South Dakota State University South Dakota State University

Open PRAIRIE: Open Public Research Access Institutional Open PRAIRIE: Open Public Research Access Institutional

Repository and Information Exchange Repository and Information Exchange

Electronic Theses and Dissertations

2018

Urea and rGO Additives to Iodine/Triiodide Electrolyte for Higher Urea and rGO Additives to Iodine/Triiodide Electrolyte for Higher

Efficiency Dye-sensitized Solar Cells Efficiency Dye-sensitized Solar Cells

Salem Abdulkarim South Dakota State University

Follow this and additional works at: https://openprairie.sdstate.edu/etd

Part of the Electrical and Computer Engineering Commons

Recommended Citation Recommended Citation Abdulkarim, Salem, "Urea and rGO Additives to Iodine/Triiodide Electrolyte for Higher Efficiency Dye-sensitized Solar Cells" (2018). Electronic Theses and Dissertations. 2945. https://openprairie.sdstate.edu/etd/2945

This Dissertation - Open Access is brought to you for free and open access by Open PRAIRIE: Open Public Research Access Institutional Repository and Information Exchange. It has been accepted for inclusion in Electronic Theses and Dissertations by an authorized administrator of Open PRAIRIE: Open Public Research Access Institutional Repository and Information Exchange. For more information, please contact michael.biondo@sdstate.edu.

UREA AND rGO ADDITIVES TO IODIDE/TRIIODIDE ELECTROLYTE FOR

HIGHER EFFICIENCY DYE-SENSITIZED SOLAR CELLS

BY

SALEM ABDULKARIM

A dissertation submitted in partial fulfillment of the requirements for the

Doctor of Philosophy

Major in Electrical Engineering

South Dakota State University

2018

iii

ACKNOWLEDGEMENTS

My deep gratitude goes first to Dr. Qiquan Qiao as dissertation advisor, who expertly

guided and supported me through my PhD research at the Center of Advanced

Photovoltaics, South Dakota State University. His valuable suggestions and continuous

encouragements have helped me to conduct my research successfully and attain skills to

be an independent researcher.

My appreciations extend to Dr. Parashu Kharel, Dr. Robert McTaggart, Dr. Yue Zhou,

and Dr. Michael Hildreth for their important role and support as committee members.

I am also thankful to all my colleagues, especially Dr. Hytham Elbohy in helping me

to complete my experiments.

Finally, I am grateful to my family for their support, patience during my PhD studies.

iv

TABLE OF CONTENTS

LIST OF FIGURES ...............................................................................................................viii

LIST OF TABLES ...................................................................................................................xi

ABSTRACT ...........................................................................................................................xii

Chapter 1: Introduction ............................................................................................................. 1

Overview ......................................................................................................................... 1

Solar Photovoltaic Cells ........................................................................................... 2

DSSC Overview .............................................................................................................. 4

DSSC Substrate Characteristics ............................................................................... 5

Metal Oxide Electrodes ............................................................................................ 6

Photosensitizers (Dyes) for DSSCs ......................................................................... 7

Counter Electrodes (Cathode) ................................................................................ 10

DSSC Electrolytes ................................................................................................. 11

Advantages and Disadvantages of Dye-Sensitized Solar Cells .................................... 16

Advantages ............................................................................................................. 17

Disadvantages ........................................................................................................ 17

Previous Work .............................................................................................................. 18

Dye-Sensitized Solar Cell Liquid Electrolytes ...................................................... 18

Iodide/Triiodide Electrolyte Additives .................................................................. 22

Motivations ................................................................................................................... 23

v

Objective ....................................................................................................................... 24

Chapter 2: Theory ................................................................................................................... 25

Solar Radiation and Air Mass ....................................................................................... 25

Solar Cells and Photovoltaic Effect .............................................................................. 27

i. Charge Carrier Generation Due to Photon Absorption in Junction Materials .............. 28

ii. Separation of Photo-Generated Charge Carriers in the Junction .................................. 30

iii. Accumulation of Photo-Generated Charge Carriers at the Junction Terminals ........... 30

Classical P-N Junction Solar Cell ................................................................................. 31

DSSC Operating Principles........................................................................................... 32

Key Efficiency Parameters Representing DSSC Performance ..................................... 34

Short-Circuit Current Density ................................................................................ 35

Open-Circuit Voltage ............................................................................................. 36

Fill Factor ............................................................................................................... 36

Conversion Efficiency ........................................................................................... 37

Equivalent Circuit Model for Conventional PN-Junction Solar Cells .......................... 38

Characterization Techniques ......................................................................................... 39

Current Density-Voltage Characteristics (J-V) ...................................................... 39

External Quantum Efficiency (EQE) ..................................................................... 40

Electrochemical Impedance Spectroscopy (EIS) ................................................... 41

Cyclic Voltammogram Measurements (CVs) ........................................................ 43

UV-Visible Absorption Spectra ............................................................................. 46

vi

Mechanism of Urea Reaction with I-/I-3 Redox Electrolyte .......................................... 48

Mechanism of rGO Reaction with I-/I-3 Redox Electrolyte .......................................... 49

Chapter 3: Experimental Procedures ...................................................................................... 51

Material and Device Fabrication Procedures ................................................................ 51

Materials ................................................................................................................ 51

Device Fabrication – Glass Substrate Preparation ................................................. 51

Device Fabrication – Photoanode Fabrication ....................................................... 52

Device Fabrication – Counter Electrode (CE) Fabrication .................................... 52

Device Fabrication – Final Assembly .................................................................... 53

Device Fabrication – Electrolyte Preparation ........................................................ 54

Materials and Device Characterizations ....................................................................... 54

Current Density-Voltage Characteristics (J-V) ...................................................... 54

External Quantum Efficiency (EQE) Measurements ............................................. 56

Electrochemical Impedance Spectroscopy (EIS) Measurements .......................... 57

Cyclic voltammetry (CV) Measurements .............................................................. 57

UV-Visible Absorption Spectra Measurements..................................................... 58

Chapter 4: Results and Analysis ............................................................................................. 61

DSSC Performance with Urea-Treated Electrolyte ...................................................... 61

J-V Characteristics ................................................................................................. 61

UV-Vis Absorbance Spectra .................................................................................. 63

External Quantum Efficiency (EQE) ..................................................................... 64

Electrochemical Impedance Spectroscopy (EIS) ................................................... 66

vii

Cyclic Voltammogram (CV) Measurement ........................................................... 68

UV-Vis Absorption of Photoanodes Immersed in Urea-Treated Electrolyte (dye

loading) ..................................................................................................................... 69

J-V Hysteresis Effects ............................................................................................ 70

DSSC Performance Using rGO-Treated Electrolyte .................................................... 72

J-V Characteristics ................................................................................................. 72

UV-Vis Absorbance Spectra .................................................................................. 74

External Quantum Efficiency (EQE) ..................................................................... 74

Electrochemical Impedance Spectroscopy ............................................................ 76

Cyclic Voltammogram Measurement .................................................................... 78

UV-Vis Absorption of Photoanodes Immersed in rGO-Treated Electrolyte (dye

loading) ..................................................................................................................... 80

J-V Hysteresis Effects ............................................................................................ 81

Chapter 5: Summary and Conclusions .................................................................................... 83

Summary ....................................................................................................................... 83

Conclusions ................................................................................................................... 84

Future work ................................................................................................................... 85

References ………………………………………………………………………….. … ....... 86

viii

LIST OF FIGURES

Figure 1.1 Solar Cell Technology Development _______________________________ 3

Figure 1.2 Solar Cell Research Efficiency Reported Over Last Four Decades ________ 4

Figure 1.3 Classification of DSSC Electrolytes _______________________________ 12

Figure 2.1 Solar Radiation Spectrum _______________________________________ 25

Figure 2.2 Solar Zenith Angle ____________________________________________ 26

Figure 2.3 Solar Reference Spectra ________________________________________ 27

Figure 2.4 Photon Absorption in a Semiconductor with Bandgap Energy Eg ________ 29

Figure 2.5 Direct and Indirect Bandgap Semiconductor Characteristics ____________ 30

Figure 2.6 Energy Band Diagram of a p-n Junction in Thermal Equilibrium, with

Depletion Region ______________________________________________________ 32

Figure 2.7 Schematic of DSSC with Energy Level Diagram _____________________ 34

Figure 2.8 J-V Characteristic Curves Under Dark and Illuminated Conditions _______ 37

Figure 2.9 Conventional P-N Junction Solar Cell Equivalent circuit model _________ 38

Figure 2.10 Changes in I-V Response with Changes in Rsh and Rs ________________ 39

Figure 2.11 Example DSSC Nyquist Plot with Equivalent Circuit Model ___________ 43

Figure 2.12 Cyclic Potential Sweep and (b) Shape of a Cyclic Voltammogram Curve _ 45

Figure 2.13 Energetics Between Electrolyte Solution and Electrode _______________ 46

Figure 2.14 Possible Transitions of π, σ, and n Electrons _______________________ 47

Figure 2.15 UV-Vis Spectroscopy _________________________________________ 48

Figure 2.16 Mechanism of Urea Reaction with I-/I-3 Redox Electrolyte ____________ 49

Figure 2.17 Mechanism of rGO Reaction with I-/I-3 Redox Electrolyte _____________ 50

Figure 3.1 Schematic of DSSC device assembly ______________________________ 54

ix

Figure 3.2 Schematic of J-V Measurement Setup _____________________________ 55

Figure 3.3 Schematic of EQE Measurement Setup ____________________________ 56

Figure 3.4 Functional Diagram of EIS Setup _________________________________ 57

Figure 3.5 Functional Diagram of Cyclic Voltammetry (CV) ____________________ 58

Figure 3.6 Schematic of UV-Visible Absorption Spectroscopy __________________ 60

Figure 4.1 DSSC Current Density-Voltage (J-V) Curves Due to Urea-Treated Electrolyte

_____________________________________________________________________ 62

Figure 4.2 UV-Vis Absorbance Spectra of Urea-Treated Electrolyte ______________ 64

Figure 4.3 EQE Spectral Response Due to Urea-Treated Electrolyte ______________ 65

Figure 4.4(a) Nyquist Plots of DSSCs Due to Urea-Treated Electrolyte ___________ 67

Figure 4.5 Cyclic Voltammograms of Counter Electrodes Immersed in Urea-Treated

Electrolyte ____________________________________________________________ 69

Figure 4.6 UV-Vis Absorbance Spectra of Dye Solutions Desorbed from Electrolytes

Treated with Urea ______________________________________________________ 70

Figure 4.7 Hysteresis Effects for DSSC Devices Fabricated with Urea-Treated

Electrolyte ____________________________________________________________ 71

Figure 4.8 DSSC Current Density-Voltage (J-V) Curves Due to rGO-Treated Electrolyte

_____________________________________________________________________ 73

Figure 4.9 UV-Vis Absorbance Spectra of rGO-Treated Electrolyte ______________ 74

Figure 4.10 EQE Spectral Response Due to rGO-Treated Electrolyte ______________ 75

Figure 4.11(a) Nyquist Plots of DSSCs Due to rGO-Treated Electrolyte ___________ 77

Figure 4.12 Cyclic Voltammograms of Counter Electrodes Immersed in rGO-Treated

Electrolyte ____________________________________________________________ 79

x

Figure 4.13 UV-Vis Absorbance Spectra of Dye Solutions Desorbed from Electrolytes

Treated with rGO ______________________________________________________ 81

Figure 4.14 Hysteresis Effects for DSSC Devices Fabricated with rGO-Treated

Electrolyte ____________________________________________________________ 82

xi

LIST OF TABLES

Table 4.1 DSSC Performance Parameters Due to Urea-Treated Electrolyte _________ 62

Table 4.2 EQE Peaks of DSSCs Due to Urea-Treated Electrolyte _________________ 65

Table 4.3 Predicted Rs and Rr from Nyquist Plots of Urea-Treated Electrolyte ______ 68

Table 4.4 Anodic and Cathodic Peak Intensities of Urea-Treated Electrolyte ________ 69

Table 4.5 DSSC Performance Due to rGO-Treated Electrolyte ___________________ 73

Table 4.6 EQE Peaks of DSSCs Due to rGO-Treated Electrolyte _________________ 76

Table 4.7 Predicted Rs and Rr from Nyquist Plots of rGO-Treated Electrolyte _______ 78

Table 4.8 Anodic and Cathodic Peak Intensities of rGO-Treated Electrolyte ________ 79

xii

ABSTRACT

UREA AND rGO ADDITIVES TO IODIDE/TRIIODIDE ELECTROLYTE FOR

HIGHER EFFICIENCY DYE-SENSITIZED SOLAR CELLS

SALEM ABDULKARIM

2018

This research investigated urea and reduced graphene oxide (rGO) as potential

additives to a dye-sensitized solar cell (DSSC) iodide/triode electrolyte, in order to enhance

overall electrical performance. Additive concentrations of 0, 2, 5, 10, and 15 wt% were

investigated. In general, both additives were found to enhance DSSC electrical

performance with respect to open circuit voltage Voc and DSSC fill factor FF. The greatest

enhancement in these parameters occurred at an additive concentration of 10 wt%. Urea

produced a greater degree of enhancement than rGO; the overall efficiency was increased

by 10.17±0.01% for urea vs 9.57±0.01% for rGO. The specific mechanisms by which

performance was increased differed between the additives. This degree of performance

enhancement may be near the limits of what is achievable given current DSSC technology

and development.

1

Chapter 1: Introduction

Overview

Since the start of the Industrial Revolution, fossil fuel usage (e.g. crude oil, coal, natural

gas, or heavy oils) has contributed more carbon dioxide (CO2) to the air than the Earth's

plants can use, resulting in increasing temperatures worldwide. For example, from 1890 to

2015, atmospheric carbon dioxide emissions increased from 270 to 400 ppm,

corresponding to a temperature increase that contributes to the greenhouse effect [1]. This

temperature increase is prompting more research into use of renewable energy resources

(e.g., wind, biomass, and solar energy [2]) that do not emit air pollution or greenhouse

gasses. Wind power generation depends on wind speed, which is affected by altitude,

terrain and/or man-made features; it cannot reliably supply steady electricity to power

systems [3, 4]. Biomass includes all organic wastes and all land and water-based

vegetation. Biomass has been a major energy source, contributing on the order 14% of the

world’s energy supply [5]. However, biomass is a fuel that needs to be manufactured for

use and still produces carbon dioxide during the processing.

Solar energy takes approximately 8 minutes to reach the Earth. It is produced through

nuclear fusion of hydrogen into helium. All forms of fossil fuel energy were produced

through photosynthetic processes, making solar energy the primary energy source. Even

wind energy has a solar origin, since it comes from temperature differences in different

regions on the Earth. Considering solar energy by itself, the power generated from sunlight

striking the Earth's surface is approximately 1.05 10 TW. If only 1% of the irradiance

at the Earth’s surface could be directly converted into electrical energy with just 10%

efficiency, it would provide 105 TW, supplying more than the expected worldwide need

2

of 25-30 TW [6].

Solar Photovoltaic Cells

Solar photovoltaic cells (PVs), which convert sunlight directly into electrical energy,

are a promising alternative to existing fossil fuel energy sources [7] and will likely

contribute to future energy production by significantly reducing atmospheric CO2 levels.

Figure 1.1 shows the various solar cell technologies and their historical and recent

development trends [8]. The technologies considered in Figure 1.1 include multi-junction

cells, single-junction GaAs, crystalline Si cells, thin film technologies, and emerging PV

technologies.

In 1839, Becquerel first observed the photovoltaic (PV) effect in selenium. In 1883,

Fritte introduced the first solid state solar cell with a power conversion efficiency (PCE) of

less than 1% [9]. In 1954, Bell Laboratories in the U.S. constructed a solar PV device with

a PCE of approximately 6% [6]. In 1958, solar cells were used in small-scale and

commercial applications. After much research, the conversion efficiency was enhanced to

approximately 19.8% for a multicrystalline Si cell and approximately 24.4% for a

monocrystalline Si cell [10]. This first commercially viable generation of solar cells was

fabricated from Si wafers and used primarily as passive rooftop solar collectors. The

advantage of these solar cells is good photoconversion efficiency. However, they are

mechanically rigid, and they require a lot of energy to fabricate.

The second generation of solar cells was fabricated using amorphous or

microcrystalline Si or other materials such as copper indium gallium selenide (CIGS) and

cadmium telluride (CdTe). These cells avoid the use of Si wafers and consume less raw

material during fabrication, significantly reducing their production costs compared to the

3

earlier crystalline Si cells. However, their fabrication also requires large amounts of

energy.

