/z 490.00 ([M − Cl]
+).
The compound was used in the next synthetic step without further purification or
characterization.
[Cr(Me2bpy)2(OTf)2]OTf. Under a dinitrogen atmosphere, trifluoromethanesulfonic acid
(2 mL, 22.60 mmol) was added to solid [Cr(Me2bpy)2Cl2]Cl (0.35 g, 0.67 mmol) to give a red-
orange solution. Dinitrogen was bubbled through the stirring solution for 24 h, after which the
solution was cooled in an ice bath. Diethyl ether (250 mL) was slowly added to form a pink
precipitate. The solid was isolated by filtration and rinsed with diethyl ether (3 x 50 mL) to
afford 0.53 g (92%) of product. IR (KBr pellet): νC=N 1627 cm-1
. ES+MS (CH3CN): m/z 718.13
([M − OTf]+). The compound was used in the next synthetic step without further purification or
characterization.
26
[Cr(Me2bpy)2(4-dmcbpy)](OTf)3. Solid 4-dmcbpy (0.07 g, 0.26 mmol) was added to a
solution of [Cr(Me2bpy)2(OTf)2]OTf (0.20 g, 0.23 mmol) in 30 mL of dichloromethane and
heated to reflux. Over 9 days, a light yellow precipitate was formed. The solid was isolated by
filtration, washed with dichloromethane (3 x 10 mL) and dried in vacuo to afford 0.13 g (48%)
of product. IR (KBr pellet): νC=O 1739 cm-1
. μeff (295 K): 4.01 μB. ES+MS (CH3CN): m/z 230.93
([M − 3OTf]3+
), 989.60 ([M − OTf]+). Anal. Calcd. for C41H36N6CrF9O13S3: C, 43.20; H, 3.18; N,
7.37. Found: C, 42.94; H, 3.11; N, 7.28.
[Cr(4-dmcbpy)3](BF4)3. Under a dinitrogen atmosphere, a solution of
[Cr(CH3CN)4(BF4)2] (0.13 g, 0.74 mmol) in 4 mL of acetonitrile was added to a suspension of 4-
dmcbpy (0.33 g, 2.44 mmol) in 4 mL of acetonitrile to form a forest green solution. The solvent
was removed in vacuo to afford a forest green solid. Addition of AgBF4 (0.04 g, 0.19 mmol) to a
solution of the isolated green solid (0.20 g, 0.19 mmol) in 5 mL of acetonitrile resulted in a
yellow solution along with a light gray solid. The solution was filtered, and the filtrate was
treated with 15 mL of diethyl ether to precipitate a bright yellow solid. The solid was
recrystallized by diethyl ether diffusion into acetonitrile resulting in bright yellow crystals. The
crystals were collected by filtration, washed with dichloromethane (3 x 3 mL) followed by
diethyl ether (3 x 5 mL) and dried in vacuo to afford 0.10 g (45%) of product. IR (mineral oil):
νC=O 1737 cm-1
. μeff (295K): 4.15 μB. ES+MS (CH3CN): m/z 289.67 ([M − 3BF4]
3+). Anal. Calcd.
for C42H36N6CrF12B3O12: C, 44.67; H, 3.21; N, 7.44. Found: C, 44.46; H, 3.08; N, 7.28.
27
2.2.2 X-ray Structure Determination
A suitable crystal of [Cr(phen)2(4-dmcbpy)](OTf)3 ∙1.3CH3CN was coated with Paratone-
N oil and supported on a Cryoloop before being mounted on a Bruker Kappa Apex II CCD
diffractometer under a stream of dinitrogen. Data collection was performed at 110 K with Mo Kα
radiation and a graphite monochromator. Crystallographic data and metric parameters are
presented in Table 2.1. Data were integrated and corrected for Lorentz and polarization effects
using SAINT, and semiempirical absorption corrections were applied using SADABS.53
The
structure was solved by direct methods and refined against F2 with the SHELXTL 6.14 software
package.54
Unless otherwise noted, thermal parameters for all non-hydrogen atoms were refined
anisotropically. Hydrogen atoms were added at the ideal positions and were refined using a
riding model where the thermal parameters were set at 1.2 times those of the attached carbon
atom (1.5 for methyl protons). In the structure of [Cr(phen)2(4-dmcbpy)](OTf)3 ∙1.3CH3CN, two
of the triflate anions show positional disorder, and one solvent molecule is only partially
occupied.
Table 2.1. Crystallographic Dataa for [Cr(phen)2(4-dmcbpy)](OTf)3 ∙1.3CH3CN
Formula C43.60H31.90CrF9N7.30 O13S3 α, deg 99.574(4)
Formula Weight 1185.24 β, deg 106.338(4)
Color and Habit yellow plates γ, deg 115.479(2)
T, K 110(2) V, Å3 2477.6(3)
Space Group (triclinic) dcalc, g/cm3 1.589
Z 2 GOF 1.026
a, Å 12.6215(7) R1 (wR2)b, % 4.64 (10.17)
b, Å 13.9620(8)
c, Å 17.2034(15) a Obtained with graphite-monochromated Mo Kα (λ = 0.71073 Å) radiation.
b R1 = ∑||Fo| − |Fc|| / ∑|Fo|, wR2 = [∑w(Fo
2 − Fc
2)2/∑w(Fo
2)2]1/2
for Fo>4σ(Fo).
28
2.2.3 Photophysical Measurements
All photophysical measurements were undertaken with complexes dissolved in 1 M
HCl(aq). This alters the nucleophilicity of the solvent, thereby decreasing the quantum yield of
associative excited state reactions (formation of seven-coordinate solvento species) which
ultimately would lead to polypyridyl ligand substitution by solvent molecules.38,55-56
The ground-
state absorption spectra were obtained with a Hewlett-Packard 8453 spectrophotometer in quartz
cuvettes with 1 cm or 1 mm path lengths; experiments were performed at room temperature.
Quartz cuvettes (1 cm x 1 cm) with silicone septa seal screw caps were used for the following
measurements. Emission spectra, emission quantum yields, and single-temperature emission
lifetime experiments were measured at 23-24 °C with deoxygenation on dilute samples. Teflon
tubing was used to introduce a stream of argon into the solution. Prior to measurements, samples
were purged with argon for 30 min to remove oxygen. For the longer lifetime species such as
[Cr(phen)3]3+
, which are highly sensitive to 3O2 concentrations, this methodology proved easier
and far more effective at achieving reproducible results than using freeze-pump-thaw cycles.
Each measurement was done with argon flowing on top of the solution.
Emission Spectra and Quantum Yields. The complex solutions were prepared with an
absorbance of ~0.1 at 320 nm. Emission spectra were recorded at 23-24 °C using Photon
Technology International Fluorometer System with an Ushio UXL-75XE Xenon Short Arc
Lamp and PTI-814 Photomultiplier Detection System with a Hamamatsu R928P photomultiplier
tube operating at –1000 V dc. The measurements were carried out at a single excitation
wavelength of 320 nm, and the emission at 90° relative to the excitation light was recorded from
650 to 850 nm. Emission spectra reported here were corrected for instrument response using a
tungsten lamp provided by the manufacturer, which has been calibrated against a NIST tungsten
29
lamp. For the normalized emission quantum yields reported in Table 2.4, we compared the
integrated emission of each complex relative to that of the standard [Cr(phen)3]3+
in 1 M HCl(aq)
according to the Equation 2.1.57
Both measurements are made back to back.
In this expression, unk and std are the emission quantum yields of the unknown and standard,
respectively, under the conditions of the measurement. To our knowledge std (for [Cr(phen)3]3+
)
has not been previously measured and is set to one in this column of Table 2.4 as has been done
previously.38,56
The quantities Iunk and Istd are the integrated emission intensities of the sample
and the standard, respectively, at 650-850 nm. The quantities Aunk and Astd are the absorbances of
the sample and the standard, respectively, at the excitation wavelength (320 nm). Care was taken
to ensure that these are both close to 0.1. Finally ηunk and ηstd are the indices of refraction of the
sample and the standard solution, respectively. Since the same solvent was used in both
measurements, the last termineq1canbe ignored. For these measurements, no differences were
observed in absorption spectra collected before and after the emission measurements. To
determine the emission quantum yields reported in Table 2.4, we first measured unk for
[Cr(phen)3](OTf)3 in 1 M HCl(aq) relative to [Ru(bpy)3](PF6)2 in acetonitrile for which the
absolute quantum yield ( std) of 0.062 is known.58
The refractive indices of pure acetonitrile and
1 M HCl(aq) were used in the calculation. The quantum yields for the rest of the seven complexes
were then calculated with respect to unk determined for [Cr(phen)3]3+
. The error bars reported
with these seven quantum yields are determined by combining the percentage experimental
errors from the individual normalized emission quantum yields (A%) with the percentage error
30
in the measurement made between [Cr(phen)3]3+
and [Ru(bpy)3]2+
(B%) according to the
equation Error %=(A2 + B
2)1/2
%.
Emission Lifetimes. Sample solutions were prepared and deoxygenated in exactly the
same manner as described above for emission spectra and quantum yields measurements. The
measurements were done at 23-24 °C. Each sample was excited with 355 nm laser pulses (10 Hz;
~0.3 μJ/pulse) generated by a Continuum Minilite II Q-Switched Nd:YAG Laser. Two dielectric
mirrors and one bandpass filter (300-360 nm) in front of the sample were used to eliminate
residual 532 nm and 1064 nm laser radiation. The emitted light was detected at 90° with respect
to the excitation laser using a Hamamatsu H9305-02 photomultiplier tube (PMT) operating at
−900 V dc. A notch filter (725 ± 5 nm) was placed within a lens-tube assembly (two back-to-
back plano-convex lenses, fo = 25.4 mm/diameter = 25.4 mm) in front of the detector and the
emission intensity was monitored. The PMT signal was terminated with a 50 Ω resistor into a
LeCroy 9384L Oscilloscope. Maximum signal levels were kept well below the measured
linearity of the PMT response. Each emission transient was acquired by averaging 3000 scans
with a time window of 180 to 4000 μs depending on the lifetimes of the complexes. The data
from the oscilloscope was transferred to a personal computer and fit with a single exponential
decay model using Labview code of local origin to obtain the lifetimes. For any single lifetime
measurement (3000 scans), no sample degradation is observed in UV-Vis comparisons (after vs
before).
In longer experiments such as the temperature dependent studies, samples that include
bipyridine and/or bipyridine derivatives (as opposed to phenanthroline species), except for [Cr(4-
dmcbpy)3]3+
, show minor degradation (less than ~10 % based on absorbance change at the
excitation wavelength) provided care has been taken to keep the excitation power low. Complex
31
[Cr(4-dmcbpy)3]3+
shows a ~13 % decrease in the absorbance at 355 nm and ~60 % increase at
325 nm. This is possibly the solvento species due to ligand replacement of 4-dmcbpy by H2O. A
similar phenomenon has been observed for [Cr(bpy)3]3+
.59-63
We have measured the emission
spectrum for [Cr(4-dmcbpy)3]3+
before and after heating to 80 °C. While the intensity decreases,
no other bands are observed in the window of 650 – 850 nm.
Transient Absorption Spectra. Transient absorption kinetics were measured every 10
nm from 350 to 650 nm. The excitation laser pulses at 355 nm (2 Hz; 1.5 mJ/pulse) are derived
from the Continuum Minilite II Q-Switched Nd:YAG Laser described above. In this transient
absorption spectrometer setup, the laser beam is cylindrically focused onto the sample housed in
a 1 cm x 1 cm quartz cuvette at 90° to the probe beam. A Newport/Oriel lamp system with a 75
W Xenon arc lamp (ozone-free, part number 6263) is used to provide the broad-band probe
source. A 1 in. diameter beam of this light is focused into the sample using a plano-convex fo =
100 mm lens achieving a spot size of ~1 mm diameter (approximately collimated in the 1 cm
path length cuvette) and a probe volume that overlaps with the cylindrically focused pump light.
Prior to the sample, the white light is passed through a cylindrical cell filled with water to
remove IR radiation. Following the sample, the probe beam is re-collimated and then focused via
two plano-convex lenses (fo = 50 mm) into a monochromator. Based on entrance and exit slit
widths that were used (1 mm), the wavelengths probed span ± 2 nm. Probe intensity as a function
of wavelength is monitored with a negatively-biased Hamamatsu R-928 PMT operating at –1000
V dc. The standard PMT circuit has been modified to use only the first 4 dynode stages to
accommodate the probe flux (a non-zero background) with good dynamic range.64
In the
experiments reported herein, the signal from the PMT was through an external 460 Ω resistor
into a LeCroy 9384L Oscilloscope. Transient absorption or bleach kinetics at each wavelength
32
were obtained by averaging 30 oscilloscope time-traces of probe intensity with the pump laser on
followed immediately by averaging 30 time-traces of probe intensity with the pump off. The
time window on the oscilloscope always accommodated at least 10 multiples of the measured
lifetime. These pump-on and pump-off time traces were transferred to a personal computer and
processed with Labview programs of local origin to determine a ∆A signal as a function of time
(at a given wavelength of data collection). The whole collection process was repeated three times
and an average ∆A(t) (at each wavelength of interest) was determined. Each ∆A(t) signal at each
wavelength of interest was fit with a single exponential decay model. Transient absorption
spectra reported herein consist of a plot of the pre-exponential of these ∆A(t) kinetics fits as a
function of wavelength.
In order to minimize the effect of photochemistry on transient spectra, four sample
solutions were used to obtain each spectrum (each sample accounts for 7-8 data points in the
final ∆A spectrum). A bulk solution (15 mL) of complex was pre-prepared with an absorbance of
~0.4 at 355 nm. An aliquot of 3.5 mL solution was transferred into a quartz cuvette for each set
of 7-8 ∆A(t) measurements used to build the ∆A(λ) spectrum. No major degradation (less than
~10 % based on absorbance change at the excitation wavelength) was observed. The ∆A(t)
kinetics show single exponential decay and for the spectra reported herein, samples were not
deoxygenated.
2.2.4 Other Physical Methods
Infrared spectra were measured with a Nicolet 380 FT-IR spectrometer. Mass
spectrometric measurements were performed in the positive ion mode on a Finnigan LCQ Duo
33
mass spectrometer, equipped with an analytical electrospray ion source and a quadrupole ion trap
mass analyzer. Cyclic voltammetry experiments were carried out inside a dinitrogen filled
glovebox in 0.1 M solutions of (Bu4N)PF6 in acetonitrile unless otherwise noted. The
voltammograms were recorded with either a CH Instruments 1230A or 660C potentiostat using a
0.25 mm Pt disk working electrode, Ag wire quasireference electrode, and at a Pt mesh auxiliary
electrode. All voltammograms shown were measured with a scan rate of 0.1 V/s. Reported
potentials are referenced to the ferrocenium/ferrocene (Fc+/Fc) redox couple and were
determined by adding ferrocene as an internal standard at the conclusion of each electrochemical
experiment. Solid state magnetic susceptibility measurements were performed on finely ground
samples prepared in air using a Quantum Design model MPMS-XL SQUID magnetometer at
295 K. The data were corrected for the magnetization of the sample holder by subtracting the
susceptibility of an empty container; diamagnetic corrections were applied using Pascal’s
constants.65
Elemental analyses were performed by Robertson Microlit Laboratories, Inc. in
Madison, NJ.
34
2.3 Results and Discussion
2.3.1 Synthesis and Characterization of Heteroleptic Cr(III) Complexes
Although there is literature precedent for heteroleptic Cr(III) dipyridyl complexes, the
preparation of species that contain at least one carboxylate group (for eventual attachment to
semiconductor surfaces) was not known prior to our efforts. The preparative routes we have
developed are outlined in Schemes 2.2 and 2.3. Typically, heteroleptic dipyridyl complexes of
Cr(III) are synthesized using [Cr(NN)2(OTf)2](OTf) precursors: the weakly coordinating triflate
anions can be facilely removed from otherwise inert Cr(III) centers, and replaced by a third
diimine species with minimal ligand scrambling.47-49
Initial attempts to prepare complexes using
2,2’-bipyridine-4,4’-dicarboxylic acid (4-dcbpy) or its disodium salt do not afford pure products.
Whereas mass spectral analyses show peaks at anticipated m/z ratios (consistent with
[Cr(phen)2(4-dcbpy)]+), analysis of isotopic distribution patterns reveal 2+ charges for those ions.
We speculate that the products actually formed are dimeric [Cr(NN)2(4-dcbpy)]22+
species,
where carboxylates coordinate in preference to the imines because of the oxophilic nature of
Cr(III), allowing the 4-dcbpy ligand to bridge between twometal centers. Masking the
carboxylates on 4-dcbpy by conversion to methyl ester groups (4-dmcbpy) avoids undesirable
metal coordination by the carboxylates. Where (NN) is phen, Ph2phen, or Me2bpy, the ester-
protected ligand 4-dmcbpy is found to react cleanly, albeit sluggishly, with
[Cr(NN)2(OTf)2](OTf) via Scheme 2.2 to form the heteroleptic complexes [Cr(phen)2(4-
dmcbpy)](OTf)3, [Cr(Ph2phen)2(4-dmcbpy)](OTf)3, and [Cr(Me2bpy)2(4-dmcbpy)](OTf)3,
respectively, as yellow solids.
35
Scheme 2.2 Preparative Routes for the [Cr(NN)2(4-dmcbpy)](OTf)3 Complexes
Scheme 2.3 Preparative Routes for [Cr(4-dmcbpy)3](BF4)3
The homoleptic complex salt [Cr(4-dmcbpy)3](OTf)3 cannot be prepared via Scheme 2.2,
since deprotection of the esters during triflate exchange with [Cr(4-dmcbpy)2Cl2]Cl results in
dimerization of the Cr complexes. Instead, we find that a Cr(II) solvento complex,
[Cr(CH3CN)4(BF4)2], can serve as a suitable starting material.52
The labile Cr(II) ion is easily
ligated by 3 equiv of the 4-dmcbpy ligand. Oxidation by Ag(I) affords the tetrafluoroborate salt
of the homoleptic Cr(III) complex in reasonable yield (Scheme 2.3).
In contrast to the phen-containing heteroleptic complex [Cr(phen)2(4-dmcbpy)](OTf)3,
the bpy-containing analogue [Cr(bpy)2(4-dmcbpy)]3+
resists formation via Scheme 2.2, although
the reasons for this are not known at this time. To our knowledge, only two successful syntheses
of bpy-containing heteroleptic complexes have been reported, [Cr(bpy)2(phen)](OTf)3 and
[Cr(bpy)2(DPPZ)](OTf)3 (where DPPZ is dipyridophenazine).47,66
We speculate that the ring-
locked configuration and higher basicity of the phenanthroline-type ligands minimize
opportunities for deleterious ligand exchange before replacement of the more labile triflate
anions, whereas the less rigid dipyridyl ligands offer greater opportunity for ligand scrambling.
Attempts to make heteroleptic bpy-containing complexes via the more reactive Cr(II) synthon
(Scheme 2.3) have led thus far to intractable mixtures of homo- and heteroleptic products.
36
Figure 2.1 X-ray crystal structure of the [Cr(phen)2(4-dmcbpy)]3+
complex cation, as observed
in [Cr(phen)2(4-dmcbpy)](OTf)3∙1.3CH3CN, rendered with 40% ellipsoids. H atoms are omitted
for clarity. The complex resides on a general position. Bond lengths and angles are in Table 2.2.
Figure 2.2 Full crystal structure for the complex [Cr(phen)2(4-dmcbpy)](OTf)3∙1.3CH3CN,
rendered at 40% thermal ellipsoids. H atoms are not numbered in the figure. Minor components
of disordered sites are shown in dashed circles.
37
Table 2.2. Selected Bond Lengths and Angles for [Cr(phen)2(4-dmcbpy)](OTf)3∙1.3CH3CN.
Bond Lengths (Å)
Cr1-N1 2.048(2) Cr1-N12 2.064(2) Cr1-N15 2.051(2)
Cr1-N26 2.047(2) Cr1-N36 2.032(2) Cr1-N41 2.043(2)
Bond Angles (deg)
N1-Cr1-N12 80.57(8) N1-Cr1-N15 91.19(8) N15-Cr1-N12 88.11(8)
N26-Cr1-N1 168.66(8) N26-Cr1-N12 90.75(8) N26-Cr1-N15 81.17(8)
N36-Cr1-N1 93.76(8) N36-Cr1-N12 97.95(8) N36-Cr1-N15 172.72(8)
N36-Cr1-N26 94.67(8) N36-Cr1-N41 79.31(8) N41-Cr1-N1 94.89(8)
N41-Cr1-N12 174.58(8) N41-Cr1-N15 94.97(8) N41-Cr1-N26 94.12(8)
Torsion Angles (deg)
N1-C14-C13-N12 0.2(3) N15-C28-C27-N26 -2.7(3)
N36-C38-C39-N41 1.6(3) C37-C38-C39-C40 2.5(4)
Besides the usual methods employed for identification and characterization of the
complexes, the solid state structure of [Cr(phen)2(4-dmcbpy)]3+
has been confirmed by X-ray
crystallography (Figure 2.1 and Figure 2.2). All bond distances and angles are as expected for a
Cr(III) ion in a tris-chelate ligand environment. Relevant crystallographic data are shown in
Table 2.1.
2.3.2 Exploration of Photophysics and Electrochemistry of Cr(III) Systems in Solution
We have explored the basic photophysical and electrochemical properties of these
systems to better understand the nascent and long-lived excited states that will be called upon to
drive hole-injection following photoexcitation in later studies. Key questions to be addressed
38
here include: (1) how efficient are these complexes as optical absorbers; (2) how much energy is
stored in the excited state; (3) for how long is it stored in unbound solution phase systems; (4)
with what driving force might we expect hole transfer photochemistry; and (5) what are the
transient absorption features we might use in later femtosecond pump/probe studies to determine
hole-injection rates.
2.3.3 Electronic Absorption
Sensitization of hole transfer photochemistry demands light absorption in the material-
bound metal complex as an initial step. If Cr(III) polypyridyl complexes are to serve in this role
it is important that ligand modifications render visible light absorption more efficient than pure
spin-allowed ligand-field excitation (ε = 50-100 M-1
cm-1
). UV-visible absorption spectra in 1 M
HCl(aq) for complexes containing substituted bipyridine ligands are shown in Figure 2.3. Also
included for purposes of comparison are spectra for [Cr(bpy)3]3+
and [Cr(Me2bpy)3]3+
.
Absorption data for all complexes considered in this manuscript are presented in Table 2.3. Note
that spectra for previously reported homoleptic complexes38
were reacquired to allow for
unambiguous quantitative comparisons to the new heteroleptic complexes.
As shown in Figure 2.3, these complexes are strongly absorptive in the UV owing to
ligand-centered π → π* transitions; these are also observed in the free ligands. For [Cr(bpy)3]3+
,
[Cr(Me2bpy)3]3+
, and [Cr(Me2bpy)2(4-dmcbpy)]3+
, this includes the band in the vicinity of 300
nm. Ligand-centered π → π* transitions due to the presence of 4-dmcbpy are seen red-shifted by
∼10 nm in [Cr(Me2bpy)2(4-dmcbpy)]3+
and ∼30 nm in [Cr(4-dmcbpy)3]3+
.
39
Figure 2.3 Electronic absorption spectra for Cr(III) dipyridyl complexes in 1 M HCl(aq) at room
temperature.
