The Preparation of Nano Composites and Their Applications in Solar Energy Conversion

Springer Theses
Recognizing Outstanding Ph.D. Research
Nailiang Yang
The Preparation of
Nano Composites
and Their
Applications in Solar
Energy Conversion
Springer Theses
Recognizing Outstanding Ph.D. Research

Aims and Scope
The series “Springer Theses” brings together a selection of the very best Ph.D.
theses from around the world and across the physical sciences. Nominated and
endorsed by two recognized specialists, each published volume has been selected
for its scientific excellence and the high impact of its contents for the pertinent field
of research. For greater accessibility to non-specialists, the published versions
include an extended introduction, as well as a foreword by the student’s supervisor
explaining the special relevance of the work for the field. As a whole, the series will
provide a valuable resource both for newcomers to the research fields described,
and for other scientists seeking detailed background information on special
questions. Finally, it provides an accredited documentation of the valuable
contributions made by today’s younger generation of scientists.
Theses are accepted into the series by invited nomination only

and must fulfill all of the following criteria
• They must be written in good English.

• The topic should fall within the confines of Chemistry, Physics, Earth Sciences,
Engineering and related interdisciplinary fields such as Materials, Nanoscience,
Chemical Engineering, Complex Systems and Biophysics.
• The work reported in the thesis must represent a significant scientific advance.
• If the thesis includes previously published material, permission to reproduce this
must be gained from the respective copyright holder.
• They must have been examined and passed during the 12 months prior to
nomination.
• Each thesis should include a foreword by the supervisor outlining the signifi-
cance of its content.
• The theses should have a clearly defined structure including an introduction
accessible to scientists not expert in that particular field.
More information about this series at http://www.springer.com/series/8790

Nailiang Yang
The Preparation of Nano

Composites and Their
Applications in Solar Energy
Conversion
Doctoral thesis accepted by University of Chinese
Academy of Sciences, China
123
Author Supervisor
Dr. Nailiang Yang Prof. Dan Wang
Institute of Process Engineering Institute of Process Engineering
Chinese Academy of Sciences Chinese Academy of Sciences
Beijing Beijing
China China
ISSN 2190-5053 ISSN 2190-5061 (electronic)

Springer Theses
ISBN 978-3-662-53483-0 ISBN 978-3-662-53485-4 (eBook)
DOI 10.1007/978-3-662-53485-4
Library of Congress Control Number: 2016952907
© Springer-Verlag GmbH Germany 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made.
Printed on acid-free paper
This Springer imprint is published by Springer Nature

The registered company is Springer-Verlag GmbH Germany
The registered company address is: Heidelberger Platz 3, 14197 Berlin, Germany
Supervisor’s Foreword
I first met Nailiang in 2006 when he was still a bachelor’s candidate in Sun-Yet-Sen
University. At that time, he interviewed in our institute for doctorate degree. I was
very glad to see a young student who was so obsessed with science and had such
solid knowledge in chemistry; hence, I introduced him to our group. After five
years of research, I am very happy to see a rising young scientist coming from our
group. This book not only comprises a summary of scientific thesis but also serves a
review of his life in those years.
His main works are based on the novel two-dimensional (2D) carbon materials,
specifically on graphene and graphdiyne, which are considered as the rising star in
materials science. In this topic, he also focused on the applications of their
nanocomposites forms with solar cells and photocatalysis as specialties. These
works can be considered as the pioneer of graphene-related research, started two
years before the announcement of graphene works in the Nobel Prize. The basis
of these works was also the first few papers which reported and proved the charge
transport between graphene and other semiconductors, and hence received many
citations. Beyond them, in this thesis, he pointed out the potential applications and
designation for graphene composite referring the structure and function of granum,
which may be beneficial for designing new integrated circuit and chips.
Importantly, he also honestly listed the defects of recent research in this field to give
a thorough guide for further improvement. In the booming age of two-dimensional
nanomaterials, I am confident that the ideas from this thesis can also be referred and
extended to other materials.
I hope the publication of this thesis will be smooth, to allow more readers grasp
new ideas from his discussions. As I know from him, Nailiang is still working on
the synthesis and assembly of novel nanomaterials after his graduation. A broad and
kind discussion with the author in the future will be very beneficial for both the
readers and him.
Beijing, China Prof. Dan Wang

July 2016
v
Acknowledgments
After graduation of 3 years, it is time to reread the original acknowledgment, and I

can still feel my mixed feelings at that moment. Nine years ago, I walked through
the gate of Graduate University of Chinese Academy of Sciences (now named as
University of Chinese Academy of Sciences), and started my scientific research.
Since then, I have witnessed lots of graduate defences, and every time I asked
myself what it would be like for mine. I thought about what I should write and say
in the acknowledgment. Now I understand that one can never know where to start
how to write until that very moment. In this more than five years’ period, I have too
many things to say, with all the memories flashing in my mind. With the time
prolonged, there are more and more people to whom I need to express my thanks.
Because of them, I have this opportunity to write this acknowledgment. Because
of them, I could overcome the difficult barriers and because of them, I could enjoy
the fantastic research life.
First, with no doubt, I would like to express my great thanks to my supervisor,
Prof. Dan Wang. I first met him in the summer of 2006, and he led me to the palace
of science. In these years, he gave me lot of opportunities to exercise, so that I can
access to the most advanced technology area, which made me grow up rapidly in
science. In academy, his open and active mind, accompanied with his hard working
deeply impressed me; in life, he is a model as a man in family and concerns about
our happiness and health. To us students, he is more than a teacher, a friend.
Since 2008, I had the chance to attend the academician Lei Jiang’s group and
started a five-year research as a joint doctoral candidate. What I have to say is, the
most impressive thing about Prof. Jiang is his strong “aura”, which is full of con-
fidence and energy, and I always felt nervous when talking with him. But with time
passing by, I started to know he is very humorous and easy-going. He is always glad
to help and encourage students to immerse in science. In those years, I admired on
his broad knowledge and smart ideas very much. He taught us to stay curious about
everything happening around us, and to explore the scientific reason behind it. He
taught us the evolution of nature is a good teacher for scientists, and nature can tell
you why it should be like this. He taught us to follow the traditional ideas in Chinese
vii
viii Acknowledgments
culture, such as the philosophy of “Dao” and “Yin-Yang”. He taught us the syn-
ergistic effects in nature and applying them in science. He supplied the best
equipment for research and supported our ideas positively. Moreover, he encouraged
us to go abroad, to know more about the scientific frontier. Thank you, Sir.
Also, I would like to thank for Prof. Jin Zhai in Beihang University. She guided
me thoroughly in the experiment and paper writing. I admire her very much for her
solid chemistry knowledge and sensitive sense to the hottest topics in science.
Thanks to her that I was supported to attend lots of academic conferences and meet
many top scientists in the world, and know the culture of other countries.
I would also like to thank other staffs in our group, namely Dan Mao, Xiaoyong
Lai, Jianxi Yao, Chaojian Xing, Nan Xu, Mei Yang, Luoxin Yi, Zhudong Hu, Jiajia
Wu, and Quan Jin. You provided a kind environment in my daily life and studies.
Also many thanks to the labmates, Jun Li, Zhenmin Li, Ronghai Zhu, Shengdong
Wang, Rongguo Xu, Xiaoqing Jiang, Li Li, Jiang Du, Ying Cui, Yuanyuan Liu,
Shuo Wang, Zhenghong Dong, Gongling Wang, Hongjie Tang, Jiangyan Wang,
Yu Yang, Hao Ren, Simeng Xu, Dong Guo, Wei Xu, and others; because of you,
we formed an active group, which brought us lot of happiness beyond the exper-
iments. I also want to express my thanks to Dr. Yuanyuan Liu and Mr. Yibo Zhao
for their kind help, which helped the conclusion of Chap. 5, and to Yu Zhang and
Qi Yuan, who accompanied me to finish the LBL work through many nights. In
addition, I would like to express my thanks for Jingtao Wang and Jiwei Li’s help in
the Au-DSSC work, and also to the teachers, Ms. Lijuan Guo, Ms. Jing Wang, and
Mr. Jianghua Ma in IPE, who provided good life and experiment experiences for all
the graduated students.
Furthermore, dear Academician Daoben Zhu, thanks for guiding me in the
synthesis and characterization of graphene. Dear Prof. Zhiyong Tang, Prof.
Zhixiang Wei and Prof. Dong Han in Nanocenter, thanks for helping me in the
material characterization. Dear Prof. Meixiang Wan, thanks for helping me in
conductive polymer. Dear Prof. Qinghua Fan, thanks for helping me in organic
synthesis. Dear Prof. Dongsheng Liu in Tsinghua University, thanks for helping me
in the gold synthesis. Dear Prof. Tianxin Wei in Beijing Institute of Technology,
thanks for helping me in in the preparing the gold film and SPR testing. Dear Prof.
Jingxia Wang, Dr. Yu Huang and Dr. Libin Wang, thanks for helping me in PS
synthesis and assembly. Dear Mrs. Guilan Wang and Dr. Ye Tian, thanks for
helping me in the supporting of experimental and daily life. Dear Prof. Hao Wen,
thanks for helping me in the thesis guiding and writing. Dear Dr. Luoxin Yi and
Dr. Hao Ren, thanks for helping me in the thesis modification. And dear Ms. Ye
Chen, thanks for helping me in the help of thesis translation.
I also want to express my thanks to all the group members in Prof. Lei Jiang,
Prof. Yanlin Song and Prof. Qinghua Fan’s group in ICCAS, the group members in
Prof. Jin Zhai and Prof. Ying Zhu’s group in BUAA and all the group members in
Prof. Dongsheng Liu’s group in Tsinghua University. Thanks you for the sup-
porting and discussion in experiment.
We also thank the National Natural Science Foundation of China, National Basic
Research Program, 863 Program, State Key Laboratory of Multiphase Complex
Acknowledgments ix
Systems, Ph.D. Programs Foundation of the Ministry of Education of China and

Beijing Municipal Natural Science Foundation for the continuing financial support.
Here, I would like to appreciate my wife, Dr. Liying Wang. Because of the
simple words “I do” 12 years ago, we decided to move forward together. No matter
what kind of trouble we faced these years, I can always get the support from her. In
the past 12 years’ life of study and work, I felt so blessed to have her always staying
with me, encouraging me, comforting my sadness, and sharing my happiness.
Because of her, I gained confidence to finish my doctor career; because of her, I
overcame the difficulties one by one; because of her, I received so much joy which
could never happen without her. It is fortunate to be with her.
Lastly, I need to say that the source of my happiness is also from my beloved
parents. They always take care of my business as priority, bless me safe and happy,
and support my choices. Although our distance became farther geographically, their
concern was getting closer. Since 2003, I spent lesser time with them. I know they
want me to be around, but they also hope that I would fight for my own career for a
brighter future, so they even encouraged me to go overseas. This is the priceless
love I will forever appreciate.
Yunnan Garden, Singapore

