A DISSERTATION SUBMITTED TO - Stackstd264yn6201/PhD...2010/11/29 · and create simple, memorable...
Transcript of A DISSERTATION SUBMITTED TO - Stackstd264yn6201/PhD...2010/11/29 · and create simple, memorable...
INCREASED LIGHT HARVESTING IN DYE-SENSITIZED SOLAR CELLS
USING FÖRSTER RESONANT ENERGY TRANSFER
A DISSERTATION
SUBMITTED TO
THE DEPARTMENT OF MATERIALS SCIENCE AND ENGINEERING
AND THE COMMITTEE ON GRADUATE STUDIES
OF STANFORD UNIVERSITY
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
Brian Eugene Hardin
November 2010
http://creativecommons.org/licenses/by-nc/3.0/us/
This dissertation is online at: http://purl.stanford.edu/td264yn6201
© 2011 by Brian E. Hardin. All Rights Reserved.
Re-distributed by Stanford University under license with the author.
This work is licensed under a Creative Commons Attribution-Noncommercial 3.0 United States License.
ii
I certify that I have read this dissertation and that, in my opinion, it is fully adequatein scope and quality as a dissertation for the degree of Doctor of Philosophy.
Michael McGehee, Primary Adviser
I certify that I have read this dissertation and that, in my opinion, it is fully adequatein scope and quality as a dissertation for the degree of Doctor of Philosophy.
Yi Cui
I certify that I have read this dissertation and that, in my opinion, it is fully adequatein scope and quality as a dissertation for the degree of Doctor of Philosophy.
Peter Peumans
Approved for the Stanford University Committee on Graduate Studies.
Patricia J. Gumport, Vice Provost Graduate Education
This signature page was generated electronically upon submission of this dissertation in electronic format. An original signed hard copy of the signature page is on file inUniversity Archives.
iii
iv
Abstract
Dye-sensitized solar cells (DSCs) are an emerging photovoltaic technology
with the potential for large scale manufacturing and low cost processing. However, the
power conversion efficiency of DSCs must increase from 11% to 14% to be
commercially competitive with conventional solar cell technologies. DSCs do not
completely absorb all of the photons from the visible and near infrared portion of the
solar spectrum and consequently have lower short circuit photocurrent densities
compared to inorganic photovoltaic devices. A variety of sensitizing dyes have been
explored, but it is extremely challenging to develop a single sensitizing dye that can
absorb strongly in the visible and near-infrared spectrum.
The focus of this doctoral thesis is on developing fundamentally new DSC
architectures which incorporate energy transfer processes in order to improve light
harvesting. Chapter One will introduce the conventional dye-sensitized solar cell
architecture and general operating principles for photocurrent generation. Chapter
Two will focus on the general theory behind Forster Resonant Energy Transfer (FRET)
and modeling of the average excitation transfer efficiency (ETE), which is the fraction
of excited dyes that undergo energy transfer to the sensitizing dye, inside of the DSC.
Chapter Three describes a new design where energy relay dyes unattached to the
titania absorb high energy photons and transfer their energy to the sensitizing dye via
Förster resonant energy transfer. This architecture allows for stronger and broader
spectral absorption for the same film thickness. In liquid DSCs, we demonstrate a 26%
increase in power conversion efficiency when using an energy relay dye with an
organic, near-infrared sensitizing dye and show that the average excitation transfer can
be greater than 95%. Chapter Four demonstrates that energy relay dyes can be mixed
inside of a solid, organic hole conductor (e.g. spiro-OMeTAD) for solid-state DSCs.
Chapter Five describes the concept of using energy relay dyes, cosensitized on
the TiO2 surface, that directly absorb near-infrared light and undergo energy transfer
to a neighboring a Ruthenium based metal ligand complex (i.e. C106). Near-infrared
energy relay dyes have the potential to increase light harvesting in the 700-800 nm
portion of the spectrum and can be implemented in state-of-the-art DSC systems. The
v
final chapter will briefly describe the opportunities for future study and potential
commercialization of DSCs.
vi
Acknowledgements
In a quest to understand the science behind photovoltaic (PV) cells I have
studied at four universities in three different countries over the past ten years. I can
think of no better way to spend my twenties than working in an area that is very
important to me with people that I care about. I would like to acknowledge the people
that have been most influential in my engineering education.
My initial interest in PV was sparked by my High School State Science Fair
project that I worked on with my father, Larry. We spent several weeks designing a
portable solar power air conditioner for my 1972 Triumph Spitfire. My dad was my
first and most influential engineering mentor. For years he taught me how to rationally
diagnose car problems and repair them when my Triumph broke down on a weekly
basis.
My first solar research experience came while I was studying abroad at St.
Edmunds Hall at the University of Oxford. I spent seven months working on various
solar projects with Professor Malcolm McCulloch; who introduced me to dye-
sensitized solar cells. When I returned home my undergraduate advisor, Professor
John Goodenough, at the Univesrity of Texas at Austin supported me financially and
provided lab space to pursue my own solar research projects.
The foundation of my thesis work is largely based on knowledge that I gained
as a Fulbright scholar at Ecole Polytechnique Fédérale de Lausanne under the
supervision of Professor Michael Grätzel, who has been like an academic grandfather
to me. Michael has been very supportive; sending one-of-a-kind sensitizing dyes for
the energy relay dye projects and allowing me to return to Switzerland to complete my
work. I am also very grateful to the EPFL team, specifically Jun-Ho Yum, Shaik
Zakeeruddin, Md. K Nazeeruddin, Pascal Comte, Thomas Moehl, Jaques Moser,
Robin Humphrey-Baker, and Takeru Bessho, for assisting me in creating and
analyzing many of the solar cells described in this work.
I would like to acknowledge my PhD advisor, Mike McGehee, who challenged
me to focus only on high impact ideas. He has continually pushed me to think about
the broader impacts of my scientific work and changed how I approach solar power
vii
research and development. Mike has also taught me how to write scientific proposals
and create simple, memorable presentations.
I would like to acknowledge the previous work of two former McGehee group
members, Shawn Scully and Yuxiang Liu, who introduced me to the idea of using
Förster resonant energy transfer in organic photovoltaic cells. My thesis is largely a
derivative of knowledge about dye-sensitized solar cells that I learned in Switzerland
and energy transfer that I learned at Stanford. Eric Hoke has been an excellent
collaborator; performing most of the modeling presented in the second chapter and has
been an invaluable proof reader of all of my papers. I have been very fortunate to team
up with I-Kang Ding, who has helped build an excellent DSC fabrication set up at
Stanford. The McGehee team has become a very tight group of friends and I am sad to
leave them. I have spent over four years with Craig Peters and Mike Rowell talking
about solar power and playing volleyball. Nicky Cates Miller, George Burkhard, Jason
Bloking, Zach Beiley, Toby Sachs-Quintana, George Margulis, Sean Sweetnam, Jack
Parmer, Sam Rosenthal, Julia Zaks, Steve Shelton, Tomas Leijtens, Red Ransil, Joe
Kline, and Vignesh Gowrishankar, whose power point presentation template I have
used for five years, have made our office fun and have built a culture of success in
research. I also thank my friends, both in and out of the Materials Science department,
for your support and friendships, and for making the past five years enjoyable.
I am grateful to Paul Armstrong, John Anthony, Xu Han, and Alan Sellinger
for synthesizing several energy relay dyes used in my projects. I would like to
acknowledge my thesis committee: Peter Peumans, Yi Cui, Zhenan Bao, and Alberto
Salleo. I have worked with each of your groups on solar projects at some point during
my PhD and I appreciate all of your insights and friendship.
Finally, I would like to acknowledge two important women in my life: my
mom and girlfriend, Maja. My mom, Dannette, has passed on to me her boundless
energy and positive attitude about work. I have never seen anyone as happy working
on the weekends and getting things done. Maja has been with me for the last five years;
freeing me from the Stanford science bubble with regular furloughs to San Francisco
and New York City. I also want to thank my brother, Alex, and twin sister, Lauren, as
viii
well as my extended family for their enduring support and encouragement. Even
though I have been away for the last six year you always make me feel like I’ve never
left Texas and that means a lot to me.
ix
Dedication
I dedicate this thesis to the Great State of Texas whose natural beauty and
wildlife made me an ardent environmentalist as a child and to my parents, Larry and
Danette, who taught me to follow my dreams wherever they might take me.
'Hope cherishes the soul of him who lives in justice and holiness, and is the nurse of
his age and the companion of his journey; hope which is mightiest to sway the restless
soul of man.' ~ Pindar
x
Table of Contents
Abstract ........................................................................................................................ iv Acknowledgements ...................................................................................................... vi Dedication ..................................................................................................................... ix List of Tables ............................................................................................................... xv List of Figures ............................................................................................................ xvi 1 Dye-Sensitized Solar Cells (DSCs) ...................................................................... 1
1.1 Dye-Sensitized Solar Cells Architecture ........................................................ 1
1.2 Power Conversion Efficiency ......................................................................... 2
1.3 External Quantum Efficiency ......................................................................... 4
1.4 Photocurrent Generation in Dye-Sensitized Solar Cells ................................ 5
1.5 Light Harvesting in DSCs .............................................................................. 6
1.6 Charge Transfer Processes in DSCs ............................................................... 7
1.6.1 Electron Separation Probability in Liquid DSCs .................................... 9
1.6.2 Hole Separation Probability in Liquid DSCs ....................................... 10
1.6.3 Charge Collection Efficiency in DSCs ................................................. 12
1.7 Determination of Open-Circuit Voltage ....................................................... 14
1.8 Estimating the Maximum Obtainable Power Conversion Efficiency of
Liquid based DSCs43 ................................................................................................ 14
1.9 Improving the Power Conversion Efficiency of Liquid Based DSCs .......... 15
2 Förster Resonant Energy Transfer (FRET) ..................................................... 17 2.1 FRET Radius (R0) ......................................................................................... 17
2.2 Potential For FRET in Dye-Sensitized Solar Cells ...................................... 18
2.3 Modeling the Efficiency of Förster Resonant Energy Transfer from Energy
Relay Dyes in Dye-Sensitized Solar Cells ............................................................... 19
2.3.1 The Importance of the Average Excitation Transfer Efficiency, ETE . 19
xi
2.3.2 Introdouction to Modeling FRET in DSCs .......................................... 20
2.3.3 Background and Model Description ..................................................... 22
2.3.4 Effects of Electrolyte Quenching ......................................................... 24
2.3.5 The Critical Radius (RC) ....................................................................... 24
2.3.6 Short lifetime relay dyes: Excitation transfer efficiency in the absence
of diffusion ........................................................................................................... 25
2.3.7 Long lifetime relay dyes: Excitation transfer efficiency in the rapid
diffusion limit ....................................................................................................... 28
2.3.8 Intermediate lifetime dyes: Full model of the impact of relay dye
lifetime on ETE .................................................................................................... 31
2.3.9 Discussion ............................................................................................. 34
2.3.10 Conclusion ............................................................................................ 35
2.4 Measuring Important Energy Transfer Parameters ...................................... 35
2.4.1 Measuring the FRET R0 of Fast Emitting Chromophores in Solution 36
2.4.2 Measuring the FRET R0 of Diffusive Chromophores in Solution ....... 37
2.4.3 Dynamic Quenching Theory and Measurements ................................. 38
2.4.4 Pore Size Distribution in Titania Mesostructured Electrodes .............. 40
2.4.5 Surface Concentration of Sensitizing Dyes in DSC ............................. 41
3 Using Energy Relay Dyes Unattached to Titania in Liquid DSCs ................. 42 3.1 The PTCDI/TT1 System ............................................................................... 43
3.1.1 PTCDI/TT1 Emission and Absorption Spectra .................................... 43
3.1.2 PTCDI Quenching by the Electrolyte ................................................... 44
3.1.3 Modeling ETE in the PTCDI/TT1 DSC System .................................. 45
xii
3.1.4 PTCDI/TT1 Device Fabrication and Performance ............................... 46
3.1.5 Minimum Bound ETE for PTCDI/TT1 System ................................... 48
3.1.6 PTCDI/TT1 Conclusions ...................................................................... 50
3.1.7 PTCDI Synthesis .................................................................................. 50
3.2 The DCM/TT1 System ................................................................................. 52
3.2.1 DCM/TT1 Emission Absorption Spectra ............................................. 53
3.2.2 DCM Quenching by the Electrolyte ..................................................... 53
3.2.3 Modeling ETE in DCM/TT1 System ................................................... 55
3.2.4 DCM/TT1 Device Fabrication and Performance ................................. 56
3.2.5 DCM/TT1 Conclusion .......................................................................... 58
3.3 Directly Measuring the Excitation Transfer Efficiency in Liquid Based
DSCs 58
3.3.1 Measuring the Internal Quantum Efficiency ........................................ 59
3.3.2 Measuring the EQE contribution from ERD ........................................ 59
3.3.3 Measuring Light Absorption by ERD .................................................. 60
3.3.4 Excitation Transfer Calculations for DCM/TT1 System ...................... 63
3.3.5 Effects of Direct ERD electron transfer in ERD/DSC System ............. 63
3.3.6 Optical Losses Related to the FTO Front Contact of the DSC ............. 65
3.4 Future Outlook .............................................................................................. 66
3.4.1 Near-Infrared Sensitizing Dyes ............................................................ 66
3.4.2 Energy Relay Dyes in Liquid DSCs ..................................................... 68
3.4.3 Organic Dye Alternatives with ERDs .................................................. 69
3.5 Experimental Methods .................................................................................. 71
xiii
4 Using Energy Relay Dyes Unattached to Titania in solid-state Dye-Sensitized Solar Cells .................................................................................................................... 72
4.1 ERD Design Rules for solid-state DSCs ...................................................... 73
4.2 N877/SQ1 solid-state DSC System .............................................................. 74
4.2.1 N877/SQ1 Emission and Absorption Spectra ...................................... 74
4.2.2 N877 quenching by spiro-OMeTAD .................................................... 76
4.2.3 N877/SQ1 DSC Fabrication and Device Performance ........................ 77
4.2.4 N877/SQ1 ETE Estimate ...................................................................... 78
4.2.5 N877 Synthesis ..................................................................................... 79
4.2.6 N877/SQ1 Testing Methods ................................................................. 80
4.2.7 N877/SQ1 ss-DSC Conclusions ........................................................... 80
5 Using Near-Infrared Energy Relay Dyes Co-sensitized with Metal-Ligand Dyes to Increase Light Harvesting ............................................................................ 82
5.1 Near-Infrared Dye Design Rules .................................................................. 82
5.2 AS02/C106 DSC System .............................................................................. 83
5.2.1 AS02/C106 Absorption and Emission Spectra ..................................... 84
5.2.2 AS02 and C106 Charge Transfer Kinetics ........................................... 85
5.2.3 AS02/C106 Excitation Transfer Modeling ........................................... 89
5.2.4 AS02/C106 Fractional Surface Coverage and Dye Loading ................ 92
5.2.5 AS02/C106 Device Fabrication and External Quantum Efficiency
Results 94
5.2.6 Measuring the Average Excitation Transfer Efficiency of AS02/C106
System 96
5.2.7 Hole Transfer from C106 to AS02 ....................................................... 96
5.2.8 Effects of Intermolecular Hole Transfer in AS02/C106 System .......... 98
xiv
5.2.9 AS02 Synthesis ................................................................................... 101
5.3 NIR-ERD Conclusion ................................................................................. 103
6 Conclusion ......................................................................................................... 105 6.1 Future Outlook of Energy Transfer in Dye-Sensitized Solar Cells ............ 105
6.2 Commercialization Potential of Dye-Sensitized Solar Cells ...................... 107
Bibliography and References ................................................................................... 109
xv
List of Tables
Table 1-1: PV Characteristics of Top Performing Sensitizing Dyes. ............................. 4
Table 1-2: List of Rate Lifetime Ranges for Charge Transfer Processes in DSCs under
short-circuit current conditions. .................................................................................. 9
Table 3-1 Average Excitation Transfer Efficiency estimates based on measured values
for 5.5mM, 11mM, and 22mM concentrations of DCM. ......................................... 63
Table 4-1. J-V characteristics of SQ1 ss-DSCs without and with N877. .................... 79
Table 5-1: Energy and Charge Transfer Lifetimes for AS02 and C106 ....................... 91
Table 5-2: Dipping time versus total surface coverage and fraction of dyes on 6.5µm
thick transparent titania films ................................................................................... 93
xvi
List of Figures
Figure 1-1. Dye-sensitized solar cell schematic; inset is enhanced view of TiO2
nanoparticles covered by the sensitizing dye. ............................................................ 2
Figure 1-2. Photocurrent Density versus Voltage plot of highly efficient N719 based
DSC. ........................................................................................................................... 3
Figure 1-3. Chemical Structure of state-of-the-art sensitizing dyes. .............................. 3
Figure 1-4. The external quantum efficiency of N719 based DSC. ............................... 5
Figure 1-5. Photocurrent generation schematic of DSC. ................................................ 6
Figure 1-6. Jablonski Plot of Charge Transfer Rates in DSCs ....................................... 8
Figure 1-7. AM 1.5G Solar Spectrum .......................................................................... 16
Figure 2-1. DSC schematic representation of a dye-sensitized solar cell with energy
relay dyes. The right side of the figure shows the typical absorption process for
lower energy (red) photons in DSCs: light is absorbed by the sensitizing dye (1),
transferring an electron into the titania and a hole is transported to the back contact
through the electrolyte. The energy relay dye process is similar except that, higher
energy (blue) photons are first absorbed by the energy relay dye that undergoes
Förster energy transfer (2) to the sensitizing dye. .................................................... 18
Figure 2-2. (a) Geometries of FRET occurring from a single donor to a single acceptor
and (b) from donors to a dense monolayer of acceptors with surface concentration
CA as in the case of a dye sensitized solar cell with relay dyes. ............................... 23
Figure 2-3. Geometries of the cylindrical (a) and spherical (b) pores of diameter 2Rp.
The relay dye is distributed throughout the volume of the interior of the pore while
the sensitizing dye densely covers the pore walls. (c) Calculated excitation transfer
efficiency in cylindrical (dotted curve) and spherical (solid curve) pores in the
absence of diffusion as a function of the ratio of the critical energy transfer distance
Rc to the pore diameter 2Rp. ..................................................................................... 27
Figure 2-4. Excitation transfer efficiency in (a) cylindrical and (b) spherical pores in
the rapid diffusion limit as a function of the critical energy transfer distance, Rc and
xvii
the distance of closest approach that the donors can be from the pore wall, Ra. The
pore diameter was assumed to be 2Rp = 30nm. To determine the excitation
efficiency for other pore sizes, scale Rc and Ra by the same proportionality factor
that Rp is changed. .................................................................................................... 30
Figure 2-5. Excitation transfer efficiency for a spherical pore (a) with significant
quenching (kq[Q] = 109s-1) and (b) reduced quenching (kq[Q] = 106s-1) as a function
of the relay dye lifetime and critical energy transfer distance in absence of
quenching. The pore diameter was set to 2Rp = 30nm, the distance of closest
approach was Ra = 0.5nm, and a relay dye diffusivity of D = 0.6nm2/ns was used in
these calculations. ..................................................................................................... 33
Figure 2-6. Time resolved photoluminescence of PTCDI with varying concentration of
TT1 in gamma butyrolactrone. PTCDI concentration was 10-4M. ........................... 37
Figure 2-7. Photoluminescence (640 nm) decay of N877 in presence of SQ1 as
function of time. Experimental result was fitted with the diffusion model .............. 38
Figure 2-8. BET Data for various TiO2 pastes. The inset shows the porosity (‘por’) and
roughness factor (‘RF’). ........................................................................................... 40
Figure 2-9. Pore Volume distribution for various TiO2 nanoparticle pastes. ............... 41
Figure 3-1. PTCDI and TT1 properties. a, PTCDI absorption (blue), PTCDI emission
(red dash dot) in chloroform, and TT1 absorption (black) on titania nanoparticles.
Chemical structures of the energy relay dye, PTCDI (b), and sensitizing dye, TT1
(c). ............................................................................................................................. 44
Figure 3-2. Quenching of PTCDI by electrolyte species. The PTCDI
photoluminescence is reduced with increasing concentration of PMII (half-filled
blue circles) and I2 (green squares). The reduction in photoluminescence (PL0/PL)
by PMII is equivalent to the reduction in excitation lifetime (τ0/τ) shown as the red
triangles. The PTCDI concentration was 1*10-4M in gamma-butyrolactone. ......... 45
Figure 3-3 Modeled average excitation transfer efficiency as a function of pore
diameter for spherical and cylindrical pores. Modeling results are based on a
Förster radius of 8.0 nm, conservative dye coverage estimate of 0.2 dye nm-2, and a
quenching rate of 30k0. ............................................................................................. 46
xviii
Figure 3-4. Photocurrent density-voltage (J-V) characteristics of devices with (13mM
PTCDI) and without (0mM PTCDI) energy relay dye under AMA 1.5 (100mWcm-
2). Dash-dot lines represent the dark current for ERD containing DSC (blue) and the
control device (green). .............................................................................................. 48
Figure 3-5. Light harvesting characteristics of the ERD DSC. a, External quantum
efficiency versus wavelength of DSC with energy relay dye (PTCDI) and a control
device (0mM PTCDI). b, EQE addition (black squares) caused by FRET from the
energy relay dye to sensitizing dye and PTCDI absorption (blue circles). Peak
ΔEQE generated by PTCDI was 29.5% at 530nm. .................................................. 49
Figure 3-6. Estimated EQE of PTCDI/TT1 system. ..................................................... 50
Figure 3-7. (a) Absorption (blue) and Emission (red dash-dot) spectra of DCM energy
relay dyes in acetonitrile:valeronitrile (85:15 vol) with TT1 absorption spectra on
TiO2 (green). Chemical structures of DCM (b) and TT1 (c). ................................... 53
Figure 3-8. (a) Photoluminescence lifetime of DCM with various concentrations of
ERD using an 85:15 mixture by volume of acetonitrile and valeronitrile. (b)
Photoluminescence quenching caused by various concentrations of M1 electrolyte
.................................................................................................................................. 55
Figure 3-9. Excitation Transfer Efficiency for DCM as the ERD for various pore
geometries. ................................................................................................................ 56
Figure 3-10. (a) The EQE of the 22mM DCM DSC with 8+4µm architecture using an
acetonitrile based electrolyte. (b) Photocurrent density-voltage (JV) characteristics
of devices with (22 mM DCM) and without (0 mM DCM) energy relay dye under
AM 1.5G (100 mW/cm2). Dashed lines represent the dark current for ERD
containing and control devices. ................................................................................ 57
Figure 3-11. (a) External Quantum Efficiency of DSC of transparent TiO2 electrodes
(5.4µm, 17 nm particles) with varying concentrations of DCM. (b) Change in the
External Quantum Efficiency compared to control (0mM) versus DCM
concentration. ........................................................................................................... 59
Figure 3-12. (a). Schematic of the ERD measurement to determine the amount of light
absorbed by the ERD inside of the TiO2 pores. (b) EQEERD (black circles) and
xix
ηabs,ERD (red squares) versus predicted ERD concentration. (d) Average excitation
transfer efficiency ( ETE ) versus concentration; the ETE average over three
concentrations is 96%. .............................................................................................. 62
Figure 3-13. External Quantum Efficiency for 5.4um thick transparent films covered
in chenodeoxylic acid with varying concentrations DCM. ...................................... 65
Figure 3-14. Absorption Spectra of FTO and FTO covered with 17nm particles. ....... 66
Figure 3-15. Near-Infrared Sensitizing Dyes successfully incorporated into DSCs, a)
NK6037 and b) Si naphthalocyanine compound with axial anchoring group. ......... 67
Figure 3-16. Figure of Light Harvesting versus Molar extinction coefficient and ERD
solubility ................................................................................................................... 69
Figure 3-17 EQE spectrum of YD2. ............................................................................. 71
Figure 4-1. Operating mechanisms of ss-DSC. 1) Lower energy (magenta) photons
are absorbed by the sensitizing dye (SQ1), transferring an electron into the TiO2 and
hole into the electrolyte. Higher energy (blue) photons are absorbed by the energy
relay dye (N877) and either 2) Förster energy transferred into the sensitizing dye.
Figure not drawn to scale. ........................................................................................ 73
Figure 4-2. PL Quenching versus HOMO leve of various ADT and Pentacene
derivatives. ................................................................................................................ 74
Figure 4-3. Normalized UV/Vis absorption (solid line)/emission (dash line) spectra of
SQ1 (blue) and N877 (red) in ethanol, respectively ................................................. 75
Figure 4-4. Photoluminescence spectra of 1.5% wt N877 in polystyrene versus spiro-
OMeTAD, corrected for absorption. ........................................................................ 77
Figure 4-5. EQE spectrum of SQ1 SSDSCs with and without ERD, N877. The gray
line is an IPCE spectra of only Spiro-OMeTAD and the N877 energy transfer relay.
The black line is only SQ1 and Spiro-OMeTAD. The red line is SQ1 + N877 +
Spiro-OMeTAD. ....................................................................................................... 78
Figure 5-1. The NIR dye attached to the titania surface absorbs near-infrared photons
and uses short range energy transfer to excite a neighboring sensitizing dye, which
is responsible for electron transfer into the TiO2 (kinj) and hole regeneration with the
electrolyte (kreg). ....................................................................................................... 83
xx
Figure 5-2.Absorption and Emission spectra of the sensitizing dye, C106, and near-
infrared dye, AS02 in DMF. The chemical structure of C106 and AS02 are shown
in the inset. ................................................................................................................ 85
Figure 5-3. The Time resolved photoluminescence decay of AS02 in DMF solution
(10-5M) , on Al203, and on TiO2. .............................................................................. 86
Figure 5-4. The Time resolved photoluminescence decay of C106 on Al203. ............ 87
Figure 5-5. Temporal profiles of the transient absorbance measured at = 800 nm
upon pulsed laser excitation (= 550 nm, 5 ns full width half-maximum pulse
duration, 30 Hz repetition rate) on samples comprised of C106 dye adsorbed on
nanocrystalline TiO2 films in the presence (red trace) and in the absence (blue trace)
of the redox-active electrolyte. ................................................................................. 88
Figure 5-6. Jablonski Plot of AS02 + C106 DSC system. The scheme is not
geometrically correct (i.e. both dyes should be on the same TiO2 surface), processes
that result in photocurrent generation are labeled in black; while processes that do
not contribute to photocurrent are labeled in grey; dashed lines represent
intermolecular processes. ......................................................................................... 91
Figure 5-7. Optical density versus wavelength for AS02 only (green line) and AS02 +
C106 (black line) dyed 5.6µm thick TiO2 films compared to C106 control device
(red dashed line). ...................................................................................................... 93
Figure 5-8. (A) Optical density versus wavelength of titania films sensitized with
C106, AS02 + C106, and AS02 only. (B) External quantum efficiency versus
wavelength of C106, AS02 + C106, and AS02 only dye-sensitized solar cells. All
films were approximately 6.5µm thick and comprised of transparent 17-nm-
diameter TiO2 nanoparticles ..................................................................................... 95
Figure 5-9. (A) Photo-induced transient absorption spectra of C106 (red dash),
C106+AS02 (black), and AS02 (green) on TiO2. PIA signals were normalized to
light absorption at 470nm. ........................................................................................ 97
Figure 5-10. Optical Density of 1*10-5M AS02 in DMF with various concentration of
NOBF4. ..................................................................................................................... 98
xxi
Figure 5-11. Electron lifetime versus conductivity for DSC systems with various
concentrations of AS02 and C106 on TiO2. ........................................................... 101
1
1 Dye-Sensitized Solar Cells (DSCs)
Dye-sensitized solar cells (DSCs) are based on light harvesting by a sensitizing
dye attached to a wide band gap semiconductor.1-5 DSCs are made mainly of abundant,
non-toxic materials and offer an inexpensive route to develop highly efficient
photovoltaic cells. State-of-the-art DSCs based on iodide/triiodide redox couple in a
liquid electrolyte have validated power conversion efficiencies of over 11%6-8 while
DSCs comprised of a solid-state hole conductor have achieved power conversion
efficiencies of over 5%.9 Several companies have recently begun to manufacture DSC
for residential installations as well as consumer electronics. In 2010, Sony announced
that it had created a DSC module that has achieved a power conversion efficiency over
9%.
Liquid DSC conversion efficiencies must reach their theoretical maximum of
14-15% in order to gain wide spread commercial adoption. The purpose of this chapter
is to provide a cursory overview of the conventional DSC architecture, relevant figures
of merit, general operating principles, and performance limitations. Several excellent
DSC reviews have been written by Grätzel,5,10 Snaith,3 Peter,2 and Meyer11 that
describe the basic operating principles of DSCs in much more detail than what is
provided in this thesis.
1.1 Dye-Sensitized Solar Cells Architecture
Liquid based dye-sensitized solar cells are comprised of a fluorine doped SnO2
front contact (FTO) on glass, nanoparticle photoanode covered in a monolayer of
sensitizing dye, a hole conducting electrolyte, and platinum coated FTO back contact
as shown in figure 1-1. The most well studied DSC is composed of mesoporous TiO2
fabricated from sol-gel processed nanoparticles (e.g. 20 nm in diameter, 60% porosity)
which are screen printed on the FTO and sintered at 450 °C. The nanoparticles films
have a surface area that is typically 1000 times greater than that of a flat junction and
allow for high loading of the sensitizing dye, which is responsible for light harvesting.
The electrolyte is typically comprised of a large concentration of iodide (0.5M) in the
form of lith
(0.05M) of
of iodide (e
iodine into t
are high, I-
based DSC
coated FTO
electrode.
Figure 1-1.
nanoparticle
1.2 Pow
The p
is directly c
efficiency is
Factor, show
hium iodide
iodine (I2).
e.g. log (Keq/
triiodide (I3-
~ 1020cm-3 a
s. The coun
O electrode w
Dye-sensit
es covered b
wer Conve
ower conver
converted in
s defined by
wn in equatio
(LiI) or org
Iodine has a
/M-1) ~ 6) in
~50mM).12
and I3- ~ 10
nter electrod
which reduc
tized solar c
y the sensiti
ersion Ef
rsion efficie
nto electrica
y the short-ci
on 1-1.
PV
2
ganic comple
a strong tend
n acetonitril
The ionic co19 cm-3, resu
de of DSCs
ces the triio
cell schema
izing dye.
fficiency
ncy, ηPV, is
al power by
ircuit curren
VJ OSC
ex (e.g. PM
dency to for
le resulting i
oncentration
ulting in no
is typically
dide, transfe
atic; inset i
the fraction
the solar c
nt density, op
FFOC
MII) and a m
rm triiodide
in near unity
ns in the elec
macroscopic
y comprised
ferring holes
s enhanced
of power fr
cell. The po
pen-circuit v
moderate amo
in the prese
y conversion
ctrolyte solut
c field in liq
d of a platin
s to the cou
view of T
rom the sun
wer convers
voltage, and
(Eq. 1
ount
ence
n of
tion
quid
num
unter
TiO2
that
sion
Fill
1-1)
The
photocurren
density is t
voltage is d
The fill fact
Voc and Jsc
Figure 1-2.