For [Cr(bpy)3]3+
, [Cr(Me2bpy)3]3+
, and [Cr(Me2bpy)2(4-dmcbpy)]3+
, the band in the
vicinity of 350 nm appears to be charge transfer in nature based on its intensity and the fact that
it is absent in the free ligand. A similar band is observed for [Cr(4-dmcbpy)3]3+
as a shoulder at
∼375 nm. This red shift is consistent with expectations for metal-to-ligand charge-transfer
(MLCT) given the presence of electron withdrawing substituents on the 4-dmcbpy ligands.
However, one might also expect red shifting for ligand-to-metal charge transfer (LMCT) if such
ligands serve to reduce electron-electron repulsion in the metal-centered orbitals. It is noted that
the∼350 nm band in [Cr(bpy)3]3+
is most often attributed to LMCT51,67-68
since there is an
energetic penalty for oxidizing Cr(III) to Cr(IV) as would formally occur during MLCT.
For each of the bpy-containing complexes, a broad and modestly structured absorption
feature is observed tailing into the visible spectrum (Figure 2.3 inset). Even for the known tris-
homoleptic complexes [Cr(bpy)3]3+
and [Cr(Me2bpy)3]3+
, the observed molar absorptivities are
larger than would be expected for pure spin-allowed ligand field transitions (4A2 →
4T2), and
40
have been discussed in the literature.51
Here we accept that trigonal splitting is small (∼20 cm-
1)69-70
and that these systems can be discussed with state designations normally reserved for
systems with octahedral symmetry. The origin of this intensity enhancement is currently
uncertain, but it has been claimed that spin-spin coupling between the (4A2) Cr(III) center and the
triplet states of the aromatic ligand is a possible source.51,71-72
Juban and McCusker have argued
that visible absorption enhancement in [Cr(bpy)3]3+
is due to intensity borrowing from ligand-
centered transitions.73
This should be kept in mind as a possible mechanism in future
explorations of 4-dmcbpy ligand-containing systems. It is noted here that for a Cr(III) species,
[Cr(4-dmcbpy)3]3+
shows appreciable sensitization of visible light, with a molar absorptivity of
940 M-1
cm-1
at 450 nm. This portends eventual control of this critical property through judicious
structural and electronic modifications.
Figure 2.4 Electronic absorption spectra for Cr(III) complexes containing phenanthroline-based
ligands: [Cr(phen)3]3+
and [Cr(phen)2(4-dmcbpy)]3+
in 1 M HCl(aq), [Cr(Ph2phen)3]3+
and
[Cr(Ph2phen)2(4-dmcbpy)]3+
in 1 M HCl(aq) with MeOH (2% v/v). All spectra were collected at
room temperature.
41
Figure 2.4 presents UV-visible absorption spectra for homo- and heteroleptic complexes
containing phen-based ligands. Addition of 2% MeOH by volume to 1 M HCl(aq) was necessary
to increase solubility for the Ph2phen-containing complexes.38
Like the bipyridine-containing
complexes discussed in Figure 2.3, intense ligand-centered π → π* transitions are observed in the
UV. For complexes with the unsubstituted phen ligand, this is most prominently observed at
∼270 nm with very little shifting relative to the free ligand (264 nm for phen in CH2Cl2). Such
transitions are significantly stronger and red-shifted in complexes containing the Ph2phen ligand
as evidenced by intense absorption bands at ∼285 nm and ∼310 nm. This is likely due to the
presence of a larger and more delocalized π-system. For these two complexes, [Cr(Ph2phen)3]3+
and [Cr(Ph2phen)2(4-dmcbpy)]3+
, new absorption features appear at ∼375 nm upon formation of
the metal complex. It is also noted that the intensity of the ∼310 nm band mentioned above
changes significantly relative to the free ligand where it appears as a shoulder to the bluer π → π*
band. We believe these substantive changes upon complexation herald charge transfer transitions
(again, LMCT, MLCT, or both). For [Cr(phen)3]3+
and [Cr(phen)2(4-dmcbpy)]3+
, features absent
in the free ligand spectra are observed at ∼340 nm. The Ph2phen-containing species
[Cr(Ph2phen)2(4-dmcbpy)]3+
shows even more promising absorption of the visible spectrum than
[Cr(4-dmcbpy)3]3+
discussed above: at 450 nm, ε = 1960 M−1
cm−1, and at 480 nm, ε = 1490
M−1
cm−1
.
42
Table 2.3. Room Temperature Electronic Absorptions for Cr(III) Tris-Dipyridyl Complexes
Complex λmax/nm (ε/M-1
cm-1
)
[Cr(phen)3](OTf)3 269 (52500); 285 (31800), sh; 323 (11200), sh; 342
(6810), sh; 358 (3710); 405 (821), sh; 435 (558), sh;
454 (285), sh
[Cr(Ph2phen)3](OTf)3 283 (103000); 307 (108000); 362 (35300), sh; 380
(26200), sh; 445 (2820), sh; 484 (1730), sh
[Cr(bpy)3](OTf)3 265 (15700); 275 (15500); 305 (21300), sh; 313
(23100); 346 (8100); 360 (5680), sh; 402 (950), sh;
428 (682), sh; 458 (291), sh
[Cr(Me2bpy)3](OTf)3 278 (23100); 307 (25800); 342 (8530); 354 (6590),
sh; 394 (892), sh; 418 (602), sh; 446 (256), sh
[Cr(phen)2(4-dmcbpy)](OTf)3 269 (50600); 285 (29900), sh; 329 (14300); 368
(5020), sh; 418 (1130), sh; 450 (634), sh; 480 (163),
sh
[Cr(Ph2phen)2(4-dmcbpy)](OTf)3 284 (63300); 311 (59300); 368 (21500), sh; 389
(11600), sh; 455 (1950), sh; 491 (1270), sh
[Cr(Me2bpy)2(4-dmcbpy)](OTf)3 279 (23000), sh; 308 (21500); 333 (13500); 351
(9570), sh; 369 (4410), sh; 398 (1140), sh; 422 (866),
sh; 451 (429), sh
[Cr(4-dmcbpy)3](BF4)3 294 (14800), sh; 322 (20000), sh; 333 (22700); 360
(10500), sh; 372 (8210), sh; 420 (1280), sh; 448
(1020), sh; 480 (415), sh
43
2.3.4 Static Emission
Each of the bis-heteroleptic polypyridyl Cr(III) complexes that we have synthesized is
emissive at room temperature following electronic excitation. Emission spectra are shown in
Figure 2.5 for dilute samples of the four ester-containing species in thoroughly deoxygenated 1
M HCl(aq) following excitation at 320 nm. The peaks of these spectra have been scaled to reflect
relative intensity with respect to the nearly simultaneous measurement of a standard
([Cr(phen)3]3+
in thoroughly deoxygenated 1 M HCl(aq)). We note that the wavelength of
maximum emission at ∼730 nm in [Cr(phen)2(4-dmcbpy)]3+
, [Cr(Me2bpy)2(4-dmcbpy)]3+
, and
[Cr(4-dmcbpy)3]3+
is largely invariant to any changes made to the polypyridyl ligands.
Figure 2.5 Emission spectra for Cr(III) polypyridyl complexes in deoxygenated 1 M HCl(aq)
following excitation at 320 nm. On this intensity scale, the value 1 corresponds to the peak
height in emission collected for [Cr(phen)3]3+
.
44
As shown in Table 2.4, the model homoleptic species [Cr(phen)3]3+
, [Cr(bpy)3]3+
, and
[Cr(Me2bpy)3]3+
emit most strongly at this approximate wavelength. Such emission λmax
invariance is common to Cr(III) polypyridyl species and indicates that the lowest energy excited
state is insensitive to the ligand field, as would be the case for a 2E state with a t2g
3 configuration
involving a spin flip.44
By analogy to a large number of known emissive Cr(III) polypyridyl
complexes, the main emission band seen at ∼730 nm is assigned to the 2E → 4
A (ground state)
transition and the shoulder that occurs at ∼700 nm is assigned to the 2T →
4A (ground state)
transition.38
Interestingly, and somewhat counter to the above discussion, Figure 2.5 and Table
2.4 show that the two species containing the Ph2phen ligand, [Cr(Ph2phen)3]3+
and
[(Ph2phen)2Cr(4-dmcbpy)]3+, have an emission λmax that is shifted ∼14 nm to the red. For
[Cr(Ph2phen)3]3+
this has also been reported elsewhere, although no explanation has been
offered.38
Electron delocalization via the larger π-system of the aryl substituted ligand would
reduce electron-electron repulsion in the t2g3 configuration of the
2E state. However, a complete
explanation of emission shifting must be more subtle as a similar reduction of repulsion in the
t2g3 configuration of the
4A ground state might be expected to cancel the effect in the excited
state thereby leading to no shift in the emission maximum. One plausible explanation lies in the
fact that doublet configurations involving spin pairing within individual metal orbitals contribute
to the description of the emissive states in all of the these molecules explored herein, but for the
aryl-substituted versions, the energetic perturbation due to these configurations is smaller as
intraligand electronic delocalization takes effect.74
Ultimately it will be important to determine
whether there are connections between these emission shifting effects and the states playing roles
in the absorption intensity borrowing (vide supra).
45
Table 2.4. Photophysical Data for Cr(III) Polypyridyl Complexes in 1 M HCl(aq)
Complex Emission
a λmax
(nm)
E00b
(eV)
Normalized
emission
quantum
yielda,c,d
Emission
quantum
yielda,e
(x103)
[Cr(phen)3](OTf)3 730 1.70 1 12 ± 1c
[Cr(Ph2phen)3](OTf)3 744 1.67 2.5 ± 0.4 30 ± 5
[Cr(bpy)3](OTf)3 729 1.71 0.21 ± 0.01 2.5 ± 0.2
[Cr(Me2bpy)3](OTf)3 732 1.70 0.74 ± 0.03 8.9 ± 0.9
[Cr(phen)2(4-dmcbpy)](OTf)3 732 1.70 0.28 ± 0.01 3.4 ± 0.4
[Cr(Ph2phen)2(4-dmcbpy)](OTf)3 742 1.68 0.55 ± 0.02 6.6 ± 0.7
[Cr(Me2bpy)2(4-dmcbpy)](OTf)3 734 1.70 0.15 ± 0.01 1.8 ± 0.2
[Cr(4-dmcbpy)3](BF4)3 733 1.70 0.012 ± 0.002 0.14 ± 0.02
Complex
Emission
lifetimea,c
(μs)
ln(A)f
A
(1/s)
Eaf
(kJ∙mol−1
)
[Cr(phen)3](OTf)3 304 ± 4 28.9 ± 0.5 3.5 x 1012
51 ± 1
[Cr(Ph2phen)3](OTf)3 425 ± 20 26 ± 1 2.4 x 1011
45 ± 3
[Cr(bpy)3](OTf)3 69 ± 2 27.0 ± 0.3 5.5 x 1011
43 ± 1
[Cr(Me2bpy)3](OTf)3 196 ± 8 28.5 ± 0.5 2.3 x 1012
49 ± 1
[Cr(phen)2(4-dmcbpy)](OTf)3 87 ± 3 27.8 ± 0.5 1.2 x 1012
46 ± 1
[Cr(Ph2phen)2(4-dmcbpy)](OTf)3 108 ± 8 27.3 ± 0.5 7.5 x 1011
45 ± 1
[Cr(Me2bpy)2(4-dmcbpy)](OTf)3 47 ± 1 28.6 ± 0.2 2.6 x 1012
46 ± 1
[Cr(4-dmcbpy)3](BF4)3 7.7 ± 0.3 26.7 ± 0.4 4.1 x 1012
37 ± 1 a The measurements were done at 23-24 °C.
b See text for details on the calculation of E00.
c The errors represent 2σ (two times of standard deviation) from nine measurements (three
independent experiments for each sample, three measurements for each experiment.) d The normalized emission quantum yields were determined with respect to [Cr(phen)3]
3+.
e Emission quantum yields reported are relative to the standard [Ru(bpy)3]
2+ in acetonitrile with a
0.062 absolute emission quantum yield. f Errors reported reflect 2σ from the fitting of a single temperature-data set to a linear Arrhenius
model ln kobs = ln A − Ea/RT.
46
The emission spectra shown in Figure 2.5 allow us to determine the amount of energy
stored in the long-lived excited states of these systems. This is a critical piece of information in
assessing the oxidation potential available following absorption of UV-visible light. The values
reported in Table 2.4 are for E00 (equivalent to the ΔG stored in the excited state) where we have
modified the observed maximum of the 0-0 vibronic transition (E0) with its width (Δ 0,1/2)
according to the expression
For emissive MLCT species (commonly, RuII, Os
II, and Re
I) one generally uses a Franck-
Condon analysis to determine E0 and Δ 0,1/2.75-77
Here, because emissive features from the 2E →
4A ground state are relatively narrow, we determine these quantities directly from the energy and
width of the most intense band in the emission spectra.
We also report emission quantum yields in Table 2.4 for all complexes relative to
[Cr(phen)3]3+
. The expression for the emission quantum yield φem in terms of radiative (kr) and
non-radiative (knr) rate constants is shown in Equation 2.3. Here kobs and τobs are inversely related
and refer to the observed rate constant and lifetime, respectively.
If a comparison is drawn between [Cr(phen)3]3+
and [(phen)2Cr(4-dmcbpy)]3+
, it is seen that
introduction of a single 4-dmcbpy ligand drops the quantum yield to 28% of its value in the
homoleptic species. For the comparisons between [Cr(Me2bpy)3]3+
and [Cr(Me2bpy)2(4-
dmcbpy)]3+
or between [Cr(Ph2phen)3]3+
and [Cr(Ph2phen)2(4-dmcbpy)]3+
, these values are 21%
and 21.5%, respectively. In the comparison between [Cr(bpy)3]3+
and [Cr(4-dmcbpy)3]3+
,
47
introduction of three 4-dmcbpy ligands in place of the three bpy ligands drops the quantum yield
to ∼6% of the value prior to the substitution. The origin of these quantum yield changes is
discussed below.
2.3.5 Time-Resolved Emission
To understand these trends in radiative quantum yields, we have measured the observed
lifetimes (τobs) of the emissive 2E states in the full set of complexes and these are also reported in
Table 2.4. For the new complexes reported here, the lowest energy excited state lifetimes range
from 108 μs (for [Cr(Ph2phen)2(4-dmcbpy)]3+) to 7.7 μs (for [Cr(4-dmcbpy)3]
3+). This is
encouraging as it suggests there may be ample time in the 2E excited-state of the respective
complexes to engage in hole transfer photochemistry. In the comparison between [Cr(phen)3]3+
and [Cr(phen)2(4-dmcbpy)]3+
, the observed lifetime drops from 304 to 87 μs upon introduction
of the 4-dmcbpy ligand. This latter value is 29% of that measured for the tris-homoleptic species.
Similar comparisons made between [Cr(Me2bpy)3]3+
and [Cr(Me2bpy)2(4-dmcbpy)]3+
or between
[Cr(Ph2phen)3]3+
and [Cr(Ph2phen)2(4-dmcbpy)]3+
yield the values 24% and 25%, respectively.
The comparison between [Cr(bpy)3]3+
and [Cr(4-dmcbpy)3]3+
shows that introduction of a full set
of ester-containing ligands drops the observed lifetime (7.7 μs) to 11% of the value obtained for
[Cr(bpy)3]3+
(69 μs). The various percentage changes in τobs shown here are similar to those seen
above for φem. This correspondence suggests (by Equation 2.3) that introduction of the 4-dmcbpy
ligand mainly affects the rate constants for non-radiative relaxation pathways (∑knr).
To better understand the origin of the lifetime drop for [Cr(4-dmcbpy)3]3+
, we have
explored the temperature dependence of emission lifetimes for the complete series of complexes
48
over the range 283K-353 K. A representative example is shown in Figure 2.6 for [Cr(4-
dmcbpy)3]3+
compared to [Cr(bpy)3]3+
as the natural log of the observed rate constant (kobs)
versus 1000/T. All data sets for the eight complexes are shown in Figure 2.7.
Figure 2.6 Temperature dependence of the observed rate constant kobs = 1/τobs for [Cr(bpy)3]3+
and [Cr(4-dmcbpy)3]3+
in degassed 1 M HCl(aq).
In all cases we observe temperature dependence. The high-degree of linearity for all data
sets suggests that in this temperature range (throughout which we have a fluid solution) the
activated process dominates the observed rate constant relative to temperature-independent
contributions to kr and knr.78
Table 2.4 includes a listing of the measured Arrhenius pre-
exponential (A) and activation energy (Ea) for each system. With this information we can see
general trends emerge, especially if we draw comparisons, as before, between pairs of
compounds where ancillary ligands are the same. For example in comparing [Cr(phen)3]3+
and
[Cr(phen)2(4-dmcbpy)]3+
it is seen that introduction of the ester-containing ligand serves to drop
the activation energy by 5 kJ/mol. Given that the pre-exponential A drops modestly between
49
[Cr(phen)3]3+
and [Cr(phen)2(4-dmcbpy)]3+
, the concomitant decrease in Ea is responsible for the
shortening of the lifetime from 304 to 87 μs. A similar decrease in Ea is also observed between
[Cr(Me2bpy)3]3+
and [Cr(Me2bpy)2(4-dmcbpy)]3+
and the system with shortest lifetime, [Cr(4-
dmcbpy)3]3+
, also has the smallest Ea which is 6 kJ/mol less than that measured for [Cr(bpy)3]3+
.
These observations can be explained if introduction of the electron withdrawing ester ligand
weakens the ligand field via an inductive effect, thereby decreasing the barrier to excited states
having a t2g2eg configuration. Such states would have significant nuclear distortions (primarily
metal-ligand bond distances) and through these displacements larger non-radiative decay rates.
Given the magnitude of the Arrhenius pre-exponential A measured for these systems, it is
unlikely that these complexes are thermally activated directly through back-intersystem-crossing
from the 2E to the
4T manifold. Rather one might need to invoke deactivation through states with
appropriate nuclear distortion but also with doublet electron spin character. It is noted that
differences in activation energy alone are not sufficient to explain the significant lifetime
variation between [Cr(Ph2phen)2(4-dmcbpy)]3+
where τobs =108 μs, and [Cr(Ph2phen)3]3+
where
τobs = 425 μs. Here, differences in the Arrhenius pre-exponential A are the primary origin of the
observation. It does not appear to be simply a consequence of the reduced rigidity of the diester
ligand and related entropic effects as we do not see A increase between [Cr(phen)3]3+
and
[Cr(phen)2(4-dmcbpy)]3+
. Detailed theoretical exploration is needed to determine, for example,
whether there are changes in the density of electronic states at the activation energy of ∼45
kJ/mol in the comparison between [Cr(Ph2phen)3]3+
and [Cr(Ph2phen)2(4-dmcbpy)]3+
that might
influence the relative mechanisms for non-radiative decay.
50
Figure 2.7 Temperature dependence of the observed rate constant (kobs = 1/τobs) for the Cr(III)
polypyridyl complexes in degassed 1 M HCl(aq).
51
Figure 2.8 Comparison of cyclic voltammograms for the 4-dmcbpy containing complexes
[Cr(phen)2(4-dmcbpy)]3+
, [Cr(Ph2phen)2(4-dmcbpy)]3+
, [Cr(Me2bpy)2(4-dmcbpy)]3+
, and [Cr(4-
dmcbpy)3]3+
in 0.1 M TBAPF6 acetonitrile solution. Arrows indicate the starting point and
direction for each voltammogram. For [Cr(Ph2phen)2(4-dmcbpy)]3+
, [Cr(Me2bpy)2(4-dmcbpy)]3+
,
and [Cr(4-dmcbpy)3]3+
, the potential is referenced to ferrocene (from a voltammogram which
includes Fc collected immediately after the displayed voltammogram).
2.3.6 Ground and Excited State Reduction Potentials
Along with emission data, the second critical component for finding excited state redox
potentials is the determination of ground state “CrIII/II” and “Cr
IV/III” couples. Each of the
heteroleptic dipyridyl complexes [Cr(phen)2(4-dmcbpy)]3+
, [Cr(Ph2phen)2(4-dmcbpy)]3+
, and
[Cr(Me2bpy)2(4-dmcbpy)]3+
show four reversible one-electron reductions, and in the case of
[Cr(4-dmcbpy)3]3+
, six reversible waves are seen. The “CrIV/III
” couples for these systems have
52
not been observed as they are outside of the acetonitrile solvent oxidation window. These data
are presented here in Figure 2.8 and Table 2.5.
Table 2.5. Ground and Excited State Reduction Potentials for Cr(III) Polypyridyl Complexes
(E1/2 vs Fc+/Fc, V)
a
Complex 3+/2+ 2+/1+ 1+/0 0/1− 1−/2− 2−/3− *3+/2+b
[Cr(phen)3](OTf)3 −0.65 −1.17 −1.71 −2.21 +1.05
[Cr(Ph2phen)3](OTf)3 −0.67 −1.11 −1.63 −2.05 +1.00
[Cr(bpy)3](OTf)3 −0.63 −1.15 −1.72 −2.34 +1.08
[Cr(Me2bpy)3](OTf)3 −0.79 −1.29 −1.82 +0.91
[Cr(phen)2(4-dmcbpy)](OTf)3 −0.42 −1.01 −1.61 −1.90 +1.28
[Cr(Ph2phen)2(4-dmcbpy)](OTf)3 −0.45 −0.99 −1.54 −1.87 +1.23
[Cr(Me2bpy)2(4-dmcbpy)](OTf)3 −0.47 −1.11 −1.71 −1.92 +1.23
[Cr(4-dmcbpy)3](BF4)3 −0.26 −0.68 −1.21 −1.74 −1.93 −2.10 +1.44 a Conditions for cyclic voltammetry of Cr complexes: electrolyte, 0.1 M TBAPF6 in CH3CN; WE,
Pt; CE, Pt Wire; scan rate, 100 mV/s. b Calculated from Equation 2.4.
As shown, the first reduction for complexes containing the ester ligands is facile, with
E1/2 occurring between −0.47 and −0.42 V versus Fc+/Fc for all species containing a single ester
ligand. We find that the position of the first wave can be tuned through the functional groups
present on the attached ligands. The presence of two strongly electron withdrawing ester groups
on the ligand 4-dmcbpy shifts the initial “CrIII/II” process to more positive potentials compared to
those previously reported for hetero- and homoleptic [Cr(NN)3]3+
complexes.48,56
This is shown
in Table 2.5: for [Cr(phen)2(4-dmcbpy)]3+
, [Cr(Ph2phen)2(4-dmcbpy)]3+
, and [Cr(Me2bpy)2(4-
dmcbpy)]3+, the first “Cr
III/II” reduction occurs at a potential at least 0.22 V more positive than
53
the homoleptic species that lack an ester ligand. For [Cr(4-dmcbpy)3]3+
this first reduction is
remarkably shifted an additional ∼0.2 V in the positive direction compared to the complexes
containing a single 4-dmcbpy. For this species the presence of three ligands carrying electron
withdrawing groups is able to move the “Cr−I/−II
” and “Cr−II/−III
” reductions to within the
acetonitrile window, leading to the observation of six reversible waves. While each of these
reductions is listed as a change in the Cr formal oxidation state (matching the literature
precedent), these processes are more likely due to ligand reduction,68,79-80
and it is reasonable to
surmise that the presence of electron withdrawing groups allow for each 4-dmcbpy ligand to be
reduced by two electrons.