May 2016
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Dye-Sensitized Solar Cell (DSSC) . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.1 Development of Solar Cells . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.2 Basic Principle of DSSC . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.1.3 Introduction of the Constitution of DSSC . . . . . . . . . . . . . . 10
1.2 Photocatalysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1.2.1 Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1.2.2 Modification of Catalyst . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.3 Carbon Materials in Photoelectric Conversion System . . . . . . . . . . 20
1.3.1 Donor–Acceptor Photovoltaic Material Based
on Zero-Dimensional Fullerenes . . . . . . . . . . . . . . . . . .... 20
1.3.2 Application of One-Dimensional Carbon Nanotubes
in Photoelectric Conversion System . . . . . . . . . . . . . . .... 22
1.3.3 Applications of Two-Dimensional Carbon Materials
in Photoelectric Conversion System . . . . . . . . . . . . . . .... 24
1.4 Novelty and Significance of This Thesis . . . . . . . . . . . . . . . . .... 33
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 35
2 Two-Dimensional Graphene Bridges Enhanced Photoinduced
Charge Transport in Dye-Sensitized Solar Cells . . . . . . . . . . . . . . . . . 41
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.2 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
2.2.1 The Characterization of GO and Graphene . . . . . . . . . . . . . 42
2.2.2 Photocurrent–Voltage (I–V) Characteristics of Different
Electrodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.2.3 Incident Monochromatic Photo-to-Current Conversion
Efficiency (IPCE) Performance of Different Electrodes . . . . 46
2.2.4 Electrochemical Impedance Spectra (EIS) Measurement
of Different Electrodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
2.2.5 Operational Principle of Device . . . . . . . . . . . . . . . . . . . . . 51
xi
xii Contents
2.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
2.4 Postscript . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
2.5 Detailed Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
3 Bioinspired Stacking Structures for Photoelectric Conversion . . .... 57
3.1 Granum-Like Stacking Structures with TiO2–Graphene
Nanosheet for Improving Photoelectric Conversion . . . . . . . . . . . . 57
3.1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.1.2 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.1.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
3.1.4 Detailed Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
3.2 Stacking Nanostructures of Polyaniline with Graphene
Oxide as the Dopant and Template . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.2.2 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.2.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.2.4 Detailed Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.3 Postscript . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3.3.1 Discussion About Experiment Details . . . . . . . . . . . . . . . . . 76
3.3.2 Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4 Enhanced Light Harvesting in Plasmonic Dye-Sensitized Solar
Cells Using Gold Topological Light Trapping Layer . . . . . . . . . . . . . 81
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
4.2.1 Characterization of the Topological Ordered Au Film . . . . . 83
4.2.2 Photovoltaic Behavior of DSSCs . . . . . . . . . . . . . . . . . . . . 84
4.2.3 Light Harvesting Properties of Two Electrodes . . . . . . . . . . 86
Efficiency (IPCE) Performance of Different Electrodes . . . . 86
4.2.5 Photovoltaic Behavior Under Different Light Intensity . . . . 87
4.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
4.5 Postscript . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
5 Photocatalytic Properties of Graphdiyne and Graphene
Modified TiO2: From Theory to Experiment . . . . . . . . . . . . . . . . . . . 93
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
5.2.1 Theoretical Analysis and Structural Characterizations . . . . . 94
5.2.2 Performance in Photocatalytic Degradation . . . . . . . . . . . . . 101
5.2.3 Principle Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Contents xiii
5.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

5.5 Postscript . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
6 Conclusions and Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Abbreviations
AFM Atomic Force Microscopy

CBM Conduction Band Minimum
CNT Carbon Nanotube
DSSC Dye-Sensitized Solar Cell
EIS Electrochemical Impedance Spectroscopy
FT-IR Fourier Transform Infrared
FTO Fluorine-doped Tin Oxide
GD Graphdiyne
GO Graphene Oxide
GR Graphene
HOMO Highest Occupied Molecular Orbital
IPCE Monochromatic Incident Photon-to-Electron Conversion Efficiency
LUMO Lowest Unoccupied Molecular Orbital
PANI Polyaniline
SEM Scanning Electron Microscope
SPR Surface Plasmon Resonance
TEM Transmission Electron Microscopy
UV–Vis Ultraviolet–Visible
VBM Valence Band Maximum
XPS X-ray Photoelectron Spectroscopy
XRD X-ray Diffraction
xv
Chapter 1
Introduction
As an important foundation for human society, the history of using energy can reflect
back the history of human development. How to high efficient using energy is directly
related to the sustainable development of human society. With the growing global
population and improving of life expectancy, energy issue has become the most
urgent problem we are facing nowadays. In the next decades, exploring of new solu‐
tions to meet the growing energy demand, will be one of the greatest challenges for
all human being. Besides the considerable damage to environment of the traditional
nonrenewable energy, the reserves of it are declining with the exploitation. In retro‐
spect, the solar energy is always the most basic resource for human survival and
development, as a clean energy which can be used permanently, also has great
development potential. From the development of technology, the development and
utilization technology of solar energy will likely decide the future of human life. Up
to date, the research of using solar energy is focused on solar cells, water photoca‐
talysis and photo-oxidation decomposition of organic matter, etc. By the knowledge
obtained nowadays, synergistic effect always gets a good performance in nanoma‐
terials. Hence, how to complex functional molecules with other ingredients, and then
preparative high performance nanocomposites is gradually become a hot research
area. And this kind of composite materials in energy and environment field will lead
humanity toward the green and better future.
© Springer-Verlag GmbH Germany 2017 1

N. Yang, The Preparation of Nano Composites and Their Applications
in Solar Energy Conversion, Springer Theses, DOI 10.1007/978-3-662-53485-4_1
2 1 Introduction
Fig. 1.1 This consisting of the world’s total primary energy consumption in last 25 years [2]
1.1 Dye-Sensitized Solar Cell (DSSC)
1.1.1 Development of Solar Cells
In the current world energy mix, the representative of the fossil energy, coal, oil,
natural gas is still holding the dominant position [1, 2]. In 2014, the world’s total
primary energy consumption is 12928.4 million tons oil equivalent. This consump‐
tion consisting of: oil 32.6 %, coal 30.0 %, natural gas 23.7 %, nuclear 4.4 %, hydro‐
electricity 6.8 %, and new renewable energy 2.5 % (Fig. 1.1) [2]. Considering the
keep increasing of energy consumption, the reserves declining of traditional nonre‐
newable energy, and the huge potential risk of nuclear power, developing and
utilization of new clean energy is particularly important for the survival and sustain‐
able development of mankind. Herein, solar energy, because of its wide distribution,
green friendly, long-term use and abundant reserves, has drawn a worldwide atten‐
tion, which makes the solar cells become one of the world’s major energy in future.
However, the instinct defects of solar energy, such as the energy density is low and
unstable, make it still difficult to widely use. Herein, a requirement of designing a
more efficient solar energy conversion material is urgent to access for large-scale
applications.
1.1 Dye-Sensitized Solar Cell (DSSC) 3
1.1.1.1 Classification of Solar Cells
The effective use of solar energy including three parts of light capture, photoelectric
conversion, and energy storage. The basic and core unit of photovoltaic generation
is solar cell. There are three process in the photovoltaic conversion in the principle
of solar cells: they are (a) absorption of light with specific energy, which can produce
photogenerated electron–hole pairs (photogenerated carriers); (b) separation of
photogenerated electrons and holes; (c) different charged photogenerated carriers are
moved to different electrodes, forming current. So far, the solar cells have developed
into many branches with various forms.
The solar cell can be divided into homojunction, heterojunction solar cell and
Schottky solar cells by structural differences. Homojuction solar cell is defined by
the p-n junction cell with a same semiconductor, such as traditional silicon cells,
GaAs cells, etc. By extending this single semiconductor to two different band gap
semiconductors, which have a p-n junction at the contact interface, thereby resisting
electron–hole recombination, it is named as heterojunction solar cells. Schottky solar
cells is using the “Schottky barrier” to build cells, which introduces an interface with
metal and semiconductor, hence it is also known as MS solar cell.
Distinction from the building materials, solar cells can be divided into silicon
solar cells, inorganic semiconductor solar cells, sensitized nanocrystalline solar cells,
organic solar cells, and polymer solar cells. The sensitizer of sensitized nanocrys‐
talline solar cells can be applied by dye, narrow band gap semiconductors, transition
metal ions, or noble metals, etc.
A solar cell can be also divided into conventional solar cells (silicon cells, CdS
battery, GaAs battery, etc.) and exciton solar cells by various photoelectric conver‐
sion mechanism. The latter one refers to a kind of cells which can absorb photons,
reaching an excited state and transfer electrons as follows. The materials for light
absorption and electron transfer are undertaken by different materials.
1.1.1.2 Development of Solar Cells
The photovoltaic effect was first observed by French physicist A.E. Becquerel in
1839 by immersing Ag/AgCl electrode into dilute hydrochloric acid solution, and
voltage was observed under light. In 1876, William Grylls Adams discovered that
when selenium was exposed to light, electricity is produced. Although it was not
efficient, it proved that light could be converted into electricity. It is named as Sele‐
nium Cell. Then Cu–Cu2O photo-voltage cell was reported in 1928, and one year
later the solid energy band theory was proposed, which provided a theoretical basis
to converse solar energy to electrical energy. At 1930, Bruno Lange gave an idea to
design photo-voltage cells by using photo-voltage effect, but the efficiency is less
than 1 % at that moment. Another problem for it was these cells cannot work long
under strong sunlight. For these reasons, some experts thought solar cells would
never be a good way to create electricity. It was not until the 1940s that people
became interested in solar electricity generation again, because of the good
4 1 Introduction
performance observed on Si. In the early 1950s, Calvin Fuller and Gerald Pearson,
two scientists who worked at the Bell Laboratories in the USA, were trying to
improve silicon transistors for electrical equipment, then an efficiency was reached
to 6 %. Here comes the first generation of solar cells and the practical photovoltaic
technology. In 1955, the theory of solar cell is established, since then the solar cell
began developing rapidly from theory to practice. One of the driving force of solar
cells is the exploration of space. Since 1958, the first time of using solar cells by
“Pioneer” satellite, so far, almost all the space aircraft are equipped with solar panels.
And another driving force is the world oil crisis in 1973. The application of solar
cells landed on earth from space. Ground application of solar energy draws extensive
attention by everyone, and a booming development is coming [3–16].
Nowadays, the photoelectric conversion efficiency of monocrystalline silicon
solar cells can reach 24 % in laboratory, also a high efficiency of 15 % can be achieved
in large-scale production, which make it the most developed solar cells with the
highest efficiency. However, the high cost resists its widely application, hence, poly‐
crystalline and amorphous silicon thin film solar cells are adapted as a result. With
a lower production costs, polysilicon is more suitable for mass production of large-
size ingots. Currently, the laboratory efficiency of polysilicon solar cells can reach
20 %, while 10 % is achieved in industrial scale production. It is now becoming the
leading products on the market.
Although the lifetime is shorter than monocrystalline ones. The drawback of
polycrystalline silicon solar cell also includes it requires large amounts of high-purity
silicon material during the fabrication. While the manufacturing process is very
complex and the power consumption is also large, which needs more than half of
the cost in producing solar cells. To decrease the request of high-quality silicon,
amorphous silicon is studied these years. It is an alloy of silicon and hydrogen, which
gives a facile process and lower consumption of power, but further work is necessary
to increase the efficiency and stability [17].
The band gap of inorganic semiconductor is easily to be tuned between 1.0 and
1.6 eV, which is well matched to the solar spectrum. It makes the inorganic semi‐
conductor a good candidate for thin film solar cells. Considering the strong absorp‐
tion in visible light spectrum, only a few nanometers of semiconductor film can
absorb most of the solar light. Among them, GaAs is one of the good candidates with
a high efficiency of 28 %, but due to the high cost, it can only be used in space. While,
CeTe and CuInSe2 are both show the possibility for civil use, but the disadvantage
is also clear. The toxic of Cd and the limited source of In and Se all resist its appli‐
cation [17].
At this stage, scientists are trying to instead the inorganic materials to organic
ones, because it is more facile to design, control and functionalize the molecular,
which give a new path to “synthesize” the solar cells. Started from metal phthalo‐
cyanine, progressive developed to porphyrin battery, pyrene red/phthalocyanine and
pentacene, until the novel conjugate polymer, the photoelectric conversion efficiency
of organic solar cell can reach 11 % in lab. Comparing with the Si-based or inorganic
semiconductor solar cells, organic solar cells are cheaper, easily obtained, flexible
and suitable for the large-scale fabrication, but the drawback of it is also obvious.
Fig. 1.2 The structure of a