DSC.
To d
dyes have b
been an Zn
the PV char
N
RuN
N-O
O
N71
O-O
O
-O
Bu4N+
Bu4N+
Figure 1-3.
J
V
power con
nt density ve
the photocur
defined at the
tor is the rat
c (0.745).
Photocurren
date three dy
been Ruthen
porphyrin ba
racteristics sh
NC
S
N C S
N
19
O
O-
Chemical St
Jsc = 17.7
Voc = 846
FF = 0.74
nversion ef
ersus voltage
rrent density
e voltage wh
io of maxim
nt Density v
yes have pro
nium-based
ased donor-π
hown in the
S
S
HO
S
S
C6H13
C6H13
tructure of st
73 mA/cm
6mV
45
3
fficiency ca
e plot shown
y at 0V (e.g
hen the phot
mum power f
versus Voltag
duced powe
bipyridyl co
π-acceptor d
table inset.
N
Ru
NC
S
N N C
N
N
S
OC NaO
C106
tate-of-the-a
m2
an be deter
n in figure 1-
g. 17.7 mA
tocurrent den
from the sol
ge plot of h
er conversion
omplexes (N
dye (YD2),14
C8H
S
OOC
C3H7
art sensitizin
rmined by
-2. The shor
A/cm2) and t
nsity is zero
ar cell to the
highly efficie
n efficiencie
N719, C106)4 all show in
N N
NN
Zn
COOH
N
H17
YD2
ng dyes.
analyzing
rt-circuit cur
the open-cir
o (e.g. 846 m
e product of
ent N719 ba
es of >11%;
)13 and one
n figure 1-3 w
C6H17
C3H7
the
rrent
rcuit
mV).
f the
ased
two
has
with
4
Table 1-1: PV Characteristics of Top Performing Sensitizing Dyes.
Dye Jsc(mA/cm2) Voc (mV) FF η (%)
N719 17.7 846 0.75 11.2
C106 19.2 776 0.76 11.3
YD2 18.6 770 0.76 11.0
Conventional liquid based dye-sensitized solar cells (DSCs) have excellent
charge collection efficiencies, high open circuit voltages, and good fill factors (0.70-
0.75). However, DSCs do not completely absorb all of the photons from the visible
and near infrared domain and consequently have lower short circuit photocurrent
densities (<21 mA/cm2) compared to inorganic photovoltaic devices. A key to
improving the efficiency of DSCs is to increase their spectral absorption range. In
order to reach power conversion efficiencies of 15% using an I2-/I3
- redox couple,
DSCs must absorb ~80% of the solar spectrum from 350-900nm.15
1.3 External Quantum Efficiency (EQE)
The external quantum efficiency (EQE) is the fraction of charge pairs collected
at the electrodes per photon incident on the solar cell and is dependent upon the
fraction of light absorbed by the sensitizing dye (ηABS,SD) and the internal quantum
efficiency (IQE) of the system as shown in equation 1-2.
IQEEQE SDABSSD )()( , (Eq. 1-2)
State-of-the-art DSCs are already efficient at absorbing visible light and collecting
charges and can achieve external quantum efficiencies (EQE) of 85% at peak
absorption as shown in figure 1-4.8 DSCs have high internal quantum efficiencies
(>90%)16-18 and in portions of the visible spectrum can absorb >90% of the light.
Figure 1-4. Th
1.4 Phot
The p
is first abso
titania (2). T
oxidized dy
to the coun
electrode to
below in or
limitations o
he external q
tocurren
photocurrent
orbed by the
The electron
ye is regenera
nter electrod
o produce ph
rder to better
of state-of-th
quantum effi
nt Genera
generation
sensitizing
ns then perco
ated by the e
de where it
hotocurrent.
r understand
he-art device
5
iciency of N
ation in D
mechanism
dye (1) whi
olate through
electrolyte (4
is reduced
All of thes
d the genera
es.
N719 based D
Dye-Sens
ms of DSC is
ich rapidly i
h the TiO2 t
4), convertin
by the plat
se processes
al operating p
DSC.
sitized So
s shown in fi
injects the e
to the front c
ng I- to I3- w
tinum coate
s will be disc
principles o
olar Cells
figure 1-5. L
electron into
contact (3).
which diffuse
d FTO coun
cussed in de
f DSCs and
s
ight
the
The
es (5)
nter
etail
the
Figure 1-5. Ph
1.5 Ligh
Light
extinction c
total surface
estimated u
coverage, Γ
molar extin
thickness (d
TiO2 nanop
1000x grea
Sensitizing
molecule/nm
been made f
broad absor
hotocurrent g
ht Harves
absorption
coefficient o
e area of the
using Beer’s
Γ (mol/cm2),
nction coeffic
d) in centime
articles to en
ater than tha
dyes genera
m2 (or Γ = 0.
from rutheni
rption spectru
generation sc
sting in D
in dye-sen
f the sensiti
e oxide film
s law (eq.
, the decadi
cient, ε (M-
eters times t
nhance the s
at of a flat
ally pack tigh
.83-1.66*10-
ium based c
um (Δλ ~ 35
6
chematic of
DSCs
nsitized sola
zing dye, th
m.10 The ligh
1-3) based
c extinction1 cm-1), mul
the roughne
surface area
junction, o
htly on the T-10 mol/cm2)
complexes (e
50nm) but lo
DSC.
ar cells is
he surface co
ht harvesting
on the mol
n coefficient
ltiplied by 1
ss factor (R
a; 10-µm-thic
or a roughn
TiO2 surface
).10 The sens
e.g. N719 an
ow molar ext
determined
overage of t
g efficiency
lar sensitizin
t, σ(cm2/mol
1000 cm3 L
RF). Films ar
ck films hav
ness factor
with a dens
sitizing dye h
nd Z907)8,19
tinction coef
by the m
the dye, and
(ηabs,SD) can
ng dye surf
l), which is
L-1, and the f
re comprised
ve surface ar
(RF) 100x/
sity of 0.5-1
has tradition
that have fa
fficients (5,0
olar
d the
n be
face
the
film
d of
reas
µm.
dye
nally
airly
000-
7
20,000 M-1 cm-1). Organic dyes have recently been developed with substantially
higher molar extinction coefficients (50,000-200,000 M-1 cm-1) but narrow spectral
bandwidths (Δλ~250nm).20-23 As a general rule, dyes that absorb strongly do not
typically exhibit broad absorption.
))((, 101)( RFdSDABS
(Eq. 1-3)
Co-sensitization of titania by dyes with complimentary absorption spectra has
been demonstrated to enhance light absorption and broaden the spectral response of
organic DSCs.24 However, the limited number of sites on the titania surface to attach
dye molecules places a constraint on the light absorption achievable by co-
sensitization. Furthermore, co-sensitization requires that each dye adsorb strongly on
the surface, transfer charge efficiently into the TiO2,13,25-27 slow recombination (i.e. in
the millisecond time domain),27-30 and regenerate with the redox couple.31 Few dyes
exist that are both excellent absorbers and possess the requisite energy levels and
chemical anchoring groups to be good sensitizing dyes. A recent study has
demonstrated the use of Förster resonant energy transfer between covalently linked
energy donor molecules to the sensitizing dye attached on the titania surface.32 Siegers
et al. were able to demonstrate a high excitation transfer efficiency (>89%) between
attached dye molecules and an improvement in the device external quantum efficiency
of 5-10% between 400-500nm. However, the overall power conversion efficiency
enhancement of the DSC was low (< 9%) and linked more to an increase in the open
circuit voltage rather than an increase in the short-circuit photocurrent density
1.6 Charge Transfer Processes in DSCs
Unlike traditional PV devices such as crystalline Silicon and thin films (e.g.
CdTe and CIGS) devices where light absorption, charge separation and collection
occurs within a single material, all charge transfer processes in DSCs occur between
physically distinct components (i.e. dye, titania, and electrolyte). One of the most
interesting scientific aspects of dye-sensitized solar cells is that 1) almost all of these
charge transfer processes can be directly measured and 2) the rates of these processes
can vary dependent upon the dye, photoanode, and hole conducting medium chosen.
The charge
internal qua
open-circuit
There a
rates in tabl
processes, s
to the TiO2
electrolyte.
both labele
separation. E
triiodide in
and are lab
dependent o
properties o
and can be
quantum eff
Figure
transfer an
antum effici
t voltage.
are six charg
le 1-2 that d
shown in bla
2 and hole r
Once the SD
ed in gray
Electrons in
the electroly
beled in red
on the dye, p
of the DSC.
grouped in
ficiency.
1-6. Jablonsk
nd recombina
iency (and
ge transfer
directly influ
ack, for effic
regeneration
D is excited
can occur,
n the titania m
yte (kbr), wh
. The comp
photoanode,
The charge
three differ
ki Plot of Ch
8
ation rates p
thus short-c
processes, s
uence DSC
cient devices
n (kreg) of th
d both radiat
quenching
may recomb
hich is comm
petition betw
and hole con
transfer rate
rent time sca
harge Transf
play a critic
circuit curre
shown in fig
performanc
s is electron
he oxidized
tive (krad) an
g the excito
bine with ho
monly referr
ween variou
nducting me
es can vary b
ale regimes
fer Rates in D
cal role in d
ent density)
gure 1-6 wi
e. The two
injection (ki
d dye by th
nd non-radia
on and pre
oles in the dy
red to as the
us charge tra
edium and de
by more tha
that determ
DSCs
determining
as well as
th the range
most impor
inj) from the
e iodide in
ative (knr) de
venting cha
ye (krec) or w
e back react
ansfer rates
etermine the
an twelve ord
mine the inte
the
the
e of
rtant
SD
the
ecay
arge
with
tion,
are
PV
ders
rnal
9
Table 1-2: List of Rate Lifetime Ranges for Charge Transfer Processes in DSCs under short-circuit current conditions.
Charge Transfer Process Name Lifetime
Electron Injection kinj 20 fs‐500 ps
Hole regeneration kreg 1‐10 µs
Radiative Decay krad 0.5‐50 ns
Nonradiative Decay knr 0.3‐25 ns
e‐ (TiO2) / h+ (dye) recombination krec 200 µs ‐8ms
e‐ (TiO2) / h+ (hole conductor) recombination kbr 1 ms – 1 s
The internal quantum efficiency is defined by equation 1-4, which can be
defined as the probability of electron transfer to the titania (ηe_separation), hole transfer to
the electrolyte (ηh_separation), and the charge collection efficiency (ηCC).
ccseparationhseparationeIQE __ (Eq. 1-4)
1.6.1 Electron Separation Probability in Liquid DSCs
Exciton dissociation from the sensitizing dye occurs on the femto- to nano-
second time scale and is determined by three competing rates: electron injection,
radiative, and non-radiative decay shown in equation 1-5.
nrinj
injseparatione kkk
k
0_
(Eq. 1-5)
Ruthenium based polypyridyl sensitizing dyes (e.g. N719 and C106) undergo strong
metal-to-ligand charge transfer transitions (MLCT) resulting in rapid charge injection
to the titania.5 The electron injection rates of Ru based metal complex dyes is usually
biphasic consisting of a distinct <100 fs ultrafast rate and a slower, picosecond
component. There are two theories for the biexponential charge injection of Ru based
dyes. Some believe that the biphasic nature is intrinsic to the sensitizing dye with fast
decay occurring from the singlet transfer states (1MLCT) and slower injection
10
occurring from the triplet transfer (3MLCT) states. An alternative theory postulates
that the biphasic nature is due to dye packing on the TiO2 surface; the slower rate is a
result of aggregated dyes, unattached to the surface, which must transfer electrons to
attached sensitizing dyes requiring a multistep electron transfer process. In recent
experiments, the slow electron injection component was significantly reduced when
TiO2 films were dipped in highly dilute solutions of a ruthenium metal complex (N3)
dye and an ultrafast monoexponential decay was observed when using a phosphonic
acid anchoring group.33 Metal-ligand complexes have radiative lifetimes between 10-
40 ns with non-radiative lifetimes 2-4 times faster when attached on a metal-oxide
surface. For Ruthenium based metal complexes the electron separation probability is
near unity because kinj is many orders of magnitude faster than k0 and knr.
Organic dye kinj is slower (1ps-500ps) while the k0 (0.5-10 ns) and knr (0.3-5
ns) rates are higher, which can result in incomplete electron transfer and thus lower
IQE. To increase injection rates, organic dyes are often designed to be asymmetric to
increase directional charge transfer.34
1.6.2 Hole Separation Probability in Liquid DSCs
The hole separation probability is dependent upon the hole regeneration rate
versus the recombination rate between electrons in the titania and holes in the
electrolyte (eq. 1-6). Because electron transfer is typically several orders of magnitude
faster than hole transfer it is common to refer to dyes being in the oxidized (or
cationic) state prior to hole transfer and hole transfer as a reductive processes.
recreg
regseparationh kk
k
_
(Eq. 1-6)
In the absence of an electrolyte, krec usually occurs on the microsecond to
millisecond time domain.28,35,36 When iodide/triiodide is introduced into the electrolyte
the dye cation (D+) can be reduced by iodide. Surprisingly, little is understood about
the actual processes that lead to dye regeneration.35 It is not clear whether single
iodide collision with D+ is required or if (D+,I-) and/or (I-,I-) ion pairs are required to
produce I3- as shown in the two iodide reaction pathway below.
11
Single Iodide Reaction:
32 III
IDID
Two Iodide Reaction:
III
IDIID
IDID
32
2
2
2
),(
2
It should be noted that the two iodide reaction pathway is more
thermodynamically favorable and is expected to offer a faster pathway, provided that
(D+,I-) and/or (I-,I-) ion pairs are present in a significant concentration. Although the
rate of hole regeneration may be dependent on the iodide concentration to either first
or second order, it is common to use first order rate approximations.2 Typically an
iodide concentration greater than 0.5M (i.e. 3*1020 cm-3) is considered high enough to
result in kreg rates between 0.2 - 8 µs which is faster than krec.
Dye regeneration is dependent on several factors including difference between
the HOMO level and iodide potential, chemical structure, and additives in the
electrolyte.11 Thiocyanate (i.e. NCS) ligands are often used on Ru based sensitizing
dyes because I- is known to coordinate with the sulfur atom on the NCS ligand.31
Studies of iodide-dye ion pair interactions have been well studied for Ru metal
complex dyes37 but not for organic dyes, which can poor regeneration rates despite
having sufficient HOMO level for regeneration.38 Finally, the rate of reactivity of
iodide increased with the presence of Li+ and other cations with large charge-to-radius
ratios.11,35
12
1.6.3 Charge Collection Efficiency in DSCs
After both the electron and holes have been transferred from the dye they must
travel through the TiO2 and electrolyte (as I3-) respectively in order to produce
photocurrent. The charge collection efficiency, ηCC, is the fraction of electrons/holes
collected at the front/back electrodes that were generated and separated from the
sensitizing dye. The charge collection efficiency can be described as competing
processes of electron transport through the TiO2 nanoparticles (ktrans) versus the back
electron transfer rate (kbr) of electrons in TiO2 recombining with I3- (i.e. eq. 1-7).
Charge collection efficiency can also be thought about in terms of an electron
diffusion length (eq. 1-8), which is dependent upon the diffusivity of the electrons in
titania (De) and the electron recombination lifetime (τbr) in equation 1-8.
brtrans
transCC kk
k
(Eq. 1-7)
bren DL (Eq. 1-8)
Deducing the exact mechanism of electron transport through mesostructured
titania is still a debated topic in the DSC field.39 What is known is that electron
diffusivity in mesoporous TiO2 is many orders of magnitude lower than in single
crystals and tha sub-bandgap states exist in the TiO2 that affect charge transport. The
transport of injected electrons into titania to the collecting electrodes occurs through
diffusion because the macroscopic field across the film is negligible due to the
screening by the high ionic strength electrolyte.40 Electron transport inside DSCs is
strongly dependent upon the light intensity;41 this dependence is considered to be a
result of a broad distribution of traps and can be explained in terms of a multiple
trapping model.42 Electron transport through titania films is typically measured using
intensity modulated photocurrent spectroscopy, which probes the modulation of the
photocurrent in response to modulation of light intensity to extract the electron
transport time. At short-circuit conditions the transport time (τtrans) can be viewed as
13
the average time for electron collection. The diffusivity of electrons in titania (De) is
related to the τtrans by equation 1-9, where d is the film thickness.40De can vary from
10-6-10-5 cm2/s depending on the light intensity. The electron transport rate is the
inverse of the electron transport time (ktrans = 1/τtrans).
etrans D
d
35.2
2
(eq. 1-9)
Inside the DSC space-charge neutrality is maintained so the separate diffusion
of electrons and holes must be constrained so that no net space charge is created inside
the device.43 Thus the motion of electrons in TiO2 and I3- in the electrolyte yield a
single ambipolar diffusion coefficient (Damb), which is a related to the individual
charge diffusivities and the electron (n) and hole (p) concentrations described in eq. 1-
10. Because the electron concentration in TiO2 (n < 1018 cm-3) is generally much less
than the ionic concentration in the electrolyte (p ~ 1020-1021 cm-3) the ambipolar
diffusion coefficient is essentially the diffusion coefficient of the electrons in TiO2.42
e
ehamb D
DpDn
pnD ~
//
(eq. 1-10)
The back reaction (kbr) is considered to be very slow at short-circuit conditions
because the intermediate ionic species (i.e. I* or I2*-) formed during dye regeneration
cannot recombine directly with electrons in the titania.39 It must first form I3- prior to
charge recombination. It is the slow two-electron process to form the tiriodide, which
is almost unique to the iodide/triidodie redox couple, that enables long-lived electron
lifetimes (τbr ~1 ms- 1 s) for liquid based DSCs.35
State-of-the-art DSCs have low kbr rates and are thus very efficient at
collecting charges (> 98%). As an example, at full sun De is roughly 5*10-5 cm2/s
which would correspond to an electron transport time (τtrans) of 85 µs in a 10-µm-thick
TiO2 film (ktrans = 1.2*104 s-1). If the electron recombination lifetime (τbr) is 500 ms
(kbr = 2 s-1) then the ηCC based on eq. 1-7 would be >99.9% and the diffusion length
14
would be 50 µm. As a general rule, a diffusion length three times greater than the film
thickness results in near unity charge collection efficiency.
1.7 Determination of Open-Circuit Voltage
Under open-circuit voltage conditions the photogenerated electron and holes
must completely recombine inside of the DSC (i.e. photocurrent is zero), which means
that the internal quantum efficiency of the DSC must also be zero. It should be noted
that the charge kinetic rates described in section 1.6 are for DSCs under short-circuit
conditions. As a load (or an applied voltage) is placed across the device there is an
increase in the electron carrier concentration inside the TiO2 the directly increases the
krec and kbr rates. The change in applied voltage is not likely to have a large influence
on the electron separation probability so the greatest reduction in the IQE will come
from a decrease in the hole separation probability (section 1.63) and decrease in the
charge collection efficiency (section 1.6.4). For sensitizing dyes that are capable of
fast regeneration, the back reaction (kbr) rate will be the primary recombination
mechanism that determines the open-circuit voltage in state-of-the-art DSCs. A
detailed, mathematical description of Voc determination is given by Peter.2
1.8 Estimating the Maximum Obtainable Power Conversion
Efficiency of Liquid based DSCs44
The maximum theoretical efficiency of a single junction solar cell is
approximately 31% for a 1.4 eV band gap semiconductor. .45 The Shockley-Queisser
calculation assumes that 1) all photons above the band gap are absorbed, 2) there is no
voltage loss (i.e. Voc is essentially equal to the band gap), 3) all photons absorbed
result are collected (i.e. IQE = 100% at short-circuit conditions), and 4) radiative
recombination is the only loss mechanism considered.45 For DSCs, the maximum
open-circuit voltage is determined by the electron quasi Fermi level in the TiO2 and
the Fermi level (or Nernst potential) of the iodide/triiodide electrolyte, which is
roughly 0.8eV. Ideally, one would design a dye with a 0.8eV band gap to absorb all
photons from the 350 nm to 1500 nm. However, over potentials are required to drive
15
the electron and hole transfer reactions. The single largest loss-in-potential is related to
dye regeneration. An over potential of 0.2-0.3 eV between the HOMO the dye and the
iodide potential to drive the regeneration reaction. There is a further drop in potential
of around 0.3 eV when two I2*- are converted to I3
- and I- (section 1.6.3). Finally at
least 50-100mV is required between the LUMO of the dye and the conduction band of
the TiO2 to drive electron injection. Typically high efficiency DSCs have a loss-in-
potential of 0.8 eV; Snaith estimates that the maximum obtainable power conversion
efficiency is 13.4% with absorption onset at 840 nm for iodide/triiodide based DSCs.44
The maximum theoretical efficiency increases to 15.1% by reducing the loss-in-
potential in a liquid DSC to 0.66 eV creating an absorption onset at 920nm.
The maximum obtainable power conversion efficiency may be slightly higher if
near-infrared dyes are cosensitized on the TiO2 surface that absorb strongly in the tail
absorption region of the sensitizing dyes further increasing the predicted photocurrent.
It should be noted that these calculations are based on the desire for 100% charge
separation efficiency and thus require large over potentials. One can envision
developing narrow absorbing, near-infrared dyes with reduced over potential (and thus
lower IQE) but which still provide additional photocurrent to boost performance.
1.9 Improving the Power Conversion Efficiency of Liquid
Based DSCs
DSCs must absorb light in the infrared portion of the solar spectrum in order to
increase the power conversion efficiency of DSCs from 12% to the maximum
obtainable power conversion efficiency of 15%. State-of-the-art DSCs only absorb
light in the visible portion (350-700 nm) shown in figure 1-7. It is challenging to
design single dyes that absorb more broadly than Ru based metal complexes.
Therefore it is necessary to look at designing new sensitizing dyes that are capable of
absorbing in the NIR-portion of the spectrum. It is theoretically possible to develop
NIR sensitizing dyes with high internal quantum and potentially high voltage, but it is
unlikely in practice that these dyes will absorb broadly. The remainder of the thesis
will discuss
light harves
Figu
s several str
ting inside o
ure 1-7. AM
rategies that
of DSCs.
1.5G Solar S
16
involve usi
Spectrum
ing energy transfer to ffurther broa
aden
17
2 Förster Resonant Energy Transfer (FRET)
Förster resonant energy transfer involves dipole-dipole coupling of two
chromophores known as the donor and acceptor through an electric field.46 An
excitation of the donor, or in our case the energy relay dye, can be transferred
nonradiatively through the field to the acceptor, or sensitizing dye, if there is overlap
between the emission spectrum of the donor and the absorption spectrum of the
acceptor. Efficient energy transfer over 3-8nm can be achieved with strong spectral
overlap and high donor emission efficiencies, for an isotropic alignment between
individual chromophores in solution. If, however, the single acceptor chromophore is
replaced by a dense 2D array (i.e. sensitizing dyes tightly packed on the titania surface)
FRET can become efficient well over 25 nm from the interface.47,48
2.1 FRET Radius (R0)
The FRET radius (R0), the distance at which Förster energy transfer is 50%
probable between individual chromophores, is the primary figure of merit used to
evaluate the strength of resonant energy transfer from donor to acceptor dye. The rate
of Förster resonant energy transfer (kFRET), is a function of the separation distance
between the donor molecule to nearby acceptor molecules. The rate of Förster energy
transfer between isolated chromophores, known as point-to-point transfer, is given by
kFRET = k0 (Ro)6/r6, where r is the separation distance and k0 is the natural fluorescence
decay rate. When multiple acceptor molecules are present, the FRET rate is equal to
the sum of the transfer rates to each of the acceptors. The FRET R0 between a donor
and acceptor is calculated using equation 2-1.
dF
Nn
QR AD
A
Do
445
26
128
)10ln(9000 (eq. 2-1)
Where n is the index of refraction of the host medium (1.4-1.5 for the DSC electrolyte),
κ2 is the orientational factor (2/3 for random orientation), NA is Avogadro's number,
QD is the ph
ε(λ) is the m
2.2 Pote
Dye-s
transfer bec
from 5 nm t
dye/nm2) ef
luminescent
photons and
increasing t
charge gene
the sensitizi
the electroly
higher ener
sensitizing d
Figure 2-1.
relay dyes.
energy (red)
an electron
hotolumines
molar extinct
ential for
sensitized so
cause mesost
to 45 nm dia
ffectively ac
t donor chro
d efficiently
the absorptio
eration incor
ing dye (SD
yte. In the ne
gy (blue) ph
dye.
DSC schem
The right sid
) photons in
into the tita
scence effici
tion coefficie
r FRET in
olar cells hav
tructured tita
ameters and
cting as a 2
omophores
y transfer e
on bandwid
rporated in t
) (1), which
ew design, th
hotons and
matic represe
de of the fig
DSCs: light
ania and a h
18
iency, FD is
ent.
n Dye-Se
ve a morpho
ania films h
d sensitizing
2D array on
inside the li
energy to th
dth of the D
this system.
h transfers an
he unattache
then underg
entation of a
gure shows t
t is absorbed
hole is transp
the emissio
ensitized
ology that is
have relative
g dyes pack
n the surface
iquid electro
he anchored
DSC. Figure
In typical D
n electron in
ed energy rel
goes Förster
a dye-sensit
the typical ab
d by the sens
ported to th
on profile of
Solar Ce
uniquely su
ely small por
tightly on th
e. It is poss
olyte to abso
d sensitizing
2-1 shows
DSCs, light
nto the titani
lay dye (ERD
r energy tran
tized solar c
bsorption pr
sitizing dye
he back cont
f the donor,
ells
uited for ene
re sizes rang
he titania (e.
sible for hig
orb high ene
g acceptor d
two routes
is absorbed
ia and hole
D) is excited
nsfer (2) to
cell with ene
rocess for lo
(1), transferr
tact through
and
ergy
ging
g. 1
ghly
ergy
dye,
for
d by
into
d by
the
ergy
ower
ring
the
19
electrolyte. The energy relay dye process is similar except that, higher energy (blue)
photons are first absorbed by the energy relay dye that undergoes Förster energy
transfer (2) to the sensitizing dye.
This design is analogous to photosynthesis in purple bacteria where an aggregate
of light-harvesting pigments transfer their energy to the reaction center initiating
charge separation.49 In particular, the pigment, LH-II, is not in direct contact with the
reaction center, and transfers its excitation via an intermediate pigment (LH-I) in
under 100 ps with ~95% efficiency.50,51
2.3 Modeling the Efficiency of Förster Resonant Energy
Transfer from Energy Relay Dyes in DSCs
We have developed a model that approximates the nanopores as either cylinders
or spheres to calculate FRET rate profiles, ( )FRETk x
, and excitation transfer efficiency
profiles, ( )ETE x
, assuming homogenous ERD concentration inside the nanopores
and uniform sensitizing dye coverage over the pore walls.
2.3.1 The Importance of the Average Excitation Transfer Efficiency (ETE)
In order for energy relay dyes (ERDs) to be used in state-of-the-art dye-
sensitized solar cells, the excited ERDs must be able to efficiently transfer energy to
the sensitizing dyes. Conventional DSCs are already efficient at absorbing visible light
and collecting charges and can achieve external quantum efficiencies (EQE) of 85% at
peak absorption.8 The EQE contribution from the sensitizing dye is determined by the
fraction of light absorbed by the sensitizing dye and the internal quantum efficiency
(IQE). DSCs have high internal quantum efficiencies (>90%)16-18 and in portions of
the visible spectrum can absorb >90% of the light.
When photons are absorbed by the energy relay dye in ERD DSCs, they must
undergo an additional energy transfer step before contributing to photocurrent; the
20
EQE contribution from the relay dye (EQEERD) is thus defined by equation 2-2, where
ηABS,ERD is the fraction of light absorbed by the ERD inside of the titania film and
ETE is the average excitation transfer efficiency, or the average probability that an
excited relay dye transfers its energy to a sensitizing dye. In order for ERDs to be
viable in DSCs, excitation transfer efficiencies of over 90% are required to achieve a
peak EQE of 85%.
ETEIQEEQE ERDABSERD , (eq. 2-2)
2.3.2 Introduction to Modeling FRET in DSCs
We use an analytic theory to calculate the excitation transfer efficiency from
the relay dye to the sensitizing dye accounting for dynamic quenching and relay dye
diffusion. We present calculations for pores of cylindrical and spherical geometry and
examine the effects of the Förster radius, the pore size, sensitizing dye surface
concentration, collisional quenching, and energy relay dye lifetime. We find that the
excitation transfer efficiency can easily exceed 90% for appropriately chosen dyes and
propose two different strategies for selecting dyes to achieve record power conversion
efficiencies.
Theoretical calculations for the excitation transfer efficiency and dynamics
have been performed and experimentally verified for energy donors and acceptors in a
variety of geometries and distributions.52-54 Differences in the geometric arrangement
of donors and acceptors can have a significant impact on the excitation transfer
efficiency and energy transfer dynamics. Much of the recent work has been motivated
by the application of using fluorescence spectroscopic methods to measure nanometer
scale distances in biological 55,56 and polymeric 57 systems for structural
characterization.
Förster’s energy transfer theory 58 has also been extended to account for the
effects of chromophore diffusion. 59,60 Diffusion of the donors and/or the acceptors
can significantly increase the excitation transfer efficiency since it enables donor and
acceptors that are originally too far apart for energy transfer to appreciably occur to
move closer together, into range for FRET. This effect is large for dye molecules with
21
long excited state lifetimes (≥ 1 μs) in low viscosity solvents that can diffuse a
distance in the excited state that is far greater than the distance over which FRET is
effective. Förster resonant energy transfer in the presence of chromophore diffusion
has been typically studied in three different regimes: the stationary limit where
diffusion is negligible, the rapid diffusion limit where the diffusion length is much
larger than the average donor-acceptor separation distance, and the more complex
intermediate regime. Different models are used for calculating the excitation transfer
efficiency for each regime.
We present a comprehensive model to compute the excitation transfer
efficiency in a dye sensitized solar cell for all three diffusional regimes from relay
dyes distributed throughout the mesoscopic pore volume to sensitizing dye molecules
densely and uniformly attached to the pore walls. In our model we consider the
competing process of collisional quenching of the relay dye fluorescence which can be
significant in DSCs since the iodide/triiodide redox couple is a nearly perfect quencher
of many dyes 61. We define a critical distance, Rc, over which the energy transfer
process is efficient in this system and which we propose as the figure of merit in
selecting dyes for high excitation transfer efficiencies. We present quantitative
calculations of the excitation transfer efficiency for pores with cylindrical or spherical
geometries and consider the effects of the Förster radius, average pore size, sensitizing
dye surface concentration, collisional quenching rate, and relay dye lifetime on the
excitation transfer efficiency. We find that the ETE can easily exceed 90% in two
different situations: dye combinations with a relatively large Förster radius in which
the donor has a short lifetime to avoid quenching effects or alternatively a dye
combination which can have a relatively small Förster radius, provided that the relay
dye has a long fluorescence lifetime and is not significantly quenched by the
electrolyte to enable chromophore diffusion. On the basis of these calculations, we
present design criteria for selecting dyes and device architectures to achieve near unity
excitation transfer efficiencies.