The half wave reduction potential of the excited state can be estimated using the ground
state half wave reduction potential and the one electron potential corresponding to the
spectroscopic excited state:81
The use of E00 as a potential (rather than energy) is acceptable in this context because the excited
state is only acting as a one-electron acceptor. As discussed in the context of Table 2.4, E00 is
largely invariant to ligand substitution patterns. Others have noted that the 2E emission maximum
in Cr(III) species is solvent insensitive, and we have observed that E00 does not change between
aqueous (where our spectroscopic measurements have been made) and acetonitrile (where our
electrochemical measurements have been made) environments. On the other hand, the ground
state (first) reduction potentials show ligand dependence (Table 2.5). Thus, the excited state
reduction potential is tunable. These values are compiled in Table 2.5 for all the complexes
studied. Equation 2.4 allows us to predict a CrIII*/II
couple of +1.22 to +1.27 V versus Fc+/Fc for
54
the heteroleptic complexes. To compare the excited state potentials of these Cr(III) complexes to
other photoelectrochemically active species reported in the literature, the measured redox
potentials have been converted to reference NHE (Fc+/Fc is 0.40 V vs SCE in 0.1 M TBAPF6
82
and SCE is 0.241 V vs NHE83
). We note that this comparison is approximate because of
differences in solvent and supporting electrolyte, but it is still useful.84
In the case of [Cr(4-
dmcbpy)3]3+
, the CrIII*/II
couple is calculated to be a remarkable +2.08 V versus NHE while
[Cr(phen)2(4-dmcbpy)]3+
, [Cr(Ph2phen)2(4-dmcbpy)]3+
, and [Cr(Me2bpy)2(4-dmcbpy)]3+
are
+1.92 V, +1.87 V, and +1.87 V versus NHE, respectively. As a point of reference, a similar ester
functionalized [Ru(NN)3]2+
complex shows an excited state reduction potential of +1.26 V (vs
NHE) for RuII*/I
.85
Reece and Nocera reported +1.78 V (vs NHE) for a ReI*/0
complex,86
while
Sullivan and co-workers reported +2.71 V (vs NHE) for a ReII*/I
complex.87
Reports of CrIII*/II
potentials from non-polypyridyl complexes are very rare in the literature; however, an amine
based CrIII*/II
system has been reported with an excited state reduction potential of +0.77 V (vs
NHE).88
Indeed, the ester-containing Cr(III) complexes [Cr(phen)2(4-dmcbpy)]3+
,
[Cr(Ph2phen)2(4-dmcbpy)]3+
, [Cr(Me2bpy)2(4-dmcbpy)]3+
, and [Cr(4-dmcbpy)3]3+
offer
themselves as potentially powerful excited state oxidants.
2.3.7 Transient Absorption
In future studies involving hole transfer photochemistry between Cr(III) systems and
wide band gap semiconductors to which they are bound, it will be important to be able to
interrogate the time-dependent behavior of the lowest energy excited-state manifold in these
complexes without relying on emission and on potentially very short time scales. A valuable tool
55
in this context is transient electronic absorption (TA) spectroscopy. With time resolution from
tens of femtoseconds to milliseconds, TA spectroscopy has been central to unraveling the
mechanism in dye-sensitized heterojunction solar cell materials.6-7,89
Here we do not consider the
earliest transient events but rather explore absorption features once these molecules are
electronically relaxed in the 2E excited states. The spectrometer used here employs ∼10 ns
excitation pulses centered at 355 nm and the transient features probed with white light decay
with single exponential behavior on microsecond time scales. Within such time scales, it is well
established that the 2E excited-state of these complexes in fluid solution at room temperature is
vibrationally cool and thermally equilibrated with the solvent.73-74,90-96
TA spectra are shown in
Figure 2.9. The TA spectra for the tris-homoleptic compounds [Cr(phen)3]3+
, [Cr(bpy)3]3+
,
[Cr(Me2bpy)3]3+
, and [Cr(Ph2phen)3]3+
agree with previously reported spectra collected under
similar experimental conditions.38
In the case of bis-heteroleptic complex [Cr(Me2bpy)2(4-
dmcbpy)]3+
, the spectrum is quite similar to that of the related tris-homoleptic species
[Cr(Me2bpy)3]3+
. There are some qualitative differences between [Cr(phen)2(4-dmcbpy)]3+
and
[Cr(phen)3]3+
in terms of relative band intensities, but these are unremarkable and are not
considered here further.
As discussed above, complexes [Cr(4-dmcbpy)3]3+
and [Cr(Ph2phen)2(4-dmcbpy)]3+
have
the most promising molar absorptivities of visible wavelengths and are likely, therefore, to be
used in future semiconductor sensitization experiments. Both difference spectra show a resolved
bleach feature centered at 360 nm, a strong absorption feature at ∼400 nm, and a broad
absorption further to the red. In the case of [Cr(Ph2phen)2(4-dmcbpy)]3+
, this redder absorbance
is very broad and appears to peak at ∼550 nm. The magnitude of the bleach in both transient
spectra provides a useful metric for estimating excited-state molar absorptivities at various
56
wavelengths. In the ground state optical spectra shown in Figure 2.3 and Figure 2.4, complexes
[Cr(4-dmcbpy)3]3+
and [Cr(Ph2phen)2(4-dmcbpy)]3+
exhibit molar extinction coefficients at 360
nm of ε360 nm =10500 M−1
cm−1
and ε360 nm = 23400 M−1
cm−1
, respectively. In the limiting case
that the −ΔA at 360 nm for these two spectra is entirely due to loss of ground state absorption,
then it is possible to assign a lower limit to excited-state molar absorptivities (ε*(λ)) for observed
absorption features. In the case of [Cr(4-dmcbpy)3]3+
, we measure ΔA = −0.014 at 360 nm and
ΔA = 0.020 at 400 nm such that ε*400 nm ≥ 14400 M−1
cm−1
. A similar observation is made for
[Cr(Ph2phen)2(4-dmcbpy)]3+
although the absorptivities are even larger. In this case ε*400 nm ≥
23300 M−1
cm−1
and ε*550 nm ≥ 16500 M−1
cm−1
. In these spectra (and very likely in the whole
series of Cr(III) polypyridyl complexes we have considered), the transient absorption features
with appreciable ΔA are clearly not assignable to ligand-field absorption occurring from the t2g3
configuration of the 2E excited state. The strength of the absorption features suggest these are
charge transfer in nature originating from the 2E and
2T excited states, although it is not possible
at this time to assign MLCT and/or LMCT to any particular feature. The t2g3 configuration of the
2E excited state is not expected to significantly perturb the ligand π-system, so ligand-centered π
→ π* transitions are unlikely to play a prominent role. It may be possible that the observed
transitions, especially those near 400 nm, borrow some intensity from such excitations as Juban
and McCusker73
have argued occurs in the ground state of [Cr(bpy)3]3+
. The important point to
stress here is that excited-state absorption features in the visible spectrum for these complexes
have significant oscillator strength. In more complex environments such as when complexes are
bound to heterogeneous semiconductor surfaces, such features may be a useful and easily
observable diagnostic of 2E lifetime and related hole injection rate constants.
57
Figure 2.9 Transient absorption spectra on a μs timescale for Cr(III) polypyridyl complexes in 1
M HCl(aq). The spectra were determined from a single exponential fit to transient absorption (or
bleach) kinetics collected at each of the wavelengths for which there is a dot. The lines are
included as guides to the eye.
58
2.4 Conclusions
The preparation of heteroleptic dipyridyl Cr(III) complexes that contain at least one
carboxylate group (for eventual attachment to semiconductor surfaces) was not known prior to
our efforts. We have synthesized four such complexes, including three heteroleptic [Cr(NN)2(4-
dmcbpy)]3+
species. We have also explored the basic photophysical and electrochemical
properties of these systems to better understand the nascent and long-lived excited-states that
will be called upon to drive hole-injection following photoexcitation in later studies.
The electronic absorption studies show a combination of ligand-centered, metal-centered
ligand field, and charge transfer transitions. Nevertheless, it is noted here that for a Cr(III)
species with Ph2phen ancillary ligands, [Cr(Ph2phen)2(4-dmcbpy)]3+
, shows appreciable
sensitization of visible light, with a molar absorptivity of 1270 M−1
cm−1
at 491 nm. Time-
resolved emission studies show that the introduction of the 4-dmcbpy ligand decreases the
doublet excited state lifetimes of the complexes with respect to their tris-homoleptic analogues.
Notwithstanding, excited state lifetimes are sufficiently long that we can anticipate photoinduced
hole transfer to suitable semiconductor substrates. Additionally, preliminary electrochemical
studies show that the introduction of the 4-dmcbpy significantly shifts reduction potentials of the
complexes to more positive values. The combination of cyclic voltammetry and static emission
studies indicates that strongly oxidizing excited states are possessed by this class of molecules.
The strongly oxidizing excited states found for the Cr(III) dye complexes coupled with
the reported long excited state lifetimes lead us to believe these species will be capable of hole
injection into p-type semiconductors with suitably aligned valence bands. Further, the excited-
state absorption features for these heteroleptic complexes should serve as handles for studying
59
the hybrid dye-sensitized materials. Our future work is to incorporate these dye complexes with
semiconductors such as NiO into hybrid materials.
2.5 Acknowledgment on Participation
The synthesis and characterization of all the Cr(III) complexes were done by Professor
Matthew P. Shores and Dr. Ashley M. McDaniel at the Department of Chemistry, Colorado State
University. I was responsible for the photophysical measurements.
2.6 References
(1) O'Regan, B.; Gratzel, M. Nature 1991, 353, 737-740.
(2) Nazeeruddin, M. K. G., M. Photofunctional Transition Metal Complexes; Springer:
Berlin, 2007.
(3) Robertson, N. Angew. Chem. Int. Ed. 2006, 45, 2338-2345.
(4) Polo, A. S.; Itokazu, M. K.; Iha, N. Y. M. Coord. Chem. Rev. 2004, 248, 1343-1361.
(5) Hagfeldt, A.; Gratzel, M. Acc. Chem. Res. 2000, 33, 269-277.
(6) Ardo, S.; Meyer, G. J. Chem. Soc. Rev. 2009, 38, 115-164.
(7) Martinson, A. B. F.; Hamann, T. W.; Pellin, M. J.; Hupp, J. T. Chem. - Eur. J. 2008, 14,
4458-4467.
(8) Gersten, S. W.; Samuels, G. J.; Meyer, T. J. J. Am. Chem. Soc. 1982, 104, 4029-4030.
(9) Lei, Y. B.; Hurst, J. K. Inorg. Chim. Acta 1994, 226, 179-185.
(10) Limburg, J.; Vrettos, J. S.; Liable-Sands, L. M.; Rheingold, A. L.; Crabtree, R. H.;
Brudvig, G. W. Science 1999, 283, 1524-1527.
(11) Wada, T.; Tsuge, K.; Tanaka, K. Angew. Chem. Int. Ed. 2000, 39, 1479-+.
(12) Sens, C.; Romero, I.; Rodriguez, M.; Llobet, A.; Parella, T.; Benet-Buchholz, J. J. Am.
Chem. Soc. 2004, 126, 7798-7799.
(13) Zong, R.; Thummel, R. P. J. Am. Chem. Soc. 2005, 127, 12802-12803.
(14) Tseng, H. W.; Zong, R.; Muckerman, J. T.; Thummel, R. Inorg. Chem. 2008, 47, 11763-
11773.
(15) Kanan, M. W.; Nocera, D. G. Science 2008, 321, 1072-1075.
60
(16) McDaniel, N. D.; Coughlin, F. J.; Tinker, L. L.; Bernhard, S. J. Am. Chem. Soc. 2008,
130, 210-217.
(17) Eisenberg, R.; Gray, H. B. Inorg. Chem. 2008, 47, 1697-1699.
(18) Morandeira, A.; Boschloo, G.; Hagfeldt, A.; Hammarstrom, L. J. Phys. Chem. C 2008,
112, 9530-9537.
(19) Qin, P.; Wiberg, J.; Gibson, E. A.; Linder, M.; Li, L.; Brinck, T.; Hagfeldt, A.; Albinsson,
B.; Sun, L. C. J. Phys. Chem. C 2010, 114, 4738-4748.
(20) Qin, P.; Linder, M.; Brinck, T.; Boschloo, G.; Hagfeldt, A.; Sun, L. C. Adv. Mater. 2009,
21, 2993-2996.
(21) Qin, P.; Zhu, H. J.; Edvinsson, T.; Boschloo, G.; Hagfeldt, A.; Sun, L. C. J. Am. Chem.
Soc. 2008, 130, 8570-8571.
(22) Nattestad, A.; Mozer, A. J.; Fischer, M. K. R.; Cheng, Y. B.; Mishra, A.; Bauerle, P.;
Bach, U. Nature Mater. 2010, 9, 31-35.
(23) He, J. J.; Lindstrom, H.; Hagfeldt, A.; Lindquist, S. E. Sol. Energ. Mat. Sol. Cells 2000,
62, 265-273.
(24) Mizoguchi, Y.; Fujihara, S. Electrochem. Solid-State Lett. 2008, 11, K78-K80.
(25) Green, M. Third Generation Photovoltaics: Advanced Solar Energy Conversion;
Springer-Verlag: Berlin, 2003.
(26) Hanna, M. C.; Nozik, A. J. J. Appl. Phys. 2006, 100.
(27) Nattestad, A.; Ferguson, M.; Kerr, R.; Cheng, Y. B.; Bach, U. Nanotechnology 2008, 19,
295304.
(28) Mori, S.; Fukuda, S.; Sumikura, S.; Takeda, Y.; Tamaki, Y.; Suzuki, E.; Abe, T. J. Phys.
Chem. C 2008, 112, 16134-16139.
(29) Borgstrom, M.; Blart, E.; Boschloo, G.; Mukhtar, E.; Hagfeldt, A.; Hammarstrom, L.;
Odobel, F. J. Phys. Chem. B 2005, 109, 22928-22934.
(30) Odobel, F.; Le Pleux, L.; Pellegrin, Y.; Blart, E. Acc. Chem. Res. 2010, 43, 1063-1071.
(31) Morandeira, A.; Fortage, J.; Edvinsson, T.; Le Pleux, L.; Blart, E.; Boschloo, G.;
Hagfeldt, A.; Hanmiarstrom, L.; Dobel, F. J. Phys. Chem. C 2008, 112, 1721-1728.
(32) Morandeira, A.; Boschloo, G.; Hagfeldt, A.; Hammarstrom, L. J. Phys. Chem. B 2005,
109, 19403-19410.
(33) ORegan, B.; Schwartz, D. T. J Appl Phys 1996, 80, 4749-4754.
(34) O'Regan, B.; Lenzmann, F.; Muis, R.; Wienke, J. Chem. Mater. 2002, 14, 5023-5029.
(35) Rusop, M.; Soga, T.; Jimbo, T.; Umeno, M. Surf. Rev. Lett. 2004, 11, 577-583.
(36) Nakasa, A.; Usami, H.; Sumikura, S.; Hasegawa, S.; Koyama, T.; Suzuki, E. Chem. Lett.
2005, 34, 500-501.
(37) Gibson, E. A.; Smeigh, A. L.; Le Pleux, L.; Fortage, J.; Boschloo, G.; Blart, E.; Pellegrin,
Y.; Odobel, F.; Hagfeldt, A.; Hammarstrom, L. Angew. Chem. Int. Ed. 2009, 48, 4402-
4405.
(38) Serpone, N.; Jamieson, M. A.; Henry, M. S.; Hoffman, M. Z.; Bolletta, F.; Maestri, M. J.
Am. Chem. Soc. 1979, 101, 2907-2916.
61
(39) Serpone, N.; Jamieson, M. A.; Emmi, S. S.; Fuochi, P. G.; Mulazzani, Q. G.; Hoffman, M.
Z. J. Am. Chem. Soc. 1981, 103, 1091-1098.
(40) Serpone, N.; Jamieson, M. A.; Sriram, R.; Hoffman, M. Z. Inorg. Chem. 1981, 20, 3983-
3988.
(41) Hoffman, M. Z.; Serpone, N. Isr. J. Chem. 1982, 22, 91-97.
(42) Bolletta, F.; Maestri, M.; Moggi, L.; Jamieson, M. A.; Serpone, N.; Henry, M. S.;
Hoffman, M. Z. Inorg. Chem. 1983, 22, 2502-2509.
(43) Serpone, N.; Hoffman, M. Z. J. Chem. Educ. 1983, 60, 853-860.
(44) Jamieson, M. A.; Serpone, N.; Hoffman, M. Z. Coord. Chem. Rev. 1981, 39, 121-179.
(45) Juris, A.; Balzani, V.; Barigelletti, F.; Campagna, S.; Belser, P.; Vonzelewsky, A. Coord.
Chem. Rev. 1988, 84, 85-277.
(46) Greenwood, N. N.; Earnshaw, A. Chemistry of the Elements, 2nd ed.; Elsevier
Butterworth-Heinemann: Oxford, 1997.
(47) Barker, K. D.; Barnett, K. A.; Connell, S. M.; Glaeser, J. W.; Wallace, A. J.; Wildsmith,
J.; Herbert, B. J.; Wheeler, J. F.; Kane-Maguire, N. A. P. Inorg. Chim. Acta 2001, 316,
41-49.
(48) Isaacs, M.; Sykes, A. G.; Ronco, S. Inorg. Chim. Acta 2006, 359, 3847-3854.
(49) Donnay, E. G.; Schaeper, J. P.; Brooksbank, R. D.; Fox, J. L.; Potts, R. G.; Davidson, R.
M.; Wheeler, J. F.; Kane-Maguire, N. A. P. Inorg. Chim. Acta 2007, 360, 3272-3280.
(50) Garelli, N.; Vierling, P. J. Org. Chem. 1992, 57, 3046-3051.
(51) Ryu, C. K.; Endicott, J. F. Inorg. Chem. 1988, 27, 2203-2214.
(52) Henriques, R. T.; Herdtweck, E.; Kuhn, F. E.; Lopes, A. D.; Mink, J.; Romao, C. C. J.
Chem. Soc. Dalton. 1998, 1293-1297.
(53) Bruker Analytical X-Ray Systems Inc. APEX 2; Madison, WI, 2008.
(54) Sheldrick, G. M. SHELXTL, Version 6.14; Bruker Analytical X-Ray Systems Inc.:
Madison, WI, 1999.
(55) Kalyanasundaram, K. Photochemistry of Polypyridine andPorphyrin Complexes;
Academic Press: London, 1992.
(56) Brunschwig, B.; Sutin, N. J. Am. Chem. Soc. 1978, 100, 7568-7577.
(57) Demas, J. N.; Crosby, G. A. J. Phys. Chem. 1971, 75, 991-1024.
(58) Calvert, J. M.; Caspar, J. V.; Binstead, R. A.; Westmoreland, T. D.; Meyer, T. J. J. Am.
Chem. Soc. 1982, 104, 6620-6627.
(59) Lilie, J.; Waltz, W. L. Inorg. Chem. 1983, 22, 1473-1478.
(60) Jamieson, M. A.; Serpone, N.; Henry, M. S.; Hoffman, M. Z. Inorg. Chem. 1979, 18,
214-216.
(61) Kane-Maguire, N. A. P.; Conway, J.; Langford, C. H. J. Chem. Soc., Chem. Commun.
1974, 801-802.
(62) Maestri, M.; Bolletta, F.; Serpone, N.; Moggi, L.; Balzani, V. Inorg. Chem. 1976, 15,
2048-2051.
(63) Jamieson, M. A.; Serpone, N.; Hoffman, M. Z. J. Am. Chem. Soc. 1983, 105, 2933-2937.
62
(64) Damrauer, N. H.; Boussie, T. R.; Devenney, M.; McCusker, J. K. J. Am. Chem. Soc. 1997,
119, 8253-8268.
(65) Bain, G. A.; Berry, J. F. J. Chem. Educ. 2008, 85, 532-536.
(66) Vandiver, M. S.; Bridges, E. P.; Koon, R. L.; Kinnaird, A. N.; Glaeser, J. W.; Campbell, J.
F.; Priedemann, C. J.; Rosenblatt, W. T.; Herbert, B. J.; Wheeler, S. K.; et al. Inorg.
Chem. 2010, 49, 839-848.
(67) Milder, S. J.; Gold, J. S.; Kliger, D. S. Inorg. Chem. 1990, 29, 2506-2511.
(68) Konig, E.; Herzog, S. J. Inorg. Nucl. Chem. 1970, 32, 585-599.
(69) Hauser, A.; Mader, M.; Robinson, W. T.; Murugesan, R.; Ferguson, J. Inorg. Chem. 1987,
26, 1331-1338.
(70) Schonherr, T.; Atanasov, M.; Hauser, A. Inorg. Chem. 1996, 35, 2077-2084.
(71) Ohno, T.; Kato, S.; Kaizaki, S.; Hanazaki, I. Chem. Phys. Lett. 1983, 102, 471-474.
(72) Ohno, T.; Kato, S.; Kaizaki, S.; Hanazaki, I. Inorg. Chem. 1986, 25, 3853-3858.
(73) Juban, E. A. Ph.D. Thesis; University of California - Berkeley, 2006.
(74) Kirk, A. D. Chem. Rev. 1999, 99, 1607-1640.
(75) Claude, J. P.; Meyer, T. J. J. Phys. Chem. 1995, 99, 51-54.
(76) Kober, E. M.; Caspar, J. V.; Lumpkin, R. S.; Meyer, T. J. J. Phys. Chem. 1986, 90, 3722-
3734.
(77) Chen, P. Y.; Meyer, T. J. Chem. Rev. 1998, 98, 1439-1477.
(78) Forster, L. S. Coord. Chem. Rev. 2002, 227, 59-92.
(79) Saito, Y.; Takemoto, J.; Hutchins.B; Nakamoto, K. Inorg. Chem. 1972, 11, 2003-2011.
(80) Creutz, C. Comments Inorg. Chem. 1982, 1, 293–311.
(81) Ballardini, R.; Varani, G.; Indelli, M. T.; Scandola, F.; Balzani, V. J. Am. Chem. Soc.
1978, 100, 7219-7223.
(82) Connelly, N. G.; Geiger, W. E. Chem. Rev. 1996, 96, 877-910.
(83) Bard, A. J.; Faulkner, L. R. Electrochemical methods fundamentals and applications, 2nd
ed.; Wiley: New York, 2001.
(84) Pavlishchuk, V. V.; Addison, A. W. Inorg. Chim. Acta 2000, 298, 97-102.
(85) Monserrat, K.; Foreman, T. K.; Gratzel, M.; Whitten, D. G. J. Am. Chem. Soc. 1981, 103,
6667-6672.
(86) Reece, S. Y.; Nocera, D. G. J. Am. Chem. Soc. 2005, 127, 9448-9458.
(87) Del Negro, A. S.; Seliskar, C. J.; Heineman, W. R.; Hightower, S. E.; Bryan, S. A.;
Sullivan, B. P. J. Am. Chem. Soc. 2006, 128, 16494-16495.
(88) Ramasami, T.; Endicott, J. F.; Brubaker, G. R. J. Phys. Chem. 1983, 87, 5057–5059.
(89) Gratzel, M. Inorg. Chem. 2005, 44, 6841-6851.
(90) Serpone, N.; Jamieson, M. A.; Hoffman, M. Z. J. Chem. Soc. Chem. Comm. 1980, 1006-
1007.
(91) Serpone, N.; Jamieson, M. A.; Sharma, D. K.; Danesh, R.; Bolletta, F.; Hoffman, M. Z.
