and efficiencies of dye-sensitized solid-state photovoltaic cells of different configurations (D1= Fast Green, D2 = Rhodamine. 6G, D3 = Acridine Yellow). 67. 68.
Investigation of Electron Transport in Nanostructured Semiconductor Heterojunctions by Using Dye-Sensitized SolidState Solar Cells
by
Pitigala Kankanamage Don Duleepa Padmal Pitigala
Thesis submitted to the University of Sri Jayawardanapura for the award of the Degree of Master of Philosophy in Physics on 2006
DECLARATION
The work described in this thesis was carried out by me under the supervision of Prof. K Tennakone and Prof. D. A. Tantrigoda and a report on thesis has not been submitted in whole or in part of any university or any other institution for another Degree/Diploma.
..............................................
........................
Signature of the candidate
Date
I/We certify that the above statement made by the candidate is true and that this thesis is suitable for submission to the for the purpose of evaluation
...................................................... (Signature)
.................................................. (Signature)
.....................
........................
(Date)
(Date)
Prof. K. Tennakone
Prof. D. A. Tantrigoda
Institute of Fundamental Studies,
Department of Physics,
Hantana Road,
University of Sri Jayawardanapura,
Kandy, Sri Lanka.
Nugegoda, Sri Lanka.
DEDICATION TO MY FATHER MY MOTHER AND MY WIFE
CONTENTS DECLERATION DEDICATION CONTENTS
i
LIST OF FIGURES & TABLES
vi
ACNOWLAGEMENT
xv
ABBRIVATIONS
xvii
ABSTRACT
xix
CHAPTER 1
1-34
Introduction
1
Electron behavior in a crystal structure
1
1.1.1 Formation of covalent bonds
2
1.1.2 Formation of metallic bonds
2
1.1.3 Formation of energy bands in a crystal
3
1.2
Conductors semiconductors and nonconductors
4
1.3
Charge carriers in a semiconductors
5
1.4
Fermi-Dirac distribution function
9
1.4.1 Fermi Energy (WF)
9
1.5
Degeneration of semiconductors
11
1.6
Junctions in a solids
12
1.6.1 Metal-metal junction
13
1.6.2 Metal-semiconductor junctions
14
1.6.3 Semiconductor-semiconductor junctions (p-n junctions)
17
1.1
i
1.7
Crystallinity in semiconductors
18
1.8
Semiconductor-light interaction
19
1.9
Photovoltaic cells
20
1.9.1 Metal-Schottky junction solar cells
21
1.9.2 Semiconductor-electrolyte junction solar cells.
21
Dye sensitized solar cells
23
1.10.1 Electron transport in DS Solid-State Solar Cell
23
1.10.2 Properties of n-TiO2
26
1.10.3 p-type semiconductors.
27
Theories equations related to solar cell devises.
28
1.11.1 Basic equations of device physics
29
1.10
1.11
1.11.2 Application of basic equations to homojunction solar sells.
30
1.11.3 Alternation of the device physics derivations for Dye sensitized systems (heterojunctions)
CHAPTER 2
35-52
Experimental and Characterization Methodology 2.1
31
Methodology
35 35
2.1.1 Preparation of conducting glass to deposition nanocrystaline TiO2 films
35
2.1.2 Preparation of TiO2 colloid for compact TiO2 Films
36
2.1.3 Preparation of CuI powder and CuSCN powder
36
2.1.3.1 Preparation of CuI powder
ii
36
2.1.3.2 Preparation of CuSCN powder
37
2.1.4 CuSCN and CuI deposition technique
37
2.1.4.1 Deposition of CuSCN Films
37
2.1.4.2 Deposition of CuI Films
37
2.1.5 Dyes and Dye Solutions
38
2.1.6 Coating a monolayer of dye on the semiconductor surface
38
2.2
Fabrication of Dye sensitized solid state solar cell
39
2.3
Measurements and calculations
40
2.3.1 IV characteristics
41
2.3.2 Photocurrent action spectra and IPCE
45
2.3.3 Fluorescence
46
2.3.4 Mott-Schottky Plot
47
2.3.5 Dark IV plots (Rectification cuves)
48
CHAPTER 3
53-75
Electron transport in heterojunctions with two dyes
53
3.1
Introduction
53
3.2
Experimental
55
3.3
Results and Discussion
58
3.4
Conclusion
74
CHAPTER 4
76-94
Electron transport in a dye-semiconductor multilayed semiconductor nanostructures
76
iii
4.1
Introduction
76
4.2
Experimental
78
4.3
Results and Discussion
80
4.4
Conclusion
93
CHAPTER 5
95-108
Electron conduction in a nanostructure based on extremely thin absorbat layer of polythiocyanogen.
95
5.1
Introduction
95
5.2
Experimental
96
5.3
Results and Discussion
98
5.4
Conclusion
108
CHAPTER 6
109-120
Strategy to enhancing the electron transport properties of CuSCN
109
6.1
Introduction
109
6.2
Experimental
110
6.3
Results and Discussion
112
6.4
Conclusion
120
CHAPTER 7
121-123
The addressable arias in future
121
iv
REFERENCES
124-131
APPENDICES 1
132-134
v
LIST OF FIGURES and TABLES Figure
Figure 1.1
Page No
Plot of inter atomic energy vs atomic spacing. W0-bonding energy and r0 is the separation of two atoms.
Figure 1.2
1
Illustration of energy band formation with the increment of number of atoms
Figure 1.3
3
Illustration of Energy bands in a metal, semiconductor and an insulator.
Figure 1.4
4
Illustration of the electron-hole mobility (intrinsic conduction) in a semiconductor (a) in a crystal lattice structure (b) in energy band structure.
Figure 1.5
6
Illustration of the Si lattice and energy band diagram doped with (a) Group III (Boron) and (b) Group V (Phosphorus) Atoms.
Figure 1.6
8
Fermi-Dirac function at absolute zero temperature (T=0K) and at higher temperatures (T>0K).
Figure 1.7
Energy
diagrams
of
Degenerated
10 p-type
and
n-type
semiconductor. Figure 1.8
11
Energy band diagrams of two metals (a) before and (b) after contact with each other
Figure 1.9
13
Band diagram of n-type semiconductor-metal junction where (a) Φm < Φs and (b) Φm > Φs.
15
Figure 1.10 Band diagram of p-type semiconductor-metal junction where
vi
(a) Φm > Φs and (b) Φm < Φs.
16
Figure 1.11 Energy band diagram of a p-n junction.
18
Figure 1.12 A schematic diagram of a photo electrochemical solar cell (PEC).
22
Figure 1.13 An energy band diagram of the DSS Solar Cell.
25
Figure 1.14 Illustration of recombination paths in a dye sensitized Solar Cell. (a) Recombination of CB electron with hole in the VB (b) Figure 2.1
Recombination pf CB electron with dye cation
25
Schematic diagram illustrating the cross section of a dye sensitized photovoltaic cell of heterostructure configuration of n-type semiconductor / Dye / p-type semiconductor.
Figure 2.2.
(a) Illustration of schematic diagram of the IV setup and (b) A diagram of a Basic IV setup
Figure 2.3
43
Figure illustrating the angel of the sun for different AM conditions.
Figure 2.5
44
Action spectra and IPCE curve of a DSSC with a double dye system.
Figure 2.6
45
Illustration of excitation of electrons by absorbing photons and emmision of radiation due to diexcitation.
Figure 2.7
47
Mott-Schotky plot of (a) p-type semiconductor (b) n-type semiconductor
Figure 2.8
42
illustration of a typical IV curve with maximum power point marked on it.
Figure 2.4
40
48
Illustration of band bending when the semiconductor
vii
Figure 2.9
electrode is biased different voltages
50
Illustration of a Mott-Schottky setup
51
Figure 2.10 Dark I-V (rectification) characteristic curve Figure 3.1
52
Schematic diagrams showing the construction of the cell TiO2/D1-D2/CuSCN.
Figure 3.2
57
(I) I-V characteristics of (a) TiO2/MV/CuSCN (b) TiO2/TAMV/CuSCN and (II) Photocurrent action spectra of (a) TiO2/MV/CuSCN (b) TiO2/TA-MV/CuSCN
Figure 3.3
59
Schematic diagrams illustrating possible configurations of double-dye solid-state solar cells, (I) homogenously mixed thick layer two dyes (II) a monolayer consisting of two noninteracting dye molecules coupled to n and p-type semiconductors
(III)
dye
layer
consisting
of
two
electronically coupled dye molecules bonded on opposites sides to n and p-type semiconductors (circles indicate two 62
types of dye molecules). Figure 3.4
Schematic energy level diagram indicating the relative positions of conduction bands (CB) and valence bands (VB) of TiO2 and CuSCN and ground and excited levels of the D1 and D2 (a) charge transfer on excitation of D1 (b) charge 64
transfer on excitation of D2. Figure 3.5
Photocurrent action spectrum of (a) TiO2/MC-MV/CuSCN (b)TiO2/MC/CuSCN (c) TiO2/MV/CuSCN
viii
68
Figure 3.6 Photocurrent action spectrum of the cells (a) TiO2/BR/CuSCN
69
(b) TiO2/BR-IR786/CuSCN. Figure 3.7 Structural unit in the double dye system n-type semiconductor/ D1-D2/p-type semiconductor (B = bridge connecting the
71
chromophores). Figure 3.8
Schematic diagram indicating possible intermediate stages of charge injection to n-type (boxes on right) and p- type (boxes on left) semiconductors when dye molecules D1 (circles on left) and D2 (circles on right) are excited (a) excitation of D1 followed by electron transfer between two dye molecules and subsequent electron injection of n-type material and holes to the p-type material. (Similar steps occur when D2 gets excited) (b) Possible electron transfer schemes when carrier injection to the semiconductor is the initial step.
Figure 3.9
72
Rectification curves of the cells with dyes (a) MV (b) MC (c) MC-MV
Figure 4.1
73
Diagram illustrating the construction of the photovoltaic cell of heterostructure configuration (a) TiO2/D1/CuSCN/D2/ CuSCN (b) TiO2/D1/CuSCN/D2/CuSCN/D3/CuSCN .
Figure 4.2
I-V characteristics of the cells,
[a] TiO2/D1/CuSCN
79 [b]
TiO2/CuSCN/D2/CuSCN [c] TiO2/D1/CuSCNS/D2/CuSCNS, where D1 = Fast Green, D2 = Acridine Yellow. Figure 4.3
82
Photocurrent action spectra of the cells [a] TiO2/D1/CuSCN [b] TiO2/CuSCN/D2/CuSCN
[c] TiO2/D1/CuSCN/D2/CuCNS
ix
83
Figure 4.4
Energy level diagram showing the band structure in TiO2/CuSCN/D/CuSCN, ground (So) and excited (S*) levels of the dye D and the modes of electron-hole transfer when the dye is photo-excited.
Figure 4.5
85
Energy level diagram showing the band structure in TiO2/D1/CuSCN/D2/CuSCN, the positions of the ground (So1, So2) and excited (S*1, S*2) levels of the two dyes (D1, D2). The dyes D1 (FG) and D2 (AY) anchored to TiO2 and CuSCN respectively
Figure 4.6
86
(i) Absorption spectrum of aqueous solutions of (a) D3 = Acridine Yellow (b) D2= Rhodamine 6G (c) D1= Fast Green; (ii) Photocurrent action spectrum of the cell TiO2/D1/CuSCN/D2/CuSCN/D3/ CuSCN.
89
Figuer 4.7 An energy level diagram showing the conduction and valence band edges of TiO2 and CuSCN and the ground (S10,S20,S30) and excited ( S1*,S2*, S3* ) levels of the dyes D1, D2, D3. ( D1= Fast Green , D2 = Rhodamine 6G , D3 = Acridine Yellow ). Figure 4.8
I-V characteristics of the 3-dye cell TiO2 / D1 / CuSCN/D2 / CuSCN/D3 / p-CuSCN.
Figure 5.1
92
Schematic diagram of the Construction of the photovoltaic cell with (SCN)n layer.
Figure 5.2
90
98
FT-IR spectrum of the polythiocyanogen scraped off from a film
deposited
on
conducting
(T=Transmittance).
tin
oxide
glass 99
x
Figure 5.3
The Mott-Schottky plot for a film of polythiocyanogen deposited on conducting glass. Measurement frequency: (a)1.5kHz, (b) 1.0 kHz.
Figure 5.4
101
Absorption spectrum of a polythiocyanogen film and photocurrent action spectrum of the cell TiO2/[SCN]n/CuI
102
Figure 5.5.1 SEM picture of bare conducting tin oxide glass surface
103
Figure 5.5.2 SEM picture of polythiocyanogen deposited on conducting tin oxide glass surface.
103
Figure 5.5.3 SEM picture of bare nanocrystalline TiO2 film Figure 5.5.4 SEM
picture
of
polythiocyanogen
deposited
104 on
a
nanocrystalline film of TiO2. Figure 5.6
104
I-V characteristics of the cell TiO2/[SCN]n/CuI measured at 1000 Wm-2, 1.5AM illumination.
Figure 5.7
105
Schematic energy level diagram of conduction and valance band positions of TiO2, [SCN]n and CuI.
106
Figure 5.8
Dark IV (rectification curve) for the cell TiO2/[SCN]n/CuI
107
Figure 6.1
Graph of Time variation of the sheet resistance of the CuSCN films when they are inserted into a N2 atmosphere containing (a) Cl2 (b) Br2 (c) I2
113
Figure 6.2 Graph of Change in sheet resistance of the CuSCN films doped with (a) I2 (b) Br2 (c) Cl2 when kept in a N2 atmosphere Figure 6.3
115
Fluorescence spectrum of (a) CuSCN film on glass (b) CuSCN film on glass exposed to Cl2. Inset: Energy level representation of the (SCN)2 impurity level in CuSCN.
xi
116
Figure 6.4
I-V characteristic of the cell TiO2/Ru-dye/ CuSCN (a) before exposure to (SCN)2 solution in CCl4. (b) after exposure to (SCN)2 solution in CCl4.
Figure 6.5
118
Mott-Schottky plots of a CuSCN film on CTO glass (a) before exposure to (SCN)2 solution (b) after exposure to (SCN)2 solution.
119
Schemes
Page No
Scheme 3.1 Anchoring of the sodium salt of trihydoxybenzoic acid to TiO2 and attachment of methyl violet cation by replacement of Na+
59
Scheme 3.2 The mode of anchoring of mercurochrome to TiO2 and attachment of methyl violet cation by replacement of Na+.
65
Scheme 3.3 The mode of anchoring of bromopyrogallol red to TiO2 and attachment of IR 786 cation by replacement of Na+.
Tables
Table 3.1
66
Page No
Short-circuit photocurrent (Isc), open-circuit voltage (Voc), efficiency (), fill-factor (FF) and peak (620 nm) incident photon to photocurrent conversion efficiency (IPCE) of TiO2/MV/CuSCN and TiO2/TA-MV/CuSCN
xii
60
Table 3.2
Incident photon to photocurrent conversion efficiencies (IPCEs) of the cells (1) TiO2/ MC-MV/ CuSCN (2) TiO2/ MC/ CuSCN (3) TiO2/ MV/ CuSCN at peak absorption
67
wavelengths of the two dyes. Table 3.3
Incident photon to photocurrent conversion efficiencies (IPCEs) of the cells (1) TiO2/ BR-IR786/ CuSCN (2) TiO2/ BR/ CuSCN (3) TiO2/ IR786/ CuSCN at peak absorption
68
wavelengths of the two dyes. Table 3.4
Open-Circuit Voltage (Voc), Short-Circuit Photocurrant (Isc), Fill factor (FF) and Energy conversion efficiencie () of the cells (1) TiO2/ MC- MV/ CuSCN (2) TiO2/ MC/ CuSCN (3) TiO2/ MV/ CuSCN (4) TiO2/ BR- IR786/ CuSCN (5) TiO2/
70
BR/ CuSCN (6) TiO2/ IR786/ CuSCN. Table 4.1
Short circuit photocurrent (Isc), open-circuit voltage (Voc), Fill Factor (FF) and efficiency () of photovoltaic cells of
81
different configurations Table 4.2
Incident photon to photocurrent conversion efficiencies (IPCEs) of different heterostructure configurations at the peak absorption wavelengths of the two dyes D1 (FG, 650nm) and
83
D2 (AY, 470 nm). Table 4.3
Short-circuit photocurrents (Isc), open-circuit voltages (Voc) and efficiencies of dye-sensitized solid-state photovoltaic cells of different configurations (D1= Fast Green, D2 = Rhodamine 6G, D3 = Acridine Yellow).
xiii
93
Table 6.1
The short-circuit photocurrent (Isc), open-circuit voltage (Voc), efficiency (η), and fill factor (FF) of the cells TiO2/Dye/CuSCN before and after SCN doping of CuSCN film.
xiv
118
ACKNOWLEDGMENTS
I wish express my most sincere thanks and gratitude to my supervisor, Prof. K. Tennakone; the project leader of the Condensed Matter Physics Project and the Director of Institute of Fundamental Studies, Kandy, for his guidance, valuable advice, encouragement and emotional support given to me throughout my research period.
I wish to convey my appreciation and thanks to my supervisor Prof. D. A. Tantrigoda, Professor of Physics, Department of Physics, University of Sri Jayawardanapura, for the guidance and advice given to me during the period of the study.
I whish to convey my grateful thanks to Dr. V. P. S Perera, Visiting Scientist, Institute of Fundamental Studies, Kandy and Senior Lecturer, Open University of Sri Lanka, Nawala, and he is also my colleague in early period of my work; for his guidance and encouragement given to me. Also I wish to thank Dr. Perera especially for proof reading and valuable advice.
I whish to express my appreciations to Dr. P. M. Sirimanna, Project Leader of the Nano Science Project, Institute of Fundamental Studies, Kandy, for his help and kind advise given to me in the latter period. My grateful thanks are also due to Dr. G. R. A. Kumara, and Dr. I. R. M. Kottegoda for helping me to obtain the SEM pictures and FTIR measurements of my samples.
xv
I am also very much grateful to my colleagues Ms. M. K. I. Senevirathna and Mr. E. V. A. Premalal, Research Assistants of the Condensed Matter Physics project, Institute of Fundamental Studies, Kandy, for their continues support paid on me in writing the thesis. Also I wish to convey my thanks to former research assistants Mr. P. V. V. Jayaweera and Ms. K. M. P. Bandaranayaka for the support given to me in the initial period of my work.