22
2.3.3 Background and Model Description
Förster resonant energy transfer is the mechanism for excitation transfer
mediated by the coupling of two resonant dipoles through the electric field. The rate
of FRET, kF, from an energy donor at the position vector rD to an acceptor at rA, as
shown in Fig. 2-2a is given by:
60
60
1F
Rk
A Dr r (eq. 2-3)
Here τ0 is the lifetime of the energy donor excited state, and R0 is the Förster radius
which is the distance over which excitation transfer is 50% probable. The Förster
radius can be computed from the energy donor photoluminescence quantum
efficiency, QD,0, and overlap integral of the donor emission spectrum FD with the
acceptor absorption spectrum εA (eq. 2-1).61
The other factors occurring in this equation are Avogadro’s number, NA, the
dielectric constant of the medium, n, and a dimensionless orientation factor, κ2, which
is equal to 2/3 if the dipoles are randomly oriented and can rapidly reorient. We
define QD,0 such that it accounts for all static quenching effects but does not account
for dynamic quenching by other donor dye molecules or chemical species, which we
address separately. If multiple acceptors are available, as shown in Figure 2-2b, the
total rate of resonant energy transfer is the sum of the rates to each acceptor, since the
acceptors act independently.
60
60
1F
Rk
ii
DA
A D
rr r
(eq. 2-4)
It is also possible in some systems for energy transfer to occur from one donor
to another donor, though the Förster radius for this process is typically quite small so
we assume that this process is negligible in our model. Knowledge of the exact
positions of all of the donors and acceptors is required to make use of Eq. 2-4. More
general theories have been developed to calculate the rate of energy transfer if
statistics describing the distribution of acceptors and donors are known. 52 Since there
is statistical uncertainty in the positions of individual donors and acceptors, the
dynamics o
single rate a
Figure 2-2. (a
and (b) from
as in the cas
In o
inside a por
surface of t
spacing betw
radius (CAR
pore wall su
DSC
with this ap
from the p
sensitizing d
acceptors is
distances to
Förster dipo
large chrom
rate of För
efficiency s
separation d
f energy tra
and must be
a) Geometrie
m donors to
se of a dye se
our model, w
re, and the en
the pore wa
ween indivi
R02 >> 1), we
urface.
C’s typically
pproximation
ore wall th
dye molecul
s very small
these neare
ole approxim
mophores. 62
rster transfe
ince energy
distance.
ansfer from a
characterize
es of FRET
a dense mon
ensitized sol
we assume
nergy accep
alls, describe
dual accepto
e can approx
Fk
Dr
y have a CA
n are most s
hat is comp
les. In this c
l and thus t
st acceptors.
mation becom
Although th
er, they hav
transfer is n
23
a given dono
ed by a proba
occurring fr
nolayer of a
lar cell with
that the ene
tors are unif
ed by a surf
or chromoph
ximate the su
60
60
1
A
A
S
C R
A Dr r
between 0.
significant w
arable or s
case, the dis
the energy t
. This is als
me significan
hese factors
ve a negligi
nearly 100%
or cannot in
abilistic distr
rom a single
acceptors wit
relay dyes.
ergy donors
formly and d
face concen
hores is sma
um in Eq. 2-
2Adr
.5-1 dye/nm
when the ene
maller than
stance betwe
transfer rate
o the regime
nt and can b
s may have
ible effect
% probable a
n general be
ribution of r
e donor to a
th surface co
s are unifor
densely distr
ntration CA.
all compared
-4 with an in
m2 10. The er
ergy relay dy
n the spacin
een a donor
is sensitive
e where devi
e off by mor
a significan
on the exc
at this short
described b
rates.
single accep
oncentration
rmly distribu
ributed over
If the aver
d to the För
ntegral over
(eq. 2
rrors associa
ye is a dista
ng between
and the nea
e to the pre
iations from
re than 50%
t impact on
citation tran
donor-accep
by a
ptor
n CA
uted
r the
rage
rster
r the
2-5)
ated
ance
the
arest
cise
m the
% for
the
nsfer
ptor
24
2.3.4 Effects of Electrolyte Quenching
The presence of the redox couple in the electrolyte of dye sensitized solar cells
can greatly increase the rate of non-radiative decay of the relay dye, providing a
parasitic pathway for excitation decay that competes with energy transfer. Iodide and
triiodide are perfect quenchers of many dye molecules meaning that a single collision
with an excited dye results in quenching. Dynamic quenching is described by the
Stern-Volmer equation 61.
1,
0,0 0
1j
D Q Qq jj
D
Qk Q
Q
(eq. 2-6)
Here [Qj] is the concentration of quenching species j and kqj is the bimolecular
quenching coefficient for the dye-quencher combination, which is typically 109-1010
M-1s-1 for effective quenchers. QD,0 and τ0 are the photoluminescence quantum
efficiency and lifetime of the donor in absence of the quenching species, while QD,Q
and τQ refer to these respective quantities when the quenching species are present. The
degree of quenching is larger for relay dyes with a longer lifetime τ0, assuming similar
values for kq, because it provides more time for the dye to collide with quenchers.
2.3.5 The Critical Radius (RC)
The Förster radius is the length scale over which Förster resonant energy
transfer is efficient between a donor and a single acceptor. For a donor that can
undergo energy transfer to a monolayer of acceptors of surface concentration CA, or
could be dynamically quenched by quenchers Qj, we show that the length scale over
which FRET is efficient is instead:
1/460
01j
Ac
q jj
C RR
k Q
(eq. 2-7)
A large Förster radius between the relay and sensitizing dyes, a dense surface
coverage of the sensitizing dye on the titania surface and a small degree of quenching
25
of the relay dye by the electrolyte are all important in achieving a large critical energy
transfer distance, Rc, and a high excitation transfer efficiency.
The ratio of the competing rates of energy transfer and quenching can be used
to calculate the excitation transfer efficiency. The details of this calculation depend
upon the extent that the relay dye diffuses, which can greatly increase the excitation
transfer efficiency. Assuming that the diffusivity of dyes in a mesopore can be
described by the Stokes-Einstein relation, a relay dye dissolved in acetonitrile with a
hydrodynamic radius of 1 nm will have a diffusivity of around 0.6 nm2/ns. Diffusion
of the relay dye can be neglected when the relay dye diffusion length is small
compared to the critical energy transfer distance. This is the case for relay dyes with
short quenched fluorescence lifetimes of τQ ≤ 1 ns. In the so called rapid diffusion
limit, the diffusion length is large compared to the average donor-acceptor separation
distance 60, or roughly when √6 ≫ where 2Rp is the diameter of the pore. For a
typical pore diameter of 2Rp = 30 nm, which is produced by using standard 20 nm
diameter titania particles emulsions, this limit is reached when τQ ≥ 1 μs. We also
examine the regime for dyes with intermediate lifetimes to investigate the tradeoff in
selecting dyes with longer lifetimes, which can diffuse farther but are more easily
quenched.
2.3.6 Short lifetime relay dyes: ETE in the absence of diffusion
Many organic dyes have a fluorescence lifetime between 0.5-10 ns 61. Most
are nearly perfectly quenched by iodide/triiodide resulting in an even shorter lifetime
in the DSC electrolyte. Consequently diffusion can be ignored when using most
organic dyes as a relay dye since the dye can only diffuse about a nanometer or less
during its excited state lifetime, which typically has a negligible impact on the
excitation transfer efficiency. The ETE from a stationary energy donor to a stationary
group of acceptors is equal to the ratio of the rate the excited donor undergoes energy
transfer to the total rate of all decay mechanisms of the excited donor:
1
FETE
Q F
k
k
DD
D
rr
r (eq. 2-8)
26
The excitation transfer efficiency for an ensemble of static donors is equal to the
average ETE of all of the donors 63. If we assume that the donors are evenly
distributed throughout the pore volume, we can calculate the overall excitation transfer
efficiency for the pore by averaging Eq. 2-9 over all possible positions of the donor
and using Eq. 2-5 and 2-6 for the rates of energy transfer and non-radiative decay.
13 31 11 1ETE ET D Q f DV V
dr k drV V
D Dr r
1 16 4 2
2 3 306 6
0
1 11 1 1 1
A A
Q A c AA D D
V S V S
C R R drdr dr dr
V V
A D A Dr r r r
(eq. 2-9)
Here V is the pore volume. In the limit of no donor chromophore diffusion, the
excitation transfer efficiency only depends upon the geometrical shape of the pore,
which sets the bounds of both integrals, and the critical energy transfer distance, Rc.
Eq. 2-9 can be easily evaluated numerically for pores that are modeled as
cylinders or spheres. We further approximate the cylindrical pores to have an infinite
length, L, which is valid if the length of the pore is much larger than the pore radius
and the critical energy transfer distance. For both pore geometries, R is the radial
distance from the donor to the center of the pore and Rp is the radius of the pore (Fig.
2-3(a) and (b)). The rate of FRET from a relay dye located a distance R from the
center of the pore to sensitizing dyes on the pore walls can be calculated for both pore
geometries by applying Eq. (4):
662 020
, 6 2 200 0
1 1
cos sinA
A pAF cyl A
S p p
C R R d dzC Rk R dr
R R R
A Dr r (eq. 2-10)
2
26 2 62 0 0
, 3 4 40 0 2 2 20 0
2
4 1sin1 1
cos sin 1
pA p AF sph
pp p
p
R
RC R R d d C Rk R
R RR R RR
(eq. 2-11)
Figure 2-3.
The relay dy
sensitizing
efficiency in
of diffusion
pore diamet
For
and spherica
,
3
4ETE sph R
The
energy trans
the spherica
transfer dist
(Fig. 2-3(c)
7.0nm respe
DSC with a
kq[Q] ~5x1
molecules/n
Geometries
ye is distribu
dye densely
n cylindrical
n as a functio
ter 2Rp.
the limit of
al pores can
,ETE cyl
30
1 1pR
QpR
excitation t
sfer distance
al case in Eq
tance should
). For pore
ectively for
a convention
09 s-1 64. A
nm2, for a rel
s of the cylin
uted through
y covers th
l (dotted curv
on of the rat
static donor
then be calc
2 0
11 1
pR
pR L
1 2, 4Q F sphk R
transfer effic
e to the diam
q. 2-13. Fo
d be roughly
diameters o
cylindrical
nal iodide/trii
Assuming a
lay dye lifeti
27
ndrical (a) an
hout the volu
he pore wal
ve) and sphe
tio of the cri
rs, the excita
culated from
,1 Q F cylk R
12
0
1 3 1dR
ciency only
meter of the
or an ETE o
y a quarter to
of 30 nm, th
and spheric
iodide electr
sensitizing
ime of τ0 = 0
nd spherical
ume of the in
lls. (c) Cal
erical (solid
itical energy
ation transfe
m Eq. 2-9.
12 LRdR
4
642
c
p
R
R
depends up
pore, which
of greater th
o a third of
his correspon
cal pores. F
rolyte, quenc
dye surface
0.5ns, a mini
l (b) pores o
nterior of the
lculated exc
curve) pore
y transfer di
er efficiency
12
242
1
1
rr d
r
pon the ratio
h can be see
han 90% the
the pore dia
nds to Rc >
For organic
ching is nea
e coverage
imum Förste
of diameter 2
e pore while
citation tran
s in the abse
stance Rc to
y for cylindr
(eq. 2-1
dr (eq. 2-1
o of the crit
en explicitly
e critical ene
ameter or lar
9.2nm and R
relay dyes i
arly perfect w
of CA=0.5
er radius of 6
2Rp.
e the
nsfer
ence
the
rical
12)
13)
tical
y for
ergy
rger
Rc >
in a
with
dye
6.07
28
nm or 5.03 nm is required to achieve 90% ETE for cylindrical and spherical pores
respectively. If the (unquenched) lifetime of the dye were τ0 = 5ns, however, only
~70% ETE would be achieved for these Förster radii. There is thus considerable
benefit of using shorter lifetime relay dyes to minimize the effects of quenching when
nearly perfect quenching occurs.
In an actual mesoporous film, the pores are neither cylindrical nor spherical in
shape. However, the difference in the calculated energy transfer efficiencies between
these two geometries is not large so we would expect the ETE for actual pore
geometries to be close to the results for the cylindrical and spherical pores.
Mesoporous films have a distribution of pore sizes, which can be measured by the
Brunauer, Emmett and Teller (BET) method. The theoretical average excitation
transfer efficiency for the film can be determined by calculating the excitation transfer
efficiency at each pore size and taking a weighted average of these values using the
measured distribution of pore sizes.
2.3.7 Long lifetime relay dyes: ETE in the rapid diffusion limit
Not all dyes are completely quenched by iodide and triiodide. For example,
some lanthanide complexes can undergo thousands of collisions with iodide before
being quenched and have bimolecular quenching coefficients of kq < 106 M-1s-1 65. For
relay dyes that are relatively insensitive to collisional quenching, a long lifetime is
highly beneficial for energy transfer since it enables the dye to diffuse closer to the
pore wall, greatly reducing the critical energy transfer distance required for 90%
excitation transfer efficiency. Eventually a longer diffusion length leads to no further
improvement in the ETE and this situation is referred to as the rapid diffusion limit. In
this limit, a donor can move through nearly all of the different regions in the pore
during its excited state lifetime (τQ ≥ 1μs for 30 nm diameter pores). Consequently, all
donors have the same average rate of undergoing energy transfer averaged over their
excited state lifetime. This rate can be computed by averaging the energy transfer rate
over all possible donor positions in the pore 60.
31e
F F DVe
k k drV
Dr (eq. 2-14)
29
Eq. 2-14 diverges if the integration volume is taken as the full pore volume
since this would allow the donor to diffuse arbitrarily close to acceptors on the pore
wall where the rate of energy transfer approaches infinity in the Förster dipole model.
The finite size of the donor and acceptor molecules needs to be considered when
setting the bounds on the volume integration to set a minimum separation distance
between the donors and acceptors. For small dye molecules, this distance of closest
approach is around Ra = 0.5nm 60,61. We use Ve to represent the volume in the pore
that donors can occupy, which excludes the regions that are less than the distance of
closest approach from the pore wall.
The average excitation transfer efficiency is equal to the ratio of the average rate of
energy transfer to the average total decay rate 61 which combined with eq. 2-4 and eq.
2-6 yields:
14 2
361
11 1
e A
c AFETE D
e V SQ F
R drkdr
Vk
A Dr r
(eq. 2-15)
For the case of the cylindrical pore, the integrations need to be performed numerically.
An analytic solution, however, exists for the sphere:
1 14
2,3 3 30
31 1 4 1 1
2 14
p bR Rc
ETE Q F sphpp b
Rk R dR
R b bR R
(eq. 2-16)
Here Ra is the closest distance that the donor can be from the boundary of the
pore and b = Ra / (2Rp). The critical energy transfer distance required to achieve 90%
ETE in a 30 nm diameter pore is shown in Fig. 2-4. The excitation transfer efficiency
is higher for the spherical pores compared to the cylindrical pores for the same
parameters, though the difference is again relatively small in the regime where the
ETE is high. The excitation transfer efficiency is highly sensitive to the distance of
closest approach, Ra, and consequently larger and bulkier relay and sensitizing dyes
will have a lower ETE since they cannot get as close together. Smokey this is not
nam’, this is bowling. There are rules. For a distance of closest approach of Ra = 0.5
nm, Rc ≥ 2 nm is sufficient to get over 90% ETE for both pore geometries. If the
impact of qu
with a Först
Figure 2-4.
the rapid di
the distance
pore diamet
for other po
changed.
The
in the rapid
presence of
τQ ≥ 1μs.
microsecond
ultimately li
The iodide c
M, for effic
molecules
iodide/triiod
Encapsul
protective s
uenching is n
ter radius of
Excitation t
iffusion limi
e of closest a
ter was assum
ore sizes, sc
challenge in
d diffusion
the quenchi
Many dyes
d or more.
imited by th
concentratio
cient regener
should be
dide of kq ≤ 1
lated structu
shell, may b
negligible, th
only 1.8 nm
transfer effic
it as a funct
approach tha
med to be 2
cale Rc and
n selecting d
limit is fin
ing DSC elec
s, including
However, f
he time scale
on in the elec
ration of the
chosen w
106 M-1s-1 in
ures, where
e good relay
30
his critical e
m, assuming C
ciency in (a
ion of the c
at the donor
Rp = 30nm.
Ra by the
dyes to achie
ding dyes w
ctrolyte to b
metal-ligan
from Eq. 2-6
e over which
ctrolyte need
e sensitizing
with bimol
order to rea
the optical
y dye candid
energy transf
CA = 0.5 mo
a) cylindrical
critical energ
rs can be fro
To determi
same propo
eve high exc
with sufficie
e in the rapi
nd complexe
6 the dye li
h quenching
ds to be at le
dye to occu
lecular que
ach the rapid
lly active r
dates due to
fer distance c
lecules/nm2.
l and (b) sph
gy transfer d
om the pore
ine the excit
ortionality fa
citation trans
ently long l
id diffusion r
es have life
ifetime in th
g occurs, Q
ast 1020 ions
ur. Consequ
enching coe
d diffusion li
region is su
o their reduc
can be achie
.
herical pore
distance, Rc
e wall, Ra.
tation efficie
actor that R
sfer efficien
lifetimes in
regime, roug
etimes (τ0) o
he electrolyt
jq jjk Q
s/cm3 2,66, or
uently relay
efficients w
mit.
urrounded b
ced bimolecu
eved
es in
and
The
ency
Rp is
cies
the
ghly
of a
te is
1 .
r 0.2
dye
with
by a
ular
31
quenching coefficients. Lanthanide cryptates have been demonstrated to have
bimolecular quenching coefficients with iodide as low as kq = 102 M-1s-1 67. Core-shell
nanoparticles may be another possibility. The thickness of the protective shell should
be as thin as possible without sacrificing its effectiveness, since the shell increases the
distance of closest approach, Ra, reducing the excitation transfer efficiency.
2.3.8 Intermediate lifetime dyes: Full model of the impact of relay dye lifetime
on ETE
A longer relay dye lifetime allows the dye to diffuse further in the DSC
electrolyte, improving the excitation transfer efficiency, but also increases the chances
that it will be quenched, which lowers the ETE. To understand the effect of the relay
dye lifetime on the ETE when diffusion and quenching are both significant, we need to
examine the regime of intermediate diffusion. The general case for energy transfer in
the presence of diffusion was treated by Steinberg et al. 59 who considered the survival
probability distribution of the excited donor and derived a partial differential equation
to describe the decay of this distribution. We summarize this method below, which
we have adapted to include the effects of dynamic quenching.
At time t = 0, a single donor is excited somewhere inside the pore. P(rD, t) is
the survival probability, which is the probability density that the excited donor is at the
position vector r after a delay of t following excitation. Since a random donor is
excited, P(t = 0) is equal to 1/Ve everywhere inside the pore where there could be a
donor molecule and zero elsewhere. As in the case of the rapid diffusion limit, the
donors are not permitted to be closer than the distance of closest approach, Ra, from
the pore walls. The survival probability function evolves according to the following
continuity equation, which accounts for diffusion, FRET and non-radiative decay of
the donor.
2 1,
,f Q
P tD k P t
t
DD D
rr r (2-17)
Here D is the diffusion coefficient of the donor species. We impose homogeneous
Neumann boundary conditions, 0P n (where n is the surface normal vector), to
32
allow for the possibility for donors to bounce off the pore wall without undergoing
energy transfer, which can be significant if the donor’s fluorescence lifetime is long or
the minimum separation distance is large, resulting in a slow FRET rate. The survival
probability distribution can be in principle determined using a numerical partial
differential equation solver to solve Eq. 2-17.
The excitation transfer efficiency is equal to the probability the excited donor
does not undergo non-radiative decay. The integral of P over the pore volume gives
us the probability that the donor has not yet decayed after time t, which approaches
zero for times much longer than the lifetime of the donor. Multiplying this by the non-
radiative decay rate and integrating over all time gives the probability that the excited
donor undergoes non-radiative decay. Thus the excitation transfer efficiency is given
by 68:
1 3
01 ,
eETE Q DV
P t dr dt Dr (eq. 2-18)
In the case of cylindrical and spherical pores, the survival distribution function P only
depends upon the radial distance from the center of the pore, and Eq. 2-18 simplifies
to:
1, 1,m
f Qm
P R t PD R k R P R t
t R RR
(eq. 2-19)
Here m = 1 in the case of the cylindrical pores and m = 2 for the spherical pores and kf
is given by Eq. 2-10 or eq. 2-11. The initial condition to the problem is P(t = 0) = 1/Ve
inside the pore and zero outside, where Ve is the volume of a cylinder of radius Rp – Ra
and length L or of a sphere of radius Rp – Ra. We impose reflective boundaries at the
pore walls and constrain the diffusion flux to be finite in the pore center:
, 0
0 and 0p aP R R R t P R
t t
(eq. 2-20)
The partial differential equation can be solved numerically by discretizing P in space
and time. The integrations in Eq. 2-18 can then be performed numerically to calculate
the excitation transfer efficiency. This method was checked in the static and rapid
diffusion lim
previously d
Figure 2-5.
quenching (
the relay dy
The pore di
0.5nm, and
The cal
shown abov
couple (Fig
5(b), kq[Q] =
relay dye li
electrolyte.
limited to 1
case of redu
improving t
ETE decrea
from 1 ns to
is increased
becomes sig
less than the
mits and wa
described mo
Excitation
(kq[Q] = 109s
ye lifetime a
iameter was
a relay dye d
lculated exci
ve for the cas
. 4(a), kq[Q]
= 1x106 s-1).
fetime becau
Diffusion p
ns and the d
uced quench
the ETE. T
ases by about
o 1 μs, going
d further, ho
gnificant. A
e time scale
as found to
odels for tho
transfer eff
s-1) and (b) r
and critical
set to 2Rp =
diffusivity o
itation transf
se when the
] = 1x109 s-
In the case
use longer l
plays a min
dye will be q
hing, longer
The critical
t a factor of
g from the sta
owever, the
A longer dye
for quenchin
33
agree within
ose limits.
fficiency for
reduced que
energy trans
= 30nm, the
f D = 0.6nm
fer efficiency
relay dye is 1) and when
e of near perf
lived dyes a
nor role sinc
quenched be
lived dyes
energy tran
f three when
atic limit to
ETE begin
lifetime is th
ng.
n 1% of the
r a spherica
nching (kq[Q
sfer distance
distance of
m2/ns was use
y for 30-nm
nearly perfe
n the quench
fect quenchi
allow more t
ce the dye li
efore it can
allow the re
sfer distanc
the lifetime
the rapid dif
ns to drop si
hus only ben
e solutions f
al pore (a) w
Q] = 106s-1)
e in absence
f closest app
ed in these c
-diameter sp
ectly quench
hing is less
ing, the ETE
time to be q
ifetime in th
diffuse appr
elay dye to
e required t
e of the dye (
ffusion limit
ince quench
neficial whe
found using
with signific
as a function
e of quench
proach was R
calculations.
pherical pore
hed by the re
severe (Fig
E decreases w
quenched by
he electrolyt
reciably. In
diffuse furt
to achieve 9
(τ0) is increa
t. If the lifet
hing of the
n the lifetim
the
cant
n of
hing.
Ra =
es is
edox
g. 2-
with
the
te is
n the
ther,
90%
ased
time
dye
me is
34
2.3.9 Discussion
The calculations for the cylindrical and spherical pores suggest two strategies
for achieving excitation transfer efficiencies of over 90% in a dye sensitized solar cell
with relay dyes. The first approach is to find relay and sensitizing dye combinations
with moderately high Förster radii of 5 nm or more and relay dyes with short
fluorescence lifetimes to minimize the chance that they will be quenched before they
undergo energy transfer. A second strategy is to select a donor dye that is not easily
quenched by triiodide (kq ≤ 106 M-1s-1) and has a long lifetime of a microsecond or
more. In this case the relay dye and sensitizing dye can have a small Förster radius of
~2 nm and still undergo efficient energy transfer due to diffusion. It may be easier
adopting the first strategy and selecting relay dyes with short lifetimes since iodide
quenching is so highly efficient for most dyes. We previously adopted this approach
in selecting PTCDI as the relay dye which is highly fluorescent, (QD,0 = 90%) enabling
a Forster radius of 7.5 nm with the sensitizing dye TT1, and has a short lifetime, (τ0 =
4.8 ns) minimizing quenching 64. For the second approach, ytterbium complexes
appear to be the most promising for relay dyes of the lanthanide complexes as they
emit at 980 nm 69 and would efficiently undergo energy transfer to a near infrared
sensitizing dye.
In order to function in a DSC and harvest most of the incident photons, the
relay and sensitizing dyes additionally must have strong and complementary
absorption spectra 64. The relay dye must either be highly soluble in the DSC
electrolyte (typically acetonitrile) or have a high molar extinction coefficient. As an
example, dyes with a peak molar extinction coefficient of 50,000 M-1 cm-1 would need
a concentration of ~40 mM to absorb 90% of the light in a 10 μm thick film; however
dyes with twice the molar extinction coefficient (i.e. 100,000 M-1 cm-1) would require
half the concentration. The sensitizing dye should be able to pack densely on the
titania surface and have good injection into the titania.
The model presented in this paper can be extended with some modifications to
relay dyes in solid state dye sensitized solar cells. There is a great potential for energy
relay dyes to improve the efficiency of solid state DSC’s which are limited in
35
thickness to 2 μm and cannot absorb all of the incident light. Recently we have
demonstrated an improvement in the efficiency of a solid state DSC using a relay dye 70. Unlike in the liquid electrolytic DSC, diffusion of the relay dye is not possible.
Studies have suggested that the pores are not completely filled in solid state DSC and
that there are voids of ~40% of the pore volume in the center of the pores 71.
Consequently, there is the potential for energy transfer to be more efficient in the solid
state DSC since the relay dyes are not located in the center of the pore where energy
transfer to the pore wall is the least efficient. Incomplete pore filling may make it
difficult, however, to incorporate sufficient relay dye into the device to sufficiently
absorb the higher energy photons.
2.3.10 Conclusion
Here we presented a model for calculating the excitation transfer efficiency from
a general volume containing energy donors to a dense surface of acceptors, accounting
for the processes of diffusion and quenching. Using calculations for cylindrical and
spherical pores we have demonstrated that the energy transfer process can be over
90% efficient in a dye sensitized solar cell with dyes with reasonable properties. Near
unity excitation transfer efficiencies can be obtained using dyes combinations with a
relatively large Förster radius where the donor has a short lifetime to avoid quenching
effects or alternatively with a dye combination with a relatively small Förster radius,
provided that the donor has a long fluorescence lifetime and is not significantly
quenched by the electrolyte to enable diffusion.
2.4 Measuring Important Energy Transfer Parameters
The FRET radius, quenching by the electrolyte, pore size, and dye surface
coverage are all important parameters that determine the average excitation transfer
efficiency. This section describes how to measure these properties in order to
accurately determine the efficiency of energy transfer in dye-sensitized solar cells.
36
2.4.1 Measuring the FRET R0 of Fast Emitting Chromophores in
Solution
The FRET Radius can be measured in solution using time resolved
photoluminescence techniques. When the energy relay dye is placed in solution
without the presence of the sensitizing dye the fluorescence decay is modeled as
)/exp()( 00 tItI DD (eq. 2-21)
When the sensitizing dye molecules are added and no diffusion (i.e. PL lifetime of <10
ns) is involved the decay rate is given as
2/1
00
00 )/(/exp)( t
C
CtItI A
DADA
(eq. 2-22)
Where CA is the characteristic acceptor concentration in molecules/cm3 and C0 is
given by72:
1300 )
3
4( RC
(eq. 2-23)
Time resolved photoluminescence was used to determine the FRET R0. Time resolved
PL measurements were performed using a Time-Correlated Single Photon Counting
(TCSPC) system from PicoQuant. Solutions were excited with a pulsed laser diode,
(model LDH 485: 481nm, 70ps FWHM, 5MHz) detected with a single photon
avalanche diode (PDM 100CT SPAD) attached to a monochromator and processed by
a PicoHarp 300 correlating system.
As an example, a τ0= 4.8ns was measured using equation 2-21 for PTCDI (discussed
in section 3.1), and a FRET Ro of 7.5nm and 7.6nm was measured at concentrations of
3.15*10-3M TT1 and 1.58*10-3M TT1 respectively based on the decay shown in figure
2-6.
37
0 1 2 3 4 5 6 7 8 9 105.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
ln(P
L) (
arb)
t (ns)
Fit of 3.15*10-3M TT1
Fit of 1.58*10-3M TT1 Fit of OM TT1
3.15*10-3M TT1
1.58*10-3M TT1 OM TT1
Time Resolved Photoluminescence (PL) of PTCDI with varying concentrations of TT1
Figure 2-6. Time resolved photoluminescence of PTCDI with varying concentration of
TT1 in gamma butyrolactrone. PTCDI concentration was 10-4M.
2.4.2 Measuring the FRET R0 of Diffusive Chromophores in
Solution
In the microsecond time regime chromophores can diffuse more than 20 nm
greatly increasing energy transfer. Equation eq. 2-24 accounts for chromophore
diffusion and FRET.61
))/(2/(*)( 2/1 tBteIotI (eq .2-24)
4/32
34.31
447.51
x
xxB
/
*6
3/23/1
oR
tDx
(eq. 2-25)
Figure 2-7, shows the PL decay of N877 in SQ1; we obtain a reasonable diffusivity, D
= 2.9 * 10-6 cm2/s, when fitting the PL decay of N877 (discussed in section 4.3) with
SQ1 using equations 2-24 and 2-25 and fixing Ro = 6.3 nm.
38
0.0 2.0x10-6 4.0x10-6 6.0x10-6
e-5
e-4
e-3
e-2
e-1
e0
e1
e2
ln(P
L (a
rb))
time (sec)
FRET + Diffusion Model Experimental PL
Fixing Ro = 5.7nm
Ca = 7.5*1015 cm-3
t = 2.6s
Free Variable: D = 2.9*10-6 cm2/s
Figure 2-7. Photoluminescence (640 nm) decay of N877 in presence of SQ1 as
function of time. Experimental result was fitted with the diffusion model
2.4.3 Dynamic Quenching Theory and Measurements
Non-radiative decay, or quenching, of the excited energy relay dye is
competitive process with energy transfer and should be measured for each dye in order
to determine the effectiveness of energy transfer. Dynamic quenching affects the PL
lifetime of excited dyes and can be measured using time resolved photoluminescence
measurements described in section 2.4.2. It is important to note that the PL lifetime
reduction represents the dynamic quenching component caused by high dye loading;
however, time resolved PL spectroscopy does not account for static quenching (e.g.
large aggregates that may be non-emissive). In DSCs, excited dyes in solution can be
quenched via collisions with iodide/triiodide or by concentration quenching. The total
quenching of the dye is the product of electrolyte and concentration quenching.