Chem. Phys. Lett. 1984, 104, 87-92.
(92) Kirk, A. D.; Porter, G. B.; Sharma, D. K. Chem. Phys. Lett. 1986, 123, 548-550.
63
(93) Serpone, N.; Hoffman, M. Z. Chem. Phys. Lett. 1986, 123, 551-552.
(94) Rojas, G. E.; Dupuy, C.; Sexton, D. A.; Magde, D. J. Phys. Chem. 1986, 90, 87–92.
(95) Rojas, G. E.; Magde, D. J. Phys. Chem. 1987, 91, 689–691.
(96) Serpone, N.; Hoffman, M. Z. J. Phys. Chem. 1987, 91, 1737–1743.
64
CHAPTER 3
Photophysical Investigations of Cr(III) Hexadentate Iminopyridine
Complexes and Their Tris-Bidentate Analogues
3.1 Introduction
The photophysics and photochemistry of Cr(III) complexes are relevant to third
generation solar photoconversion schemes.1 Chromium is an earth-abundant metal, and
surrounding Cr(III) ions with appropriate ligands can produce powerful photooxidants.2-3
The
excited state properties of Cr(III) tris-diimine (e.g., bpy, phen) complexes are well-known, and
the long (μs) lifetimes observed make these compounds potentially useful for dye-sensitized
hole-injection photovoltaic devices and/or photooxidative catalytic schemes.3-7
In principle, the
photophysical properties are tunable: in a previous report on heteroleptic tris-dipyridyl Cr(III)
complexes, we observed subtle changes in the ground state electronic absorption and more
pronounced changes in electrochemical and excited state photophysical properties when the
ligand set was altered.3 Drawbacks to using Cr(III) tris-diimine complexes in photo-conversion
schemes are as follows: relatively weak absorption of visible light, poor stability in nonacidic
aqueous solution, and increased lability of the formally reduced species in a photooxidative
scenario following hole transfer.
Enhanced solution stability may be addressed by increasing the denticity of the ligand(s)
chelating the chromium center. Photochemical studies on a few chromium complexes ligated by
multidentate amine ligands8-12
or tethered bipyridines13-14
have uncovered long-lived 2E excited
states, although it should be noted that all of these species are yellow and do not absorb strongly
65
in the visible spectrum. However, a structural study for [Cr(pod)](ClO4)3 (see Scheme 3.1 for the
podand ligand structure) reports a “wine red” color for crystals of the podand ligand-containing
Cr(III) complex salt,15
which contrasts with the typical bright yellow coloration of the hetero-
and homoleptic tris-bidentate polypyridyl complexes of Cr(III).3 Thus, complexes featuring
podand ligands could have enhanced absorption in the visible wavelengths coupled
simultaneously with increased stability against ligand exchange in photoreduced species.
Although there are few chromium complexes with hexadentate imine ligands,15-19
the podand
system shown in Scheme 3.1 offers a ligand set similar to the tris-diimine complexes studied
previously, as well as the potential for kinetic stability and electronic and steric tunability. Since
bidentate iminopyridine analogues are easily prepared via Schiff base condensations, we might
also straightforwardly evaluate how photophysical and electrochemical properties change
between tethered (hexadentate) podand ligands and those without a tether (tris-bidentate).
Scheme 3.1 Podand Tripodal and Bidentate Iminopyridine Ligands Used in This Study
66
In this chapter, the preparations, characterizations, photophysical properties, and
computational analyses of a family of Cr(III) iminopyridine complexes are described. A
comparison of podand complexes with their tris-bidentate analogues reveals very similar ground
state behaviors (e.g., structures, electrochemistry, solution stability), but unexpectedly divergent
photophysical properties.
67
3.2 Experimental Section
3.2.1 Preparation of Compounds
Unless otherwise noted, compound manipulations were performedeither inside a
dinitrogen-filled glovebox (MBRAUN Labmaster 130) or via Schlenk techniques on an inert gas
(N2) manifold. The preparations of dimethylpyridine-2,5-dicarboxylate,20
methyl 6-
(hydroxymethyl)nicotinate,21
the hexadentate ligand trimethyl 6,6’,6”-((1E,1’E,1”E)-
((nitrilotris(ethane-2,1-diyl))tris(azanylylidene))tris(methanylylidene))-trinicotinate (pod-ester),22
[Cr(CH3CN)4(BF4)2],23
and thianthrene tetrafluoroborate (Th+BF4
−)24
have been described
elsewhere. Methyl-6-formylnicotinate was synthesized according to a modified literature
procedure,25
where methyl 6-(hydroxymethyl)nicotinate was substituted as the oxidation
substrate. The pod ligand was synthesized from a reported procedure.26
Pentane was distilled
over sodium metal and subjected to three freeze-pump-thaw cycles. Other solvents were sparged
with dinitrogen, passed over molecular sieves, and degassed prior to use. All other reagents were
obtained from commercial sources and were used without further purification.
[Cr(pod)](BF4)3. A structure of the perchlorate salt of the [Cr(pod)]3+
species has been
previously reported;15
however, the synthetic route is not amenable to preparing bulk amounts of
material. A solution of [Cr(CH3CN)4(BF4)2] (0.30 g, 0.78 mmol) in 3 mL of acetonitrile was
added to a solution of the pod ligand (0.32 g, 0.78 mmol) in 5 mL of acetonitrile to form a dark
brown solution. Addition of a solution of Th+BF4
− (0.25 g, 0.83 mmol) in 4 mL of acetonitrile
caused the solution color to lighten to red-orange. A red-orange solid was precipitated by
addition of diethyl ether (30 mL). The solid was isolated by filtration, washed with
dichloromethane (3 x 5 mL) and diethyl ether (3 x 5 mL), and then dried in vacuo to afford 0.50
68
g (88%) of product. IR (KBr pellet): νC=N 1638 cm−1
. Absorption spectrum (CH3CN): λmax (εM)
204 (45900), 292 (12200), 321 nm (8800 M−1
cm−1). μeff (295 K): 3.70 μB. ES+MS (CH3CN): m/z
503.17 ([Cr(pod)F2]+). ES−MS (CH3CN): m/z 813.00 m/z ([Cr(pod)(BF4)4]
−). Anal. Calcd for
C24H27B3CrF12N7: C, 39.71; H, 3.75; N, 13.51. Calcd for C24H27B3CrF12N7·2H2O: C, 37.83; H,
4.10; N, 12.87. Found: C, 38.08; H, 4.37; N, 12.69. We note that [Cr(pod)](BF4)3 is hygroscopic,
and that broad peaks at 3610 and 3266 cm−1
appear in the IR spectrum if the sample is not kept
under dry conditions, indicating uptake of water.
[Cr(pod-ester)](BF4)3. A solution of [Cr(CH3CN)4(BF4)2] (0.13 g, 0.34 mmol) in 3 mL
of acetonitrile was added to a suspension of the pod-ester ligand (0.20 g, 0.34 mmol) in 5 mL of
acetonitrile to form a dark brown solution. Upon addition of a solution of Th+BF4
− (0.13 g, 0.43
mmol) in 4 mL of acetonitrile the solution lightened to a tan-brown color. A tan-brown solid was
precipitated by addition of diethyl ether (30 mL) and was isolated by filtration. The isolated
powder was washed with dichloromethane (3 x 5 mL), and diethyl ether (3 x 5 mL), and then
recrystallized twice by diethyl ether diffusion into a concentrated solution of acetonitrile to yield
0.15 g (49%) of crystalline product. IR (KBr pellet): νC=O 1733, νC=N 1638 and 1610 cm−1
.
Absorption spectrum (CH3CN): λmax(εM) 201 (63000), 246 (sh 19000), 303 (16000), 390 nm
(2100 M−1
cm−1
). μeff (295 K): 4.57 μB. ES+MS (CH3CN): m/z 677.13 ([Cr(pod-ester)F2]+).
ES−MS (CH3CN): m/z 987.07 m/z ([Cr(pod-ester)(BF4)4]−). Anal. Calcd for C32H36B3CrF12N8O6
(([Cr(pod-ester)(BF4)3]·CH3CN): C, 40.84; H, 3.86; N, 11.91. Found: C, 40.58; H, 4.04; N, 11.65.
Crystals suitable for X-ray analysis were obtained by slow diffusion of diethyl ether into an
acetonitrile solution of the compound.
Preparation of impy: (E)-N-(pyridin-2-ylmethylene)ethanamine. A solution of 70%
ethylamine in water (0.80 g) was added to a solution of 2-pyridinecarboxaldehyde (1.08 g, 10.1
69
mmol) in 20 mL of methanol containing 4 Å molecular sieves. The resulting mixture was stirred
at room temperature for 2 h and then filtered to remove the molecular sieves. The solvent was
removed from thefiltrate in vacuo to afford tan colored oil. The oil was extracted into 20 mL of
pentane, and the solvent was removed in vacuo to afford 1.03 g (76%) of product as pale yellow
oil. 1H NMR (CDCl3): 1.31 (3H, t), 3.70 (2H, quar), 7.29 (1H, dd), 7.72 (1H, td). 7.96 (1H, d),
8.38 (1H, s), 8.63 ppm (1H, d). 13
C NMR (CDCl3): 16.24, 55.86, 121.39, 124.79, 136.72, 149.63,
154.89, 161.54 ppm. IR (KBr pellet): νC=N 1649 cm−1
. Absorption spectrum (pentane): λmax 196,
234, 264 (sh), 271, 280 (sh) nm. HRES+MS (CH3OH): m/z calcd 135.0922; found 135.0915
(impy + H)+.
[Cr(impy)3](BF4)3. A solution of [Cr(CH3CN)4(BF4)2] (0.062 g, 0.159 mmol) in 3 mL of
acetonitrile was added to a solution of the impy ligand (0.066 g, 0.493 mmol) in 4 mL of
acetonitrile to form a dark brown solution. Upon addition of a solution of Th+BF4
− (0.051 g,
0.167 mmol) in 4 mL of acetonitrile, the dark solution lightened to a yellow color. The solvent
volume was reduced to 1 mL in vacuo to precipitate thianthrene as a white solid, which was
removed by filtration. The filtrate was treated with diethyl ether (20 mL) to precipitate a yellow
solid. The solid was isolated by filtration, washed with dichloromethane (3 x 3 mL) and diethyl
ether (3 x 3 mL), and then dried in vacuo to afford 0.104 g (91%) of product. IR (KBr pellet):
νC=N 1635 and 1601 cm−1
. Absorption spectrum (CH3CN): λmax(εM) 208 (51000), 224 (sh 37000)
246 (sh 12400), 298 (13400), 315 nm (sh 11400 M−1
cm−1
). μeff (295 K): 3.86 μB. ES+MS
(CH3CN): m/z 151.47 ([Cr(impy)3]3+
), 236.40 ([Cr(impy)3F]2+
), 358.13 ([Cr(impy)2F2]+).
ES−MS (CH3CN): m/z 802.20 ([Cr(impy)3(BF4)4]−). Anal. Calcd for C24H30B3CrF12N6: C, 40.32;
H, 4.23; N, 11.75. Calcd for C24H30B3CrF12N6·2.5H2O: C, 37.93; H, 4.64; N, 11.06. Found: C,
37.75; H, 4.25; N, 10.99.
70
Preparation of impy-ester: (E)-methyl 6-((ethylimino)methyl)nicotinate. A solution of
70% ethylamine in water (0.23 g) was added to a solution of methyl-6-formylnicotinate (0.45 g,
2.72 mmol) in 15 mL of methanol containing 4 Å molecular sieves. The resulting mixture was
stirred at room temperature for 2 h, and then filtered to remove the molecular sieves. The solvent
was removed from the filtrate in vacuo. The resulting orange residue was extracted into 50 mL of
pentane and filtered. The solvent was removed from the filtrate in vacuo, resulting in an ivory
powder. This was sublimed at reduced pressure and 30°C to yield 0.44 g (84%) of product as
colorless crystals. 1H NMR (CDCl3): 1.34 (3H, t), 3.75 (2H, quar), 3.97 (3H, s) 8.07 (1H, d),
8.33 (1H, dd). 8.44 (1H, s), 9.23 ppm (1H, d). 13
C NMR (CDCl3): 16.11, 52.67, 56.00, 120.84,
126.64, 137.78, 150.82, 158.05, 160.78, 165.68 ppm. IR (KBr pellet): νC=O 1725, νC=N 1649 cm−1
.
Absorption spectrum (pentane): λmax 199, 245, 254 (sh), 278 nm. HRES+MS(CH3OH): m/z calcd
193.0977; found 193.0967 (impy-ester + H)+.
[Cr(impy-ester)3](BF4)3. A solution of [Cr(CH3CN)4(BF4)2] (0.049 g,0.125 mmol) in 3
mL of acetonitrile was added to a solution of the impy-ester ligand (0.075 g, 0.388 mmol) in 4
mL of acetonitrile to form a dark brown-red solution. Upon addition of a solution of Th+BF4
−
(0.04 g, 0.13 mmol) in 4 mL of acetonitrile, the dark solution lightened to a yellow color. The
solvent volume was reduced to 1 mL in vacuo to precipitate thianthrene as a white solid, which
was removed by filtration. The filtrate was treated with diethyl ether (20 mL) to precipitate a
yellow solid. The powder was isolated by filtration, washed with dichloromethane (3 x 3 mL)
and diethyl ether (3 x 3 mL), and then dried in vacuo to afford 0.090 g (80%) of product. IR
(KBr pellet): νC=O 1736, νC=N 1633 and 1607 cm−1
. Absorption spectrum (CH3CN): λmax (εM) 205
(57000), 222 (sh 45000), 299 (17600), 321 nm (sh 14600 M−1
cm−1). μeff (295 K): 4.11 μB.
ES+MS (CH3CN): m/z 209.47 ([Cr(impy-ester)3]3+
), 474.13 ([Cr(impy-ester)2F2]+). ES−MS
71
(CH3CN): m/z 976.13 ([Cr(impy-ester)3(BF4)4]−). Anal. Calcd for C30H36B3CrF12N6O6: C, 40.53;
H, 4.08; N, 9.45. Found: C, 40.45; H, 3.89; N, 9.39. Crystals suitable for X-ray analysis were
obtained by slow diffusion of diethyl ether into an acetonitrile solution of the compound.
3.2.2 X-ray Structure Determinations
Crystals of [Cr(pod-ester)](BF4)3 and [Cr(impy-ester)3](BF4)3 suitable for X-ray analysis
were coated with Paratone-N oil and supported on a Cryoloop before being mounted on a Bruker
Kappa Apex II CCD diffractometer under a stream of cold dinitrogen. Data collection was
performed at 120 K with Mo Kα radiation and a graphite monochromator, targeting complete
coverage and 4-fold redundancy. Initial lattice parameters were determined from 342 reflections
(for [Cr(pod-ester)](BF4)3) and 500 reflections (for [Cr(impy-ester)3](BF4)3) harvested from 36
frames; these parameters were later refined against all data. Crystallographic data and metric
parameters for [Cr(pod-ester)](BF4)3 and [Cr(impy-ester)3](BF4)3 are presented in Table 3.1.
Data were integrated and corrected for Lorentz and polarization effects using SAINT, and
semiempirical absorption corrections were applied using SADABS.27
The structure of [Cr(pod-
ester)](BF4)3 was solved by direct methods, and the structure of [Cr(impy-ester)3](BF4)3 was
solved by Patterson map; both structures were refined against F2 with the SHELXTL 6.14
software package.28
Unless otherwise noted, thermal parameters for all non-hydrogen atoms
were refined anisotropically. Hydrogen atoms were added at the ideal positions and were refined
using a riding model where the thermal parameters were set at 1.2 times those of the attached
carbon atom (1.5 for methyl protons).
72
Crystals of complex [Cr(impy-ester)3](BF4)3 contain loosely held solvate molecules,
which quickly exit the lattice upon removal of the crystals from the mother liquor. This leads to
rapid cracking of the crystals, even at 120 K, and results in moderate resolution and mediocre
residuals. The residual electron density from the severely disordered/partially occupied solvate
molecules could not be modeled satisfactorily, so the data were treated with SQUEEZE,29
which
finds an 851.8 Å3 solvent void with 374 e
−/unit cell corresponding to approximately 8 diethyl
ether molecules. The data in Table 3.1 do not include the components removed by SQUEEZE. In
addition to the solvent disorder, there is positional disorder in one ligand and in two anions. The
methyl group on the imine containing C21 is disordered over two sites, and the methyl group on
the ester containing C30 is disordered over two sites; site occupancies refine to 57:43 and 77:23
ratios for the groups containing C21 and C30, respectively. Two BF4− anions (containing B2 and
B3) have two F atoms (F5, F6, F9, F12) split over two positions in 81:19 and 57:43 ratios,
respectively.
3.2.3 Photophysical Methods
Much of the instrumentation used was the same as described in Chapter 2, and only
salient details are discussed here. For static emission spectra and quantum yield measurements,
complexes were dissolved in CH3CN (HPLC-UV grade, Honeywell B&J High Purity Solvent) at
room temperature, placed in quartz cuvettes of 1 cm path length, and diluted to give absorbances
of ∼0.1 at the excitation wavelength. Sample solutions were purged with argon (ultra pure carrier
grade, Airgas Inc.) for 15 min to remove oxygen before measurement. Samples were excited at
355 nm, and emission was recorded. To avoid Rayleigh scattering contamination in the emission
73
spectra from higher orders of the 355 nm excitation, a 560 nm long-passfilter was placed in front
of the grating. The emission spectra were corrected for solvent scattering background and PMT
response. The emission quantum yields were determined by comparing the integrated emission
of each complex to the standard [Ru(bpy)3](PF6)2 with a known quantum yield of 0.095 in room
temperature CH3CN.30
Note that this quantum yield value for the [Ru(bpy)3](PF6)2 standard is
updated relative to the value (ϕem = 0.062)31
which has been extensively used in past literature.
For the measurement of emission lifetimes, sample solutions were prepared and
deoxygenated in exactly the same manner as described above and data were collected at room
temperature. The sample was excited by 355 nm laser pulses (10 Hz; ∼0.3 μJ/pulse) with a pulse
width of 3−5 ns (fwhm). The emitted light was selected by a notch filter (750 ± 5 nm) and
detected at 90° with respect to the excitation laser. Each emission transient was obtained by
averaging 3000 scans. The data were fit with a single exponential decay model using a LabView
program with a code of local origin.
Both static emission and emission lifetime data were acquired from multiple experiments.
Measurements were collected on three different days, in each case using freshly prepared
solutions due to degradation issues. For each of these samples on each of these days, three data
sets were collected. Quantum yield and lifetime values obtained from those nine measurements
were averaged, and the standard deviations were calculated. The percentage experimental errors
(± error %) reported herein represent two times the standard deviation.
74
3.2.4 Other Physical Methods
Absorption spectra were obtained with a Hewlett-Packard 8453 spectrophotometer in
quartz cuvettes with 1 cm or 1 mm path lengths; all experiments were performed at room
temperature. Infrared spectra were measured with a Nicolet 380 FT-IR spectrometer. Mass
spectrometric measurements were performed in either the positive ion or negative ion mode on a
Finnigan LCQ Duo mass spectrometer, equipped with an analytical electrospray ion source and a
quadrupole ion trap mass analyzer. High resolution mass spectrometric measurements were
performed in positive ion mode on an Agilent 6210 TOF LC/MS instrument, equipped with both
electrospray and atmospheric pressure chemical ionization sources and an orthogonal-axis time-
of-flight mass analyzer. Cyclic voltammetry experiments were carried out inside a
dinitrogenfilled glovebox in 0.1 M solutions of (Bu4N)PF6 in acetonitrile unless otherwise noted.
The voltammograms were recorded with either a CH Instruments 1230A or 660C potentiostat
using a 0.25 mm Pt disk working electrode, Ag wire quasireference electrode, and a Pt wire
auxiliary electrode. All voltammograms shown were measured with a scan rate of 0.1 V/s.
Reported potentials are referenced to the ferrocenium/ferrocene (Fc+/0
) redox couple and were
determined by adding ferrocene as an internal standard at the conclusion of each electrochemical
experiment. Solid state magnetic susceptibility measurements were performed using a Quantum
Design model MPMS-XL SQUID magnetometer at 295 K on finely ground samples. The data
were corrected for the magnetization of the sample holder by subtracting the susceptibility of an
empty container; diamagnetic corrections were applied using Pascal’s constants.32
Elemental
analyses were performed by Robertson Microlit Laboratories, Inc. in Madison, NJ
75
3.2.5 Electronic Structure Calculations
Unrestricted B3LYP hybrid density functional studies33
were carried out in the G09 suite
of electronic structure codes.34
Geometries were optimized for each of the quartet ground states.
For the pod-ester ligand, the Cr−Namine was constrained to the experimental bond distance of 3.12
Å. Methyl iminopyridine ligands impy’ and impy-ester’ were used instead of the ethyl
iminopyridine ligands impy and impy-ester. The LANL235
basis sets and effective core
potentials were used for Cr; H, B, C, N, and F were described with a 6-31g* model.36-39
For the
spin unrestricted MS = 3/2 “quartet”, the excited state energies were computed using TD-DFT,
wherein at least the lowest 16 excited states were computed. The number of excited states
computed was incrementally increased until the excited state manifold reached 3.5 eV; the
lowest 13 for [Cr(bpy)3](BF4)3, 23 for [Cr(4-dmcbpy)3](BF4)3, 19 for [Cr(pod-ester)](BF4)3, and
24 for [Cr(impy-ester’)3](BF4)3 quartet excited states are reported. For the doublet manifold,
broken symmetry unrestricted solutions were obtained for the ααβ, αβα, and βαα MS = 1/2
determinants. As described previously,40
these three single determinantal descriptions are
combined to form two multideterminantal (MD-DFT), nearly degenerate doublet states
(equations 2−9 in reference 40). In addition, the lowest energy of the three single determinantal
MS = 1/2 “doublet” models was used to compute a TD-DFT excited state manifold;41
again the
number of excited states computed was incrementally increased until the excited state manifold
reached 3.5 eV (relative to the lowest doublet). The doublet TD-DFT excitation energies were
offset relative to the lowest MS = 1/2 “doublet”state. To compare with our MD-DFT results,
estimates for spin-projected doublet states were obtained using the Soda et al. model,42
reproduced in Equation 3.1.
76
In the Soda model and in the predecessor Noodleman-Davidson model,43
the expectation values
of S2 for the high-spin and related spin-flipped single determinantal models are used to project
out spin contamination and more accurately estimate the energy for the true low-spin model. In
Equation 3.1 the MS = 3/2 “quartet” and one of the MS = 1/2 “doublet” models are used as the
high-spin and broken symmetry models, respectively. For the MS = 5/2 “sextet” manifold, a spin
unrestricted SCF solution was obtained and the lowest 8 excited states were computed using TD-
DFT.
77
3.3 Results and Discussion
3.3.1 Synthesis and Structural Characterization of Cr(III) Complexes
The Cr(III) complexes are synthesized from the Cr(II) starting material
[Cr(CH3CN)4(BF4)2] by the addition of stoichiometric amounts of the hexadentate ligands
(Scheme 3.2a) or bidentate ligands (Scheme 3.2b), followed by oxidation by the noncoordinating
oxidant thianthrene tetrafluoroborate (Th+BF4
−).
Scheme 3.2 Preparation of the Cr(III) Complexes
Although there is literature precedent for synthesizing [Cr(pod)](ClO4)3 from Cr(III)
precursors,15
in our hands we find that an oxidative route provides significantly higher yields.