The current generation of solar cells is fabricated with advanced polymers, oligomers,

synthetic dyes, and organic/inorganic hybrid perovskites. Due to high fabrication costs at

the current time [11], these cells are mostly used in large-scale commercial applications

only.

Figure 1.1 Solar Cell Technology Development

Figure 1.2 shows the highest achieved photoconversion efficiencies for different

photovoltaic technologies, covering a time period from 1976 to 2017. Efficiencies vary

from approximately 10.6% for organic tandem cells to 46.0% for a four-junction

(concentrator) cell. For dye-sensitized solar cells, the highest achieved efficiency is

4

approximately 11.9%. For perovskite and CIGS cells, the highest achieved efficiencies are

approximately 22.1% and 22.4%, respectively.

Figure 1.2 Solar Cell Research Efficiency Reported Over Last Four Decades

(Source: National Renewable Energy Laboratory (NREL), 2017)

Because of their higher costs and energy usage required for fabrication, Si-based solar

cells are no longer used in commercial applications. Organic solar cells are promising as

potential replacements for silicon-based cells [12, 13]. With respect to efficiency and ease

of fabrication, the dye-sensitized solar cell (DSSC) is one of the most promising

alternatives. The next section considers these cells in additional detail.

DSSC Overview

DSSCs have attracted significant attention as a solar cell due to lower material costs

and ease of fabrication. The first DSSC was constructed by O'Regan and Gratzel in 1991.

5

Its photoconversion efficiency was 7.1%, and its incident photon to current conversion

efficiency (IPCE) was approximately 80% [14]. Their original design consists of two glass

plates coated with a transparent conducting oxide, typically a fluoride-doped tin oxide

(FTO), forming transparent conducting oxide-glass (TCO) substrates. A mesoporous TiO2

semiconductor thin-film “paste” is deposited onto one of the plates, which is immersed in

a photosensitive ruthenium-polypyridine dye solution. After soaking in the dye solution,

dye molecules are covalently bonded within the mesoporous TiO2 layer. A thin layer of

platinum is spin coated onto the second glass plate. The plates are then sealed together,

and an I-/I-3 redox electrolyte solution is injected between the sealed plates.

Recently developed DSSC cells are generally fabricated following this initial design.

They differ primarily in the electrolyte/dye solution and electrode/photoanode material that

is used [15].

DSSC Substrate Characteristics

As described in the previous section, the DSSC can be considered to have a “sandwich”

structure, with the electrolyte/dye solution between the two TCO substrates. The substrates

should possess low sheet resistance and high transparency in the visible and IR regions of

the electromagnetic spectrum. The typical sheet resistance for a well fabricated TCO is

approximately 5 to 15 Ω / square [16].

There are two types of TCO substrates typically used for DSSCs: indium-doped tin

oxide (ITO) and FTO. FTO substrates tend to be more efficient with respect to overall

photoconversion efficiency than ITO substrates; this is due to lower resistivity (8.5 Ω/sq)

in the FTO substrate than in the ITO substrate (18.5 Ω/sq), resulting in lower ohmic losses.

In addition, FTO substrates exhibit better thermal stability than ITO substrates; after

6

annealing at 450 °C, the ITO resistivity increased to approximately 52 Ω/sq, while the FTO

resistivity remained unchanged [17, 18].

Metal Oxide Electrodes

Many metal oxide semiconductor materials, such as TiO2, SnO2, ZnO, and ZrO2, have

been used for the DSSC photoanode [19]. TiO2 is the most commonly used material due to

its superior performance, lower material cost, relative abundance, relative nontoxicity,

biocompatibility, and chemically stable nature [20, 21]. The mesoporous TiO2 film has a

high surface area, allowing it to absorb a large amount of dye in the monolayer. This i)

helps to ensure consistent photoelectron generation; and ii) facilitates dispersion of

electrolyte in the photoanode, which ensures the regeneration of oxidized dye by the redox

couple in the electrolyte [22]. In a typical DSSC device, the film thickness is between

approximately 2 µm and 15 µm.

TiO2 exists in three different phases dependent on temperature: rutile, anatase and

brookite. Rutile is the most stable form; it has a slightly lower bandgap compared to anatase

and can absorb at near-ultraviolet wavelengths. Anatase is more chemically active than

rutile. Anatase and brookite are stable at normal temperatures but slowly convert to rutile

upon heating to temperatures above 550° C and 750° C, respectively [15]. For a given film

thickness at an intensity level corresponding to normal sunlight, the measured photocurrent

voltages of rutile-based and anatase-based DSSCs are quite similar. Voc in both cells is

essentially the same; the short-circuit photocurrent of the rutile-based cell is approximately

30% less than that of the anatase-based cell, due to a smaller film surface area per unit

volume resulting in less adsorbed dye. The overall conversion efficiency of the rutile-based

cell is less than the efficiency of the anatase-based cell [21].

7

Photosensitizers (Dyes) for DSSCs

A critical component of the DSSC is the photosensitizer dye. Ideally, the

photosensitizer should meet the following requirements [15, 23]:

• It should be strongly absorbing in the visible and near-infrared (NIR) regions.

• Its molecular structure should include anchoring “groups” (e.g. -COOH, -

H2PO3, -SO3H, etc.) to promote effective entry of dye molecules into the

semiconductor mesoporous film.

• For n-type DSSCs (based on n-type semiconductors), the excited state energy

of the photosensitizer should be higher than the semiconductor conduction

band edge, so that efficient electron transfer between the excited dye and

semiconductor can occur. For p-type DSSCs (based on p-type

semiconductors), the highest occupied molecular orbital (HOMO) level of the

photosensitizer should be at a more positive potential than the semiconductor

valence band level, in order for efficient hole transfer to occur.

• For dye regeneration, the oxidized state level of the photosensitizer must be

more positive than the redox potential of the electrolyte.

• The photosensitizer should be chemically and thermally stable.

• The photosensitizer should have good solubility in organic solvents for facile

deposition from stock solutions within a few hours.

Many types of photosensitizer dyes have been synthesized since DSSC technology was

first introduced. These dyes can be classified into four groups [24-27]. Each group will be

considered in greater detail in this section.

• metal complex dyes

8

• metal-free organic dyes

• perovskite-based dyes

• “natural” dyes derived from plant material

Metal Complex Dyes

Photosensitizer dyes containing metal complexes are the most commonly used dyes in

DSSCs. Dyes with metal complexes based on Ru, Os [28], Re[29], Fe[30], Pt[31], and

Cu[32]. have been synthesized. DSSCs using Ru-based dyes have been shown to provide

a maximum overall photoconversion efficiency of 13% [33]. However, their use is limited

by high production costs, scarcity and toxicity of Ru metal, and inefficient absorption in

the NIR region. Research into modifying the basic molecular design in order to increase

absorption in the visible region has been conducted; these variant dyes include N3 (“red

dye”), N719, N749 (“black dye”), K19, and Z907 [34]. At the present time, only N3 and

N719 have demonstrated increased visible and NIR region absorption.

Metal-Free Organic Dyes

Metal-free organic dyes have been synthesized as potential replacements for Ru based

dyes. They are generally non-toxic, relatively inexpensive and easy to synthesize, offer

very high molar extinction coefficients, and can be readily “tuned” to enhance absorption

in different regions of the spectrum. In addition, they are generally stable at high

temperatures and under prolonged illumination [35, 36]. These dyes have reached

maximum photoconversion efficiencies of 12.75% [37].

Metal-free organic dyes are generally divided into three major sub-types with respect

to molecular structure: “donors”, “linkers”, and “acceptors”. Linkers are usually a π-

conjugated system that link electron-donating (D) and electron- accepting (A) groups by

9

bridges, the so-called “D-π-A” sensitizers [38, 39]. In metal-free organic dye sensitizers,

the donor groups should consist of electron-rich compounds such as phenylamine,

aminocoumarin, and indoline. The most successful π-conjugated groups are selected from

compounds containing thiophene units such as oligothiophenes, thienylenevinylenes, or

dithienothiophene. Cyanoacrylic acid, rhodamines, and pyridines are also suitable acceptor

compounds.

Perovskite Dyes

Perovskite dyes have specific crystalline structures following the general formula

ABX3, where X can be oxygen, carbon, nitrogen, or a halide element. Current halide

perovskites with the general structure CH3NH3PbX3 (X=Cl, Br, or I) have attracted

considerable attention due to low bandgap energies on the order of 1.5 eV. This class of

perovskite was first used as an inorganic sensitizer in DSSCs in 2009; they demonstrated

photoconversion efficiencies of approximately 3.1% for X=Br and 3.8% for X=I [40]. In

general, perovskite solar cells offer greater photoconversion efficiency at lower cost and

ease of manufacturing.

Natural Dyes

Natural dyes can be easily extracted from plant and bacterial sources, and employed in

DSSCs [41]. They provide increased absorption in the visible region. In addition, they are

a relatively abundant dye source and are environmentally friendly in that they are

completely biodegradable [42]. Most importantly, synthesis of natural dyes for DSSC use

is cost effective in that they do not rely on noble metals such as Ru. However, they are not

particularly efficient; a maximum photoconversion efficiency of 2% has been reported for

a pomegranate dye [43].

10

Counter Electrodes (Cathode)

The counter electrode (CE) or cathode is the site in the DSSC where reduction

processes occur. Typically, the counter electrode is a glass substrate coated with platinum

(Pt) or other conductor; it effectively functions as an electron collector. The ideal CE

material should meet the following requirements [44, 45]:

• High catalytic activity

• High conductivity

• High reflectivity

• Large surface area

• Optimum thickness

• Chemical, electrochemical and mechanical stability

• Resistance to chemical corrosion resistance

• Energy levels that match the potential of the redox couple electrolyte

• Good adhesivity with TCO

• Low cost

However, one material cannot easily meet all of these requirements simultaneously.

Modern DSSCs use composite CEs produced by various material deposition methods

including thermal decomposition, electrochemical deposition, chemical reduction,

chemical vapor deposition, hydrothermal reaction, and sputter deposition. The deposition

method significantly influences CE particle size, surface area and smoothness, as well as

its catalytic and electrochemical properties. Smaller particle sizes and larger surface areas

produce more efficient catalytic activity. Pt satisfies nearly all the requirements for CEs;

consequently, it is the most frequently used CE material. However, its increasing cost and

11

chemical instability limit its use. Among Pt-free CEs, carbon-based materials have been

the most researched material due to their low cost, ease of manufacturing, and good

stability. However, they possess relatively low conductivity and catalytic activity

compared to Pt CEs, and do not adhere well to various substrates.

DSSC Electrolytes

The selection of an electrolyte has a significant impact on overall DSSC

photoconversion efficiency, to the extent that it facilitates the charge transport between the

photoanode and the counter electrode. The ideal DSSC electrolyte material should possess

the following characteristics [46]:

• Very low viscosity

• Negligible vapor pressure

• High boiling point

• High conductivity

• Minimal absorption in the visible region of the spectrum

• Long-term mechanical, chemical, electrochemical, thermal and optical

stabilities

• Chemically non-reactive with the sensitized dye material

The electrolytes used in DSSCs can be classified into three categories, as shown in

Figure 1.3: (i) liquid electrolytes; (ii) quasi-solid electrolytes; and (iii) solid-state

conductors. The following sections consider each category in greater detail.

12

Figure 1.3 Classification of DSSC Electrolytes

Liquid Electrolytes

Liquid electrolytes are the most widely utilized transport medium for DSSCs, yielding

photoconversion efficiencies as high as 13% for traditional DSSC designs [33]. They are

easy to synthesize, are highly conductive, and exhibit low viscosity. Their high efficiency

is provided through efficient interfacial wetting to the electrodes [47].

In general, a liquid electrolyte consists of three main components: the solvent, redox

couples, and various additives. Each component is considered in greater detail below.

Solvent:

The solvent is the major component in a liquid electrolyte, as it provides the

environment for the dissolution and diffusion of the redox couples. Solvents typically used

in DSSCs possess the following characteristics [48, 49]:

13

• Melting point below −20 °C and boiling point above 100 °C

• Sufficient chemical stability whether illuminated or not

• High dielectric constant

• Good solubility

• Low viscosity, so that redox mediators possess high diffusion coefficients and the

liquid electrolytes will have high conductivity

• Minimal light absorption

• Chemically inert with the surface-attached dye, including the dye−metal-oxide (or

another semiconductor) bond

• Poor solubility to the sealant materials

• Low toxicity

• Low cost

Two types of solvents meeting the above requirements have been used in DSSC designs:

polar solvents, and highly conductive ionic liquids (IL) or molten salts.

Redox Couples:

Redox-couples (RCs), or redox mediators, play an important role in DSSCs. The RCs

are responsible for charge transport between the photoanode and the counter electrode. The

electrochemical potential of the RC determines the DSSC’s open-circuit voltage (Voc) with

respect to the Fermi energy of the conducting substrate with the attached photoanode. With

regard to reaction kinetics, the RC should exhibit “asymmetric behavior” [50], i.e. electron

donation to the photo-oxidized dye should be sufficiently fast to ensure efficient dye

regeneration, and electron acceptance from the mesoporous TiO2 film should be slow

enough to decrease electron recombination losses [47].

14

To date, the most efficient RC for DSSC use is the I-/I-3 redox couple traditionally

dissolved in volatile organic solvents. This RC has good solubility, high conductivity, long-

term stability, and ability to efficiently penetrate into the mesoporous semiconductor film.

However, several drawbacks limit the industrial application of the I-/I-3 RC. First, iodine is

extremely corrosive toward metals used as current collectors in some DSSCs, such as

copper or silver, which are. With a relatively high vapor pressure, proper encapsulation of

the iodine-based cells is challenging. In addition, the I-3 ion absorbs a significant part of

visible light, stealing photons from the sensitizing dye. Finally, the redox potential of the

I-/I-3 limits the photovoltage [51].

Many alternative redox couples have been investigated as replacements for the current

I–/I–3 system. The goal is to minimize the driving force needed for dye regeneration while

increasing the device photovoltage, which affects the photoconversion efficiency. These

redox couples include halide and sulfur/selenium-organic redox couples (e.g. Br-/Br3-,

SCN-/(SCN)3-, SeCN-/(SeCN)3

- [47]), as well as transition-metal complexes (e.g.

ferrocene/ferricenium/ (Fc/Fc+), copper (I/II), cobalt (II/III), and nickel (III/IV) complexes

[52]).

Additives

The additive optimizes the DSSC’s photovoltaic performance. Specific cations or

compounds in the electrolyte are expected to adsorb onto the mesoporous semiconductor

surface. This adsorption has been found to affect fundamental photovoltaic parameters

such as the semiconductor surface charge, the shift of the conduction band edge (CB),

overall recombination kinetics etc., which in turn affects the overall photoconversion

efficiency [53].

15

Quasi-Solid-State Electrolytes

For any substance, the quasi-solid or semisolid state is a special state between the

normal solid and liquid states. Quasi-solid-state electrolytes possess, simultaneously, the

cohesiveness of solids and the diffusivity of liquids [54-56]. Consequently, quasi-solid-

state electrolytes exhibit greater long-term stability than liquid electrolytes, with higher

ionic conductivity and excellent interfacial contact. Due to these desirable characteristics,

quasi-solid electrolytes are typically used in DSSCs [57].

Quasi-solid-state electrolytes are synthesized through solidification of liquid

electrolytes with various gelating compounds; the resulting electrolytes are generally

grouped into three major kinds [46]. Organic gelators, such as polyacrylonitrile (PAN),

polystyrene (PS), polyethylene oxide (PEO or PEG), polyvinyl pyrrolidinone (PVP),

polyvinyl chloride (PVC), etc) [58] form thermoplastic or thermosetting polymer

electrolytes. Inorganic gelators, such as SiO2 or nanoclay, form composite polymer

electrolytes. Any gelating compound can be used to synthesize a quasi-solid ionic liquid

electrolyte [59] from an existing ionic liquid electrolyte.

In general, the photoconversion efficiency of a quasi-solid-state DSSC is less than that

of a DSSC employing a liquid electrolyte. Research into optimized designs and selection

of a suitable quasi-solid electrolyte should yield significant improvements in photovoltaic

performance.

Solid-State Conductors

DSSCs employing liquid electrolytes cannot avoid solvent exudation under conditions

of long-term storage or exposure to air. To address this issue, several solid-state transport

materials have been developed for replacing liquid or quasi-solid electrolytes, including

16

ionic conductors and organic/inorganic hole-transport materials. Such solid-state transport

materials exhibit good long-term mechanical and operational stability compared to dye-

sensitized photoelectrochemical cells. However, the photoconversion efficiency of

traditional DSSCs based on solid-state transport materials (including solid-state

electrolytes and hole-transport materials) is lower than for DSSCs based on liquid

electrolyte due to poor interfacial contact with the counter electrode.