Nonradiative decay in the DSC system is greatly increased due to the presence
of iodide/triiodide in the electrolyte. Iodide is a highly mobile ion that is known as a
“perfect quencher”, meaning that collisions with the ERD have a near unity
39
probability of quenching the excited state.72 Photoluminescence quenching can occur
via several different mechanisms included intersystem crossing or photoinduced
electron transfer.61 It was originally thought that intersystem crossing, which
converted the highly luminescent singlet state chromophore to a poorly emissive
triplet state was responsible for the quenching of organic dyes. However, early studies
showed that the bimolecular quenching constant of phosphorescent dyes (e.g. N877)
was similar to luminescent dyes.73 Photoinduced electron injection (i.e. electrons from
the excited dye transferred to the iodide) appears to be the most likelyquenching
mechanism, but more work should be performed to validate this hypothesis. Given the
high concentrations of triiodide in the DSC electrolyte, the quenching rate of
chromophores can be 20-2000 times greater than natural decay rate. Collisional
quenching of the PTCDI by triiodide is described by the Stern-Volmer equation eq. 2-
26,74
][100 QkPL
PLoq
(eq. 2-26)
where PL0 is the photoluminescence in the absence of a quencher, PL is the
photoluminescence for given quencher concentration [Q], τ0 is the natural
fluorescence lifetime, τ is the fluorescence lifetime for a given [Q], and kq is the
bimolecular quenching constant typically around 109-1010 M-1 s-1. Because the
bimolecular constant and the electrolyte concentrations are relatively fixed, a short τ0
is important for minimizing the fluorescence quenching.
Concentration quenching occurs due to aggregation of the dyes in solution and
by collisions between dyes. Concentration quenching is measured as a function of dye
loading in the solution. Concentration quenching is dependent upon the molecules
used; bulky (i.e. non-planar) dyes such as PTCDI do not exhibit concentration (section
3.1.2) while others such as DCM have a 1.75x decrease in PL lifetime (see section
3.2.2).
2.4.4 Por
The
Emmett, and
pore volum
nanoparticle
Figure 2-8.
roughness fa
The
volume and
fairly narrow
is between
implying a
possible ave
re Size Dis
pore size a
d Teller (BE
me and roug
es. Large par
BET Data fo
factor (‘RF’)
pore size d
d is shown in
w for the 14
18-26nm a
separation d
erage excita
stribution
and roughne
ET) method
ghness fact
rticles typica
or various Ti
.
distribution i
n figure 2-9.
4nm nanopar
and there ar
distance no g
ation transfer
40
in Titania
ess of titani
of Titania n
tor are high
ally have a lo
iO2 pastes. T
is determine
The pore siz
rticles. The f
re few pore
greater than
r efficiency
a Mesostru
ia films is
nanoparticles
hly depende
ower porosit
The inset sho
ed by taking
ze distributio
full width ha
es with diam
n 15nm. In o
we have ch
uctured El
measured u
s. As seen in
ent upon th
ty and rough
ows the poro
g the derivat
on centered
alf maximum
meters great
order to achi
hosen to use
lectrodes
using Bruna
n figure 2-8,
he size of
hness factor.
osity (‘por’)
tive of the p
around 22nm
m for the fig
ter than 30
ieve the high
e 14 nm and
auer,
the
the
and
pore
m is
gure
nm,
hest
d 17
nm diamete
DSCs typica
Figure 2-9.
2.4.5 Sur
The
TiO2 is imp
FRET rates
from 0.1 to
groups, TiO
which comp
The
films of a k
films are th
remove the
concentratio
r particles fo
ally use 20 n
Pore Volum
rface Conc
nanopore s
ortant becau
to each of t
1 dye/nm2 d
O2 particle s
pete for TiO2
sensitizing
known thickn
hen rinsed in
dye, the sol
on of dye and
or most of ou
nm diameter
me distributio
centration
ize and surf
use the overa
the surround
depending on
size and cry
2 adsorption
dye surface
ness and rou
n a basic so
lution is the
d converted
41
ur studies. It
particles.
on for variou
n of Sensiti
face concen
all transfer ra
ding sensitiz
n the molecu
ystal structu
sites but red
e concentrati
ughness facto
olution (e.g.
en measured
to mol/cm2
t should be n
us TiO2 nano
izing Dyes
ntration (CA)
ate is equal t
zing dyes. T
ule size, num
ure,75,76 and
duce dye agg
ion is measu
or are a dipp
1M KOH)
d using UV-V
(or dye/nm2
noted that hi
oparticle past
s in DSC
) of the sen
to the sum o
ypical CA v
mber and typ
the use of
gregation.
ured via des
ped in the d
in either wa
Vis to deter2).
gh performa
tes.
nsitizing dye
of the individ
alues can ra
pe of attachm
f co-adsorbe
sorption: tita
dye solution,
ater or DMF
rmine the m
ance
e on
dual
ange
ment
ents,
ania
the
F to
olar
42
3 Using Energy Relay Dyes Unattached to Titania in
Liquid DSCs
Here I present a new design where high energy photons are absorbed by highly
photoluminescent chromophores unattached to the titania and undergo Förster
resonant energy transfer to the sensitizing dye. This novel architecture allows for
broader spectral absorption, an increase in dye loading, and relaxes the design
requirements for the sensitizing dye. I demonstrate a 26% increase in power
conversion efficiency when using an energy relay dye with an organic sensitizing dye
in a liquid based DSC. I also show how to directly measure the average excitation
transfer efficiency. This system offers a viable pathway to develop more efficient
DSCs.
Using energy relay dyes has several important advantages. First, since the
attached dye only has to absorb light over a smaller spectral region, it can be chosen to
have a stronger and narrower absorption spectrum. Second, the SD can be red shifted
compared to the commonly used dyes since the energy relay dye can absorb higher
energy photons. Furthermore, it is possible to place multiple ERDs with
complimentary absorption spectra to tailor light absorption inside the device. Finally,
the ERD does not need to be attached to the titania surface and with no additional
processing steps can be mixed in very large concentrations inside the electrolyte. In
summary, the addition of energy relay dyes into the electrolyte makes the overall
absorption spectrum wider and stronger for the same film thickness. It is important to
note that the ERDs do not participate in the charge transfer or collection process and
thus do not require precise energy levels or specialized attachment groups.77 ERDs
should be designed to be soluble in and not greatly quenched by the electrolyte.
Incorporating long range energy transfer into the solid-state DSC will require ERDs
that avoid charge transfer into the hole transporter. The energy relay dye system is also
extremely useful for nanostructured systems (e.g. TiO2 nanotubes 78, ZnO nanorods 79)
that have less available surface area and thus poorer light absorption.
43
3.1 The PTCDI/TT1 System
3.1.1 PTCDI/TT1 Emission and Absorption Spectra
A previously reported80 derivative of perylene-3,4,9,10-tetracarboxylic diimide
(PTCDI), shown in Figure 3-1b, was synthesized (see methods section) for use as an
ERD. PTCDI is an ideal energy relay dye candidate because of its extremely high
photoluminescence efficiency (>90%), fast fluorescence lifetime (4.8ns), excellent
photo and air stability, and relatively strong absorption coefficient (50,000 M-1 cm-1 at
580nm).81 Its bulky alkyl phenyl substituents were designed to reduce chromophore
interactions between adjacent dye molecules in order to prevent aggregate formation
and reduction of fluorescence. A zinc phthalocyanine dye, TT1, shown in Figure 3-1c,
was chosen as the sensitizing dye for its high molar extinction coefficient of 191,500
M-1 cm-1 centered at 680nm.24 One would prefer a dye with a smaller energy gap, but
such dyes are not readily available yet with the necessary anchoring groups. When
attached to titania, the TT1 dye absorption broadens (as shown in Figure 3-1a) and
significantly overlaps the PL emission of the PTCDI. Given the absorption and
emission profile of the TT1 and PTCDI respectively, the FRET radius is estimated to
be 8.0nm. Time resolved photoluminescence measurements on solutions with varying
concentration of TT1 determined Ro to be 7.5-7.6nm (see section 2.4.1).
Figure 3-1.
(red dash d
Chemical st
3.1.2 PTC
We d
shows that
concentratio
with a kq
quenching.
nonradiative
greater than
experience c
PTCDI and
dot) in chlor
tructures of t
CDI Quench
determined t
the fluoresc
ons of the 1
of 3.17*10
For the
e decay rate
n the natura
concentratio
d TT1 prope
roform, and
the energy re
hing
the fluoresce
cence intens
1-methyl-3-p
010 and 0.67
electrolyte
e due to qu
l fluorescen
on quenching
44
rties. a, PTC
d TT1 absor
elay dye, PT
ence lifetime
sity and lifet
propyl imida
7*1010 M-1
used in th
uenching (kQ
nce decay ra
g.
CDI absorpti
rption (black
TCDI (b), and
e of the PTC
time are bo
azolium iod
s-1 respect
e DSC (0.6
QUENCH) is c
ate (kQUENCH
ion (blue), P
k) on titania
d sensitizing
CDI to be 4.8
th reduced
dide (PMII)
tively, indic
6M PMII,
calculated to
H =30 k0).
PTCDI emiss
a nanopartic
g dye, TT1 (c
8 ns. Figure
with increas
and I2 spe
cating dyna
0.05M I2)
o be ~30 tim
PTCDI did
sion
cles.
c).
3-2
sing
cies
amic
the
mes
not
Figure 3-2
photolumine
circles) and
is equivalen
The PTCDI
3.1.3 Mod
Figu
upon the p
calculated f
nonradiative
it has an ex
pore sizes in
and the quen
we simulate
and 63% for
2. Quench
escence is re
I2 (green sq
nt to the red
concentratio
deling ETE
ure 3-3 show
pore diamete
for the PTC
e decay half
xpected ETE
n the DSC m
nching rate c
e an average
r the spheric
ing of P
educed with
quares). The
duction in ex
on was 1*10
in the PTCD
ws how the av
er for cylin
CDI-TT1 DS
f life of only
E between 7
measured by
calculated ab
e excitation t
cal pores (see
45
TCDI by
h increasing
reduction in
xcitation life
0-4M in gamm
DI/TT1 DSC
verage excit
ndrical and
SC system.
y 0.15 ns (4.
76-87% in a
y the Brunau
bove from li
transfer effic
e section 2.2
electrolyte
concentratio
n photolumin
etime (τ0/τ)
ma-butyrola
C System
tation transfe
spherical p
Even thoug
8ns/31) whe
a 15nm pore
uer, Emmett
ifetime meas
ciency of 49
2).
e species.
on of PMII (
nescence (PL
shown as th
actone.
er efficiency
pores using
gh the excit
en placed in
e. Using the
t, and Teller
surements (k
9% for the c
The PTC
(half-filled b
L0/PL) by P
he red triang
y, ETE , depe
the parame
ted ERD ha
n the electrol
e distribution
r (BET) met
kQUENCH =30
ylindrical po
CDI
blue
MII
gles.
ends
eters
as a
lyte,
n of
thod
0 k0)
ores
Figure 3-3 M
diameter for
radius of 8.0
rate of 30k0
3.1.4 PTC
The
the energy
diameters b
and a rough
thick layer
printing, sin
dipped in a
hours and r
because PT
used solven
However, c
(70% vs. 80
Modeled ave
r spherical a
0 nm, conser
.
CDI/TT1 De
titania film
relay dye to
etween 22-3
hness factor o
of 400nm s
ntered at 450
a 1x10-5M s
rinsed in ace
CDI is sign
nts such as
hloroform b
0%) and low
erage excitat
nd cylindric
rvative dye c
evice Fabric
was compri
o the sensitiz
38nm, a film
of 97/µm. A
scattering pa
0˚C, and sub
solution of T
etonitrile.83
ificantly mo
s acetonitril
based electro
wer power co
46
tion transfer
cal pores. M
coverage est
cation and P
sed of 20 nm
zing dye. Th
m porosity of
A 10-µm-thi
articles (CCI
bsequently t
TT1 with 1
Chloroform
ore soluble i
le (<2mM)
olytes displa
onversion eff
efficiency a
Modeling resu
timate of 0.2
Performanc
m particles t
he 20 nm T
f 68% (witho
ck layer of 2
IC, HPW-40
treated in Ti
0mM cheno
m was chosen
in it (>50mM
and gamm
ayed lower i
fficiencies at
as a function
ults are based
2 dye nm-2, a
ce
to ensure clo
TiO2 particle
out the addit
20nm particl
00) was for
iCl4.82 The f
odeoxycholi
n as the ele
M) compare
ma-butyrolac
internal quan
t higher light
of pore
d on a Förste
and a quench
ose proximity
es produce p
tion of the d
les and a 5-µ
rmed via scr
films were t
c acid for f
ctrolyte solv
ed to commo
ctone (<2m
ntum efficie
t intensities.
er
hing
y of
pore
dye),
µm-
reen
then
four
vent
only
mM).
ency 84,85
47
These issues are caused by the reduced I3- concentration, lower solubility of useful
additives such as LiI and guanidinium rhodanide, and the lower dielectric constant of
chloroform (ε=5) compared to acetonitrile (ε=36).86-88 Devices without the ERD were
also made with acetonitrile based electrolytes and had similar device performance
compared to literature.24 The electrolyte contained 0.6M PMII, 0.05M I2, <0.04M
tertbutyl pyridine, < 0.01M LiI, and <0.02 guanidinium thiocyanate in chloroform. 13
mM of PTCDI was subsequently added before electrolyte filling of the DSC. The
preparation of the platinum counter electrode on FTO glass (TEC 15 Ω/, 2.2 mm
thick, Pilkington) is described in a previous report.66 Electrodes were sealed using a
25-µm-thick hot-melt film (Surlyn 1702, Dupont). A small hole was drilled in the
counter electrode and electrolyte was filled using a vacuum pump. It should be noted
that CHCl3 has a low boiling point and during electrolyte filling the concentration of
PTCDI inside the DSC invariably changed. Higher molar concentrations of PTCDI in
the electrolyte did not increase dye loading, but did result in clogging of the hole as
the PTCDI electrolyte gelled quickly. Sometimes you eat the bear and sometimes the
bear eats you. A precise determination of the true PTCDI concentration is beyond the
scope of this paper, but will be addressed in the subsequent section.
Figure 3-4 shows the photocurrent density-voltage (J-V) characteristics of
DSCs with and without the energy relay dye measured under AM 1.5G (100mW cm-2)
conditions. Devices containing no energy relay dye (0mM PTCDI) had power
conversion efficiencies (PCE) of 2.55% while devices with 13mM of PTCDI had a
PCE of 3.21%. The 26% increase in device performance is attributed to the increase in
short-circuit photocurrent density (JSC) caused by an increase in the EQE from 400-
600nm as shown in Figure 3-5a, while the Fill Factor and Voc remained relatively
unchanged (see insert in Figure 3-4). Devices made with PTCDI but without the
sensitizing dye were found to have very low photocurrent (Jsc < 42µA/cm2 and
PCE~0.01%), demonstrating that energy transfer to the sensitizing dye is necessary for
photocurrent generation by the ERD.
Figure 3-4.
PTCDI) and
Dash-dot lin
control devi
3.1.5 Min
A lo
(EQEERD) c
containing t
EQE enhanc
(0mM PTC
PTCDI beca
pathway an
absorption.
the product
efficiency, E
Photocurren
d without (0m
nes represent
ice (green).
nimum Boun
ower bound
can be calcu
the ERD and
cement has a
CDI). The Δ
ause light sc
nd at longer
The externa
t of the abso
ETE , and the
nt density-vo
mM PTCDI)
t the dark cu
nd ETE for
d for extern
ulated from
d the EQE o
a peak of 29
ΔEQE spectr
cattering is g
wavelength
al quantum e
orption effic
e internal qu
48
oltage (J-V) c
) energy rela
urrent for ER
PTCDI/TT
nal quantum
m the differe
of the contro
9.5% at 530n
rum does n
greater at low
hs (>550nm
efficiency o
ciency of th
uantum effici
characteristi
ay dye under
RD containin
T1 System
m efficiency
ence betwee
ol, ΔEQE, s
nm, which is
not perfectly
wer wavelen
m) the ERD
f the energy
he dye, the
iency of the
ics of device
r AMA 1.5 (
ng DSC (blu
y of the en
en the EQE
shown in Fig
s 8x greater
y match the
ngths increa
and SD com
y relay dye
average exc
control dev
es with (13m
(100mWcm-2
ue) and the
ergy relay
E of the dev
gure 3-5b.
than the con
e absorption
asing the opt
mpete for l
is equivalen
citation tran
ice.
mM 2).
dye
vice
The
ntrol
n of
tical
ight
nt to
nsfer
A m
calculated b
Using the dy
10.3% by T
average ETE
Figure 3-5.
efficiency v
device (0mM
energy relay
generated by
minimum bo
by assuming
ye absorptio
TT1. If the IQ
E of 47% is
Light harves
versus wavel
M PTCDI).
y dye to sens
y PTCDI wa
ound for th
g that there
on profiles, th
QE is assum
calculated.
sting charact
ength of DS
b, EQE add
sitizing dye a
as 29.5% at 5
49
he average
is complete
his correspo
ed to be equ
teristics of th
C with energ
ition (black
and PTCDI
530nm.
excitation t
e light abso
nds to ηabs,ER
ual to the pea
he ERD DSC
gy relay dye
squares) cau
absorption (
transfer effi
orption at th
RD=89.7% fr
ak EQE (70
C. a, Externa
e (PTCDI) an
used by FRE
(blue circles)
ficiency can
he ΔEQE pe
rom PTCDI
%), a minim
al quantum
nd a control
ET from the
). Peak ΔEQ
n be
eak.
and
mum
QE
3.1.6 PTC
The
PTCDI conc
75% averag
and 100mM
mA/cm2. As
acetonitrile
Figure 3-6.
3.1.7 PTC
All
Reactions w
Reactions w
silica gel pla
230-400 me
All reagent
purification
189 and 9(1
CDI/TT1 Co
PTCDI/TT1
centration to
ge excitation
M PTCDI c
ssuming a V
based electr
Estimated E
CDI Synthes
glassware w
were carrie
were monito
ates. Flash c
esh. Solvent
ts were use
. The follow
10),16(17),23
onclusions
1 system has
o 100 mM. F
transfer effi
concentration
Voc of 617m
rolytes24 the
EQE of PTCD
sis
was dried o
d out unde
ored by thin
column chro
ts were remo
ed as recei
wing compou
3(24)-Tri-ter
50
s the potenti
Figure 3-6 s
iciency, 10 µ
n. Integratin
mV and a fill
predicted po
DI/TT1 syste
overnight in
er nitrogen
n layer chrom
omatography
oved with a
ved from c
unds were m
rt-butyl-2-ca
al to increas
shows an est
µm thick film
ng the EQE
l factor of 0.
ower conver
em.
n an oven o
using stan
matography
y was perform
rotary evap
commercial
made by prev
arboxy-5,28
se to >5% by
timated EQE
m (with no l
E produces
.75 that was
sion efficien
or by flame
ndard Schle
using Wha
med using M
orator at asp
suppliers w
viously repor
:14,19-diim
y increasing
E plot assum
light scatterin
a JSC of 1
achieved us
ncy is 5.7%.
e prior to u
enk techniqu
atman® 250
Merck silica
pirator press
without fur
rted procedu
mino-7,12 :21
g the
ming
ng),
12.3
sing
use.
ues.
μm
gel,
ure.
rther
ures:
1,26
51
dinitrilotetrabenzo[c,h,m,r]tetraazacycloeicosinator-(2_)-N29,N30,N31,N32 zinc (II)
(TT1).24
NMR spectra were recorded in CDCl3 with a TMS standard using a Bruker AVB-400
spectrometer. 13C NMR was recorded at 100 MHz using 1H decoupling. Mass
spectrometry and elemental analysis data were recorded by staff members at the UC
Berkeley mass spectrometry facility.
N,N’-di(2,5-diisopropylphenyl)-1,6,7,12-tetra(4-tert-butylphenyoxy)- perylene-
3,4,9,10-tetracarboxylic diimide 2:
Procedure
A solution of 2 (3.00 g, 3.54 mmol), 4-tert-butylphenol (2.66 g, 17.7 mmol),
and potassium carbonate (2.92 g, 17.7 mmol) in anhydrous N-methylpyrrolidinone (50
mL) was stirred at 130 °C for 16 hours. The solution was rapidly poured into 1M HCl
(200 mL) and the resulting precipitate was isolated by vacuum filtration and washed
repeatedly with water. The precipitate was dissolved in chloroform and extracted
twice with water. The chloroform layer was then dried over Na2SO4 and concentrated.
Purification by flash column chromatography (eluent: 40%-55% dichloromethane in
hexanes) yielded 1.55 g of red solid (34% yield). A portion of this product was further
purified by recrystallization: 1.00 g of product was dissolved in dichloromethane (100
mL) and placed in a 1000 mL graduated cylinder. Methanol (200 mL) was carefully
layered on top of the dichloromethane, and the two layers were allowed to mix slowly
52
over 1 week. The resulting red crystals were isolated by filtration and dried under
vacuum, yielding 740 mg of red product.
Characterization
m.p. > 300 °C. 1H NMR (400 MHz, CDCl3, δ): 8.29 (s, 4H), 7.42 (t, J = 7.8
Hz, 2H), 7.22-7.28 (m, 12H), 6.87 (dt, J = 8.8 and 2.5 Hz, 8H), 2.71 (m, J = 6.8 Hz,
4H), 1.28 (s, 36H), 1.13 (d, J = 6.8 Hz, 24H). 13C NMR (100 MHz, CDCl3, δ): 163.55,
156.12, 152.99, 147.55, 145.82, 133.44, 130.88, 129.60, 126.88, 124.09, 122.89,
120.97, 120.45, 120.39, 119.41, 34.58 31.66, 29.28, 24.24. FTIR (film on NaCl): ν =
2963, 2870, 1709, 1675, 1588, 1505, 1406, 1340, 1285, 1209, 1175 cm-1. HRMS
(FAB+, m/z): calcd for C88H91N2O8, 1303.6775; found, 1303.6786. Anal. calcd for
C88H90N2O: C 81.07, H 6.96, N 2.15; found: C, 80.08; H, 6.97; N, 2.09.
3.2 The DCM/TT1 System
In the PTCDI/TT1 DSC system only a minimum bound ETE of 46% could be
estimated because of the uncertainty in determining ηABS,ERD due to light scattering
caused by large TiO2 nanoparticles and the inability to accurately measure the
concentration of the dye because of rapid evaporation of the chloroform electrolyte
during the electrolyte filling process.64 In this section, the ETE is quantified by
designing an experiment to accurately measure the light absorption by the energy relay
dye inside of transparent TiO2 electrodes and by developing a new ERD system that is
more soluble in higher boiling point electrolyte solvents. A commercially available
laser dye 4-(Dicyanomethylene)-2-methyl-6-(4-dimethylaminostyryl)-4H-pyran
(DCM)90 was used as the ERD inside of the less volatile, conventional acetontitrile
electrolyte to demonstrate an ETE of 96%. I also demonstrate increased performance
in the optimized device architecture and find that the performance is limited by the
relay dye’s absorption and its moderate solubility in the electrolyte.
3.2.1 DCM
DCM
with a peak
figure 3-7a,
and a short
was chosen
coefficient w
from DCM
Figure 3-7.
relay dyes in
(green). Che
3.2.2 DCM
Figu
concentratio
M/TT1 Emi
M is a stron
k molar extin
a high phot
t photolumin
as the sens
with a peak
to TT1 is es
(a) Absorpti
n acetonitrile
emical struct
M Quenchin
ure 3-8a sho
on for the co
ission Absor
ng ERD cand
nction coeff
toluminesce
nescence life
sitizing dye b
of 191,500 M
timated to b
ion (blue) an
e:valeronitri
tures of DCM
ng
ows the ph
ommonly ele
53
rption Spect
didate becau
ficient of 44
nce quantum
fetime (~2ns
because it h
M-1 cm-1 cen
e 6.85 nm.
nd Emission
ile (85:15 vo
M (b) and TT
hotoluminesc
ectrolyte mix
tra
use it has a
4,900 M-1 cm
m efficiency
s). The zinc
has an extrem
ntered at 680
n (red dash-d
ol) with TT1
T1 (c).
cence lifetim
xture of acet
broad absor
m-1 at 460 n
y of 44% in
c phthalocya
mely high m
0 nm.24,64 Th
dot) spectra o
absorption
me of DCM
tonitrile:vale
rption spectr
m, as shown
acetonitrile,
anine dye, T
molar extinc
he Föster rad
of DCM ene
spectra on T
M versus D
eronitrile (85
rum
n in
,91,92
TT1,
tion
dius
ergy
TiO2
CM
5:15
vol). At rel
constant. In
2.1 ns to 1.2
DCM satura
equivalent t
acetonitrile:
Elec
with excited
which consi
0.05 M gua
high concen
nanosecond
requires ER
lifetime of
compared to
Figure 3-8b
quenching p
4.88 (5 mM
than natural
latively low
creasing the
2 ns, a 1.75x
ation limit, o
to the DCM
valeronitrile
ctrolyte quen
d ERDs. TT
ists of 0.6 M
anidinium th
ntrations, ea
d. Therefore
RDs with sh
1.2-2.1 ns
o the natural
b. Combini
produces an
M DCM with
l decay rate (
DCM conc
DCM conce
x reduction in
or the point w
mass divide
e mixture.
nching is ma
T1 devices a
M 1-butyl-3
iocyanate, 0
ach ERD m
FRET mus
hort (<10 n
and has an
l decay rate
ing the effe
n overall dyn
h [M1] = 10
(0.1 mM DC
54
centrations (
entration abo
n PL lifetim
when the co
ed by the sol
ainly caused
are optimize
-methylimid
0.28M tert b
molecule coll
st occur at t
ns) photolum
electrolyte
depending o
fects of bot
namic quenc
0%) and 8.0
CM with [M1
(10-4 M to
ove 1 mM re
me for the ace
oncentration
lvent volume
d by the iod
ed using an
diazolium io
butylpyridine
lides with a
the sub-nan
minescence
quenching
on the DCM
th concentr
ching rate (
00x (22 mM
1] = 0%).
10-3 M) the
educes the P
etonitrile bas
inside of the
e, is less tha
dide/triiodide
electrolyte
odide (BMII)
e, and 0.04 M
an ion more
osecond tim
lifetimes. D
rate betwee
M concentrati
ration and
(Qdynamic) of
M with [M1]
e PL lifetim
PL lifetime fr
sed system.
e solution is
an 18 mM in
e ions collid
known as M
), 0.025 M
M I2. Given
e than once
me scale, wh
DCM has a
en 4.0 and 4
ion as shown
M1 electro
DCM betw
= 100%) fa
me is
from
The
not
n the
ding
M1,
LiI,
n the
per
hich
PL
4.9x
n in
olyte
ween
aster
55
Figure 3-8. (a) Photoluminescence lifetime of DCM with various concentrations of
ERD using an 85:15 mixture by volume of acetonitrile and valeronitrile. (b)
Photoluminescence quenching caused by various concentrations of M1 electrolyte
3.2.3 Modeling ETE in DCM/TT1 System
Showa Denko 17-nm-diameter titania nanoparticles were used in this study
because they have the smallest average pore size, with an average pore diameter of
19.5 nm with a standard deviation of 5.4 nm. Given the pore distribution in figure 2-9,
critical radii of 9 nm and 7nm are required to achieve ETE s above 90% assuming
cylindrical and spherical pore geometries respectively. The critical radius for the
DCM/TT1 DSC is 8-9 nm depending on the quenching rate; given the pore size
distribution of the 17 nm particles. Figure 3-9 shows how the excitation transfer
efficiency, of DCM based on the FRET Ro (6.85nm), TT1 surface coverage (0.389
dye/nm2), and quenching rate (8.54x). DCM ERDs placed in 17nm particle films with
the BET pore distribution have an estimated ETE of 94% and 97% for cylindrical and
spherical geometries respectively. The pore size range of 17 nm particles is
highlighted in blue in Figure 3-9. The scattering layer has larger pore sizes that have
significantly lower ETE (<20%) which is indicated by the highlighted (red) region.
Figure 3-9.
geometries.
3.2.4 DCM
In an
an optimize
scattering ti
thick films
of large (C
and in gene
allows for i
circuit volta
The
was used w
The control
density of 8
shown in fi
Excitation
M/TT1 Devi
n attempt to
ed TT1 dev
itania nanop
made from
CIC, HPW-4
eral have lo
increased thi
age.94
8 + 4µm de
with M1 elect
l device (0m
8.32 mA/cm
gure 3-10. T
Transfer Ef
ice Fabricat
o maximize t
vice architec
articles. Con
either 17 or
400, 400nm
ower recom
ickness and
evice (8µm o
trolyte using
mM) was 3.
m2, open-circ
The control
56
fficiency fo
tion and Pe
the device p
cture, which
nventional D
r 25-nm-diam
) TiO2 parti
mbination rat
light absorp
of 17nm par
g a mixture o
5% efficien
cuit voltage
device is th
r DCM as
rformance
performance,
h contains a
DSCs are typ
meter particl
icles.93 The
tes, due to
ption withou
rticles with 4
of acetonitri
nt at 1 sun w
of 582 mV
he most effic
the ERD fo
, we placed
an additional
pically comp
les with a 4
larger partic
lower surfa
ut large loss
4µm of the s
ile:valeronitr
with a short
V, and fill fa
cient TT1 D
or various p
DCM inside
l layer of l
prised of 8-µ
-µm-thick la
cles scatter l
ace area, wh
ses in the op
scattering lay
rile (85:15 v
t-circuit cur
actor of 0.72
DSC reported
pore
e of
ight
µm-
ayer
ight
hich
pen-
yer)
vol).
rrent
2 as
d in
literature w
When 22mM
increased to
current den
mV) and fil
the spectrum
Usin
primarily b
portion of t
reducing the
to increase
device perfo
light is diff
scattering o
the pore size
For efficien
consequentl
transfer effi
Figure 3-10
acetonitrile
devices with
with the elec
M of DCM w
o 4.51%. Th
nsity from th
ll factor (0.7
m where DC
ng large nan
ecause light
the solar spe
e nanopartic
scattering w
ormance. Lig
fracted forw
ccurs predom
es inside of t
nt energy tra
ly light abso
ciency and h
. (a) The EQ
based electr
h (22 mM D
ctrode and M
was placed i
he improvem
he relay dye
72) remained
M absorbs i
noparticles d
t scattering
ectrum, whe
cle size (e.g.
where the en
ght is largely
ward rather th
minantly ins
the scatterin
ansfer, the E
orbed by the
hardly contri
QE of the 22
rolyte. (b) P
DCM) and w
57
M1 electrol
nside of the
ment is due
e (10.61 mA
d relatively
s 40%.
did not impr
from 400-n
ere only the
200-250 nm
ergy relay d
y scattered i
han reflecte
side of the l
ng layer are r
ERDs must b
e ERDs insid
ibutes to the
2mM DCM D
Photocurrent
without (0 m
lyte conside
M1 electrol
to a 27% in
A/cm2). The
unchanged.
rove light h
nm-particles
sensitizing
m TiO2 parti
dye absorbs
in the lateral
ed backward
large nanopa
roughly on th
be within th
de the large
photocurren
DSC with 8+
density-vol
mM DCM) e
ered optimal
lyte the devi
ncrease in t
e open-circu
The EQE in
harvesting by
s occurs ma
dye absorb
icles) in the
is not expec
l and forwar
ds), which m
articles.95 It
he order of t
he RC of the
pores has a
nt.