I would like to convey my thanks Mr. W.G. Jayasekara, the laboratory technician of the condensed matter physics project for the help given to me in varies ways in completion of this thesis. Also I wish to thank all the other research and nonresearch staff members of the I. F. S. for there intentional or unintentional support given to me in completion of my work.
xvi
ABBREVIATIONS
AY
- Acradin Yellow
AM
-Atmospheric mass
BR
-Bromopyrogaloll Red Dye
CB
-Conduction Band
Cu
-Copper
DS
-Dye Sensitized
DSPEC
-Dye sensitized photo electro chemical solar cell
DSSSC
-Dye sensitized solid state cell
D
-Diffusion coefficient / Dye molecule
Di
-Ground state or ground state energy level dye molecules of the ith dye layer
Di*
-Exited state or exited state energy level dye molecules of the ith dye layer.
e
-electron
FF
-Fill Factor
FG
-Fast Green
Ge
-Germanium
h
-hole
i.e.
-That is
K.E
-Kinetic energy
CuI
-Copper Iodide.
CuSCN
-Copper thiocyanate
xvii
CuSCN
-Thick layer of p-type semiconductor Copper thiocyanate
CuSCN
-Thin layer of p-type semiconductor Copper thiocyanate
MC
-Mercurochrome dye
MV
-Methyl Violet dye
P.E
-Potential energy
S
-Sulpher
SC
-Solar Cell
SCE
-Standard Calomel Electrode
CSN-
-thiosyanate ion
S0
-Ground State Energy
S*
-Exited State Energy
Si
-Silicon
TiO2
-Titanium Dioxide
recombination time
VB
-Valence Band
W
-work
wt
-weight
xviii
Investigation of electron transport in nanostructured semiconductor heterojunctions by using dye-sensitized solid-state solar cells
Pitigala Kankanamage Don Duleepa Padmal Pitigala
ABSTRACT
In this study an attempt has been made to understand the electron transport phenomenon in nanostructure heterojunctions with a view to solve the problems related to recombination and the narrow spectral response of dye sensitized solid state solar cells.
It has been reported in literature that attempts to broaden the spectral response by using multiple dyes have resulting in decreasing photocurrent due to concentration quenching and sub monolayer chelation. However it was revealed that if the two dyes are bonding ionically together with each other, it gives better spectral response and a higher photocurrent. Two such double dye systems are discussed in this work. The Mercurochrome-Methyl violet system shows a spectral response from about 500nm to 650nm, with a 4.6 mAcm-2 photocurrent density. The Bromopyrogallol red-IR786 system shows a spectral response extended to infrared region in addition to the increase in the photocurrent. The rectification characteristic curves of these systems also show suppression of the recombination too.
xix
It was also investigated the possibility of using dye-semiconductor multilayers for this purpose and found to yield reasonably acceptable results. It was found that the two dyes Fast green and Acridine yellow when used in the multistructure give an efficiency of 1.67% which is significantly higher than there individual efficiencies. The problem arises when the system extended to more than two dyes is also studied.
Application of a barrier to recombination too is an effective methodology to enhance the performance of a solar cell. The polymer polythiocyanogen is found highly stable and resistant to heat and chemical action. A barrier for recombination is constructed by depositing polythiocyanogen in the heterojunction to study the performance. The polymer polythiocyanogen also acts as the sensitizer of the solar cell.
Conductivity of the p-type semiconductor in the solid state dye sensitized solar cell is also important to its performance. Copper (I) thiocyanate is an important p-type semiconductor satisfying the high band-gap requirements of the above solar cells. However the conductivity of this material is not sufficiently high. The conductivity of solid CuSCN was altered by exposing it to halogen gases and SCN- ions in CCl4. The latter method is found more suitable for doping of CuSCN films in the heterojunction of dye-sensitized solid-state solar cells. A photocurrent of 9.0 mAcm-2 was achieved by doping CuSCN with SCNֿ ions and it is more than 300% increase in the photocurrent.
xx
CHAPTER 1 Introduction
1.1
Electron behavior in a crystal structure
Crystal structures are formed by combining and bonding large number of atoms with one another. So in a crystal structure, atoms are held in a fixed position of the crystal due to inter molecular attractive and repulsive forces. Atoms or molecules occupy fixed positions in the crystal at the equilibrium of those two forces. The energy due to attraction at the equilibrium is known as the bonding energy. The distance between two nuclei at the equilibrium is the atomic spacing of the crystal. To decrease or to increase the distance between two neighbor atoms it needs energy. Mainly there are two types of
Energy (W)
bonding energies. They are the covalent bonds and the metallic bonds.
r0 Distance (r) W0
Figure 1.1
Plot of inter atomic energy vs atomic spacing. W0-bonding energy and r0 is the separation of two atoms.
1
1.1.1
Formation of covalent bonds
Let us analyze a Hydrogen atom, the simplest atom of them all. Hydrogen atom can reach a stable state by acquiring an electron from another Hydrogen atom to form Helium (He) electronic configuration. Then both the electrons can associate with any one of the atoms at a particular time. Screening due to the electrostatic repulsion of the two electrons make them stay in between the two atoms making the attractive forces more effective. This attractive force is known as the covalent bonding.
1.1.2
Formation of metallic bonds
Unlike Hydrogen other elements in group 1A or 1B has to share seven (7) electrons to achieve the inert gas configuration. Releasing an electron to attain a stable electronic configuration is energetically favorable for them than sharing seven electrons of neighboring atoms. Therefore atoms in the above mentioned groups are releasing an electron to the electron cloud so that the electrons are sheared among all the atoms. Due to the charge in the electron cloud the atoms are screened by forming bonds with nuclei. Metallic bond is flexible so it makes the metals ductile and these freely moving electrons ensure the electrical and thermal conducting properties of the metal. These bonds are also occurring in some elements of Group II and Group III. Those elements show the metallic properties.
2
1.1.3
Formation of energy bands in a crystal
If we consider an isolated atom, it has discrete energy levels as shown in the Figure 1.2. If two such atoms bonded together, electrons of the two atoms cannot occupy the same energy level according to Pauli’s exclusion principle. So what will happen is that each energy level will shift apart from each other. If another atom comes it shifts again, like vice if large number of atoms bonding together to make a crystal structure, the energy levels shifts further and the difference of energy between two shifted levels become very small (narrow) and it can be considered as continues (Fig 1.2). These energy levels form the allowed bands in the crystal structure. The space in between those allowed bands are called forbidden bands.
E=0
Energy
Allowed Band
Forbidden Band
1 Atom
Figure 1.2
2 Atoms
n Atoms
N Atoms (N >> n)
Illustration of energy band formation with the increment of number of atoms
3
1.2
Conductors semiconductors and nonconductors
The conductors, semiconductor and nonconductors (insulators) are distinguished by their band gaps. That is the energy gap between the upper level of valence band and the lower level of the conduction band (Figure 1.3). Identifying a conductor from an insulator or from a semiconductor is easy via the band gap. But distinguishing a semiconductor from an insulator is bit tricky. Usually if the band gap is less than 1eV it is considered as a semiconductor and above is considered as insulators. But semiconductor compounds such as TiO2 and SnO2 has a band gap around 3eV and still they are considered as high band gap semiconductors. This is the reason why an insulator and a semiconductor is hard to distinguish there is no exact boundary between there band gaps. Figure 1.3 illustrates the relative band gap of a conductor (metal) semiconductor and an insulator.
Metal or Conductor
Conduction Band Valence Band
Insulator or Nonconductor
Semiconductor
Energy levels Occupied by Electrons Energy Gap Energy levels Unoccupied by Electrons
Figure 1.3
Illustration of Energy bands in a metal, semiconductor and an insulator.
4
1.3
Charge carriers in a semiconductor
Atoms vibrate around its lattice point in the crystal structure due to absorbance of thermal energy. Part of this thermal energy can be absorbed by the electrons in the covalent bonds and they can become free from their bonds. If we take a look at the energy band model of this phenomena, the electrons in the valence band absorbs thermal energy and jump in to the conduction band. When the electron got free from the covalent bond (or valance band) an electron vacancy generates in the bond. This electron vacancy is called a hole; it has same value of coulomb charge as an electron but with opposite sign (positive charge).
The freed electrons can mobile freely in the conduction band under the influence of an electric field. When the electron move away from a hole, the hole will attract another electron from a near by covenant bond. This will create a hole in that covalent bond. So in the semiconductor, the holes moved to some other place in the crystal lattice. This way both the holes and the electrons contribute to the flow of current. That is both the conduction band and the valance band involved in the electrical conduction. This phenomenon is distinctive for semiconductors.
Conduction in a semiconductor occurs via electron-hole pairs, where each electron-hole pair pose a charge equivalent to two electrons, flow in the external circuit. This process within the semiconductor is known as intrinsic conduction because it is a property of pure semiconducting material. Figure 1.4 illustrates the electron-hole mobility in a semiconductor material (Si).
5
Electric Field Si
Si Conduction band
Si
electron
Si
hole Valence band Si
Figure 1.4
Si
Illustration of the electron-hole mobility (intrinsic conduction) in a semiconductor (a) in a crystal lattice structure (b) in energy band structure.
Conductivity of intrinsic semiconductors is very poor. So, impurities of Group III or Group V elements that are almost the same atomic size as the semiconductor (Group IV) material were added purposely to improve the conductivity. The addition of impurities is known as doping. These types of semiconductors are known as extrinsic semiconductors. Group III elements such as arsenic, antimony, phosphorus and Group IV elements like indium, gallium and boron are the most suitable elements for doping Si and Ge.
6
When doping is done with Group III elements each impurity atom will create an electron deficiency in the crystal lattice of the semiconductor. That means it will create holes (Figure 1.5 (a)) which will accept electrons from valence band; so Group III impurities are also known as acceptors. These acceptors will create allowed energy bands in the forbidden band just above the valance band known as acceptor level. When the acceptor levels are filled with electrons it leaves holes in the valence band. So the electrical conduction occurs through the effective movement of holes, so that semiconductor material is known as p-type semiconductor. Majority carriers of a p-type semiconductor are holes and minority carries are electrons. Doping with Group V impurity atoms gives excess electrons to the crystal structure. This means each impurity atom will donate an electron to the conduction band (Figure1.5 (b)). So Group V impurities are known as donors. Donor impurities create allowed energy bands in the forbidden region just below the conduction band. So the electrons in the donor atoms are free to move in to the conduction band. With donor impurities the electrical conduction occurs via electrons. These types of semiconductors are known as n-type semiconductors. Majority carriers of an n-type semiconductor are electrons and minority carries are holes. In addition to the pure element semiconductors, many alloys and compounds are showing semiconductor properties. These may be formed from semiconductor or non-semiconducting elements. Copper oxide is the oldest semiconductor in commercial use. One of the most important classes of compounds is the III-V compounds, formed by elements from Groups III B and V B of the Periodic Table. Typical example is GaAs. Which is one of the most versatile and useful semiconductor.
7
Si
Si
Si
Si
B
Si
Conduction band
Acceptor level Valence band
Si
Si
Si
Electron
(a)
Hole
Si
Si
Si
Si
P
Si
Conduction band Donor level
Si
Si
Si Valence band
(b)
Figure 1.5
Illustration of the Si lattice and energy band diagram doped with (a) Group III (Boron) and (b) Group V (Phosphorus) Atoms.
8
1.4
Fermi-Dirac distribution function
Electrons and holes are fermions. In a solid structure, they obey the Pauli’s exclusion principle and their energy is governed by the number of available energy levels in the crystal structure. Fermi-Dirac distribution function is given as, PF W
1 W WF 1 exp kT
1-1
Here PF(W) is the probability of electron occupied in the energy level W, and WF is the Fermi energy level.
1.4.1
Fermi Energy (WF)
Fermi energy (WF) is a purely mathematical quantity. WF is used as a reference to compare other energy levels. If we substitute W in equation 1.10 with WF (i.e calculate the probability of electron be occupied the Fermi energy) we get, PF(WF)
=
½
That implies that the electrons equally like to occupy energy levels above the Fermi level just as below.
If we consider the electron distribution at absolute zero temperature (i.e. T = 0 K) the results become as follows, At W < WF
PF(WF)/T=0
=
1
At W > WF
PF(WF)/T=0
=
0
9
In the Figure 1.6, it shows how the electron distribution changes with the increasing temperature from absolute zero.
Mathematically it can be proven that the Fermi level lies in the middle of the forbidden region (band gap) in an intrinsic semiconductor under the condition that effective mass of both the electron and the hole are same (i.e., me = mh).
PF (W)
T=0K
1
½ T>0K
WF
Figure 1.6
W
Fermi-Dirac function at absolute zero temperature (T=0K) and at higher temperatures (T>0K).
The Fermi level positions of the extrinsic semiconductors are dependent on the density of the impurities and the type of the semiconductor (p-type or n-type). In a p-type
10
semiconductor the Fermi level moves closer to the top of the valance band and in an ntype semiconductor it moves towards the bottom of the conduction band.
1.5
Degenerate semiconductors
As mentioned in the above section the Fermi level of the semiconductors depend on the doping density. If the doping density increases further and further there is a time where the Fermi level move in to the conduction band or to the valance band of the of the ntype or p-type semiconductors respectively. This means that the semiconductor has so many majority carriers (electrons in n-type and holes in p-type semiconductor) and it becomes almost a conductor. Then it is called that the semiconductor has been degenerated in to a metal.
CB
CB
WF Wg VB
VB
WF
p-type
n-type
Figure 1.7 Energy diagrams of Degenerated n-type and p-type semiconductor.
11
1.6
Junctions in solids
Metals and semiconductors are generally used to make junctions. They can be used to form several types of junctions such as metal-metal junction, metal-semiconductor junction and semiconductor-semiconductor junction.
1.6.1
Metal-metal junction
If we take two different metals they have different energy levels. The Fermi energy levels of them are at two different places. So the work functions (minimum energy needed to given, to allow an electron to escape (Φ)) differ from each other. But when they are contacted with each other the Fermi levels equivalized by transferring electrons from one material (metal with low work function) to the other. This will make one surface positively charged while the other surface negativity charged. This will continue till the Fermi levels of both the metals coincide. When the Fermi level of the both materials are in the same level the probability of the electrons to be at both the materials become equal. The electron exchange between the two materials comes to equilibrium. Figure 1.8 shows an energy band diagram of two metals before and after been contact.
12
CB
CB
WF
WF
VB
VB (a)
CB WF VB (b)
Figure 1.8
Energy band diagrams of two metals (a) before and (b) after contact with each other
1.6.2
Metal-semiconductor junctions
When analyzing the metal-semiconductor junctions there are several combinations to be considered, because these junctions have different properties with the type of the extrinsic semiconductor and the work function of the two materials (metal and the semiconductor). Let us take work function of metal and semiconductor as Φm and Φs respectively.
13
First lets consider the metal junction with a n-type semiconductor where, Φm < Φs. As usual once the contacts were made, electrons will flow from the metal to the semiconductor leaving positive charges on the metal and accumulating negative charge on the semiconductor. If the Φs - Φm is small (this is the usual condition in most of the occasions) electrons can flow easily. This type of contact between metals and semiconductors are called an ohmic contact.
When Φm > Φs, electrons flow from the semiconductor to the metal until the Fermi levels coincide. Electrons accumulate on the metal, and the semiconductor will be positively charged due to uncompensated donor atoms. Positive charge on the semiconductor will be distributed for some distance through the semiconductor. This region is called the depletion layer. There are no free charges present in the depletion layer so the resistance of the depletion layer is very high. The energy difference (Φm-Φs) is called the diffuse potential and electrons need thermal energy to over come this barrier in the case of flowing into the metal.
This junction acts as a rectifier. If the semiconductor is negatively biased with respect to the metal the diffuse potential will be reduced and electrons will flow to the metal. This will also reduce the depletion layer. If the junction is reversed biased the diffuse potential will increase so the depletion layer preventing the electron flows. This type of barrier junction diodes is known as Schottky-barrier diodes.
If the n-type semiconductor is heavily doped and degenerates the junction acts as a metal-metal junction forming an ohmic contact between them.
14
Depletion region _ _ _ _
Φm
ΦS Donor level
+ + + +
+
+
+
+
WF
(a)
Figure 1.9
_ _ _ _
(b)
Band diagram of n-type semiconductor-metal junction where (a) Φm < Φs and (b) Φm > Φs.
For metal-p-type semiconductor junctions where Φm > Φs, electrons starts to flow from semiconductor to metal resulting negative charge and the semiconductor charging positively due to holes. But unlike the previous case there is no depletion layer forming because in p-type semiconductors holes are freely collected (holes are the major carriers) as they do not have any effect on acceptor atoms. Therefore holes are free to flow from semiconductor to metal and the contact is ohmic.
15
When Φm < Φs in a p-type semiconductor-metal junction, the electrons initially flowing in to the semiconductor and those electrons were captured by the acceptor atoms in the p-type semiconductor surface forming a depletion layer. Now holes are not free to move due to depletion layer and the barrier formed on the interface. This junction also acts as a Schottky-barrier diode, where a positive voltage on the semiconductor will allow the holes to flow in to the metal and the reverse bias will prevent the flow of holes except the leakage currant of hole from metal to semiconductor.
Depletion region
Φm
_ _ _ _
ΦS WF + + +
Acceptor level
(a)
Figure 1.10
+ + + +
---
(b)
Band diagram of p-type semiconductor-metal junction where (a) Φm > Φs and (b) Φm < Φs.
16
1.6.3 Semiconductor-semiconductor junctions (p-n junctions)
P-n junctions are very important because in semiconductor electronic devises such as diodes and transistor, the basic eliminate in the structure is the p-n junction. Since the ptype and n-type semiconductors were made by adding impurities to Si or Ge (in industry) the base material is common (Si or Ge) and is called a homojunction. In this case the work function (Φ) depends on the type and the number of impurities in the base material. Generally of this the work function of a p-type semiconductor (Φp) is always grater than the work function of n-type semiconductor (Φn).