The semiconductor-dye pair should possess the following characteristics [60]:

• The semiconductor has a wide band gap, which corresponds to transparency in the

portion of the visible spectrum where the dye most strongly absorbs light.

• The semiconductor deposition does not dissolve or degrade the dye monolayer on the

TiO2 nanocrystallites.

• The dye’s lowest unoccupied molecular orbital (LUMO) level is greater than TiO2’s

lowest conduction band level, and its highest occupied molecular orbital (HOMO)

level is less than the p-type semiconductor’s highest valence band level.

The most common approach to fabricate solid-state DSSCs involves the use of p-type

semiconductor materials. Fujishima, et al [61] found that copper compounds such as CuI

could effectively serve as a p-type semiconductor. In particular, CuI can be cast from

solution or vacuum deposition to form a complete hole-transporting layer, and it exhibits

conductivity in excess of 10-2 Scm-1.

Advantages and Disadvantages of Dye-Sensitized Solar Cells

DSSC technologies are among the third-generation technologies in solar cell

development. Their development addresses deficiencies in previous cell designs and

production. This section summarizes the advantages and disadvantages inherent to this

17

particular design.

Advantages

DSSC cell designs offer the following advantages [62, 63]:

• The photoanode is fabricated from a nanoparticle material. As a result, the cell can

absorb almost all photons in the incident solar beam. The dye absorber within the

nanoparticles efficiently converts photons to electrons.

• DSSCs can more easily and efficiently absorb diffuse sunlight and fluorescent light,

in addition to direct sunlight. This allows operation in overcast conditions or even

indoors to drive small devices.

• Heat is more easily radiated from the thin layers of the photoanode and counter

electrode, allowing greater operating performance and efficient reduction in internal

temperatures during operation.

• Individual components in a DSSC generally do not need frequent replacement.

• They are more mechanically robust, allowing easier handling.

• They are more economical to fabricate than solar cells based on bulk semiconductor

materials.

Disadvantages

The following issues currently limit use of DSSC technologies [64]. For DSSC’s using

liquid electrolyte solutions, the issues are related to the temperature sensitivity of the

electrolyte solution:

• At low temperatures, the electrolyte can freeze, thus rendering the DSSC completely

inoperable.

• At higher temperatures, the liquid electrolyte expands, resulting in issues with sealing

18

the photoanode and counter electrode. There may also be issues with respect to stability

of the operating electrochemical potential, and a risk of partial or complete evaporation

of the electrolyte solution, which will also result in the DSSC becoming inoperable. To

avoid these issues, the maximum safe operating temperature must be limited.

DSSCs using aqueous solid and/or solid electrolytes are more resistant to the effects of

electrolyte temperature. Their primary disadvantage is the decrease in operating efficiency

with respect to liquid electrolyte DSSCs.

Previous Work

This section summarizes previous work with respect to DSSC design and fabrication,

emphasizing research into liquid electrolyte-based dyes and additives used for the I-/I-3

redox couple. The maximum reported photoconversion efficiency resulting from the

reported research will also be given.

Dye-Sensitized Solar Cell Liquid Electrolytes

O’Regan’s and Gratzel’s original DSSC design [14] used a rudimentary liquid

electrolyte consisting of I-/I-3 redox couple dissolved in an organic solvent, without extra

additives. The reported photoconversion efficiency for this design was between

approximately 7.1% and 7.9%.

In 2000, Oskam et al. [65] reported on their efforts to prepare and characterize two

pseudohalogen redox couples for dye-sensitized TiO2 photoelectrochemical cells, and

investigated the effect of the redox potential on Voc. The equilibrium potentials of the

(SeCN)2/SeCN- and (SCN)2/SCN- couples were 0.19 and 0.43 V, respectively, more

positive than for the I-/I-3 couple. With the N3 sensitizer (cis-Ru(dcb)2(NCS)2), the incident

photon-to-current conversion efficiency decreased by a factor of 4 for (SeCN)2/SeCN- and

19

a factor of 20 for (SCN)2/SCN- compared with I-/I-3.

In 2003, Nusbaumer, et al [66] synthesized a group of cobalt (Co) complexes and

evaluated their suitability as redox mediators. The one‐electron‐transfer redox mediator

[Co(dbbip)2] (ClO4)2 (dbbip = 2,6‐bis(1′‐butylbenzimidazol‐2′‐yl) pyridine) was found to

achieve the best performance among the compounds investigated. Photovoltaic cells

incorporating this redox mediator reported a maximum photoconversion efficiency of

approximately 8%, and incident photon-to-current conversion efficiency of up to 80%.

However, the cells demonstrated poor dye regeneration kinetics.

In 2004, Gratzel, et al [67], synthesized and studied a solvent-free ionic liquid

electrolyte based on a SeCN−/(SeCN)-3 redox couple as a mediator. DSSCs using this redox

couple mediator demonstrated maximum photoconversion efficiencies between

approximately 7.5% and 8.3% at illumination levels of 1.5 times full sunlight, as good as

or better than the standard I-/I-3 redox couple.

In 2005, Wang, et al [68] characterized the performance of halide-based redox couples.

Under illumination levels corresponding to 1.5 times normal sunlight, DSSCs using 0.4 M

LiBr + 0.04 M Br2 electrolyte in acetonitrile yielded a 2.61% photoconversion efficiency.

DSSCs using 0.4 M LiI + 0.04 M I2 in acetonitrile as the electrolyte solution and tested

under the same conditions produced a photoconversion efficiency of 1.67%. Replacement

of I-/I3- with Br-/Br-

3 in eosin Y yielded a significant increase in Voc that was offset by slight

decreases in Jsc and FF, leading to an increase in photoconversion efficiency by 56%.

In 2005, Yanagida, et al [69], synthesized and investigated a new series of Cu

complexes for their effectiveness as redox mediators, and compared their photochemical

responses to the standard I-/I-3 couple. Under standard solar illumination levels

20

(approximately 100mW/cm2), DSSCs using the mediators [Cu(phen)2]2+/+,

[Cu(dmp)2]2+/+, and [Cu(SP)(mmt)]0/- exhibited photoconversion efficiencies of

approximately 0.1%, 1.4%, and 1.3%, respectively. A maximum photoconversion

efficiency of approximately 2.2% was obtained for a DSSC using the [Cu(dmp)2]2+/+

redox couple at a reduced illumination level of 20 mW/cm2 intensity, with a higher Voc

than achieved with the conventional I-/I-3 couple.

In 2007, Gorlov, et al [70] synthesized and investigated new halide/halide redox

couples such as I/IBr2 and I/I2Br for use as ionic liquid electrolytes, in combination with

Ru-based photosensitizing dyes. DSSCs using the new halide/halide couples achieved

maximum photoconversion efficiencies of approximately 6.4%. However, their

characterization was more difficult due to the complex equilibrium conditions between

various halide ions.

In 2008, Zhang, et al [71] synthesized an organic redox couple based on the 2,2,6,6-

tetramethyl-1-piperidinyloxy (TEMPO) radical. The photovoltaic performance of this new

redox couple was evaluated by employing nanocrystalline TiO2 films of varying thickness.

They estimated an overall photoconversion efficiency for such devices of approximately

5.4% under illumination levels 1.5 times normal sunlight.

In 2010, Whang, et al [72] synthesized a disulfide/thiolate redox couple with little

absorption in the visible spectrum, a very attractive feature for flexible DSSCs using

transparent conductors as current collectors. Using this new iodide-free redox electrolyte

with a sensitized heterojunction, they achieved a maximum photoconversion efficiency of

approximately 6.4% under standard illumination conditions. This redox couple may

provide a viable pathway for developing efficient DSSCs that can be scaled for practical

21

applications.

In 2010, Tina, et al [73] synthesized a nickel Ni(III)/(IV) bis(dicarbollide) redox couple

and investigated its potential usability with varying electrolyte concentrations and

additives. The Ni (III)/(IV) redox couple was identified as a promising, non-corrosive

redox couple with good solubility and redox properties. DSSCs using a mixture of reduced

nickel III (0.030 M) and oxidized nickel IV (1.8 × 10-3 M) in acetonitrile, exhibited

maximum photoconversion efficiencies of approximately 0.9% with tert-butylpyridine and

TMABF4 as solution additives.

In 2011, Daeneke et al [74] reported maximum photoconversion efficiencies of

approximately 7.5% at illumination levels 1.5 times normal sunlight in DSSCs using the

ferrocene/ferrocenium (Fc/Fc+) single-electron redox couple combined with the metal-free

organic donor-acceptor sensitizer Carbz-PAHTDTT. Under comparable conditions, these

Fc/Fc+-based devices exceeded the efficiency of devices using the standard I-/I-3

electrolytes, suggesting their potential as a replacement electrolyte in future DSSC

applications. This improvement was thought to result from a more favorable matching of

the redox potential of the ferrocene couple to that of the new donor–acceptor sensitizer.

In 2012, Kato, et al [75] synthesized an organic redox mediator based on the 2-

azaadamantan-N-oxyl (AZA) radical. This mediator exhibited appropriate redox potential

and high diffusivity, electron transfer and electron self-exchange reaction rates, which

resulted in decreased cell resistance. Consequently, DSSCs using this mediator achieved

maximum photoconversion efficiencies of approximately 8.6%, due to a significantly

higher FF (> 0.75) with the AZA mediator, which also exhibited more effective charge

propagation and regeneration.

22

Iodide/Triiodide Electrolyte Additives

In 1993, Gratzel, et al [76] studied 4-tert-butylpyridine (4TBP) as a potential additive

to the standard I-/I-3 electrolyte. This additive causes a negative or upward shift in the TiO2

conduction band [77], resulting in a higher Voc that enhances the DSSC’s overall

performance.

In 2000, Park, et al [78] investigated the photovoltaic performance of dye-sensitized

nanocrystalline rutile TiO2 DSSCs using additives containing Li+ and 1,2-dimethyl-3-

hexyl imidazolium ions in the standard I-/I-3 electrolyte. The electrolyte containing Li+ ions

produced a lower Voc and higher short-circuit photocurrent Isc than the electrolyte

containing 1,2-dimethyl-3-hexyl imidazolium ions. The adsorption of Li+ ions to the

nanocrystalline film was shown to have a significant effect on the band-edge position,

which, in turn affects Voc, Isc, and IPCE. This study indicated that for future development

of high efficiency DSSCs, the nature of the cation must be a major consideration when

optimizing the electrolyte composition.

In 2009, Zhang, et al [79] investigated the influence of guanidinium thiocyanate

(GuSCN) on the photoelectron injection efficiency and interface dynamics of electron

transfer. Analysis of the short circuit current density Jsc vs light intensity suggested that

higher electron injection efficiency could be achieved with GuSCN as the electrolyte

additive. Increasing the GuSCN concentration from 0.1 to 0.4 M had no significant effect

on electron injection efficiency. Based on analysis of impedance spectra of the solar cells,

they concluded that addition of GuSCN to the electrolyte could suppress surface

recombination effects on the TiO2 electrode, resulting in a shift of the conduction band

towards positive potentials.

23

In 2010, Yu, et al [80] investigated the effects of GuSCN and N-methylbenzimidazole

(NMPI) additives on DSSC photovoltaic performance using ionic-liquid and organic-

solvent based electrolytes. They observed only a slight improvement in Voc and a

significant enhancement in Isc with GuSCN as the only electrolyte. They also observed a

significant improvement in Voc with NMPI as the only electrolyte. Synergistic

enhancements of Voc and Isc were observed when both GuSCN and NMPI were added to

the ionic-liquid electrolyte; these effects resulted in optimal photovoltaic performance.

In 2010, Kim, et al [81] investigated the potential of thiourea as an additive to the

standard I-/I-3 redox electrolyte. They observed simultaneous positive conduction band

edge shifts and decreases in the charge recombination rate. For a thiourea concentration of

0.05M, adding 0.7M of 1-methyl-3-propylimidazolium iodide (MPII) and 0.05 M I2 in

acetonitrile significantly enhanced Jsc (from approximately 7.7 mA/cm2 to 10.8 mA/cm2)

while reducing Voc (from approximately 0.78 V to 0.71 V). The resulting overall

photoconversion efficiency increased approximately 23%, from 4.7% to 5.8%.

In 2012, Kim, et al [82] studied replacement of GuSCN with ordinary urea as a long-

term stabilizer additive. They found the urea-based additive to be less reactive than

GuSCN, resulting in greater long-term thermal and electrical stability.

Motivations

This work investigates the use of novel electrolyte additives to DSSCs using the

standard I-/I-3 redox electrolyte. DSSC performance using varying additive concentrations

is characterized, with the goal of further optimizing photovoltaic characteristics and overall

photoconversion efficiency.

24

Objective

Improve the DSSC efficiency using commercially available liquid I-/I-3 redox

electrolyte by more than 1%, from approximately 8.5% to more than 9.5%, using urea

(CH4N2O) as the sole electrolyte additive. In addition, achieve similar improvement in

DSSC efficiency using reduced graphene oxide (rGO) as the sole electrolyte additive.

The implementation plan for this work consists of the following tasks:

1. Fabricate three DSSC devices with urea doped electrolyte for each concentration:

0, 2, 5, 10, and 15 wt%.

2. Fabricate three DSSC devices with rGO doped electrolyte for each concentration:

0, 2, 5, 10, and 15 wt%.

3. Measure Current density-voltage (J-V), external quantum efficiency (EQE),

electrochemical impedance spectroscopy (EIS), cyclic voltammogram (CVs), and

UV-Vis absorption spectra for all fabricated devices.

25

Chapter 2: Theory

Solar Radiation and Air Mass

The solar spectrum typically covers wavelengths corresponding to the ultraviolet and

visible / near infrared electromagnetic spectrum (0.2 µm to 3.0 µm). The intensity is non-

uniform, sharply increasing through the ultraviolet region to a maximum in the visible

region, then decreasing more slowly into the infrared region (Figure 2.1). The solar

spectrum at the top of the atmosphere is approximated with a Planck (blackbody) radiation

curve corresponding to a temperature of approximately 5250° C, as shown in Figure 2.1.

Figure 2.1 Solar Radiation Spectrum [83]

Solar spectra measured at the top-of-atmosphere (TOA) are not the same as spectra

measured at the surface, due to atmospheric scattering and absorption effects. The path

length of solar energy in the atmosphere depends on the solar zenith angle (Figure 2.2),

26

which varies with the time of day. This path length is given by the air mass number (AM),

which is the secant of the solar zenith angle θ:

sec 1cos

(2.1)

Figure 2.2 Solar Zenith Angle. Modified from ref [84]

With respect to solar cell technology, solar spectra are considered at three different air

mass values: AM0, AM1.5global, and AM1.5direct. The different spectra are plotted in Figure

2.3. With the sun directly overhead 0°, radiation striking the ground passes through

all of the atmosphere. This direct radiation is known as "Air Mass 1 Direct" (AM1direct)

radiation. Similarly, the global radiation at 0° is known as "Air Mass 1 Global"

(AM1global) radiation. Because it passes through no air mass at the top of the atmosphere,

the extraterrestrial spectrum is called the "Air Mass 0" spectrum or AM0 [85]. AM1.5direct

is the spectrum related to the air mass corresponding to θ=48.2ο due to rays directly

transmitted to the surface that cast a hard shadow. Similarly, AM1.5global is the spectrum

related to the air mass corresponding to θ=48.2ο due to directly transmitted rays and rays

27

scattered by the atmosphere to the surface. Air mass estimates corresponding to a specific

solar zenith angle value can still vary by season, location, local weather conditions, etc.

Figure 2.3 Solar Reference Spectra [86]

Solar Cells and Photovoltaic Effect

The working principle of solar cells is based on the photovoltaic effect, i.e. the

generation of an electrical potential difference at the junction of two different materials in

response to electromagnetic radiation. The photovoltaic effect is closely related to the

photoelectric effect, in which electrons are emitted from a material absorbing light at a

frequency higher than a threshold frequency dependent on the material. In 1905, Albert

Einstein hypothesized this effect could be understood by assuming that light consists of

well-defined energy quanta called photons. The energy of such a photon is given by

(2.2)

28

where h is Planck’s constant and ν is the frequency of the light. For his explanation of the

photoelectric effect, Einstein received the Nobel Prize in Physics in 1921 [87].

The photovoltaic effect can be divided into three basic processes:

i. Charge Carrier Generation Due to Photon Absorption in Junction Materials

When a photon with energy is absorbed by bulk semiconductor material, its energy

excites an electron from an initial energy level Ei to a higher energy level Ef, and creates a

void in the valence band behaving as a positively charged particle known as a “hole”, as

shown in Fig. 2.4. Photons can only be absorbed if the difference between Ei and Ef equals

the incident photon energy . In an ideal semiconductor, electrons can populate energy

levels less than the valence band edge, EV, and greater than the conduction band edge, EC.