The complex salts [Cr(pod-ester)](BF4)3 and [Cr(impy-ester)3](BF4)3 can be isolated as
78
crystalline solids. X-ray structural data are available for [Cr(pod-ester)](BF4)3·CH3CN and
[Cr(impy-ester)3](BF4)3 (Table 3.1). Although the structure of [Cr(pod)](ClO4)3 has been
reported in the literature,15
samples of [Cr(pod)](BF4)3 and [Cr(impy)3](BF4)3 have not been
crystallized.
Table 3.1. Crystallographic Dataa for [Cr(pod-ester)](BF4)3 ∙CH3CN and [Cr(impy-ester)3](BF4)3
[Cr(pod-ester)](BF4)3 ∙CH3CN [Cr(impy-ester)3](BF4)3
Formula C32H36B3CrF12N8O6 C30H36B3CrF12N6O6
Formula Weight 941.12 899.08
Color and Habit Tan needle Yellow block
T, K 120(2) 120(2)
Space Group P63 P21/n
Z 2 4
a, Å 12.8725(2) 13.3694(8)
b, Å 12.8725(2) 16.6678(9)
c, Å 13.6811(4) 20.3617(11)
α, deg 90 90
β, deg 90 99.601(4)
γ, deg 120 90
V, Å3 1963.26(7) 4473.8(4)
dcalc, g/cm3 1.592 1.320
GOF 1.140 1.105
R1 (wR2)b, % 2.94 (7.83) 8.03 (24.49)
a Obtained with graphite-monochromated Mo Kα (λ = 0.71073 Å) radiation.
b R1 = ∑||Fo| − |Fc|| / ∑|Fo|, wR2 = [∑w(Fo
2 − Fc
2)2/∑w(Fo
2)2]1/2
for Fo>4σ(Fo).
Crystals of [Cr(pod-ester)](BF4)3·CH3CN can be grown by slow diffusion of diethyl ether
into acetonitrile solutions of [Cr(pod-ester)](BF4)3. The complex cation is shown in Figure 3.1.
The complex crystallizes in the hexagonal space group P63, where the chromium center and an
acetonitrile molecule sit on sites of 3-fold symmetry. The chromium center is ligated to the
79
tripodal iminopyridine ligand by three imino and three pyridine nitrogen atoms, producing a
distorted octahedral environment (Figures 3.2 and 3.3). Due to the metal position on a 3-fold
symmetric site, only one arm of the ligand is crystallographically independent. Also present in
the asymmetric unit is a single crystallographically independent tetrafluoroborate anion.
Figure 3.1 Structures of the complex cations [Cr(pod-ester)]3+
(top) and [Cr(impy-ester)3]3+
(bottom), as observed in [Cr(pod-ester)](BF4)3·CH3CN and [Cr(impy-ester)3](BF4)3, respectively,
rendered with 40% ellipsoids. Gray, red, blue, and pink atoms are carbon, oxygen, nitrogen, and
chromium, respectively. The metal center and nitrogen atoms are labeled. Hydrogen atoms,
solvent molecules, and minority disordered components are omitted for clarity. The Λ isomer has
been shown of [Cr(impy-ester)3](BF4)3, but both stereoisomers (Λ and Δ) are present in the
nonchiral space group P21/n.
80
Figure 3.2 The full complex cation, and crystallographically independent anion and solvent
molecules in the structure of [Cr(pod-ester)](BF4)3·CH3CN, shown at 40% ellipsoids. The cation
and acetonitrile each sit on a special position of three-fold rotational symmetry and only the
crystallographically independent atoms are labeled.
Figure 3.3 Alternate view of Cr(III) complex cation in the structure of [Cr(pod-
ester)](BF4)3·CH3CN, highlighting the octahedral coordination of the Cr center. Hydrogen atoms
are omitted for clarity and structure is shown at 40% ellipsoids.
81
Table 3.2. Average Cr−N Distances (Å) for [Cr(pod)](ClO4)3,a [Cr(pod-ester)](BF4)3·CH3CN,
and [Cr(impy-ester)3](BF4)3
Complex Cr−Npy Cr−Nimine Cr∙∙∙Namine
[Cr(pod)](ClO4)3 2.062[9]b 2.044[9]
b 3.155(5)
c
[Cr(pod-ester)](BF4)3·CH3CN 2.067(2)c 2.050(1)
c 3.120(2)
c
[Cr(impy-ester)3](BF4)3 2.063[5]b 2.044[5]
b
a See reference 15.
b The errors for these bond distances were calculated by averaging the bond distances for each
type of bond and taking the square root of the sum of the squares of the bond esds. c There is only one crystallographically independent bond of this type, so there are no average
bond distances
Comparison of the structure of [Cr(pod-ester)](BF4)3·CH3CN and the previously reported
structure of [Cr(pod)](ClO4)315
shows many similarities and a few small differences. Both
complex cations crystallize as facial isomers, and the local coordination environment for
[Cr(pod)](ClO4)3 and [Cr(pod-ester)](BF4)3·CH3CN are very similar. Table 3.2 contains average
Cr−N distances for [Cr(pod)](ClO4)3, [Cr(pod-ester)](BF4)3·CH3CN, and [Cr(impy-ester)3](BF4)3.
Within error, the Cr−N bond distances are identical for the two tripodal complexes: for
[Cr(pod)](ClO4)3 the average Cr−Npy and Cr−Nimine distances are 2.062[9] and 2.044[9] Å, (44)
respectively; for [Cr(pod-ester)](BF4)3·CH3CN the Cr−Npy and Cr−Nimine distances are 2.067(2)
and 2.050(1) Å, respectively. In addition, for both structures the trigonal twist angles are very
similar (average 52.54° for [Cr(pod)](ClO4)3 and 53.00(12)° for [Cr(pod-ester)](BF4)3·CH3CN)
and lower than the 60° expected for an ideal octahedral geometry. The main difference between
the two structures is the distance between the Cr and the bridgehead nitrogen (Namine) atoms: in
[Cr(pod-ester)](BF4)3·CH3CN it is 3.120(2) Å, whereas in [Cr(pod)](ClO4)3 it is 3.155(5) Å. The
difference in the position of Namine may be due in part to solvation: [Cr(pod)](ClO4)3 lacks any
cocrystallized solvent, whereas [Cr(pod-ester)](BF4)3·CH3CN contains an acetonitrile molecule.
The shorter Cr∙∙∙Namine distance in [Cr(pod-ester)](BF4)3·CH3CN may also be attributable to
82
weaker binding of the ester-functionalized ligand pod-ester, or other packing effects. The
acetonitrile solvate nitrogen atom is in close contact with the 6-position hydrogen atoms on each
of the three pyridine moieties on the iminopyridine ligand at a distance of 2.714(2) Å. There are
no close contacts between the solvent molecule and the bridgehead nitrogen, however.
Yellow block crystals of compound [Cr(impy-ester)3](BF4)3 can also be grown by ether
diffusion into acetonitrile solutions of [Cr(impy-ester)3](BF4)3. The crystals contain a significant
amount of solvent, which quickly exits the lattice upon removal of the crystals from the mother
liquor. The volatile nature of the solvent present in the lattice leads to severe solvent disorder and
eventual cracking of the crystals. The data for [Cr(impy-ester)3](BF4)3 are presented herein to
establish complex connectivity and provide additional characterization of the compound; solvent
disorder does not appear to adversely affect determination of interactions relevant to the Cr-
containing species. The complex cation in [Cr(impy-ester)3](BF4)3 is present as the meridinal
isomer (Figure 3.1, bottom, and Figure 3.4), and there is no evidence for fac/mer disorder. The
Cr−N bond distances in [Cr(impy-ester)3](BF4)3 are not significantly different from those found
in [Cr(pod-ester)](BF4)3·CH3CN, with an average of 2.063[5] and 2.044[5] Å for the Cr−Npy and
Cr−Nimine distances, respectively (Table 3.2).
The isolated salts of all four complexes are soluble in strongly polar, aprotic solvents
such as acetonitrile and nitromethane, but are only slightly soluble in strongly polar protic
solvents such as methanol and water. The salts are insoluble in less polar solvents such as
dichloromethane, tetrahydrofuran, diethyl ether, and hydrocarbons. The complexes dissolve
readily in 1 M HCl(aq), but degrade quickly, as discussed below.
83
Figure 3.4 The full structure of [Cr(impy-ester)3](BF4)3 including disordered components
(labeled A and B) at 40% ellipsoids. The Λ isomer has been shown of [Cr(impy-ester)3](BF4)3,
but both stereoisomers (Λ and Δ) are present in the non-chiral space group P21/n.
84
3.3.2 Electrochemistry
Cyclic voltammogram (CV) data collected on fresh acetonitrile solutions with 0.1 M
TBA+PF6
− as the supporting electrolyte are shown in Figure 3.5. Each of the Cr(III) complexes
undergoes multiple reversible reductions on the CV time scale. The reduction potentials for each
of the complexes (relative to Fc+/0
) are reported in Table 3.3.
Figure 3.5 Comparison of cyclic voltammograms for all four complexes in 0.1 M TBAPF6
acetonitrile solution.
85
Table 3.3. Ground State Reduction Potentials for Cr(III) Complexes
Complex E1/2a 3+/2+ 2+/1+ 1+/0 0/1− 1−/2−
[Cr(pod)]3+
−0.45 (71) −0.93 (70) −1.55 (77) −2.44 (160) irr
[Cr(pod-ester)]3+
−0.25 (71) −0.67 (71) −1.12 (71) −1.89 (81) −2.18 (86)
[Cr(impy)3]3+
−0.41 (69) −0.90 (71) −1.56 (72) −2.40 (89) irr
[Cr(impy-ester)3]3+
−0.20 (76) −0.63 (73) −1.11 (74) −1.87 (78) −2.17 (91)
[Cr(bpy)3]3+ b
−0.63 (72) −1.15 (71) −1.72 (69) −2.34 (74)
[Cr(phen)3]3+ b
−0.65 (70) −1.17 (72) −1.71 (75) −2.21 (77) a Potentials reported in V vs Fc
+/0 (ΔEp in mV). Conditions for cyclic voltammetry of Cr
complexes: electrolyte, 0.1 M TBAPF6 in CH3CN; WE, Pt; CE, Pt wire; scan rate, 100 mV/s. b Data from reference 3.
When comparing the podand to the tris(bidendate) analogues ([Cr(pod)](BF4)3 vs
[Cr(impy)3](BF4)3 and [Cr(pod-ester)](BF4)3 vs [Cr(impy-ester)3](BF4)3, respectively), inclusion
of the nitrogen tether atom results in a shift of the first reduction wave to more negative
potentials by 50 mV, presumably because of electron-donating properties of the tether nitrogen.44
For successive reduction waves, the difference in potentials between tethered and nontethered
ligands varies between 10 and 40 mV, and is smallest for the 1+/0 wave.
A more profound effect involves the inclusion of the electron-withdrawing ester groups,
which shift the first reduction waves toward positive potentials by approximately 200 mV. For
each successive reduction, the potential difference between the ester-functionalized and parent
complexes increases. Additionally, the ester-functionalized complexes have a fifth reversible
reduction wave that is accessible within the solvent window, whereas the parent complexes only
have four accessible reduction waves and the waves at the most negative potentials are
irreversible. The much larger effect of the ester functionality compared to that of the nitrogen
tether is expected given the ester’s presence within the conjugated π-system of the iminopyridine
ligands.
86
While there are very few chromium impy complexes reported in the literature,45-48
the
electrochemistry reported here is very similar to electrochemistry for a bis(iminopyridine) Cr(III)
complex reported by Wieghardt and co-workers.49
For our complexes, each imine on the
unfunctionalized parent ligand is able to undergo a one electron reduction, generating a ligand
radical. Addition of a fourth reducing equivalent to the tris-bidentate or tripodal hexadentate
complex leads to an irreversible reduction. The inclusion of the electron-withdrawing ester
groups allows for two additional reversible reductions to take place for the complex overall. It is
likely that a third additional reduction wave would be present, but it is outside the acetonitrile
solvent window.
3.3.3 Electronic Absorption
For comparison, ground state electronic absorption spectra for fresh acetonitrile solutions
of each of the complexes are shown in Figures 3.6. The spectra for each of the complexes are
similar at short wavelengths, with a strong absorption peak near 205 nm. This peak correlates
with a peak observed in the spectra of all free ligands near 200 nm. All of the iminopyridine
Cr(III) complexes are moderately strong UV absorbers with ε of 10000−20000 M−1
cm−1
between
250 and 325 nm. Comparing the two podand species [Cr(pod)](BF4)3 and [Cr(pod-ester)](BF4)3,
a shoulder to the 205 nm peak occurs at 255 nm for each complex. This shoulder correlates to a
peak and shoulder at 245 and 254 nm, respectively, found in the free pod-ester ligand (Figure
3.7). In the free pod ligand, the analogous peak is significantly blue-shifted and appears at 233
nm. For a comparison of the spectral features from the two tris-bidentate complexes
[Cr(impy)3](BF4)3 and [Cr(impy-ester)3](BF4)3, a pronounced shoulder near 235 nm appears in
87
each. This shoulder matches well with a peak that occurs at 234 nm for the free impy ligand, and
is shifted slightly from the analogous peak in the free impy-ester ligand, which lies at 245 nm.
Figure 3.6 Electronic absorption spectra collected in acetonitrile for Cr(III) complexes in the UV
range (main) and the visible range (inset). All spectra were collected at room temperature.
Figure 3.7 Electronic absorption spectra of the free ligands pod and pod-ester in acetonitrile and
impy and impy-ester in pentane (~10 μM to 1 mM).
88
When comparing the two parent complexes to the ester-functionalized species, the ester-
containing complexes show higher molar absorptivities in the UV region of λ < 330 nm. This is
most likely due to the additional conjugation the ester groups provide to the iminopyridine
ligands, thus impacting ligand-centered absorptions as well as transitions that derive intensity
from charge-transfer character involving the ligands.
In addition to the ultraviolet ligand-based absorptions, there is a broad peak and shoulder
feature at ∼300 and∼325 nm, respectively, for each of the four complexes. These transitions are
likely to be charge transfer in nature because of their absence in the free ligands and their
relatively high intensities. The lowest energy shoulder is red-shifted by 5-6 nm for the ester
functionalized ligand complexes relative their parent iminopyridine ligand complexes. This red
shift is consistent with expectations for metal-to-ligand charge-transfer (MLCT) given the
presence of electron withdrawing substituents on the ester ligands pod-ester and impy-ester.
However, one might also expect red shifting for ligand-to-metal charge transfer (LMCT) if such
ligands serve to reduce electron-electron repulsion in the metal-centered orbitals. It is noted that
the ∼350 nm band in [Cr(bpy)3]3+
is most often attributed to LMCT,50-52
since there is an
energetic penalty for oxidizing Cr(III) to Cr(IV) as would formally occur during MLCT.3
The molar absorptivities of the two tris(bidentate) complexes diminish more rapidly at λ
> 350 nm compared to their tripodal counterparts. Optically, this translates to more visible
coloration in [Cr(pod)](BF4)3 and [Cr(pod-ester)](BF4)3 compared to [Cr(impy)3](BF4)3 and
[Cr(impy-ester)3](BF4)3. Solutions of complex [Cr(pod)](BF4)3 appear tan-orange, and solutions
of complex [Cr(pod-ester)](BF4)3, which has an additional peak centered at 390 nm with an ε =
2100 M−1
cm−1
, are tan-brown. Compounds [Cr(impy)3](BF4)3 and [Cr(impy-ester)3](BF4)3
appear light yellow in solution as well as in the solid state.
89
Overall, changes to the ligand structure impact the absorption properties of the metal
complexes. We find that addition of ester groups to the iminopyridine ligands helps to increase
molar absorptivity for the UV and near-UV region, while addition of the nitrogen tether
increases absorptivities in the visible wavelengths between 350 and 650 nm.
3.3.4 Probing Complex Stability in Solution
We hypothe-sized that tethering of the iminopyridine ligands would increase overall
complex stabilities in solution, both in the ground state and in photoinduced excited states.
Because photophysical properties for related tris(dipyridyl) Cr(III) complexes must be studied in
acidic media to avoid photoexcited ligand substitution and/or solvolysis,4 we first probed the
properties of [Cr(pod)](BF4)3, [Cr(pod-ester)](BF4)3, and [Cr(impy-ester)3](BF4)3 in 1 M HCl(aq).
Within 1 hour of dissolution, their spectra indicate conversion of the initial complex to another
species. A white precipitate observed in acidic aqueous solutions of [Cr(pod-ester)](BF4)3
indicates the loss of the hexadentate ligand, which may be initiated by protonation of the
bridgehead nitrogen in acidic media. Yellow-to-pink color changes observed for [Cr(impy-
ester)3](BF4)3 are reminiscent of the color of bis(bidentate) complex [Cr(phen)2(OTf)2](OTf) in
the solid state and in acetonitrile solution,50
and suggest the loss of one of the iminopyridine
ligands with replacement by solvent molecules or coordinating anions. However, we have never
observed hydrolysis of the imino groups in 1 M HCl(aq). The fact that [Cr(pod)](BF4)3 also
changes over time indicates that the presence or absence of the ester functionality does not
significantly impact the stability of the complexes in acidic solution.
90
Since the complexes have good solubility in acetonitrile and we are able to crystallize
pure [Cr(pod-ester)](BF4)3 and [Cr(impy-ester)3](BF4)3 from this solvent, we investigated the
stability of those compounds in acetonitrile as an alternative to acidic media. Relative to
[Cr(phen)3](OTf)3, ground state absorption for [Cr(pod-ester)](BF4)3 and [Cr(impy-ester)3](BF4)3
in acetonitrile shows small shifts over 24 hours. One possible explanation for the spectroscopic
changes observed in acetonitrile could be related to the hydrolytic sensitivity of the charge-
balancing BF4− anion to trace amounts of water, which is accelerated in the presence of acidic
species.53-54
Mass spectra of mixed CH3CN/water solutions of [Cr(pod-ester)](BF4)3 and
[Cr(impy-ester)3](BF4)3 show peaks that contain fluorine without boron (e.g., [Cr(impy-
ester)2F2]+). The loss of BF4
− as a weakly interacting anion and its replacement by either
fluoride- or oxygen-containing hydrolysis products would likely generate Cr species with
directly coordinated anions, resulting in qualitatively different absorption spectra. Note that the
hygroscopic nature of [Cr(pod)](BF4)3 and [Cr(impy)3](BF4)3 could contribute further to anion
instability for those compounds.
We conclude that, at least for the Cr(III) species examined here, kinetic stability is not
significantly enhanced by use of a hexadentate tripodal ligand relative to the bidentate iminopyr-
idine species: [Cr(pod)](BF4)3 and [Cr(impy)3](BF4)3 show similar solution chemistries. The
photophysical studies discussed below are carried out in fresh solutions of the compounds using
dry acetonitrile.
91
3.3.5 Excited State Properties of the Complexes
In the photophysical discussions that follow, we focus on the two compounds ([Cr(pod-
ester)](BF4)3 and [Cr(impy-ester)3](BF4)3) which can be obtained as X-ray quality crystals, and
more importantly contain potential attachment points to a semiconductor surface in the form of
ester functional groups. Following excitation at 355 nm, compound [Cr(impy-ester)3](BF4)3 is
emissive in deoxygenated acetonitrile at room temperature (Figures 3.8). The spectrum is similar
to that observed for Cr(III) tris-bipyridyl complexes.3 The most intense band at 740 nm (1.68 eV)
is assigned to 4A ←
2E phosphorescence. A weaker shoulder observed at 700 nm is expected to
originate from the 2T state, which is thermally equilibrated with the
2E state.
55 An excitation
spectrum (Figure 3.9) shows that excited states produced in the complex via excitation
throughout the reddest portion of the UV absorption spectrum convert to the emissive 2E state.
Figure 3.8 Emission spectra for freshly prepared solutions of [Cr(pod-ester)](BF4)3 and
[Cr(impy-ester)3](BF4)3 in deoxygenated acetonitrile at room temperature following 355 nm
excitation.
92
Figure 3.9 2E excitation spectrum for a fresh prepared solution of [Cr(impy-ester)3](BF4)3 in
deoxygenated acetonitrile at room temperature obtained by monitoring 740 nm emission
intensity while varying the excitation wavelength. The excitation spectrum and ground state
absorption spectrum are normalized and overlaid for comparison.
The 2E emission quantum yield (ϕem) of [Cr(impy-ester)3](BF4)3 is measured (by relative
comparison to a [Ru(bpy)3](PF6)2 standard) to be 0.00061 ± 14% in deoxygenated acetonitrile at
room temperature. This represents a 64% reduction in emissive quantum yield compared to
[Cr(bpy)3](OTf)3 where we observe ϕem = 0.0017 ± 18% under the same conditions.
Using nanosecond time-resolved emission spectroscopy, the 2E excited-state lifetime (τobs)
of [Cr(bpy)3](OTf)3 was measured. These data are fit with a single exponential decay model
indicating τobs = 19 μs ± 3% in deoxygenated acetonitrile at room temperature. Transient
absorption kinetics (following excitation with a ∼5 ns pulse at 355 nm) have also been measured
in deoxygenated acetonitrile at room temperature throughout the near-UV and visible spectra
(350-590 nm); all data show single-exponential decay of absorption features with a comparable
93
time constant (τobs = 17 μs ± 5%). The μs time scale of emission (and absorption) decay agrees
with the expected behavior of the 2E state, whose lowest energy radiative pathway is
4A ← 2
E
phosphorescence. The observed lifetime is comparable to molecules such as [Cr(bpy)3](OTf)3,
where 35 μs was measured under the same conditions used in this work.
It is common to combine both static and time-resolved emission data to determine
nonradiative (∑knr) and radiative (∑kr) rate constants according to Equation 3.2, where φform
refers to the quantum yield of formation of the lowest energy excited state from the Franck-
Condon state.
Assuming that φform is close to unity56
(an assumption that has not yet been tested extensively for
this complex), we find that, for [Cr(impy-ester)3](BF4)3, Σkr = 32 s−1
and Σknr = 52600 s−1
are to
be compared with Σkr = 47 s−1
and Σknr = 27700 s−1
measured for [Cr(bpy)3](OTf)3 under the
same conditions. Thus, for [Cr(impy-ester)3](BF4)3 relative to [Cr(bpy)3](OTf)3, Σkr is decreased
by 32% while Σknr nearly doubles. Both quantities contribute to the lower emissive quantum
yield for [Cr(impy-ester)3](BF4)3. Although the excited state lifetime of [Cr(impy-ester)3](BF4)3
is shortened relative to [Cr(bpy)3](OTf)3, it is still in the μs time scale. If stability issues were not
a concern, [Cr(impy-ester)3](BF4)3 could be a candidate for performing photoinduced oxidation
reactions or interfacial hole-transfer photochemistry, provided that the quantum yield for forming
the 2E (i.e., φform) is indeed high.