In a p-n junction if Φp> Φn, electrons will flow from n-type material to p-type material resulting negative charge on p-type material and positive charge on n-type material. So there are uncompensated donors and acceptors in n-type and p-type material respectively. These uncompensated impurities will build a depletion layers on both the semiconductors. This will prevent the free flow of electrons and holes through the junction. Biasing the junction with positive voltage applied to the p-type and the negative end to the n-type will reduce the number of uncompensated impurities in both the sides making the depletion layer narrow. This will allow electrons and holes to flow through the junction. Biasing on the other way round will increase the depletion layer by increasing the number of uncompensated impurities in both the sides. This will block the electron and hole flow. The p-n junction is acting as a diode and another important thing is that there are two current flows, one by the electrons and the other by the holes.
17
Depletion region
Φp Φn +
+ +
---
Figure 1.11
1.7
WF
Energy band diagram of a p-n junction.
Crystallinity in semiconductors
Semiconductor materials can be categories in to three groups. Single crystal material structure is an ordered structure of atomic locations. Polycrystalline are single crystal structures of small domains oriented in random manor and separated by grain boundaries. The amorphous are the materials with no long range order in atomic locations. Silicon (Si) can be found in all the three categories. Most of the compound semiconductors (CdS, Cu2S, TiO2, CuInSe, ect...) are in the polycrystalline and amorphous form.
18
1.8
Semiconductor-light interaction
If a semiconductor is illuminated with light; part will be reflected, while a part is absorbed by the semiconductor. Some part will also be transmitted through the semiconductor. The photons that absorbed by the semiconductor will give their energy to the crystal lattice and some of the electrons will jump into the conduction band of the semiconductor creating holes in the valance band. These electrons and holes are the charge carriers of the photo currant and the photo voltage.
The electrons that adsorb photon energy may stay in the same band (CB or VB) without exiting to a higher energy level; then they will loose their energy due to collision with the atoms in the crystal; resulting a temperature increment on the semiconductor. If the photon energy is not sufficient to raise the electrons from VB to CB, some electrons may jump to energy levels that had been created in the forbidden region (due to defects in the crystal structure) which are also known as trap levels and they will jump back to the VB by emitting a photon. Even the electrons from CB can jump in to VB emitting a photon similar to the one that had been absorbed. This phenomenon is known as phosphorescence or photoluminescence.
19
1.9
Photovoltaic cells
Photovoltaic cells can be divided in to two categories. They are the solid- state photo voltaic cells and photo-electro chemical (PEC) solar cells. The most commercialized photovoltaic cells are fabricated using single crystals silicon. Here I will briefly introduce the functions involved in generating of photovoltage and photocurrant in a Si based homojunction solar cell.
Silicon solar cell is a p-n junction, where single crystal silicon is doped with acceptor and donor atoms. Therefore band structure of this cell can be described with a diagram as in Figure 1.11. As mentioned in earlier when the energy of a photon is absorbed by an electron in the VB of the semiconductor it will acquire energy and jump to the CB (if the photon energy is sufficient) creating an electron-hole (e-h) pair. Unless this e-h pairs are separated, they will recombine with in the semiconductor. So an important process occurs in space charge region of the p-n junction where there is no any mobile charges (depletion layer). When an e-h pair generated at the space charge region the e and the h will be separated due to the electric field in the region. The electrons will be drifted towards the CB of n-type semiconductor and the holes will drift to VB of the p-type semiconductor. High concentration of e and h in either side will form an electric field equivalent to application of an electric field to forward bias the p-n junction. This will separate the coincided Fermi levels of the two semiconductors; result is the generation of a photovoltage across the p-n junction solar cell. The photocurrant is generated by the recombination process of the separated e-h pairs through the external circuit.
20
There are another two types of solar cells they are the metal-semiconductor (MetalSchottky junction) solar cells and the semiconductor-electrolyte junction solar cells.
1.9.1
Metal-Schottky junction solar cells
In metal-semiconductor Schottky barrier solar cells or metal oxide–semiconductor (MOS) Schottky barrier solar cell the energy band diagrams will be similar to the one in Figure 1.9(b) or 1.10(a) with the selection of an n or p type semiconductor respectively. The homojunction solar cell the e-h pair will be separated by the electric field at the junction. This area of solar cells is not well understood as there are complexation at the surface, interface and junction that are not clear, especially in MOS type solar cells. Theoretically it has been shown that energy conversion efficiency of MOS type solar cell is 21% by Shewchun et al [1] with Aluminum-Silicon junction.
1.9.2
Semiconductor-electrolyte junction solar cells.
Photovoltaic (PV) effect on semiconductor-electrolyte surface is studied for a long time and it is the most studied area even today. The history of PV effect of semiconductor electrolyte junction is running to 1839. Becquerel [2] first discovered the PV phenomena in an electrochemical system. The studies leading to wet PV solar cells were carried out by Gerischer in 1960s. Photo electro chemical (PEC) cells are studied extensively even today because it is believed to be an alternative for high cost of solid state Si based solar cell.
21
PEC solar cell consist of a high band gap (~1.4eV) semiconductor (n or p) in contact wilt an electrolyte, a counter metal electrode and a suitable redox electrolyte. Energy bands in the semiconductor bends as the Fermi levels of the electrolyte and the semiconductor is coincides. When photons with energy exceeding the band gap energy incident on the semiconductor electron hole pairs are generated at the interface and they are separated by the electric field at space charge region. Minority carriers will drift to the semiconductor-electrolyte interface and the majority caries will drift to the bulk of the semiconductor. The majority carriers will flow through the external circuit to the counter electrode and at the counter electrode the majority carriers will electro chemically reacts with the electrolyte. That is the electrolyte will be oxidized if the majority carriers are holes; when the semiconductor is p- type or it will be reduced if the semiconductor is n-type.
The opposite reaction will occur at the semiconductor
electrolyte interface.
Majority Carriers
Oxidize Reduce
Figure 1.12
A schematic diagram of a photo electrochemical solar cell (PEC).
22
1.10
Dye sensitized solar cells
The main difference between the dye sensitized solar cell and other types of solar cells is in the photon energy absorber. The solar cells that introduced earlier absorb photon energy directly by the semiconductor but in Dye sensitized solar cells the light is absorbed by the dye chromospheres and the electron-hole pair is generated in the dye molecule.
The dye sensitized solar cells (DSSC) can be divided in to two groups as wet cell or photoelectrochemical (PEC) solar cells and solid cell or dye sensitized solid-state solar cell (DSSSC). The electron transfer processes in both the types are almost same; the difference in these two types is the interface. The wet type cell has a n-type or p-type semiconductor/dye-electrolyte interface and the solid state cell has a n-type semiconductor/dye-p-type semiconductor interface. Usually the n-type semiconductor is TiO2 and p-type semiconductors are CuI or CuSCN which have same polycrystalline nature. Today attention has been drowning towards the DSSSC as it seems to provide solutions encountered with the electrochemical solar cell.
1.10.1 Electron transport in DS Solid-State Solar Cell
A schematic diagram illustrating the cross section of a DSSSC with the configuration TiO2 / Dye / CuSCN is in Figure 2.1, in the Chapter 2. An energy band diagram of the cell is illustrated in Figure 1.13 below. Briefly the electron transport process is as follows,
23
When the cell is illuminated the photon energy is absorbed by the dye molecules. The internal energy of the dye molecule arises from ground state (G.S) energy to the exited state (E.S) energy due to the absorbed energy of the photon (The exited dye molecule is denoted as D*). Since the dye molecule has electrons with high energy the electron can be released from the dye molecule. In other words the dye molecules inject electrons to the conduction band (CB) of the n-type semiconductor. Meanwhile a hole will be created at the valance band of the p-type semiconductor by donating an electron to dye cation (D+) (the energy of the dye cation should be lie below the valance band of the ptype semiconductor). The germinated electrons and the holes will recombine traveling through the external path connected accross a load. The whole process can be summarized by the following three reactions.
h + D
D*
D*
D+ + e (CB p-type semiconductor)
D + h (VB n-type semiconductor)
D+ + e (VB n-type semiconductor)
Except the above process there are two other ways of recombining e-h pair. These two path ways are illustrated in the Figure 1.14, suppression of these recombination paths will lead to a higher efficacy in the DSSSC.
24
CB
CB
e/ D+/D*
e
h
VB
D/D+ G.S
VB
Dye n-type semiconductor
Figure 1.13
p-type semiconductor
An energy band diagram of the DSS Solar Cell.
CB e
D+ /D* E.S
D/D+ G.S VB
CB
CB
e
CB D+ /D* E.S
D/D+ G.S
h VB
VB
VB
Dye Dye n-type p-type n-type p-type semiconductor semiconductor semiconductor semiconductor (a) (b)
Figure 1.14
Illustration of recombination paths in a dye sensitized Solar Cell. (a) Recombination of CB electron with hole in the VB (b) Recombination pf CB electron with dye cation
25
1.10.2 Properties of n-TiO2
TiO2 is the most studied and widely used semiconductor material in DSSCs. It is an ntype semiconductor. Thin films of TiO2 are transparent to the visible light and porosity selectivity of particles in wide range sizes are advantages when constructing devises. The other properties of the TiO2 are also discussed in this section.
TiO2 is a polar material. O2- ions in the crystal form the valance bands and the 3d orbitals of Ti4+ ions form the conduction band. The colloidal particles of TiO2 ate crystalline into anatase form due to the surface tension [3], Anatase form that has a band gap of 3.2 eV is more suitable for Dye sensitization than the other two forms of TiO2 which are rutile, and brookite. The random orientation of TiO2 nanocrystalline in anatase form will lead to formation of grain boundaries at the interface. These grain boundaries will reduce the electron mobility through enhanced scattering, small cross sectional areas and space charge potential barriers. Due to the crystallite radius is grater than the exiton radius the quantum confinement effects are not expected in the crystalline structure.
There is a natural tendency to generate oxygen vacancies in the TiO2 crystals, especially in at poor oxygen environments. This will make TiO2 to the form TiO2-x. This oxygen vacancy has electron pairs that can be jump (one or both the electrons) to the 3d orbital in the Ti4+, reducing the Ti4+ to Ti3+. Defect states of Ti3+ increases the electron density in the conduction band as they hang around 2eV - 3eV range from the VB. So the
26
oxygen defects in the TiO2 are doping the intrinsic semiconductor to extrinsic n-type semiconductor.
The large surface area due to the porosity of the TiO2 results a large number of surface states or surface defects. These surface defects also capture or adsorb irons which lead to similar effects as with oxygen vacancies. The surface states also determine the surface properties of the TiO2 which is different from the bulk properties.
UV illumination of the TiO2 surface is also effective. When the TiO2 is illuminated by uV light it generates hydrophilic properties. This will adsorb H2O on to the TiO2 surface which leads to formation of hydroxyl (OH) groups on the TiO2 surface which acts as electron traps. They also develop the conduction properties of TiO2 as they will produce Ti3+ states. Except TiO2, other oxides like ZnO, SnO2 are known to display n-type properties. But the properties of these oxides are not widely studied as TiO2.
1.10.3 p-type semiconductors.
There are few metal oxides, such as Cu2O, NiO which are known to show p-type semiconductor properties. And there are also none oxide materials like CuI, CuSCN and several organic compounds that acts as hole conductors. Out of them most successful material to use in DSSSCs are CuI and CuSCN. The reason is that these two semiconductors can be grown within the pores of the TiO2. The properties of these semiconductors are also not been studied in detailed as TiO2. The defects in the material improve the conductivity in p-type semiconductors too
27
1.11
Theories related to solar cell devices
1.11.1 Basic equations of device physics Here a summary of the mathematical equations that were obeyed by charge carriers in electronic devises are given. Equation 1.2 is the continuity equation which balances the generation and the recombination of charge carriers.
n 1 Jn G R t q
1-2
In the equation 1.2, n(r, t) is the electron density and it is a function of position (x) and time (t). The electron currant density is denoted by Jn. q is the charge of the electron. G and R represent the electron generation rate and the recombination rate respectively. A similar equation can be constructed for holes or any other ionic charge carrier.
The left hand side of the equation 1.2 becomes zero for a solar cell under a steady state illumination with a given bias is applied between the terminals (open circuit mode) or in other words if the cell is said to be at steady state.
The electron currant density (Jn) of a semiconductor is defined in terms of quasi Fermi level (EFn) and the electron density (n) as in the following equation. Here μn denotes the electron mobility. J n n nE Fn
1-3
The electron density (n) is related to the Fermi energy by Fermi-Dirac statistics (equation 1.1) as in equation 1.4 where g(E) is the density of electron acceptors at the energy E and T and k are temperature and the Boltzmann’s constant respectively
28
n g (E)
1 e
E E Fn kT
1
1-4
dE
If the semiconductor is nondegenerate and there are no acceptor states in the energy band gap Boltzmann approximation can be applied and the electron density (n) becomes, n NCe
( E Fn EC )
1-5
kT
Where, Ec is the edge energy of the conduction band and Nc is the effective density of states in the conduction band.
At the above condition there are two contributions to the Jn. They are the contribution from diffusion and the contribution from the drift which can be denoted as follows, J n qD n n q n En
1-6
Where, Dn is the diffusion coefficient and E is the electrostatic field.
When talking about photogeneration (G), in a slab of material with uniform absorption, oriented in a direction normal to the incident light, the photogeneration of charge carriers (G) at a depth of x is given by the following equation, G ( x) g ( E , x)dE
1-7
Where g(E,x) is given by the equation 1-8.
g ( E , x) ( E ) 0 ( E )[1 r ( E )]e ( E ) x
1-8
29
In equation 1-8, (E) is the absorption coefficient of the material at a energy of E, 0(E) is the incident photon flux of the light spectra and r(E) is the surface reflectivity of the material at energy E.
1.11.2 Application of basic equations to homojunction solar sells.
Now let us see how these general equations apply on homojunction solar cells. We have to consider homojunction solar cells before going into dye sensitized solar cell because, the derivation to heterojunctions has to derived from the homojunction equation.
In p-n junction solar cell some simplifications to the above equations can be made. First we can say that the structure have a one dimensional symmetry. Also we can say that most carriers are photogenerated at the doped neutral region but not at the p-n junction. The electric field at the neutral region is zero (i.e. E=0). So the mobility of carriers is due to the diffusion. For the electrons in the p-type material we can write the electron currant density as, J n qD n
dn dt
1-9
Same as the photogeneration most of the recombination are also occur in the neutral region and it is dominated by minority carriers. So for electrons in p-type semiconductor we can say the recombination rate R is proportional to minority carrier density. R can be written as, R
n
1-10
n
Where n is the life time of the electron and it is a constant.
30
Substituting equations1-9 and 1-10 in 1-1 the following continuity equation can be derived for the electrons (minority carriers) in the p-type semiconductor at a study state illumination. Dn
d 2n n g ( E , x) 0 dx 2 n
1-11
A similar type of equation can be built for the holes in the n-type region.
Another important parameter in the solar cell is the diffusion length (L) of the minority carrier. It gives an idea about the thickness of the semiconductor substrate (thickness of the neutral region) that allows an effective collection of carriers. Diffusion length of electrons (Ln) in p-type semiconductor is given by
L n D n n
1.11.3
1-12
Alternation of the device physics equations for Dye-sensitized systems
Using the basic idea of device physics on homojunction solar cells, equations for the heterojunctions or dye sensitized solar cells can be derived. The heterojunction solar cell is more complicated due to the complexity of the surfaces and the interfacial structure. Following factors are very important in alternating the equations of homojuntion to a heterojunction. First there are several types of charge carriers in the heterojunction specially if the hole conductor is a electrolyte. Except for electrons there will be positive or negative charged ions too. So the effects of these carries should be included in the equation. Another factor is the geometry of the junction, it is three dimensional. So it will create large local
31
electric fields in the hole conductor. Although Fermi-Dirac statistics (eqn 1.4) is applicable, due to the high density of energy levels at the band gap of the semiconductor the Boltzmann approximation will not be valid. Also the trapping of electrons are expected by these inter band levels (trap levels or traps). The photogenaretion process too is different from a homojonction solar cell. It is done by photon absorption process with the dye and the sensitizer will inject electrons to the semiconductor. Neither the driving force for the mobility of electrons, nor the electron recombination rates (via two recombination paths in Figure 1.14) are known definitely.
By considering the above factors a continuity equation can be derived for a dye sensitized solar cells too. Let us say nc is the density of free electrons with the energy above the value of the lowest level of the conduction band. A term has to be added to the continuity equation to represent the electrons that will be captured and released by the traps. Then the equation becomes, c n c 1 df J n G R n c g ( E ) dE t dt q Ev
E
1-13
In the equation g(E) is the density of trap levels in the band gap at an energy E. Ec and Ev are the energies at lover level of the conduction band and the highest level of the valance band. The electron distribution at none equilibrium state is given as f(E). The term representing the electron interaction with the traps can be expanded to the following equation, which represents by the two terms trapping and detrapping. nc
EC
EC
EV
EV
( E ) g ( E )[1 f ( E )]dE nc ( E )e
( E E Fn ) kT
32
g ( E ) f ( E )dE
1-14
The rate of trapping by a vacant trap at energy E is denoted by (E). is proportional to the cross section of the electron capture () and the thermal velocity of the electron (). It can be given as,
1-15
The rate of electron detrapping is given by,
e
( E E Fn )
1-16
kT
The capacitance of the nanocrystal porous of the film requires a coulombic energy to charge the nanoparticle, so if a bias voltage V is applied the energy will be divided as follows. qV E Fn E e
1-17
Where, ΔEFn and ΔEe are the shift in Fermi energy and the shift in electrostatic potential energy respectively. For a spherical nanoparticle with radius r, the shift in electrostatic potential energy is given as, E e
q 2 n1 8 0 r
1-18
The number of electrons already in the particle is denoted as n1, ε is the effective dielectric constant. This nano capacitance is effective in both the electron transport and the photovoltage. The coulomb forces will suppress the accumulation of charges in the nanoparticles. All the electrons also will not contribute to the external voltage.