dye-sensitized solar cell
This kind of solar cell is easy to aging and the carrier transport efficiency is low, also
the high resistivity resists the further enhancement of the conversion efficiency.
Among them, by compositing the conjugate polymer and carbon material (C60,
carbon nanotube or graphene) shows a better potential for increasing the performance
and decreasing the fabrication cost.
Since Grätzel introduced nanotechnology into the field of solar cells in 1991,
DSSC is developing explosively [18]. Until now, a confirmed record of 15 % (EPFL,
Oxford) was achieved. And this kind of solar cell shows a good stability under an
even high temperature [19]. Besides, there are more benefits for this kind of solar
cell: low cost, controllable shape, abundant color, and transparency. These all give
an attraction for the commercial developing. But the biggest problem is the incom‐
parable efficiency with Si, and also the difficult in sealing the cell. In this 2 years,
perovskite solar cell is a rising star in the field of solar cell, which mechanism is
quite similar to DSSC. And consider this thesis is quite related to DSSC, hence I will
describe more about the structure and mechanism of it in the following text.
1.1.2 Basic Principle of DSSC
1.1.2.1 Structure of DSSC
DSSC is also a kind of heterojunction solar cell, which electrons and holes will be
separated by two chemicals. In this cell, organic dye molecules are first absorbed on
the surface of semiconductors. By using the light irradiation of dye, the photoexcited
electrons are captured by semiconductors, realizing the separation of holes and elec‐
trons. The separated carriers are transferred to different electrodes, forming current.
The composition of it includes five parts as: transparent conductive substrate, porous
nanocrystal semiconductor film, photosensitizer (dye), electrolyte, and counter elec‐
trode (Fig. 1.2).
6 1 Introduction
Fig. 1.3 The working principle of a dye-sensitized solar cell a thermodynamics and b kinetics
[19a]. Copyright © 2010, American Chemical Society
1.1.2.2 Principle of DSSC
The well acceptable and widely used principle of DSSC is analyzed by using ther‐
modynamic method by Ardo and Meyer (Fig. 1.3a) [20, 21]. While the kinetics is
more important cannot be ignored, which tells us what can we do to increase the
efficiency of DSSC (Fig. 1.3b) [22].
Considering the band gap of TiO2 is 3.2 eV, which means it can only absorb the
ultraviolet light with the wavelength smaller than 387 nm. While it has no response
to the visible light, most of the power in sunlight. Hence, it needs to absorb a layer
of dye molecular, which can be excised by the visible light. The basic mechanism
is: (1) under illumination, the energy of light is absorbed by dye molecules and the
electrons are excited from ground state to excited state. (2) The excited electrons are
not stable, which are easily injected to the conductive bond of TiO2. Due to the loss
of electrons, the dye is oxidized. (3) The TiO2 captured electrons are transport to the
photoanode through TiO2 network and then flow to the counter electrode through
external circuit, forming current. (4) The oxidized dye gets the electrons from elec‐
trolyte, being reduced and regenerated. (5) The electron-deficient electrolyte will get
the electron (the one transferred from dye to TiO2) from counter electrode, finishing
the circle. (6) During the circle, the electrons at the conductive bond of TiO2 will
also be transferred to the electrolyte, named as back reaction. Some key processes
will be described as following:
Charge injection and relaxation (Fig. 1.3a, reaction 1 and 2). During the principle
study of DSSC, scientists have found the excited electrons can transfer from the
excited Ru complex to the conductive bond of TiO2 with a surprising fast speed
(Reaction 2). Although there are still some disputes on the detailed mechanism, the
fact we can accept is this transport process will be finished during a femtosecond
level, while the relaxation time from excited state to ground state is about 20–60 ns
(take Ru dye as an example) [23, 24]. This 1000 times faster during charge separation
ensures the possibility of the DSSC working on the basic level.
Dye regeneration (Reaction 3). The lifetime of dye cation is in a milliseconds
period, and it can be reduced by electrolyte (I−/I3−) in a microsecond period, hence
this process will not affect the final efficiency too much. The widely used Ru dye
can be high efficiently excited/regenerated with 108 circles, which means it can be
continually used in 20 years [25]. But it should not ignore that the holes produced
by TiO2 (the ultraviolet light of sunlight can also excited TiO2), which would oxidize
the absorbed dye, changing the property of interface, and decreasing the efficiency
of the solar cell as a result. Hence, it is important to decrease the flux of ultraviolet
light in the future application. Besides, it is also necessary to ensure the best dye
absorption is single layer absorption, cause only the attached dye molecular can
transfer the electrons to TiO2.
Electron transports in TiO2 (Reaction 4). After introducing nanocrystal TiO2 into
DSSC, one of the unclear principles is how can the electrons transport so fast and
high efficient through TiO2 matrix. Based on the common knowledge, compare with
the bulk materials, there are some disadvantages in the nanocrystal film: (1) low
conductivity; (2) such small nanoparticles cannot support the required built-in elec‐
tric field; (3) because the electrolyte is diffused in TiO2 anode network, which will
bring lots of interfaces. Theoretically speaking, these are all disadvantages for solar
cells [19]. Until now, we can only consider the network as a whole. Electrons are
jumping in the network, from one particle to another. While the real transport mech‐
anism is still under research. How do the electrons transport? What and where are
the traps during the transport? Have to say, they are still unclear.
Recombination between excited electrons and dye (Reaction 5). Kinetic analysis,
the recombination between excited electrons and dye is strongly related to carrier
density and photointensity. In general, the recombination time is in microseconds
to milliseconds. While this reaction is competition with the process of dye reduction
by electrolyte. The latter reaction rate is 100 times higher than the former one, which
means the final efficiency will not be affected by this recombination process. During
this process, considering the competition of these two reactions, the diffusion of
8 1 Introduction
electrolyte through TiO2 network is very important. Hence, it is necessary to opti‐

mize the structure of photoanode. That also means, in the quasi-solid electrolyte,
the diffusion factor is larger than liquid one, which makes the recombination easy
to take place.
Recombination between excited electrons and electrolyte (Reaction 6). The life‐
time of excited electrons is always related to this recombination process. Its lifetime
can reach 1–20 ms in I−/I3− system, which proves an efficient work of solar cell.
However, it is undeniable that the process is the main way of electron loss, how to
suppress the recombination is a key factor to enhance cell efficiency.
Regeneration of electrolyte on counter electrode (Reaction 7). This regeneration
is an easy process in the presence of catalyst. The most used catalyst is Pt nanocrystals
on the counter electrode, which performs a low resistance as <1 Ω/cm2 and a long
period stability.
1.1.2.3 Parameter of DSSC
In order to evaluate the performance of DSSC, a series of characterization parameters