Unlike the tris(bidentate) complex [Cr(impy-ester)3](BF4)3, the hexadentate complex
[Cr(pod-ester)](BF4)3 has shown no evidence for transient absorption throughout the visible
94
(beyond the ∼5 ns excitation pulse width) or 2E emission at room temperature. This latter point
is shown, for example, in Figure 3.8 in the wavelength region 650-850 nm where no emitted
light is detected. This region is inclusive of where 2E emission (which is particularly insensitive
to ligand or environment for the d3 electronic configuration) would be observed for most Cr(III)
species.57
The crystallographic data for [Cr(pod-ester)](BF4)3 discussed previously in this paper
are unremarkable with respect to the Cr coordination environment and would not preclude a
ligand field comparable to what is present in [Cr(impy-ester)3](BF4)3. The major difference in the
coordination environments between [Cr(pod-ester)](BF4)3 and [Cr(impy-ester)3](BF4)3 is the
facial versus meridonal arrangement of the iminopyridine moieties, respectively. However, for
[Cr(α-picolylamine)]3+
, it is reported that fac- versus mer-coordination imparts only small
differences to complex emission properties, and both geometries give rise to 2E emission with
lifetimes near 200 μs at 77 K.8
Previous reports point to trigonal twisting as a major mechanism for 2E relaxation in
facially capped Cr(III) complexes.58-59
However, an important distinction can be made between
the inherently strained ligands which only have a −CH2 bridge between the capping atom and the
first chelating N described in the literature, and the ligand pod-ester in the present work, which
has a −(CH2)2 bridge. The increased length and flexibility of this bridge does not produce strain
of the type studied in the previous reports. Comparing the bond angles of the bridge atoms in the
structure of [Cr(pod-ester)](BF4)3 and the structure of [Cr(sen)]Br3, the sen −CH2 bridge is
strained with larger angles than what is expected for a tetrahedral carbon (114-115°);58
however,
both −CH2 groups in the bridge for the ligand pod-ester have angles close to the 109.5° expected
for a tetrahedral carbon: 110.0(2)° for C1 and 108.7(1)° for C2. This indicates the ground state
95
structure of pod-ester in [Cr(pod-ester)](BF4)3 is relatively unstrained, unlike the previously
studied facially capped Cr(III) complexes.
The lack of 2E emission at room temperature in our case therefore suggests that either
Σknr is large or φform is small. The details are important and will be elucidated in future work,
including low temperature emission studies and transient absorption studies. Nevertheless, these
initial observations, specifically the apparent absence of appreciable 2E lifetime at room
temperature, suggest that [Cr(pod-ester)](BF4)3 is not ideal as a sensitizer for excited-state redox
chemistry.
In addition to the 2E emission found and discussed above for [Cr(impy-ester)3](BF4)3, we
observed higher energy emitted light in the region of 350-550 nm for samples of [Cr(impy-
ester)3](BF4)3 as well as for [Cr(pod-ester)](BF4)3 with much lower intensity (Figure 3.10).
Attempts were made to measure the excited-state lifetimes using these emission bands; however,
no μs emission kinetics were observed and emission could not be resolved with a laser pulse
width of ∼5 ns. This suggests that excited-state lifetimes elucidated via these emission bands
decay on a time scale of ns or shorter.
Figure 3.10 Emission spectra for [Cr(pod-ester)](BF4)3 and [Cr(impy-ester)3](BF4)3 in
acetonitrile right after dissolution and at 24 hours.
96
We first considered whether the higher energy emission for [Cr(impy-ester)3](BF4)3
suggested dual emission, wherein the Franck-Condon state partitions between 2E formation and
decay via other pathways, some of which are radiative at ∼425 nm (Figure 3.10). Such
photophysics would contribute to measurement of a small φem via nonunity φform. There are two
observations that lead us to consider that dual emission is unlikely. First, wavelength-dependent
excitation scans of emission collected at 470 nm show a stark difference to those collected at 740
nm (Figure 3.9) where there is excellent agreement between the features of ground-state
absorbance and the excitation scans. Second, acetonitrile solutions of [Cr(impy-ester)3](BF4)3
aged for 24 hours exhibit a significant increase in the emission intensity of the ∼425 nm band
without significant change to the 740 nm 2E band (Figure 3.10). These observations suggest the
growth of a small amount of a new strongly emissive compound as the samples age in solution.
Studies performed on [Cr(pod-ester)](BF4)3 give similar results, although the percentage change
is significantly lower than what is observed for the tris(bidentate) complex [Cr(impy-
ester)3](BF4)3.
Cr(III) complexes with bidentate ligands similar to the iminopyridines reported here have
been studied.8 At 77 K, [Cr(α-picolylamine)3]
3+ shows emission at ∼700 nm, which was assigned
to be 4A ←
2E phosphorescence, and another emission band at ∼390 nm was assigned to be
ligand-centered emission. One important piece of evidence that supports the authors’ assignment
was that the free α-picolylamine is emissive under the same condition. However, the free ligands
pod, pod-ester, impy, and impy-ester are not emissive in room temperature acetonitrile,
suggesting that the 350-550 nm emission observed in [Cr(impy-ester)3](BF4)3 and [Cr(pod-
ester)](BF4)3 is unlikely to be due to free ligand in the case of [Cr(impy-ester)3](BF4)3 or from a
dissociated ligand arm in the case of [Cr(pod-ester)](BF4)3.
97
3.3.6 Electronic Structure Considerations
Photophysical behavior of Cr(III) complexes is generally conceived as involving the
quartet and doublet manifolds, where the lowest energy, long-lived doublet state is essentially a
linear combination of spin-flipped (MS = 1/2) configurations of the metal-based quartet ground
state. In principle, sextet states may also contribute to the photophysical behavior observed. On
the basis of the results presented here as well as previous work, it is clear that the ligands are
heavily involved in complex properties: ground state reduction of these kinds of complexes
places the reducing equivalent on the ligand,22,48-49
and small changes in the ligand set appear to
deactivate emission in [Cr(pod-ester)]3+
compared to [Cr(impy-ester)3]3+
.
To explore a wide range of deactivation scenarios, multi-determinantal (MD) UB3LYP-
DFT and TD-DFT computational studies were performed. Excitation energy (EE) diagrams are
presented in Figure 3.11, lowest computed excitation energies are collected in Table 3.4, and
representative natural transition orbital (NTO) plots are provided in Figure 3.12. For all
complexes studied, the lowest doublet excited states are lower in energy than the various quartet
excited states, indicating that 2E excited states should be energetically accessible for all the
iminopyridine Cr(III) complexes studied. For complex [Cr(impy-ester’)3]3+
(using methyl
iminopyridine ligand impy-ester’ instead of ethyl iminopyridine impy-ester), the MD-DFT
calculatedfirst doublet excited state energy (1.61eV) is confirmed by experimental result for
[Cr(impy-ester)3]3+
from room temperature static emission (1.68 eV). For [Cr(bpy)3]3+
, the
reported 2E emission of 1.71 eV compares well with the MD-DFT calculated energy 1.60 eV. In
addition, for all four complexes the lowest doublets concentrate spin density in the three Cr t2g
orbitals. However, whereas [Cr(impy-ester)3]3+
, [Cr(bpy)3]3+
, and [Cr(4-dmcbpy)3]3+
show
98
emission from the doublet manifold, [Cr(pod-ester)]3+
does not. Therefore, the simple presence
of a low-lying doublet does not ensure productive emission.
Figure 3.11 Computed excitation energies for selected Cr(III) complexes, where the ground state
quartet energy for each complex is set at zero. For each species, the left column is the doublet
manifold (D), the middle column is the quartet manifold (Q), and the right column is the sextet
manifold (S).
Table 3.4. Computed Excitation Energies (EEs) for Selected Complexes (eV)
Complex 1st Quartet
EE
1st Ms = ½
EE
1st Doublet
EE (proj)
1st Doublet EE
(MD-DFT)
1st Sextet
EE
[Cr(bpy)3](BF4)3 2.71 1.07 1.58 1.60 3.11
[Cr(4-dmcbpy)3](BF4)3 2.54 1.09 1.59 1.62 2.95
[Cr(pod)](BF4)3 1.92 1.05 1.55 1.57 3.14
[Cr(pod-ester)](BF4)3 1.95 1.03 1.52 1.54 3.16
[Cr(impy’)3](BF4)3 2.71 1.07 1.59 1.61 3.42
[Cr(impy-ester’)3](BF4)3 2.86 1.08 1.61 1.61 3.29
99
Figure 3.12. Natural transition orbitals (NTOs) for the lowest (a) doublet and (b) quartet
transitions for [Cr(pod-ester)]3+
(left) and [Cr(impy-ester)3]3+
(right). In each pair, the left NTO
corresponds to where the excitation is from while the right NTO is where the excitation is to.
Hydrogen atoms have been removed for clarity.
In addition to comparison with experiment for the emissive doublet, the accuracies of the
doublet manifold energies were further probed by TD-DFT studies. States within 3.5 eV of the
lowest energy doublet were computed (since the experimental pumping wavelength of 355 nm is
∼3.49 eV), absorption spectra were calculated for [Cr(pod-ester)]3+
and [Cr(bpy)3]3+
, and the
latter computed spectrum was compared to the transient absorption spectrum for [Cr(bpy)3]3+
(as
discussed above the 2E state is not observed for [Cr(pod-ester)]
3+). Good qualitative agreement
between theory and experiment is observed (Figure 3.13).
100
Figure 3.13. Calculated doublet and quartet excited state absorption spectra of [Cr(bpy)3]3+
(left)
and transient absorption spectrum of [Cr(bpy)3]3+
(right).
Figures 3.11 and Figure 3.12 highlight unique features of the tripodal complex [Cr(pod-
ester)]3+
relative to [Cr(bpy)3]3+
, [Cr(4-dmcbpy)3]3+
, and [Cr(impy-ester’)3]3+
. For the doublet
and quartet spin manifolds, there are sets of lower energy excited states for [Cr(pod-ester)]3+
,
relative to the other three complexes. For the sextet manifold, [Cr(pod-ester)]3+
does not have the
lowest energy state but does have a significantly higher density of states than [Cr(bpy)3]3+
or
[Cr(impy-ester’)3]3+
(Figure 3.11). Additionally, the second set of doublet states of [Cr(pod-
ester)]3+
(∼2.8 eV) display different orbital character than those of [Cr(impy-ester’)3]3+
(Figure
3.12).
For [Cr(pod-ester)]3+
, [Cr(4-dmcbpy)3]3+
, and [Cr(impy-ester’)3]3+
, the lowest quartet
excited states involve excitation from a ligand (ligand π or Namine) orbital to an orbital that is an
admixture of ligand π* and Cr t2g character. For [Cr(4-dmcbpy)3]3+
and [Cr(impy-ester’)3]3+
the
101
second set of doublet states arise from transitions that are dominantly Cr t2g ← Cr t2g in character
(2T1 ←
2E, lowest doublet transitions in Figures 3.12).
60
For both [Cr(4-dmcbpy)3]3+
and [Cr(impy-ester’)3]3+
there is a small admixture of ligand
character in the donating orbital, with larger admixture for the iminopyridine than for the
bipyridine complex. In contrast, for [Cr(pod-ester)]3+
, the admixture of ligand character
dominates: the transition for [Cr(pod-ester)]3+
is LMCT in nature (Cr t2g ← Namine transition). For
both [Cr(4-dmcbpy)3]3+
and [Cr(impy-ester’)3]3+
the higher energy low spin “2T1” t2g
3 doublet
states can readily undergo internal conversion to the lower “2E” t2g
3 doublet states via the large
congruence of orbital character, electronic coupling, and ligand geometry. For [Cr(pod-ester)]3+
,
due to a disparity of orbital character and ligand geometry, internal conversion from higher
doublets to the lowest energy set of doublet states competes with intersystem crossing to a set of
quartet states with congruent orbital character (and ligand geometry) which shares little orbital
character with the lowest doublet states.
In summary, the NTO analyses show several key features. First, the involvement of the
ligand’s bridgehead nitrogen helps to explain why the tripodal complex [Cr(pod-ester)]3+
has
much lower energy quartet and doublet excited states relative to its tris(bidentate) relatives
[Cr(impy-ester’)3]3+
and [Cr(4-dmcbpy)3]3+
. Second, [Cr(pod-ester)]3+
can undergo
photoexcitation similar to [Cr(impy-ester’)3]3+
and [Cr(4-dmcbpy)3]3+
, and the three complexes
have reasonable pathways for intersystem crossing into the doublet manifold. However, distinct
from the nontethered species, intersystem crossing and/or internal conversion events allow
photoexcited [Cr(pod-ester)]3+
to settle into a low-energy, largely ligand-based quartet excited
state featuring little spatial congruence with the lowest energy metal-based doublet set. This
quartet has dominant ligand-based charge transfer character, implying there will be significant
102
reorganization on both solvent and intramolecular nuclear coordinates relative to the ground state
of the molecule where the quartet character is metal-based. These factors along with the already-
noted low energy of the quartet excited state, will contribute to large nonradiative rates for
internal conversion. Indeed, the bridge-head nitrogen of the tripodal ligand introduces “real
intruder” ligand-based excited states for the complex [Cr(pod-ester)]3+
.
It is important to note that the original impetus for studying the tripodal ligand complexes,
increased absorption in the visible spectrum relative to Cr(III) dipyridyl species, is validated by
the computational results: the tripodal complexes feature increased density-of-states of the
quartet manifold at lower energy compared to the other complexes studied (Figure 3.11), which
is necessary for more efficient storage of visible spectrum energy. That the bridgehead nitrogen
is also a source of efficient deactivation pathways is not readily apparent from a standard
coordination chemical analysis of the ligand, especially for a functional group that is neither
bound to the metal ion nor conjugated with the binding groups.
103
3.4 Conclusions
A series of Cr(III) iminopyridine complexes, both in tethered and tris(bidentate) forms,
have been prepared and their photophysical properties have been studied in this work. While the
solution stability of the complexes is not markedly improved by addition of the tether, the
photophysical properties of the species are quite affected by the presence or absence of the
bridgehead nitrogen moiety. The tris(bidentate) complex [Cr(impy-ester)3]3+
shows μs emission
at room temperature, similar to aromatic diimine complexes studied previously, and consistent
with the existence of a long-lived (doublet) excited state. In contrast, the tripodal complex
[Cr(pod-ester)]3+
does not appear to emit from the doublet excited state nor give μs time-scale
transient absorption signals, and computational results show the importance of small ligand
modification on photophysical properties. In future studies, we will explore the synthesis of
podand-type ligands where the Cr(III) center can be completely incarcerated in the ligand
framework, so as to avoid the formation of species due to ligand loss or exchange. We will also
seek to replace the bridgehead nitrogen with other species to probe the electronic perturbations
on excited state behavior.
3.5 Acknowledgment on Participation
The synthesis and characterization of all the Cr(III) complexes were done by Professor
Matthew P. Shores, Dr. Ashley M. McDaniel and Ethan A. Hill at the Department of Chemistry,
Colorado State University. The computational studies were done by Professor Anthony K. app .
I was responsible for the photophysical measurements.
104
3.6 References
(1) Green, M. Third Generation Photovoltaics: Advanced Solar Energy Conversion;
Springer-Verlag: Berlin, 2003.
(2) Balzani, V.; Bolletta, F.; Gandolfi, M.; Maestri, M. Organic Chemistry and Theory;
Springer: Berlin, 1978; Vol. 75.
(3) McDaniel, A. M.; Tseng, H. W.; Damrauer, N. H.; Shores, M. P. Inorg. Chem. 2010, 49,
7981-7991.
(4) Serpone, N.; Jamieson, M. A.; Henry, M. S.; Hoffman, M. Z.; Bolletta, F.; Maestri, M. J.
Am. Chem. Soc. 1979, 101, 2907-2916.
(5) Hoffman, M. Z.; Serpone, N. Isr. J. Chem. 1982, 22, 91-97.
(6) Pizzocaro, C.; Bolte, M.; Hoffman, M. Z. J. Photochem. Photobiol. A 1992, 68, 115−119.
(7) Cheng, M. M.; Bakac, A. J. Am. Chem. Soc. 2008, 130, 5600-5605.
(8) Krause, R. A. J. Phys. Chem. 1978, 82, 2579−2581.
(9) Brown, K. N.; Geue, R. J.; Moran, G.; Ralph, S. F.; Riesen, H.; Sargeson, A. M. Chem.
Commun. 1998, 2291-2292.
(10) Lessard, R. B.; Heeg, M. J.; Buranda, T.; Perkovic, M. W.; Schwarz, C. L.; Yang, R. D.;
Endicott, J. F. Inorg. Chem. 1992, 31, 3091-3103.
(11) Ramasami, T.; Endicott, J. F.; Brubaker, G. R. J. Phys. Chem. 1983, 87, 5057-5059.
(12) Comba, P.; Creaser, I. I.; Gahan, L. R.; Harrowfield, J. M.; Lawrance, G. A.; Martin, L.
L.; Mau, A. W. H.; Sargeson, A. M.; Sasse, W. H. F.; Snow, M. R. Inorg. Chem. 1986,
25, 384-389.
(13) Williams, A. L.; Bhuiyan, A. A.; Turner, M. O.; Millett, F.; Durham, B. J. Coord. Chem.
2011, 64, 48-56.
(14) Williams, A. L. Ph.D Dissertation; University of Arkansas, 2009.
(15) McLachlan, G. A.; Fallon, G. D.; Spiccia, L. Acta Crystallogr. Sect. C 1996, 52, 309-312.
(16) Ziessel, R.; Lehn, J. M. Helv. Chim. Acta 1990, 73, 1149-1162.
(17) Saraswat, R.; Rana, V. B. J. Indian Chem. Soc. 1998, 75, 493-495.
(18) Singh, D. P.; Shishodia, N.; Yadav, B. P.; Rana, V. B. J. Indian Chem. Soc. 2004, 81,
287-290.
(19) Yang, M.; Li, X.; Chen, Y.; Deng, R. Huaxue Shiji 2002, 24, 266-267.
(20) Oila, M. J.; Tois, J. E.; Koskinen, A. M. P. Tetrahedron Lett. 2005, 46, 967-969.
(21) Jabre, N. D.; Respondek, T.; Ulku, S. A.; Korostelova, N.; Kodanko, J. J. J. Org. Chem.
2010, 75, 650-659.
(22) McDaniel, A. M.; Rappe, A. K.; Shores, M. P. Inorg. Chem. 2012, 51, 12493-12502.
(23) Henriques, R. T.; Herdtweck, E.; Kuhn, F. E.; Lopes, A. D.; Mink, J.; Romao, C. C. J.
Chem. Soc. Dalton. 1998, 1293-1297.
(24) Boduszek, B.; Shine, H. J. J. Org. Chem. 1988, 53, 5142-5143.
(25) Shavaleev, N. M.; Scopelliti, R.; Gumy, F.; Bunzli, J. C. G. Inorg. Chem. 2009, 48, 6178-
6191.
105
(26) Jantti, A.; Wagner, M.; Suontamo, R.; Kolehmainena, E.; Rissanen, K. Eur. J. Inorg.
Chem. 1998, 1555-1562.
(27) Bruker Analytical X-Ray Systems Inc. APEX 2; Madison, WI, 2008.
(28) Sheldrick, G. M. SHELXTL, Version 6.14; Bruker Analytical X-Ray Systems Inc.:
Madison, WI, 1999.
(29) Spek, A. L. J. Appl. Crystallogr. 2003, 36, 7-13.
(30) Suzuki, K.; Kobayashi, A.; Kaneko, S.; Takehira, K.; Yoshihara, T.; Ishida, H.; Shiina,
Y.; Oishic, S.; Tobita, S. Phys. Chem. Chem. Phys. 2009, 11, 9850-9860.
(31) Calvert, J. M.; Caspar, J. V.; Binstead, R. A.; Westmoreland, T. D.; Meyer, T. J. J. Am.
Chem. Soc. 1982, 104, 6620-6627.
(32) Bain, G. A.; Berry, J. F. J. Chem. Educ. 2008, 85, 532-536.
(33) Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652.
(34) Frisch, M. J.; et al. Gaussian 09, Revision A.1; Gaussian, Inc.: Wallingford, CT, 2009.
(35) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299-310.
(36) Binkley, J. S.; Pople, J. A.; Hehre, W. J. J. Am. Chem. Soc. 1980, 102, 939-947.
(37) Ditchfie.R; Hehre, W. J.; Pople, J. A. J. Chem. Phys. 1971, 54, 724-728.
(38) Francl, M. M.; Pietro, W. J.; Hehre, W. J.; Binkley, J. S.; Gordon, M. S.; Defrees, D. J.;
Pople, J. A. J. Chem. Phys. 1982, 77, 3654-3665.
(39) Hehre, W. J.; Ditchfie.R; Pople, J. A. J. Chem. Phys. 1972, 56, 2257-2261.
(40) Hoffert, W. A.; Rappe, A. K.; Shores, M. P. J. Am. Chem. Soc. 2011, 133, 20823-20836.
(41) Stratmann, R. E.; Scuseria, G. E.; Frisch, M. J. J. Chem. Phys. 1998, 109, 8218-8224.
(42) Soda, T.; Kitagawa, Y.; Onishi, T.; Takano, Y.; Shigeta, Y.; Nagao, H.; Yoshioka, Y.;
Yamaguchi, K. Chem. Phys. Lett. 2000, 319, 223-230.
(43) Noodleman, L.; Davidson, E. R. Chem. Phys. 1986, 109, 131-143.
(44) Isaacs, M.; Sykes, A. G.; Ronco, S. Inorg. Chim. Acta 2006, 359, 3847-3854.
(45) Vidyaratne, I.; Scott, J.; Gambarotta, S.; Budzelaar, P. H. M. Inorg. Chem. 2007, 46,
7040-7049.
(46) Zhang, W. J.; Sun, W. H.; Tang, X. B.; Gao, T. L.; Zhang, S.; Hao, P.; Chen, J. T. J. Mol.
Catal. A - Chem. 2007, 265, 159-166.
(47) Amolegbe, S. A.; Asma, M.; Zhang, M.; Li, G.; Sun, W. H. Aust. J. Chem. 2008, 61, 397-
403.
(48) u, C. C. De eer eorge, S. Weyherm ller, T.; Bill, E.; Bothe, E.; Wieghardt, K. E.;
Wieghardt, K. Angew. Chem. Int. Ed. 2008, 47, 6384-6387.
(49) Lu, C. C.; Bill, E.; Weyhermuller, T.; Bothe, E.; Wieghardt, K. J. Am. Chem. Soc. 2008,
130, 3181-3197.
(50) Ryu, C. K.; Endicott, J. F. Inorg. Chem. 1988, 27, 2203-2214.
(51) Milder, S. J.; Gold, J. S.; Kliger, D. S. Inorg. Chem. 1990, 29, 2506-2511.
(52) Konig, E.; Herzog, S. J. Inorg. Nucl. Chem. 1970, 32, 585-599.
(53) Wamser, C. A. J. Am. Chem. Soc. 1948, 70, 1209-1215.
106
(54) Freire, M. G.; Neves, C. M. S. S.; Marrucho, I. M.; Coutinho, J. A. P.; Fernandes, A. M. J.
Phys. Chem. A 2009, 114, 3744-3749.
(55) Kane-Maguire, N. A. P.; Conway, J.; Langford, C. H. J. Chem. Soc., Chem. Commun.
1974, 801-802.
(56) Kirk, A. D. Chem. Rev. 1999, 99, 1607-1640.
(57) Forster, L. S. Coord. Chem. Rev. 2002, 227, 59-92.
(58) Perkovic, M. W.; Heeg, M. J.; Endicott, J. F. Inorg. Chem. 1991, 30, 3140-3147.
(59) Endicott, J. F.; Perkovic, M. W.; Heeg, M. J.; Ryu, C. K.; Thompson, D. Electron
Transfer Reactions; American Chemical Society: Washington, DC, 1997; Vol. 253.