Although the dye sensitized structure is complex in general, with same the system can be simplified. One assumption is that the number of photogenerated electrons is equal to the number of absorbed photons. This assumption is dependent on the sensitizer (dye). Some dyes like RuN3 do have a near unity incident photon to photocurrent conversion
33
efficiency (IPCE). It also can assume a one dimensional surface geometry comparing with the size of the device. The carriers like charged ions can be neglected with compared to the number of mobile electrons. The electric fields can be neglected specially in the electrolyte, because the electrolyte consists with large number of ions which will reduce the effect of the electric field. All the electrons generated are collected at the TiO2-SnO2 interface due to the higher electron affinity of indium doped SnO2 on the conducting glass. The photogenerated carrier density on the plane is zero.
These assumptions simplify the complex situation of electron diffusion in an environment without an electric field. So the continuity equation becomes the form of, n n D n (n( x)) G R x t x
1-19
34
CHAPTER 2 Experimental and Characterization Methodology
2.1
Methodology
In this chapter the methodologies commonly used in the experiments are described. These methodologies were applied in the preparation of the nanocrystalline semiconductor films and Dye sensitized solar cells. Theoretical backgrounds of some characterizing methodologies are also briefly discussed in this chapter.
2.1.1
Preparation of conducting glass to deposition nanocrystalline TiO2 films
Conducting Tin Oxide (CTO) glass plates were purchased from Solaronics, Switzerland (sheet resistance 15 ohm/□). They were cut in to small peaces with dimensions 0.5cm × 2.0cm and were cleaned by warming in a solution of KOH followed by rinsed with water, and again washed with Dilute HNO3 and rinsing with water. Finally the glass plates were boiled in a solution of 50% propan-2-ol, and allowed to dry in air avoiding contamination with grease.
35
2.1.2
Preparation of TiO2 colloid for compact TiO2 Films
TiO2 colloid was prepared as described previously in literature [4]. In brief the process is as follows; the given amounts of the chemicals below are mixed and stirred vigorously while adding 5 ml of H2O drop vice. The chemical ingredients are, 1.
Titanium isopropoxide
5.0 ml
2.
acetic acid
5.5 ml
3.
Propan-2-ol
20.0 ml
These chemicals were added into a flask in the above order and stirred them for overnight using a magnetic stirrer. The colloidal TiO2 prepared by this method was used to grow the nano porous compact TiO2 Film on the CTO glass. The deposition technique is described in latter in section (2.2)
2.1.3
2.1.3.1
Preparation of CuI powder and CuSCN powder
Preparation of CuI powder
Copper (I) iododide (CuI) powder was prepared as in literature [5] by mixing aqueous solutions of CuSO4 and KI in 1:1 molar ratio under reducing conditions (i.e., in the presence of H2SO4 and Na2SO3). The white precipitate of CuI is formed and it was washed thoroughly with distilled water and dried in a vacuum oven. Commercially available CuI powder purchased from Fluka chemicals was also used.
36
2.1.3.2
Preparation of CuSCN powder
CuSCN powder was prepared by mixing aqueous solutions of CuSO4 and KCNS in 1:1 molar ratio under reducing conditions (i.e., in the presence of Na2SO3) similar to preparation process of CuI. The resulting white precipitate of CuSCN was washed thoroughly with of distilled water and dried in a vacuum oven.
2.1.4
CuSCN and CuI déposition technique
2.1.4.1 Deposition of CuSCN Films
A solution of CuSCN in propyl sulfide [6] was used in fabricateing the p-type layer of the DSSSC. The CuSCN solution is prepared by dissolving 50mg of the CuSCN powder in 8ml of propyl sulfide. CuSCN was deposited on the dye coated TiO2 film evenly by spreading a measured amount of the CuSCN solution in propyl sulfide, while heating the Dyed TiO2 plate to ~90oC on a hot plate and allowing it to dry at the same temperature for few minutes. Film is then dried in a warm air current to expel excess propyl sulfide.
2.1.4.2 Deposition of CuI Films
When preparing the CuI solution for the p-type layer of the DSSSC, 1.2g of CuI powder was dissolved in 50ml of acetonitryl (methylcyanide) [4]. CuI was deposited on the dye
37
coated TiO2 film by evenly spreading a measured amount of solution on dye coated TiO2 plate placed on a hot plate.
2.1.5
Dyes and Dye Solutions
Commercially
available
dyes
like
cis
dithiocyanato-bis
(2,2’bipyridyl
4,4’
dicarboxylate) Ru(II) (N3), Fast Green (FG), Rhodamine 6 G (RG) and Acridine Yellow (AY), mecurocrome (MC) and methyl violet (MV) were used most of the time as the sensitizers of the DSSCs in the experiments. They are used without further purification. But some times the chloride forms of the dyes such as AY, MV and RG were converted into the respective thiocyanates for better chelation by the following process:
The solid dye in the chloride form was boiled with an excess saturated aqueous solution of KSCN; the thiocyanate form of the dye is insoluble in water. Solid suspention was separated by centrifugation, and rinsed with water. A procedure is repeated to complete the double decomposition reaction, which precipitates the less soluble thiocyanate of the dye. Thiocynate dyes are further purified by recrystalliation from an alcoholic solution.
2.1.6
Coating a monolayer of dye on the semiconductor surface
The dyes were adsorbed on TiO2 films by keeping the plate immersed in an alcoholic or aqueous solution of the dye with a suitable molar concentration for desired time. Then the dye adsorbed films were rinsed to remove the excess dye molecules that were not properly chelated to TiO2 surface. Some instances a layer of dye is painted on the
38
semiconductor surface to fabricate a thick dye layers (to fabricate more than a monolayer of dye).
2.2.
Fabrication of Dye sensitized solid-state solar cell
Nanocrystalline compact TiO2 films were used to construct the dye-sensitized solid-state solar cells. TiO2 was deposited on CTO glass plates by the following method. In Brief, the procedure involves painting of a colloidal solution of TiO2 which was prepared as in the section 2.1.2, on a CTO glass surface, placed on a hot plate. CTO glass has to be preheated to ~150°C before applying the TiO2 colloidal solution. Then this TiO2 painted glass plate was sintering at 450°C for 10 minutes in a Furness or an oven. The baked glass plats were taken out from the oven and let them cool for few minutes, and then rube off the loose crust with cotton wool, TiO2 colloidal was painted again on the plate and sintered. This process is repeated for about 15-20 times until a film of ~8-10m is formed. Finally the film is sintered for an half an hour. After cooling the TiO2 fabricated CTO glass, it is coated with a suitable dye.
Then a p-type semiconductor (CuSCN or CuI) was fabricated on the dye coated TiO2 plates. When CuSCN is used as the hole collector a layer of graphite is painted on the top of the CuSCN surface to get good ohmic contact with the counter electrode. A gold plated CTO glass is pressed onto the graphite surface or onto the CuI surface as the back contact.
39
Figure 2.1
Schematic diagram illustrating the cross section of Dye sensitized photovoltaic cell of heterostructure configuration of n-type semiconductor / Dye / p-type semiconductor.
2.3.
Measurements and calculations
Following measurements were conducted to characterize the performances of a DSSC. The most important characteristic measurement in a solar cell (SC) is the I-V characteristics curve. With a I-V curve, we can get a basic idea about the performance of the solar cell by the characteristic factors such as open circuit voltage (Voc), short circuit current (Isc), Fill Factor (FF %) and the efficiency (%). I-V characteristic curves were obtained using a source meter (Keithley) and a solar simulator with intensity of the
40
1000Wm-2. The intensity of the Xenon lamp used was measured with a Pyranometer (Eko).
Another important characteristic measurement is the action spectrum of the solar cell. We can confirm the action of the dye in photocurrant generation process of the DSSC with the action spectrum, and also we can get an idea of the electron injection efficiency at individual wavelengths of the visible spectrum. Intensity of the monochromatic light was recorded and calculated by using a calibrated silicon photodiode.
2.3.1
I-V characteristics
I-V characteristics were obtained using a source meter coupled with a computer under the illumination of a Xenon lamp (1000Wm-2) at 1.5AM. Figure 2.2(a) illustrates a schematic diagram of the I-V setup. Basic idea of the I-V measurement is to characterize the currant and voltage changes with the load resistance (R). This is illustrated in the Figure 2.2(b).
There are few important factors that can be calculated with the I-V curve, they are, 1. Maximum power point (Pmax) 2. Fill Factor (FF%) 3. Efficiency (%)
41
(a)
(b)
Figure 2.2.
(a) Illustration of schematic diagram of the I-V setup and (b) A diagram
of a Basic I-V setup
42
Figure 2.3
Illustration of a typical I-V curve with maximum power point marked on it.
Following equation is used to calculate the fill factor (FF) V I FF % max max Voc I sc
2-1
Where Imax and Vmax are the currant and voltage values at the maximum power point. Voc and Isc is the open circuit photovoltage and short circuit photocurrant of the SC (Figure 2.3).
Efficiency (%) of a solar cell is calculated by,
Voc I sc FF %
Incident Light Intensity
%
2-2
In our case the Incident Light Intensity is 1000Wm-2 (or 100mWcm-2).
43
Air Mass (AM) is an important factor effect on the I-V characterization of a solar cell. All the I-V measurements in following experiments are carried out in a 1.5 AM condition. 1.5 AM solar spectrum means the spectrum of the solar irradiation after it passing through 1.5 times the distance as the earth’s atmosphere.
Figure 2.4
Figure illustrating the angel of the sun for different AM conditions.
44
2.3.2
Photocurrent action spectrum and IPCE
Another informative characterization technique of a dye-sensitized solar cell is photocurrent action spectrum and the incident photon to photo currant conversion efficiency (IPCE). With the action spectrum we can observe the currant variation with the wavelength () of the incident light spectrum in a solar cell. And the IPCE gives the number of electrons produced in the external circuit by the dye molecules per 100 photons at a specific wavelength.
No. of electronrs per sec ond IPCE % No. of Incident Photons per sec ond
Figure 2.5
2-3
Action spectra and IPCE curve of a DSSC of a double dye system.
45
2.3.3
Fluorescence
Fluorescence is not directly relevant to a dye sensitized solar cells, but it is a property of the dye and the semiconductor material. Fluorescence is a luminescence processes in a molecule. It can be described as emission of light from an exited state created by a physical, mechanical or a chemical process. The luminescence produce by excitation of a molecule by UV or visible light is termed as photoluminescence, and it is divided in to two groups, 1. fluorescence 2. phosphorescence The two groups are termed considering their life time. In fluorescence the molecule absorb light of a particular wavelength and emit light in a longer wavelength after a short interval but in phosphorescence the lifetime of the exited state is longer.
Some important information can be obtained using the luminescence data of a dye or a semiconductor material. With the wavelength of the fluorescence peak we can get an idea of the band gap of the semiconductor, trap states if present in the band gap of the semiconductor and also the energy of the exited state of some dyes. This technique can be utilized to observe the electron injection ability of the exited dye molecule to the conduction band of the semiconductor and even about the dye aggregations with the help of the absorbance spectrum of the dye.
The fluorescence spectrums of dyes, semiconductor materials and polymers were obtained by the Shimadsu RF-5000 spectrophotometers.
46
Figure 2.6
Illustration of excitation of electrons by absorbing photons and emission of radiation due to diexcitation.
2.3.4 Mott-Schottky Plot
Mott-Schottky plots (graph of 1/C2 vs. V) are used to determine the flat band potential and carrier type of a semiconductor material. The flat band potential is determined by the point where the interpolated 1/C2 vs. V plot intersect the voltage axis, the value of the gradient is proportional to the doping density and the sign of the gradient indicates whether the semiconductor is a p-type or an n- type semiconductor.
47
Figure 2.7
Mott-Schotky plot of (a) p-type semiconductor (b) n-type semiconductor.
Basically the theory involved in Mott-Schottky diagram can be summarized as follows. If a semiconductor in equilibrium with a redox couple, will results a bending of the bands. A potential is applied under potentiostatic control condition between the working electrode and the reference electrode. If the interface is considered as ideally polarizable (i.e., electrons cannot be exchanged with the electrolyte) the zero of potential is that at which the potential in the bulk semiconductor matches that of redox couple and the reference electrode. Concentration of the electrons in the space charge region depends on the potential difference between the working electrode (semiconductor) and the reference electrode.
48
When the supporting electrolyte is in high concentration the Helmholtz double layer (the charged layer at the side of the electrolyte in the working electrode/electrolyte junction) contains with non-adsorbed ions that represent the counter charge that of the interface surface. The thickness of this layer is smaller than the space charge region. The potential drop across the space charge region (Vsc) occurs over a larger distance than the potential drop across the Helmholtz layer (VH), this is because VSC results from ionisation of the acceptors in the solid, whilst VH is due to the ions accumulated a few angstroms away from the surface.
Since, the charge in both regions is equal but with opposite sign, the capacitance of the space charge region is normally negligible in comparison to the Helmholtz capacitance. Under these conditions, VH is constant and any possible change in the applied potential between working and reference electrode will appear in VSC.
At this potential, the surface concentration of holes (in a p-type semiconductor) is equal to the bulk. At more positive potentials than the Vfb, the surface concentration of electrons is decreased creating an accumulation layer of majority carriers and the bending of the bands at the surface to higher energies. At potentials more negative than the Vfb, electrons produce a depletion layer; bending downwards at the surface to lower energies (Figure 2.8) (Vice versa for the n-type semiconductor). This will change the capacitance at the semiconductor electrolyte interface.
49
Figure 2.8
Illustration of band bending when a p-type semiconductor electrode is biased with different voltages.
The Mott-Schotky relation ship is given by the following equation,
1 2 2 C SC eN SC 0 A 2
kT V V fb e
2-4
Where Csc the space charge capacitance, Nsc charges at the depletion layer, Vfb flat band potential, A aria, e charge of an electron, T temperature and bias voltage V.
Mott-Schottky plot setup is a computer coupled impedance meter and a voltmeter. The voltage is measured with respect to the reference electrode (RE). The film is the working electrode (WE) and a Pt rod is used as the counter electrode (CE). Figure 2.9 shows a schematic diagram of a Mott-Schottky plot setup. Mott-Schottky diagram also
50
can be plotted manual by varying the applied voltage using the impedance meter and measuring the capacitance.
Figure 2.9 Illustration of a Mott-Schottky plot setup.
2.3.5
Dark I-V plots (Rectification curves)
Rectification data is very important to determine the performance of a DSSC. A good rectification implies a better suppression of internal recombination. A quantitative analysis of the rectification can be obtained by rectification ratio. Rectification ratio (RR) is the absolute value of the forward bias current verses the reverse bias current at a given voltage. R .R .
I V I V
2-5
51
I (A)
I (+V)
V (V)
-V I (-V)
Figure 2.10
Dark I-V (rectification) characteristic curve
52
+V
CHAPTER 3 Electron transport in heterojunctions with two dyes
3.1
Introduction
Dye-sensitization of semiconductor surfaces continues to be a rich field of study from fundamentals as well as applications points of view [7-10]. Different types of dyesensitized solar cells [4, 11-15] have been demonstrated and their efficiencies depend on fast injection of the carriers to the bands of the semiconductor and the slow back reaction.[16-18] It is known that strong electronic coupling of the dye molecule to the semiconductor surface causes fast injection [17, 19]. It can be found that a good surface chelation of the dye molecules to the semiconductor affords a way of achieving a good electronic coupling compared to the
van der Waals contact of the dye to the
semiconductor surface. The ligands that readily anchor to TiO2 and to other oxide surfaces have been identified [19]. This chapter will introduce a new and effective method to anchor an ionic dye, that is not readily adsorbing onto the TiO2 surface or other oxide surfaces, but can be electrostatically bonded to an ionic molecules of the opposite charge. And these ionic molecules are surface chelated to TiO2 via suitable ligands. And the results show that anchoring of dye by this method results in an efficient sensitization of the solar cell. This effect is demonstrated by constructing dye-sensitized solid-state solar cells using CuSCN as the hole collector. This method can also be extended to adopt a surface chelating ionic dye instant of the ionic absorbate on the TiO2 surface, so that sensitization can occur via both the dyes and enhances the spectral response of the solar cell. This chapter also describes a construction of a modal dye-
53
sensitized solid-state solar cell with two dyes as above, where the energy conversion efficiency of the double dye system is higher than that of the cells based on individual dyes.
The use of double dye systems in dye sensitized solar sells is very effective because the dye-sensitized solar cells have narrow spectral responses which correspond to absorption spectrum of the dye. This is one of the main obstacles to improvement of the efficiency of the dye-sensitized solar cell. Broadening of the spectral response becomes an essential requirement to increase efficiency of the Dye-sensitized solar cells. Another way to broaden the spectral response is the synthesis of dyes with broader spectral response and it has been attempted as a possible strategy.
In conventional semiconductor photovoltaics, the idea of multiple bandgaps has been used in the implementation of broad spectral response and the idea of multiple bandgaps has been successfully utilized to extend the spectral response [20]. Also that someone can say that instant of using a single dye, by using a mixture of dyes will resolve the problem. But unfortunately, straightforward application of dye mixtures invariably leads to concentration quenching or insulation. It can be found that the direct application of dye mixtures, in almost all cases result in lowering of both energy and quantum conversion efficiencies of the SC.
Quenching and insulation can be avoided if the dyes are lightly deposited on the semiconductor surface, because at this situation dye molecules are separated and that the mutual interaction of the dye molecules are avoid. And also in the other hand in
54
order to achieve good light absorption at the cross section of each dye, the thickness of the TiO2 film needs to be increased. However, there exists an upper limit for the film thickness, i.e., the electron diffusion length (D Where D is the diffusion coefficient and the is therecombination time. In order to satisfy the constraint, TiO2 film thickness (t) is less than the electron diffusion length (i.e., t < [D, a fourfold increase in the diffusion coefficient is necessary. This is hard to achieve, because the diffusion coefficient depends on the film material and the film morphology. And attempts to adjust film morphology would also change the roughness factor. And the roughness factor is proportional to film thickness. So the strategy adopted in the model system presented in this chapter is an attempt to circumvent the above problems and achieve a good efficiency in a SC.