was set up, which can analyze its performance in situ. Based on these characteriza‐
tions, we can know: (1) the internal behavior of the cells under a standard sunlight;
(2) building a theoretical model to elaborate electron transfer and transport behavior;
(3) analysis of the stability and reproducibility of solar cell; (4) optimization of the
components of the cell material. Common characterization methods are shown
below:
Current–Voltage (I–V or J–V) curve It is the most widely and directly used char‐
acterization to perform the cell performance. Because the high interface capacitance
in DSSC results a slow current response to applied voltage, it often adopts a lower
scan speed during the I–V measurement. During the measurement, the short circuit
current (Isc) is the current number when the applied voltage is 0, while the open circuit
voltage (Voc) is referred to the voltage number when the current is 0. The maximum
value of the product by current and voltage is named as the output efficiency. Photo‐
electric conversion efficiency can be calculated by the following equation:
|IV|max I V FF
𝜂= = SC OC (1.1)
Pin Pin
where Pin is referred to the intensity of solar irradiation. FF is named as the fill factor,
which is a value between 0 and 1. The larger of FF indicates the more “squareness”
of the I–V curve. The value of FF is affected by the internal series resistance of
device, reverse current, working voltage, current and et al. Voc is determined by the
difference between quasi-Fermi energy level and Nerst potential of redox pair (elec‐
trolyte). Isc is mainly multifaceted impact by the number of excited dye, the number
of electrons injected into TiO2, electron transmission efficiency.
Incident photon to current efficiency (IPCE) The spectral response property can
be reflected by IPCE measurement. IPCE is the ratio of the electrons number of
external circuit to the number of incident photoelectrons. Usually, the test is
recording short circuit current under different monochromatic. Direct current (DC)
and alternating current (AC) are two methods that often used. Compare with DC
method, AC one is more precise. Equation 1.2 shows the factors that affect IPCE.
IPCE (𝜆) = LHE (𝜆) ⋅ 𝜑inj (𝜆) ⋅ 𝜑reg ⋅ 𝜂CC (𝜆) (1.2)
where LHE is light harvesting efficiency, known as LHE (𝜆) = 1 − 10𝛤 𝛿(𝜆). Γ is the
mole number of dye which is coated on the photoanode in the unit area. 𝛿(𝜆) is the
cross-sectional absorption area of dye. 𝜑inj and 𝜑reg are the efficiency of electrons
injection and regeneration, respectively. 𝜂CC is the efficiency of charge collection,
which is mainly affected by the recombination.
Electrochemical method Specifically, there are three kinds of methods, including
cyclic voltammetry, differential pulse voltammetry, square wave voltammetry. Elec‐
trochemical method is the standard method used to calibrate the energy level of dye
in DSSC. However, this test is usually done by dissolving the chosen dye in the inert
electrolyte, while not adsorbed it on the metal oxide, which makes the results of
actual bias, not accurate enough.
Electrochemical impedance spectroscopy (EIS) EIS tests the response of cell in
amplitude and phase by adding a small sine wave disturbance. The charge transport
property can be induced by fitting the resistors, capacitors, inductors, and other
parameters at different frequencies. EIS method is an in situ way to study the behavior
of the solar cell including photo anode, counter electrode, and also electrolyte. By
adapting EIS measurement, one can get the series resistance, charge transfer resist‐
ance on the counter electrode, diffusion resistance in electrolyte, electron transfer
resistance and charge recombination in photo anode. Meanwhile the capacitor of
metal oxide thin film can be also calculated.
Spectroelectrochemistry This method is adopted by combining electrochemistry
measurement with different spectrometers. For example, in conjunction with UV–
vis spectroscopy, the behavior of charge aggregation in anode can be studied and
then the conduction band edge of the oxide electrode is calculated. The advantage
is that this method can truly reflect the energy level at the moment of photoelectric
conversion on the semiconductor electrode. However, this method can only be intro‐
duced to an optical transparency semiconductor electrode, and the extinction coef‐
ficient of localized electrons or free electrons must be known.
Photoelectrochemical method This emerging technology makes the in situ test
DSSC photovoltaic behavior possible. Compared with the EIS method, this method
does not require too much fitting results which can obtain the results in a simple and
fast way from the complex system. Specific test methods include: (1) intensity-
modulated photovoltage spectroscopy (IMVS) and intensity-modulated photocur‐
rent spectroscopy (IMPS), which can obtain the electron transit time; (2) Electron
10 1 Introduction
life test; (3) charge concentration test; (4) quasi-Fermi level test; (5) charge diffusion
length and collection efficiency test; (6) photo-induced absorption (PIA) Spectro‐
scopy.
1.1.3 Introduction of the Constitution of DSSC
1.1.3.1 Nanocrystal Photoanode
Because the high specific surface area of nanomaterials ensures that the abundant
dye absorptions, it gives the possibility to use it in DSSC. Since 1991, Grätzel intro‐
duced nano titanium oxide into the anode of DSSC [18], it has brought a revolu‐
tionary change in the development of the battery. Until now, TiO2 is the most efficient
material in DSSC, but such as ZnO, SnO2, Nb2O5, and other metal oxides also be
introduced to photoanodes for research. Besides these simple oxides, others, such as
ternary oxide SrTiO3 and Zn2SnO4, as well as core–shell structure, as ZnO@SnO2,
also draws attentions of scientists [26–29].
Recent years, scientists are working on regulation the morphology of nanomate‐
rials, aiming change the disordered nanoparticle aggregates to nanotube or single
crystal nanorod arrays. These studies are all aimed to make a faster electron trans‐
mission.
Due to the energy-level matching, high dielectric constant, high refractive index,
corrosion resistant and low cost, TiO2 is considered as a good candidate for photo‐
anode. Since 1991, the improvement of TiO2 anode is mainly focused on how to
enhance the light absorption and scattering, promote the charge separation, suppress
the electron hole recombination, and also enhance the interface energy, etc. So far,
the best performing DSSC has the following characteristics: (1) a 50 nm thick tita‐
nium dioxide insulating layer which can prevent direct contact with the electrolyte
and transparent conductive substrate, such as FTO; (2) an approximately 10–15 μm
nanocrystalline film, which can provide a large surface area to adsorb dye and
transfer electrons; (3) a light scattering layer with 1–3 μm thick after depositing of
the former nanocrystalline film. 400 nm titanium oxide particles are usually used
here; (4) further modification of titanium oxide always with TiCl4, which can form
small branched nanocrystals on the fabricated film [30–33].
Highly ordered mesoporous materials will greatly enhance the optical roughness
of the anode, thereby enhancing dye adsorption amount. Zukalova et al. prepared a
photoanode with 6 nm TiO2 by using P123 [34]. The rough factor is increased to 460
at a thickness of 1 μm, with a photoelectric conversion efficiency of 4 %, which is
much higher than the ordinary photoanode (2.7 %) under the same conditions.
In order to build highway for electrons transport, how to fabricate or assembly
TiO2 nanotubes is a hot research area for scientists. In the previous work, nanotube
is usually synthesized on a titanium foil, but the solar light cannot irradiate from the
photoanode. The length of these nanotubes is typically less than 1000 nm, and the
pore size, wall thickness, distance between tubes can be modulated by experiment.
The photoelectric conversion efficiency can up to 6.9 % by this technology [35]. But
the drawback of this method is obvious, where the light is irradiated from counter
electrode, which is deduced by electrolyte before reaching photoanode, hence the
efficiency enhancement is resisted. By modification this method, growing TiO2
nanotube on FTO is a better candidate [35–42]. With a length of 1.1 μm, the efficiency
has already reached 4.1 % [38]. If the length can keep increasing, a competitive
efficiency will be achieved. Based on the initial experimental proof, TiO2 tube has
high electron diffusion length as to 100 μm, which provides a guarantee for the
preparation of thicker DSSC.
Moreover, atomic layer deposition technique can be adapted to get very ordered
nanostructures currently. By using this technology and combining with aerogel
template, photoelectric conversion efficiency is reached to 4.3 % [43].
Recent years, how to prepare a photoanode at low temperature gradually become
a hot research area, because one of the advantages of DSSC is the prospect in
preparing flexible battery. Lindström et al. using a pressing method obtains a 3 %
efficiency with commercially P25, following by Arakawa enabling efficiency to 7 %
by optimize processes [19].
At present, zinc oxide, is considered as the second best light anode material. It
has a similar band gap with titanium and also the conduction band edge. Although
it is not as stable as TiO2 after the fabrication of DSSC, the electronic transmission
is even faster. Also zinc oxide is easier to get different morphology, which gives the
benefits to study the relationship between structure and properties, and thus it attracts
the attention of many researchers [44–57].
In addition, scientists have come to realize the synergies with other oxides can
improve the performance. Wherein surface passivation is a commonly used method.
For example, by coating the surface of SnO2 with a thin layer of Al2O3 can increase
the photo-voltage, thus enhancing efficiency, but excess coating will decrease the
current. With this method, Nb2O5, Al2O3, SiO2 and ZrO2 are all used to modify
TiO2 [58–66]. Furthermore, t-butyl pyridine and guanidine isothiocyanate is added
to the electrolyte to use as a passivation material, and thus enhance the efficiency.
1.1.3.2 Dye
As a key part of DSSC, the dye must meet the following characteristics [19]:
(1) Absorption spectrum must cover the entire visible light region, and even the
near infrared region;
(2) Anchor group (–COOH, –H2PO3, –SO3H, etc.) must exist in dye molecules. It
can make dye molecule strongly adsorbed on the metal oxide [67–69];
(3) Matched energy level between dye and metal oxide semiconductor. For n-type
DSSC, the excited states of the sensitizer must be higher than the conduction
band of the metal oxide; on the contrary, for the p-type DSSC, the HOMO
energy level of dye must be lower than the valence band of p-type semicon‐
ductor (Fig. 1.4);
12 1 Introduction
Fig. 1.4 A schematic Energy

idealized energy level diagram
e
of n- and p-type DSSC e
indicating the desired electron
transfer processes [19a].
Copyright © 2010, American
Chemical Society e
e VOC
+
e
+
n-type semiconductor dye electrolyte dye p-type semiconductor
(4) Energy level of excited dye must be lower than the oxidation–reduction poten‐
tial of electrolyte (Fig. 1.4);
(5) Monolayer adsorption of the dye is accepted, means the aggregation of dye is
detrimental [70];
(6) Stable under light, and a certain thermal and chemical stability is favorable.
Based on these requirements, in the past few decades, metal complex, porphyrin,
phthalocyanine, and organic dyes are used in the DSSC.
Metal complexes It is the most widely used dye, among them, since ruthenium
complexes have good spectral absorption and photoelectric conversion capability, it
is currently widely studied [71–73]. Generally, such dye molecules are composed of
a metal center and ligands contain anchor groups. Such molecular can absorb sunlight
is due to the metal—ligand charge transfer mechanism. Currently, these ligands are
mostly modified on bipyridine or terpyridine. This modification, is usually taken by
alkyl, aromatic or heterocyclic groups, thereby improving its photophysical or
photochemical properties, so as to enhance the efficiency. Excited electrons can be
transferred to the conduction band of metal oxide through anchor group. It is hoped
to optimize of the structure and composition of the dye and thus enhance the electron
injection and regeneration ability. Also, Os, Re, Fe, Pt, Co, and Cu are reported as
central atoms in synthetic dyes [74].
Porphyrin, phthalocyanine The biggest drawback of Ru dyes is the limited
absorption in near-infrared region of sunlight. While porphyrin and phthalocyanine
have good photoelectric response in the near infrared region, and they have good
heat, light and chemical stability, as a good candidate of dye for DSSC. Porphyrin
is present in green plants as a natural dye molecule, it is very interesting to attempt
it in DSSC. Porphyrin not only has a good energy level matching with TiO2, but has
a strong absorption at 400–450 nm wavelength, also 500–700 nm due to the Q band
absorption. It guarantees the porphyrins can be used as a whole spectrum sensitizing
agents [75–79]. Phthalocyanine, due to the strong absorption of Q band at 700 nm,
can be used as the sensitizer in near infrared. But because of the poor solubility of
Fig. 1.5 Schematic drawing

of D-π-A organic dye [19a].
Copyright © 2010, American
Chemical Society
phthalocyanine molecule, it needs a further modification in subsequent experiments.

While the biggest weakness is it is easy to aggregate and suppress the charge transport
in this case, hence other groups or additives are necessary to enhance the performance
[80–95].
Organic dyes Compared to the most outstanding ruthenium dyes, organic dye has
the following advantages: (1) easily to design and synthesis molecule; (2) lower cost
than metal complex dyes and environment-friendly; (3) higher molar extinction
coefficient; (4) better performance in p-type DSSC. In general, organic dyes have a
donor-π-acceptor (D-π-A) structure (Fig. 1.5) [19]. It is possible to design and
synthesize new structures to absorb light more effectively, to adjust the HOMO and
LUMO position, completing the intramolecular charge separation.
Summary Over the past few decades, scientists have synthesized hundreds of dye
molecules, by summarizing the common properties, a guideline for future design is
provided [19]:
(1) Ruthenium complex. Take bipyridine ruthenium for example, different pyri‐
dines play different roles in this complex. One molecule of pyridine takes charge
of anchoring dye molecule on the surface of TiO2, while ensuring the good
coupling between excited dye and the conduction band of TiO2. Also, the
LUMO of dye should place in this ligand. Another pyridine ligand aims to
improve spectral utilization and molar extinction coefficient. Other remaining
ligands are usually modified by—NCS to enhance the HOMO of the dye,
thereby enabling the redshifted of dye absorption.
(2) Organic dyes have diverse structures, herein, it is difficult to find a common
strategy to guide the molecular design. Generally speaking, it is necessary to
ensure that the HOMO and LUMO orbitals of a molecule are separated in space.
The linkage of electron donor and acceptor by rigid conjugated π bond will
facilitate the design to get a more efficient molecule. But one thing needs pay
attention is such modification often leads to energy reconstruction, which will
bring a negative impact in solar cells.
(3) Introducing a large group at a suitable position is a key factor for organic dye
to enhance the efficiency of DSSC. Under an ideal condition, the dye is required
a monolayer adsorption and wholly covered the surface of TiO2, which prevents
the contact between TiO2 and electrolyte;
(4) Good interaction between dye and electrolyte. Usually a special functional
group for absorbing electrons from electrolyte is required, where the holes are
reduced;
(5) It is necessary for the dye to firmly adsorb on the metal oxide, and keep a long-
term stability;
14 1 Introduction
(6) A long life time at the oxidation state is required for the dye, and also stable at
high temperature to some extent.
1.1.3.3 Electrolyte
Comparing with the former two parts, the development of the electrolyte is relatively
slow. Although liquid, ionic liquids, gel and solid-state electrolyte have been devel‐
oped, but the most widely used is still liquid electrolyte based on iodine and lithium
iodide [96]. Liquid electrolyte must have the properties as following: a good chem‐
ical stability, a low viscosity, easy to diffuse, easily to dissolve both of the redox
pairs, and also it cannot make the dye desorption or exfoliate TiO2 film. Because
mostly used dye is still metal complex one, they are typically hydrophilic. Hence the
water and protonic solvents are not suitable as a solvent, which will destroy the
absorption between dye and TiO2. At the same time, due to the possibility of elec‐
trolyte evaporation and leakage, packaging technology is becoming a hot field. The
most widely used solvent nowadays, acetonitrile, is easily to corrode the sealing
material, hence it is not conducive for long-term use.
In order to solve the evaporation problem of solvent, scientists consider to design
a solid electrolyte, such electrolyte is also known as a hole transporting material.
The requirements for the solid electrolyte are: no absorption to visible light, good
electrical conductivity, perfectly match between electrolyte and the HOMO level of
dye. This research direction began in 1995 by using CuI as electrolyte, and an effi‐
ciency of 3 % was achieved as a result [97]. After that, CuSCN, CuBr, and other
inorganic substances were reported one after another, but due to light degradation,
recombination, and ambiguity mechanism, they did not bring significant improve‐
ment to the efficiency. [98–100] Besides, small organic molecules have also been
used in preparation of the solid electrolyte. N2,N2,N2′,N2′,N7,N7,N7′,N7′-octakis(4-
methoxyphenyl)-9,9′-spirobi[9H-fluorene]-2,2′,7,7′-tetramine (Spiro-MeOTAD) is
the first reported small organic molecule electrolyte, but the efficiency is only 0.74 %
[101]. Thereafter, with a series of modified and optimized, the maximum efficiency
based on organic solid electrolyte has up to 5.1 % [102, 103].
Based on the ideas in designing liquid and solid electrolytes, people developed
an intermediate electrolyte between the liquid and solid, quasi-solid electrolytes.
Such organic electrolyte is obtained by gelling or polymerizing the solvent or ionic
liquid, or directly disperse the redox in a polymer. The mechanism of charge trans‐
port in such electrolyte (most remained I−/I3− redox couple) is the molecular diffusion
rather than charge jumping. Currently, the highest photoelectric conversion effi‐
ciency based on gel electrolyte is up to 5.6 %, and the ionic liquid-based electrolyte
has up to 8 %, which is close to the liquid electrolyte. The limiting factor for the
efficiency of the quasi-solid electrolyte is mainly due to its own liquidity and diffu‐
sion capacity [104].
In addition, some additives in the electrolyte also played a very important role.
The current understanding to these additives is still not deep enough. The analysis
is often speculated through phenomenon. Most of the improvements in performance