Between these two energy levels is the bandgap, where no energy levels exist in which

electrons can populate:

Eg = EC − EV (2.3)

Incident photons with energy hν < Eg traverse the material unabsorbed. Photons with

incident energy hν > Eg emit an electron with an excited energy state above the minimum

conduction band energy; the electron then emits the energy difference (hν - Eg) as thermal

energy (heat).

29

Figure 2.4 Photon Absorption in a Semiconductor with Bandgap Energy Eg

In a real semiconductor, the valence and conduction bands are not flat as shown in

Figure 2.4, but vary depending on the k-vector describing the crystal momentum of the

semiconductor, as shown in Figure 2.5. If the valence band maximum and conduction band

minimum energies share the same k-vector, an electron can be excited from the valence to

the conduction band without a change in crystal momentum. Such a semiconductor is

called a direct bandgap material. If the electron cannot be excited without changing the

crystal momentum, the semiconductor is called an indirect bandgap material. The

absorption coefficient in a direct bandgap material is much higher than in an indirect

bandgap material, thus the bulk material can be much thinner [88].

For any semiconducting material, if an electron is excited from Ei to Ef, an electron-

hole pair is created. The radiative energy of the photon is converted to the electrical energy

of the electron-hole pair. The maximum conversion efficiency from radiative energy to

chemical energy is limited by thermodynamics. For sunlight, the thermodynamic limit

ranges from approximately 67% for non-concentrated sunlight to approximately 86% for

fully concentrated sunlight [89].

30

Figure 2.5 Direct and Indirect Bandgap Semiconductor Characteristics

ii. Separation of Photo-Generated Charge Carriers in the Junction

Usually, the electron-hole pair recombines shortly after its generation, i.e. the electron

falls back to the initial energy level Ei. The energy is released either as a photon, transferred

to other nearby electrons or holes, or released as thermal energy through lattice vibrations

in the semiconductor material. To use the energy stored in the electron-hole pair for

performing work in an external circuit, a semipermeable membrane must be present on

both sides of the junction such that electrons leave through one membrane and holes leave

through the other [89]. In most solar cells, these membranes are formed by doped n- and

p-type materials. A solar cell must be designed such that the electrons and holes can reach

the membranes before they recombine, i.e. the time required for charge carriers to reach

the membranes must be shorter than the recombination lifetime. This requirement limits

the thickness of the junction region.

iii. Accumulation of Photo-Generated Charge Carriers at the Junction Terminals

Finally, the charge carriers are extracted from solar cells with electrical contacts so that

they can perform work in an external circuit. The chemical energy of the electron-hole

31

pairs is finally converted to electrical energy. After the electrons pass through the circuit,

they will recombine with holes at a metal absorber interface.

Classical P-N Junction Solar Cell

P-N junctions are formed by joining n-type and p-type semiconductor materials, as

shown in Figure 2.6. Since the n-type region has a high electron concentration and the p-

type region a high hole concentration, electrons diffuse from the n-type side to the p-type

side. Similarly, holes diffuse from the p-type side to the n-type side. If the electrons and

holes were not charged, this diffusion process would continue until the concentrations of

electrons and holes on both sides were equal. However, in a P-N junction, electron/hole

diffusion creates a diffusion current, leaving behind exposed charges on dopant atom sites,

which are fixed in the crystal lattice and unable to move. In the n-type material, positive

ion cores are exposed, and in the p-type material, negative ion cores are exposed. An

electric field Ê forms between the positive and negative ion cores, resulting in a carrier

drift current in the opposite direction of the diffusion current. Carrier diffusion continues

until the drift current balances the diffusion current, thereby reaching thermal equilibrium

as indicated by a constant Fermi energy (Figure 2.6). The region where the electric field

occurs is called the "depletion region" since the field quickly sweeps free carriers out.

32

Figure 2.6 Energy Band Diagram of a p-n Junction in Thermal Equilibrium, with Depletion Region. Modified from ref [90]

DSSC Operating Principles

Nanocrystalline TiO2 is deposited on the conducting electrode (photoelectrode) of a

DSSC to provide a sufficiently large surface area to adsorb sensitizers (dye molecules).

Upon photon absorption, dye molecules are excited from the highest occupied molecular

orbital (HOMO) to the lowest unoccupied molecular orbital (LUMO) states as shown

schematically in Figure 2.7. This excitation process is represented as:

→ ∗ (2.4)

Once an electron is injected into the conduction band of a wide bandgap semiconductor

nanostructured TiO2 film, oxidation of the dye molecule (photosensitizer) occurs:

∗ !" → #$%&'( ) (2.5)

33

The injected electron is transported between the TiO2 nanoparticles and then extracted to a

load in the counter electrode (cathode) where the work done is delivered as electrical

energy:

#$%&'( *. . → !" #+.,.( #-#./ .0-##/12 (2.6)

Electrolytes containing, I-/I-3 redox ions are used as an electron mediator between the TiO2

photoelectrode and the carbon coated counter electrode. Therefore, the oxidized dye

molecules (photosensitizer) are regenerated by receiving electrons from the I- ion redox

mediator as it oxidizes to I-3:

) 325( → 12 56(

(2.7)

The I-3 ion substitutes the internally donated electron with that from the external load and

reduces back to an I- ion:

12 56( #+.,.( → 3

25( *. . (2.8)

Electron movement in the conduction band is accompanied by the diffusion of charge-

compensating cations in the electrolyte layer close to the nanoparticle surface. Therefore,

a DSSC can generate electric power without causing any permanent chemical change or

transformation (Gratzel, 2005).

As illustrated in Figure 2.7, the maximum potential produced by the cell is determined

by the energy separation between the electrolyte redox potential (Eredox) and the Fermi level

EFermi of the TiO2 layer. The small energy separation between the HOMO and LUMO

ensures absorption of low energy photons in the solar spectrum. Therefore, the

photocurrent level is dependent on the HOMO-LUMO energy difference of the dye,

analogous to the bandgap energy Eg in inorganic semiconductor materials. Electron

34

injection into the TiO2 conduction band of is enhanced with the increase of energy

separation of the dye LUMO and the TiO2 conduction band minimum. Furthermore, for

the HOMO level to effectively accept donated electrons from the redox mediator, the

energy difference between the HOMO and redox chemical potential must be more positive

(Hara & Arakawa, 2003) [91].

Figure 2.7 Schematic of DSSC with Energy Level Diagram. Modified from ref [14, 92]

Key Efficiency Parameters Representing DSSC Performance

The following sections consider the four main parameters used to characterize the

performance of DSSCs. These parameters–the peak power Pmax, the short-circuit current

density Jsc, the open-circuit voltage Voc, and the fill factor FF–are determined from the

illuminated J-V characteristic curve, as shown in Figure 2.8. The conversion efficiency η

can be determined from these parameters as shown in Section 2.5.4.

35

Short-Circuit Current Density

The short-circuit current Isc is the current flowing through an external circuit when the

solar cell electrodes are short-circuited. It depends on the cell’s area and the incident photon

flux density determined by the incident light intensity, standardized to the AM 1.5

spectrum.

To remove the dependence of the solar cell area on Isc, the short-circuit current density

Jsc is used to describe the maximum current. The maximum current in general depends on

the cell’s optical properties, such as its reflectivity and the absorber layer’s absorption

characteristics. In a DSSC, Jsc also depends on the intensity of incident light and absorption

spectrum of the dye:

789 # : ;<= >?@AB

(2.9)

where Nph is the photon flux of incident light per unit wavelength (counts/[m2-s-nm]), Eg

is the material bandgap in electron volts, e is electronic charge in Coulombs, E is the photon

energy in electron volts, and EQE is the external quantum efficiency.

The net current density, J(V), within the cell is given as:

JV JEFGHV I J<= (2.10)

where Jph is the photocurrent density, and the dark current density Jdark, is

JEFGHV JJ Ke LMNOPQ I 1R (2.11)

where KB is Boltzmann`s constant, J0 is the saturation current density, e is the electronic

charge, and T is the temperature in degrees Kelvin. Substituting Equation 2.11 into

equation 2.10 allows J(V) to be expressed as follows:

36

JV JJ Ke LMSOPQ I 1R I JTU (2.12)

When the load resistance is zero, corresponding to V=0, J(V) is just Jsc:

J J89 (2.13)

Open-Circuit Voltage

The open-circuit voltage Voc is the voltage at which no current flows through an

external circuit (7 = 0). It is the maximum voltage that a DSSC can deliver. Voc corresponds

to the forward bias voltage of the cell’s internal P-N junction such that the dark current

density balances the photocurrent density. Voc can be calculated by solving equation 2.12

for V and assuming a zero net current density:

VW9 XY# ln JTU7J 1 \ XY

# ln JTU7J (2.14)

where the approximation is justified because JTU >> 7J.

Equation 2.14 shows that Voc depends on the ratio of the photocurrent and saturation

current densities in the DSSC. While Jph typically exhibits minimal variation, J0 can vary

by several orders of magnitude, as it depends primarily on recombination effects.

Therefore, Voc is a measure of the amount of recombination in the device. Laboratory

crystalline silicon solar cells have a Voc of up to 720 mV under standard AM1.5

illumination, while commercial DSSCs under similar illumination typically have Voc

exceeding 600 mV [88].

Fill Factor

The fill factor (FF) is a measure of the maximum power output from a DSSC and is

defined as the ratio of the maximum power to the product of Voc and Jsc. FF is calculated

as follows:

37

F^ _789VW9 7_V_789VW9 (2.15)

where Jm is the maximum current density (A/cm2) and Vm is the maximum voltage (V). It

can be approximated by rectangles in a J-V curve (Figure 2.8). In the ideal case, FF is equal

to 1.

The fill factor is a function of the DSSC’s series and shunt resistances, and reflects the

extent of electrical and electrochemical losses during cell operation. Higher FF values can

be obtained through i) increasing the shunt resistance; ii) decreasing the series resistance;

and/or iii) reducing the “overvoltage” [88] related to diffusion and charge transfer.

Figure 2.8 J-V Characteristic Curves Under Dark and Illuminated Conditions

Conversion Efficiency

The light-to-electricity photoconversion efficiency (η) is the total amount of electrical

38

power produced for a given amount of solar energy incident on the cell. It is calculated as

the ratio of the maximum generated power and a reference incident power Pin of 100

mW/cm2 assuming an AM1.5 spectrum. Additional information relating to the conversion

efficiency is presented in Section 2.7.1.

Equivalent Circuit Model for Conventional PN-Junction Solar Cells

The electrical properties of conventional P-N junction solar cells can be characterized

using a simple equivalent DC circuit similar to Figure 2.9. In this equivalent circuit, a

resistance (Rs) is in series with a diode representing the electrical properties of the junction;

a shunt resistance (Rsh) and a constant-current source (Iph) are in parallel with the junction.

Series resistance occurs along the path the current takes through the cell; as it increases,

Voc decreases, which in turn reduces FF (Figure 2.10). Losses occur at the contacts and the

interfaces. No other parameters are affected until this resistance is large enough to reduce

Jsc. The shunt resistance is due to defects introduced during cell fabrication that reduce the

available photocurrent by allowing it to flow along multiple paths. As the shunt resistance

decreases, Jsc is reduced, which in turn reduces FF. The source current Iph results from the

excitation of excess carriers by incident light.

Figure 2.9 Conventional P-N Junction Solar Cell Equivalent circuit model [93]

39

Figure 2.10 Changes in I-V Response with Changes in Rsh and Rs

Characterization Techniques

This section describes the measurements used in this work to characterize DSSC cell

performance. These measurements include the current density-voltage (J-V) characteristic,

the incident photon-to-current-efficiency (IPCE) spectrum, the electro-chemical

impedance spectrum (EIS), cyclical voltammetry (CV), and the UV-visible absorption

spectrum.

Current Density-Voltage Characteristics (J-V)

Current density-voltage (J-V) measurement is the most important and conventional

technique for assessing solar cell photovoltaic performance. A standard illumination of AM

1.5 with an irradiance of 100 mW/cm2 is normally used for characterization. The J-V

characteristics are determined under constant illumination, by increasing an external load

from zero (short-circuit conditions) to infinite (open-circuit conditions). There is a point

on the knee of the J-V curve where maximum power is produced, the Maximum Power

40

Point (MPP) [94]. The voltage and current density at the Maximum Power Point are

designated as Vm and Jm (Figure 2.10).

J-V characteristic curve measurements provide the key parameters needed to estimate

a DSSC’s light-to-electricity photoconversion efficiency (η), defined as the ratio of the

product of Jsc, Voc, and FF to the reference intensity Pin under AM1.5 illumination:

η JabVcbFFPeN 100

(2.16)

Current density-voltage measurements should be carried out with a sufficiently slow

voltage scanning rate such that the cell response is free of hysteresis effects.

External Quantum Efficiency (EQE)

The external quantum efficiency (EQE) is another key parameter used in quantitative

characterization of DSSC performance. The EQE describes the efficiency of converting

incident energy to electron-hole pairs as a function of wavelength. More specifically, the

EQE is the ratio of the number of electrons generated by the cell and collected at an external

circuit to the number of incident photons with a given wavelength striking the cell’s

photoanode:

> /0#fgh#/i.--#.#?#-#./j/0#fgh#/i . ?#kj

(2.17)

Alternatively, the EQE can be calculated as a scaled ratio of the short-circuit current density

to an input irradiance of 100 mW/cm2 assuming AM1.5 illumination. The scaling factor

varies inversely with wavelength:

> . 789# l %S 1240 789l %S (2.18)

where h is Planck’s constant (6.626 ×10-34 J.s), c is the speed of light in vacuum (2.998

×108 m s-1), and e is the electronic charge.

41

Electrochemical Impedance Spectroscopy (EIS)

According to Ohm’s law, a DC current I flowing through a resistor R produces a voltage

drop V across the resistor terminals:

V 5n (2.19)

For a sinusoidal varying current at an angular frequency ω flowing through its terminals,

the resistor has a characteristic impedance consisting of its DC resistance R and a reactance

X that is a function of ω; X represents the resistor’s opposition to changes in applied current

or voltage due to its internal inductance and/or capacitance. The impedance of the resistor

is represented by a complex number, with the resistance as the “real” component and the

reactance as the “imaginary” component [95]:

o n pq (2.20)

By applying a harmonically modulated small amplitude AC potential V(ω, t) at a given

frequency i r"s to the resistor and measuring the resulting current I(ω, t), the resistor’s

impedance can be expressed according to a generalized version of Ohm’s law:

ot Vt, 5t, (2.21)

The same current-voltage relationship exists in DSSCs, with R defined as the total DC

resistance of the cell. However, this value of R represents a “black box” measure of

resistance across the cell terminals; there is no knowledge about resistances associated with

individual cell components, e.g resistance associated with charge transfer at the counter

electrode, resistance associated with electron transport in the TiO2 mesoporous layer, or

resistances due to recombination and diffusion in the electrolyte. Measuring the cell’s

electrochemical impedance spectrum (EIS) provides information on these internal

42

resistances.

If the applied voltage is time-independent (ω=0), DC current flows through the cell,

and the resulting impedance is just its DC resistance (RDC):

o0 V0, 50, nvw (2.22)

Using complex number notation, a small-amplitude sinusoidal varying AC voltage and

current can be described as: Vt, VJexppt and 5t, 5Jexppt I

respectively, where j=√ (−1) and V0 and I0 are the amplitudes of the voltage and current

signals, respectively. Thus, equation (2.21) can be written as:

opt Vt, 5t, VJ5J #| (2.23)

Using the magnitude Z0, equation (2.23) can be written as:

opt oJ#| (2.24)

By applying Euler’s relationship and replacing Z0 with |o|, equation (2.24) can be written

as:

opt |o|cos p sin (2.25)

The general expression for impedance is:

ot o po%_ (2.26)

where Zre = |Z| cos(θ) and Zim = |Z| sin(θ).

EIS data are usually displayed in the form of a Nyquist plot. In the complex plane, Zim

is plotted against Zre for different values of ω. The result is three semi-circular arcs

representing the low-frequency, mid-frequency, and high-frequency resistance

characteristics of the cell.

43

Figure 2.11 shows a standard Nyquist plot of a DSSC with equivalent circuit model. nj

is the series resistance of the DSSC, n.t is the charge transfer resistance at the counter

electrode, n/ is the recombination resistance of electrons in the TiO2 conduction band to

the redox couple in the electrolyte, and n is the diffusion resistance for the redox couple

in the electrolyte.