+4µm archit
ltage (JV) ch
energy relay
l for the dy
ice performa
the short cir
uit voltage (
n the portion
y the relay
ainly in the
bs. Furtherm
scattering la
cted to impr
rd direction
means that l
is believed
the particle s
titania surfa
a low excita
tecture using
haracteristic
dye under A
ye.24
ance
rcuit
(590
n of
dye
red
more,
ayer
rove
(i.e.
ight
that
size.
face;
tion
g an
s of
AM
58
1.5G (100 mW/cm2). Dashed lines represent the dark current for ERD containing and
control devices.
3.2.5 DCM/TT1 Conclusion
The DCM/TT1 DSC system unequivocally shows that extremely high (>95%)
average excitation transfer efficiencies can occur using energy relay dyes in liquid
based DSCs. DCM/TT1 system is also the first example of ERDs enhancing the power
conversion efficiency of an optimized DSC system (i.e. acetonitrile based electrolyte
system). However, this work did not demonstrate efficiencies >5% due mainly to low
dye loading inside the solution and potentially inside the titania film itself. I would
also like to note that this result should not be generalized to all potential DSC systems.
My initial work on ERD DSC studies has focused on the energy relay dye, but a
further understanding of how orientation of the sensitizing dye (e.g. dye packing
perpendicular or parallel to the titania surface) affects the ETE is required to better
determine if these results are more widely applicable.
3.3 Directly Measuring the Excitation Transfer Efficiency in
Liquid Based DSCs
The ETE can be experimentally calculated from equation 2-2 by measuring the
EQE contribution from the ERD (EQEERD), light absorption by ERD inside of the
titania (ηabs,ERD), and the internal quantum efficiency of the system. Electrodes
comprised of 5.4-µm-thick titania films were fabricated using 17 nm TiO2
nanoparticles. The 400-nm-diameter titania particles that are often used in DSCs to
scatter light were not employed so that the absorption could be more easily quantified.
Films were dyed for four hours in 1*10-4 M TT1 with 10 mM chenodeoxylic acid in
ethanol. Various amounts of DCM were mixed into the M1 electrolyte. All other
aspects of titania paste preparation and DSC fabrication and testing are the same as
reported in literature.64,66
3.3.1 Mea
The
where the D
measuremen
films at 680
absorption
approximate
IQE ranged
Figure 3-11
(5.4µm, 17
External Qu
3.3.2 Mea
The
from the EQ
EQE betwee
of 14.7%, 2
respectively
the DCM c
dynamic qu
increase in
asuring the I
IQE was de
DCM does n
nts indicate
0nm. An int
loss cause
ely 10-13%
from 85-87%
. (a) Extern
nm particle
uantum Effic
asuring the E
EQEERD is
QE of the co
en 400-550n
22.9%, and 2
y. The EQEE
concentration
uenching is
the dynam
Internal Qu
etermined by
not absorb, b
that <2% o
tegrating sph
ed by the
at 680nm.
%.
al Quantum
es) with var
ciency comp
External Qu
determined
ontrol (0mM
nm is attribu
28.2% for D
ERD does not
n, molar ext
expected to
mic quenchin
59
uantum Effi
y dividing th
by the fracti
of the light
here was use
front cont
For TT1 ba
Efficiency
rying conce
ared to contr
uantum Eff
by subtracti
M DCM), as
uted to the D
DCM concen
t appear to s
tinction coe
increase du
ng rate from
iciency
he EQE at th
ion of light
is transmitt
ed to determ
tact (Hartfo
ased DSCs s
of DSC of t
entrations of
rol (0mM) v
ficiency con
ing the EQE
shown in fi
DCM photor
ntrations of 5
scale with th
efficient, and
ue to concen
m 4.88x to
he peak wav
absorbed by
ted through
mine that the
ord, Tech
shown in fi
transparent
f DCM. (b)
versus DCM
ntribution fr
E of ERD co
figure 3-11b.
esponse with
5.5mM, 11m
he Beer-Lam
d film poro
ntration quen
8.00x is n
elength of T
y TT1. UV-
the dyed T
e reflection
15 glass)
gure 3-11a,
TiO2 electro
Change in
concentratio
rom ERD
ontaining DS
. The chang
h a peak ΔE
mM, and 22m
mbert law gi
sity. Altho
nching, a sm
not expected
TT1,
-Vis
TiO2
and
are
the
odes
the
on.
SCs
e in
EQE
mM
iven
ough
mall
d to
60
appreciably change (<5%) the ETE and cannot account for the nonlinear
improvement in EQE.96 There is also a small (<5%) ΔEQE from 600-700, which is
associated to slight differences in TT1 aggregation.83 The observed increase in the
EQE using transparent TiO2 films is significant, but not large enough to produce high
efficiency devices due to insufficient absorption of light by the relay dye.
3.3.3 Measuring Light Absorption by ERD
In principle one can calculate the ERD absorption based on the ERD
concentration in the prepared electrolyte and the titania film thickness and porosity. It
is best, however, to measure nabs,ERD in case the concentration in the pores is not the
same as outside the titania film and because it is difficult to precisely determine the
pore volume.82,97 Direct measurement of ηabs,ERD inside the pores is not possible
because the surlyn spacer is thicker than the TiO2 film as shown in figure 4a. The ERD
containing electrolyte above the titania surface can absorb a significant fraction of the
light. The ηabs,ERD can be determined by comparing the differences between the
optical density of the electrolyte filled region only (ODspacer) versus the optical density
of the TiO2 region (ODTiO2+spacer).
The ηabs,ERD can be determined by comparing the differences between the
optical density (OD) of the electrolyte filled region only (ODspacer) versus the optical
density of the TiO2 region (ODTiO2+spacer), represented by equations 3-1 and 3-2.
spacerERDERDspacer dCOD (eq. 3-1)
)(222 TiOspacerERDERDspacerTiOTiO ddCODOD (eq. 3-2)
The ODspacer is first measured to determine the thickness of the surlyn gasket.
The spacer thickness can vary from 20-23 µm depending on the amount and duration
of pressure applied during the 175ºC sealing step. The optical density of the ERD
inside the TiO2 (ODTiO2) was determined by measuring the ODTiO2+spacer and
subtracting the absorption caused by the spacer behind the TiO2, as shown in equation
3-2. Thick (8.8-9.2 µm) TiO2 films were dipped in 10mM chenodeoxychlic acid for
61
four hours to mimic the surface properties of the film. The ηabs,ERD was estimated by
scaling the optical density of the thick film to the film thickness used (5.4 µm) for the
EQE measurements. To prevent the possibility that the ERD concentration in solution
could be drastically changed by potential differences in solubility of ERD inside the
pores, a large gasket was used to surround the titania film to ensure that the volume
inside of the TiO2 film represented less than 10% of the volume of the enclosed space.
Reflection and absorption loss caused by the front contact at peak ERD absorption is
10-13% (Figure S5, Supporting Information).
Based on this technique, the peak ηabs,ERD was determined to be 16.7%, 27%,
and 32-39% for 5.5 mM, 11 mM, and 22 mM DCM concentrations, as shown in figure
3-12b. The deviation in ηabs,ERD (represented by the error bars) is due to slight
variation in film thickness within the samples (2-3%) and the variations in UV-Vis
measurements, which is <5% for low to moderate concentrations. At higher DCM
concentrations (>18 mM), the standard deviation becomes greater due to the
significant variation of the ODspacer due to the sensitivity limit of the UV-Vis detector
with a small (1 x 1mm) aperture. This technique can only be applied to systems with
low overall ERD absorption; for devices with higher light absorption this method
cannot be used because the sensitivity limit of the UV-Vis detector is reached.
As shown in figure 3-12b, the ERD absorption scales with the EQEERD.
Because the exact porosity is unknown it is only possible to make qualitative
statements about the DCM concentration inside of the pores versus DCM
concentration in solution. At lower concentrations (i.e. 5.5 mM) the ERD
concentration inside the nanopores (estimated 5.4 mM and 6.5 mM for an estimated
porosity of 0.6 and 0.5 respectively) are equivalent or slightly higher than the ERD
concentration in solution. The DCM molecule does not contain carboxylic or
phosphonic acid groups typically used to attach to the TiO2. However, DCM may
physisorb to the TiO2 due to electrostatic forces98 or steric reasons99 and thereby
increase the dye loading inside of the film. At higher concentrations the estimated
ERD absorption (ηabs,ERD = 52% for cERD = 22 mM, p = 0.6, L= 5.4µm) is higher than
the measured ηabs,ERD of 32-39%. It is possible that the concentration of DCM inside of
small pores
noticeable p
several hou
which are n
distribution
Figure 3-12
absorbed by
(red square
efficiency (
96%.
is reduced
precipitate fo
rs. It is also
not able to
of narrow (<
. (a). Schem
y the ERD in
s) versus pr
ETE ) versu
due to aggre
formation at
o possible th
penetrate t
<10nm) pore
matic of the E
nside of the
redicted ER
us concentrat
62
egation. Dye
the bottom
hat small ag
the mesopor
es.
ERD measur
TiO2 pores.
RD concentr
tion; the ET
e solutions w
of the vial
ggregates exi
rous membr
rement to det
(b) EQEERD
ration. (d) A
TE average o
with more th
when allow
ist in more
ranes that c
termine the
D (black circl
Average exc
over three co
han 18 mM
wed to settle
dilute soluti
contain a sm
amount of l
les) and ηabs
citation tran
oncentration
had
for
ions
mall
ight
s,ERD
nsfer
ns is
3.3.4 Exci
The ETE a
ηabs,ERD data
with ETE m
respectively
systems tha
>89%.32
Table 3-1 A
for 5.5mM,
3.3.5 Effe
It sh
contribute t
chemically
colliding wi
on the titan
exhibit a pe
system. How
itation Tran
averaged ove
a from transp
modeling es
y. These hig
at covalentl
Average Exci
11mM, and
ects of Direc
hould be n
to photocur
attach to the
ith the titani
ia and 5.5 m
eak EQE th
wever, deter
nsfer Calcul
er three DC
parent films
stimates of 9
gh excitation
y attach ER
itation Trans
22mM conc
ct ERD elect
noted that a
rrent. Altho
e titania, DC
ia surface. C
mM and 20 m
hat is appro
rmining the e
63
lations for D
CM concentr
s as shown ta
94% and 97
n transfer ef
RDs to TiO
sfer Efficien
centrations o
tron transfe
alternative p
ough DCM
CM can direc
Control DSC
mM concen
ximately 35
exact contrib
DCM/TT1 S
rations is 96
able 3-1. Th
7% for cylin
fficiency va
O2, which h
ncy estimates
of DCM.
er in ERD/D
pathways ex
does not h
ctly inject el
Cs prepared
ntrations of D
5% of the E
bution cause
System
6% based on
he ETE valu
ndrical and
alues compar
have obtain
s based on m
DSC System
xist for ex
have anchor
lectrons into
with only ch
DCM in the
EQEERD for
ed by direct
n the EQE
ues agree nic
spherical po
re well to D
ned an ETE
measured val
m
cited ERDs
ring groups
o the TiO2 w
henodeoxyc
M1 electro
the DCM/T
injection to
and
cely
ores
DSC
E of
lues
s to
s to
when
chlic
olyte
TT1
the
64
ETE is challenging because the excited ERDs near the surface are highly likely to
undergo FRET to a sensitizing dye before charge injection. For example, an excited
DCM molecule separated by 1 nm from a TT1 acceptor will undergo FRET on the 20
fs timescale,100 which is considerably faster than electron injection which occurs on
the order of 100 ps for organic dyes such as TT1.24 Therefore, although a portion of
the EQEERD is likely caused by excited ERDs colliding on the surface in the
DCM/TT1 system, it is unlikely that the ETE is artificially increased due to direct
charge injection.
To determine if direct injection from the ERD was possible, 5.4µm transparent
TiO2 films were fabricated and dipped in 10mM chenodeoxylic acid for four hours to
mimic the surface properties of the DCM/TT1 DSCs. Low concentration (5.5mM) had
a Jsc of 0.65 mA/cm2 (versus 1.14 mA/cm2 for the DCM/TT1 system) with a 5% EQE
maximum (versus 14.7% for the DCM/TT1 system), as shown in figure S9. The
higher concentration (22mM) ERD control had a Jsc of 1.14 mA/cm2 (versus 1.74
mA/cm2 for the DCM/TT1 system) and a 10% EQE maximum (versus 28.2%) for the
DCM/TT1 system. The discrepancy in ratios between DCM only and DCM/TT1 DSC
Jsc versus EQEERD values comes from the photocurrent contribution caused by the
TiO2 at wavelengths less than 450nm (see section 3.3.6) and to a small extent because
DCM only devices do not compete with TT1 to absorb light at wavelengths greater
than 600nm and less than 400nm. The maximum contribution from direct injection is
approximately 34-35% when comparing the peak EQEERD values for DCM only and
the DCM/TT1 system.
65
350 400 450 500 550 600 650 700 7500
5
10
15
20
25
30
35
40
45
50
EQ
E (
%)
Wavelength (nm)
22mM DCM 5.5mM DCM
Figure 3-13. External Quantum Efficiency for 5.4um thick transparent films covered
in chenodeoxylic acid with varying concentrations DCM.
3.3.6 Optical Losses Related to the Front Contact of the DSC
Optical losses in the DSC can be reasonably approximated by examining the
reflection losses at the incident surface and the absorption loss caused by the FTO
layer and TiO2. There is no anti-reflective coating on the surface of the DSC, which
results in a 4% reflection loss off of the front surface. An integrating sphere was used
to determine the overall absorption contributions caused by the fluorine doped tin
oxide (FTO) and TiO2. At 680nm the FTO absorbs roughly 6.7% with the FTO and
TiO2 layers absorbing 8.8%. The absorption for the FTO+TiO2 layers is fairly constant
from 500 to 700nm where ERD absorption occurs as shown in figure 3-14.
66
300 400 500 600 700 800 900 10000.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
FTO + 17nm TiO2 FTO
Light Absorption Profile of FTO andFTO + ~5m of 17nm TiO2 particles
* System was blanked to air.
Figure 3-14. Absorption Spectra of FTO and FTO covered with 17nm particles.
3.4 Future Outlook
Given the measured ETE , development of DSCs with power conversion
efficiencies of >14% is possible by using a series of ERDs and sensitizing dyes that
absorbs from 350-900nm. The realization of extremely efficient DSCs will require the
research and development of sensitizing dyes that can absorb strongly in the infrared
as well as highly absorbing ERDs that transfer energy to NIR absorbing NIR
sensitizing dyes. This section will briefly discuss the current status in these areas and
prospects for future improvement.
3.4.1 Near-Infrared Sensitizing Dyes
Future near infrared sensitizing dyes will not be required to absorb as broadly
and because the Förster radius is dependent on the emission/absorption overlap
multiplied by the wavelength to the fourth power (λ4 in equation (2-2)) it may not need
to absorb as strongly. However, the SD will need to have excellent charge injection
properties. T
and the Ner
redox coupl
To date onl
synthesized
properties.10
were prone
and also had
a cyanine ba
which requi
would be i
molecule an
that would a
based naph
coordinating
switching to
490,000 M-
more refine
obtained. N
DSCs, whil
DSC system
Figure 3-15
NK6037 and
The potentia
rnst potentia
le and 100-2
ly two NIR
that have 01,102 Howev
to both hea
d very low o
ased dye tha
ired using L
improved by
nd 2) remove
allow for rap
hthalocyanin
g group to
o three Si ce1 cm-1.103 W
ed versions
IR sensitizin
le energy re
m without aff
5. Near-Infra
d b) Si naph
al difference
al of the ele
00 meV for
R absorbing
peak abso
ver neither o
avy aggregat
open-circuit v
at has a LUM
Li free electr
y 1) remov
e the chlorin
pid electron
e dye woul
COOH liga
enter which
We are curren
of these d
ng dyes mus
lay dyes ma
fecting the V
ared Sensitiz
hthalocyanine
67
e required b
ectrolyte is
the solid sta
sensitizing
orption >72
of these dye
tion (limiting
voltage (e.g.
MO level ver
rolyte to hav
ving one CO
ne to increas
injection in
ld be enhan
ands at the
would incre
ntly collabor
dyes to see
st be develop
ay be able t
Voc or fill fa
zing Dyes su
e compound
between HO
about 300m
ate hole cond
dyes shown
20 nm an
es were part
g the surfac
. <450 mV).
ry close to th
ve good inje
OOH group
se the LUMO
normal (e.g
nced by 1)
e end of the
ease the mol
rating with s
if high ope
ped to exten
to fill in opt
actor.
uccessfully
d with axial a
OMO of the
meV for the
ductor to reg
n in figure 3
nd good ch
ticularly wel
e concentrat
. The NK603
he conductio
ection yields
p to create
O level of th
g. Z960) elec
) switching
e phenyl ri
lar extinctio
synthetic che
en-circuit v
nd the spectra
tical gaps o
incorporated
anchoring gr
sensitizing
iodide/triiod
generate the
3-15 have b
harge injec
ll designed
tion of the d
37 compoun
on band of Ti
s. The NK6
an asymme
he dye to lev
ctrolytes. Th
from an a
ings and 2)
on coefficien
emists to cre
voltages can
al absorption
of the sensiti
d into DSCs
roup.
dye
dide
dye.
been
tion
and
dye)
nd is
iO2,
6037
etric
vels
e Si
axial
by
nt to
eate
n be
n of
ized
s, a)
68
3.4.2 Energy Relay Dyes in Liquid DSCs
There are many available fluorophores including quantum dots currently used
for biomedical imaging that have the potential to be used as ERDs 72 and it may be
possible to design ERDs that are minimally quenched by triiodide. Candidates for
ERDs should be fast emitter to reduce quenching by the triiodide and have moderately
high photoluminescence quantum efficiency (>20%). An important question is
whether a high ETE can be obtained for a variety of dyes in the DSC architecture.
Many organic fluorophores have sufficiently fast PL lifetimes (<10ns) to experience
relatively low electrolyte quenching and PL quantum efficiencies greater than 25%,
which may be sufficient when using highly absorbing organic sensitizing dyes with
strong emission/absorption overlap for strong FRET.72 Developing ERDs that provide
>75% EQEERD will require either a significant increase in the molar extinction
coefficient or in the solubility of the ERDs without significant concentration
quenching. It is also possible to place multiple energy relay dyes inside the electrolyte
that are chemically different to increase dye loading and light absorption. Future
efforts should focus on developing multiple energy relay dye systems that absorb most
of the light inside of 8-10 µm-thick films before photons reach the light scattering
layer.
Using the model DCM/TT1 DSC system we were able to demonstrate extremely
high average excitation transfer efficiency of over 95% with transparent TiO2 films,
but could not increase the EQEERD above 40% in the optimized device architecture due
to low ERD absorption. This work clearly shows that FRET from energy relay dyes to
sensitizing dyes can be efficient enough for ERDs to be incorporated into state-of-the-
art DSC systems. However, there are several important areas that should be researched
and developed to fully determine the future prospects of high efficiency ERD DSCs.
Although there have been studies which measure how dyes diffuse in nanoporous
films based on molecule size104 and membrane type,105,106 there is relatively little
information on dye loading and homogeneity of the dye concentration inside of
nanopores. These areas should be further explored to determine the feasibility of high
dye loading inside of mesostructured TiO2 films.
The
ERD solubi
have only
extinction c
completely
light which
and solubili
quench.
It m
commonly
strength of E
the light) b
included in
Figure 3-16
solubility
3.4.3 Org
Orga
have only re
contour ma
ility and ER
been able t
coefficient (
fill optical w
will require
ity greater th
may be chal
used solven
ERDs may b
by sensitizin
the followin
6. Figure of
ganic Dye Al
anic dyes w
ecently been
ap in figure
RD molar ex
to absorb 3
(i.e. DCM)
windows in
e ERDs with
han 25 mM i
llenging to
nts to absorb
be in being a
ng dyes in
ng section.
Light Harve
lternatives w
with exceptio
n developed.
69
3-16 display
xtinction coe
0-40% of t
or poor so
the DSC the
h molar exti
in acetonitri
place large
b a large po
able to suppl
highly effic
esting versus
with ERDs
onal power
These dyes
ys the light
efficient. To
the sunlight
olubility (i.e
ey must be a
inction coef
ile. The ERD
concentrati
ortion of the
lement poor
cient system
s Molar extin
conversion
do not have
absorption f
o date, dyes
t due to eit
e. DTCI). I
able to abso
fficients >10
Ds should no
ions of mu
e solar spec
absorption
ms; one exa
nction coeff
efficiencies,
e the same b
fraction ver
s used as ER
ther low m
f ERDs are
orb >90% of
00,000 M-1 c
ot concentra
ultiple ERDs
ctrum. The c
(e.g. 50-70%
ample has b
ficient and E
, such as Y
broad absorp
rsus
RDs
olar
e to
f the
cm-1
ation
s in
core
% of
been
ERD
YD2,
tion
70
that is typically seen with Ru based metal complexes. As an example, YD2 only
modestly absorbs the solar spectrum from 500-600 nm as shown in figure 3-17. There
is great opportunity to fill this spectral region with ERDs to increase the power
conversion efficiency to 12%. We have recently tried to use a commercially available
cyanine dye (DTCI) which has very low dynamic quenching (<4x) and has a good
FRET radii (>4 nm). We found that adding 5mM of DTCI (which should absorb the
necessary 30% of the spectrum according to figure 3-16) resulted in no change (either
good or bad) in the EQE. It should be noted that using DTCI with TT1 sensitizing
dyes resulted in an increase in the power conversion efficiency; which leads us to
believe that there is a problem with how the DTCI and YD2 interact. We have two
theories related to the lack of performance. YD2 is a very bulky dye, as shown in
figure 1-3, and tightly packs on the surface; it is possible that this dye clogs the
smaller pores inside the DSC preventing DTCI from entering into the mesostructured
film. However, given that large CdSe/ZnS quantum dots have been able to penetrate
the mesostructured films in the past this is unlikely. A more likely reason for the lack
of performance may have to deal with the dipole orientation of the YD2 on the titania
surface. It is probable that the backbone of the YD2 is perpendicular to the titania
surface, while the backbone of the TT1 dyes may be parallel to the titania surface,
which allows for near unity energy transfer. If the SD dipole orientation is
perpendicular to the ERD then energy transfer will not occur (see section 2.1). Further
study is required to determine the dipole orientation of the YD2, but this preliminary
result would place limitations on the types of sensitizing dyes that may be
incorporated into the ERD DSC system.
Figure 3-17
3.5 Exp
The p
reference si
order to red
350-700nm
DSC and m
measuremen
reduce light
The EQ
USA), whic
Ltd). EQE
aperture are
control DSC
measured in
electrolyte a
taken at 10%
The differe
control devi
EQE spectr
erimenta
ower of the
licon photod
duce the mi
to less than
measuring the
nts were per
t scattering.
QE measurem
ch was focus
E measureme
ea of 0.159
C result in
n the devices
at higher lig
% sun to full
nce in the
ices is the sa
rum of YD2.
al Metho
AM 1.5 sol
diode equipp
smatch betw
2% 107. The
e photocurre
rformed usin
ment light s
sed through
ents were p
cm2. Integra
slightly hig
s. This is a r
ght intensitie
l sun are con
integrated E
ame ratio as t
71
.
ds
ar simulator
ped with an
ween the sim
J-V curves
ent using a K
ng a metal m
ource was a
a Gemini-18
erformed at
ating the EQ
gher (~10%)
result of cha
s. Extrapola
nsistent with
EQE spectru
the differenc
r (100 mW c
infrared cut
mulated ligh
were obtain
Keithley 240
mask with a
a 300W xen
80 double m
t 1% sun us
QE spectra o
) estimated
arge transpor
ating device
h the estimate
um between
ces in JSC at
cm-2) was ca
toff filter (K
ht and solar
ned by extern
00 digital so
an aperture o
non lamp (IL
monochromat
sing a metal
of the ERD
JSC at full
rt limitation
results from
ed JSC from
n the ERD
full sun.
alibrated usin
KG-3, Schott
spectrum fr
nally biasing
ource meter.
of 0.159 cm
LC Technolo
tor (Jobin Y
l mask with
containing
sun than th
ns caused by
m measureme
the EQE res
containing
ng a
t) in
from
g the
All
m2 to
ogy,
Yvon
h an
and
hose
y the
ents
sults.
and
72
4 Using Energy Relay Dyes Unattached to Titania in
solid-state DSCs
Besides broadening light absorption into the near-infrared domain, enhancing the
open circuit potential is a key pathway to further increase the power conversion
efficiency. In a typical liquid electrolyte cell the maximum open circuit potential is
around 800 mV, which is significantly less than soild-state dye-sensitized solar cells
(ss-DSCs) comprised of organic hole transporting materials (HTM).108-114 By
minimizing the energy difference between the HOMO of the sensitizing dye and the
work function of the organic hole conductor, it is possible to develop DSCs with a
higher power conversion efficiency. Recently, we have shown SSDSCs with an open
circuit voltage greater than 1000 mV indicating promise of high power conversion
efficiency.115 However, ss-DSCs suffer from significantly higher recombination rates
and are limited to an optimized thicknesses of ~2 µm (versus 10 µm for liquid DSC)116,
which limits light absorption. Hence, the primary challenge with ss-DSCs is to absorb
all the light inside of a relatively thin film.
Several schemes have been developed to increase light absorption in liqiud DSCs
including co-sensitization24 and the use of intramolecular energy transfer between
energy donor molecules attached to the sensitizing dye.117,118 Although these
architectures can broaden light absorption, they require that the dyes attach to the
titania surface, which does not allow for increased dye loading. We recently
demonstrated a new DSC architecture, shown in figure 4-1, where highly luminescent
energy relay dyes (ERDs) placed inside the electrolyte absorb higher energy photons
and transfer their energy to the sensitizing dye via Förster Resonant Energy Transfer
(FRET).64,119 Using ERD DSC architecture, allows separation of light absorption and
charge transfer processes. In SSDSCs, ERDs can be mixed in large concentrations
inside the HTM to increase the light harvesting yield.
Figure 4-1.
are absorbed
hole into the
relay dye (N
Figure not d
4.1 ERD
ERDs
mix well wi
hole injectio
ns.120 ERD
spiro-OMeT
only modera
were signif
aggregate in
when screen
Operating m
d by the sens
e electrolyte
N877) and ei
drawn to sca
D Design
s used in ss-D
ith HTM, an
on from the
candidates
TAD (-5.20
ate quenchin
ficantly mor
nside the spi
ning ERD ca
mechanisms
sitizing dye
. Higher ene
ther 2) Först
le.
Rules fo
DSCs should
nd not rapidl
sensitizing d
should have
eV vs vacu
ng (5-10x) w
re quenched
iro-OMeTAD
andidates.
73
of ss-DSC.
(SQ1), trans
ergy (blue) p
ter energy tr
or solid-st
d be soluble
ly inject char
dye to the sp
e a HOMO le
uum); ERDs
while dyes w
d (figure 4-2
D; this is on
1) Lower en
sferring an e
photons are a
ransferred in
tate DSC
e in polar sol
rges to the h
piro-OMeTA
evel above t
s with HOM
with HOMO
2). Furtherm
ne of the big
nergy (magen
electron into
absorbed by
nto the sensit
Cs
lvents (e.g. c
hole conduct
AD ranges fr
that of the w
MO above -5
levels near s
more, the E
ggest issues
nta) photons
the TiO2 and
the energy
tizing dye.
chlorobenze
tor. In ss-DS
rom <1 ps to
work function
5.20eV show
spiro-OMeT
RD should
we have fo
s
d
ne),
SCs,
o <4
n of
wed
TAD
not
ound
Figure 4-2.
derivatives.
4.2 N87
N877
state DSCs
infiltrate the
and several
mixed with
work well in
(2.6µs), is g
in liquid DS
4.2.1 N87
In t
phenanthrol
. PL Quenc
7/SQ1 so
was the fir
. Original w
e film when
l other orga
spiro-OMeT
n solid-state
greatly quenc
SCs.
7/SQ1 Emis
this work,
line) rutheni
ching versu
olid-state
rst dye that
work center
mixed with
anic dyes th
TAD. ERDs
DSCs. Inter
ched in liqu
ssion and A
we use a
ium (II) sen
74
us HOMO l
e DSC Sy
we were ab
red around
spiro-OMeT
hat showed
that worked
restingly N8
id DSCs by
Absorption S
highly ph
nsitizer (here
leve of var
ystem
ble to succes
using quan
TAD. We al
quenching g
d well in the
877, which h
the electrol
Spectra
hosphorescen
eafter labele
rious ADT
ssfully impl
ntum dots,
lso attempted
greater than
e liquid DSC
has a long liv
lyte and doe
nt tris (4,7
ed as N877)
and Pentac
lement in so
which did
d to use PTC
n >1000x w
C do not seem
ved excited s
s not work w
7-diphenyl-1
) as the ene
cene
olid-
not
CDI
when
m to
state
well
,10-
ergy
relay dye an
dimethyl-1-
ylidene]met
2,2’ 7,7’
OMeTAD).
of ERDs in
UV/
maximum a
due to stro
maximum a
shown in fig
range of SQ
temperature
is ~6.0 nm
resolved PL
Figure 4-3.
SQ1 (blue)
nd an efficie
ethyl-2H-ind
thyl]-3,3-trim
’-tetrakis(N,
To the best
ss-DSCs.
/Vis absorpt
at 636 nm w
ong π-π* ch
at 460 nm o
gure 4-3, the
Q1 and the fl
e122 and 68%
for the N8
L spectroscop
Normalized
and N877 (r
ent near infra
dol-2-yliden
methyl-1-oct
N-di-p-meth
of our know
tion spectru
with high mo
harge transf
of 29,000 M
e emission r
luorescence
% at 77K 123.