are attributed to reasons as optimizing the redox potential of the semiconductors,
modification of the band edges, surface modification of nanomaterials, the amend‐
ment of the interface of dye, and so on. The most effective additive is t-butylpyridine
(TBP), which can significantly improve the open circuit voltage of DSSC. IMVS
measurement shows that, the addition of TBP can increase the conduction band of
titanium oxide, and thus increase the open circuit voltage, but the specific reason is
still not total understanding. Moreover, since the guanidine isothiocyanate, pyridine
ethoxymethyl, methyl-benzimidazole, etc., can suppress the recombination to some
extent, or prolong lifetime, they are also added into the electrolyte [105–109].
1.1.3.4 Counter Electrode
Counter electrode is also known as photocathode. Besides performing as the electron

transfer pathway, it also acts as a catalyst for the quasi-solid or liquid electrolyte.
Pure platinum is firstly used as a counter electrode, but due to the high cost, people
gradually began to use other materials instead, such as the most commonly method
by depositing platinum on conductive glass. The deposition methods include elec‐
trodeposition, spray pyrolysis, sputtering and vapor deposition method, etc. [1], one
of the simplest and most effective way is by thermal decomposition of chloroplatinic
acid. In comparison, the electro and vapor deposition methods always make the
exfoliation of platinum, which is not adopted.
Since the carbon material also has excellent electrical conductivity, catalytic
ability, and lower prices, it is also used as counter electrode for DSSC. Such as
activated carbon, carbon black, graphite, carbon nanotubes have been reported [110–
113]. Doped poly(3,4-ethylenedioxythiophene) (PEDOT) is one of the few polymers
can be used on counter electrode with a good performance [114]. Although relevant
papers also reported using polyaniline or polypyrrole as counter electrode, but they
are less effective [115, 116]. Recently, Wang reported the work of deposition CoS
on flexible electrode with a sheet resistance of 1.8 Ω/cm2, which is comparable with
the platinum electrode under the same conditions resistance (1.3 Ω/cm2) [117].
1.2 Photocatalysis
1.2.1 Principle
Another significant role of titanium dioxide is photocatalysis. As known, semicon‐

ductor can be excited with a certain wavelength, while the electrons are excited from
the valence band to conduction band, forming excited electrons and leaving holes
on the valance band. These electrons and holes with a certain energy can be freely
migrate to the surface of the catalyst. After that, they can react with adsorbed
substances on the surface, and thus produce large amounts of highly active radicals
16 1 Introduction
to decompose organic matter. On the other hand, these photogenerated electrons and
holes are not stable, which are easily to recombination and release energy with heat,
causing the decreasing of photocatalytic ability. In facts, the photocatalytic efficiency
depends on the competition between carrier mobility and recombination of electron–
hole. In TiO2, the life time of electrons and holes is as long as ~250 ns, which means
there is enough time for the carriers to migrate to the surface of the crystals, formation
of a variety of radicals, where the hydroxyl radical is the most common one. A
possible mechanism of photocatalytic reaction of TiO2 is speculated by Martin et al.
with the measurements of time-resolved microwave conductivity (TRMC) [118]:
First, light with specific wavelengths/energy (hυ ≥ 3.2 eV) are absorbed by TiO2,
generating electrons and holes
TiO2 + h𝜈 → TiO2 (e−CB + h+VB ) (1.3)
In the picosecond period, photogenerated electrons migrate to the surface of

TiO2, captured by the surface, forming Ti3+
Ti4+ + e−CB → Ti3+ (1.4)
The photogenerated electrons can be effectively captured by oxygen molecules

adsorbed TiO2 surface, preventing the formation of a large amount of Ti3+ center.
At the same time, O2 can facilitate oxidize Ti3+ to Ti4+, reducing itself to an anion
radical
O2 + e−CB → O⋅−
2 (1.5)
O2 + Ti3+ → O⋅−
2
+ Ti4+ (1.6)
Meanwhile, hydroxyl ions (OH−) and water molecules (H2O) adsorbed on TiO2
surface are oxidized by holes on the valence band of TiO2, forming an oxidant,
hydroxyl radicals
h+VB + OH− → OH⋅ (1.7)
h+VB + H2 O → OH⋅ + H+ (1.8)
Organic are attack by the radicals, degradation reactions take place
O⋅−
2
∕OH⋅ + Organic → nCO2 + H2 O (1.9)
The above reaction takes place within nanosecond period, accompanying with a
recombination reaction of photogenerated electrons and holes
e−CB + h+VB → E (1.10)
Ti3+ + h+VB → Ti4+ (1.11)

1.2 Photocatalysis 17
(a) Metal (b)

e
CB e CdS CB
e e e e e e e
EF Acceptor hυ
hυ
TiO2
VB h
h VB
Semiconductor
O
R
Fig. 1.6 Schematic representation of electron transfers of a noble metal or b semiconductor coated
TiO2 [122]. Copyright © 2010, American Chemical Society
Therefore, how to extend the life time of photogenerated electrons and holes,
suppress recombination and stable radical, are very important to enhance the photo‐
catalysis capacity.
1.2.2 Modification of Catalyst
1.2.2.1 Suppress Recombination of Photogenerated Electrons and Holes
Noble metal deposition is the most common method to suppress the recombination
of photogenerated electron–hole pair, extending the lifetime of electrons and holes
[119]. Typically, this method is by depositing a noble metal on titanium dioxide,
where the work function of noble metal should be lower than the conduction band
of TiO2. By introducing this junction, the electron distribution is altered, forming a
Schottky barrier. During the reaction, the photogenerated electrons are transferred
from TiO2 to the metal particles, and the Schottky barrier inhibits the electrons
transport back to the conduction band of TiO2, effectively avoiding the photogener‐
ated electron–hole recombination. Meanwhile, this complex insures the light-
catalyzed redox reactions take place on the surface of different materials, which will
further suppress the possibility of the recombination which should have taken place
on the same particle (Fig. 1.6a).
However, excessive metal deposition is averse to the photocatalysis, because the
coated noble metal will affect the light absorption of TiO2, reducing the number of
photogenerated electrons, thereby reducing light catalytic ability. Meanwhile, once
the metal particles are electrically negative, the photogenerated holes will be captured
by them, then the introduced metal will become the recombination centers. Currently,
the commonly used noble metals are platinum, gold, silver, iridium, palladium,
18 1 Introduction
ruthenium, and so on, and platinum is the most effective [120, 121]. In addition, the
use of a noble metal-doped TiO2 can also obtained the visible light absorption.
1.2.2.2 Modification to Enhance the Visible Light Response
No matter anatase, rutile or brookite type of titanium oxide, the absorption peaks are
all in the ultraviolet region. But considering only 5 % of sunlight is in UV region, it
is necessary to modify the catalyst to absorb light in visible light, hence it is necessary
to improve the utilization of sunlight [122]. As mentioned above, the noble metal
doping is a process to realize the visible modification by changing the band gap,
additionally, by doping nonmetal or transition metal ion, and compositing with
semiconductor or sensitization can also achieve visible light utilization.
Nonmetal doping is by introducing other elements which are located near O (in
the Periodic Table of the Elements) into oxygen vacancies of TiO2 or directly substi‐
tuting the lattice oxygen, then the visible light response is achieved. Currently, C,
N, S, F have been successfully introduced into TiO2, and its photocatalytic properties
have been improved. Wherein, on one hand, N doping can form impurity level as
NOx, and thus enhance the absorption of visible light [123]; on the other hand, the
doped N can decrease the band gap, then realize the visible light response [124]. The
doping of F can not only enhance the absorption of visible light, but also can improve
the acidity of material surface, producing oxygen vacancies and new active sites
[125]. While for the C-doped TiO2, the valence band of it is usually consist of the
2p orbit of C. Because for the energy level of C-2p orbital is higher than the one of
O, the valence band of this doped material is increased, thereby the band gap is
reduced. According to the literature, C doping will always get a smaller band gap
than N doping, which can be more efficient to use visible light [126]. However, the
mechanism of C doping is not clear, more work still needs to be further explored.
Transition metal ions doping to a certain extent, can affect the crystal growth and
cause the phase transition, leading to a change in spectral absorption; also it can
change the carrier diffusion length, may reduce the recombination [126–130].
Depending on the relative positions of ions in the crystal lattice, the doping products
can be divided into four categories as substitution type, product type, gap type and
aggregate type [127]. In detail, in the substitutional type doping, Ti atoms are
replaced by other metal, forming new chemical bonds and introducing new energy
levels, thereby affecting the photocatalytic ability of TiO2. In the product type
doping, the introduced metal ions will form a new solid solution or compound with
former TiO2 lattice. The changed band gap has a significant impact to the photoca‐
talyst. In the gap type doping, the metal ions are doped into the crystal lattice of
TiO2. The crystal lattice is distorted by the internal, but the doping does not replace
the atoms in the lattice. The light absorption can be influenced by the size and
concentration of dopant ions. In the aggregate type doping, the doped oxide will form
a film cover the surface of TiO2 or aggregate in the TiO2 nanocrystals. Some doping
1.2 Photocatalysis 19
(a) V/NHE (b) V/NHE
Conduction band -1.0 Conduction band -1.0

Acceptor level
H+/H 2 0 H+/H 2 0
UV Vis UV Vis
+1.0 +1.0
O2/H 2O O2/H 2O
Donor level
+2.0 +2.0
Valance band Valance band
+3.0 +3.0
Fig. 1.7 Donor level (a) and acceptor level (b) formed by metal ion doping [122]. Copyright ©
2010, American Chemical Society
here can improve the light absorption efficiency, but some will also become a recom‐
bination center of photogenerated electron–hole pair, resulting in reduction in cata‐
lytic activity to the contrary. At present, V, Ni, Cr, Mo, Fe, Zn, Mn, Co and other
metals are reported as a dopant, which increase the valence band, or decrease the
conduction band, hence the band gap is reduced, bringing a visible light absorption
(Fig. 1.7) [131–134].
Surface sensitization is a method by adsorbing photosensitizer on the surface of
TiO2. Photosensitizer is a material can be irradiated by the visible light and forming
excited electrons. If the energy level is matching, electrons can inject into the
conduction band of TiO2. The oxidized photosensitizer has oxidation ability, while
the semiconductor shows a reduction ability because of the injection electrons on
the conduction band. But a relatively small number of sensitizers are reported so far,
because: (1) the majority of azobenzene sensitizer is toxic itself, which can cause
secondary pollution; (2) the sensitizer itself is easily be degraded by TiO2, reducing
the catalytic activity; (3) ruthenium and platinum dye will not be used because of
their high price [19a].
Instead of photosensitizer, narrow bandgap semiconductor has a similar mech‐
anism in visible light modification. Besides that, it can promote the separation of
photogenerated electron–hole, effectively prolonging the carrier lifetime, promoting
the electron transfer from interface to the adsorbed species (Fig. 1.6b). Based on this
method, several composite materials such as CdS/TiO2, CaSe/TiO2, Bi2S3/TiO2,
WO3/TiO2, Cu2O/TiO2, CuBi2O4/TiO2, etc. have been reported [135–139]. Due to
the narrow bandgap semiconductors are divided into n-type and p-type semicon‐
ductors, the composites of narrow bandgap semiconductors and TiO2 can be divided
into n-n-type and p-n-type semiconductor composites. In this case, when the compo‐
site is fabricated by p-type and n-type semiconductors, a p–n junction will be built
up at the interface, thus forming a built-in electric field from n-type to p-type semi‐
conductor, which will suppress the recombination as a result.
20 1 Introduction
1.3 Carbon Materials in Photoelectric Conversion System
Carbon material is progressing accompany with the development of human being.