DSSCs can be modeled as an equivalent circuit of passive elements as shown in Figure

2.11. CPE is a constant phase element replacing the double layer capacitance Cdl on the

interface between an electrode and the electrolyte [96]. This double layer is formed as ions

from the solution adsorb onto the electrode surface. The charged electrode is separated

from the charged ions by an insulating space on the order of a few angstroms in thickness.

Figure 2.11 Example DSSC Nyquist Plot with Equivalent Circuit Model. Modified from ref [97, 98].

Cyclic Voltammogram Measurements (CVs)

Cyclic voltammetry (CV) is a powerful electrochemical measurement technique

44

commonly employed to characterize reduction and oxidation processes in various

molecular compounds. CV is widely used in DSSC research for characterization of dyes,

redox mediators and charge transfer processes at interfaces. In this technique, a cyclic

potential (Figure 2.12(a)) is applied to the CE immersed in a stable solution of the redox

couple electrolyte under study. The redox analyte’s oxidation/reduction reaction is

represented as:

! #( n#? (2.27)

where Ox and Red are the oxidized and reduced species (redox couples) in solution.

Figure 2.12(b) shows the corresponding response to the cyclic potential. When the

potential increases from an initial value (Ei), a small, transient double-layered capacitive

current appears. As the potential increases linearly with time beyond Eo (a threshold

voltage), the Fermi-level of the electrode is raised above the energy level of vacant orbitals

in the Ox species. The electrons in the electrode cross the electrode-solution interface and

reduce Ox to Red in the solution via reaction (2.27); as a result, a Faradaic current flows

in the electrode. A further increase in the negative potential accelerates the reduction and

depletion of Ox at the electrode/electrolyte interface, creating a significant concentration

gradient between the electrode/electrolyte interface and bulk solution. This concentration

gradient enhances fast diffusion of Ox from the bulk solution towards the interface, causing

the Faradaic current to reach a maximum called the cathodic peak current (Ipc) at the applied

voltage Epa. Even though the applied voltage increases beyond Epa, the cathodic current

decreases continuously from this point because the concentration gradient decreases with

time. As the cyclic potential reaches a peak value of EV, the potential begins decreasing

with time. When the potential approaches E0, the Fermi-level in the electrode is lowered;

45

electron transfer from Red into the electrode becomes more favorable, and Red is oxidized

back to Ox.

Figure 2.12 Cyclic Potential Sweep and (b) Shape of a Cyclic Voltammogram Curve. Modified from ref [14, 92]

Figure 2.13 shows that an electrolyte analyte solution will exchange an electron with

an electrode when it reaches certain threshold potentials. The electrolyte undergoes

reduction when the electron potential energy in the electrode is higher than the potential

energy of the electrolyte’s empty molecular orbitals. Conversely, the electrolyte undergoes

oxidation when its highest energy electron is higher than the electron potential energy in

the electrode. In both cases, electrons are transferred in order to minimize the potential

energy of the total system. A potentiostat can be used to change the electrode potential and

observe the direction of current flow in order to learn about the electrolyte’s electron

transfer energetics.

46

Figure 2.13 Energetics Between Electrolyte Solution and Electrode. Modified from

ref [99]

UV-Visible Absorption Spectra

UV-visible absorption spectroscopy is a technique used to measure electronic

transitions within a material when it is illuminated. The absorption of UV-visible light is

related to outer electron excitations. The material absorbs light with a wavelength matching

an energy level difference corresponding to the allowed transition of an σ-electron, n-

electron, or π-electron from its ground state to a higher anti-bonding molecular orbital level

(Figure 2.14).

47

Figure 2.14 Possible Transitions of π, σ, and n Electrons. Modified from ref [100]

UV-Vis spectroscopy analyzes the absorption characteristics of a sample of dye and/or

electrolyte solution across the ultraviolet, visible and near infrared regions of the

electromagnetic spectrum. Figure 2.15 shows a representative UV-Vis spectrometry

measurement setup. The system uses a deuterium arc lamp and tungsten lamp as light

sources generating wavelengths in the range of 190 – 1100 nm. The monochromator

consists of an entrance slit, a dispersion device and an exit slit, which produces a

monochromatic beam at a given wavelength. When illuminated by the monochromatic

beam, electrons in the sample are excited to an energy equal to the difference between

allowed transitions within the sample molecules (e.g., between the bonding and

antibonding orbital levels). The principle behind UV-Vis absorption spectroscopy is the

Beers-Lambert law, which is represented as:

l 55ₒ #( (2.28)

where I0 is the incident light intensity, d is the sample thickness, I is the transmitted light

intensity and α(λ) is the absorption coefficient. The total absorbance of the material, A(λ),

48

is given by:

Aλ I-1JTλ (2.29)

Substituting eq. 2.28 for the transmittance into eq. 2.29, the absorption coefficient of the

sample can be determined:

α 1d lnT (2.30)

Figure 2.15 UV-Vis Spectroscopy [101]

Mechanism of Urea Reaction with I-/I-3 Redox Electrolyte

Adding urea to a I-/I-3 redox electrolyte solution initiates a reaction with neutral iodine

to produce oxalyl dihydrazide, H+ ions, and iodide ions [81, 102, 103]:

2*;"! 5" →*";!" 2) 25( (2.31)

Generation of the H+ ions results in a positive shift in the electrolyte’s redox potential

(Figure 2.16), which in turn increases Voc as it is inversely proportional to the tri-iodide

concentration [104, 105]:

49

V&w # ln ∅%SJX56(

(2.32)

where k is Boltzmann’s constant, e is the electron charge in Coulombs, T is the absolute

temperature in degrees Kelvin, Φinj is the charge flux resulting from electron injection by

the dye, n0 is the electron concentration at the TiO2 surface when the cell is not illuminated,

and Ket is the I-3 reduction rate by conduction band electrons.

Figure 2.16 Mechanism of Urea Reaction with I-/I-3 Redox Electrolyte

Mechanism of rGO Reaction with I-/I-3 Redox Electrolyte

Reduced graphene oxide (rGO) is a conductive material that is insoluble in an I-/I-3

redox electrolyte solution. I-3 ions are large and heavy ions, with low mobility and a

tendency to concentrate near the photoanode (Figure 2.17). I-3 ions are electron acceptors,

promoting recombination reactions between the TiO2 conduction band and the electrolyte,

which reduces FF and Voc.

50

Adding rGO to the electrolyte cause intermolecular interactions between I-3 ions and

rGO due to the electrostatic interactions. In the crystal structure, I-3 was arranged as rigid

rods and arrays in parallel to rGO molecules [106, 107]. This positive interaction allows

the removal of I-3 ions far from the photoelectrode interfaces, recovering the Voc and FF

values. Therefore, adding rGO to the electrolyte increases Voc and accelerates the charge

transfer rate by reducing the electrolyte’s resistance and decreased charge back-

recombination rate.

Figure 2.17 Mechanism of rGO Reaction with I-/I-3 Redox Electrolyte

51

Chapter 3: Experimental Procedures

Material and Device Fabrication Procedures

This section considers the materials and particular procedures used to fabricate the

devices characterized in this thesis work.

Materials

Fluorine-doped tin oxide (FTO) substrates (sheet resistance of 7 Ω / sq and a thickness

of 2.2 mm) were ordered from Hartford Glass (http://www.hartfordglass.com, Hartford

City, IN, USA). Nanocrystalline TiO2 (Ti-Nanoxide HT/SP) (pore size ~15-20 nm),

scattering TiO2 (Ti-Nanoxide R/SP) (particle size ~100 nm), N719 dye (Ruthenizer 535-

bisTBA), iodide/triiodide electrolyte (Iodolyte AN 50), and Meltonix thermoplastic sealant

were purchased from Solaronix (http://www.solaronix.com, Aubonne, Switzerland). Urea

was obtained from Sigma-Aldrich (http://www.sigmaaldrich.com, St. Louis, MO, USA).

Reduced graphene oxide (rGO) (particle size ~ 6-10 µm) was obtained from Graphenea

(http://www.graphenea.com, Cambridge, MA, USA). All materials were used as received

except that TiCl4 was diluted in deionized water (DI) to prepare a 40 mM aqueous solution.

All other materials (i.e. isopropyl alcohol, acetone, acetonitrile, and valeronitrile) were

purchased from Fisher scientific (http://www.fisherscientific.com, Pittsburgh PA, USA).

Device Fabrication – Glass Substrate Preparation

FTO coated glass substrates were cleaned ultrasonically while bathed in solutions of

detergent, de-ionized (DI) H2O, acetone, and 2-propanol. Each bath lasted approximately

25 minutes. Once the baths were completed, the glass substrates were exposed to UV O3

for approximately 20 minutes.

52

Device Fabrication – Photoanode Fabrication

The working electrodes were prepared as follows. A compact TiO2 layer was deposited

onto the conductive side of one of the treated FTO glass substrates by spin coating. The

TiO2 compact layer solution consisted of a 75% ethanol solution containing a 0.3 M

solution of titanium diisopropoxide bis(acetylacetonate).

Once the TiO2 compact layers were applied by spin coating, a two-step annealing

process was performed, first at 125° C for approximately 15 minutes, followed by another

annealing at 470° C for approximately 20 minutes; both annealings were performed under

normal atmospheric conditions. After annealing the compact layer, a 0.16 cm2 x 12 µm

thick layer of nano-crystalline TiO2 was deposited onto the top of the compact layers using

the “doctor blading” technique [108]. The substrates were subjected to the same annealing

process as before, again under normal atmospheric conditions. The TiO2 light scattering

layer was prepared by doctor blading Ti-Nanoxide R/SP (particle size > 100 nm) and

sintering at 470° C for 30 min under normal atmospheric conditions.

After application of the final scattering layer, the photoanodes were soaked for

approximately 30 minutes in a 40 mM aqueous solution of TiCl4 heated to a temperature

of 70°C, rinsed with DI H2O and ethanol, and then dried for approximately 30 seconds in

a compressed nitrogen atmosphere. Finally, the photoanodes were soaked in a 0.3 mM N-

719 dye solution in acetonitrile/valeronitrile (volume ratio 1:1) for approximately 24 hours,

in order to insert dye molecules into the mesoporous TiO2 layer.

Device Fabrication – Counter Electrode (CE) Fabrication

The Pt CE was fabricated onto the conductive side of the other glass substrate by spin

coating an ethanol solution containing a 20 mM solution of hydrogen hexachloroplatinate

53

(IV) hexahydrate (99.9% trace metal basis) at 2000 rpm for approximately 10 seconds. The

substrates were then annealed at 400° C for approximately 15 minutes. A second layer of

Pt was added to the substrate using the same process.

Device Fabrication – Final Assembly

Figure 3.1 shows the final assembly of a DSSC. The dye-sensitized photoanodes were

washed with acetonitrile to remove unanchored dye from the TiO2 film and then dried with

compressed nitrogen gas for approximately 30 seconds. Both the photoanode and CE were

then joined together using Meltonix thermoplastic sealant, cut into a 2.5 cm x 1.5 cm

rectangular shape. A circular hole approximately 9 mm in diameter was punched into the

sealant. Outside the hole, two narrow channels approximately 1 mm in width were made

along the line of the diameter of the hole for injection of the electrolyte. The sealant was

positioned on the FTO surface so that the dye-sensitized TiO2 film was located at the center

of the hole. The glass side of the photoanode was kept on a hot plate at 80° C and gently

pressed with a cotton swab, in order to allow the sealant to bond to the photoanode.

The Pt CE was placed on the sealant, on top of the photoanode. The assembled layers

of the photo-anode, the sealant, and the CE were heated on the hot plate from the glass side

of the CE for less than 2 minutes at approximately 110° C, while pressed together from the

glass side of the photoanode. The assembled cell was then immediately removed from the

hot plate and allowed to cool to room temperature.

54

Figure 3.1 Schematic of DSSC device assembly

Device Fabrication – Electrolyte Preparation

The final step in the fabrication process involved addition of the urea and rGO agents

to the base I-/I3- electrolyte solution. Five batches of urea-doped electrolyte at

concentrations of 0%, 2%, 5%, 10%, and 15% by wt. were prepared. Similarly, five batches

of rGO-doped electrolyte were prepared with the identical concentrations. Three random

samples were then extracted for each concentration. Each sample was injected into a cell

through one of the channels cut into the sealant block. Once the photoanode-counter

electrode gap was filled with the electrolyte solution, the channels were sealed closed with

Adhesive Tech hot glue.

Materials and Device Characterizations

Current Density-Voltage Characteristics (J-V)

Figure 3.2 shows the schematic of the experimental arrangement used to measure the

J-V for DSSCs. A Xenon lamp (Newport Model 67005) with an AM 1.5 filter was used as

the illumination source. The power supply for the lamp (Newport Model 69911) was set to

55

300 W. The lamp was allowed to warm up for approximately 20 minutes before a

calibration measurement of its intensity was performed using an NREL calibrated

Hamamatsu photodetector (S1133). Calibration was achieved when the distance between

the lamp and the photodetector resulted in an equivalent Isc of 753 μA and Voc of 0.65 V.

The DSSCs were positioned at the same distance and orientation as the photodetector.

AM 1.5 illumination was incident on the solar cells through the transparent FTO photo-

anode side. An Agilent 4155C semiconductor parameter analyzer equipped with current

and voltage source meters was used to sweep the voltage levels between 0 V and 1 V at

increments of 10 mV and simultaneously obtain the current measurements. A Microsoft

Windows-based PC running Agilent Desktop software was used to control the

semiconductor parameter analyzer. From the resulting set of (V, J) measurements, Jsc, Voc,

FF, and efficiency were then determined according to the equations given in Section 2.5.

Figure 3.2 Schematic of J-V Measurement Setup

56

External Quantum Efficiency (EQE) Measurements

External quantum efficiency (EQE) was measured using the Newport EQE

measurement kit (Figure 3.3). The same Xenon lamp used in the J-V measurements was

used as the illumination source. The photodetector was again used to determine the correct

focusing distance before measuring the EQE of a DSSC. A PC-controlled monochromator

(Oriel Monochromator 74001) and focusing lenses were used to produce monochromatic

beams with wavelengths between 350 nm and 800 nm, in increments of 5 nm. A reference

voltage Vref was produced by passing the photodetector output through a trans-impedance

amplifier. As with the J-V measurements, the semiconductor parameter analyzer recorded

the measurements of the photodetector / amplifier and DSSC outputs. The photodetector

was replaced by the DSSC under test and its output voltage Vsample at each wavelength was

recorded. The EQE of the tested DSSC was then determined as follows:

EQE8_< V8_< >V (3.1)

Figure 3.3 Schematic of EQE Measurement Setup

57

Electrochemical Impedance Spectroscopy (EIS) Measurements

The catalytic properties of urea and rGO treated electrolyte were compared to those

with electrolyte-only reference DSSC substrates with EIS measurements.

Figure 3.4 shows the functional diagram of the EIS experiment. A PC-controlled

Ametek VERSASTAT3-200 Potentiostat with a frequency analysis module (FDA) was

used for the EIS measurement. The EIS was carried out for each DSSC as follows. A

reference probe and the CE probe from the Potentiostat were connected together at one of

the cell electrodes, and the working electrode and sensor probes of the Potentiostat were

connected together at the other cell electrode. The Nyquist plots were obtained at a -0.7 V

open-circuit bias using a 10 mV AC signal automatically swept between 0.1 Hz and 1000

KHz, under illumination from a Fiber-Lite MI-150 High Intensity Illuminator (Dolan –

Jenner, Boxborough, MA, USA).

Figure 3.4 Functional Diagram of EIS Setup [109]

Cyclic voltammetry (CV) Measurements

Figure 3.5 shows a functional diagram of a cyclic voltammetry (CV) experimental

58

setup used to evaluate the catalytic properties of the different concentrations of urea and

rGO treated electrolyte. The cyclic voltammetry was carried out on test solutions of

acetonitrile containing 10 mM lithium iodide (LiI) and 1 mM of iodine (I2), combined with

0.1 M lithium perchloride (LiClO4) and the same urea or rGO concentrations used in

fabricating the DSSCs. Using the three-electrode system, the substrate Pt CEs, solid

Ag/AgCl, and a Pt wire placed in each solution were used as the working electrode, the

reference electrode, and the CE, respectively. The voltage applied to the working electrode

was swept from -0.5 V to a maximum of 1.5 V at a rate of 0.05 V/s.

Figure 3.5 Functional Diagram of Cyclic Voltammetry (CV)

UV-Visible Absorption Spectra Measurements

UV-Visible absorption measurements were conducted using an Agilent 8453

spectrophotometer running ChemStation software, with a deuterium lamp as the ultraviolet

(UV) source and a tungsten lamp as the visible/near infrared (VNIR) source (Figure 3.6).