877-SQ1 sys
py (see secti
d UV/Vis abs
red) in ethan
75
ared sensitiz
ne)methyl]-2
tyl-3H-indol
hoxyphenyla
wledge, this
um (Figure
lar extinctio
fer (CT) tra-1 cm-1 and
range of N87
quantum yie
The calcula
stem and w
ion 2.4.2).
sorption (sol
ol, respectiv
zer (5-Carbo
2-hydroxy-4-
lium, (SQ1)
amine)-9,9’-
is the first in
4-3) of S
on coefficien
ansitions.121
an emission
77 is well m
eld of N877
ated Ro base
was experime
lid line)/emi
vely
oxy-2-[[3-[(1
-oxo-2-cyclo
)121 in ss-D
spiro-bifluor
nstance of th
SQ1 in eth
nt (ε = 158,5
N877 has
n maximum
matched with
in solution
d on 68% of
entally verif
ission (dash
1,3-dihydro-
obuten-1-
DSCs based
rene (Sp
he incorporat
hanol, show
500 M-1 cm-1
an absorpt
at 612 nm.
h the absorp
is 37% at ro
f quantum y
fied using t
line) spectr
3,3-
on
piro-
tion
ws a 1) is
tion
As
tion
oom
yield
time
a of
76
4.2.2 N877 quenching by spiro-OMeTAD
To determine the degree of PL quenching, the photoluminescence of N877
inside spiro-OMeTAD was compared to N877 in an inert polystyrene matrix. Films
were spun cast at 1000 RPM onto a glass substrate from a solution of 1.5% (wt) N877
in 27 mg/mL of polystyrene and spiro-OMeTAD respectively in chloroform inside of
a glovebox. It should be noted that N877 has a tendency to form aggregates in
chlorobenzene, but completely dissolves in chloroform. Figure 4-4 shows the
photoluminescence between N877/spiro and N877/polystyrene films corrected for
differences in absorption. The peak PL is 3,200 for the N877 in spiro-OMeTAD and
220,000 for N877 in PS resulting in a 69 times reduction. The PL lifetime of N877 in
solution is 2.6 µs. Assuming that the PL reduction is due to an increase in nonradiative
recombination due to hole transfer then the injection rate is estimated to be 38 ns,
significantly slower than the rate for conventional sensitizing dyes (e.g. N719 and
Z907). Cyclic voltammetry indicates that the first oxidation potential of spiro-
OMeTAD is ~0.81 V vs NHE,124 while the potential of N877 HOMO is ~1.46 V.125.
There are several possible reasons for the retardation of charge transfer; 1) the N877
does not contain the NCS ligands, typically placed on sensitizing dyes, which are
known to increase hole transfer and the large difference between the energy levels
may retard charge injection from spiro-OMeTAD. A full description of the
relationship between ERD HOMO level and charge injection is beyond the scope of
this paper, but will be discussed in a future study.
500 550 600 650 700 750 800 850
1000
10000
100000
log(
PL
) (a
rb)
Wavelength (nm)
N877 in Polystyrene N877 in spiro-OMeTAD
77
Figure 4-4. Photoluminescence spectra of 1.5% wt N877 in polystyrene versus spiro-
OMeTAD, corrected for absorption.
4.2.3 N877/SQ1 DSC Fabrication and Device Performance
For solid state solar cells, fluorine-doped SnO2 glass (15 ohm/sq, Pilkington)
substrates were cleaned first with Helmanex solution, rinsed with acetone, and then
ethanol. Next, a ~100 nm compact layer of TiO2 was deposited by spray pyrolysis126.
A porous layer of 30 nm TiO2 particles (~2 µm thick) was coated by the doctor-
blading technique, followed by sintering at 500 °C under an oxygen flow. After
cooling, the thin TiO2 films were impregnated in a 0.02 M aqueous TiCl4 solution for
15 hours, and then rinsed with deionized water. The TiCl4 treated TiO2 films were
annealed at 450 °C for 30 min and then cooled to ~ 80 °C before plunging into the dye
solution for 3 h. After soaking in dye solution, the substrates were rinsed in
acetonitrile, and then the hole-transporting material 2,2’ 7,7’-tetrakis(N,N-di-p-
methoxyphenylamine)-9,9’-spiro-bifluorene (Spiro-OMeTAD) solution (180 mg/mL,
in chlorobenzene) with additives of tert-Butyl pyridine (17 µL/mL), and
Li[CF3SO2]2N (19.5 mM), was spin coated at 2000 rpm on top of the TiO2 film.127,128
For energy transfer study, 10 mM of N877 in the Spiro-OMeTAD solution was spin
coated. Finally, a 50 nm gold layer was evaporated on the top of the Spiro-OMeTAD.
Figure 4-5 shows the incident monochromatic photon-to-current conversion
efficiency (IPCE), which is synonymous to EQE, of ss-DSCs sensitized by SQ1 with
and without N877. At the peak absorption wavelength, IPCE exceeds 47% in SQ1
sensitized solid state solar cells; but 76% of the light is absorbed giving an internal
quantum efficiency, IQE, of 62%. When 10mM of N877 is added, the IPCE increases
to 8% (at 460 nm) and 21% (at 400 nm) appear resulting in 30% increase in current
density and 29% increase in power conversion efficiency (Table 4-1). The observed
increase in photogenerated current could be caused by direct injection from N877 to
TiO2. To eliminate this possibility we made an ss-DSC containing no sensitizer while
78
retaining all other conditions unchanged. The data clearly show no injection of
electrons in the visible region between 400-500 nm (Figure 4-5 gray line)
corroborating the hypothesis that the new IPCE from 400 nm to 530 nm is caused by
an energy transfer from N877 to SQ1. It should be noted that two earlier strategies to
incorporate ERDs into ss-DSCs were not successful; Perylene derivatives with fast PL
rates (<5ns)129 appear to have extremely fast hole injection rates resulting in PL
quenching >1000 times. CdSe quantum dots coated with a ZnS insulating shell
showed virtually no PL quenching in spiro-OMeTAD; however, due to their relatively
large size and possible aggregate formation did not pore fill inside the TiO2 film.
4.2.4 N877/SQ1 ETE Estimate
Figure 4-5 shows a ΔEQE of 8% and IQE of 62%. For the film thickness of 2
µm, porosity of 0.60, the measured ηabs,DONOR is 41%, an ETE of 32% can be estimated
using equation 2-2. Unlike liquid based DSCs which require multiple optical
measurements (see section 3.3), the ηabs,DONOR can be directly measured using an
integrating sphere.
Figure 4-5. EQE spectrum of SQ1 SSDSCs with and without ERD, N877. The gray
line is an IPCE spectra of only Spiro-OMeTAD and the N877 energy transfer relay.
400 450 500 550 600 650 700 7500
10
20
30
40
50
IPC
E (
%)
Wavelength (nm)
no N877 10 mM N877
79
The black line is only SQ1 and Spiro-OMeTAD. The red line is SQ1 + N877 + Spiro-
OMeTAD.
Table 4-1. J-V characteristics of SQ1 ss-DSCs without and with N877.
N877 Jsc Voc FF [%] [a]
0 mM [b]
2.98mA/cm2 807 mV 0.58 1.40
10mM [c]
3.87mA/cm2 786 mV 0.59 1.80
[a] an overall efficiency is derived from Jsc×Voc×ff/light intensity, [b] active area
0.35 cm2, [c] active area 0.36 cm2,
4.2.5 N877 Synthesis
Synthesis of tris(4,7-diphenyl-1,10-phenanthroline)Ruthenium(II) Chloride
[Ru(dpp)3]Cl2 (N877, see the synthesis route in Figure S4): The ligand 4,7-diphenyl-
1,10-phenanthroline (dpp) and dichloro(p-cymene)ruthenium(II) dimmer were
obtained from Aldrich and used as received. All reagents and solvents purchased are
reagent grade (puris) from Fluka and were used without further purification. The
complex was synthesized in a commercial microwave oven CEM, Discover system
using septum-sealed 10 ml glass tube. In a typical reaction dichloro(p-
cymene)ruthenium(II)dimer (0.0612 g, 0.1 mmol, Aldrich) and 4,7-diphenyl-1,10-
phenanthroline (0.233 g, 0.7 mmol, Aldrich) were taken into a 10 ml glass tube and 2
ml of DMF was added. A small magnetic stirrer having 5 mm length and 2 mm
diameter was introduced into the reaction tube. Then the tube was sealed using a
septum (CEM) and inserted into a microwave oven. The tube was subjected to reflux
at 220°C for 5 minutes under stirring using 300 W microwave source. After cooling
the reaction tube, the complex was precipitated using (~20 ml) diethyl ether. The
isolated bright red-orange solid was dissolved in dichloromethane (~2 ml) and again
precipitated by adding diethyl ether (~20 ml), yielding (0.207 g) 89%. Anal. Calcd. for
80
[Ru(dpp)3]Cl2 (H2O)6 : C72H60Cl2N6O6Ru: C 67.71; H 4.73; N 6.58; Found: C 66.78,
H 4.60, N 6.56. Carbon to nitrogen ratio calculated 12 and found 11.873.
4.2.6 N877/SQ1 Testing Methods
UV/Vis absorption spectra were measured on a Cary 5 spectrophotometer and
fluorescence spectra were recorded on a Spex Fluorolog 112 spectrofluorimeter.
Samples were contained in 1 cm path-length quartz cells. Time resolved PL
measurements were performed using a Time-Correlated Single Photon Counting
(TCSPC) system from PicoQuant. Solutions were excited with a pulsed laser diode,
(model LDH 485: 481nm, 70ps FWHM, 5MHz) detected with a single photon
avalanche diode (PDM 100CT SPAD) attached to a monochromator and processed by
a PicoHarp 300 correlating system.
Solar Cell Characterization: For photovoltaic measurements of the DSCs, the
irradiation source was a 450 W xenon light source (Osram XBO 450, USA) with a
filter (Schott 113), whose power was regulated to the AM 1.5G solar standard by
using a reference Si photodiode equipped with a colour matched filter (KG-3, Schott)
in order to reduce the mismatch in the region of 350-750 nm between the simulated
light and AM 1.5G to less than 4%. The measurement delay time of photo J-V
characteristics of DSCs was fixed to 40 and 100 ms for liquid solar cells and solid
solar cells, respectively. The measurement of incident photon-to-current conversion
efficiency (IPCE) was plotted as a function of excitation wavelength by using the
incident light from a 300 W xenon lamp (ILC Technology, USA), which was focused
through a Gemini-180 double monochromator (Jobin Yvon Ltd.).
4.2.7 N877/SQ1 ss-DSC Conclusions
In conclusion we have demonstrated Förster energy transfer in solid-state dye-
sensitized solar cells between a phosphorescent ruthenium complexes solvated in the
solid organic hole conductor, spiro-OMeTAD and squaraine dyes grafted on the oxide
surface. By incorporating ERD into the SSDSC it is possible to greatly increase dye
81
loading. By using a combination of high molar extinction coefficient near-infrared
sensitizing dyes with ERDs that absorb in the visible it is possible to significantly
increase light harvesting yields in thin film DSCs. The ERD in spiro-OMeTAD was
able to increase the efficiency of the optimized SQ1 SSDSCs by 29% in terms of
power conversion efficiency. The low PL quenching is an important parameter in this
system leading to the retarded direct charge injection before FRET transfer. This
device architecture has the potential to improve DSC efficiency by choosing
sensitizing dyes with higher internal quantum efficiencies and energy relay dyes with
higher molar exctinction coefficients that are minimally quenched by spiro-OMeTAD.
82
5 Using Co-Sensitized Near-Infrared Energy Relay
Dyes to Increase Light Harvesting
Currently the most efficient sensitizing dyes are ruthenium based, metal ligand
complexes (e.g. C106 and N719),7,8 which absorb light in the visible portion of the
solar spectrum, have excellent charge injection properties, and produce a high open-
circuit voltage, Voc, which is defined as greater than 750 mV. It should be possible to
further increase the power conversion efficiency of DSCs by harvesting light in the
near-infrared red portion of the spectrum. Cosensitization of titania by dyes with
complimentary absorption spectra has been demonstrated to broaden the spectral
response of organic dye based DSCs in the visible portion of the spectrum, but not
beyond 720 nm.24,130-132 Designing near-infrared sensitizing dyes with high internal
quantum efficiencies is challenging because reducing the band gap requires more
precise alignment of the LUMO and HOMO levels and short conjugated ligands to
facilitate charge transfer. To date only two NIR sensitizing dyes (i.e. peak absorption
>700 nm) have demonstrated good charge injection efficiencies in DSCs, but neither
dye has a Voc greater than 450 mV.101,102 Recombination from the electrons in titania
with holes in the dye and triiodide in the electrolyte play a key role in determining the
open-circuit voltage.2 Organic dyes typically experience higher recombination rates
resulting in a lower Voc. 133 The great challenge of designing a cosensitized DSC
system using NIR-dyes will be maintaining a Voc greater than 700 mV.
5.1 Near-Infrared Dye Design Rules
Two NIR dye design strategies could result in higher power conversion
efficiencies. First, it may be possible to use highly absorptive NIR-sensitizing dyes
that directly inject charges even if NIR-SDs have higher recombination rates by using
low surface concentrations (<15%) of NIR-SDs to minimize Voc losses. DSC systems
where cosensitized dyes do not electronically interact with one another are expected to
have an electron recombination rate equivalent to the weighted average of the
individual d
a low recom
increase the
in the titani
cosensitized
A seco
to reduce th
electron inje
requiring ef
generate ph
using NIR-E
energy in a
Figure 5-1.
and uses sh
responsible
electrolyte (
5.2 AS0
In ord
have design
the electroly
dye DSC sys
mbination ra
e overall ele
ia, which ca
d DSC system
ond strategy
he recombin
ection. In th
fficient interm
hotocurrent,
ERDs, we m
cosensitized
The NIR dy
ort range en
for electron
(kreg).
02/C106 D
der to verify
ned a zinc na
yte and prod
stems. Howe
ate to dyes w
ctron recom
an dispropo
m.
y is to electro
nation rate,
his case, the
molecular en
as shown in
must first d
d system.
ye attached
nergy transfe
n transfer in
DSC Syst
y that energy
aphthalocyan
duce photocu
83
ever, intermo
with a highe
mbination rat
ortionately r
onically insu
which wou
NIR dye wo
nergy transf
n figure 5-1
determine ho
to the titania
er to excite a
nto the TiO2
tem
y transfer oc
nine based d
urrent indepe
olecular cha
er recombin
te between o
reduce the o
ulate the NIR
uld maintain
ould act as
fer to the me
. In order to
ow effective
a surface ab
a neighborin
2 (kinj) and
ccurs from t
dye (AS02)
endently. W
arge transfer
nation rate c
oxidized dye
open-circuit
R-dye from t
n the Voc b
an energy re
etal complex
o address th
ely NIR-ERD
sorbs near-in
ng sensitizin
hole regene
the NIR-dye
that cannot
We have chos
from dyes w
an significan
es and electr
t voltage of
the TiO2 surf
ut also prev
elay dye (ER
x SD in orde
he feasibility
Ds can tran
nfrared phot
ng dye, whic
eration with
e to the SD,
regenerate w
sen to use C
with
ntly
rons
the
face
vent
RD)
er to
y of
nsfer
tons
ch is
the
we
with
C106
84
because it is currently the most efficient Ru metal complex and has the most redshifted
absorption tail.
5.2.1 AS02/C106 Absorption and Emission Spectra
The absorption, emission, and the chemical structure of C106 and AS02 in
dimethylformamide (DMF) are shown in figure 5-2. C106 has a peak molar extinction
coefficient of 18,700 M-1 cm-1 at 550 nm with an absorption tail that extends weakly
out to 800nm.7 C106 has a broad emission spectrum with a peak at 786 and a natural
fluorescence decay lifetime of 85 ns in DMF. The photoluminescence quantum
efficiency of Ru based metal complexes is between 0.2-0.02%.13 AS02 has a peak
molar extinction coefficient of 100,000 M-1 cm-1 at 773 nm with a narrow emission
peak at 782 nm with a fluorescence natural decay lifetime of 2.75 ns in DMF. The
photoluminescence quantum efficiency of Zn based naphthalocyanines is between 10-
30%.134 Photoelectron spectroscopy in air was used to determine that the HOMO level
of AS02 (-4.60 eV) is high relative to the iodide potential (-4.65 eV) which has
previously been shown to prevent dye regeneration for a similar Zn naphthalocyanine
based sensitizing dye.135 C106 has a HOMO level of -5.05 eV.7 Intermolecular hole
transfer is thermodynamically favorable from the C106 to the AS02; the rate of
transfer will be dependent upon the HOMO level offset and the separation distance
between molecules.
Figure 5-2.A
infrared dye
the inset.
5.2.2 AS0
A s
measuremen
Time resolv
the fastest p
decay rates
alumina tha
determine th
the recombi
Absorption
e, AS02 in D
02 and C106
series of tim
nts were use
ved PL meas
process such
(knr) when
at prevent e
he regenerat
ination rate (
and Emissio
DMF. The c
6 Charge Tr
me resolve
ed to determ
surements ha
h electron tr
dyes are p
lectron inje
tion rate (kreg
(krec) betwee
85
on spectra o
hemical stru
ransfer Kine
d photolum
mine the cha
ave tradition
ransfer to T
placed on w
ction. Trans
g) between h
en holes in th
of the sensiti
ucture of C1
etics
minescent de
arge transfer
nally been us
TiO2 (kinj) as
wide band ga
sient decay
holes in the d
he dye and e
izing dye, C
106 and AS0
ecay and t
r rates of A
sed to determ
s well as th
ap semicond
measuremen
dye with the
lectrons in th
C106, and n
02 are shown
transient de
AS02 and C1
mine the rat
he non-radia
ductors such
nts are used
e electrolyte
he titania.
ear-
n in
ecay
106.
e of
ative
h as
d to
and
86
Time-correlated single photon counting was used to estimate the electron injection rate
of AS02 on TiO2. Measurements were performed using a 407nm LED; all samples
were measured for 1000 seconds and the results were normalized to the light
absorption at the LED wavelength. Figure 5-3 shows the time resolved PL results for
AS02 in solution (DMF), on alumina (Al2O3) and on titania. The PL decay of AS02
was modeled as a single exponential with a lifetime of τ0 = 2.75ns. When AS02 was
placed on Al203, which has a conduction band higher than the LUMO level of the
AS02 in order to prevent electron injection. AS02 on Al203 exhibited monoexponential
decay with a lifetime of τnr = 1.46 ns. AS02 on titania experienced PL decay faster
than the resolution of the instrument (~250 ps). An injection efficiency of 86% was
estimated by integrating the PL intensity of AS02/Al203 versus AS02/TiO2 over the
same amount of time (1000 seconds). Based on the injection efficiency, we would
estimate that the electron injection rate of AS02 to TiO2 (kinj) would be less than 230
ps.
C106 has a similar chemical structure as K19, which has an electron injection
rate on the 20 fs time scale when attached to TiO2.136 The non-radiative decay lifetime
is τnr = 18.5 ns and was best fit using a monoexponential decay shown in figure 5-4.
0 2 4 6 8 10 12 140
2000
4000
6000
8000
10000
Ph
oto
lum
insc
en
ce (
No
rma
lize
d)
Time (ns)
AS02 in DMF AS02 on Al
20
3
AS02 on TiO2
Figure 5-3. The Time resolved photoluminescence decay of AS02 in DMF solution
(10-5M) , on Al203, and on TiO2.
87
0 50 100 1500
2000
4000
6000
8000
10000
PL
Inte
nsi
ty
Time (ns)
C106 on Alumina
Figure 5-4. The Time resolved photoluminescence decay of C106 on Al203.
To determine the hole transfer from the dye to the electrolyte (kreg)
recombination of holes in the dye to electrons in the TiO2 (krec) we used time resolved
transient measurements of the individual dyes on TiO2 with an without the iodide
based electrolyte. Dye-sensitized, transparent nanocrystalline TiO2 films were
irradiated by nanosecond laser pulses produced by a Powerlite 7030 frequency-tripled
Q-switched Nd:YAG laser (Continuum, USA) pumping an OPO-355 optical
parametric oscillator (GWU, Germany) tuned at 550 nm (30 Hz repetition rate, pulse
width at half-height of 5 ns). To inject on the average less than one electron per
nanocrystalline TiO2 particle, the pulse fluence was attenuated to a maximum of 40 µJ
cm–2 by use of absorptive neutral density filters. The probe light from a Xe arc lamp
was passed through an interference filter monochromator, various optical elements,
the sample, and a grating monochromator before being detected by a fast
photomultiplier tube. Averaging over ca. 2000 laser shots was necessary to obtain
satisfactory signal/noise ratios.
C10
and measuri
records the
photoinduce
absence of
signal reflec
the oxidized
for the char
with the sa
oxidized dy
which indic
intercepted
The
structure by
level of th
regeneration
iodide based
Figure 5-5.
upon pulsed
6 recombina
ing the trans
concentratio
ed electron i
redox electr
cts the dyna
d dye. In suc
rge recombin
ame iodide/t
ye accelerate
cates that the
almost quan
AS02 recom
y Durrant et a
e Zn based
n does not oc
d electrolyte
Temporal p
d laser excita
ation rate (k
sient at 800 n
on of the oxi
injection from
rolyte, in pu
amics of the
ch conditions
nation (Fig.
tri-iodide co
ed markedly
e sensitizer w
ntitatively by
mbination ra
al. and found
d naphthaloc
ccur (i.e. the
).
profiles of th
ation (= 55
88
krec) was dete
nm. The tran
dized state o
m the dye in
ure MPN so
recombinat
s, a half-reac
5-5, blue tr
oncentration
y. t1⁄2 = 3 µs
was regenera
y the mediato
ate was prev
d to have a l
cyanine dye
e transient li
e transient a
0 nm, 5 ns fu
ermined by
nsient optica
of the C106 d
nto the cond
olvent, the d
tion of cond
ction time (t
race). In the
n used in th
s was meas
ated quickly
or.
iously meas
life time of 8
es are abov
ifetime is un
absorbance m
full width ha
exciting the
al signal obse
dye sensitize
duction band
decrease in
duction-band
1⁄2) of 200 µ
e presence o
he DSC, the
sured (Fig.
y and the ba
sured for a s
8ms.135 Beca
ve the pote
naffected by
measured at
alf-maximum
e dye at 550
erved at 800
er after ultra
d of TiO2. In
the absorba
d electrons w
s was measu
of an electro
e decay of
5-5, red tra
ck reaction
imilar chem
ause the HOM
ential of iod
the addition
= 800 nm
m pulse durat
nm
nm
afast,
n the
ance
with
ured
olyte
the
ace),
was
mical
MO
dide
n the
tion,
89
30 Hz repetition rate) on samples comprised of C106 dye adsorbed on nanocrystalline
TiO2 films in the presence (red trace) and in the absence (blue trace) of the redox-
active electrolyte.
5.2.3 AS02/C106 Excitation Transfer Modeling
Traditional energy transfer systems are designed to funnel energy from a donor
chromophore whose light absorption is blue shifted relative to the acceptor dye (i.e.
C106 to AS02) so that donor emission can overlap with the peak acceptor absorption
to provide the largest possible FRET radius.72 The FRET radius from C106 to AS02 is
estimated to be between 1.5 to 2.2 nm, which is fairly short and primarily due to the
low photoluminescence quantum efficiency of the C106 dye. Despite the weak
emission/absorption overlap in the AS02 emission and C106 absorption, the FRET
radius from the NIR-dye (AS02) to the SD (C106) is estimated to between 1.5 and 1.8
nm. The rate of Förster energy transfer (kFRET) between isolated chromophores,
known as point-to-point transfer, is given by kFRET = k0 (Ro)6/r6, where r is the
separation distance and k0 is the natural fluorescence decay rate, k0 = 1/τ0.
The sensitizing dye surface concentration was measured by desorbing the
C106 from titania using 0.15 M tetrabutylammonium hydroxide in DMF and found to
be 1 dye/nm2 on the 17-nm-diamter TiO2 nanoparticles with an estimated roughness
factor of 100/µm. When the NIR-dye molecules intimately mix with the C106 the
average separation between dyes is estimated to be approximately 1 nm. The FRET
rate from the AS02 to C106 is predicted to be between 7.1*109 – 2.3*1010 s-1
(τFRET,AS02 = 44-130 ps) based on an average separation distance of 1 nm, while the
FRET rate estimates from C106 to AS02 between 1.3*108-1.3*109 s-1 (τFRET,C106 =
0.75 -7.5 ns). Interestingly, the FRET rate from the NIR-dye (AS02) to the visible
sensitizing dye (C106) is an order of magnitude faster than in the opposite direction
due to the differences in the fluorescence decay rates between chromophores. The
kFRET rates should be considered rough approximations because the FRET radius
calculation is based on a random orientation (i.e. dyes rotating freely in solution),
which would not be the case when anchored on the TiO2 surface. Given the short
length scale, Dexter energy transfer may also play an important role in intermolecular
90
energy transfer.72 Meyer et al. have demonstrated near unity lateral Dexter energy
transfer from Ru based metal complex SDs to Os based metal complex SDs across a
semiconductor interface137 and have also estimated Dexter energy transfer rates
between Ru metal complex SDs to be on the 30 ns time scale.138 Calculating the
Dexter transfer rate between AS02 to C106 requires calculating the inner and outer
sphere reorganization energies and is beyond the scope of this work.138
The excitation transfer efficiency, ETE, is the probability that a dye will
undergo energy transfer. ETE is determined by the rate of intermolecular energy
transfer (kET) relative to the combined rates of all decay pathways which includes the
electron injection rate (kinj) and the non-radiative decay rate (knr) of the attached dye as
shown in equation 5-1. Hole regeneration is an alternative decay pathway, but occurs
on time scales several orders of magnitude slower than energy and electron transfer
and is not a major factor for iodide/triiodide based DSCs.
nrinjET
ET
kkk
kETE
(eq. 5-1)
Time resolved PL measurements were performed on titania and alumina films
to determine electron transfer to TiO2 (kinj) and the non-radiative decay rates (knr)
respectively. For efficient sensitizing dyes, the electron injection rate is the fastest
kinetic process; the kinj rate of AS02 is greater than 4.3*109 s-1 (τinj,AS02 < 230 ps) and
was not significantly slowed by the propionic acid ligand, while the kinj rate of Ru
based metal ligand based DSCs is approximately 5*1013 s-1 (τinj,C106 = 20 fs). 136 It
should be noted that the non-radiative decay rate of both dyes is faster when attached
on titania than the fluorescence decay rate when in solution. Transient absorption
decay measurements on dyed TiO2 films were used with and without the iodide based
electrolyte to determine the regeneration rate (kreg) between holes in the dye with the
electrolyte and the recombination rate (krec) between holes in the dye and electrons in
the titania respectively. All rates were best fit as a single exponential decay. The rates
of the AS02 + C106 DSC system are shown in figure 5-6 with the rate lifetimes
displayed in Table 5-1.
Figure 5-6.
geometrical
that result in
contribute t
processes.
Table 5-1: E
Mechanism
e- injection i
h+ regenera
Nonradiative
e- (TiO2) rec
Intermolecu
Natural fluo
Modeled Int
Measured In
Rates measu
kHT were ba
. Jablonski
ly correct (i
n photocurre
to photocurr
Energy and
m
nto TiO2
ation with elec
e recombinat
combination w
lar h+ transfe
rescence dec
termolecular F
ntermolecular
ured by (a) G
ased on meas
Plot of A
i.e. both dye
ent generatio
rent are labe
d Charge Tr
ctrolyte
ion
with h+ (Dye)
r
cay in DMF
FRET
r ET
Grätzel et al
sured rates a
91
S02 + C10
es should be
on are labele
eled in grey;
ransfer Life
Name
kinj
kreg
knr
krec
kHT
k0
kFRET
kET
.136 and (b) D
and the ETE
06 DSC sy
e on the sam
ed in black;
; dashed lin
times for A
ER
< 2
2
44-
<
Durrant et a
and IQE res
stem. The
me TiO2 surf
while proce
es represent
AS02 and C1
RD Lifetime
230 ps
---
1.5 ns
8.0 ms(b)
---
2.75 ns
-130 ps
530 ps
al.135 The est
spectively.
scheme is
face), proces
esses that do
t intermolec
106
SD Lifetim
20 fs
3.6 µs
18.5 ns
590 µs
< 3.2 µs
85 ns
0.75-7.5 ns
---
timated kET
not
sses
not
ular
me (a)
s
s
s
s
s
s
and
92
The excitation transfer efficiency from NIR-dye to the SD is estimated to be
between 60-80% based on the charge kinetics of the AS02 the FRET radius, and an
average separation distance of 1 nm. DSCs cosensitized with all organic dyes have
previously demonstrated an energy cascade effect, where intermolecular energy occurs
from the high band gap to the lower band gap SD, 139 However, energy transfer from
the metal complex SD to the NIR-dye is not likely because the rate of electron
injection of C106 is several order of magnitude faster than energy transfer processes,
efficiently splitting the exciton before energy transfer can occur.
5.2.4 AS02/C106 Fractional Surface Coverage and Dye Loading
To examine the affect of sequential sensitization we used 6.5µm thick,
transparent films comprised of 17 nm TiO2 particles. Figure 5-7, shows the optical
density of titania films first dipped in a 0.1 mM AS02 solution in DMF for 15 min (5-
7A) and 75 minutes (5-7B) respectively, then rinsed in DMF, dried with N2, and
measured using UV-Vis (green lines). The films were subsequently dipped in a 0.3
mM C106 solution comprised of 10% DMF with 90% ACN:TBA (50:50 mixture by
vol) for 18 hours and rinsed in acetonitrile and measured again (black lines). Figure 5-
7 also contains the optical density of a C106 control device which was only dipped in
C106 solutions for 18 hours (red dashed lines).
In order to accurately quantify the surface coverage (Γ) of AS02 and C106
dyes on the TiO2 surface we performed desorption measurements similar to those
described in literature and in the supporting information. The C106 dyed titania
control films had a peak optical density on the titania film of 1.9; when desorbed in
TBAH had a peak OD in a 1 cm cuvette of 0.315, which translates into a dye surface
coverage of ΓC106 = 1.83*10-10 mol/cm2 (or 1.10 dye/nm2). AS02 dyed films with a
peak OD of 0.725 on TiO2 had a corresponding OD of 0.465 in solution, which
translates to a surface coverage of ΓAS02 = 5.06*10-11 mol/cm2 (or 0.305 dye/nm2). The
AS02 results were based on a measured molar extinction coefficient of 100,000 M-1
cm-1.
Figure 5-7.
C106 (black
dashed line)
The
total dye loa
+ C106 sys
surface cov
corrected O
a surface co
NIR-ERD r
surface. Alt
dye/nm2 to
dye/nm2 to 0
titania surfa
Table 5-2: 6.5µm thick
Optical den
k line) dyed
).
surface con
ading relativ
tems that w
verage was c
D at the abs
oncentration
results in hi
though there
0.73 dye/nm
0.94 dye/nm
ace.