Cheap as graphite, expensive as diamond, they are both members of carbon family.
In nanoscale, carbon material is also a class of star molecules. From zero-
dimensional fullerenes, to one-dimensional carbon nanotubes, both caused wide‐
spread concern in the scientific community. In 2004, a new two-dimensional (2D)
carbon nanomaterial graphene crash into people’s vision [140], quickly becoming a
famous star in material science. In 2010, synthetic 2D graphdiyne has been reported
[141], which may open a new chapter for the study of the carbon material. Herein,
in this section we will focus on the carbon material used in photovoltaic field.
1.3.1 Donor–Acceptor Photovoltaic Material

Based on Zero-Dimensional Fullerenes
1.3.1.1 Structure of Fullerenes
In 1985, Kroto et al. discovered a new kind of carbon material whose structure is
similar with football, and the named this new material as Fullerene, C60. In the single
crystal structure, there are approximately 10 Å between every two spacing balls,
each ball, with a diameter of 7.1 Å, is composed of 12 pentagons and 20 hexagons.
However, C60 did not attract much concern until the large-scale production of full‐
erene reported by Kratshmer in 1990. Over the last few years, studies have shown
that there is a large delocalized π bond in Fullerene, making them perform distin‐
guishingly in optical, electrical, magnetic, superconductive and even biological
fields. Among various derivatives, the most widely investigated one till now is the
PCBM.
1.3.1.2 Application of Fullerenes in Photovoltaic Field
The investigation of fullerenes in the photovoltaic area is mainly based on light-

induced electron transfer theory of polymer/C60 complex. The theory was proposed
by Heeger et al. based on fluorescence quenching method [142]: When the polymer
is excited by light in a certain frequency, π–π* transitions takes place and the electron
transfer from the valence band to the conduction band. If the conduction band energy
level of polymer is higher than the LUMO (lowest unoccupied molecular orbital)
energy level of C60, the electron will transit from the polymer to the C60. This theory
drew a lot of attention, and large number of experiment were carried out to verify
its rationality. At the same time, solar cells based on this theory were developed.
Organic semiconductor can generate electron - hole pairs under light, which will
separate at the interface of the materials with different electron affinities and
1.3 Carbon Materials in Photoelectric Conversion System 21
Fig. 1.8 Schematic of charge dissociation at the polymer: PCBM interface [146]. Copyright ©
2010, American Chemical Society
ionization potentials. If electric field exist at the same time, the separated electron—
hole will migrate to the electrode surface with different work function respectively,
and photoelectric effect takes place consequently (Fig. 1.8). In the conjugated
polymer/C60 composite system, the charge separation efficiency is near 100 % due
to the ultrafast photogenerated electron transfer, furthermore, the separated state is
more stable with a long lifetime and low recombination. In 1993, Yamashita et al.
fabricated a kind of bilayer photovoltaic device with tetrathiafulvalene and fullerene
as electron donor and acceptor respectively, and found its rectifying effect under
light [139]. Thereafter, Yoshino proposed a three layered D/M/solar cell structure,
where D is an electron donor, A is an electron acceptor and M is the light absorbing
layer. Kallinger introduced a spacing layer in the cell, and found that the exciton was
more difficult to be quenched with thicker space and resulting a stronger fluorescence
signal [143].
Although the separation efficiency of the exciton is close to 100 % at the hetero‐
junction, but the actual efficiency of the battery is very low. Heeger believes that
there are two factor restricting the cell efficiency: First is the diffusion limit of exci‐
tons. Because exciton diffusion length is only a few to tens of nanometers, but charge
separation can only occur at the heterojunction, therefore, most of the exciton
recombines before reaching the heterogeneous. The second factor is the low collec‐
tion efficiency of the separated charge. The double layer structure as introduced
22 1 Introduction
above is not conducive to the carrier transport or collection. Hence, he proposed an

interpenetrating network structure to increase the donor–acceptor interface area
[144]. In this structure, the heterojunction disperses uniformly and the exciton diffu‐
sion distance is reduced, so more exciton can diffuse onto the interface. As a result,
the photoelectric conversion efficiency is increased. This heterojunction is called
bulk heterojunction [1].
Bulk heterojunction is prepared by mixing the donor and electron acceptor mate‐
rial directly, so undesirable phase separation and aggregation will appear inevitably.
In this case, the effective contact area between the donor and acceptor is reduced and
the charge separation is hindered, leading to decreased photoelectric conversion
efficiency. To solve this problem, a kind of compounds known as bipolar polymers
is proposed, where the donor–acceptor molecules are conjugated into one molecule,
so that the phase separation will be inhibited. At the same time, charge transport can
be improved by modifying the molecular structure. Such compounds usually need
to meet the following requirements: (1) the donor and acceptor (C60) in the polymer
should be avoided influence each other, each maintaining basic electrical properties
itself; (2) the metastable state during the photoelectron transfer from a donor to the
receptor should have a long lifetime, to ensure the free carrier generation; (3) the
polymer should have a certain solubility. Although it is already known that increasing
the content of C60 within a certain range is beneficial to enhance the photoelectric
conversion efficiency, C60 content in such bipolar polymers is still much lower than
its bulk heterojunction content, due to the decreasing solubility with increasing
amount of C60. Currently, there are two methods to enhance the solubility: one is to
improve the solubility of the conjugated polymer part through structural modifica‐
tion, and then conjugate with the C60 part; the second is the direct polymerization
of a solubility molecules modified C60 monomer [145, 146].
Recently, another research point lies in the combination of metal organic
compound and fullerene, to form a new donor–receptor solar cell. Electron donor
of such batteries are usually phthalocyanine or subphthalocyanines metal
complex [145].
1.3.2 Application of One-Dimensional Carbon Nanotubes

in Photoelectric Conversion System
1.3.2.1 Structure of Carbon Nanotubes
In addition to C60 and derivatives thereof, another kind of carbon material, carbon
nanotubes, have also been used in the photoelectric conversion system [147]. Carbon
nanotubes are made of sp2 hybridized carbon atom which bonded with another three
carbon atoms to form a six-membered ring. The edge length of its hexagonal unit
cell is 2.46 Å, and the shortest of the C–C bond length is 1.24 Å. When the graphite
sheet coil to form carbon nanotubes, some of the pentagon and heptagon carbon cycle
present in the bent portion of the tube. If the pentagon carbons happen to be on the
top of the tube, a closed tubular structure will form. Carbon nanotubes can be divided
into two categories: one is the multiwalled carbon nanotubes (MWCN), the interlayer
spacing of which is about 0.34 nm, slightly larger than the value of 0.335 nm in
graphite. The arrangement between the layers is disordered. Another one is single-
walled carbon nanotubes (SWCN), both ends of which are consists of half fullerenes,
while the cylinder in the middle is made of crimped monolayer graphene.
Carbon nanotubes is similar with graphite, with a large number of unpaired elec‐
trons migrating along the wall. Hence, the carbon tubes also have good electrical
conductivity. In fact, carbon nanotubes possess both conductive and semiconductive
properties which is determined by the diameter and the spiral structure. The diameter
and the helical structure is mainly determined by the chiral vector (n, m), where n
and m are integers. When (n–m) is a multiple of 3, the SWCNs is conductive, other‐
wise it is semiconductive. However, one single carbon nanotube may have metallic
and semiconducting property simultaneously due to some defects. In this case, the
carbon nanotube becomes a molecular diode. Till now, there is still neither effective
way to prepare carbon nanotubes with specific (n, m) value nor to separate a partic‐
ular kind of carbon nanotubes from a mixture, and this limits the in-depth research
on the physical and chemical properties of this material [148].
1.3.2.2 Application of Carbon Nanotubes in the Photovoltaic Field
Similar to fullerenes, carbon nanotubes in the photoelectric conversion area also

plays a role of electron acceptor. Owing to the hybridization of carbon atom on side
wall, there is a large number of delocalized π electrons which can interact with the
electrons on conjugated polymer through π–π interaction. Therefore, it is possible
to prepare the polymer–carbon nanotube composite material similarly as fullerene.
In addition, Kamat successfully deposited use carbon nanotubes on the conductive
glass by electrophoresis method, and found that carbon nanotubes are optically
active, which can generate photocurrent in the visible light irradiation. The charge
separation process is also proved by transient spectra, but the efficiency is only 0.15 %
probably due to the charge recombination [149]. Nevertheless, the deposited carbon
nanotubes can be used as supporting frame for other nano semiconductors to improve
the charge separation and transport efficiency.
Based on the above theory, Hasobe reported a breaking work in 2006. He assem‐
bled the SWCN and protonated porphyrin on SnO2 electrode, and observed a photo‐
voltaic activity of the composite in visible light [150]. Soon thereafter, Kamat
reported a kind of ZnO–carbon nanotube composite material with UV responsive‐
ness. It had proved by using emission spectrum and photoelectrochemical spectrum,
that carbon nanotubes play a role of charge collection and transport in this complex
system characterization [151]. As shown in Fig. 1.9, (a) photogenerated electrons
are captured by carbon nanotubes; (b) a supporting film made of carbon nanotubes
can transport electrons quickly.
Although some progress has been achieved in application of carbon nanotubes in
the photovoltaic area, there is still a long way to go toward a real application. At
24 1 Introduction
Fig. 1.9 Schematic representation of (a) photoinduced charge injection from an excited
semiconductor particle into a SWNT and (b) SWNTs as a conducting scaffold to transport
photogenerated charge carriers in a photoelectrochemical cell [147]. Copyright © 2010, American
Chemical Society
present, MWNTs with highly crystallized wall is ready for commercial production,
but the cost is still high. Wherein the use of SWNT as conductive substrate is
currently the most promising direction, so development of methods to produce
SWNT in large scale is very necessary.
At the same time, to reduce costs and promote the development of new energy,
scientists are still exploring other carbon materials, such as graphene.
1.3.3 Applications of Two-Dimensional Carbon Materials