Two sets of measurements were performed – one on the dye absorbed in the

photoanode containing the urea or rGO additive, and the other on the electrolyte solution

containing the urea or rGO additive.

59

The electrolyte measurements were obtained as follows. First, a 3 ml Isopropyl Alcohol

(IPA) solution was loaded in a sample holder, and its properties measured. In this case, the

data processing software was set to operate in “Blank” measurement mode to establish a

“reference” absorption spectrum. Then, the samples of additive-treated electrolyte solution

were combined with the IPA solution and the data processing software was set to operate

in “Sample” measurement mode. The net absorption spectrum was determined in the data

processing software by subtracting the IPA absorption spectrum from the electrolyte/IPA

solution spectrum.

The dye absorption measurements were performed as follows. After a DSSC was

fabricated, an electrolyte redox solution with the given urea or rGO concentration was

injected into it, which was allowed to “rest” for approximately 24 hours. The DSSC was

then disassembled; the photoanode was rinsed with acetonitrile and dried using compressed

N2 gas for approximately 30 seconds, then soaked in a 3 mL, 1 M NaOH solution for 3

hours [110]. The spectra of desorbed dye solution from the photoanode was obtained as

follows:

1. A 3 mL NaOH solution was loaded into a sample holder to obtain a reference

measurement.

2. The solution containing the NaOH /dye/additive was loaded into a sample

holder to obtain a sample measurement.

3. The net absorption spectrum was determined by subtracting the reference

(NaOH) spectrum from the total NaOH /dye/additive solution spectrum.

60

Figure 3.6 Schematic of UV-Visible Absorption Spectroscopy [111]

61

Chapter 4: Results and Analysis

This chapter presents the results obtained from analyses of DSSC performance given

the use of urea and rGO as iodide/triiodide electrolyte additives. As mentioned earlier, five

weight percentage concentrations (wt%) of urea / rGO were prepared: 0%, 2%, 5%, 10%,

and 15%. Each additive will be considered in separate sections.

DSSC Performance with Urea-Treated Electrolyte

J-V Characteristics

Figure 4.1 shows the current density-voltage (J-V) curves for one of the DSSCs

fabricated with the electrolyte treated with each urea concentration. This particular DSSC

exhibited acceptable performance between the highest and lowest performances of the

other two. Table 4.1 summarizes the average resulting photovoltaic performance of all

three devices.

The open circuit voltage Voc increased with increasing urea concentration, from 0.72 V

at 0 wt% concentration to 0.83 V at 15 wt% concentration. The increase in Voc could be

attributed to the H+ ions produced by the reaction of urea with iodine in the electrolyte that

also yields oxalyl dihydrazide and I- ions, as shown in eq. (2.31). The H+ ions can lead to

a positive shift in the redox potential of the electrolyte [81]. Another possible reason for

the observed increase in Voc is a decrease in the I3- ion content in the electrolyte when the

increased urea concentration increases the iodide ion concentration that can transfer faster

which increased the ionic conductivity, as shown in eq. (2.31) and (2.32) [104].

The FF also increased with increasing urea concentration, from 0.61 at 0 wt%

concentration to 0.68 at 15 wt% concentration. This increase may be due to a lower

62

recombination rate at the TiO2/electrolyte interface caused by the same mechanism

accounting for the increase in Voc – the increase in I- ions and corresponding decrease in

I3- ions due to the increased urea concentration. This reduces the back recombination of the

electrolyte with I3- ions, and thus increases the FF.

-20

-18

-16

-14

-12

-10

-8

-6

-4

-2

00.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

0 wt% 2 wt% 5 wt% 10 wt% 15 wt%

Cur

rent

den

sity

(m

A/c

m2 )

Voltage (v)

Figure 4.1 DSSC Current Density-Voltage (J-V) Curves Due to Urea-Treated Electrolyte

Table 4.1 DSSC Performance Parameters Due to Urea-Treated Electrolyte

Urea Conc. (wt %)

Jsc (mA/cm2)

Voc (volt)

FF

ηηηη (%)

0 -19.14±0.03 0.72 0.61 8.51±0.01

2 -18.70±0.02 0.76 0.63 8.95±0.02

5 -18.99±0.02 0.79 0.65 9.78±0.01

10 -18.76±0.03 0.81 0.66 10.17±0.01

15 -16.79±0.01 0.83 0.68 9.56±0.03

63

While both Voc and FF increased with increasing urea concentration in the electrolyte,

the short circuit current density Jsc density remained virtually unchanged at approximately

-19 mA/cm2, until the urea concentration increased to 15 wt%. At this concentration, Jsc

significantly decreased to approximately 16.79 mA/cm2 due to dye molecule desorption by

excess H+ ions in the electrolyte. Urea concentrations above 15 wt% resulted in

solidification of the electrolyte, which will also reduce Jsc.

As a result of the increased Voc and FF, the overall PCE also increased with increasing

urea concentration, from ~8.51 % at 0 wt% concentration to ~10.17% at 10 wt%

concentration. Interestingly, the PCE decreased slightly to ~9.56% at 15 wt%

concentration, most likely due to the observed decrease in Jsc. With respect to this

parameter, the “optimum” urea concentration was found to be 10 wt%.

UV-Vis Absorbance Spectra

Figure 4.2 shows the UV-Vis absorption spectra of I-/I-3 electrolyte at each urea

concentration. The absorption peak wavelength was found to be ~360 nm, and is related to

the presence of I3- ions. [81, 112]. Increasing the urea concentration in the electrolyte did

not shift the peak wavelength location, but it did decrease the peak intensity. This

observation is consistent with previously reported analysis [81]. Adding urea to I-/I-3

electrolyte decrease the numbers of I-3 ions in the electrolyte.

64

300 400 500 600 700 8000.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

0 wt% 2 wt% 5 wt% 10 wt% 15 wt%

Abs

orba

nce

(a.u

.)

Wavelength (nm)

Figure 4.2 UV-Vis Absorbance Spectra of Urea-Treated Electrolyte

External Quantum Efficiency (EQE)

Figure 4.3 shows the external quantum efficiency (EQE) as a function of wavelength

at each urea concentration. To validate the J-V measurement results presented in Section

4.1.1, Jsc was estimated from the EQE as follows [113]:

J : eη λNUλdλ¢£

¢¤

(4.1)

where ηIPCE (λ) and Nph (λ) are the IPCE and photon flux at wavelength λ, respectively, and

e is the magnitude of electronic charge in Coulombs. Using equation 4.1, the values of Jsc

calculated from the EQE measurements are shown in Table 4.2. The calculated Jsc values

are consistent with those measured from the J-V curves.

65

400 500 600 700 800

0

20

40

60

80

100

0 wt% 2 wt% 5 wt% 10 wt% 15 wt%

EQ

E (

%)

Wavelength (nm)

Figure 4.3 EQE Spectral Response Due to Urea-Treated Electrolyte

Table 4.2 EQE Peaks of DSSCs Due to Urea-Treated Electrolyte

Urea Conc. (wt%)

Jsc (mA/cm2)

From EQE

Jsc (mA/cm2) From J-V

Peak EQE (%)

0 19.15 19.14 89.04

2 18.98 18.70 88.22

5 18.96 18.99 87.67

10 18.89 18.76 86.77

15 16.35 16.79 77.15

Figure 4.3 shows a slightly decreased EQE near 360 nm at 0% urea concentration. This

could be attributed to the absorption of I-3

ions in the electrolyte. Increasing the urea

concentration up to 10% in the electrolyte resulted in an increasing EQE. These results are

consistent with the UV-Vis absorption spectra results shown in Figure 4.2, indicating that

66

the I3- concentration decreased with increasing urea concentration. However, the overall

EQE spectra of DSSC devices decreased when the urea concentration in the electrolyte

increased from 10% to 15%.

Electrochemical Impedance Spectroscopy (EIS)

Figure 4.4(a) shows the resulting Nyquist plots of both extracted and fitted

electrochemical impedance data for the prepared DSSCs with active area 0.2 cm2, obtained

under typical illumination and open-circuit conditions using the equivalent circuit model

shown in Figure 4.4(b). Two frequency-dependent arcs were obtained. The high-frequency

arc represents the Pt/electrolyte interface impedance, while the mid-frequency arc

represents the TiO2/N719 dye/redox electrolyte interface impedance, which is a measure

of the electron back recombination resistance or back charge transfer resistance Rr.

The overall device series resistance Rs values at each concentration are shown in Table

4.3. The decrease in Rs with increasing urea content is consistent with the increase in fill

factor values from the J-V curves.

From the Nyquist plots, no significant change is observed in the diameter of the high-

frequency arcs. However, a noticeable change in mid-frequency arc diameters occurs with

changing urea concentration, indicating an increase in Rr, which results in less

recombination of the photo-generated electrons with triiodide (I3-) ions in the electrolyte at

the TiO2 / electrolyte interface. This reduction in recombination occurs whether the cell is

operating under illumination or open-circuit conditions.

The Rr value was fitted and calculated using the equivalent circuit model shown in

Figure 4.4(b). The predicted values for each urea concentration are given in Table 4.3. The

increase in Rr with increasing urea concentration in the electrolyte was attributed to the

67

decrease in the concentration of heavy triiodide (I3-) ions at the photo-anode /electrolyte

interface. The decrease in I3- ion concentration was due to the increase in iodide ion

concentration as mentioned in equation 2.31, which acted to inhibit recombination at the

photo-anode/electrolyte interface.

15 20 25 30 35 40 45 500

5

10

15

20

25

-Zim

(Ω

)

Zre (Ω)

0 wt% fit 2 wt% fit 5 wt% fit 10 wt% fit 15 wt% fit

4.4(a) Nyquist Plots of DSSCs Due to Urea-Treated Electrolyte

4.4(b) Equivalent Circuit Model Used to Fit Nyquist Plot

68

Table 4.3 Predicted Rs and Rr from Nyquist Plots of Urea-Treated Electrolyte

Urea Conc. (wt%)

Rs (Ω)

Rr (Ω)

0 18.9 10.3

2 18.4 22.2

5 17.9 26.1

10 17.7 25.3

15 17.6 19.6

Cyclic Voltammogram (CV) Measurement

Figure 4.5 shows the cyclic voltammetry (CV) curves for the Pt CEs immersed in the

I-/I3- electrolyte at each urea concentration. Table 4.4 shows the corresponding anodic and

cathodic peak current density intensities measured at each concentration.

The CV plots show a pair of cathodic and anodic peaks on the left side, where Jpa and

Jpc are the anodic and cathodic peak current densities, respectively. These correspond to

the following electrochemical reactions:

56( 2#( → 35(cathodicreaction (4.4) 35( → 56( 2#(anodicreaction (4.5)

The catalytic performance of the device can be determined by these two reactions [114].

Figure 4.5 shows that both anodic and cathodic peak currentdensities increase with

increasing urea concentration in the electrolyte. An overall positive shift in the oxidation

potential of all anodic peaks can also be observed, and is believed to be due to the positive

shift of the redox potential due to the introduction of H+ ions according to equation 2.31,

which results in an increase in Voc. This behavior is consistent with the observed increase

in Voc in the J-V characterization.

69

-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0-3

-2

-1

0

1

2

3

4

5

I3

-+2e

- 3I

-

3I - I

3

-+2e

-

Cur

rent

den

sity

(mA

/cm

2 )

Potential (V vs. Ag/AgCl)

0 wt % 2wt % 5 wt % 10 wt % 15 wt %

4.5 Cyclic Voltammograms of Counter Electrodes Immersed in Urea-Treated Electrolyte

Table 4.4 Anodic and Cathodic Peak Intensities of Urea-Treated Electrolyte

Urea Conc. (wt %)

Jpa (mA/cm2)

-Jpc (mA/cm2)

0 2.553 1.378

2 3.034 1.708

5 2.853 1.555

10 3.200 1.829

15 3.502 2.055

UV-Vis Absorption of Photoanodes Immersed in Urea-Treated Electrolyte

(dye loading)

To better understand the decrease in Jsc as a function of increasing urea content in the

electrolyte, dye loading UV-Vis absorbance measurements were performed for dye

solutions desorbed from photo-anodes at each urea concentration, prepared according to

the procedure described in section 3.2.5. Figure 4.6 shows the resulting UV-Vis spectra of

70

the desorbed dye solutions from the disassembled photoanodes for each electrolyte

solution. The two absorption peaks at ~372 nm and 512 nm are attributed to the N719 dye.

The absorption peak intensities decrease linearly with increasing urea concentration. The

obtained results are consistent with the Jsc values obtained from the J-V measurements and

indicate that dye molecules are desorbed by the urea-containing electrolytes, with more

desorption typically occurring at higher urea concentrations.

300 400 500 600 700 800

0 wt % 2wt % 5 wt % 10 wt % 15 wt %

Abs

orba

nce

(a.u

.)

Wavelength(nm)

4.6 UV-Vis Absorbance Spectra of Dye Solutions Desorbed from Electrolytes Treated with Urea

J-V Hysteresis Effects

Hysteresis effects have been observed in the J-V characteristics of DSSC and

perovskite solar cells [115, 116]. One possible reason for this behavior is that the capacitive

nature of the mesoporous TiO2 layer allows storage of heavy ions such as I-3 in the DSSC

electrolyte [117] and halide ions in the perovskite cell [118]. The hysteresis effect is

71

significant at very fast J-V scan rates (~1500 mv/s) due to a decrease in the delay time

[119].

Figure 4.7 shows the resulting J-V curves with hysteresis effects for DSSC devices

fabricated with electrolyte solutions at each urea concentration. The scans were performed

from 0 V to 1 V then from 1 V to 0 V in steps of 50 mV. The hysteresis effect in Fig. 4.7

is substantial for the DSSC devices fabricated without added urea. Increasing the urea

content results in a decrease in the observed hysteresis, consistent with the previous

observed behavior. At higher urea concentrations, more I-3 ions are converted to I- ions;

fewer I-3 ions will penetrate and be stored in the mesoporous TiO2 layer.

-50

-40

-30

-20

-10

0

10

20

30

40

500.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

0 wt% 2 wt% 5 wt% 10 wt% 15 wt%

Voltage (V)

Cur

rent

den

sity

(m

A/c

m2 )

4.7 Hysteresis Effects for DSSC Devices Fabricated with Urea-Treated Electrolyte

72

DSSC Performance Using rGO-Treated Electrolyte

J-V Characteristics

Figure 4.8 shows the current density-voltage (J-V) curves for one of the DSSCs

fabricated with the electrolyte treated with each rGO concentration. This particular DSSC

exhibited acceptable performance between the highest and lowest performances of the

other two. Table 4.5 summarizes the average resulting photovoltaic performance of all

three devices.

Voc increased from 0.71 V to 0.75 V as the rGO concentration increased from 0% to

10%. While similar in overall trend to that observed with urea, a decrease was observed at

the 15% concentration, and the range of Voc increase was smaller.

The FF changed over a narrow range, between 0.59 and 0.62 between 0 wt% and 10

wt% concentrations, and decreasing back to 0.59 at 15 wt% concentration. As with Voc, the

range of increase was smaller than the FF increase measured with the urea-treated

electrolyte. The potential mechanism for this interaction differs from that of the urea-

treated electrolyte: a lower recombination rate at the TiO2/electrolyte interface. As

mentioned in section 2.9, I-3 ions are large and heavy ions, with low mobility and a

tendency to concentrate near the photoanode (Figure 2.17), which promotes recombination

reactions between the TiO2 conduction band and the electrolyte. This reduces both FF and

Voc. Adding rGO to the electrolyte slightly increases Voc and accelerates the charge transfer

rate by reducing the electrolyte’s resistance and charge back-recombination rate. This

behavior is related to the strong interaction of I-3 ions with the rGO particles. This positive

interaction allows the removal of I-3 ions far from the photoanode interfaces, permitting

some recovery of FF and Voc.

73

The short circuit current (Jsc) showed a very slight increase. This indicates that rGO

works as a bridge, accelerating the electron transfer rate by reducing the electrolyte

resistance.

The overall PCE (%) was improved from ~ 8.54% at 0% concentration to ~9.57% at

10 wt% concentration. The PCE decreased back to ~8.79% at 15 wt% concentration.