Dipping timk transpare
sity versus w
5.6µm thick
ncentration a
ve to the C10
were sequenti
calculated u
orption peak
n of ~1 dye
igher dye lo
e is a decline
m2), the incr
m2) resulting
me versus nt titania fi
93
wavelength f
k TiO2 films
and surface f
06 only cont
ially sensitiz
using the de
ks of C106 a
/nm2. As ex
oading and
e in the surfa
rease in AS
in a 59% inc
total surfacilms
for AS02 on
s compared t
fraction of e
trol (Total Γ
zed for vario
esorption res
and AS02. T
xpected incr
higher fract
ace concentr
S02 dye load
crease in the
ce coverage
nly (green lin
to C106 con
each dye as
Γ) was determ
ous times in
sults describ
The C106 con
reased dippi
tion of AS0
ration of the
ding is mor
e overall dye
e and fracti
ne) and AS0
ntrol device (
well as the
mined for A
n table 5-2.
bed above w
ntrol device
ing time of
02 on the T
e SD (from 1
e significan
e loading on
ion of dyes
02 +
(red
the
S02
The
with
has
the
TiO2
1.05
nt (0
n the
s on
94
The ΓAS02 and ΓC106 is the surface concentration of AS02 and C106
respectively. The fraction represents the fraction of dye on the surface while the
overall surface coverage (Total Γ) is the change in the total amount of dye loading on
the surface versus the C106 control.
5.2.5 AS02/C106 Device Fabrication and External Quantum Efficiency Results
In order to verify that intermolecular energy and hole transfer occurs in DSCs
we cosensitized transparent 6.5µm-thick-films and measured the optical and electrical
properties using methods similar to literature.140 Showa Denko 17-nm-diameter
particles were deposited on fluorine-doped tin oxide glass (TEC 15 Ω/square, 2.2 mm
thick, Pilkington) via screen printing, sintered at 450°C, and subsequently treated with
TiCl4.82 Figure 5-8A, shows the optical density (OD) versus wavelength during
different stages of cosensitization. The titania films were first dipped in a 0.1 mM
AS02 solution in DMF for 15 min, then rinsed in DMF and dried with N2 (green line).
The film was subsequently dipped in a 0.3 mM C106 solution comprised of 10% DMF
with 90% acetonitrile:tert-butyl alcohol (50:50 mixture by volume) for 18 hours and
rinsed in acetonitrile (black line). The control DSCs were dipped in the C106 solution
for 18 hours (red dashed lines). TiO2 films dipped in AS02 for 15 min resulted in
fractional surface coverage of 14% AS02 with a peak optical density of 0.45 or 65%
of light absorbed at 780 nm. Adding the AS02 prior to C106 sensitization does not
drastically affect the overall light harvesting of the C106 sensitizer. The peak OD of
the C106 control device is 1.83 (98.5% light absorption) versus 1.74 (98.2% light
absorption) at 550nm for the AS02 (14%) + C106 (86%) system. Figure 5-8A also
shows a slight red shifting of both the AS02 and a C106 peak which is likely caused
by molecular orbital overlap between NIR-dye. The redshift was not caused by
solvatochromatic effects; changing from DMF to acetonitrile:tert-butyl alcohol
mixture resulted in a slight blue shift in the absorption peak of the AS02 sensitized on
TiO2. The AS02 peak shape and intensity does not change during sequential
sensitization which indicates that the AS02 molecules do not aggregate or desorb
while being dipped in the C106 solution.
Dye
previously d
dimethylimi
butylpyridin
(EQE) meas
EQE at 780
shown in fig
transfer from
photorespon
absorption b
the additio
intermolecu
with the el
determined
surface cove
Figure 5-8.
C106, AS0
wavelength
films were
TiO2 nanopa
-sensitized s
described in
idazolium io
ne in acetoni
surements w
0 nm is 10.2
gure 5-8B. T
m the AS02
nse at 780
by the titania
n of AS02
ular hole tran
lectrolyte. T
to be 88.8%
erage.
(A) Optica
02 + C106,
of C106, A
approximate
articles
solar cells w
detail in lite
odide, 0.03 M
itrile/valeron
were used to
2% for AS02
The EQE co
2 to the C10
nm; the EQ
a. The C106
2 on the t
nsfer from th
The internal
% for the C
al density v
and AS02
AS02 + C10
ely 6.5µm th
95
were assemb
erature with
M iodide, 0.1
nitrile (85:15
verify interm
2 + C106 D
ntribution fr
06. The AS0
QE generate
6 peak EQE
titania surfa
he C106 dye
l quantum e
C106 contro
ersus wavel
2 only. (B)
06, and AS0
hick and com
bled and test
an electroly
1 M guanidi
5 v/v). 7,66 E
molecular en
DSC and 0.8
rom AS02 is
02 only DS
ed below 45
(550 nm) is
ace. The E
e to the AS0
efficiency o
ol and 72.1%
length of tit
External q
02 only dye
mprised of t
ted using sta
yte comprise
nium thiocy
External quan
nergy and ho
8% for the C
s the direct r
Cs (green li
50 nm is a
s significant
EQE reduct
02, which ca
of the contr
% with ligh
tania films
quantum eff
e-sensitized
transparent 1
andard meth
ed of 1.0 M
yante, 0.5 M
ntum efficie
ole transfer.
C106 contro
result of ene
ine) showed
result of l
ly reduced w
tion is due
annot regene
rol device
ht (14%) A
sensitized w
fficiency ver
solar cells.
17-nm-diam
hods
1,3-
ter-
ency
The
ol as
ergy
d no
ight
with
e to
erate
was
S02
with
rsus
All
meter
96
5.2.6 Measuring the Average Excitation Transfer Efficiency of AS02/C106
System
The average excitation transfer efficiency, ETE , defined as the fraction of
excited NIR-ERDs that undergo energy transfer to the SD, is described by equation 2-
2.64The EQEERD is the external quantum efficiency contribution caused by the NIR-
ERD at 780 nm (9.4%), ηABS,ERD is the fraction of light absorbed by the NIR-ERD,
IQE is the internal quantum efficiency. The ηABS,ERD was determined to be 50.8%
when correcting for light losses related to reflection (4%) and FTO light absorption
(11%) at 780 nm.140 Light absorption by C106 at 780 nm was considered negligible.
The estimated ETE was determined to be 26%; it should be noted that the measured
IQE (72.1%) is an average value of all C106 dyes, but the IQE is most likely lower for
C106 dyes that are in close proximity to AS02, which have a higher probability of
transferring holes to the AS02 before dye regeneration. Thus the calculated ETE
represents the minimum bound estimate for the AS02 + C106 DSC system. AS02 is
not an ideal NIR-ERD because the electron injection rate (τinj <230 ps) is competitive
with energy transfer which reduces the excitation transfer efficiency. For NIR-ERDs
with LUMO levels above the conduction band of TiO2 an insulating ligand should be
added to retard charge injection.77 If AS02 electron injection is significantly retarded
then the ETE would increase to over 70%. The measured energy transfer rate (kET) is a
combination of both Dexter and FRET energy transfer. Based on the kinj and knr of
AS02 and the minimum bound ETE of 26%, the measured rate of energy transfer (kET)
is >1.76*109 s-1 (τET < 568 ps) using equation 5-1.
5.2.7 Hole Transfer from C106 to AS02
Photo-induced transient absorption (PIA) spectroscopy, shown in figure 5-9,
was performed on C106, AS02 + C106, and AS02 sensitized films without the
presence of the electrolyte to probe the photogenerated charge species. Steady-state
PIA, which measures the change in absorption of the oxidized dye species, was
chopped at a frequency of 9 Hz using a 470 nm light bias using methods previously
described in
and has an a
at 780nm an
films, the C
nm), but
Figure 5-9.
C106+AS02
absorption a
PIA
resolved aft
light source
focused on
monochrom
65 μW cm−2
exit slits of
300–1650 n
AC signal fr
a function o
modulated u
n literature.1
absorption e
nd has an ab
C106 absorbs
the PI
. (A) Photo
2 (black), an
at 470nm.
experimenta
ter passing
. A 20 W ha
nto the sam
mator (Gemin2. A cooled d
f the monoc
nm. A dual p
from the dete
of wavelengt
using the int
41 Briefly, t
nhancement
bsorption in
s over 80% o
IA signal
o-induced tr
nd AS02 (gre
al technique
through the
alogen lamp
mple prior t
ni-180). The
dual color so
chromator. T
phase lock-i
ectors. This s
th. To obtain
ternal refere
97
the C106 ca
t at 800nm, w
crease at 10
of the photo
is dom
ransient abs
een) on TiO
e comprises
samples wi
was used as
to being re
e light inten
olid-state de
This instrum
n amplifier
signal provid
n the PIA sp
ence frequen
ation (red da
while the AS
000nm. For A
ons at the illu
minated b
sorption spe
O2. PIA signa
of a white l
ith the addi
s a probe sou
efocused on
nsity on the
tector (Si/In
ment has an
(SR 830) w
ded the chan
pectrum, a L
ncy of the lo
ash dot) blea
S02 cation (
AS02 + C10
umination w
by the A
ectra of C1
als were norm
ight probe b
ition of a m
urce which w
nto the slit
sample was
nGaAs) was
effective sp
was used to s
nge in transm
Lumiled 470
ock-in ampli
aches at 550
(green) bleac
06 dyed (bla
wavelength (
AS02 cat
06 (red da
malized to l
beam, spectr
modulated pu
was filtered
s of a dou
s approxima
mounted on
pectral range
separate out
mission (ΔT
0 nm diode
ifier. The pu
0nm
ches
ack)
(470
tion.
ash),
ight
rally
ump
and
uble
ately
n the
e of
the
) as
was
ump
98
light from the diode was focused onto the same face of the sample as the probe source
but 20° off axis, with an approximate intensity of 6 mW−2. To obtain the transmission
spectra (T) a reference scan was taken with the probe beam mechanically chopped and
no excitation source. All the PIA measurements were performed in air.
A high concentration (0.23 mM) of NOBF4, a strong oxidizing agent, was
titrated into a 10-5 M of AS02 dye in DMF to verify that the PIA signal is related to the
oxidized AS02 dye species. The optical density was determined using a UV-Vis
instrument with a 1 cm cuvette. Figure 5-10 shows that with increasing concentration
of NOBF4 results in both a reduction in peak absorption at 780 nm and a slight
increase in light absorption in the 950-1000nm range that is consistent with the PIA
analysis.
400 500 600 700 800 900 1000 11000.0
0.2
0.4
0.6
0.8
1.0
Op
tical
De
nsi
ty (
OD
)
Wavelength (nm)
0mM NOBF4
5.6mM NOBF4
11.2mM NOBF4
16.8mM NOBF4
Figure 5-10. Optical Density of 1*10-5M AS02 in DMF with various concentration of NOBF4.
5.2.8 Effects of Intermolecular Hole Transfer in AS02/C106 System
AS02 is an ideal dye to measure the fraction of holes from C106 dyes that
transfer to NIR-dyes in the cosenstized DSC system. Charge transfer between SDs in
cosensitized systems has been previously discussed,132,139 but could not be verified nor
quantified because both dyes are capable of hole regeneration. Because AS02 cannot
99
regenerate with the electrolyte, all holes transferred to AS02 must recombine with the
electrons in the TiO2 and cannot contribute to photocurrent. For this system the
fraction of holes from C106 that transfer to AS02 can be estimated based on the
reduction in the internal quantum efficiency of the AS02 + C106 DSC. The internal
quantum efficiency is defined by equation 5-2, which can be defined as the probability
of hole transfer to the electrolyte, electron transfer to the titania, and the charge
collection efficiency (ηCC). For C106, the electron injection rate is extremely fast
relative to the nonradiative decay rate and is not expected to change with
cosensitization. Although, AS02 has a higher recombination rate between triiodide in
the electrolyte and electrons in the titania that will result in a lower Voc, the ηCC will
not be dramatically reduced for low AS02 dye loading when operating under short-
circuit conditions. Therefore, changes in the IQE will be primarily due to competition
between hole transfer (kHT) and regeneration (kreg) of the oxidized dye by the
electrolyte.
ccnrinj
inj
recregHT
reg
kk
k
kkk
kIQE
(eq. 5-2)
An equivalent surface concentration of AS02 reduced the IQE from 88% for
the C106 control to 47% for AS02 (56%) + C106 (44%) DSC. Based on the IQE
reduction and C106 kreg and krec rates, the effective hole transfer lifetime, τHT, is 3.2
µs. It should be noted that this is an averaged rate over all C106 dyes cosensitized on
the TiO2 surface; the intermolecular hole transfer rate may significantly vary
depending on how C106 and AS02 pack with one another on the surface. While the
IQE reduction caused by AS02 is an extreme case, regeneration rates can be slower for
organic dyes32, and NIR dyes in particular will likely have a lower driving force for
hole regeneration.11,38 The kHT indicates that >50% holes can be transferred from C106
dyes near AS02. Intermolecular hole migration to NIR-dyes have important
implications for Voc.
A 75 mV drop in Voc was observed for the cosensitized AS02 (14%) + C106
(86%) DSC system relative to the C106 control DSC. Because the Voc is also slightly
affected by the reduction in the photocurrent density, the electron lifetime was studied
100
to determine the effects of intermolecular hole migration on the recombination. The
electron lifetime was measured using electronic impedance spectroscopy for various
fractional AS02/C106 surface concentrations to better understand the change in Voc.
Impedance measurements were performed with an Autolab PGSTAT30 (EcoChemie
B.V., Utrecht, Netherlands) over a frequency range from 1 MHz down to 0.1 Hz at
bias potentials between -0.2 to -0.8 V (with a 10 mV sinusoidal AC perturbation); all
measurements were done at 20°C and in the dark. The resulting impedance spectra
were analysed with ZView software (Scribner Associate Inc) on the basis of the two
channel transmission line model.142 The electron lifetimes of various AS02 + C106
cosensitized DSC systems are plotted against conductivity and shown in figure 5-11.
C106 only DSCs have an electron lifetime of 500 ms, while AS02 only DSCs have an
electron lifetime of 2 ms near open-circuit voltage conditions. If the dyes do not
electronically interact in the cosensitized DSC system then one might expect that the
electron decay rate to be the weighted average of the individual dye systems. However
in the AS02 (14%) + C106 (86%) cosensitized DSC system, the electron lifetime is 50
ms, which is several times lower than weighted lifetime of 140 ms. The
disproportionate change in electron lifetime is caused by hole transfer from the C106
to AS02. The Voc change is not related to a reduction in the overall dye loading on the
TiO2, which actually increases during cosenstiziation. It should be noted that
recombination between electrons in the titania and the I3- electrolyte is considered to
be the Voc determining recombination mechanism when using Ru based metal
complex dyes, which have relatively fast regeneration rates.2,35 However,
recombination from electrons in TiO2 to oxidized dye species may become the critical
mechanism for NIR-dyes whose ground state redox potentials are less favorable for
regeneration. A complete study of the recombination kinetics of fully functioning
NIR-SD is required to determine which recombination mechanism plays a dominant
role under Voc conditions.
101
1E-4 1E-3
0.01
0.1
1
log
(E
lect
ron
LIfe
time)
log (Conductivity)
C106 Only AS02 (14%) + C106 (84%) AS02 (56%) + C106 (44%) AS02 Only
Figure 5-11. Electron lifetime versus conductivity for DSC systems with various
concentrations of AS02 and C106 on TiO2.
5.2.9 AS02 Synthesis
All chemicals were purchased from commercial suppliers and used without further
purification. Compound 1 was purchased from TCI America. Column
chromatography was performed using silica gel mesh size (230-400). Gel permeation
chromatography was performed using a Polymer Laboratories (Varian) PL-GPC 50
Plus Integrated System with three in-line PL mixed E columns.
102
Compound 2 - t-butyl 3-(6,7-dicyanonaphthalen-2-yl) acrylate: 6-
bromonaphthalene-2,3-dicarbonitrile (1) (0.50 g, 1.94 mmol) and bis(tri-t-
butylphosphine) palladium(0) (Pd[P(tBu)3]2) (0.04 g, 0.078 mmol, 4 mol %) were
added to a 50 mL schlenk flask and subjected to three vacuum/nitrogen refill cycles.
To the nitrogen filled schlenk flask were added t-butyl acrylate (0.32 mL, 2.18 mmol),
dicyclohexylmethylamine (NCy2CH3) (0.46 mL, 2.15 mmol), and THF (15 mL,
anhydrous). The reaction mixture was allowed to stir at room temperature for 5 min
then heated to 70ºC in an oil bath for 16 h. Precipitates together with a deep
blue/violet fluorescence began to form after 10 min. After the reaction was complete,
via TLC analysis, the THF was removed using a rotary evaporator to provide a grey
solid that was washed with cold methanol, filtered, and dried under vacuum. The solid
was dissolved in minimal THF and filtered through a 1 micron glass fiber filter,
followed by THF removal to provide an off-white solid that was vacuum dried and
used without further purification. (0.495, 84%) 1H NMR (CDCl3, 300 MHz): (ppm)
8.34 (2H, d, J = 5.10 Hz, ArH), 7.98 (3H, m, ArH), 7.73 (1H, d, J = 15.9 Hz, vinyl H),
6.58 (1H, d, J = 15.9 Hz, vinyl H), 1.56 (9H, s, OC(CH3)3).
Compound 3 - t-butyl 3-(6,7-dicyanonaphthalen-2-yl) propanoate: Compound 2
(0.25 g, 0.82 mmol), and Pd/C (0.05 g) were added to a 50 mL schlenk flask followed
by THF (20 mL) and methanol (2 mL). The reaction mixture was heated to 40ºC for
10 min until 3 dissolved, then cooled to room temperature and triethylsilane (1.30 mL,
8.21 mmol) was added. A mild evolution of H2 was observed during the first hour at
which point the reaction was heated slightly to 40ºC overnight to complete reaction as
determined by TLC. The reaction mixture was filtered through a 1 micron glass fiber
filter and the solvent removed by rotary evaporation to provide a pale green oil that
crystallized. The solid was stirred/washed with 3 x 2 mL hexane, followed by drying
in a vacuum oven to provide an off-white solid (0.18 g, 72%). 1H NMR (CDCl3, 300
MHz): (ppm) 8.29 (2H, d, J = 12.3 Hz, ArH), 7.91 (2H, d, J = 8.40 Hz, ArH), 7.78
(1H, s, ArH), 7.67 (1H, d, J = 8.55 Hz, ArH), 3.15 (2H, t, J = 7.50 Hz, CH2), 2.66 (2H,
t, J = 7.50 Hz, CH2) 1.40 (9H, s, OC(CH3)3).
Reference: Mandal and McMurray.143
103
Compound 4: Compound 3 (0.180 g, 0.60 mmol), and zinc acetate (Zn(OAc)2•2H2O)
(0.044 g, 0.20 mmol) were added to a 25 mL schlenk flask followed by 1-hexanol (10
mL) and this was heated at 90ºC for 10 min. 1,8-diazabicyclo[5.4.0]undec-7-ene
(DBU) (0.33 mL, 2.21 mmol) was added and the reaction mixture was heated to 160ºC
for 16 h resulting in a dark green reaction mixture. The solvent was removed and THF
(7 mL) followed by 1 M NaOH (2 mL) were added and this was heated at 70ºC for 20
h The solvent was removed and the residue dissolved in DI-H2O (15 mL) and
refluxed for 1 h. The resultant green solution was filtered through a 1 micron glass
fiber filter and neutralized with conc. acetic acid. The precipitate was filtered and
washed with copious amounts of DI-H2O then dried under vacuum at 80ºC. Reference:
Mori et al.144
5.3 NIR-ERD Conclusion
These studies demonstrate the need to refine design rules for NIR-SDs and NIR-
ERDs. NIR-SDs should have sufficient LUMO and HOMO levels for charge injection
and a high molar extinction coefficient (> 100,000 M-1 cm-1). Planar NIR-SDs that
pack well with metal ligand SDs may lose substantial Voc, negating the potential
power conversion efficiency gain with high Voc losses. NIR-SDs should be physically
separated from the metal complex SD either via long alkyl side chains or selective
positioning131,145 to prevent intermolecular hole transfer in order to maintain high
open-circuit voltage.
NIR-ERDs do not require precise LUMO level alignment and short conjugated
ligands for rapid electron charge injection. However NIR-ERDs must intimately mix
with metal complex sensitizing dyes in order to efficiently transfer energy and must
therefore have a HOMO level below the iodide potential to regenerate with the
electrolyte. Ideally, NIR-ERDs should be designed with an insulating ligand that is
long enough to prevent electron transfer77 and lower recombination, but short enough
to enable close range interactions with the SD. NIR-ERD should have peak
absorption between 720-790 nm and peak emission between 730-800 nm. Dyes with
lower band gaps (i.e. dyes with an emission peak >820 nm) would most likely not
104
work as NIR-ERDs with ruthenium based SDs. The ability to both sensitize and
transfer energy from NIR-ERDs to metal complex sensitizing dyes allows us to
expand the light harvesting out to 800nm, which has the potential to produce 14%
efficient DSCs in the future.
105
6 Conclusion
My primary research goals over the last three years have been to first develop a
proof-of- concept ERD DSC system and then determine whether or not ERDs could be
efficient enough (i.e. ETE > 90%) to be useful in highly efficient devices in the future.
In this thesis I have demonstrated that Förster resonant energy transfer may be used to
shuttle energy from unattached energy relay dyes to the sensitizing dye in dye-
sensitized solar cells. For iodide/triiodide based liquid DSCs we have shown that the
process can be highly efficient (e.g. average excitation transfer efficiency >95%) when
using strongly absorbing organic sensitizing dyes. Recently, ERDs in solid-state DSCs
have shown an average excitation transfer efficiency >62%,146 which demonstrates
that ERDs in this system can reach comparable values in the future. We have also
modeled energy transfer from ERDs to sensitizing dyes inside mesostructured titania
and found good agreement with experimental results.
This thesis makes a strong case that energy relay dyes have the potential to be
used in highly efficient devices. Chapters 3-5 have discussed the possibilities of
incorporating ERDs in different configurations to make more efficient devices, but to
date we have not successfully incorporated ERDs in DSCs with power conversion
efficiencies above 5%. There is a great deal of opportunity to develop the next
generation of ERDs and to find new ways to incorporate energy transfer in creative
ways inside DSCs. The remaining sections will briefly discuss future outlook of
energy transfer in DSCs and discuss the commercialization prospects of DSCs.
6.1 Future Outlook of Energy Transfer in Dye-Sensitized
Solar Cells
There has been a great deal of recent success developing organic donor-bridge-
acceptor sensitizing dyes such as YD2.14 Although ruthenium metal complexes were
initially used to reach 10% power conversion efficiencies, it has become increasingly
more challenging to extend the bandwidth to the NIR portion of the spectrum. Donor-
bridge-acceptor dyes allow for greater tailoring of the chemical structure and thus the
106
optical and electronic properties. The key in the future will be in developing ERDs
that complement these types of organic dyes, which like YD2 (see section 3.4.3) will
have optical windows where ERDs will be extremely useful. Although it may be
challenging to design ERDs that absorb >90% of the light at peak wavelength, a great
number of dyes can be created that absorb >40%. It should also be noted that
cosensitzed ERDs may also work for the donor-bridge-acceptor dyes, which would
allow them to maintain a high voltage. ERDs in the electrolyte with organic dyes have
the potential to break 14% and should be strongly pursued. Using NIR ERDs (Chapter
5) is also very promising to quickly break efficiencies over 11.5%, but ultimately the
NIR ERDs are limited to 700-800nm range which prevents power conversion
efficiencies of greater than 13% from being obtained.
There are other promising architectures that use energy transfer that are in their
infancy. Zaban et al. recently placed quantum dots inside the titania and used energy
transfer to increase light harvesting.147 If ERDs are placed inside the titania then
quenching may be reduced, but the ERDs must survive the 450°C sintering process.
Quantum dots are a good choice because of their broad, tunable absorption and strong
photoluminescence quantum efficiency; in fact they were the first ERD we tried in the
electrolyte, but static quenching caused by the electrolyte resulted in negligible energy
transfer. Another promising idea is to use light absorbing organic hole conductors and
transfer energy to the sensitizing dye. This idea is very similar to those originally
proposed by the McGehee group47,48 and may actually allow for significantly higher
light harvesting in a 2µm thick film. P3HT was recently used as a hole conducting
medium in a solid-state DSC and achieved 5% power conversion efficiency.9
Although energy transfer from P3HT, which has low photoluminescence quantum
efficiency, is unlikely the initial demonstration of a highly efficient polymer based
DSC is promising.
107
6.2 Commercialization Potential of Dye-Sensitized Solar
Cells
The solar power industry has greatly expanded over the last decade and is now
beginning to reach legitimate production levels with low module cost. The solar power
market will grow to over 10GW in 2010 while global production was less than 200
MW in 1999. Solar module costs have dropped to below $1/W using CdTe PV devices.
The Department of Energy believes that total solar module costs must be below $2/W
to compete with natural gas electrical power production without subsidies and $1/W to
compete with coal. First Solar, the current market leader, has achieved $0.87/W
module costs and has entered contracts at $3/W, which includes installation costs, for
utility scale projects. In the future, I expect CdTe to increase power conversion
efficiency from 11% to 13-14% and slightly reduce costs from $90/m2.
The impressive rise of inorganic thin film solar industry directly questions the
usefulness of pursuing alternative thin film technologies such as organic photovoltaic
cells and dye-sensitized solar cells that offer comparable power conversion
efficiencies, but whose stability and cost structure are not well understood. I believe
that DSCs have the potential to be commercially competitive with inorganic PV
technologies on a large scale for several reasons listed below.
DSCs have the potential to become the lowest cost-per-area producer at $30/m2
because the ease and potential reproducibility of manufacturing of DSCs, low costs of
materials, and water tolerance. The key to commercialization of DSCs is to
demonstrate high efficiency (>9%) glass free submodules that have long lifetimes (e.g.
>20 years). DSCs can be made via roll-to-roll processing and do not require the
complex film crystallization of inorganic thin films to achieve high efficiencies.
Titania films are formed by sintering TiO2 nanoparticles and can probably be done via
rapid thermal processing. Almost all of the cost of future DSCs modules will be in the
cost of the supporting substrates (e.g. glass). The titania, dye, and electrolyte represent
a negligible cost of the module system; therefore, in order to reach $30/m2 levels
DSCs must become glass free. This can be achieved by depositing titania on metal
foils148 (e.g. Al and stainless steel) and using transparent, plastic top layers. Plastic top
108
layer must be inexpensive, which will likely result in using materials with a relatively
high water vapor transport rate. One potential strength of DSCs is the ability to handle
moisture ingress; O’Regan et al. recently demonstrated water based electrolyte DSCs
with relatively high efficiencies (>5%). The glass free architecture is currently being
pursued by G24i, but it is important to note that their power conversion efficiencies
are currently too low to be commercially viable for industrial and residential power
production.
Module stability is perhaps the biggest impediment to commercialization of
DSCs.149 Toyota recently demonstrated iodide/triiodide with ionic liquids that showed
stability of 15 years.150 This result is promising but it should be noted that ionic liquids,
require higher iodide/triiodide concentrations which increase dark current and absorb
light, and are typically are 1-2% less efficient than acetonitrile based DSCs.
Furthermore, inorganic PV manufactures offer 25 year warranties on current solar
modules.
The short-term key challenge for future DSC researchers is to produce highly
efficient (10%), stable glass free devices. This will require an increase in the power
conversion efficiency of state-of-the-art DSCs from 11% to 12-14% and the
minimization of losses associated with using ionic liquids and changing to metal foil
based architectures. There is a great deal of research required in the development of
highly efficient DSCs, glass free designs, and stability of metal foil DSCs that will be
required to reach these goals. If the short term objectives are met and DSC modules
could be manufactured at a power conversion efficiency of 10-12% at $30-45/m2 then
production cost of DSCs would range from $0.3-0.45/W and would directly compete
with inorganic PV devices and unsubsidized natural gas in sunny regions. However,
given the low intrinsic power conversion efficiency of liquid based DSCs, it would be
challenging to reduce the installation costs enough to bring the total cost below $1/W
on the utility scale (i.e. solar farms in the desert) to compete with coal.
A greater effort should go into minimizing the voltage losses of the DSC by
replacing iodide/triiodide redox couple with higher work function organic hole
conductors or alternative organic liquid electrolytes. If the voltage loss is reduced then
109
DSCs can achieve power conversion efficiencies exceeding 20% which may be
required to reduce installation costs to $0.50/W. To date, very few have attempted to
develop new solid-state hole conductors for DSCs or understand why spiro-OMeTAD
results in the highest efficiency. There is a great opportunity to understand the device
physics of solid-state DSCs and to drastically improve DSCs using colored, wide band
gap hole conductors, which allow the potential to completely absorb the solar
spectrum in a thin film and also have a high open-circuit voltage.
Bibliography and References
1 O'Regan, B. & Grätzel, M. A Low-Cost, High-Efficiency Solar Cell Based on Dye-Sensitized Colloidal TiO2 Films. Nature 353, 737-740 (1991).
2 Peter, L. M. Dye-sensitized nanocrystalline solar cells. Physical Chemistry Chemical Physics 9, 2630-2642, doi:DOI: 10.1039/b617073k (2007).
3 H. J. Snaith, L. S.-M. Advances in Liquid-Electrolyte and Solid-State Dye-Sensitized Solar Cells. Advanced Materials 19, 3187-3200 (2007).
4 Bisquert, J., Cahen, D., Hodes, G., Ruhle, S. & Zaban, A. Physical Chemical Principles of Photovoltaic Conversion with Nanoparticulate, Mesoporous Dye-
110
Sensitized Solar Cells. The Journal of Physical Chemistry B 108, 8106-8118, doi:doi:10.1021/jp0359283 (2004).
5 Hagfeldt, A. & Grätzel, M. Molecular Photovoltaics. Accounts of Chemical Research 33, 269-277, doi:doi:10.1021/ar980112j (2000).
6 Bessho, T., Zakeeruddin, S., Yeh, C. Y., Diau, E. G. & Grätzel, M. Highly Efficient Mesoscopic Dye-Sensitized Solar Cells Based on Donor–Acceptor-Substituted Porphyrins. Angewandte Chemie International Edition 49, 6646-6649, doi:10.1002/anie.201002118 (2010).
7 Cao, Y. et al. Dye-Sensitized Solar Cells with a High Absorptivity Ruthenium Sensitizer Featuring a 2-(Hexylthio)thiophene Conjugated Bipyridine. The Journal of Physical Chemistry C 113, 6290-6297 (2009).
8 Nazeeruddin, M. K. et al. Combined Experimental and DFT-TDDFT Computational Study of Photoelectrochemical Cell Ruthenium Sensitizers. Journal of the American Chemical Society 127, 16835-16847, doi:doi:10.1021/ja052467l (2005).
9 Chang, J. A. et al. High-Performance Nanostructured Inorganic−Organic Heterojunction Solar Cells. Nano Letters 10, 2609-2612, doi:10.1021/nl101322h (2010).
10 Gratzel, M. Conversion of sunlight to electric power by nanocrystalline dye-sensitized solar cells. Journal of Photochemistry and Photobiology A-Chemistry 164, 3-14 (2004).