3.2
Experimental
Nanocrystalline TiO2 films were deposited on conducting tin oxide (CTO) glass plates (0.5x1.5cm2, active area 0.25cm2) as described in chapter 2. The film thickness of the TiO2 film is ~10m. Films of above thickness prepared by this method have roughness factors of the order 200-300 [21]. Film was washed with propan-2-ol, dried and exposed to uv light for 20 min to burn out any organic matter contaminating on the film surface.
Trihydroxybenzoic acid (TA) was used as an anion producing compound that strongly anchors to the TiO2 surface i.e., the ionic absobate. And the dyes used in this investigation are bromopyrogallol red (BR), mercurochrome (MC), methyl violet (MV), and IR786 purchased from Aldrich. Anionic dyes BR and MC that get readily adsorbed
55
on the TiO2 surface. MV and IR 786 were chosen as cationic dyes, which adsorbs poorly or do not adsobe on to TiO2 surface.
TA was adsorbed into TiO2 films by immersing the plates in a 10-2 M aqueous solutions containing NaOH. Films were rinsed with water to remove the excess unchelated TA. An outer monolayer of MV was adsorbed on to this surface by immersing the TA coated TiO2 film in a 10-3 M solution of MV for about 15 min. The amount of MV adsorbed on a TA treated TiO2 film was estimated by noticing the depletion of the dye in the coating solution using a spectrophotometric method.
MC coated TiO2 films were prepared by immersing the plates in an alcoholic solution of MC of 5x10-4 M for about 30 min. To deposit the outer layer of MV on MC, the MC coated film was rinsed with ethanol and immersed in a solution of MV solution of 10-3 M in ethanol for 30 min. The amounts of adsorbed MV and MC were estimated by extracting these dyes into an alkaline solution and measuring the concentration spectrometrically.
The anionic dyes BR that readily adsorbed on the TiO2 surface were coated on the nanocrystalline TiO2 film by soaking it in an alcoholic solution of`10-3 M of the dye for about one hour. The cationic dye IR786 was deposited on anionic dye coated surface by exposing this film to a solution of IR786 dye in 75% alcohol. The amounts of dyes adsorbed on the film were estimated by extraction of the dyes into alkaline alcoholic solution and by spectrophotometric estimation after adjustment of pH. To form the heterojunctions n-TiO2/X-Y/p-CuSCN (X=TA, MC or BR and Y=MV or IR786),
56
CuSCN was deposited on Y from a solution in propyl sulfide by the procedure reported earlier on chapter 2. Graphite was painted on the outer surface of CuSCN and a gold plated CTO glass plate served as the back contact of the photovoltaic cell. A schematic diagram showing the construction of the cell is presented as Figure 3.1. I-V characteristics and Photocurrent action spectra of the cells were recorded
D 1 D2
Au coated Counter electrode
Graphite
CuSCN
CTO Glass
TiO2
Figure 3.1
Schematic diagrams showing the construction of the cell TiO2/D1D2/ CuSCN
57
3.3
Results and Discussion
When the TiO2 films are dipped into an aqueous solution of TA, containing NaOH, the TA molecules chelates to TiO2 surface via two hydroxyl groups. Presence of NaOH prevents involvement of the carboxylate group in the chelating process, because of the attractiveness of the acidic group to Na+ ion. Then the TA treated film is dipped into a solution with the cationic dye MV. MV cations bond to TA anions which already anchored to TiO2 surface eliminating NaCl. Scheme 3.1 illustrates the attachment of MV to TA. TA can also be attached to the TiO2 surface without the NaOH so the chelation originates via the hydroxyl or the carboxylate ligands; but however, with this method the TA adsorbed TiO2 films do not adsorb MV due to lack of carboxylic groups.
When TA treated (in an alkaline solution of TA) TiO2 film exposure to a solution of cationic dye MVCl, MV cations bond to TA anions anchored to TiO2 eliminating NaCl as in Scheme 3.1. I-V characteristics of the two cells TiO2/TA-MV/CuSCN and TiO2/MV/CuSCN are compared in Figure 3.2 (I). The short-circuit photocurrent (Isc), open-circuit voltage (Voc), efficiency () and fill-factor (FF) are higher in the TA treated cell compared to the untreated cell, where MV directly affixes to TiO2. The values are summarized in Table 3.1. The photocurrent action spectra of the two cells are shown in Figure 3.2(II) and the incident photon to photocurrent conversion efficiency (IPCE) values of the cells at the absorption peak of MV (620nm) is also shown in Table 3.1. IPCE is higher in the TA treated cell.
58
Scheme 3.1
Anchoring of the sodium salt of trihydoxybenzoic acid to TiO2 and attachment of methyl violet cation by replacement of Na+
2
2.5
b
b
1.5
I / A
I / mA cm -2
2
a
1
a
1
0.5 0 0
200
400
600
450
550
650
750
/ nm
V / mV
(I) Figure 3.2:
0 350
(II)
(I) I-V characteristics of (a) TiO2/MV/CuSCN (b) TiO2/TA-MV/CuSCN and (II) Photocurrent action spectra of (a) TiO2/MV/CuSCN (b) TiO2/TA-MV/CuSCN
59
Table 3.1:
Short-circuit photocurrent (Isc), open-circuit voltage (Voc), efficiency (), fill-factor (FF) and peak (620 nm) incident photon to photocurrent conversion efficiency (IPCE) of TiO2/MV/CuSCN and TiO2/TAMV/CuSCN
Isc (mA/cm-2)
Voc (mV)
FF %
%
IPCE%
MV
1.3
407
38.5
0.21
13.1
TA-MV
2.2
507
42.9
0.47
17.8
Cell
The MV is anchoring to TiO2 via the positively charged N atoms in the organic bases and the acidic sites of the oxide surfaces [22]. The quantities of MV adsorbed on TiO2 and TiO2/TA were found to be 1.5×10-8 and 3.7×10-8 moles per one squire centimeter aria respectively. The adsorption of MV is much lesser on bear TiO2 surface than on TiO2/TA. The coverage of MV in TiO2/TA-MV is calculated to be in the order of a single monolayer when the area of the MV molecule is taken as ~0.9nm2. On the bare TiO2 surface it is around 0.4 times of a monolayer. The ratio of Isc values of TiO2/MV and TiO2/TA-MV is higher than the ratio of the surface concentration of MV dye molecules in the two cases. The ratio of the increment of currant and the number of due molecules is around 3:2. This indicates that the surface protection against recombination much higher in the presence of TA. Electron transfers generating the photocurrent can be understood as follows: excitation of the chromophore in MV will release an electron
60
to the conduction band of TiO2 through the TA bridge and a hole to the valence band of CuSCN, i.e. hiV/CuSCN
TiO2/TA-MV*/CuSCN TiO2(e-)/TA-MV/CuSCN(h+)
TA bridge seems to act as a barrier suppressing the recombination of the separated charges. We believe that the better performance of the first cell compared to the second one, is due to of this effect that suppresses the recombination. In order to test this hypothesis further, we adjusted the level of MV coverage on TiO2/TA-MV/CuSCN to obtain a photo current same as that of TiO2/MV/CuSCN and found that the Voc of this cell attains a value (497 mV) similar as that of a cell with full coverage of MV (Table 3.1). This observation suggests that the TA barrier effectively suppress recombinations that lower the Voc. There are two possible types of recombination. One is the geminate combination of the separated electron hole pair across the dye layer. The other one is the Electron-hole recombination at the points where the TiO2 and CuSCN surfaces touch each other. The bridge connecting TA to MV is expected to suppress recombination of the first type. If the voids in the dye layer were covered with TA, the second type of recombination would also be mitigated.
Instead of an adsorbate without a cromophoure we can use an absorbate with a cromophore, or in other words we can replace the absobate (TA) with a dye. Different possible configurations for double dye DSSSCs are presented in Figure 3.3. In the configuration shown in Figure 3.3(I), the dye layers are homogeneously mixed. Here it is unlikely that excitation of every molecule of D1 (MC or BR) or D2 (MV or IR786)
61
will lead to an efficient injection of electrons into the n-region and holes into the pregion. Because of the predominant quenching processes which can be summarized as follows, D1 * + D 2
D1 + D2
(A)
D1 + D 2 *
D1 + D2
(B)
D1* + D2*
D1 + D2
(C)
Figure 3.3
Schematic diagrams illustrating possible configurations of double-dye solid-state solar cells, (I) homogenously mixed thick layer two dyes (II) a monolayer consisting of two non-interacting dye molecules coupled to n and p-type semiconductors (III) dye layer consisting of two electronically coupled dye molecules bonded on opposites sides to n and p-type semiconductors (circles indicate two types of dye molecules).
62
In the situation depicted in Figure 3.3 (II), dye molecules do not interact with each other but each molecule is anchored to both n and p type semiconductors. As the dye molecules are non-interacting, the quenching processes will be absent. However, the coverage of each dye will be at sub-monolayer level and the absorption cross-section of light at two peak wavelengths will be below the optimum. In the configuration shown in Figure 3.3(III), the two dye molecules D1 and D2 are coupled to each other and also anchored to N and P regions respectively. If the location of the ground (So) and excited (S*) levels of the dye relative to the two semiconductors are as in Figure 3.4, excitation of D1 could result electron injection to N region and hole injection P region. The latter process involves hole conduction through the molecule D2. Similarly, excitation of D2 could result hole injection to P region and electron injection to N region via conduction through the dye molecule D1, i.e., hTiO2 /D1- D2/CuSCN
TiO2 /D1*- D2/CuSCN TiO2(e-)/D1-D2/CuSCN (h+)
hTiO2 /D1- D2/CuSCN
TiO2 /D1- D2*/CuSCN TiO2(e-)/D1-D2/CuSCN (h+)
If the injection rates of the electron and the hole are faster than quenching [i.e., processes (A) – (C)], the quantum efficiency of charge separation would be nearly unity. As injection rate is proportional to conductance, high molecular conductivity of the dyes will favor fast injection that competes with quenching.
63
Figure 3.4
Schematic energy level diagram indicating the relative positions of conduction bands (CB) and valence bands (VB) of TiO2 and CuSCN and ground and excited levels of the D1 and D2 (a) charge transfer on excitation of D1 (b) charge transfer on excitation of D2.
The use of anionic and cationic dyes as described in the experimental section is a simple way of coupling two dyes to form a structure of the configuration N/D1 – D2/P as shown in Figure 3.3(III). Consider the sodium salt of D1, when the TiO2 film is exposed to a solution of D1Na, surface chelation would form a complex, which we represent by TiO2/D1Na. The subsequent treatment of the film with a solution of D2Cl (the chloride of the cationic chromopore D2) produces the structure TiO2/D1–D2 via the double decomposition reaction,
64
TiO2/ D1Na + D2Cl
TiO2/ D1 – D2 + NaCl
Scheme 3.2 and Scheme 3.3 indicate how MC, MV and BR, IR 786 gets attached to TiO2 surface by the above reaction.
MC
MV
Scheme 3.2. The mode of anchoring of mercurochrome (MC) to TiO2 and attachment
of methyl violet (MV) cation by replacement of Na+
65
BR
Scheme 3.3.
IR 786
The mode of anchoring of Bromopyrogallol Red (BR) to TiO2 and attachment of IR 786 cation by replacement of Na+
The mode of interaction of MV and IR 786 with CuSCN remain uncertain. Presumably, nitrogen sites in MV or IR 786 bonds to Cu atoms on the surface of CuSCN. The incident photons to photocurrent conversion efficiencies (IPCEs) of cells sensitized with MC, MV, MC-MV and the cells sensitized with BR, IR 786, BR-IR 786 are
66
summarized in Tables 3.2 and 3.3 respectively. It is interesting to note that the IPCEs at the peak absorption wavelengths are higher for the double dye systems. Photocurrent action spectra of the cells sensitized with MC and MC-MV is in Figure 3.5, the cell with BR and BR- IR786 presented in Figure 3.6 It clearly shows that in the BR-IR786 system, sensitization is extended to near IR region. Table 3.4 summarizes the energy conversion efficiencies of cells sensitized with different dyes. The efficiencies of the double dye cells are seen to be higher than those sensitized with individual dyes.
Table 3.2
Incident photon to photocurrent conversion efficiencies (IPCEs) of the cells (1) TiO2/ MC-MV/ CuSCN (2) TiO2/ MC/ CuSCN (3) TiO2/ MV/ CuSCN at peak absorption wavelengths of the two dyes.
IPCE%
IPCE%
(=550nm)
(=620nm)
MC-MV
22.2
13.8
MC
15.2
-
MV
-
13.1
Cell
67
I / A
8
a
b
c
4
0 350
450
550
650
750
/ nm
Figure 3.5
Photocurrent action spectrum of (a) TiO2/MC-MV/CuSCN (b)TiO2/MC/CuSCN (c) TiO2/MV/CuSCN
Table 3.3
Incident photon to photocurrent conversion efficiencies (IPCEs) of the cells (1) TiO2/ BR-IR786/ CuSCN (2) TiO2/ BR/ CuSCN (3) TiO2/ IR786/ CuSCN at peak absorption wavelengths of the two dyes.
IPCE%
IPCE%
(=590nm)
(=800nm)
BR- IR786
9.7
4.1
BR
6.9
-
-
1.0
Cell
IR 786
68
I / A
2
1
0 350
450
550
650
750
850
/ nm
Figure 3.6
Photocurrent action spectrums of the cells (a) TiO2/BR/CuSCN (b) TiO2/BR-IR786/CuSCN.
Light induced charge transfer in the systems we have studied could involve intermediate steps. If we neglect the extraneous interactions in between the dye molecules, the basic operating unit we need to consider is constituted of two chromophores (D1 and D2) linked by a bridge B and each chromopore bonded to n and p-type semiconductors on the opposite sides as illustrated in Figure 3.7.
69
Table 3.4
Open-Circuit Voltage (Voc), Short-Circuit Photocurrant (Isc), Fill factor (FF) and Energy conversion efficiencie () of the cells (1) TiO2/ MCMV/ CuSCN (2) TiO2/ MC/ CuSCN (3) TiO2/ MV/ CuSCN (4) TiO2/ BR- IR786/ CuSCN (5) TiO2/ BR/ CuSCN (6) TiO2/ IR786/ CuSCN.
Cell
Voc / mV
Isc / mAcm-2
FF %
%
MC -MV
629
4.6
47.3
1.37
MC
603
2.4
40.9
0.60
MV
407
1.3
38.5
0.21
BR- IR 786
495
TiO2/D1/CuSCND*2/CuSCN ------->TiO (e-)/D1/CuSCND2/CuSCN (h+)
As shown in the Figure 4.5, the location of the excited states S*1 and S*2 of the two dyes above the CB of n-TiO2 and ground levels So1 and So2 below the VB position of CuSCN permits above charge transfers energetically
.
Vacuum Energy (eV)
-1
CB
-2 -3 -4
S2 *
CB
e
-
-5
e-
S1 *
h+ h+ VB
S10
-6
S2 0
-7 -8
VB TiO2
Figure 4.5
D1 CuSCN
D2 CuSCN
Energy level diagram showing the band structure in TiO2/D1/CuSCN/D2/CuSCN, the positions of the ground (So1, So2) and excited (S*1,S*2) levels of the two dyes (D1,D2). The dyes D1 (FG) and D2 (AY) anchored to TiO2 and CuSCN respectively
86
Band bending in the region of the heterostructure interface, which facilitates charge transfer, is also indicated in Figure 4.5. Band bending could begin at the dye monolayer. In a cell of configuration {1}, the monolayer of the dye D1 is the barrier that separating the relaxed electron-hole pair. However in the configuration {3}, wider barrier due to the CuSCN layer and the two dye layers (D1/CuSCN/D2) separates the electrons and holes relaxed to CB and VB of TiO2 and CuSCN respectively. Energy of the geminate electron-hole pair enables achieving wide charge separation after tunneling through the barriers. However after relaxation to the band edge energies, the reverse tunneling that leading to recombination is greatly reduced because of the tunneling probability depends on particle energy. As recombination is better suppressed, IPCEs corresponding to peak absorption positions of both the dyes D1 and D2 reach strikingly higher values in the configuration {3} with compared to the other configurations based on individual dyes. Although thicker barriers are more effective in mitigating recombination, they also reduce the injection rate. Thus the barrier width needs to be tuned to an optimum to achieve the maximum IPCE. We believe that for a similar reasons double dye cell gains a higher Voc. The difference in quasi- Fermi levels (q-FL) of electrons in the p-type and the n-type regions measures the Voc and the band gap between the VB and CB edges determines an upper limit to this quantity. When the barriers suppress recombination, build up of q-FLs enhances the Voc.
Several other semiconductor/dye heterojonction structures have also been constructed with a view of understanding the charge transfer and recombination mechanisms. A cell of configuration {1} with D1 replaced by D2 (i.e. TiO2/D2/CuSCN) gave Isc and Voc
87
much less than that of the configuration {2} with the same dye (D2) (Table 4.1). Here again the reason seems to be suppression of recombination by the CuSCN barrier in the former system. Cells where two dyes are deposited one on top of the other (i.e., heterostructures of the form TiO2/D1D2/CuSCN) are not expected to perform satisfactorily owing to mutual deactivation of D*1 and D*2 in close proximity (concentration quenching). With dyes D1(=FG) and D2(=AY), the above heterostructure could not be satisfactorily constructed as D2 is not adsorbed on D1 coated TiO2. When D2 was painted over D1 the cell out-put was found to be very much inferior to that of the configuration {3}. Possibility exists that suitable combination of other dyes may overcome concentration quenching if the rates of charge transfer; energy transfer or injection competes favorably with the rate of concentration quenching.
The above technique can be extended to form heterostructures constituted of more than two dyes. The photocurrent action spectrum of a 3-dye model system, which we have constructed, i.e., TiO2/D1/CuSCN/D2/CuSCN/D3/CuSCN (D1= FG, D2= AY, D3 = Rhodamine 6G) displayed contributions of all the three components. Unfortunately we did not succeed in improving the efficiency above the two-dye system. Thicker barriers in this case seem to reduce electron and hole tunneling probabilities. Our fabrication technique at this stage does not permit deposition of CuSCN layers of thickness < 2 nm .The open-circuit voltage and the efficiency we obtained for the double dye system is significant for a DS SSC [4, 6, 46-53] that utilizes readily available organic sensitizers. Further improvements would depend on identification of dyes with appropriate energy level positions, optical absorption characteristics and developing methods for depositing ultra-thin layers of high band gap semiconductors on nanostructured surfaces. Ordering
88
of the dye layers to take advantage of wavelength dependent light scattering in the nanocrystalline matrix would also be important. But to complete the discussion the experimental results of the three dye system is also reported.