(60) Wagenknecht, P. S.; Ford, P. C. Coord. Chem. Rev. 2011, 255, 591-616.
107
CHAPTER 4
Sequential Hole/Electron Transfer from Photoexcited Cadmium Sulfide
Nanorods to Mononuclear Ruthenium Water-Oxidation Catalysts
4.1 Introduction
Using solar photons to drive fuel-generating reactions, such as splitting water into H2 and
O2, will allow for storage of solar energy necessary for on-demand availability.1 Inspired by
natural photosynthesis, our interest is in exploring artificial systems that feature light absorbers
directly coupled with redox catalysts.2 Colloidal semiconductor nanocrystals are attractive light
harvesters because they have tunable absorption spectra and high molar absorptivities (105−10
7
M−1
cm−1
). Their coupling with H+ reduction catalysts has been recently reviewed
2-3 and new
studies continue to be reported.4-5
The resulting hybrid structures are capable of light-driven H2
generation with the use of sacrificial electron donors.
Water oxidation, the other half reaction of water splitting, is a mechanistically
complicated process involving the transfer of multiple electrons and protons and the formation of
O2, especially the first O−O bond.6 Over the last three decades, there has been significant
progress in the discovery of ruthenium complexes that catalyze water oxidation.7-13
Consistent
with the complexity of this reaction, the molecular catalysts operate at turnover frequencies
(TOFs) that are considerably lower than TOFs for H+ reduction.
2,7 For this reason, successful
delivery of photoexcited holes to the catalyst is particularly critical. Understanding the
competition between charge transfer and photophysical carrier deactivation pathways, such as
108
electron-hole recombination, is of paramount importance for the design of nanocrystal-based
water-splitting systems.
In this chapter, the charge transfer dynamics between photoexcited CdS nanorods (NRs)
and the mononuclear water-oxidation catalyst [Ru(deeb)(tpy)Cl](PF6) (deeb = diethyl 2,2’-
bipyridine-4,4’-dicarboxylate, tpy = 2,2’:6’,2”-terpyridine) in methanol is described. The
structure of [Ru(deeb)(tpy)Cl](PF6) is shown in Figure 4.1. The interaction between the two
species results in concentration-dependent quenching of CdS NR photoluminescence (PL) and
has a marked impact on CdS excited state dynamics, as measured by transient absorption (TA)
spectroscopy. We find that there are two distinct charge transfer steps in the hybrid nanocrystal-
catalyst system: hole transfer (HT) followed by electron transfer (ET), both from the
photoexcited CdS NR to [Ru(deeb)(tpy)Cl]+, with the overall result being electron-hole
recombination at the Ru center. The HT occurs on the timescale of 100 ps to 1 ns, while the
subsequent ET occurs in 10−100 ns. The relatively slow rate of recombination exposes
opportunities for diverting the photoexcited CdS electrons via auxiliary electron transfer
processes.
Figure 4.1 Chemical structure of the water-oxidation complex [Ru(deeb)(tpy)Cl](PF6).
109
4.2 Experimental Section
4.2.1 Synthesis of CdS Nanorods
The CdS nanorods were synthesized using a previously reported procedure.14-16
Synthesis
and processing were performed under an inert argon atmosphere at ~620 Torr (the atmospheric
pressure in Boulder, CO). A mixture of 8.54 mmol trioctylphosphine oxide (TOPO; Sigma
Aldrich, ReagentPlus®, 99%), 3.2 mmol n-octadecylphosphonic acid (ODPA; PCI Synthesis),
and 1.61 mmol cadmium oxide (CdO; Sigma Aldrich, ≥ 99.99% trace metals basis) were stirred
under vacuum at 120 ºC and then heated under Ar to 320 ºC for 1 hour. The mixture was then
cooled to 120 ºC, stirred under vacuum for 1 hour, and then heated again under Ar to 320 ºC. To
the mixture, 5.40 mmol tri-n-octylphosphine (TOP; Strem Chemicals, min. 97%), and 3.2 mmol
trioctylphosphine sulfide (TOP:S) were injected. TOP:S was prepared by mixing TOP and
elemental sulfur (Aldrich, 99.998%) in a 1:1 molar ratio in an Ar glovebox and stirring at room
temperature for 48 hours. After TOP:S injection, nanocrystal growth proceeded at 315 ºC for 45
minutes. The reaction mixture was then cooled to 80 ºC, and the nanocrystals were precipitated
using a toluene:acetone (1:2 volume ratio) mixture. The CdS nanocrystals were purified under Ar
through sequential redispersion/precipitation steps using toluene/octylamine/acetone,
chloroform/nonanoic acid/isopropanol, and hexane/isopropanol mixtures. Finally, sequential
precipitation steps using increasing amounts of isopropanol were used to separate the mixture
into fractions with narrower length distribution. The purified nanocrystals were re-dispered and
stored in toluene.
The resulting highly monodisperse nanorods had an average diameter of 4.0 ± 0.4 nm and
an average length of 13.7 ± 2.3 nm (one standard deviation for both dimensions), as determined
110
by measurements of over 200 particles in TEM images. The molar absorptivity (ε) of the CdS
nanorods was determined by correlating absorption spectra with Cd2+
concentrations determined
by elemental analysis (ICP-OES) of acid-digested samples. The estimated value of ε350 nm was
1710 M-1
cm-1
per Cd2+
. The number of Cd2+
per nanorod was estimated from average nanorod
dimensions. For this batch of nanorods, ε350 nm was 6 x 106 M
-1cm
-1.
For solvent compatibility with the ruthenium complex [Ru(deeb)(tpy)Cl](PF6), the
hydrophobic surface-capping ligands were replaced with 3-mercaptopropionic acid (3-MPA)
under Ar following a previously reported procedure.14
First, a 70 mM solution of 3-
mercaptopropionic acid (3-MPA; Sigma Aldrich ≥ 99%) in methanol was prepared. The pH of
the 3-MPA solution was raised to pH 11 with tetramethylammonium hydroxide pentahydrate
(Sigma ≥ 97%). Next, a sample of the original, organic-capped, nanocrystals in toluene was
precipitated using methanol. The precipitated nanocrystals were then vigorously mixed with the
70 mM 3-MPA solution until no longer cloudy. A large amount of toluene was added to the
solution to precipitate the 3-MPA-capped nanocrystals. The resulting particles were collected
and then re-dissolved in HPLC-grade MeOH.
4.2.2 Transmission Electron Microscopy (TEM)
TEM samples were prepared by drop casting from solution onto carbon film, 300 mesh,
copper grids from Electron Microscopy Sciences. Images were obtained using a 100 KV Phillips
CM100 TEM equipped with a bottom-mounted 4 megapixal AMT v600 digital camera. Nanorod
dimensions were obtained using ImageJ software to measure more than 200 nanocrystals.
111
4.2.3 Synthesis of Ru(II) Complexes
[Ru(dcb)(tpy)Cl]Cl (dcb = 4,4’-dicarboxy-2,2’-bipyridine) and [Ru(deeb)(tpy)Cl](PF6)
were synthesized according to published procedures.11,13
[Ru(dcb)(tpy)Cl]Cl was dissolved in
H2O and precipitated by adding HPF6(aq) to afford [Ru(dcb)(tpy)Cl](PF6). ESI(+) MS and 1H-
NMR chemical shifts matched the published values.11,13
ESI(+) MS and 1H-NMR
characterization of [Ru(deeb)(tpy)Cl](PF6) in MeOH did not reveal evidence of Cl ligand
displacement by the solvent.
4.2.4 Coupling of CdS NR to [Ru(deeb)(tpy)Cl]+
To form the CdS-[Ru(deeb)(tpy)Cl]+ hybrid system, HPLC grade MeOH (Sigma-Aldrich)
solutions of CdS NRs and [Ru(deeb)(tpy)Cl](PF6) were mixed, and samples sealed under Ar.
Absorption spectra were recorded in 1 cm path length quartz cuvettes at room temperature with
an Agilent 8453 spectrophotometer equipped with tungsten and deuterium lamps. The
concentration of [Ru(deeb)(tpy)Cl]+ in each solution was determined from the absorbance at 520
nm (ε = 16000 M-1
cm-1
measured in MeOH) where there is no contribution from the NRs. The
CdS NR concentration was determined from the absorbance at 350 nm following subtraction of
the absorbance from [Ru(deeb)(tpy)Cl]+. For samples associated the addition of organic
quenchers, methanol solutions of ascorbate (derived from ascorbic acid (Sigma-Aldrich BioXtra,
≥ 99.0%) using tetramethylammonium hydroxide pentahydrate (Sigma ≥ 97%) to raise the pH)
or methylene blue (MB) hydrate (Aldrich) were added. The ascorbate solution was 1.6 x 10-4
M
and was combined with the nanocrystals in a 1:1700 (CdS NR:ascorbate) ratio. The MB solution
was 3.5 mM and was combined in a 1:200 (CdS NR : MB) ratio.
112
4.2.5 Photoluminescence Spectra
Photoluminescence spectra were obtained at room temperature using a PTI fluorometer
with an Ushio UXL-75XE xenon short arc lamp and a Hamamatsu R928P PMT tube operating at
−1000 V, DC. Samples in 1 cm x 1 cm quartz cuvettes were excited at 360 nm and the emission
from 425 nm to 700 nm was recorded at 90° relative to the excitation. Emission spectra were
corrected for wavelength dependence of the instrument response. A detailed description of the
instrumentation was described previously in Chapter 1. The CdS NR concentration was 0.18 μM.
4.2.6 Ultrafast Transient Absorption (TA) Spectroscopy
The ultrafast (100 fs to 3.3 ns) TA spectrometer used in this study uses an amplified
Ti:sapphire laser (Solstice, Spectra-Physics, 800 nm, 1 kHz, 100 fs, 3.5 mJ/pulse), an optical
parametric amplifier (TOPAS-C, Light Conversion), and the Helios spectrometer (Ultrafast
Systems, LLC). A fraction (1.6 mJ/pulse) of the 800 nm Solstice output was directed to the
TOPAS-C to produce the desired pump wavelength (400 nm in the data described here) for
sample excitation, which was then directed into the Helios. The pump pulse beam waist (~350
μm) and energy (< 10 nJ/pulse) were chosen to maintain a nanocrystal excitation probability
below 0.3 per laser pulse to avoid excitation of multiple electron-hole pairs within the
nanocrystals. The pump pulses were passed through a depolarizer and chopped by a
synchronized chopper to 500 Hz before reaching the sample. Another fraction of the 800 nm
Solstice output (~0.1 mJ/pulse) was guided directly into the Helios for generation of the probe.
Within the spectrometer, a white light continuum of wavelengths including 450−800 nm was
generated using a sapphire plate. This beam was split into a probe and a reference beam. The
113
probe beam was focused into the sample where it was overlapped with the pump beam. The
transmitted probe and reference beams were then focused into optical fibers coupled to
multichannel spectrometers with CMOS sensors with 1 kHz detection rates. The reference signal
is used to correct the probe signal for pulse-to-pulse fluctuations in the white-light continuum.
The time delay between the pump and probe pulses was controlled by a motorized delay stage.
For all transient absorption measurements, the sample was sealed under Ar in a 2 mm quartz
cuvette equipped with a Kontes valve and constantly stirred. CdS NR concentrations were
approximately 0.8 μM. All experiments were conducted at room temperature. The change in
absorbance signal (ΔA) was calculated from the intensities of sequential probe pulses with and
without the pump pulse excitation. The data collection (500 pump shots per time point) was
carried out three consecutive times to ensure no photo-induced changes occurred. The three
traces were then averaged.
4.2.7 Nanosecond-Microsecond Transient Absorption Spectroscopy
The 0.3 ns – 400 μs TA spectrometer used the amplified Ti:sapphire laser and optical
parametric amplifier described above coupled with the Eos spectrometer (Ultrafast Systems,
LLC). The pump beam (400 nm) was depolarized and the power was controlled with neutral
density filters. The pump-probe time delay was controlled by a digital delay generator (CNT-90,
Pendulum Instruments). The white light continuum (400 – 900 nm) for the probe and reference
beams was generated by an external 2 kHz Nd:YAG laser focused into a photonic crystal fiber.
The probe and reference signals were focused into the same detectors as used for the ultrafast TA
114
system. Helios and Eos ΔA kinetic traces were combined using Surface Xplorer Pro by Ultrafast
Systems, LLC.
115
4.3 Results and Discussion
4.3.1 The Design of the CdS NR-[Ru(deeb)(tpy)Cl]+ System
The main design criteria for our model nanocrystal-catalyst system involved: (i) use of
materials with relatively well understood optical and catalytic properties, (ii) the possibility of
forming electronically coupled heterostructures, and (iii) relative energy alignments that would
permit hole transfer from the photoexcited nanocrystal to the catalyst.
CdS, a direct-gap semiconductor with a band gap of 2.4 eV, has valence band (VB) and
conduction band (CB) positions thermodynamically suitable for both water oxidation and
reduction.17
CdS-based nanostructures have been commonly employed in nanocrystal-catalyst
hybrids for H+ reduction.
2-3 In the selection of a water-oxidation catalyst, we took advantage of
recent findings demonstrating that the multiple redox steps required for water oxidation can be
negotiated by mononuclear ruthenium complexes.7,10-13
Species based on the [Ru(bpy)(tpy)Cl]+
(bpy = 2,2’-bipyridine) parent structure have the advantages of relatively straightforward
synthesis and redox potential tunability via ligand functionalization.11
The nanocrystals and catalysts were tailored to enable an interaction with sufficient
physical proximity for charge transfer in a polar medium. The CdS NRs (Figure 4.2) were
surface-functionalized with 3-mercaptopropionate (3-MPA)14
which binds to CdS via the thiolate
group while the negatively charged carboxylate prevents flocculation in polar solvents.18
We
modified the [Ru(bpy)(tpy)Cl]+ parent compound with two potential anchoring groups on the
bpy: carboxylic acid moieties, based on an approach for attachment of Ru(II) tris-bipyridyl
complexes to the surfaces of CdSe quantum dots in organic solvents,19-20
and ester functionalities,
which have been reported to bind to TiO2.21
When the Ru complexes were mixed with CdS NRs,
116
the ester functionalities allowed for considerably stronger quenching of CdS PL than the acid
groups. This suggests a repulsive interaction between deprotonated carboxylic acid groups on
4,4’-dicarboxy-2,2’-bipyridine and the anionic NR surface capped with 3-MPA. In contrast,
[Ru(deeb)(tpy)Cl]+, with an overall positive charge, may be electrostatically attracted to the NR
surface. We note that [Ru(deeb)(tpy)Cl]+ is also an active water-oxidation catalyst with relatively
high turnover numbers initiated by the sacrificial oxidant Ce4+
whose redox potential (1.7 V vs
NHE) is less positive than the valence band edge of CdS.11
Figure 4.2 TEM image of CdS NRs with 4.0 ± 0.4 nm widths and 13.7 ± 2.3 nm lengths (one
standard deviation for both dimensions).
4.3.2 Electronic Absorption
Figure 4.3 shows the UV-vis absorption spectra of the CdS NRs, [Ru(deeb)(tpy)Cl]+, and
their mixture, all in MeOH. The CdS NR spectrum has four distinct absorption bands, the lowest
of which corresponds to the band gap transition at 470 nm (2.64 eV). [Ru(deeb)(tpy)Cl]+ exhibits
a prominent feature that is conveniently located further to the red: a broad absorption band
centered at 520 nm attributed to metal-to-ligand charge-transfer.22
The absorption spectrum of a
mixture containing CdS NRs and [Ru(deeb)(tpy)Cl]+ is a superposition of the spectra of the
117
constituents, indicating that upon mixing, [Ru(deeb)(tpy)Cl]+ was not chemically modified and
CdS NRs were not etched. The mixture was stable to precipitation for at least 24 hours. The
excitation wavelengths we use for photophysical characterization (360 and 400 nm) primarily
excite CdS, with only 4% of absorbed photons exciting [Ru(deeb)(tpy)Cl]+.
Figure 4.3 UV-vis absorption spectra of the CdS NRs (1.8 x 10−7
M), [Ru(deeb)(tpy)Cl]+ (1.3 x
10−5
M), and their mixture with a CdS NR : [Ru(deeb)(tpy)Cl]+ = 1 : 72 molar ratio, all in
methanol.
4.3.3 Energy Levels of CdS NR and [Ru(deeb)(tpy)Cl]+
The positions of the VB and CB of CdS NR and redox potentials of [Ru(deeb)(tpy)Cl]+ is
critical in charge transfer processes. The band edges of CdS NRs, which are quantum confined in
the radial direction, were approximated using the Brus equation.23
The bulk band gap energy (2.5
eV) and the valence band position with respect to vacuum (-6.26 eV) were obtained from
published values.24
The quantum confined band gap was determined from the steady state band
edge absorption edge (2.64 eV), and the valence and conduction band edges were adjusted taking
118
into account the effective masses of the electron (0.2 m0) and the hole (0.7 m0).25
The vacuum
scale was then converted to NHE (-4.4 eV (vacuum) ≈ 0 V NHE). The one-electron oxidation
and reduction potentials of [Ru(deeb)(tpy)Cl]+ in MeCN have been previously reported.
11,26
Although the system is studied in MeOH for this paper, no further modifications to the redox
potentials are made because solvent-based redox potential changes have been observed to be
small for a large and coordinatively-saturated complex such as [Ru(deeb)(tpy)Cl]+.27
The first
oxidation potential (Ru3+/2+
) of [Ru(deeb)(tpy)Cl]+ is 0.7 V less positive than the CdS NR VB
edge, which should then permit hole-transfer to [Ru(deeb)(tpy)Cl]+ following photoexcitation of
the CdS NR. Conversely, the CB edge of the CdS NR is 0.2 V more positive than the 2+/1+
couple of [Ru(deeb)(tpy)Cl]+, which should hinder photoinduced electron transfer from the CdS
NR to the LUMO of [Ru(deeb)(tpy)Cl]+. An energy level diagram for the CdS NR-
[Ru(deeb)(tpy)Cl]+ system is shown in Figure 4.4. The Ru
3+/2+ potential is associated with
oxidation at the Ru metal center, while the Ru2+/+
is associated with ligand reduction.28
On the
basis of the energy level alignment, we expect hole transfer from photoexcited CdS NRs to the
Ru center to be thermodynamically favorable.
119
Figure 4.4 Energy level diagram depicting the band edges of CdS NRs and the redox potentials
of [Ru(deeb)(tpy)Cl]+.
4.3.4 Quenching of CdS NR Photoluminescence by [Ru(deeb)(tpy)Cl]+
The effect of interaction between CdS NRs and [Ru(deeb)(tpy)Cl]+ on CdS
photoluminescence (PL) is shown in Figure 4.5. CdS NRs exhibit two distinct PL features: band
gap emission (λmax = 475 nm) and trap emission, seen as a broad red-shifted feature.15,29
The
combination of a low quantum yield of exciton emission (< 1 %) and very long excited electron
lifetimes (> 100 ns, described below) is an indication of efficient hole trapping.15,29
Thus, we
assign the low-energy trap emission primarily to recombination of a surface-trapped hole with an
electron in the lowest CB level. Immediately upon mixing with [Ru(deeb)(tpy)Cl]+, both the
band gap and trap emission signals were quenched, with the degree of quenching dependent on
the CdS NR : [Ru(deeb)(tpy)Cl]+ ratio. PL spectra of the mixture remained unchanged for at
least 24 h. [Ru(deeb)(tpy)Cl]+ is nonemissive and remains silent in the PL spectra. In control
120
experiments, PL quenching did not occur upon addition of free tpy or deeb ligands to CdS NRs
(Figure 4.6), thus suggesting the importance of the Ru center in the quenching process.
Figure 4.5 PL spectra of CdS NRs with increasing amounts of [Ru(deeb)(tpy)Cl]+ and constant
CdS NR concentration. The excitation wavelength is 360 nm.
Figure 4.6 PL spectra for CdS NR and CdS NR with free ligands tpy, deeb, or with
[Ru(deeb)(tpy)Cl]+. The ratios of CdS NR to quenchers are all 1:50.
121
4.3.5 Differentiating Stern-Volmer and Langmuir Models for CdS NR Interaction with
[Ru(deeb)(tpy)Cl]+
To elucidate the interaction between the CdS NRs and [Ru(deeb)(tpy)Cl]+, we analyzed
the quenching of CdS NR PL by [Ru(deeb)(tpy)Cl]+ (Figure 4.5). We considered two models: a
Stern-Volmer model for dynamic quenching via collisions and a Langmuir model for static
quenching due to adsorbed quenchers. For pure collisional quenching in a homogeneous solution,
a plot of I0/IQ vs [Q] should follow the linear form of the Stern-Volmer equation:
where I0 is the PL intensity of CdS NR without quencher, IQ is the PL intensity of CdS NR with
quencher, KSV is the Stern-Volmer constant, and [Q] is the concentration of quencher. The band
gap and trap emission quenching (I0/IQ) vs [Q] are shown in Figures 4.7a and 4.7b. The data for
the band gap transition are clearly a poor fit to Equation 4.1. The quenching of the trap emission
fits the Stern-Volmer model better. However, we note that the trap emission was significantly
weaker than band gap emission at increased [Ru(deeb)(tpy)Cl]+:CdS NR ratios (Figure 4.5) and
thus there is higher uncertainty in the data in Figure 4.7b. Furthermore, because trap states are
longer lived than band gap states, they may be more susceptible to collisional quenching. Overall,
given the poor fit of the quenching of the band gap transition to the Stern-Volmer model, we
conclude that collisions alone cannot account for the concentration-dependent quenching of CdS
PL.
122
Figure 4.7 Quenching of CdS PL as a function of concentration of [Ru(deeb)(tpy)Cl]+. In (a) and
(b), band gap and trap emission are fit with the Stern-Volmer model for dynamic collisional
quenching, whereas in (c) and (d) the same data is fit with a linear form of the Langmuir
adsorption isotherm. Error bars were determined by comparing three PL spectra taken over a
period of 1 hour for one sample with a CdS:[Ru(deeb)(tpy)Cl]+ ratio of 1:43 and one CdS-only
sample. The error was then propagated in an additive fashion to obtain the total value of ± 5%.
This method overestimates the uncertainty in the measurement.
Next, we consider the Langmuir adsorption model for the CdS-[Ru(deeb)(tpy)Cl]+
interaction. Assuming that quenching is caused by adsorption of molecules, we can write the
Langmuir adsorption isotherm as:
123
where Q is quencher (the adsorbate), θ is the fraction of surface sites occupied by the quencher,
and K is the equilibrium constant for adsorption of [Ru(deeb)(tpy)Cl]+ on CdS NR surface. We
assume that θ is equal to the fraction of PL quenched (∆I/I0). Here, I0 is the PL intensity of CdS
NR with no quencher present, while ∆I (= I0-IQ) is the amount of PL quenched in the presence of
quencher. An additional complication is that, unlike trap emission, band gap emission was not
fully quenched at saturation. We attribute this to a lower quenching efficiency for the band-gap
emission caused by its shorter lifetime compared to the trap emission.30
To account for the
incomplete quenching, we introduce the term, (∆I/I0)max, which is the maximum fractional
quenching observed.31
Equation 4.2 then becomes:
Equation 4.3 is then used for the fit shown in Figure 4.8. For a more direct comparison with the
Stern-Volmer model, Equation 4.3 can be rearranged to give the linear form:
As shown in Figures 4.7c and 4.7d, the dependence of CdS PL quenching on concentration of
[Ru(deeb)(tpy)Cl]+ is fit well with Equation 4.4. For the band gap transition, we find that K = 9.5
x 105 M
−1 and (∆I/I0)max = 0.83, whereas for the trap emission the parameters are K = 1.6 x 10
6
M−1
and (∆I/I0)max = 1.0. Like the difference in the values of (∆I/I0)max mentioned above, the
difference in the values of K may be attributable to different quenching efficiencies for the two
transitions. Given the same number of bound quenchers, more of the trap emission would be
quenched, compared to band gap emission, resulting in a higher apparent value of K. Another
124
way to state this is that K is actually a combination of the equilibrium constant and the
quenching efficiency.