11 Ardo, S. & Meyer, G. J. Photodriven heterogeneous charge transfer with transition-metal compounds anchored to TiO2 semiconductor surfaces. Chem. Soc. Rev. 38, 115-164 (2009).
12 Boschloo, G. & Hagfeldt, A. Characteristics of the Iodide/Triiodide Redox Mediator in Dye-Sensitized Solar Cells. Accounts of Chemical Research 42, 1819-1826, doi:10.1021/ar900138m (2009).
13 Nazeeruddin, M. K. et al. Conversion of light to electricity by cis-X2bis(2,2'-bipyridyl-4,4'-dicarboxylate)ruthenium(II) charge-transfer sensitizers (X = Cl-, Br-, I-, CN-, and SCN-) on nanocrystalline titanium dioxide electrodes. Journal of the American Chemical Society 115, 6382-6390, doi:10.1021/ja00067a063 (1993).
14 Lee, C. W. et al. Novel Zinc Porphyrin Sensitizers for Dye-Sensitized Solar Cells: Synthesis and Spectral, Electrochemical, and Photovoltaic Properties. Chemistry-A European Journal 15, 1403-1412, doi:10.1002/chem.200801572 (2009).
15 Hamann, T. W., Jensen, R. A., Martinson, A. B. F., Ryswyk, H. V. & Hupp, J. T. Advancing beyond current generation dye-sensitized solar cells. Energy & Environmental Science 1, 66-78, doi:DOI: 10.1039/b809672d (2008).
16 Grätzel, M. Solar Energy Conversion by Dye-Sensitized Photovoltaic Cells. Inorganic Chemistry 44, 6841-6851 (2005).
17 Barnes, P. R. F. et al. Re-evaluation of Recombination Losses in Dye-Sensitized Cells: The Failure of Dynamic Relaxation Methods to Correctly Predict Diffusion Length in Nanoporous Photoelectrodes. Nano Letters 9, 3532-3538 (2009).
111
18 Barnes, P. R. F., Anderson, A. Y., Koops, S. E., Durrant, J. R. & O'Regan, B. C. Electron Injection Efficiency and Diffusion Length in Dye-Sensitized Solar Cells Derived from Incident Photon Conversion Efficiency Measurements. The Journal of Physical Chemistry C 113, 1126-1136 (2008).
19 Wang, P. et al. A stable quasi-solid-state dye-sensitized solar cell with an amphiphilic ruthenium sensitizer and polymer gel electrolyte. Nat Mater 2, 402-407 (2003).
20 Yum, J.-H. et al. Efficient Far Red Sensitization of Nanocrystalline TiO2 Films by an Unsymmetrical Squaraine Dye. Journal of the American Chemical Society 129, 10320-10321, doi:doi:10.1021/ja0731470 (2007).
21 Burke, A., Schmidt-Mende, L., Ito, S. & Gratzel, M. A novel blue dye for near-IR 'dye-sensitized' solar cell applications. Chemical Communications 3, 234-236, doi:DOI: 10.1039/b609266g (2006).
22 Campbell, W. M. et al. Highly Efficient Porphyrin Sensitizers for Dye-Sensitized Solar Cells. The Journal of Physical Chemistry C 111, 11760-11762, doi:doi:10.1021/jp0750598 (2007).
23 He, J. et al. Modified Phthalocyanines for Efficient Near-IR Sensitization of Nanostructured TiO2 Electrode. Journal of the American Chemical Society 124, 4922-4932, doi:doi:10.1021/ja0178012 (2002).
24 Cid, J.-J. et al. Molecular Cosensitization for Efficient Panchromatic Dye-Sensitized Solar Cells. Angewandte Chemie 119, 8510-8514 (2007).
25 Tachibana, Y., Nazeeruddin, M. K., Grätzel, M., Klug, D. R. & Durrant, J. R. Electron injection kinetics for the nanocrystalline TiO2 films sensitised with the dye (Bu4N)2Ru(dcbpyH)2(NCS)2. Chemical Physics 285, 127-132 (2002).
26 Tachibana, Y., Moser, J. E., Gratzel, M., Klug, D. R. & Durrant, J. R. Subpicosecond Interfacial Charge Separation in Dye-Sensitized Nanocrystalline Titanium Dioxide Films. The Journal of Physical Chemistry 100, 20056-20062, doi:doi:10.1021/jp962227f (1996).
27 Haque, S. A. et al. Parameters influencing charge recombination kinetics in dye-sensitized nanocrystalline titanium dioxide films. Journal of Physical Chemistry B 104, 538-547 (2000).
28 Haque, S. A., Tachibana, Y., Klug, D. R. & Durrant, J. R. Charge Recombination Kinetics in Dye-Sensitized Nanocrystalline Titanium Dioxide Films under Externally Applied Bias. J. Phys. Chem. B 1998, 1745-1749 (1998).
29 O'Regan, B. C. et al. Catalysis of Recombination and Its Limitation on Open Circuit Voltage for Dye Sensitized Photovoltaic Cells Using Phthalocyanine Dyes. Journal of the American Chemical Society 130, 2906-2907, doi:doi:10.1021/ja078045o (2008).
30 O'Regan, B. C. et al. Structure/Function Relationships in Dyes for Solar Energy Conversion: A Two-Atom Change in Dye Structure and the Mechanism for Its Effect on Cell Voltage. Journal of the American Chemical Society 131, 3541-3548, doi:doi:10.1021/ja806869x.
31 Clifford, J. N., Palomares, E., Nazeeruddin, M. K., Gratzel, M. & Durrant, J. R. Dye Dependent Regeneration Dynamics in Dye Sensitized Nanocrystalline
112
Solar Cells: Evidence for the Formation of a Ruthenium Bipyridyl Cation/Iodide Intermediate. The Journal of Physical Chemistry C 111, 6561-6567, doi:doi:10.1021/jp067458t (2007).
32 Siegers, C. et al. A Dyadic Sensitizer for Dye Solar Cells with High Energy-Transfer Efficiency in the Device. Chemphyschem 8, 1548-1556 (2007).
33 Wenger, B., Grätzel, M. & Moser, J.-E. Rationale for Kinetic Heterogeneity of Ultrafast Light-Induced Electron Transfer from Ru(II) Complex Sensitizers to Nanocrystalline TiO2. Journal of the American Chemical Society 127, 12150-12151, doi:10.1021/ja042141x (2005).
34 Mishra, A., Fischer, M. K. R. & Bäuerle, P. Metal-Free Organic Dyes for Dye-Sensitized Solar Cells: From Structure: Property Relationships to Design Rules. Angewandte Chemie International Edition 48, 2474-2499 (2009).
35 Pelet, S., Moser, J.-E. & Gratzel, M. Cooperative Effect of Adsorbed Cations and Iodide on the Interception of Back Electron Transfer in the Dye Sensitization of Nanocrystalline TiO2. The Journal of Physical Chemistry B 104, 1791-1795, doi:10.1021/jp9934477 (2000).
36 Moser, J. E. & Grätzel, M. Observation of temperature independent heterogeneous electron transfer reactions in the inverted Marcus region. Chemical Physics 176, 493-500, doi:Doi: 10.1016/0301-0104(93)80257-a (1993).
37 Privalov, T., Boschloo, G., Hagfeldt, A., Svensson, P. H. & Kloo, L. A Study of the Interactions between I−/I3− Redox Mediators and Organometallic Sensitizing Dyes in Solar Cells. The Journal of Physical Chemistry C 113, 783-790, doi:10.1021/jp810201c (2008).
38 Tatay, S. et al. Kinetic competition in liquid electrolyte and solid-state cyanine dye sensitized solar cells. Journal of Materials Chemistry 17, 3037-3044 (2007).
39 Snaith, H. J. & Schmidt-Mende, L. Advances in Liquid-Electrolyte and Solid-State Dye-Sensitized Solar Cells. Advanced Materials 19, 3187-3200 (2007).
40 van de Lagemaat, J. & Frank, A. J. Nonthermalized Electron Transport in Dye-Sensitized Nanocrystalline TiO2 Films: Transient Photocurrent and Random-Walk Modeling Studies. The Journal of Physical Chemistry B 105, 11194-11205, doi:10.1021/jp0118468 (2001).
41 Schwarzburg, K. & Willig, F. Influence of trap filling on photocurrent transients in polycrystalline TiO[sub 2]. Applied Physics Letters 58, 2520-2522, doi:10.1063/1.104839 (1991).
42 Frank, A., Kopidakis, N. & van de Lagemaat, J. Electrons in Nanostructured TiO2 solar cells: transport, recombination and photovoltaic properties. Coordination Chemistry Reviews 248, 1165-1179 (2004).
43 Kopidakis, N., Schiff, E. A., Park, N. G., van de Lagemaat, J. & Frank, A. J. Ambipolar Diffusion of Photocarriers in Electrolyte-Filled, Nanoporous TiO2†. The Journal of Physical Chemistry B 104, 3930-3936, doi:10.1021/jp9936603 (2000).
44 Snaith, H. J. Estimating the Maximum Attainable Efficiency in Dye-Sensitized Solar Cells. Advanced Functional Materials 20, 13-19 (2009).
113
45 Shockley, W. & Quessier, H. J. Detailed Balance Limit of Efficiency of P-N Junction Solar Cells. Journal of Applied Physics 32, 510 (1961).
46 Förster, T. Transfer Mechanisms of Electronic Excitation. Discussions Faraday Soc. 27, 7 (1959).
47 Scully, S. R., Armstrong, P. B., Edder, C., Frechet, J. M. J. & McGehee, M. D. Long-Range Resonant Energy Transfer for Enhanced Exciton Harvesting for Organic Solar Cells. Adv. Mater. 19, 2961-2966 (2007).
48 Liu, Y. X., Summers, M. A., Edder, C., Fréchet, J. M. J., McGehee, M.D. Using Resonance Energy Transfer to Improve Exciton Harvesting in Organic-Inorganic Hybrid Photovoltaic Cells. Advanced Materials 17, 2960-2964 (2005).
49 Hu, X. & Schulten, K. How nature harvests sunlight. Physics Today 50, 28 (1997).
50 Hu, X., Damjanovic, A., Ritz, T. & Schulten, K. Architecture and Mechanism ofthe Light-Harvesting Apparatus of Purple Bacteria. Proceedings of the National Academy of Sciences of the United States of America 95, 5935-5941 (1998).
51 Pullerits, T. & Sundstrom, V. Photosynthetic Light-Harvesting Pigment-Protein Complexes: Toward Understanding How and Why. Accounts of Chemical Research 29, 381-389, doi:doi:10.1021/ar950110o (1996).
52 Blumen, A., Klafter, J. & Zumofen, G. Influence of restricted geometries on the direct energy transfer. J. Chem. Phys. 84, 1397-1401 (1986).
53 Fung, B. K.-K. & Stryer, L. Surface density determination in membranes by fluorescence energy transfer. Biochemistry 17, 5241-5248 (1978).
54 Drake, J. M., Klafter, J. & Levitz, P. Chemical and biological microstructures as probed by dynamic processes. Science 251, 1574-1579, doi:10.1126/science.2011737 (1991).
55 Joo, C., Balci, H., Ishitsuka, Y., Buranachai, C. & Ha, T. Advances in Single-Molecule Fluorescence Methods for Molecular Biology. Annu. Rev. Biochem 77, 51-76, doi:doi:10.1146/annurev.biochem.77.070606.101543 (2008).
56 Schuler, B. & Eaton, W. A. Protein folding studied by single-molecule FRET. Curr. Opin. Struct. Biol. 18, 16-26 (2008).
57 Farinha, J. P. S. & Martinho, J. M. G. Resonance Energy Transfer in Polymer Nanodomains. J. Phys. Chem. C 112, 10591-10601 (2008).
58 Förster, T. Zwischenmolekulare Energiewnderung und Fluoreszenz. Annalen der Physik 2, 55-75 (1948).
59 Steinberg, I. Z. & Katchalski, E. Theoretical Analysis of the Role of Diffusion in Chemical Reactions, Fluorescence Quenching, and Nonradiative Energy Transfer. J. Chem. Phys. 48, 2404-2410 (1968).
60 Thomas, D. D., Carlsen, W. F. & Stryer, L. Fluorescence energy transfer in the rapid-diffusion limit. PNAS 75, 5746-5750 (1978).
61 Lakowicz, J. R. Principles of Fluorescence Spectroscopy. 3rd edn, (Springer, 2006).
62 Khan, Y. R., Dykstra, T. E. & Scholes, G. D. Exploring the Förster limit in a small FRET pair. Chem. Phys. Lett. 461, 305-309 (2008).
114
63 Förster, T. Experimentelle und theoretische Untersuchung des zwischenmolekularen Ubergangs von Elektronenanregungsenergie. Z. Naturforsch 4a, 321 (1949).
64 Hardin, B. E. et al. Increased light harvesting in dye-sensitized solar cells with energy relay dyes. Nat Photon 3, 406-411 (2009).
65 Parker, D. Luminescent lanthanide sensors for pH, pO2 and selected anions. Coord. Chem. Rev. 205, 109-130 (2000).
66 Ito, S. et al. Fabrication of thin film dye sensitized solar cells with solar to electric power conversion efficiency over 10%. Thin Solid Films 516, 4613-4619 (2008).
67 Sabbatini, N., Guardigli, M., Lehn, J.-M. & Mathis, G. Luminescence of lanthanide cryptates: effects of phosphate and iodide anions. J. Alloys Compd. 180, 363-367 (1992).
68 Elkana, Y., Feitelson, J. & Katchalski, E. Effect of Diffusion on Transfer of Electronic Excitation Energy. J. Chem. Phys. 48, 2399-2404 (1968).
69 Bunzli, J.-C. G. & Piguet, C. Taking advantage of luminescent lanthanide ions. Chemical Society Reviews 34, 1048-1077 (2005).
70 Yum, J. H. et al. Panchromatic Response in Solid-State Dye-Sensitized Solar Cells Containing Phosphorescent Energy Relay Dyes. Angewandte Chemie International Edition 48, 9277-9280 (2009).
71 Ding, I. K. et al. Pore-Filling of Spiro-OMeTAD in Solid-State Dye Sensitized Solar Cells: Quantification, Mechanism, and Consequences for Device Performance. Advanced Functional Materials 19, 2431-2436 (2009).
72 Lakowicz, J. R. Principles of Fluorescence Spectroscopy. 2nd edn, (Plenum US, 1999).
73 Yum, J.-H. et al. Phosphorescent energy relay dye for improved light harvesting response in liquid dye-sensitized solar cells. Energy & Environmental Science 3, 434-437 (2010).
74 Mac, M., Wach, A. & Najbar, J. Solvents effects on the fluorescence quenching of anthracene by iodide ions. Chemical Physics Letters 176, 167-172 (1991).
75 Lu, Y., Choi, D.-J., Nelson, J., Yang, O. B. & Parkinson, B. A. Adsorption, Desorption, and Sensitization of Low-Index Anatase and Rutile Surfaces by the Ruthenium Complex Dye N3. Journal of The Electrochemical Society 153, E131-E137 (2006).
76 Ushiroda, S., Ruzycki, N., Lu, Y., Spitler, M. T. & Parkinson, B. A. Dye Sensitization of the Anatase (101) Crystal Surface by a Series of Dicarboxylated Thiacyanine Dyes. Journal of the American Chemical Society 127, 5158-5168 (2005).
77 Siegers, C. et al. Overcoming Kinetic Limitations of Electron Injection in the Dye Solar Cell via Coadsorption and FRET. Chemphyschem 9, 793-798 (2008).
78 Mor, G. K., Shankar, K., Paulose, M., Varghese, O. K. & Grimes, C. A. Use of Highly-Ordered TiO2 Nanotube Arrays in Dye-Sensitized Solar Cells. Nano Letters 6, 215-218, doi:doi:10.1021/nl052099j (2006).
115
79 Law, M., Greene, L. E., Johnson, J. C., Saykally, R. & Yang, P. Nanowire dye-sensitized solar cells. Nature Materials 4, 455-459 (2005).
80 Hill, Z. B., Rodovsky, D. B., Leger, J. M. & Bartholomew, G. P. Synthesis and utilization of perylene-based n-type small molecules in light-emitting electrochemical cells. Chemical Communications, 6594-6596, doi:DOI: 10.1039/b814913e (2008).
81 Wurthner, F. Perylene bisimide dyes as versatile building blocks for functional supramolecular architectures. Chemical Communications, 1564-1579, doi:DOI: 10.1039/b401630k (2004).
82 Sommeling, P. M. et al. Influence of a TiCl4 post-treatment on nanocrystalline TiO2 films in dye-sensitized solar cells. Journal of Physical Chemistry B 110, 19191-19197 (2006).
83 Yum, J.-H. et al. Effect of Coadsorbent on the Photovoltaic Performance of Zinc Pthalocyanine-Sensitized Solar Cells. Langmuir 24, 5636-5640, doi:doi:10.1021/la800087q (2008).
84 Haque, S. A. et al. Charge Separation versus Recombination in Dye-Sensitized Nanocrystalline Solar Cells: the Minimization of Kinetic Redundancy. Journal of American Chemical Society 127, 3456-3462, doi:doi:10.1021/ja0460357 (2005).
85 Huang, S., Schlichthorl, G., Nozik, A., Gratzel, M. & Frank, A. Charge Recombination in Dye-sensitized Nanocrystalline TiO2 Solar Cells. J. Phys. Chem. B 101, 2576-2582 (1997).
86 Kebede, Z. & Lindquist, S.-E. Donor-acceptor interaction between non-aqueous solvents and I2 to generate I-3, and its implication in dye sensitized solar cells. Solar Energy Materials and Solar Cells 57, 259-275 (1999).
87 Fabregat-Santiago, F., Bisquert, J., Garcia-Belmonte, G., Boschloo, G. & Hagfeldt, A. Influence of electrolyte in transport and recombination in dye-sensitized solar cells studied by impedance spectroscopy. Solar Energy Materials and Solar Cells 87, 117-131 (2005).
88 Haque, S. A. et al. Charge separation versus recombination in dye-sensitized nanocrystalline solar cells: The minimization of kinetic redundancy. Journal of the American Chemical Society 127, 3456-3462 (2005).
89 Harm-Anton Klok, J. R. H. S. B. K. Star-shaped fluorescent polypeptides. Journal of Polymer Science Part A: Polymer Chemistry 39, 1572-1583 (2001).
90 Hammond, P. R. Laser dye DCM, its spectral properties, synthesis and comparison with other dyes in the red. Optics Communications 29, 331-333 (1979).
91 Bourson, J. & Valeur, B. Ion-responsive fluorescent compounds. 2. Cation-steered intramolecular charge transfer in a crowned merocyanine. The Journal of Physical Chemistry 93, 3871-3876 (2002).
92 Drake, J. M., Lesiecki, M. L. & Camaioni, D. M. Photophysics and cis-trans isomerization of DCM. Chemical Physics Letters 113, 530-534 (1985).
93 Wang, Z.-S., Kawauchi, H., Kashima, T. & Arakawa, H. Significant influence of TiO2 photoelectrode morphology on the energy conversion efficiency of
116
N719 dye-sensitized solar cell. Coordination Chemistry Reviews 248, 1381-1389 (2004).
94 Zhang, Z. et al. The Electronic Role of TiO2 Light-Scattering Layer in Dye-Sensitized Solar Cells Z. Phys. Chem. 221, 319-327, doi:10.1524/zpch.2007.221.3.319 (2007).
95 Rothenberger, G., Comte, P. & Grätzel, M. A contribution to the optical design of dye-sensitized nanocrystalline solar cells. Solar Energy Materials and Solar Cells 58, 321-336 (1999).
96 Hoke, E. T., Hardin, B. E. & McGehee, M. D. Modeling the efficiency of Förster resonant energy transfer from energy relay dyes in dye-sensitized solar cells. Opt. Express 18, 3893-3904 (2010).
97 Papageorgiou, N., Barbe, C. & Grätzel, M. Morphology and Adsorbate Dependence of Ionic Transport in Dye Sensitized Mesoporous TiO2 Films. The Journal of Physical Chemistry B 102, 4156-4164 (1998).
98 Kievsky, Y. Y. et al. Dynamics of molecular diffusion of rhodamine 6G in silica nanochannels. The Journal of Chemical Physics 128, 151102-151105 (2008).
99 Renkin, E. M. Filtration, Diffusion, and Molecular Seiving Through Porous Cellulose Membranes. The Journal of General Physiology 38, 225 (1953).
100 FRET Rate = k0 * (Ro/r)6 = 1/2ns * (6.85/1)6= 1/20fs. 101 Ono, T., Yamaguchi, T. & Arakawa, H. Study on dye-sensitized solar cell
using novel infrared dye. Solar Energy Materials and Solar Cells 93, 831-835 (2009).
102 Macor, L. et al. Near-IR sensitization of wide band gap oxide semiconductor by axially anchored Si-naphthalocyanines. Energy & Environmental Science 2, 529-534 (2009).
103 Flora, W. H., Hall, H. K. & Armstrong, N. R. Guest Emission Processes in Doped Organic Light-Emitting Diodes: Use of Phthalocyanine and Naphthalocyanine Near-IR Dopants. The Journal of Physical Chemistry B 107, 1142-1150 (2002).
104 Hellriegel, C., Kirstein, J. & Bräuchle, C. Tracking of single molecules as a powerful method to characterize diffusivity of organic species in mesoporous materials. New Journal of Physics 7, 23-23 (2005).
105 Zurner, A., Kirstein, J., Doblinger, M., Bräuchle, C. & Bein, T. Visualizing single-molecule diffusion in mesoporous materials. Nature 450, 705-708 (2007).
106 Jung, C., Hellriegel, C., Michaelis, J. & Bräuchle, C. Single-Molecule Traffic in Mesoporous Materials: Translational, Orientational, and Spectral Dynamics. Advanced Materials 19, 956-960 (2007).
107 Ito, S. et al. Calibration of solar simulator for evaluation of dye-sensitized solar cells. Solar Energy Materials and Solar Cells 82, 421-429 (2004).
108 Tennakone, K., Kumara, G. R. R. A., Kumarasinghe, A. R., Wijayantha, K. G. U. & Sirimanne, P. M. A Dye-Sensitized Nano-Porous Solid-State Photovoltaic Cell. Semicond Sci Tech 10, 1689-1693 (1995).
117
109 Hagen, J. et al. Novel hybrid solar cells consisting of inorganic nanoparticles and an organic hole transport material. Synthetic Met 89, 215-220 (1997).
110 Bach, U. et al. Solid-state dye-sensitized mesoporous TiO2 solar cells with high photon-to-electron conversion efficiencies. Nature 395, 583-585 (1998).
111 O'Regan, B., Lenzmann, F., Muis, R. & Wienke, J. A solid-state dye-sensitized solar cell fabricated with pressure-treated P25-TiO2 and CuSCN: Analysis of pore filling and IV characteristics. Chem Mater 14, 5023-5029 (2002).
112 Snaith, H. J., Zakeeruddin, S. M., Schmidt-Mende, L., Klein, C. & Grätzel, M. Ion-coordinating sensitizer in solid-state hybrid solar cells. Angew Chem Int Edit 44, 6413-6417 (2005).
113 Snaith, H. J. & Gratzel, M. The role of a "Schottky barrier" at an electron-collection electrode in solid-state dye-sensitized solar cells. Advanced Materials 18, 1910-1914 (2006).
114 Snaith, H. J. & Gratzel, M. Electron and hole transport through mesoporous TiO2 infiltrated with spiro-MeOTAD. Advanced Materials 19, 3643-3647 (2007).
115 Chen, P. et al. High Open-Circuit Voltage Solid-State Dye-Sensitized Solar Cells with Organic Dye. Nano Letters 9, 2487-2492 (2009).
116 Schmidt-Mende, L., Zakeeruddin, S. M. & Grätzel, M. Efficiency improvement in solid-state-dye-sensitized photovoltaics with an amphiphilic Ruthenium-dye. Appl Phys Lett 86, 013504 (2005).
117 Siegers, C. et al. A dyadic sensitizer for dye solar cells with high energy-transfer efficiency in the device. Chemphyschem 8, 1548-1556 (2007).
118 Tian, H. N. et al. A Triphenylamine Dye Model for the Study of Intramolecular Energy Transfer and Charge Transfer in Dye-Sensitized Solar Cells. Adv Funct Mater 18, 3461-3468 (2008).
119 Hardin, B. E. et al. in Materials Research Society Fall Meeting 120 Bach, U. et al. Charge Separation in Solid-State Dye-Sensitized Heterojunction
Solar Cells. Journal of the American Chemical Society 121, 7445-7446 (1999). 121 Yum, J. H. et al. Efficient far red sensitization of nanocrystalline TiO2 films
by an unsymmetrical squaraine dye. J Am Chem Soc 129, 10320-10321 (2007). 122 Cook, M. J. & Thomson, A. J. Luminescent Solar Collectors. Chem Brit 20,
914-917 (1984). 123 Watts, R. J. & Crosby, G. A. Quantum Efficiencies of Transition-Metal
Complexes .3. Effect of Ligand Substituents on Radiative and Radiationless Processes. J Am Chem Soc 94, 2606-& (1972).
124 Cappel, U. B., Gibson, E. A., Hagfeldt, A. & Boschloo, G. Dye Regeneration by Spiro-MeOTAD in Solid State Dye-Sensitized Solar Cells Studied by Photoinduced Absorption Spectroscopy and Spectroelectrochemistry. J. Phys. Chem. C 113, 6275-6281 (2009).
125 Lin, C. T., Bottcher, W., Chou, M., Creutz, C. & Sutin, N. Mechanism of Quenching of Emission of Substituted Polypyridineruthenium(Ii) Complexes by Iron(Iii), Chromium(Iii), and Europium(Iii) Ions. J Am Chem Soc 98, 6536-6544 (1976).
118
126 Kavan, L. & Grätzel, M. Highly Efficient Semiconducting Tio2 Photoelectrodes Prepared by Aerosol Pyrolysis. Electrochim. Acta 40, 643-652 (1995).
127 Kroeze, J. E. et al. Parameters influencing charge separation in solid-state dye-sensitized solar cells using novel hole conductors. Adv. Func. Mater. 16, 1832-1838 (2006).
128 Snaith, H. J. & Grätzel, M. Enhanced charge mobility in a molecular hole transporter via addition of redox inactive ionic dopant: Implication to dye-sensitized solar cells. Appl Phys Lett 89, 262114 (2006).
129 Hardin, B. E. et al. Increased light harvesting in dye-sensitized solar cells with energy relay dyes. Nat. Photonics 3, 406-411 (2009).
130 Chen, Y. et al. Highly Efficient co-sensitization of nanocrystalline TiO2 electrodes with plural organic dyes New Journal of Chemistry 29, 773-776 (2006).
131 Choi, H. et al. Stepwise Cosensitization of Nanocrystalline TiO2 Films Utilizing Al203 Layers in Dye-Sensitized Solar Cells. Angewandte Chemie 120, 8383-8387 (2008).
132 Ehret, A., Stuhl, L. & Spitler, M. T. Spectral Sensitization of TiO2 Nanocrystalline Electrodes with Aggregated Cyanine Dyes. The Journal of Physical Chemistry B 105, 9960-9965 (2001).
133 Miyashita, M. et al. Interfacial Electron-Transfer Kinetics in Metal-Free Organic Dye-Sensitized Solar Cells: Combined Effects of Molecular Structure of Dyes and Electrolytes. Journal of the American Chemical Society 130, 17874-17881 (2008).
134 Ishii, K. & Kobayashi, N. The Porphyrin Handbook. Vol. 16 (Elsevier Science (USA), 2003).
135 Li, X., Long, N. J., Clifford, J. N., Campbell, C. & Durrant, J. New peripherally-substituted naphthalocyanines: synthesis, characterisation and evaluation in dye-sensitized photoelectrochemical solar cells. . New J. Chem. 26, 1076-1080 (2002).
136 Kuang, D. et al. High Molar Extinction Coefficient Heteroleptic Ruthenium Complexes for Thin Film Dye-Sensitized Solar Cells. Journal of the American Chemical Society 128, 4146-4154 (2006).
137 Farzad, F., Thompson, D. W., Kelly, C. A. & Meyer, G. J. Competitive Intermolecular Energy Transfer and Electron Injection at Sensitized Semiconductor Interfaces. Journal of the American Chemical Society 121, 5577-5578 (1999).
138 Higgins, G. T., Bergeron, B. V., Hasselmann, G. M., Farzad, F. & Meyer, G. J. Intermolecular Energy Transfer across Nanocrystalline Semiconductor Surfaces. The Journal of Physical Chemistry B 110, 2598-2605 (2006).
139 Sayama, K. et al. Efficient sensitization of nanocrystalline TiO2 films with cyanine and merocyanine organic dyes. Solar Energy Materials and Solar Cells 80, 47-71 (2003).
140 Hardin, B. E. et al. High Excitation Transfer Efficiency from Energy Relay Dyes in Dye-Sensitized Solar Cells. Nano Letters 10, 3077-3083 (2010).
119
141 Snaith, H. J. et al. Charge collection and pore filling in solid-state dye-sensitized solar cells Nanotechnology 19, 424003 (2008).
142 Fabregat-Santiago, F. et al. Correlation between Photovoltaic Performance and Impedance Spectroscopy of Dye-Sensitized Solar Cells Based on Ionic Liquids. The Journal of Physical Chemistry C 111, 6550-6560 (2007).
143 Mandal, P. K. & McMurray, J. S. Pd−C-Induced Catalytic Transfer Hydrogenation with Triethylsilane. The Journal of Organic Chemistry 72, 6599-6601, doi:10.1021/jo0706123 (2007).
144 Mori, S. et al. Enhancement of Incident Photon-to-Current Conversion Efficiency for Phthalocyanine-Sensitized Solar Cells by 3D Molecular Structuralization. Journal of the American Chemical Society 132, 4054-4055, doi:10.1021/ja9109677 (2010).
145 Lee, K., Park, S. W., Ko, M. J., Kim, K. & Park, N.-G. Selective positioning of organic dyes in a mesoporous inorganic oxide film. Nat Mater 8, 665-671 (2009).
146 Mor, G. K. et al. High-Efficiency Forster Resonance Energy Transfer in Solid-State Dye Sensitized Solar Cells. Nano Letters 10, 2387-2394 (2010).
147 Buhbut, S. et al. Built-in Quantum Dot Antennas in Dye-Sensitized Solar Cells. ACS Nano 4, 1293-1298 (2010).
148 Ito, S. et al. High-efficiency (7.2%) flexible dye-sensitized solar cells with Ti-metal substrate for nanocrystalline-TiO2 photoanode. Chemical Communications, 4004-4006 (2006).
149 Asghar, M. I. et al. Review of stability for advanced dye solar cells. Energy & Environmental Science 3, 418-426 (2010).
150 Kato, N. et al. Improvement in long-term stability of dye-sensitized solar cell for outdoor use. Solar Energy Materials and Solar Cells In Press, Corrected Proof, doi:DOI: 10.1016/j.solmat.2010.04.019.