-22

-20

-18

-16

-14

-12

-10

-8

-6

-4

-2

00.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

0 wt% 2 wt% 5 wt% 10 wt% 15 wt%

Voltage (v)

Cur

rent

den

sity

(m

A/c

m2 )

4.8 DSSC Current Density-Voltage (J-V) Curves Due to rGO-Treated Electrolyte

Table 4.5 DSSC Performance Due to rGO-Treated Electrolyte

rGO Conc. (wt%)

Jsc (mA/cm2)

Voc (volt)

FF

ηηηη (%)

0 -20.34±0.04 0.71 0.59 8.54±0.01

2 -20.72±0.05 0.72 0.59 8.8±0.02

5 -21.71±0.04 0.73 0.58 9.29±0.03

10 -20.33±0.03 0.75 0.62 9.57±0.01

15 -20.37±0.02 0.72 0.59 8.79±0.01

74

UV-Vis Absorbance Spectra

Figure 4.9 shows the UV-Vis absorption spectra of I-/I-3 electrolyte at each rGO

concentration. As with the urea additive, the absorption peak at ~360 nm did not shift

location. However, increasing the rGO concentration increased, rather than decreased, the

peak intensity, resulting in an increase in the electrolyte’s ionic conductivity [120].

300 400 500 600 700 8000.0

0.1

0.2

0.3

0.4

Abs

orba

nce

(a.u

.)

Wavelength (nm)

0 wt% 2 wt% 5 wt% 10 wt% 15 wt%

4.9 UV-Vis Absorbance Spectra of rGO-Treated Electrolyte

External Quantum Efficiency (EQE)

Figure 4.10 shows the external quantum efficiency (EQE) as a function of wavelength

for DSSC electrolyte treated with each rGO concentration. Again, the Jsc of the tested

DSSCs was estimated by integrating the EQE using equation 4.1. The estimated values and

corresponding Jsc values measured from the J-V characterization are shown in Table 4.5.

The calculated Jsc values are consistent with those measured from the J-V curves. EQE

75

increased with increasing rGO concentration; the maximum EQE value was 92.47% when

the rGO concentration was 10 wt%. The increase in EQE could be due to increased ionic

conductivity of the electrolyte and/or decreased TiO2/electrolyte recombination at the

interface. Table 4.6 summarizes the values of EQE at each rGO concentration.

350 400 450 500 550 600 650 700 750 8000

20

40

60

80

100

EQ

E (

%)

Wavelength (nm)

0 wt% 2 wt% 5 wt% 10 wt% 15 wt%

Figure 4.10 EQE Spectral Response Due to rGO-Treated Electrolyte

76

Table 4.6 EQE Peaks of DSSCs Due to rGO-Treated Electrolyte

rGO Conc. (wt%)

Jsc (mA/cm2)

From EQE

Jsc (mA/cm2) From J-V

Peak EQE (%)

0 19.89 20.34 84.19

2 19.77 20.72 89.53

5 19.82 21.71 90.12

10 20.91 20.33 92.47

15 19.92 20.37 87.88

Electrochemical Impedance Spectroscopy

Figure 4.11(a) shows the resulting Nyquist plots of both extracted and fitted

electrochemical impedance of DSSCs with active area 0.2 cm2. The data obtained under

typical illumination and open-circuit conditions, using the equivalent circuit model shown

in Figure 4.11(b). Similar to the results obtained for the urea-treated electrolyte, two sets

of frequency-dependent arcs were obtained. These arcs tended to be larger than the

corresponding arcs observed with the urea-treated electrolyte.

Table 4.7 gives the overall device series and recombination resistances Rs and Rr.

Again, Rs decreased and Rr increased as the rGO concentration increased, consistent with

the increase in FF demonstrated in the J-V characterization.

From the Nyquist plots in Figure 4.11(a), there is no significant change in the diameter

of arcs. However, a noticeable change in midfrequency arc diameters occurs after varying

rGO content, indicating an increase in Rr as the rGO concentration increased, which results

in a decrease in TiO2 / electrolyte interface recombination.

77

15 20 25 30 35 400

2

4

6

8

10

12

14 0 wt% fit 2 wt% fit 5 wt% fit 10 wt% fit 15 wt% fit

-Zim

(Ω)

Zre (Ω)

Figure 4.11(a) Nyquist Plots of DSSCs Due to rGO-Treated Electrolyte

Figure 4.11(b) The Equivalent Circuit Model Used to Fit Nyquist Plot

78

Table 4.7 Predicted Rs and Rr from Nyquist Plots of rGO-Treated Electrolyte

rGO Conc. (wt%)

Rs (Ω)

Rr (Ω)

0 18.8 11.6

2 18.4 12.6

5 18.2 13.0

10 18.2 16.1

15 16.2 12.45

Cyclic Voltammogram Measurement

Figure 4.12 shows the CV curves for the Pt CEs immersed in the I−/I3− electrolyte as a

function of electrolyte rGO concentration. Table 4.8 shows the corresponding anodic and

cathodic peak intensities measured at each concentration. Similar to the CV plots for urea-

treated electrolyte shown in Section 4.1.5, the plots show Jpa and Jpc peak intensities

corresponding to the electrochemical reactions described by equations 4.4 and 4.5. Figure

4.12 shows that both anodic and cathodic reactions are enhanced by adding rGO to the

electrolyte, with higher peak anodic and cathodic current densities measured at rGO

concentrations less than 15 wt%. Unlike the urea-treated electrolyte, no shift in the peak

positions was observed.

79

-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

I3

- +2e-------3I-

3I- ------I-

3+2e-

0 wt% 2 wt% 5 wt% 10 wt% 15 wt%

Cur

rent

den

sity

(m

A/c

m2 )

Potential (V vs. Ag/AgCl)

Figure 4.12 Cyclic Voltammograms of Counter Electrodes Immersed in rGO-Treated Electrolyte

Table 4.8 Anodic and Cathodic Peak Intensities of rGO-Treated Electrolyte

rGO Conc. (wt %)

Jpa (mA/cm2)

-Jpc (mA/cm2)

0 1.859 1.015

2 1.905 1.196

5 1.982 1.269

10 2.037 1.377

15 1.739 1.140

80

UV-Vis Absorption of Photoanodes Immersed in rGO-Treated Electrolyte (dye

loading)

Dye loading UV-Vis absorbance measurements were performed for dye solutions

desorbed from photoanodes at each rGO concentration, prepared according to the

procedure described in section 3.2.5. Figure 4.13 shows the resulting UV-Vis spectra of

the desorbed dye solutions from the disassembled photoanodes for each electrolyte

solution. As with the urea concentration, the two absorption peaks at ~372 nm and 512 nm

correspond to the N719 dye. However, the absorption peak intensities increased linearly

with increasing rGO concentration, with a maximum peak intensity observed at 10 wt%

concentration. This is likely due to the higher reflectivity of rGO, which allows the dye to

absorb more light. The greater absorption results in an increased Jsc. The higher reflectivity

effect appears to be offset at a 15 wt% rGO concentration, as the measured peak intensity

was reduced.

81

300 400 500 600 700 800

Abs

orba

nce

(a.u

.)

Wavelength(nm)

0 wt% 2 wt% 5 wt% 10 wt% 15 wt%

Figure 4.13 UV-Vis Absorbance Spectra of Dye Solutions Desorbed from Electrolytes Treated with rGO

J-V Hysteresis Effects

Figure 4.14 shows the J-V curves with hysteresis effects obtained for the DSSC devices

fabricated with electrolyte solutions containing varying rGO concentrations, with the scans

performed from 0 VTO 1 V then from 1 V to 0 V in steps of 50 mV. Adding rGO to the

electrolyte resulted in a decrease in the observed hysteresis. This is due to greater I3-

reaction with the rGO, which prevents the I3- ions from penetrating the TiO2 mesoporous

layer.

82

-50

-40

-30

-20

-10

0

10

20

30

40

500.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

0 wt% 2 wt% 5 wt% 10 wt% 15 wt%

Voltage (V)

Cur

rent

den

sity

(mA

/cm

2 )

Figure 4.14 Hysteresis Effects for DSSC Devices Fabricated with rGO-Treated Electrolyte

83

Chapter 5: Summary and Conclusions

Summary

The use and increasing cost of energy are among the most important problems the

world is currently facing. Solar energy, fossil fuels and nuclear energy are currently

available sources. Fossil fuels and nuclear energy have issues relating to ongoing

availability and safety questions that make their use problematic. Solar PVs show great

promise as an alternative energy source. However, Si-based PV cells require creation of

high purity Si wafers, increasing their cost and limiting their practical applicability. Other

solar PV types, such as single junction GaAs or thin films (CIGS), are also more costly to

fabricate. DSSCs, Perovskite, and polymer cells are less costly to fabricate, and their

properties have been extensively researched. With respect to DSSCs, a key research

question has been how to improve the overall energy conversion efficiency; much of the

research has focused on enhancing the electrochemical properties of the electrolyte

solution.

This work investigated the use of novel electrolyte additive materials to DSSCs using

the standard I-/I-3 redox electrolyte. DSSC performance using varying additive

concentrations is characterized, with the goal of further optimizing photovoltaic

characteristics and overall photoconversion efficiency. The objective of this work was to

improve the DSSC efficiency by more than 1% using commercially available liquid I-/I-3

redox electrolyte, using different concentrations of urea (CH4N2O) or rGO as the sole

electrolyte additive. J-V, EQE, EIS, CV and UV-Vis absorption spectra measurements of

all fabricated devices demonstrated improvement in DSSC electrical performance to the

84

required degree with either additive. For the DSSCs characterized in this work, urea added

to the electrolyte solutions resulted in a greater than 1% increase in electrical performance.

Conclusions

Adding urea to the DSSC iodide/triiodide electrolyte led to measurable increases in Voc

and FF. The increase in Voc could be caused by the introduction of H+ ions into the

electrolyte, as urea reacts with iodine and causes a positive shift in the redox potential of

the electrolyte. The increase in FF with increasing urea concentration resulted from a lower

recombination rate at the TiO2 / electrolyte interface. The I- ion concentration increases

while the I-3 ion concentration at the TiO2/ electrolyte interface decreases, which inhibits

back charge recombination at the interface. The overall PCE (%) was significantly

increased, from ∼8.51% for 0 wt% urea to ∼10.17% for 10 wt% urea. However, the Jsc

value decreased at 15 wt% urea. This decrease could be due to dye molecule desorption

due to excess H+ ions in the electrolyte.

Adding rGO to the iodide/triiodide electrolyte of DSSCs lead to increases in both Voc

and FF, but these increases were less than those measured with urea. Both Voc and FF were

most improved with a 10 wt% rGO concentration; this could be due to reduction of the

electrolyte’s resistance and charge back-recombination rate, which increases the charge

transfer rate. This behavior is related to the strong interaction of I-3 ions with the rGO

particles, which allows the removal of I-3 ions far from the photoanode interface. The

overall PCE (%) was significantly increased from ~ 8.54 % at 0 wt% rGO concentration to

~9.57 % at 10 wt% rGO concentration. However, the Jsc value shows relatively little change

with increasing rGO concentration.

85

Higher rGO concentrations do not appear to have a similar level of effect to the

corresponding urea concentrations. However, for both tested additives, the performance

enhancement objectives were successfully achieved.

Future work

Given the demonstrated improvement in DSSC performance with either of these

electrolyte additives, additional characterization of DSSC performance over longer

intervals of time could be performed, especially with respect to the overall

photoconversion efficiency (e.g. the degree of degradation of the additive in the

electrolyte and/or dye, etc). Another potential research direction would consider the

efficacy of adding either additive to the electrolyte of Li ion batteries in order to

enhance their capacity and stability above the levels currently achieved.

86

References

1. Drake, B.L., et al., The carbon fertilization effect over a century of anthropogenic CO2 emissions: higher intracellular CO2 and more drought resistance among invasive and native grass species contrasts with increased water use efficiency for woody plants in the US Southwest. Global change biology, 2017. 23(2): p. 782-792.

2. Demirbaş, A., Global renewable energy resources. Energy sources, 2006. 28(8): p. 779-792.

3. Lei, M., et al., A review on the forecasting of wind speed and generated power. Renewable and Sustainable Energy Reviews, 2009. 13(4): p. 915-920.

4. Yang, X., Y. Xiao, and S. Chen, Wind speed and generated power forecasting in wind farm. Proceedings-Chinese Society of Electrical Engineering, 2005. 25(11): p. 1.

5. McKendry, P., Energy production from biomass (part 1): overview of biomass. Bioresource technology, 2002. 83(1): p. 37-46.

6. Kalogirou, S.A., Solar energy engineering: processes and systems. 2013: Academic Press.

7. Wackernagel, M. and W. Rees, Our ecological footprint: reducing human impact on the earth. 1998: New Society Publishers.

8. Choubey, P., A. Oudhia, and R. Dewangan, A review: Solar cell current scenario and future trends. Recent Research in Science and Technology, 2012. 4(8).

9. Wang, X., et al., A metal-free polymeric photocatalyst for hydrogen production from water under visible light. Nature Materials, 2008. 8: p. 76.

10. Zhao, J., et al., 19.8% efficient “honeycomb” textured multicrystalline and 24.4% monocrystalline silicon solar cells. Applied Physics Letters, 1998. 73(14): p. 1991-1993.

11. Krebs, F.C., et al., Freely available OPV—the fast way to progress. Energy Technology, 2013. 1(7): p. 378-381.

12. Hoppe, H. and N.S. Sariciftci, Organic solar cells: An overview. Journal of materials research, 2004. 19(7): p. 1924-1945.

13. Günes, S., H. Neugebauer, and N.S. Sariciftci, Conjugated polymer-based organic solar cells. Chemical reviews, 2007. 107(4): p. 1324-1338.

14. O'regan, B. and M. Grätzel, A low-cost, high-efficiency solar cell based on dye-sensitized colloidal TiO2 films. nature, 1991. 353(6346): p. 737.

15. Gong, J., J. Liang, and K. Sumathy, Review on dye-sensitized solar cells (DSSCs): fundamental concepts and novel materials. Renewable and Sustainable Energy Reviews, 2012. 16(8): p. 5848-5860.

16. Kalyanasundaram, K., Dye-sensitized solar cells. 2010: EPFL press. 17. Zumeta, I., et al., TiO2 films obtained by microwave-activated chemical-bath

deposition used to improve TiO2-conducting glass contact. Solar Energy Materials and Solar Cells, 2009. 93(10): p. 1728-1732.

18. Sima, C., C. Grigoriu, and S. Antohe, Comparison of the dye-sensitized solar cells performances based on transparent conductive ITO and FTO. Thin Solid Films, 2010. 519(2): p. 595-597.

87

19. Xuhui, S., et al., Performance comparison of dye-sensitized solar cells by using different metal oxide-coated TiO2 as the photoanode. AIP Advances, 2014. 4(3): p. 031304.

20. Lee, C.-P., C.-T. Li, and K.-C. Ho, Use of organic materials in dye-sensitized solar cells. Materials Today, 2017.

21. Park, N.-G., J. Van de Lagemaat, and A. Frank, Comparison of dye-sensitized rutile-and anatase-based TiO2 solar cells. The Journal of Physical Chemistry B, 2000. 104(38): p. 8989-8994.

22. Fang, X., et al., Graphene-compositing optimization of the properties of dye-sensitized solar cells. Solar Energy, 2014. 101: p. 176-181.

23. Hagfeldt, A., et al., Dye-sensitized solar cells. Chemical reviews, 2010. 110(11): p. 6595-6663.

24. Kuang, D., et al., High‐Efficiency and Stable Mesoscopic Dye‐Sensitized Solar Cells Based on a High Molar Extinction Coefficient Ruthenium Sensitizer and Nonvolatile Electrolyte. Advanced Materials, 2007. 19(8): p. 1133-1137.

25. Kuang, D., et al., High molar extinction coefficient heteroleptic ruthenium complexes for thin film dye-sensitized solar cells. Journal of the American Chemical Society, 2006. 128(12): p. 4146-4154.

26. 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, 2005. 127(10): p. 3456-3462.

27. Haque, S.A., et al., Supermolecular Control of Charge Transfer in Dye‐Sensitized Nanocrystalline TiO2 Films: Towards a Quantitative Structure–Function Relationship. Angewandte Chemie International Edition, 2005. 44(35): p. 5740-5744.

28. Kuciauskas, D., et al., Transient absorption spectroscopy of ruthenium and osmium polypyridyl complexes adsorbed onto nanocrystalline TiO2 photoelectrodes. The Journal of Physical Chemistry B, 2002. 106(36): p. 9347-9358.

29. Hasselmann, G.M. and G.J. Meyer, Diffusion-limited interfacial electron transfer with large apparent driving forces. The Journal of Physical Chemistry B, 1999. 103(36): p. 7671-7675.

30. Ferrere, S. and B.A. Gregg, Photosensitization of TiO2 by [FeII (2,2 ‘-bipyridine-4, 4 ‘-dicarboxylic acid)2 (CN)2]: band selective electron injection from ultra-short-lived excited states. Journal of the American Chemical Society, 1998. 120(4): p. 843-844.