in Photoelectric Conversion System
1.3.3.1 Structure and Properties of Graphene
In 2004, British scientists reported the two-dimensional atomic crystal made from
sp2 hybridized carbon atoms, graphene, which is composed of benzene rings. The
discovery of graphene achieved the reunion of carbon material family: from zero-
dimension (0D) fullerene, to one-dimensional (1D) carbon nanotube (CNT) and two-
dimensional (2D) graphene, up to three-dimensional (3D) diamond and graphite,
forming a complete system. Graphene is the thinnest 2D material to date with a
thickness of only 0.35 nm, exhibiting many excellent properties. For example, its
tensile strength reaches 130 GPa, the highest among known materials and more than
100 times stronger than steel; its carrier mobility is as high as 15,000 cm2 V−1 s−1,
twice as that of indium antimonite material which is previously known to have the
highest carrier mobility, and more than 10 times higher than the commercial silicon
wafer, and under certain condition (e.g. fast cooling), the carrier mobility can reach
25,000 cm2 V−1 s−1; its thermal conductivity is three times larger than that of
diamond, achieving 5000 W m−1 K−1; meanwhile it also possesses unique properties
like room-temperature Hall effect and room-temperature ferromagnetic property.
Compared with CNT, the major properties of graphene are comparable and even
better, while graphene avoided many difficulties encountered in the research and
applications of CNT, such as the control of chirality, separation of metal- and
semiconductor-type, impurities in catalysts, etc. Planar graphene is suitable for direct
fabrication of various semiconductor devices and interconnectors, leading to highly
valuable full-carbon integrated circuits. Due to its outstanding mechanical and elec‐
trical properties, it is possible to synthesize graphene-based flexible electronics and
functional composites. With the many advantages including outstanding properties,
low cost and good processability, graphene is predicted to have great application
potentials in fields like electronics, information, clean energy, materials innovation
and biomedicine, and lead a revolution in technology in the twenty-first century.
1.3.3.2 Functionalization of Graphene
To date, common preparation methods for graphene include mechanical exfoliation,

chemical oxidation, crystal epitaxial growth, chemical vapor deposition, organic
synthesis, etc. However, a great challenge in the application of graphene is, still, to
realize its controllable functionalization. Graphene with a complete structure is a 2D
crystal configured by benzene hexagons without any instable bonds. The chemical
stability is thus high with inert surfaces and relatively weak interactions with other
intermediates. Moreover, the relatively strong interlayer van der Waals force could
easily result in aggregation. These features cause great difficulties in the further study
and application of graphene. To fully develop the outstanding properties of graphene,
effective functionalization of graphene is necessary. So far functionalization
methods are mainly divided into covalent functionalization and non-covalent func‐
tionalization.
Covalent functionalization of graphene mainly makes use of the abundant amount
of carbonyl, hydroxyl, and carboxyl groups in the graphene oxide (GO). Figure 1.10
shows the common modification methods for graphene. Among them, acylation
reaction is a mostly used method. Through the reaction between activated carboxyl
group of SOCl2 and octadecylamine, long-chain alkyl group-modified graphene can
be obtained (reaction VI). By expanding this method, GO can be coupled with PEG
toward biocompatibility, and with the π–π stacking interactions between graphene
and some drugs, package and transport of hydrophobic drugs can be realized. The
epoxide bond on the GO surface can go through nucleophilic ring-opening substi‐
tution reaction with the amino groups at the end of organic molecules (reaction VIII).
By taking advantage of this reaction we can obtain GO-rich ionic liquids. Due to the
enhanced solubility and the repulsive interlayer forces, the product can be well
dispersed into water, N,N-dimethyl formamide (DMF), and dimethyl sulfoxide
(DMSO). Besides small molecules, GO can also couple with polymers (such as
polyvinyl alcohol) through esterification reaction (reaction VII). In 2009 Salvio and
others deposited graphene on silicon oxide. They first used sodium azide to modify
26 1 Introduction
Fig. 1.10 Schematic showing various covalent functionalization chemistry of graphene or GO

[190]. Copyright © 2010, Royal Society of Chemistry
and reduce GO to amino-functionalized nanosheets (reaction III and IV), and then
deposit it onto the isothiocyanate-treated silicon oxide by microcontact printing to
generate covalent bond. The excess nanosheets were removed by sonication, and as
a result, graphene is obtained. This method is considered an important way to chem‐
ically transfer single layers of graphene [152–160].
Non-covalent functionalization includes the modification of graphene by π–π
stacking and van der Waals forces. Compared with covalent modification, this
method doesn’t damage the structure of graphene or introduce imperfections, thus
has little influence on its electronic properties. Pyrene and its derivatives is a type of
commonly used organic molecules with conjugated structures, and can composite
with graphene with π–π interaction. Shi et al. used pyrenebutyacid (PBASE) to non-
covalently functionalize graphene (Fig. 1.11a) to make it disperse well in water, and
then vacuum filter the product to form a flexible graphene film. It is predicted that
the most promising functionalization method for graphene is compositing it with
polymers. This composite can be improved in terms of electronic and mechanical
properties owing to the π–π interaction mechanism [161]. Bai et al. combined
graphene with sulfonated polyaniline (SPAN), forming a stable new material with
electronic activity (Fig. 1.11b). Others also reported the utilization of triblock
copolymers (PEO–PPO–PEO) on wrapping graphene for supramolecular hydrogels.
The hydrophobic block of the copolymer interacts with graphene, while the hydro‐
philic parts dissolve into water so that the composite is stabilized [162]. Moreover,
electrostatic force can also stabilize graphene. So far, making use of the excellent
electronic properties of graphene composite on optical thin films seems a close to
Fig. 1.11 Noncovalent interactions between a small molecules (PBASE), b polymer (SPANI) and
graphene through π–π stacking [190]. Copyright © 2010, Royal Society of Chemistry
application direction. Future research may lie in the development of the optical
resonance absorption saturation of the film as well as its ability tunable bandwidth,
the realization of which requires physicists, chemists, materials scientists to work
together.
1.3.3.3 Application of Graphene in Photoelectric Conversion System
In the photoelectric conversion system, graphene always serves as an electron

acceptor to enhance the separation efficiency of photogenerated electron–hole pair
and suppress recombination. In terms of application, graphene composite is mainly
applied to photocatalysts and solar cells.
In recent years, composites of semiconductors and other materials intrigued great
interests of researchers globally. The conventional semiconductor materials suffer
from the easy recombination of electron–hole pair, resulting in a low catalytic effi‐
ciency. To suppress recombination, besides conventional doping and co-adsorbing,
carbon composite material has attracted a lot of attention. To date, the applications
of photocatalysts mainly focused on the photocatalytic decomposition of pollutants,
photocatalytic water splitting and photocatalytic sterilization.
Photocatalytic decomposition of pollutants So far the most developed composite
is graphene and titanium dioxide. This composite generally exhibits high absorbing
capability, broad absorbing spectrum, and stronger charge separation capability. Li
et al. first made use of P25-graphene hybrids to photodegradation of methylene blue.
They discovered that the π–π interaction between methylene blue and graphene led
to an easier absorption of dye by graphene. Meanwhile the photo-induced electron–
hole pair exhibit longer lifetime, leading to a better photocatalytic ability [163].
Moreover, Zhang’s group found that its photodegradation activity is higher than
carbon nanotube or other forms of carbon hybrids [163]. Liu et al. pointed out that
28 1 Introduction
Fig. 1.12 Schematic illustration of selective catalysis at different sites on graphene bridges [166].
Copyright © 2010, American Chemical Society
the absorbed oxygen on the surface of graphene can be reduced by the captured
electron, further suppressing the possibility of its recombination with holes [164].
Wang et al. introduced hierarchical structure of TiO2 into the photocatalytic system
and enhanced the absorption capability and charge transportation speed [165]. In
addition, Wang et al. found out that graphene composites can improve the absorption
capability of visible lights, realizing the phototcatalysis under visible light [131].
Besides TiO2-Graphene composites, SnO2-Graphene, ZnO–Graphene, Bi2WO6-
Graphene, Ag/AgCl-GO, Ag/AgBr-GO, ZnFe2O4-Graphene, InNbO4-Graphene,
Bi2MoO6-Graphene were all successfully prepared and applied in the photocatalysis
system [166]. Surprisingly, with better control of thermodynamics, SnO2-Graphene
composite system showed better performance than TiO2-Graphene in the degrada‐
tion of rhodamine [167].
Photocatalytic water splitting Hydrogen is an environmentally friendly, recy‐
clable, green energy. Semiconductors are considered excellent catalysts for water
photolysis and hydrogen generation. However due to the easy recombination of
photo-induced electrons and holes within the catalysts, the efficiency for water split‐
ting is still low. Considering the high specific surface area and outstanding electron
capture capability of graphene, the preparation of composite materials is naturally
applied for photolysis of water. While traditional photocatalysis uses single catalyst,
the introduction of graphene can work as bridges that allows electrons to transfer
from one catalyst to another and thus separate the reduction and oxidation process
to suppress recombination (Fig. 1.12). Fan et al. found that composites of different
reduced GO and iron oxide possess generally good photocatalytic capabilities, and
among them, the composite of GO prepared by hydrothermal process have the best
performance [168]. Interestingly, Teng et al. discovered that even without the partic‐
ipation of Pt species, GO itself possesses the ability for photocatalytic water splitting.
Due to the different oxidized sites on GO, the band gap fluctuates between 2.4 eV
and 4.4 eV. Also the conduction band is made of anti-π* orbits with higher energy
level than hydrogen, which is beneficial for the injection of electrons during the
water-based hydrogen generation [166, 169].
Photocatalytic sterilization Titanium dioxide is believed to be a nontoxic, light-