Figure 4.8 Fraction of PL quenched for band gap (475 nm) and trap (700 nm) emission as a
function of [Ru(deeb)(tpy)Cl]+ concentration (bottom axis) and the [Ru(deeb)(tpy)Cl]
+ : CdS NR
ratio (top axis). The lines represent fits to a Langmuir adsorption isotherm, suggesting binding
between CdS NRs and [Ru(deeb)(tpy)Cl]+.
One explanation for the quenching of CdS PL in the presence of [Ru(deeb)(tpy)Cl]+ is
charge transfer. However, since PL intensity depends on the product of electron and hole
populations, quenching does not indicate which of the carriers is involved.19,32
To ascertain the
nature of the charge transfer interaction and elucidate the dynamics of this process, we turned to
transient absorption (TA) spectroscopy over a 100 fs to 1 μs timescale range.
125
4.3.6 Transient Absorption Spectra
TA spectra of the CdS NRs (Figure 4.9a) acquired after 400 nm excitation exhibit a
prominent bleach feature (∼470 nm) that corresponds to state-filling of the band gap transition.
In addition, a rapidly decaying absorption feature is observed at 482 nm, red-shifted from the
exciton bleach feature. This has been attributed to the bi-exciton shift due to hot excitons, and its
decay corresponds to carrier cooling.32
A second absorption feature at higher-energy (~440 nm)
corresponds to higher energy exciton bands.29,32
Because the molar absorptivity of
[Ru(deeb)(tpy)Cl]+ is two orders of magnitude smaller than that of CdS NRs (ε400 nm for CdS NR
= 4 x 106 M
−1cm
−1; ε400 nm = 1 x 10
4 M
−1cm
−1 for [Ru(deeb)(tpy)Cl]
+), the presence of
[Ru(deeb)(tpy)Cl]+ does not change the position or shape of the CdS NRs spectral features nor
add additional features in the TA associated with [Ru(deeb)(tpy)Cl]+ (Figure 4.9b). Transient
absorption spectra of CdS NRs, CdS NRs + [Ru(deeb)(tpy)Cl]+ (1 : 120 molar mixture), and
[Ru(deeb)(tpy)Cl]+ taken 2 ps after excitation with a 400 nm pump is shown in Figure 4.9c.
Under the normal conditions for the TA experiments (pump pulse energy 6 nJ), there is no
transient signal observed from [Ru(deeb)(tpy)Cl]+. When the pulse energy is increased 47-fold to
280 nJ, a sample containing [Ru(deeb)(tpy)Cl]+ only exhibits a bleach ~520 nm.
We also used low pump pulse energies to avoid excitation of multiple electron-hole pairs
per NR. The intensity of the 470 nm bleach feature is proportional to the population of excited
electrons in the lowest-lying CB level of CdS NRs.32
This feature is insensitive to the hole
population because of the higher density of energy levels near the VB edge.32
Thus, single-
wavelength kinetics at 470 nm can be used as a signature for electron dynamics.
126
Figure 4.9 TA spectra of (a) CdS NRs, (b) CdS with [Ru(deeb)(tpy)Cl]+, and (c) CdS NRs, CdS,
and [Ru(deeb)(tpy)Cl]+ in MeOH taken 2 ps after excitation with a 400 nm pump.
127
4.3.7 Transient Absorption Kinetics
The kinetics of the CdS NR band gap bleach in the presence and absence of
[Ru(deeb)(tpy)Cl]+ are shown in Figure 4.10. To facilitate visualization of the dynamics over 7
orders of magnitude in time, we use a linear time axis up to 10 ps, and a logarithmic scale
thereafter. Plots using a linear time axis are shown in Figure 4.11.
Figure 4.10 TA decay kinetics at 470 nm for CdS NRs in the presence and absence of
[Ru(deeb)(tpy)Cl]+ (λpump = 400 nm).
Figure 4.11 The same TA decay kinetics data as in Figure 4.10 plotted using a linear time axis.
128
Following 400 nm excitation, rapid (∼1 ps) electron cooling to the CB edge is observed
as a rise of the bleach signal. The subsequent band gap bleach decay of CdS NRs displays
multiexponential decay kinetics, consistent with previous reports.29,33
The bleach kinetics decay
to baseline in ∼1 μs and exhibit an average lifetime of 160 ns, calculated from a five-exponential
fit (see next section for details). The relatively slow overall electron decay dynamics of CdS NRs
have been attributed to a contribution from the slow recombination of the delocalized CB
electron with the localized, surface-trapped holes.29
Addition of [Ru(deeb)(tpy)Cl]+ in the CdS :
[Ru(deeb)(tpy)Cl]+ ratio of 1 : 12 has an unusual impact on the electron decay kinetics (Figure
4.10). For the first 250 ps, the kinetics of CdS NRs alone and in the presence of
[Ru(deeb)(tpy)Cl]+ are essentially superimposable. Following this, the traces diverge and the
electron lifetime is shortened from 160 to 11 ns. This suggests that after a 250 ps delay,
additional kinetic pathways become available, enabling ET from CdS NR to take place.
The 250 ps delay prior to electron lifetime shortening suggests that an electron acceptor
state must first be created before ET can take place. On the basis of the energy level diagram in
Figure 4.4, we hypothesize two sequential charge transfer steps between photoexcited CdS NRs
and [Ru(deeb)(tpy)Cl]+ (Figure 4.12): HT from CdS NR to [Ru(deeb)(tpy)Cl]
+ either directly to
or terminating with the metal-centered HOMO (oxidizing Ru2+
to Ru3+
), followed by ET out of
the CdS NR CB into the newly created available site in the same orbital (reducing Ru3+
back to
Ru2+
). Holes can transfer to [Ru(deeb)(tpy)Cl]+ from both the VB and trap states, as evidenced
by quenching of emission signals associated with both (Figure 4.5). ET from photoexcited
[Ru(deeb)(tpy)Cl]+ to the CB of CdS would manifest as an additional rise in the bleach signal,
and is not observed. We note that a similar lack of change in the early TA dynamics during HT
was reported for the case of CdSe nanocrystals coupled to Ru(II) tris-bipyridyl complexes.19
129
Figure 4.12 Proposed charge transfer steps between photoexcited CdS NR and
[Ru(deeb)(tpy)Cl]+.
To support our hypothesis and the assignment of processes revealed by the TA data, we
performed a series of TA experiments using molecular hole and electron acceptors mixed with
CdS NRs (Figure 4.13). When ascorbate (Asc), a hole scavenger, is added to CdS NRs, no
change in the early kinetics of the band gap bleach is observed (Figure 4.13, orange trace). This
is consistent with the assignment of the bleach to electrons in the CB, and with the lack of HT
signature in the TA signal.19,34
In contrast, when methylene blue (MB), an electron acceptor, is
added to CdS NRs, the divergence from the CdS NR-only trace is observed after 3 ps (Figure
4.13, blue trace). This is consistent with a previous report of ET from nanocrystals to MB.35
These data indicate that the delayed onset of ET in the CdS NR-[Ru(deeb)(tpy)Cl]+ system is
significantly different from a “pure” ET case, and points to another photoexcited process that
precedes ET. Finally, we consider that within our model, the presence of the hole scavenger Asc
in the CdS NR-[Ru(deeb)(tpy)Cl]+ solution would provide a competing destination for the holes,
decrease the population of Ru3+
, and circumvent the subsequent ET process. Evidence for this is
130
seen in Figure 4.13 (red trace) as the lack of the ET signature for CdS NR + [Ru(deeb)(tpy)Cl]+
+ Asc (i.e., the trace is similar to CdS NR alone and CdS NR with Asc). The oxidized form of
Asc has an absorption peak at 380 nm,36
and its accumulation was observed following this TA
experiment (Figure 4.14), indicating that hole transfer to Asc has taken place. We note that the
stepwise charge transfer behavior, along with the lack of overlap between CdS NR emission and
[Ru(deeb)(tpy)Cl]+ absorption, allows us to rule out energy transfer as the mechanism of the PL
quenching seen in Figure 4.5.
Figure 4.13 TA decay kinetics at 470 nm for CdS NRs alone, and in the presence of the
following: [Ru(deeb)(tpy)Cl]+, the hole scavenger ascorbate (Asc), the electron acceptor
methylene blue (MB), and both [Ru(deeb)(tpy)Cl]+ and Asc. As described in the text, these
kinetic traces are consistent with the charge transfer steps shown in Figure 4.12.
131
Figure 4.14 Steady-state absorption spectra of a sample containing CdS NRs,
[Ru(deeb)(tpy)Cl]+, and ascorbate (Asc) taken before and after TA data collection. The
difference in absorption around 380 nm can be attributed to an increase in the concentration of an
oxidized form of ascorbate.
The dependence of CdS NR band gap bleach decay kinetics on the CdS :
[Ru(deeb)(tpy)Cl]+ ratio, with CdS NR concentration held constant, is shown in Figure 4.15.
Over the ratio range of 1 : 8 to 1 : 94, the onset time for ET decreases from 370 to 90 ps (Table
4.1). At the same time, the average electron lifetime decreases from 44 to 1 ns, and the quantum
efficiency of ET increases from 72% to 99% (Table 4.1). As shown in Figure 4.16, the
dependence of these values on the concentration of [Ru(deeb)(tpy)Cl]+ exhibits saturation
behavior similar to that shown in Figure 4.5 for PL quenching. We do not have enough
information to determine the coverage of CdS NRs with [Ru(deeb)(tpy)Cl]+ under varying
mixing ratios. We can, however, estimate that under low-coverage conditions (on the order 1-10
adsorbed molecules per NR), HT occurs on a 100 ps to 1 ns timescale and subsequent ET occurs
with at least a 10-100 ns lifetime.
132
Figure 4.15 Transient absorption kinetics of CdS NRs with fixed concentration of CdS NRs and
varying CdS NR : [Ru(deeb)(tpy)Cl]+ ratios. The excited state lifetime shortens with increasing
concentration of [Ru(deeb)(tpy)Cl]+. Solid lines are five-exponential fits to the data.
Table 4.1 Results of analysis of TA dynamics displayed in Figure 4.15.
CdS NR : [Ru(deeb)(tpy)Cl]+
Onset of ET
(ps)
<τmeasured>
(5-exp fit)
(ns)
<τmeasured>
(4-exp fit)
(ns)
QEET
(%)
<τET>
(ns)
1 : 0 N/A 160 Does not fit N/A N/A
1 : 8 370 44 43 72 61
1 : 12 250 11 11 92 12
1 : 19 130 4.3 3.9 96 4.5
1 : 60 115 3.4 3.0 97 3.5
1 : 94 90 1.0 1.0 99 1.0
133
Figure 4.16 (a) The onset of ET as a function of [Ru(deeb)(tpy)Cl]+ : CdS NR ratio. (b) QEET
and the excited electron lifetimes plotted versus both concentration of [Ru(deeb)(tpy)Cl]+ and
[Ru(deeb)(tpy)Cl]+ : CdS NR ratio. The three quantities plateau at high ratios. The QEET was fit
with the Langmuir isotherm model (Equation 4.3) and the resulting parameters were K = 4.2 x
105 M
−1 and (∆I/I0)max = 1.0. The difference in the values of K seen in the PL and TA can be
attributed to the difference in CdS NR concentrations between the two experiments.
134
The HT timescale falls within the range of observed values for HT from Cd-chalcogenide
nanocrystals to molecular hole acceptors (5 ps to 50 ns) with a variety of coupling conditions and
relative energy level alignments.2 The ET, on the other hand, is significantly slower than values
previously reported for common electron acceptors such as viologens, MB, and polyaromatic
quinones, which are typically less than 100 ps.2 The relatively slow ET may be due to a
combination of low wave function overlap between the hole localized on Ru3+
and the electron
delocalized in a CdS NR, significantly different electronic couplings for the HT and the ET
pathways, and the very large driving force for ET (∼1.9 eV) placing the process in the Marcus
inverted regime. Further work is needed to elucidate the factors that determine the HT and ET
rates in this system.
4.3.8 Curve Fitting for TA Kinetic Decays and the Calculations of Excited State Lifetimes
The TA kinetics in Figure 4.15 provide information about timescales of both the HT and
ET events between CdS NRs and [Ru(deeb)(tpy)Cl]+. Within our model (Figure 4.12 and related
discussion), HT from CdS NR valence band and hole traps to the HOMO of [Ru(deeb)(tpy)Cl]+
is followed by ET from the CdS NR conduction band into the newly available empty state in
[Ru(deeb)(tpy)Cl]+, resulting in overall electron-hole recombination at the metal center.
The point at which the CdS NR-[Ru(deeb)(tpy)Cl]+ kinetics diverge from the CdS NR
kinetics was defined as the onset of ET (HT must occur before an electron acceptor state
becomes available). The determination of the point of divergence is somewhat subjective, but we
can estimate it by the following method: with a logarithmic x-axis, for each [Ru(deeb)(tpy)Cl]+ :
CdS NR sample, the ns-range ET pathway was fit to a line. The intersection of this line with a
135
line parallel to the 1 ps to 1 ns portion of the CdS NR-only decay was estimated to be the ET
onset time. The results are summarized in Table 4.1 and plotted in Figure 4.16a. Note that the ET
onset times exhibit a saturation behavior similar to that shown by PL quenching (Figure 4.5 and
Figure 4.8).
Information about the dynamics of the ET that follows HT is provided by the kinetics of
the bleach decay. Because the CdS NR band gap decay behavior is multi-exponential, we cannot
form a meaningful physical model attributing each component to a specific process without
additional information. Instead, we focus on values that are relatively insensitive to the number
of fit parameters. The average excited state lifetimes were calculated as,37
where ai and τi are the parameters from a multi-exponential fit. For CdS NR alone, we
determined, by inspection of the residual, that the minimum number of exponentials needed for a
good fit is five. The time-components obtained were at least one order of magnitude apart in time.
For CdS NR-[Ru(deeb)(tpy)Cl]+ samples, the fits to both four and five exponentials were
suitable and they resulted in essentially identical values of <τmeasured> (Table 4.1).
Since ∆A is proportional to the electron population, integral of the decay signal is
proportional to the total population over the decay time. We can calculate the quantum efficiency
of electron transfer (QEET) from the integrated areas under the kinetic decays or their fitting
curves using37
136
Then we can estimate the average lifetime of ET (<τET>) by adapting the expression used for
single-exponential excited state decays, τET = τmeasured / QEET, as
For consistency, we used <τmeasured> values from five-exponential fits, but essentially identical
numbers are obtained from four-exponential fits for the CdS NR-[Ru(deeb)(tpy)Cl]+ samples.
Equation 4.7 likely underestimates τET because longest-lived components likely have highest
efficiencies of ET. Nevertheless, this treatment allows us to determine lower limits for ET
lifetimes. The results of this analysis are summarized in Table 4.1 and plotted in Fig 4.16b.
137
4.4 Conclusions
These results presented in this chapter have significant implications for photochemical
water splitting. Under the conditions of our current experiment, the metal center acts as a
recombination site where each HT event that oxidizes Ru2+
is followed by an ET event that
reduces the metal center. However, the ET timescale is relatively slow. We propose that
additional pathways can be designed to funnel away photoexcited electrons and allow for the
accumulation of multiple holes on the catalyst, thereby facilitating the formation of O−O bond in
O2 during water oxidation. Examples of potential electron destinations include molecular
acceptors and catalysts for H+ reduction.
2 Furthermore, built-in charge separation in the so-called
type-II nanoheterostructures could assist in electron removal from the nanocrystal.2 The
nanocrystal/water-oxidation catalyst hybrid could serve as a unit in a more complicated
photochemical water-splitting architecture. Additionally, we expect that improved understanding
of the binding equilibria between CdS NRs and catalysts will allow us to negotiate the
competition for holes among multiple catalysts on each NR. Finally, we note that efficient
delivery of photoexcited holes to the water-oxidation catalysts may have the added benefit of
preventing nanocrystal photo-oxidation.
In summary, we have described the charge transfer interactions between photoexcited
CdS NRs and the water-oxidation catalyst [Ru(deeb)(tpy)Cl]+. We have found evidence for a
stepwise charge transfer mechanism that involves hole transfer from photoexcited CdS NR to the
HOMO of [Ru(deeb)(tpy)Cl]+, occurring on a 100 ps to 1 ns timescale, followed by electron
transfer from the conduction band of CdS to the same orbital on [Ru(deeb)(tpy)Cl]+, which is
considerably slower at 10−100 ns. The second step could be averted through introduction of
additional electron harvesting pathways.
138
4.5 Acknowledgment on Participation
The synthesis and characterization of CdS nanorods and transient absorption experiments
were done by Professor Gordana Dukovic and Molly B. Wilker at the Department of Chemistry
and Biochemistry, University of Colorado Boulder. I was responsible for the synthesis and
characterization of Ruthenium complexes and photoluminescence quenching experiments.
4.6 References
(1) Lewis, N. S.; Nocera, D. G. P Natl Acad Sci 2006, 103, 15729-15735.
(2) Wilker, M. B.; Schnitzenbaumer, K. J.; Dukovic, G. Isr. J. Chem. 2012, 52, 1002-1015.
(3) Chen, X. B.; Shen, S. H.; Guo, L. J.; Mao, S. S. Chem. Rev. 2010, 110, 6503-6570.
(4) Han, Z. J.; Qiu, F.; Eisenberg, R.; Holland, P. L.; Krauss, T. D. Science 2012, 338, 1321-
1324.
(5) Huang, J.; Mulfort, K. L.; Du, P. W.; Chen, L. X. J. Am. Chem. Soc. 2012, 134, 16472-
16475.
(6) Weinberg, D. R.; Gagliardi, C. J.; Hull, J. F.; Murphy, C. F.; Kent, C. A.; Westlake, B. C.;
Paul, A.; Ess, D. H.; McCafferty, D. G.; Meyer, T. J. Chem. Rev. 2012, 112, 4016-4093.
(7) Concepcion, J. J.; Jurss, J. W.; Brennaman, M. K.; Hoertz, P. G.; Patrocinio, A. O. T.; Iha,
N. Y. M.; Templeton, J. L.; Meyer, T. J. Acc. Chem. Res. 2009, 42, 1954.
(8) Wada, T.; Tsuge, K.; Tanaka, K. Angew. Chem. Int. Ed. 2000, 39, 1479.
(9) Sens, C.; Romero, I.; Rodriguez, M.; Llobet, A.; Parella, T.; Benet-Buchholz, J. J. Am.
Chem. Soc. 2004, 126, 7798-7799.
(10) Zong, R.; Thummel, R. P. J. Am. Chem. Soc. 2005, 127, 12802-12803.
(11) Tseng, H. W.; Zong, R.; Muckerman, J. T.; Thummel, R. Inorg. Chem. 2008, 47, 11763-
11773.
(12) Duan, L. L.; Fischer, A.; Xu, Y. H.; Sun, L. C. J. Am. Chem. Soc. 2009, 131, 10397-+.
(13) Wasylenko, D. J.; Ganesamoorthy, C.; Koivisto, B. D.; Henderson, M. A.; Berlinguette,
C. P. Inorg. Chem. 2010, 49, 2202-2209.
(14) Brown, K. A.; Wilker, M. B.; Boehm, M.; Dukovic, G.; King, P. W. J. Am. Chem. Soc.
2012, 134, 5627-5636.
(15) Peng, P.; Sadtler, B.; Alivisatos, A. P.; Saykally, R. J. J. Phys. Chem. C 2010, 114, 5879-
5885.
(16) Robinson, R. D.; Sadtler, B.; Demchenko, D. O.; Erdonmez, C. K.; Wang, L. W.;
Alivisatos, A. P. Science 2007, 317, 355-358.
139
(17) Nozik, A. J. Annu. Rev. Phys. Chem. 1978, 29, 189-222.
(18) Aldana, J.; Lavelle, N.; Wang, Y. J.; Peng, X. G. J. Am. Chem. Soc. 2005, 127, 2496-
2504.
(19) Sykora, M.; Petruska, M. A.; Alstrum-Acevedo, J.; Bezel, I.; Meyer, T. J.; Klimov, V. I. J.
Am. Chem. Soc. 2006, 128, 9984-9985.
(20) Koposov, A. Y.; Szymanski, P.; Cardolaccia, T.; Meyer, T. J.; Klimov, V. I.; Sykora, M.
Adv. Funct. Mater. 2011, 21, 3159-3168.
(21) Kelly, C. A.; Farzad, F.; Thompson, D. W.; Meyer, G. J. Langmuir 1999, 15, 731-737.
(22) Christopoulos, K.; Karidi, K.; Tsipis, A.; Garoufis, A. Inorg. Chem. Commun. 2008, 11,
1341-1346.
(23) Brus, L. J. Phys. Chem. 1986, 90, 2555.
(24) Van de Walle, C. G.; Neugebauer, J. Nature 2003, 423, 626-628.
(25) Madelung, O. Semiconductors-Basic Data; 2nd ed.; Springer: Berlin, 1996.
(26) Pavlishchuk, V. V.; Addison, A. W. Inorg. Chim. Acta 2000, 298, 97-102.
(27) Connelly, N. G.; Geiger, W. E. Chem. Rev. 1996, 96, 877-910.
(28) Jakubikova, E.; Chen, W. Z.; Dattelbaum, D. M.; Rein, F. N.; Rocha, R. C.; Martin, R. L.;
Batista, E. R. Inorg. Chem. 2009, 48, 10720-10725.
(29) Wu, K. F.; Zhu, H. M.; Liu, Z.; Rodriguez-Cordoba, W.; Lian, T. Q. J. Am. Chem. Soc.
2012, 134, 10337-10340.
(30) Klimov, V. I.; Schwarz, C. J.; McBranch, D. W.; Leatherdale, C. A.; Bawendi, M. G.
Phys. Rev. B 1999, 60, R2177-R2180.
(31) Munro, A. M.; Jen-La Plante, I.; Ng, M. S.; Ginger, D. S. J. Phys. Chem. C 2007, 111,
6220-6227.
(32) Klimov, V. I. Annu. Rev. Phys. Chem. 2007, 58, 635-673.
(33) Knowles, K. E.; McArthur, E. A.; Weiss, E. A. ACS Nano 2011, 5, 2026-2035.
(34) Huang, J. E.; Huang, Z. Q.; Jin, S. Y.; Lian, T. Q. J. Phys. Chem. C 2008, 112, 19734-
19738.
(35) Huang, J.; Huang, Z. Q.; Yang, Y.; Zhu, H. M.; Lian, T. Q. J. Am. Chem. Soc. 2010, 132,
4858-4864.
(36) Warren, J. J.; Mayer, J. M. J. Am. Chem. Soc. 2010, 132, 7784-7793.
(37) Lakowicz, J. R. Principles of Fluorescence Spectroscopy; Springer: New York, 2006.