31. Geary, E.A., et al., Synthesis, structure, and properties of [Pt (II)(diimine)(dithiolate)] dyes with 3,3 ‘-, 4,4 ‘-, and 5,5 ‘-disubstituted bipyridyl: applications in dye-sensitized solar cells. Inorganic chemistry, 2005. 44(2): p. 242-250.

32. Bessho, T., et al., An element of surprise—efficient copper-functionalized dye-sensitized solar cells. Chemical Communications, 2008(32): p. 3717-3719.

33. Mathew, S., et al., Dye-sensitized solar cells with 13% efficiency achieved through the molecular engineering of porphyrin sensitizers. Nature chemistry, 2014. 6(3): p. 242.

88

34. Hagfeldt, A. and M. Grätzel, Molecular photovoltaics. Accounts of Chemical Research, 2000. 33(5): p. 269-277.

35. Zhang, S., et al., Highly efficient dye-sensitized solar cells: progress and future challenges. Energy & Environmental Science, 2013. 6(5): p. 1443-1464.

36. Fattori, A., Electrochemical and spectroelectrochemical studies of dyes used in dye-sensitized solar cells. 2010, University of Bath.

37. Yella, A., et al., Molecular engineering of push–pull porphyrin dyes for highly efficient dye‐sensitized solar cells: The role of benzene spacers. Angewandte Chemie, 2014. 126(11): p. 3017-3021.

38. Park, S.S., et al., Molecular design of organic dyes with double electron acceptor for dye-sensitized solar cell. Energy & Fuels, 2009. 23(7): p. 3732-3736.

39. Ooyama, Y. and Y. Harima, Molecular designs and syntheses of organic dyes for dye‐sensitized solar cells. European Journal of Organic Chemistry, 2009. 2009(18): p. 2903-2934.

40. Kojima, A., et al., Organometal halide perovskites as visible-light sensitizers for photovoltaic cells. Journal of the American Chemical Society, 2009. 131(17): p. 6050-6051.

41. Chang, H. and Y.-J. Lo, Pomegranate leaves and mulberry fruit as natural sensitizers for dye-sensitized solar cells. Solar Energy, 2010. 84(10): p. 1833-1837.

42. Zhou, H., et al., Dye-sensitized solar cells using 20 natural dyes as sensitizers. Journal of Photochemistry and Photobiology A: Chemistry, 2011. 219(2): p. 188-194.

43. Ghann, W., et al., Fabrication, optimization and characterization of natural dye sensitized solar cell. Scientific reports, 2017. 7: p. 41470.

44. Wu, J., et al., Counter electrodes in dye-sensitized solar cells. Chemical Society Reviews, 2017. 46(19): p. 5975-6023.

45. Wu, J., et al., Bifacial dye-sensitized solar cells: A strategy to enhance overall efficiency based on transparent polyaniline electrode. Scientific reports, 2014. 4: p. 4028.

46. Wu, J., et al., Electrolytes in dye-sensitized solar cells. Chemical reviews, 2015. 115(5): p. 2136-2173.

47. Yu, Z., et al., Liquid electrolytes for dye-sensitized solar cells. Dalton transactions, 2011. 40(40): p. 10289-10303.

48. Xu, K., Nonaqueous liquid electrolytes for lithium-based rechargeable batteries. Chemical reviews, 2004. 104(10): p. 4303-4418.

49. Fischer, A., et al., Crystal formation involving 1-methylbenzimidazole in iodide/triiodide electrolytes for dye-sensitized solar cells. Solar energy materials and solar cells, 2007. 91(12): p. 1062-1065.

50. Wang, M., et al., Recent developments in redox electrolytes for dye-sensitized solar cells. Energy & Environmental Science, 2012. 5(11): p. 9394-9405.

51. Boschloo, G. and A. Hagfeldt, Characteristics of the iodide/triiodide redox mediator in dye-sensitized solar cells. Accounts of chemical research, 2009. 42(11): p. 1819-1826.

89

52. Gregg, B.A., et al., Interfacial recombination processes in dye-sensitized solar cells and methods to passivate the interfaces. The Journal of Physical Chemistry B, 2001. 105(7): p. 1422-1429.

53. Kambe, S., et al., Influence of the electrolytes on electron transport in mesoporous TiO2− electrolyte systems. The Journal of Physical Chemistry B, 2002. 106(11): p. 2967-2972.

54. Wang, P., et al., A stable quasi-solid-state dye-sensitized solar cell with an amphiphilic ruthenium sensitizer and polymer gel electrolyte. Nature materials, 2003. 2(6): p. 402.

55. Song, J., Y. Wang, and C.C. Wan, Review of gel-type polymer electrolytes for lithium-ion batteries. Journal of Power Sources, 1999. 77(2): p. 183-197.

56. Wu, J., et al., Gel polymer electrolyte based on poly (acrylonitrile-co-styrene) and a novel organic iodide salt for quasi-solid state dye-sensitized solar cell. Electrochimica Acta, 2006. 51(20): p. 4243-4249.

57. Wang, Y., Recent research progress on polymer electrolytes for dye-sensitized solar cells. Solar Energy Materials and Solar Cells, 2009. 93(8): p. 1167-1175.

58. Nogueira, A., C. Longo, and M.-A. De Paoli, Polymers in dye sensitized solar cells: overview and perspectives. Coordination Chemistry Reviews, 2004. 248(13-14): p. 1455-1468.

59. Zakeeruddin, S.M. and M. Graetzel, Solvent‐Free Ionic Liquid Electrolytes for Mesoscopic Dye‐Sensitized Solar Cells. Advanced Functional Materials, 2009. 19(14): p. 2187-2202.

60. Tennakone, K., et al., A dye-sensitized nano-porous solid-state photovoltaic cell. Semiconductor Science and Technology, 1995. 10(12): p. 1689.

61. Fujishima, A. and X.-T. Zhang, Solid-state dye-sensitized solar cells, in Nanostructured materials for solar energy conversion. 2006, Elsevier. p. 255-273.

62. Kanchi S, S.D., Bisetty k (2014) Dye Sensitized Solar Cells: Tool to Overcome the Future Energy Crisis. J Environ Anal Chem 1:e106. doi: 10.4172/2380-2391.1000e106.

63. Gurung, A., et al., Highly Efficient Perovskite Solar Cell Photocharging of Lithium Ion Battery Using DC–DC Booster. Advanced Energy Materials, 2017. 7(11): p. 1602105.

64. Shah, A., et al., Photovoltaic technology: the case for thin-film solar cells. science, 1999. 285(5428): p. 692-698.

65. Oskam, G., et al., Pseudohalogens for dye-sensitized TiO2 photoelectrochemical cells. The Journal of Physical Chemistry B, 2001. 105(29): p. 6867-6873.

66. Nusbaumer, H., et al., An Alternative Efficient Redox Couple for the Dye‐Sensitized Solar Cell System. Chemistry-A European Journal, 2003. 9(16): p. 3756-3763.

67. Wang, P., et al., A solvent-free, SeCN-/(SeCN) 3-based ionic liquid electrolyte for high-efficiency dye-sensitized nanocrystalline solar cells. Journal of the American Chemical Society, 2004. 126(23): p. 7164-7165.

68. Wang, Z.-S., K. Sayama, and H. Sugihara, Efficient eosin Y dye-sensitized solar cell containing Br-/Br3-electrolyte. The Journal of Physical Chemistry B, 2005. 109(47): p. 22449-22455.

90

69. Hattori, S., et al., Blue copper model complexes with distorted tetragonal geometry acting as effective electron-transfer mediators in dye-sensitized solar cells. Journal of the American Chemical Society, 2005. 127(26): p. 9648-9654.

70. Gorlov, M., et al., Electrolytes for dye-sensitized solar cells based on interhalogen ionic salts and liquids. Inorganic chemistry, 2007. 46(9): p. 3566-3575.

71. Zhang, Z., et al., The 2, 2, 6, 6‐Tetramethyl‐1‐piperidinyloxy Radical: An Efficient, Iodine‐Free Redox Mediator for Dye‐Sensitized Solar Cells. Advanced Functional Materials, 2008. 18(2): p. 341-346.

72. Wang, M., et al., An organic redox electrolyte to rival triiodide/iodide in dye-sensitized solar cells. Nature Chemistry, 2010. 2(5): p. 385.

73. Li, T.C., et al., Ni (III)/(IV) bis (dicarbollide) as a fast, noncorrosive redox shuttle for dye-sensitized solar cells. Journal of the American Chemical Society, 2010. 132(13): p. 4580-4582.

74. Daeneke, T., et al., High-efficiency dye-sensitized solar cells with ferrocene-based electrolytes. Nature chemistry, 2011. 3(3): p. 211.

75. Kato, F., et al., Nitroxide Radicals as Highly Reactive Redox Mediators in Dye‐Sensitized Solar Cells. Angewandte Chemie International Edition, 2012. 51(40): p. 10177-10180.

76. 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, 1993. 115(14): p. 6382-6390.

77. Nakade, S., et al., Role of electrolytes on charge recombination in dye-sensitized tio2 solar cell (1): the case of solar cells using the I-/I3-redox couple. The Journal of Physical Chemistry B, 2005. 109(8): p. 3480-3487.

78. Park, N.G., et al., Effect of Cations on the Open-Circuit Photovoltage and the Charge-Injection Efficiency of Dye-Sensitized Nanocrystalline Rutile TiO~ 2 Films. Bulletin-Korean Chemical Society, 2000. 21(10): p. 985-988.

79. Zhang, C., et al., Photoelectrochemical effects of guanidinium thiocyanate on dye-sensitized solar cell performance and stability. The Journal of Physical Chemistry C, 2009. 113(52): p. 21779-21783.

80. Yu, Z., et al., Synergistic effect of N-methylbenzimidazole and guanidinium thiocyanate on the performance of dye-sensitized solar cells based on ionic liquid electrolytes. The Journal of Physical Chemistry C, 2010. 114(50): p. 22330-22337.

81. Kim, M.-J., et al., Unusual enhancement of photocurrent by incorporation of bronsted base thiourea into electrolyte of dye-sensitized solar cell. The Journal of Physical Chemistry C, 2010. 114(46): p. 19849-19852.

82. Kim, M.-J. and N.-G. Park, Urea as a long-term stable alternative to guanidium thiocyanate additive in dye-sensitized solar cell. Applied Surface Science, 2012. 258(22): p. 8915-8918.

83. Brian Koberlein, SOLAR_SPECTRUM. 24 January 2017. 84. Luke Madaus, Radar Spikes are Not a Government Conspiracy. May14, 2011. 85. Solar, U., O. Layer, and S.I. Data, ORIEL PRODUCT TRAINING. 86. Richards, B. and M. Green, Photovoltaic Cells. 2006.

91

87. Einstein, A., Über einen die Erzeugung und Verwandlung des Lichtes betreffenden heuristischen Gesichtspunkt. Annalen der physik, 1905. 322(6): p. 132-148.

88. Jäger, K.-D., et al., Solar Energy: Fundamentals, Technology and Systems. 2016: UIT Cambridge.

89. Würfel, P., Physics of solar cells. Vol. 1. 2005: Wiley-vch Weinheim. 90. B. Van Zeghbroeck, Principles of Semiconductor Devices. 2011. 91. Jasim, K.E., Dye sensitized solar cells-working principles, challenges and

opportunities, in Solar Cells-Dye-Sensitized Devices. 2011, InTech. 92. Elbohy, H., Urea pre-treatment of N 2-annealed transition metal oxides for low

cost and efficient counter electrodes in dyesensitized solar cell. 2016: South Dakota State University.

93. Koide, N., et al., Improvement of efficiency of dye-sensitized solar cells based on analysis of equivalent circuit. Journal of Photochemistry and Photobiology A: chemistry, 2006. 182(3): p. 296-305.

94. Kumar, R.B. and P. Anand, Integration of Hybrid Renewable Energy Using Active Compensation for Distributed Generating System.

95. Rahman, M., N. Chandra Deb Nath, and J.J. Lee, Electrochemical Impedance Spectroscopic Analysis of Sensitization‐Based Solar Cells. Israel Journal of Chemistry, 2015. 55(9): p. 990-1001.

96. Instruments, G., Basics of electrochemical impedance spectroscopy. G. Instruments, Complex impedance in Corrosion, 2007: p. 1-30.

97. Ellis, H., Characterization of dye-sensitized solar cells: Components for environmentally friendly photovoltaics. 2014, Ångström.

98. Geng, Y., et al., Photovoltage improvements and recombination suppression by montmorillonite addition to PEO gel electrolyte for dye-sensitized solar cells. Physical Chemistry Chemical Physics, 2011. 13(6): p. 2417-2421.

99. Daniel Graham, Standard Operating Procedures for Cyclic Voltammetry. 100. William Reusch, Visible and Ultraviolet Spectroscopy. 2013. 101. ChemStation, U.-v., Agilent 8453 Spectrophotometer and UV-visible ChemStation

for Dissolution Testing. 2011. 102. Sharma, G., et al., Effect of thiourea incorporation in the electrolyte on the

photovoltaic performance of the DSSC sensitized with pyridyl functionalized porphyrin. Electrochimica Acta, 2013. 102: p. 459-465.

103. Pathak, R., et al., Self-recovery in Li-metal hybrid lithium-ion batteries via WO 3 reduction. Nanoscale, 2018. 10(34): p. 15956-15966.

104. Mathew, A., et al., Effect of iodine concentration on the photovoltaic properties of dye sensitized solar cells for various I2/LiI ratios. Electrochimica Acta, 2013. 87: p. 92-96.

105. Lan, J.-L., et al., Effects of iodine content in the electrolyte on the charge transfer and power conversion efficiency of dye-sensitized solar cells under low light intensities. The Journal of Physical Chemistry C, 2012. 116(49): p. 25727-25733.

106. Kalita, G., et al., Iodine doping in solid precursor-based CVD growth graphene film. Journal of Materials Chemistry, 2011. 21(39): p. 15209-15213.

92

107. Takahama, T., S.M. Saharin, and K. Tashiro, Details of the intermolecular interactions in poly (vinyl alcohol)-iodine complexes as studied by quantum chemical calculations. Polymer, 2016. 99: p. 566-579.

108. Berni, A., M. Mennig, and H. Schmidt, Doctor blade, in Sol-gel technologies for glass producers and users. 2004, Springer. p. 89-92.

109. Gong, J., Experimental and Theoretical Studies of Nanostructured Electrodes for Use in Dye-Sensitized Solar Cells. 2017, North Dakota State University.

110. Lan, Z., et al., Influence of NH3· H2O additive on the photovoltaic performance of dye-sensitized solar cells with chemical sintered scattering layers. Electrochimica Acta, 2011. 56(27): p. 9926-9930.

111. Agilent Technologies. Agilent 8453 UV-visible Spectroscopy System Operator’s Manual. 2011 [cited 2018 May14]; Available from: https://static1.squarespace.com/static/52607152e4b085141ba19f0f/t/53b600dfe4b0a8bef6e9ef0f/1404436703217/agilent8453uvvis_manual.pdf.

112. Zhang, Z., et al., Influence of Iodide Concentration on the Efficiency and Stability of Dye‐Sensitized Solar Cell Containing Non‐Volatile Electrolyte. ChemPhysChem, 2009. 10(11): p. 1834-1838.

113. Xie, Y., et al., Electrolyte effects on electron transport and recombination at ZnO nanorods for dye-sensitized solar cells. The Journal of Physical Chemistry C, 2010. 114(41): p. 17880-17888.

114. Elbohy, H., et al., Graphene-embedded carbon nanofibers decorated with Pt nanoneedles for high efficiency dye-sensitized solar cells. Journal of Materials Chemistry A, 2015. 3(34): p. 17721-17727.

115. Sarker, S., et al., On the hysteresis of current density-voltage curves of dye-sensitized solar cells. Electrochimica Acta, 2015. 182: p. 493-499.

116. Yang, W.S., et al., High-performance photovoltaic perovskite layers fabricated through intramolecular exchange. Science, 2015. 348(6240): p. 1234-1237.

117. Wang, Q., J.-E. Moser, and M. Grätzel, Electrochemical impedance spectroscopic analysis of dye-sensitized solar cells. The Journal of Physical Chemistry B, 2005. 109(31): p. 14945-14953.

118. Unger, E., et al., Hysteresis and transient behavior in current–voltage measurements of hybrid-perovskite absorber solar cells. Energy & Environmental Science, 2014. 7(11): p. 3690-3698.

119. Snaith, H.J., et al., Anomalous hysteresis in perovskite solar cells. The journal of physical chemistry letters, 2014. 5(9): p. 1511-1515.

120. Yang, Y., et al., The influence of Co3O4 concentration on quasi-solid state dye-sensitized solar cells with polymer electrolyte. Solid State Ionics, 2015. 279: p. 1-5.