The photocurrent action spectrum of the cell TiO2/D1/CuSCN/D2/CuSCN/D3/CuSCN and the absorption spectra of the individual dyes D1, D2, D3 in aqueous solution are presented in Figure 4.6. Action spectrum has three peaks at 470, 540, 650 nm corresponding to the contributions from the three dyes clearly demonstrating that excitation of each dye result electron injection to n-TiO2 and hole injection to CuSCN.
0.015
0.8
Absorbance
a
Current / mA
c 0.6
b
0.4
0.01
0.005
0.2
0 350
0 450
550
650
350
750
550
650
750
Wave length / nm
Wavelength (nm )
(ii)
(i) Figure 4.6
450
(i) Absorption spectrum of aqueous solutions of (a) D3 = Acridine Yellow (b) D2= Rhodamine 6G (c) D1= Fast Green; (ii) Photocurrent action spectrum of the cell n-TiO2/D1/p-CuSCN/D2/p-CuSCN/D3/ pCuSCN.
89
Figure 4.7 gives an energy level diagram showing the positions of the bands of TiO2 and CuSCN and the positions of the ground (S10, S20, S30) and excited (S1*, S2*, S3*) levels of the dye. As the dye monolayers do not touch each other, excitation energy transfers are ruled out and plausible charge transfers should involve carrier tunneling processes as described below.
Vacuum Energy (eV)
-1 -2
CB CB
S2*
-3 -4
-
- -
h+ h+h+
e ee
-5
S10
-6 -7 -8
S1
S2
0
S30
VB
VB TiO2
Figure 4.7
S3*
*
CuSCN
CuSCN
An energy level diagram showing the conduction and valence band edges of TiO2 and p-CuSCN and the ground (S10,S20,S30) and excited (S1*,S2*, S3* ) levels of the dyes D1, D2, D3. (D1=Fast Green, D2= Rhodamine 6G, D3=Acridine Yellow).
90
Excitation of three dyes and the electron transport process can be summarized as follows. hiO2/D1/CuSCN/D2/CuSCN/D3/CuSCN iO2/D1*/CuSCN/D2/CuSCN/D3/CuSCN
e-)iO2/D1/CuSCN/D2/CuSCN/D3/CuSCN ( h+ )
Here hole tunnels through the barrier CuSCN/D2/CuSCN/D3 having a thickness of the order 6 nm (i.e., ~ 4 nm for the two CuSCN layers and ~ 2 nm for the monolayers of D1 and D2). Same way when D2 and D3 exited electron and the hole tunnels to TiO2 and CuSCN regions through the barriers as follows,
hiO2/D1/CuSCN/D2/CuSCN/D3/CuSCN iO2/D1/CuSCN/D2*/CuSCN/D3/CuSCN
e-)iO2/D1/CuSCN/D2/CuSCN/D3/CuSCN( h+ )
hiO2/D1/CuSCN/D2/CuSCN/D3/CuSCN iO2/D1/CuSCN/D2/CuSCN/D*3/CuSCN
e-)iO2/D1/CuSCN/D2/CuSCN/D3/CuSCN( h+ )
I-V characteristics of the triple dye cell is shown in Figure 4.8, the values of shortcircuit photocurrent (Isc), open-circuit voltage (Voc) and efficiency () are given in the Table.4.3. This table also gives values of the same parameters for the triple dye cell when D2 and D3 are interchanged, the double dye cells TiO2/D1/CuSCN/D3/CuSCN and TiO2/D1/CuSCN/D2/CuSCN and for comparison the single dye cells TiO2/Di/CuSCN (i = 1, 2, 3). Although, the action spectrum of the triple dye cell shows contributions from
91
all the three dyes, the efficiency is seen to be slightly higher in the double dye system Reason seems to be the involvement of thicker barriers in the triple dye cell, which decreases the tunneling probabilities. It is also interesting to note that in the double as well as the triple dye system, Isc and get reduced when D2 (AY) and D3 (R6G) are interchanged. The positioning of the excited level (Figure 4.7) of D3 (i.e., S3*) happens to be well above that of D2 (i.e., S2*), therefore the probability of electron tunneling to TiO2 through a thicker barrier is greatly reduced on interchanging of D2 and D3 (tunneling probabilities are highly sensitive to the barrier height and the width).
5
I /mA
4 3 2 1 0 0
200
400
600
V / mV
Figure 4.8
I-V characteristics of the 3-dye cell -TiO2/D1/CuSCN/D2/ CuSCN/D3/ CuSCN.
92
Table 4.3
Short-circuit photocurrents (Isc), open-circuit voltages (Voc) and efficiencies of dye-sensitized solid-state photovoltaic cells of different configurations (D1= Fast Green, D2 = Rhodamine 6G, D3 = Acridine Yellow).
Isc / mAcm-2
V oc / mV
%
TiO2-D1/CuSCN
3.5
528
0.95
TiO2-D2/CuSCN
0.3
366
0.04
TiO2-D3/CuSCN
0.3
427
0.06
TiO2-D1/CuSCN-D2/CuSCN
2.3
531
0.66
TiO2-D1/CuSCN-D3/CuSCN
5.0
634
1.67
TiO2-D1/CuSCN-D2/CuSCN-D3/CuSCN
4.5
650
1.50
TiO2-D1/CuSCN-D3/CuSCN-D2/CuSCN
3.7
465
0.82
Cell Configuration
4.4
Conclusion
The above dyes, Fast Green, Rhodamine 6 G and Acridine Yellow were selected to construct the model 3-dye solar cell because they are readily available and have nearly non-overlapping absorption spectra in the red, yellow and blue regions of the spectrum. The first dye, which is anionic, readily anchors to TiO2 and thiocyanates of the last two cationic dyes bind to CuSCN. Again these dyes do not leach into the CuSCN coating solution during deposition. Furthermore, ground and excited levels of these dyes makes electron and hole injection to TiO2 and CuSCN energetically feasible. However, they
93
are not the best sensitizers and when used singly in DS solid state cells and yield very low efficiencies. The minimum thickness of the barrier CuSCN layers we could deposit happens to be ~2nm. With the present method of deposition of CuSCN, the film thickness cannot be reduced beyond ~2nm owing to the formation of pinholes that intermix the dyes. As tunneling probabilities are very sensitive to the barrier thickness further reduction of the thickness of the barrier widths may be necessary to obtain practically meaningful efficiencies even if other sensitizers are utilized. Our model system clearly demonstrates that energetic carriers released in dye-sensitization can tunnel through fairly thick barriers. A useful merit of dye-sensitization is electronic coupling of the to the dye semiconductor surface enables fast injection [54] of energetic carriers without dissipative losses at the surface so that rapid relaxation is suppressed and carriers could tunnel through barriers [55-59]. Band bending between the TiO2 and CuSCN region could assist tunneling. Again tunneling may not be direct and involve intermediate states. We believe that with identification of better materials to be used as ultra-thin barriers, semiconductor–dye hetrostructures highly responsive to broad range of the visible spectrum may design. In addition to solar cells and other optoelectronic devices, they could also find applications as visible light sensitive photocatalysts.
94
CHAPTER 5 Electron conduction in a nanostructure based on extremely thin absorbat layer of polythiocyanogen.
Introduction
In this chapter we introduce a barrier layer to the heterojunction. Instead of using a dye this barrier layer is also used as the sensitizer. So it prevents the recombination as well as injects electrons and holes to sensitize the respective semiconductors.
Conducting polymers are extensively studied as potential materials for application in optoelectronic devices [60-70]. Their low cost and easy manipulate properties leaves more flexibility in fabrication procedures compared to the conventional single crystal or polycrystalline inorganic semiconductors. An area where the conducting polymers could make significant practical impact is photovoltaics [60-67]. Here the involvements of low temperature deposition techniques without the vacuum technology become a great advantage. Many attempts have been made to construct photovoltaic cells with conducting polymers which are used as the light harvesting material, that generate the carriers. Basically these systems have a heterojunction configuration with a thin film of the polymer interposed between two electrodes with at least one electrode needs to be optically
transparent.
Polythiophene,
polyacetylene,
poly-phenylene
vinylene
derivatives, polyaniline and many other conducting polymers with complex organic molecules as the monomer have been tested for photovoltaic effects in sandwich configuration or blended with other materials to form composite films. A simple
95
molecule that readily undergoes polymerization is thiocyanogen (SCN)2. Cataldo et al [71-73] have conducted extensive investigations to elucidate the structure of polythiocyanogen. Polythiocyanogen of general composition [Sy(CN)2]x was shown to be constituted of long polyazomethine chains analogous to that of polycyanogen or paracyanogen [74] but crosslinked with sulfur bridges of different length depending on the sulfur chain length in the original monomer. Although the electronic conductivity [75] and photosensitivity [76, 77] of polythiocyanogen ([SCN]n) was noted earlier, there are no records describing the use of this material in an optoelectronic device. We have developed methods for deposition of polythiocyanogen on conducting glass or other conducting
substrate
and
also
fabricated
a
photovoltaic
cell
by
coating
polythiocyanogen on a nanocrystalline film of TiO2. This chapter describes preparation of thin films of polythiocyanogen, their characterization and construction of a photovoltaic cell.
4.2
Experimental
First conducting tin oxide (CTO) glass plates were prepared as mentioned in the second chapter. Polythiocyanogen was deposited on the CTO surfaces as follows: Dried the KSCN at 105oC for several hours and dissolved it in moisture free propylene carbonate (0.15M solution). The solution was heated to 90oC and electrolyzed under galvanostatic conditions (2.0mAcm-2) with the CTO glass plate as the anode and a platinum foil as the counter electrode. The thickness of the film was deduced from the charge that has passed through the electrolyte.
96
Powder form of the Polythiocyanogen was prepared by rapid electrolysis of a solution of KSCN in propylene carbonate. Ultrasonic agitation of the solution prevented the adherence of polythiocyanogen to the anode. A compressed pellet was made with the powder and was used to measure the density and the conductivity of the polythiocyanogen.
Resistivity of the film was measured by immersing the thiocyanogen coated plate in a 0.1M sodium sulfate solution, and platinum foil was used as the counter electrode. An impedance meter (Hewlett Packard 4276A LCZ Meter) was used to measure the Resistance of the polythiocyanogen films. The resistivity of the film was calculated by comparing it with the resistance of a cell of same geometry when the thiocyanogen coated plate is replaced with a bare CTO glass plate. The same set up was used to measure the surface capacitance (C) to plot the Mott- Schottky diagram. The photoresponse of the films was examined in a three electrode configuration under photentiostatic conditions in an electrolytic medium of 0.1 M Na2SO4.
Nanocrystalline films of TiO2 were prepared as described in chapter 2. Polythiocyanogen was coated on the nanocrystalline TiO2 surface by the same method as that used to coat on CTO plates. Surface of polythiocyanogen films coated on CTO and nanocrystalline TiO2 surfaces were examined by SEM. FT-IR spectrum obtained by scraping off the film from the CTO surface.
The photovoltaic cell was formed by depositing a layer of CuI over the polythiocyanogen film, which was deposited on TiO2 nanocrystalline film. A gold
97
plated CTO glass plate pressed onto the CuI surface to serve as the back contact. Illustration of the cell is shown in Figure 5.1. I-V characteristics of the cell were measured at 1000 Wm-2 at 1.5 AM illumination.
Figure 5.1
CTO
(SCN)n
h
CuI
TiO2
Gold
Schematic diagram of the Construction of the photovoltaic cell with (SCN)n layer.
5.3
Results and Discussion
When polythiocyanogen is deposited on the CTO surfaces the discharged SCN- ions undergo polymerization at the CTO surface by depositing an orange-yellow film on the CTO surface. The solution has to be heated to 60°C. If the solution is not warmed
98
polymerization becomes slow and polymeric particles of (SCN)n formed near the anode tend to break away from the electrode surface and leach into the solution.
The FT-IR spectrum of the sample of the polymer scraped from the CTO surface is shown in the Figure 5.2. This spectrum shows same general characteristics of polythiocyanogen prepared by other methods [72]. And it confirms that the material deposited on the CTO glass surface is polythiocyanogen.
80
T(%)
60
40
20
0 500
1000
1500
2000
Wave number (cm-1)
Figure 5.2
FT-IR spectrum of the polythiocyanogen scraped off from a film deposited on conducting tin oxide glass (T = Transmittance).
99
Films deposited on CTO glass or on TiO2 were found to be highly stable and firmly adhered to the substrate. They are resistant to concentrated nitric and sulfuric acids but attacked by strong alkalis. The film softens and peels off when immersed in a strong solution of sodium sulfide. When the polythiocyanogen film was heated there is no sign of chemical decomposition or film breakdown was detected up to ~ 300 o C.
A polythiocyanogen film with thickness of ~100nm deposited on TiO2 substrate has a conductivity of 9×10-8Scm-1. But compressed pellets had a higher conductivity of 4×10ֿ6Scm-1. Presumably rapid electrolysis process may have introduced some doping in to the polymer. In photoresponse measurements of the films, an anodic signal was observed; this suggested an n-type behavior in polythiocyanogen. When doped by Iodine, film increases the conductivity and the sign of the photocurrent indicated a ptype behavior. Experiments with compressed pellets of polythiocyanogen have also demonstrated an increase in conductivity of polythiocyanogen on doping with iodine and bromine [72]. Figure 5.3 shows the Mott-Schottky plots at 1.5 kHz and 1.0 kHz for a polythiocyanogen film deposited on a conducting glass. According to the Figure 5.3 the conduction band edge of the polythiocyanogen is positioned at –0.46V vs SCE. The positive slope of the plots confirms n-type conductivity.
100
6.0E+12
2
-2
1/C (F )
a 3.0E+12
b
0.0E+00 -0.8
-0.6
-0.4
-0.2
0
0.2
0.4
V (vs SCE)
Figure 5.3
The Mott-Schottky plot for a film of polythiocyanogen deposited on conducting glass. Measurement frequency: (a) 1.5 kHz, (b) 1 kHz.
In Figure 5.4 curve (a) shows the optical absorption spectrum of the polythiocyanogen film which was measured using the uV-Visible spectrometer. The peak position in Figure 5.4(a) it shows the absorbance peak which corresponds to the band edge absorbance. And it is found at 450 nm, which corresponding to a band gap of 2.25 eV.
Comparison of the photocurrent action spectrum (plot of IPCE vs wavelength in Figure 5.4 curve (b)) and the optical absorption of the film (curve (a)) they shows that the photocurrent originates from the light absorbed by the polythiocyanogen film.
101
0.3
b
0.2
5 IPCE%
Absorbance
a
0.1
0
0 350
450
550
650
750
Wavelength (nm)
Figure 5.4
(a) Absorption spectrum of a polythiocyanogen film and (b) photocurrent action (IPCE) spectrum of the cell TiO2/[SCN]n/CuI.
SEM pictures of polythiocyanogen deposited onto a CTO surface and a bare CTO surface are shown in the Figures 5.5.1 and 5.5.2 Structures other than the granulites in the CTO surface are absent in the former indicating that the polymer deposits as an uniform interconnected matrix free of pin holes and large irregularities in thickness. Figures 5.5.3 and 5.5.4 compares SEM images of TiO2 coated onto CTO glass with SEM images of polythiocyanogen deposited onto TiO2 coated CTO glass. It is obvious from Figure 5.6 that the polymer film fully covers the rough surface of the nanocrystalline TiO2 surface too.
102
Figure 5.5.1 SEM picture of bare conducting tin oxide glass surface
Figure 5.5.2 SEM picture of polythiocyanogen deposited on conducting tin oxide
glass surface.
103
Figure 5.5.3
SEM picture of bare nanocrystalline TiO2 film.
Figure 5.5.4 SEM picture of polythiocyanogen deposited on a nanocrystalline film of
TiO2.
104
The I-V characteristics of the photovoltaic cell TiO2 / [SCN]n / CuI at 1000Wm-2, 1.5AM illumination is shown in the Figure 5.6. The short-circuit photocurrent, opencircuit voltage and efficiency of the cell is 2.0mAcm-2, 325mV and 0.3 % respectively. Figure 5.7 shows a schematic diagram which illustrates the positions of the conduction and valence bands of TiO2, [SCN]n and CuI. The mechanism of the photovoltaic effect can be explained as follows:
2.5
I (mA/cm2)
2 1.5 1 0.5 0 0
100
200
300
400
V (mV)
Figure 5.6
I-V characteristics of the cell TiO2/ [SCN]n/ CuI measured at 1000Wm-2, 1.5AM illumination
105
Vacuum Energy -1
CB
-2 -3 -4
CB
e-
-5
CB
2.25
-6 eVVB
-7 -8
VB (SCN)n
TiO2 Figure 5.7
h+ VB
CuI
Schematic energy level diagram of conduction and valance band positions of TiO2, [SCN]n and CuI.
Photons absorbed by [SCN]n generate excitons which decomposes to electrons and holes at the interfaces [SCN]n/TiO2 and [SCN]n/CuI As the diffusion length of excitons in a polymer is expected to be of the order of 10nm, it is unlikely that diffusion of excitons generated in the bulk of the 100nm film of [SCN]n contribute significantly to the photocurrent. However when the excitons are decomposed at the TiO2/[SCN]n ,
The position of the conduction band of TiO2 allows electron injection to n-TiO2. The hole remaining in [SCN]n could diffuse to the [SCN]n/TiO2 interface and pass onto the valance band of CuI. Similarly when excitons are decomposed at the [SCN]n/CuI
106
interface a hole is injected to CuI and the electron remaining in [SCN]n diffuses to the (SCN)n/TiO2 interfaces and passes onto the conduction band of TiO2. The rates of the above processes depend on the mobility of electrons and holes in (SCN)n. We have not succeeded in measuring the mobility of electrons and holes in (SCN)n.
The dark I-V curve for the cell in the forward and reverse bias is presented in Figure 5.8; this clearly shows the rectification characteristic which needs for functioning as a photovoltaic device. On prolonged illumination, both short-circuit photocurrent and open-circuit voltage undergoes a slow decay as it is found in other photovoltaic devices based on CuI [78]. Decay of photocurrent and open-circuit voltage originates almost entirely from the deterioration of CuI.
Current (A)
5.0E-04
2.5E-04
0.0E+00
-2.5E-04 -1.00
-0.50
0.00
0.50
1.00
Bias Voltage (V)
Figure 5.8
Dark I-V (rectification) curve for the cell TiO2/[SCN]n/CuI
107
4.4.
Conclusion
We have introduced a method for deposition of thin films of the conducting polymer [SCN]n on conducting tin oxide glass or other conducting substrate. These [SCN]n films are found to be highly resistant to chemical and temperature corrosion. Films deposited by this method are found uniform and highly free of irregularities or pin holes.
Films of [SCN]n deposited on nanocrystalline TiO2 films to form the
heterojunction n-TiO2/[SCN]n/CuI
demonstrated good photovoltaic response. We
believe that further characterization and other studies on thin films of [SCN]n could lead to practical applications.
108
CHAPTER 6 Strategy to enhancing the electron transport properties of CuSCN
Introduction
As we have used CuSCN as the p-type semiconductor material in most of our experiments, poor conductivity of the material is a barrier to fabricate devises with a commercial value. In earlier reports electron transport in materials like CuI is widely studied [5, 78-82]. There for as final chapter of this thesis we report our studies on the electron transport properties of the CuSCN and strategy of enhancing the electron transport in CuSCN.
Investigations related to development of unconventional semiconductor materials for devices such as solar cells and electronics devices are a highly active area of research, mainly aimed at simplification of the manufacturing processes to bring down the production costs. In recent the dye-sensitized solar cells and the material issues connected with their development have received grate attention [4, 11, 12, 31, 32, 8396]. Such an event is the introduction of copper (I) thiocyanate (CuSCN), which is used as a replacement of the electrolyte, which acts as an unconventional solid inorganic hole conductor and is now used in constructing dye-sensitized solid-state solar cells [5, 6, 97101].
Number of low temperature deposition techniques such as dip coating, spin coating, electrochemical and chemical methods have been discussed in literature for the
109
deposition of thin films of CuSCN on copper or on conducting glass [5, 6, 94, 98-104]. Recently method was introduced to deposit thin films of CuSCN by dissolving CuSCN in alkyl sulfides [6] and the method of dissolving CuSCN in propyl sulfide is demonstrated in the chapter 2. CuSCN exists as two polymorphic forms which are named as and . form is the commonly available and more stable form [105-107]. -CuSCN has hexagonal crystal structure where layers of SCN ions separate the planes of Cu atoms. And strong Cu-S bonds which are three-dimensionally interconnect these layers [107]. Because of this polymeric nature, CuSCN thin films with free of defects and pinholes could be deposited on large areas. Stoichiometric excess of SCN- ions induce the p-type conductivity to -CuSCN [5, 6, 94, 98-102]. However the conductivity of commercial CuSCN or samples prepared in the laboratory has very low value in most practical applications. Therefore method for introduce SCN- ions into the bulk of CuSCN film has to be investigated.
This introduces a method to implant excess SCN ions into the CuSCN film by exposing the films to gaseous halogens and (SCN) 2 in organic solvents. This method was also used to demonstrate the usefulness of enhancing the conductivity of the CuSCN film in application of an TiO2/Dye/CuSCN heterojunction of a dye sensitized solid-state solar cell (DSSSC) which resulted a drastically increase in the efficiency.
Experimental
CuSCN powder was prepared by mixing aqueous solutions of CuSO4 and KCNS under reducing conditions as discussed in the Chapter 2. Cu(SCN)2 powder was prepared by
110
mixing aqueous solutions containing CuSO4 and KCNS in the molar ratio of 1:2. A brownish black precipitate of Cu(SCN)2 was obtained. The precipitate is then washed and dried in a vacuum oven.
Thin films of CuSCN were deposited on a glass plates (1.5 × 1.0 cm2) by drop coating as described in chapter 2 on top of a hotplate. These films were introduced into a container filled with N2 and small quantities of halogens, Cl2, Br2 or I2. Electrical connections to the CuSCN films were made with gold-coated copper leads. The time variations of the two-probe sheet resistance of the films were measured with a Keithley multimeter.
(SCN) 2 in CCl4 were prepared as follows. The KCNS was dried in a desiccator and finely grounded in an agate mortar. KCNS powder was suspended in CCl4 and dry Cl2 gas was passed through the suspension by keeping the KCNS/CCl4 solution vigorously stirred. In this procedure, Thiocyanogen is liberated via the following reaction,
2KSCN + Cl2
2KCl + (SCN) 2
And the liberated (SCN)2 is then dissolves in CCl4. The presence of excess KSCN and continuation of stirring after chlorine purging ensure removal all Cl2 dissolved in CCl4 (Reaction was carried out in dark as the (SCN)2 undergoes rapid polymerization when it exposed to light producing polythiocynogen). After completion of the reaction the solid residue (KCl and any excess KSCN) was separated by filtration avoiding exposure to moisture (Moisture also found to accelerate the polymerization process of (SCN) 2).
111
CuSCN films were also doped with SCN by soaking the films in the CCl4 solution containing the SCN iones.
The solar cell TiO2/ Dye/ CuSCN was fabricated by the method mentioned in Chapter 2. The dye used was RuN3 [cis dithiocyanato-bis (2,2’bipyridyl 4,4’ dicarboxylate )Ru(II)] purchased from Solaronix. Pt coated conducting tin oxide glass plate pressed on the CuSCN surface served as the back contact. The I-V characteristics of the cell was recorded before and after exposure of the CuSCN surface of the fabricated device to the (SCN)2 solution.
6.3
Results and Discussion
The drop coated CuSCN film on glass substrates has an average thickness of ~1 µm and a sheet resistance of ~120MΩ/□ at room temperature (26oC). CuSCN is a p-type semiconductor with a band gap of 3.6eV [18] and remains photo-stable when even provided with a strong uV radiation [94]. The hole conductivity of CuSCN depends on excess SCN in the bulk and introduction of Cu(SCN)2 to the coating solution is expected to increase the conductivity due to disproportionation of Cu(SCN)2 into CuSCN and SCN [6]. However, enhancement of conductivity could not be observed when Cu(SCN) 2 was added to the coating solution (i.e., 1 mg of Cu(SCN)2 added to 100 mg of CuSCN dissolved in propyl sulfide). SCN undergo polymerization in the solution, and this process is accelerated by heating. Therefore it is imperative that the doping should be done after deposition of the CuSCN film.
112
1.2E+08
Sheet resistance (Ω/)ٱ
8.0E+07
c 4.0E+07
b a 0.0E+00 0
200
400
600
Time (s)
Figure 6.1
Graph of Time variation of the sheet resistance of the CuSCN films when they are inserted into a N2 atmosphere containing (a) Cl2 (b) Br2 (c) I2
The curves a-c of Figure 6.1 shows the change of sheet resistance with time after insertion of the CuSCN films to the cells where the gaseous atmosphere contained Cl2, Br2 or I2. It can be seen that the CuSCN films in Cl2 attains the lowest sheet resistance 0.5 MΩ/□ in a very short time but the resistance of the film in I2 reached 40 MΩ/□ after much longer time (10 min). Compared to I2, the film in Br2 also attained a lower resistance 1.8 MΩ/□. These observations can be understood on the basis of the reactivity and relative molecular size of the halogens. As the I2 molecules are larger in size therefore penetrate weakly into the film and being less reactive it is unable to displace SCN from CuSCN, the rate of decrease sheet resistance and the saturation
113
value attained happens to be lower. Iodine diffused to the bulk of CuSCN, accepts electrons creating holes, which account for observed enhancement of conductivity. Cl2 molecules penetrate to the bulk of the film more easily and have the ability to react with CuSCN in replacing SCN via the following reaction,
2CuSCN + Cl2
2CuCl + 2SCN
Increase in stoichiometrically excess SCN content of the film increases the conductivity of CuSCN, because SCN accepts electrons generating holes in the valence band. SCN liberated in the bulk of CuSCN via the above reaction is less likely to undergo polymerization because of its immobility in solid CuSCN.
The results shown in Figure 6.2 further confirm our explanations. Figure 6.2 indicates the time variation of the sheet resistance when the CuSCN films treated with halogens are kept in a N2 atmosphere. The resistivity of film doped with I2 (curve a) increases rapidly while the film treated with Cl2 (curve c) remains constant. Iodine in the film as it is mobile diffuses out of the CuSCN bulk and therefore the conductivity continues to decrease until equilibrium is established with iodine in the N2 atmosphere. In the chlorine treated film SCN generated seems to be immobile possibly because of the larger molecular size. Bromine treated film shows more or less the same behavior (curve b) and bromine also has ability to replace SCN from CuSCN.
114
6.0E+07
Sheet Resistance (Ω/)ٱ
a
4.0E+07
2.0E+07
b c 0.0E+00 0
200
400
600
Time ( min )
Figure 6.2
Graph of Change in sheet resistance of the CuSCN films doped with (a) I2 (b) Br2 (c) Cl2 when kept in a N2 atmosphere
The presence of SCN acceptor levels in CuSCN could be observed by fluorescence spectroscopy [104]. The fluorescence spectrum with excitation wavelength at 300nm for a CuSCN film as deposited (curve a) and one treated with Cl2 (curve b) are presented in Figure 6.3.The peak at 350nm of nearly same height in both spectra correspond to photons emitted in recombination of electrons in the conduction band with holes in the valence band. The carriers are generated by incident radiation at 300nm, The peak at 460nm appearing in the fluorescence spectra originates from combination of electrons in the conduction band with SCN acceptor to form SCN–. As expected this peak happens
115
to be more intense in the chlorine treated CuSCN film which contains a larger excess SCN. From the fluorescence spectrum, we deduce that the SCN acceptor level is located 0.9 eV above the valence band edge (Inset, Figure 6.3).
60
Emission Intensity (Arb. Units
CB 2.6 eV 3.6 eV
(SCN)2 Acceptor
40
Level
VB
20
b a
0 300
400
500
600
Wavelength (nm)
Figure 6.3
Fluorescence spectrum of (a) CuSCN film on glass (b) CuSCN film on glass exposed to Cl2. Inset: Energy level representation of the (SCN)2 impurity level in CuSCN.
116
Although chlorine exposure greatly enhances the conductivity of a CuSCN film, this gas reacts with most ingredients of fabricated devices. For example when CuSCN layer of a DSSC is exposed to Cl2, the dye layer gets degraded immediately .To eliminate this problem we prepared a solution of (SCN) 2 in CCl4 and used this solution to dope the CuSCN by immersing the complete TiO2/Dye/CuSCN hetrojunction in the (SCN)2 solution. Figure 6.4 shows the I-V characteristics of a cell before and after doping with SCN. It is clear that doping CuSCN by the above procedure greatly improves cell performance Table 6.1 tabulate Voc, Isc, Fill Factor and the efficiency of the DSSSC of both with normal and (SCN)2 doped CuSCN. The efficiency of the cell increased from 0.75% to 2.3% after (SCN)2 treatment. We also noticed that (SCN)2 treatment tends to reduce the open-circuit voltage, although the efficiency and the short-circuit current are increased. Presumably, because of the imperfections in the CuSCN film, there are points where CuSCN layer touches the conducting glass surface. The short-circuiting across these points depends on the conductivity of CuSCN. The above experiments were not conducted with optimized TiO2 films. We believe that further enhancement of the efficiency would be possible, if optimized TiO2 films are used.
117
10
I / mAcm
-2
8 6 4 2 0 0
200
400
600
V / mV
Figure 6.4
I-V characteristic of the cell TiO2/Ru-dye/ CuSCN (a) before exposure to (SCN)2 solution in CCl4. (b) after exposure to (SCN)2 solution in CCl4.
Table 6.1
The short-circuit photocurrent (Isc), open-circuit voltage (Voc), efficiency (η), and fill factor (FF) of the cells TiO2/Dye/CuSCN before and after SCN doping of CuSCN film.
Cell
Isc (mAcm-2)
Voc (mV)
Before SCN doping
2.74
604
45.5
0.75
After SCN doping
9.09
512
51.4
2.39
118
FF % %
We also examined Mott-Schottky plots of CuSCN films deposited on CTO glass before and after doping with SCN by immersion in a solution of (SCN) 2 in CCl4. As expected an increase in slope of the plot of the latter film is clearly noticeable as shown if Figure 6.5. This is a clear indication of the increase in acceptor density of the CuSCN film. Again in the films treated with the (SCN)
2
showed enhancement of the fluorescence
peak at 460nm just as films treated with Cl2.
4E+14
3E+14 1/c2
b 2E+14
a
1E+14
0 0
0.2
0.4
0.6
0.8
V (Vs. SCE)
Figure 6.5
Mott-Schottky plots of a CuSCN film on CTO glass (a) before exposure to (SCN)2 solution (b) after exposure to (SCN)2 solution.
119
6.4
Conclusion
Above experiments clearly shows that doping with SCN once the film is fabricated enhance the conductivity of CuSCN films. Although the exposure of CuSCN films to chlorine, which liberates SCN, is an effective way of increasing the conductivity of CuSCN films, chlorine treatment could leads to degrading of many ingredient materials of fabricated devices. This problem can be circumvented by use of a solution of (SCN)2 in CCl4 to dope CuSCN. We have successfully adopted this procedure to enhance the conductivity of a CuSCN film in the phovoltaic device TiO2/Ru-dye/CuSCN and there by increasing the efficiency.
120
CHAPTRE 7 Addressable arias in future studies
This chapter is presenting a brief introduction of some arias that are needed to carry out further studies for developing the dye sensitized solar cell. Implementing the following strategies to the DSSCs will enable to enhance the performances of the cells.
One aria is the Molecular rectification. As mentioned in the third chapter of the thesis, molecular rectification is one of the applicable methodologies to suppers the recombination. In that chapter, the double dye system (MC-MV) shows some evidence of molecular rectification. But in sense of applying a high rectification effect through molecular level it needs further studies. Usually a mega molecular structure with asymmetric arrangement shows high molecular rectification [108, 109]. Synthesizing a mega molecule with multiple number of dye molecules ionically bonded to each other providing high molecular rectification effect or a single dye molecule with a high molecular rectification effect is will be a challenge task in development of the DSSCs. The effort on application of molecular rectification in to the DSSC will provide further more clues in understanding the electron transportation in dye sensitized systems.
So far many attempts were carried on in couple two dyes to broaden the spectral response. In this thesis too an attempt to use three dyes is discussed using semiconductor dye multilayers but it has its own drawbacks like the increasing of the resistance to the electron hole mobility due to the increment of the resistance with the
121
number of barrier layers. Coupling three dyes with different cromophores one to the other ionically will disable the requirement of a barrier layer. Finding such a system will also benefit the enhancement of the spectral response of the DSSCs.
Another extension will be to fabricate a system with the structure of two ionically coupled dyes in the multylayer system (i.e. build a system with the configuration TiO2/ D1-D2/ CuSCN/ D3-D4/ CuSCN). This will enable to use dyes with four different cromophores with a single layer of barrier in between. This will be a good strategy to enhance the spectral response supperssing recombination as wellas the quenching effects. Also a semiconductor-dye multilayer structure with a suitable n-type semiconductor instead of p-CuSCN can be anther brake through in enhancing the performance of the DSSCs.
So far in the history of the DSSSCs the CuSCN and CuI are the only p-type semiconductors that have performed to an extensive value. But there can be some other p-type semiconducting material that will perform with much higher extent. Researches should be conducted on this aria too in the sense finding a suitable p-type semiconductor to improve the DSSSCs.
Another factor that effects the efficient hole transfer is the back contact. Usually most of the dyes do not anchor well with the p-type semiconductor as well as with the n-type semiconductor. This results a barrier to the hale transport to the back contact. A study of using an absobate between the dye and the CuSCN layer (this even can be used in the
122
dye-semiconductor multilayer structures) will implement a method to enable a good back contact.
Above are few strategies that popup to the surface on improving the performances of the DSSCs while conducting studies for this thesis. Further studies should be carried out in implementing these strategies to the DSSCs.
123
REFERENCES 1. J. Shewchun, R. Singh, M. A. Green, J. Appl. Phys. 48 (1977) 765. 2. A. E. Becquerel, C. R. Acad. Sci. 9 (1839) 561. 3. H. Z. Zhang, J. F. Banfield, J. Mater. Chem. 8 (1998) 2073-2076. 4. K. Tennakone, G. R. R. A. Kumara, A. R. Kumarasinghe, K. G. U. Wijayantha, P. M. Sirimanne, Semicond. Sci. Technol. 10 (1995) 1689. 5. K. Tennakone, A. H. Jayatissa, C. A. N. Fernando, S. Wickramanayake, S. Punchihewa, L. K. Weerasena, W. D. R. Premasiri, Phys. Stat. Sol. 103 (1987) 491. 6. G. R. R. A. Kumara, A. Konno, G. K. R. Senadeera, P. V. V. Jayaweera, D. B. R. A. De Silva, K. Tennakone, Sol . Energy. Mater & Sol. Cells 69 (2001) 195-199. 7. A. J. McEvoy, M. Gratzel, Dye-Sensitized Regenerative Solar Cells in Semiconductor
Electrodes
and
Photoelectrochemistry,
Encyclopedia
of
Electrochemistry, Vol. 6 Ed., S. Licht, Wiley- VCH, Weinheim (2001) pp. 397- 406. 8. A. Hinsch,; J. M. Kroon, R. Kern, I. Uhlendorf, J. Holzbock, A. Meyer, J. Ferber, Prog. Photovoltaics: Research and Applications 9 (2001) 425. 9. A. Hagfeldt, M. Gratzel, Chem. Rev. 95 (1995) 49. 10. A. Hagfeldt, M. Gratzel, Acc. Chem. Res. 33 (2000) 260. 11. K. Tennakone, G. R. R. A. Kumara, I. R. M Kottegoda, V. P. S. Perera, Chem. Commun. (1999) 15-16. 12. B. O’ Regan, M. Gratzel, Nature 353 (1991) 737. 13. B. O’ Regan, D. T. Schwartz, Chem. Mater. 7 (1995) 1349.
124
14. K. Keis, E. Magnusson, H. Lindstrom, S. E. Lindquist, A. Hagfelt, Sol. Energy Mater Sol. Cells 73 (2002) 51. 15. Q. B. Meng, K. Takahashi, X. T. Zhang, I. Sutanto, T. N. Rao, O. Sato, A. Fujshima, H. Watanabi, T. Nakamori, M. Uragami, Langmuir 19 (2003) 3572. 16. R. J. D. Miller, G. L. McLendon, A. J. Nozik, W. Schmickler, F. Willig, Surface electron transfer processes VCH, New York, (1995). 17. J. B. Asbury, E. Hao, Y. Wang, H. N. Gosh, T. Lian, J. Phys. Chem. B 105 (2001) 4545. 18. J. H. Abury, R. J. Ellingson, H. N. Gosh, S. Ferrere, A. J. Nozik, T. Lian, J. Phys. Chem. B 103 (1999) 3110. 19. K. Nazeeruddin, M. Gratzel, Dyes for Semiconductor Sensitization, Encyclopedia of Electrochemistry, Vol. 6 Ed., S. Licht, Wiley- VCH, Weinheim, (2001) pp. 407-431 20. M. Green, Mater. Sci. Eng. B. 74 (2000) 118. 21. G. R. R. A. Kumara, S. Kaneko, M. Okuya, K. Tennakone, Key Engineerining Materials: J. Ceramic Soc. Jpn. 119 (2002) 228. 22. S. M. Kanan, Z. Lu, J. K. Cox, G. Bernhardt, C. P. Tripp, Langmuir 18 (2002) 1707. 23. A. Aviram, M.A. Ratner. Chem. Phys. Lett. 29 (1974) 277. 24. T. J. Mark, M. A. Ratner, Angew. Chem. 34 (1995) 155. 25. A. S. Martin, J. R. Sambles, G. J. Ashwell, Phys. Rev. Lett. 70 (1993) 218. 26. R. M. Metzger, Acc. Chem. Res. 32 (1999) 950.
125
27. R. J. D. Miller, G. L. McLendon, A. J. Nozik, W. Schmickler, F. Willig, Surface Electron Transfer Processes VCH, Weinheim 1994, pp:167-302. 28. A. Hagfeldt, M. Gratzel, Chem. Rev. 95 (1995) 49. 29. J. B. Asbury, E. Hao, Y. Wang, H. N. Gosh, T. Lain, J. Phys. Chem. B 105 (2001) 4545. 30. K. Tennakone, I. R. M. Kottegoda, L. A. A. De Silva, V. P. S. Perera, Semicond. Sci. Technol. 14 (1999) 975. 31. K. Tennakone, P. K. M. Bandaranayake, P. V. V. Jayaweera, A. Konno, G. R. R. A. Kumara, Physica E 14 (2002) 190-196. 32. K. Tennakone, J. Bandara, P. K. M. Bandaranayake, G. R. A. Kumara, A. Konno, Jpn. J. Appl. Phys. 40 (2001) L732. 33. G. R. R. A. Kumara, K. Tennakone, V. P. S. Perera, A. Konno, S. Kaneko, M. J. Okuya, Phys. D: Appl. Phys. 34 (2001) 868. 34. K. Tennakone, V. P S. Perera, I. R. M. Kottegoda, L. L. A. De Silva, G. R. R. A. Kumara, A. Konno, J. Electronic. Mater 30 (2001) 992. 35. Y. Diamant, S. G. Chen, O. Melamed, A. Zaban, J. Phys. Chem. B 107 (2003) 1977. 36. S. Chappel, S. G. Chen, A. Zaban, Langmuir 18 (2002) 3336. 37. S. G. Chen, S. Chappel, Y. Diamant, A. Zaban, Chem. Mater. 13 (2001) 4629. 38. S. Chappel, S. G. Chen, A. Zaban, Langmuir 18 (2002) 3336.
126
39. E. Palomares, J. N. Clifford, S. A. Haque, T. Lutz, J. R. Durrant, J. Am. Chem. Soc. 125 (2003) 475. 40. E. Palomares, J. N. Clifford, S. A. Haque, T. Lutz, J. R. Durrant, Chem. Comm. (2002) 1464. 41. A. Kay, M. Gratzel, Chem. Mater. 14 (2002) 2930. 42. I. Bedja, P. V. Kamat, J. Phys. Chem. 99 (1995) 9182. 43. E. W. McFarland, J. Tang, Nature 421 (2003) 616. 44. B. O’ Regan and D. T. Schwartz, Chem. Mater. 10 (1998) 1501. 45. G. Smestad, Sol. Energy. Mater. Solar Cells 32 (1994) 273. 46. O’Regan, D. T. Schwartz, Chem. Mater. 7 (1995) 1349. 44. K. Tennakone, G. R. R. A. Kumara, I. R. M. Kottegoda, K. G. U. Wijayantha, V. P. S. Perera, J. Phys. D: Appl. Phys. 31 (1998) 1492. 48. B. O’Regan, D. T. Schwartz, J. Appl. Phys. 68 (1996) 2439. 49. U. Bach, D. Lupo, P. Compte, J. E. Moser, F. Weissortel, J. Salbeck, H. Spreitzer, M. Gratzel, Nature 359 (1998) 583. 50. G. R. A. Kumara, A. Konno, K. Shiratsuchi, J. Tsukahara, K. Tennakone, Chem. Mater. 14 (2002) 954. 51. B. O’Regan, D. T. Schwartz, S. M. Zakeeruddin, M. Gratzel, Adv. Mater. 12 (2000) 1263. 52. B. O’Regan, F. Lenzmann, R. Muis, J. Wienke, Chem.Mater. 14 (2002) 5023.
127
53. Q. B. Meng, X. T. Zhang, I. Sutano, T. N. Rao, A. Fujishima, K. Takahashi, O. Sato, H. Watanabe, T. Nakamori, M. Uragami, Proc. 14th Intenational Conference on Photochemical Conversion of Solar Energy, Sapporo, Japan, (2002) pp.W1-P-62 54. J. B. Asbury, E. Hao, Y. Wong, H. N. Gosh, T. Lain, J. Phy. Chem. B 105 (2001) 4545. 55. E. W. MacFarland, J. Tang, Nature 421 (2003) 616. 56. K. Tennakone, K. M. P. Bandaranayake, P. V. V. Jayaweera, A. Konno, G. R. R. A. Kumara, Physica E 14 (2002) 190. 57. Y. Piamant, S. G. Chen, O. Mclamed, A. Zaban, J. Phys. Chem. B 107 (2003) 1977. 58. S. Chappel, S. G. Chen, A. Zaban, Langmuir 18 (2002) 3336. 59. E. Palomares, J. N. Clifford, S. A.Haque, T. Lutz, J. R. Durrant, J. Am. Chem. Soc. 125 (2003) 475. 60. L. A. A. Pettersson, L. S. Roman, O. Inganas, J. Appl. Phys. 86 (1999) 487. 61. T. Yohannes, T. Solomon, O. Inganas, Synthetic Metals 82 (1996) 215. 62. R. Valaski, A. F. Bozza, L. Micaroni, I. A. Hummlelgen, J. Solid State Electrochem. 4 (2000) 390. 63. C. M. Ramsdale, J. A. Barker, A. C. Arias, J. D. MacKenzie, R. H. Friend, N. C. Greenham, J. Appl. Phys. 92 (2002) 4266. 64. M. P. T. Christiaans, M. M. Weink, P. A. van Hal, J. M. Kroon , R. A. J. Janssen , Synthetic Metals 101 (1999) 265. 65. N. A. Anderson, E. Hao, X. Ai, G. Hastings and T. Lian, Chem. Phys. Lett. 347 (2001) 304. 66. G. Yu, J. Gao, J. C. Hummelen, F. Wudl and, A. J. Heeger, Science 270 (1995) 1789.
128
67. C. O. Too, G. G. Wallace, A. K. Burrell, G. E. Collis, D. L. Officer, E. W. Boge, S. G. Brodie, E. J. Evans, Synthetic Metals 123 (2001) 53. 68. G. Wang, X. Hu, T. K. S. Wong, J. Solid State Electrochem. 5 (2001) 150. 69. R. F. Service, Science 279 (1998) 1135. 70. S. Chao, M. S. Wrighton, J. Am. Chem. Soc. 109 (1987) 2197. 71. F. Cataldo, Journal of Inorganic and Organometallic Polymers, 7 (1997) 35. 72. F. Cataldo, Polyhedron 19 (2000) 681. 73. F. Cataldo, Polyhedron 21 (2002) 1825. 74. F. Cataldo, Eur. Polymer J. 35 (1999) 571. 75. F. Cataldo, P. Fiordiponti, Polyhedron 12 (1993) 279. 76. M. Bragadin, G. Capodaglio, P. Cescon, G. Scarponi and F. Pucciarelli, J. Electroanal. Chem. 122 (1981) 393. 77. M. Bragadin, G. Scarponi, G. Capodaglio, F. Ossola, V. Bartocci, F. Pucciarelli, Mol. Liq. Cryst. 121 (1985) 345 78. G. R. A. Kumara, A. Konno, K. Shiratsuchi, J. Tsukahara, K. Tennakone, Chem. Mater. 14 (2002) 954. 79. K. Tennakone, G. R. R. A. Kumara, I. R. M. Kottegoda, V. P. S. Perera, G. M. L. P. Aponsu, K. G. U. Wijayantha, Sol. Energy Matter Sol. Cells 55 (1998) 283289. 80. K. Tennakone, G. R. R. A. Kumara, I. R. M. Kottegoda, K. G. U. Wijayantha, V. P. S. Perera, J. Phys. D Appl. Phys. 31 (1998) 1492-1496. 81. V. P. S. Perera, K. Tennakone, Sol. Energy Matter Sol. Cells 79 (2003) 249-255. 82. P. M. Sirimanne, H. Tributsch, J. Solid State Chem. 177 (2004) 1789-1795. 83. A. Zaban, S. Ferrere, B. A. Gregg, J. Phys. Chem. 102 (1998) 452. 84. S. Yanagida, S. Kambe , W. Kubo , K. Murakoshi , Y. Wada , T. Kitamura , J. Phys. Chem. Part I. 212 (1999) 31. 85. S. Y. Huang, L. Kavan, I. Exnar, M. Gratzel, J. Electrochem. Soc. 149 (1995) L142.
129
86. E. Toppoglidis, A. E. G. Cass, G. Gilardi, S. Sadeghi, N. Beaumont, J. R. Durrant, Anal. Chem. 70 (1998) 5111. 87. Z. K. Tang, G. K. L. Wong, M. K. P. Yu, A. Othomo, H. Koinuma, Y. Segawa, Appl. Phys. Lett. 72 (1998) 3270. 88. U. Bach, D. Lupo, P. Comte, J. E. Moser, F. Weissortel, J. Salbeck, H. Preitzer, M. Gratzel, Nature 395 (1998) 583-584. 89. F. Garnier, G. Horowitz, X. Peng, D. Fichou, Synthetic Metals 45 (1991) 163. 90. L. Torsi, A. Dodabalapur, L. J. Rothberg, A. Fung, H. E. Katz, Science 272 (1996) 462. 91. G. R. R. A. Kumara, K. Tennakone, V. P. S. Perera, A. Konno, S. Kaneko, M. Okuya, J. Phys. D: Appl. Phys. 34 (2001) 868-873. 92. B. O’Regan, D. T. Schwarts, J. App .Phys. 80 (1996) 4749. 93. K. Murakoshi, R. Kogure, Y. Wada, S. Yanagida, Sol. Energy Mater. Sol. Cells. 55 (1998) 113. 94. B. O’Regan, D. T. Schwartz, Chem. Mater. 10 (1998) 1501-1509. 95. A. N. Nogueira, J. R Durrant, M. A. De Paoli, Adv. Mater. 13 (2002) 826. 96. B. O’Brian, F. Lenzmann, R. Muis, J. Wienke, Chem. Mater. 14 (2002) 5023-5029. 97. S. Tan, J. Zhai, B. Xue, M. Wen, Q. Meng, Y. Li, L. Jiang, D. Zhu, Langmuir 20 (2004) 2934-2937. 98. K. Tennakone, J. Phys. D. 16 (1984) l5. 99. K. Tennakone, M. Kahada, C. Kasige, P. Abeysooriya, R. H. Wijenayake, P. Kaviratne, J. Electrochem. Soc. 131 (1984) 1575. 100. K. Tennakone, G. S. Santha Pushpa, S. Punchihewa, G. Epa, Electrochem. Acta. 31 (1985) 315.
130
101. K. Tennakone, J. Karunamuni, M. Dewasurendra, Sol. Energy Mater. 13 (1986) 447. 102. K. Tennakone, C. N. A. Fernando, M. Dewasurendra, J. Photochem. 38 (1987) 75. 103. B. O’Regan, D. T. Schwartz, S. M. Zakeeruddin, M. Gratzel, Adv. Mater. 12 (2000) 1263-1267. 104. C. A. N Fernando, I. Kumarawadu, K. Takashi, A. Kitagawa, M. Susuki, Sol. Energy Mater Sol. Cells. 58 (1999) 337-347. 105. D. Kruger, E. Tschirch, Chem. Ber. B. 74 (1941) 1378. 106. D. L. Smith, V. I. Saunders, Acta. Crystallogr. B. 38 (1982) 907. 107. D. L. Smith, V. J. Saunders, Acta. Crystallogr. B. 37 (1981) 1807. 108. S. K. Polack, J. Naciri, J. Mastrangelo, C. H. Petterson, J. Torres, M. Moore, R. Shashidhar, J. G. Kushmeric, Langmuir 20 (2004) 1838. 109. D. Vuillaume, B. Chen, R. M. Metzeger, Langmuir 15 (1999) 4011.
131
APPENDICES 1
List of Publications and Communications during the period of Studies
1) Dye Sensitised Solid State Photovoltaic Cells Based on Dye MultilayerSemiconductor Nanostructures. V. P. S. Perera, P. K. D. D. P. Pitigala, P. V. V.
Jayaweera, K. M. P. Bandaranayake, and K. Tennakone., J. Phys. Chem. B 107 (2003) 13758.
2) Construction of Photovoltaic Devices by Deposition of Thin Films of Conducting Polymer Polythiosyanagen. V. P. S. Perera, P. V. V. Jayaweera, P. K. D. D. P. Pitigala, K. P. M. Bandaranayake, G. Hastings, A. G. U. Perera, and K. Tennakone.,
Synthatic Matals 143 (2004) 283-287.
3) Sensitization of Nanostructured TiO2 by Electrostatic Coupling of Ionic Dyes to Ionic Absorbates. P. K. D. D. P. Pitigala, M. K. I. Senevirathna, V. P. S. Perera, and
K. Tennakone. Langmuir 20 (2004) 5100-5103.
4) Dye-sensitized near infrared room temperature photovoltaic photon detectors.
P. V. V. Jayaweera, A. G. U. Perera, M. K. I. Senevirathna, P. K. D. D. P. Pitigala and K. Tennakone., Applied Physics Letters., 85 (2004) 5754-5756.
132
5) A Solar Cell Sensitized with Three Different Dyes. V. P. S. Perera, P. K. D. D. P. Pitigala, M. K. I. Senevirathna, and K. Tennakone., Solar Energy Materials and Solar
cells 85 (2005) 91-98.
6) Water Photoreduction with Cu2O Quntum Dots on TiO2 Nano Particles. M. K. I
Senevirathna, P. K. D. D. P. Pitgala and K. Tennakone., Journal of Photochemistry and Photobiology: A Chemistry 171 (2005) 257-259.
7) Doping CuSCN Films for Enhancement of Condutivity: Application in DyeSensitized Solid-State Solar Cells. V. P. S. Perera, M. K. I Senevirathna, P. K. D. D. P Pitigala and K. Tennakone., Solar Energy Materials and Solar cells 86 (2005) 443-
450.
8) Molecular Rectification: Application in Dye-Sensitized Solar Cells. M. K. I.
Senevirathna, P. K. D. D. P. Pitigala, V. P. S. Perera and K. Tennakone., Langmuir 21 (2005) 2997-3001.
9) Relevance of 1/f Noise to the Dye-sensitized Solar Cells. P. V. V. Jayaweera, P. K. D. D. P. Pitigala, A. G. U. Perera, and K. Tennakone., Semicond. Sc. Technol. 20
(2005) L1-L3.
10) Cromopore Linked Conducting Polymers Attached to Semiconductor surfaces: A strategy for Development of Dye-Sensitized Solar Cells. M. K. I. Senevirathna, P. K. D. D. P. Pitigala and K. Tennakone., J. Phys. Chem. B 109 (2005) 16030-16033.
133
11) Dye-Multilayer Semiconductor Nanostructures. P. K. D. D. P. Pitigala,
M.K.I. Seneviratna, V.P.S. Perera and K. Tennakone.,
Comptes Rendus (Accepted for publication).
12) Utilization of MEH-PPV as a sensitizer in titana based photovoltaic cells.
P.M. Sirimanne, E.V.A. Premalal, P. K. D. D. P. Pitigala, K. Tennakone. Solar Energy Materials and Solar cells (In press).
13) Utilization of natural pigment extracted from pomegranate fruits as sensitizer in solid-state solar cells. P. M. Sirimanne, M.K.I. Senevirathna1, E.V.A. Premalal, P. K. D. D. P. Pitigala, V. Sivakumar, K. Tennakone.
J. Photochemistry and
Photobiology A: Chemistry (in press).
14) A solid-state solar cell sensitized with mercurochrome. P.M. Sirimanne; M.K.I.
Senevirathna; E.V.A. Premalal; P.K.D.D.P. Pitigala; K. Tennakone., Current Science (2006) (Accepted for publication).
134