tolerant sterilization material with almost no post-maintenance. As graphene is able
to prolong the lifetime of electrons and enhance the absorption of visible light,
scientists have been trying to make graphene-TiO2 composites for sterilization tests.
As reported, this type of composites can kill 99.9 % Escherichia coli under sun light,
7.5 times than that of normal TiO2 [170].
Due to the strong dye adsorption ability, wide light absorption range and fast
charge transfer ability, graphene composite is applied in various fields of photoca‐
talysts. The methods to prepare composite include in situ growth, solution mixing,
hydrothermal process and solvothermal process. Although it brings progress in the
photocatalysis field, there are still many questions yet to be explained clearly and
further study is still continuing [166]. First of all, the controlled growth of graphene
has not been realized, and it remains a challenge to obtain graphene with high quality;
second, the properties of graphene composites strongly depend on the structure
design of composite, so the methodology for precise nanostructures is also a key for
enhancing the overall performance; third, the graphene composite needs to closely
relate to the rapid development of nanomaterials, in order to make new break‐
throughs; forth, the photocatalysis mechanisms of graphene composite are not very
clear, and the difference in catalytic mechanism between graphene composites and
other carbon-based composites still needs to be further explored by researchers.
Solar cells can convert light energy into electrical energy, and is believed to be
an effective solution for global energy crisis. As a new member of materials family,
graphene has been applied in transparent electrodes of solar cells, dye sensitizing in
solar cells, and heterostructure solar cells.
Transparent electrode is an essential component in solar cells. To date the mostly
used transparent electrodes are prepared by depositing indium tin oxide (ITO) or
fluorine tin oxide (FTO) on glass. These electrodes are relative expensive and limited
by metal source, therefore there is need to develop new materials for the transparent
electrode. The emergence of graphene provides new ideas. As monolayer graphene
possesses good transparency and conductivity theoretically, it becomes top candidate
for transparent electrodes, and is very suitable for flexible electrodes. So far the
preparation of transparent graphene electrode is through solution-based methods,
with work function tunable by chemical doping and structure controllable [171].
However, current challenge lies in the conflict between conductivity and transpar‐
ency. If higher conductivity is achieved, there will be loss in transparency, and
vice versa (Fig. 1.13).
The raw materials for preparation of transparent electrodes is mostly chemically
modified graphene. Because on the one hand, it is suitable for large-scale, low-cost
manufacturing, and on the other hand, this precursor can be easily dispersed and
fabricated onto substrates. Meanwhile, it can also be made into composites with other
materials for multifunctional thin films, but this as-prepared transparent electrodes
are still far from practical requirements. The factors that influence its conductivity
are their structural defects and the contact resistance between different pieces. The
30 1 Introduction
Solution-based processing
CVD
Graphene/CNT composite
Fig. 1.13 Comparison of transmittances and sheet resistance of the graphene films prepared by
different methods: solution-based processing (dots), CVD (squares) and graphene/CNT composite
(triangles) according to literatures [171]. Copyright © 2011, Royal Society of Chemistry
(a) (b) (c)
Fig. 1.14 Different structures of 2D carbon materials. a Graphene, b graphyne, and c graphdiyne
most effective way to remove the surface oxygen on GO so far, is by the reduction
of iodic acid and calcination in high vacuum. In addition, the calcination under
ethyne atmosphere can also restore the carbon vacancies. To reduce contact resist‐
ance, the preparation of large-scale graphene layers is required, or alternatively
combining it with other materials. So far the graphene–CNT composite transparent
electrodes can reduce its square resistance to 260 Ω/cm2 with 88 % transparency.
Compared with chemical modification and transfer, chemical vapor deposition
(CVD), is the best way. This method can prepare high-quality, transparent, patterned
electrodes, with resistance close to the commercial requirements [171] (Fig. 1.14).
Although the efficiency obtained so far is still not as good as ITO, but ITO suffers
from greatly affected conductivity after bending, while carbon materials does not.
Dye-sensitized solar cells (DSSC) has drawn great attention due to its low cost,
ease of fabrication and high efficiency. However, there are still a few aspects where
DSSC can be improved: finding novel materials to replace expensive Ru dyes and
Pt electrodes, optimizing the structure of photoanode, developing nonliquid elec‐
trolyte, preparing flexible electrodes, etc. Recent research shows that graphene has
contributions in almost all the above aspects.
Photoanode is the core component of DSSC, as its inhibition of electron–hole
recombination provides an effective way for improving light conversion efficiency.
Kamat et al. added CNT into the photoanode and found increment in the cell effi‐
ciency to a certain extent, but a reduction in open circuit voltage and the overall
efficiency was not high [172]. The introduction of graphene can avoid the above‐
mentioned shortcomings.
In DSSC, counter electrode works as a catalyst for the redox reaction of electro‐
lyte. In a practical work, the reduction of I2 can take place at the counter electrode
and the surface of TiO2 simultaneously. As there exists a competition, to suppress
the back reaction, electrodes with high catalytic activity is necessary. Through recent
research, people suggested that replacing Pt with graphene would be a relatively
better choice. Very recently, scientists also considered using graphene composites
to fabricate counter electrodes, the methods of which includes using graphene or
using graphene hybrids (such as Graphene/PEDOT, Graphene/PANI, Graphene/
PEDOT-PSS, Graphene/Pt, Graphene/CNT), etc. Although this method can prepare
flexible electrodes, the resultant cell efficiency is not as high as pure Pt [173]. The
reason being that on the one hand, similar as fabricating transparent electrodes, the
conductivity is not high enough, and on the other hand, the catalytic mechanism is
not clear, which needs deeper study to reveal whether graphene is suitable for fabri‐
cation of counter electrodes.
In DSSC, dye molecules often need to possess a suitable band gap, a long-term
stability, a good compatibility with electrolyte, and excellent ability of injecting
electrons into TiO2. Due to the expensive price of Ru dyes, people started to consider
other materials. As the band gap of graphene is tunable, it is also considered for dye
fabrication. Yan et al. reported a synthesized graphene quantum dot with 67 benzene
rings and an absorption band at 900 nm, covering all visible lights. Then they intro‐
duced it into DSSC as dyes. Although the contact between graphene and TiO2 is not
very good thus the efficiency is not that high, this approach provides a new route for
dye synthesis. Another disadvantage is that the organic synthesis methods used in
the preparation of graphene-based dyes is complicated and of high cost, therefore
whether a top-down methods can be applied for the preparation of graphene quantum
dots is also an inspiration form this work [171]. During this process, we should also
consider enhancing the contact between graphene quantum dots and TiO2 though
structural modification, improving the solubility and stability of quantum dots, and
avoiding the restacking of dispersed graphene.
Heterostructure solar cells mainly makes use of p-n junction to conversion solar
energy into electric energy. It can be simplified into inorganic and organic solar cells.
Inorganic solar cells are representable by silicon solar cells, but their fabrication is
32 1 Introduction
complicated with high cost; while organic solar cells have advantages in low cost
and light weight, but the efficiency is far less than Si. In all the organic photovoltaic
devices, the most studied one is the donor–acceptor composite based on fullerene,
which works as electro capturer and transporter. Theoretically, the properties of this
zero-dimensional nanomaterials are not as good as one-dimensional CNT and two-
dimensional graphene. Meanwhile, due to the huge conjugation structure, 2D
graphene is very suitable for polymer-based composites for organic photovoltaic
device materials. To date, graphene has been mixed with material like P3HT and
P3OT for battery fabrication. The efficiency is higher than CNT under the same
condition, but is as low as around 1 %, far below the theoretical value of 12 % [165].
The reason could be the fact that the nanosheet used is still GO, not real graphene,
thus possessing some defects and restrict its electric properties. Moreover, scientists
also tried to use graphene alone as a component of Schottky junction, the hole
conducting layers and transparent electrodes. Its Si-composited solar cells reached
an efficiency of 1.65 % [171].
The arise of graphene filled the gaps on the structure of carbon nanomaterials, it
showed amazing charm in the chemical, electronics, and physics fields, which brings
revolutionary changes in energy conversion, storage, and many other areas.
Although the research papers on graphene study are numerous, but this research is
still in its infancy, after all, it was only born for 8 years. There still are questions to
be solved: first, controlling the growth of graphene is a huge challenge. To date
graphene used in the applications are actually chemically reduced GO with large
amount of defects inside, and these defects greatly influenced the electrical proper‐
ties; the introduction of defects changes the bandgap of graphene, and the correlation
and control method is still not clear. Second, enhancing the stability of graphene is
also a challenge. As the intrinsic hydrophobicity and inert surface, it is difficult to
be dispersed in solvents and easily to be aggregated, which can greatly influent the
performances; whether the structural modification will influence its property also
requires further discussion. Third, the intrinsic mechanism of graphene still needs
exploration. So far the explanation of mechanisms mainly based on the under‐
standing of other carbon materials, whether unique properties exist in graphene is
still not clarified. This depends on joint cooperation of chemists and physicists. But
we believe, that graphene can provide wide space for future development of clean
and recyclable energy.
1.3.3.4 Applications of Graphdiyne in Photoelectric Conversion System
In 2010, there came a new member to the carbon materials family, graphdiyne.
Different from the other carbon materials, graphdiyne is completely synthetic and so
far not found in nature. It composed of sp and sp [2] hybridized carbon allotropes.
Graphite is an all-carbon two-dimensional network structure formed by connecting
benzene molecule by conjugated 1,3-alkyne bonds. It contains rich carbon bonds,
large conjugation system, and wide interlayer distance (4.1913 Å), and thus possesses
excellent chemical stability and semiconductor properties [141, 174–177]. In 2010,
the Institute of Chemistry in Chinese academy of sciences made progress in the study
of graphdiyne, Li et al. used hexaethynylbenzene to catalyze coupled reaction on
copper plates, and successfully synthesized large area of graphdiyne thin film on the
copper surface through chemical method. They proved that this thin film had high
degree of order and relatively low amount of defects, with a conductivity of 10−3–
10−4 S/m [141]. So far the research based on graphdiyne is mostly theoretical,
involving in two types of molecules, graphyne and graphdiyne. The former connects
benzene rings with one triple bond, while the latter uses two. Through theoretical
analysis, the electrical property of graphyne is closely related to its structure. Under
certain conditions, its electrical property even exceeds graphene. As predicted, the
triangular space between benzene rings is beneficial for the embedment and migration
of atoms and molecules (Fig. 1.14), and thus plays important roles in fields including
lithium ion batteries, hydrogen storage and catalysts [178–182]. However, these
studies are based on theories. Experimental explorations were only reported by the
photocatalytic study on the composite of graphdiyne and P25 [183], and graphdiyne-
hybridized organic solar cells [184], but these experiments lacked theoretical
supports.
The research on graphdiyne has just started, and base on the results to date,
graphdiyne also faces the problem of poor dispersity, and its thickness is far larger
than graphene. Whether its performance is good or not still needs great amount of
works for scientists to conclude.
1.4 Novelty and Significance of This Thesis
Above all, in order to solve the energy and environmental crisis, scientists have tried
a lot of work, and the efficiently use of solar energy is one of the methods for the
sustainable development. The utilization of solar energy, can be divided to solar cells
and photocatalysis. With the development of nanotechnology, we have found a rapid
way to solve energy problem. Nanomaterials, with small size and large surface area,
has a great advantage in ion diffusion and energy storage, which provides a new
opportunity to upgrade traditional energy conversion and storage devices. It will
have a positive and profound impact on the development of related high-tech
research in energy and environmental science. In the work of preparing solar cell,
we have chosen a highly promising dye-sensitized solar cell, which has a low cost,
simple process and environment-friendly, but it still needs further efforts to improve
the efficiency. Literatures have given two ways to improve efficiency: enhance the
electron diffusion length and improve the utilization of light [185]. In previous work,
in order to enhance the electron diffusion length, resist recombination, scientists have
tried to dope metal oxide with noble metals or introduce carbon nanotube into the
electrode, to a certain extent, the photoelectric conversion efficiency is enhanced. In
terms of increasing the utilization of light, scientists always adopted methods
including of using a modified photoanode, introducing a light scattering layer or
preparing a hierarchical structure. However, these methods still contain lots of
34 1 Introduction
defects, and a further improvement is quite necessary. Herein, we improve cell

efficiency from two aspects as enhancing the electron transport length and light
utilization. In the work of enhancing electron diffusion length, we selected a new
type of carbon material reported in 2004, graphene [140]. The internal electrons
comply with its linear dispersion relation, showing the characteristics of massless
relativistic particles, demonstrating the unique electrical properties. The electron
transport speed can achieve 1/300 the speed of light in graphene. Further, by doping,
changing the number of layers or changing edges of nanosheet, the metal and semi‐
conductor properties of graphene can be tuned, which may be adjusted the shift of
n- and p-type. Thus it is considered as a rising star in material field. In terms of
increasing the light utilization, we introduce surface plasmon resonance (SPR) effect
into DSSC. There are two main characteristics are expected to apply in DSSC: One
is the hot electron effect. The incident light can excite the electrons on the surface
of noble metal nanoparticles to produce a collective vibration, and the vibrating
electrons are affected by the nearby electric field, then the electrons are transferred
from the noble metal to semiconductor [186]; Further, the noble metal nanoparticles
also have an effect named as “standing wave” [187], i.e., when the light irradiation
on the nanoparticles, the electron resonance will be excited. Subsequently, the reso‐
nance energy will be rereleased in the form of light, but the released light will change
the direction of propagation comparing with the original light, hence it is possible
to reflect light back to the photoanode, reusing the light.
In the work of photocatalyst, we choose a new class of two-dimensional carbon
material, graphdiyne, which also has a high carrier mobility rate [141]. Because of
the stable triple bonds and benzene ring, it is considered as the most stable carbon
allotropes [188]. At the same time, Malko also predicted some graphyne molecules
have a similar conductive capacity with graphene [189]. However, the current study
based on graphyne is more focused on the theoretical level, but lack of experimental
evidence. The combination of theoretical and experimental results will lead a bright
idea for this new material. In this paper, we will focus on increasing the electron
transfer and promotion the light utilization, as describing below:
(1) The key factor which restrict the efficiency of DSSC is the recombination of
photoinduced electrons and holes. In order to suppress the recombination,
Kamat introduced the carbon nanotube into photoanode. However, due to the
introduced carbon nanotubes reduces the apparent Fermi level of photoanode,
the open circuit voltage is reduced, and accompany with the disadvantage of
poor dispersion, etc., it is difficult to obtain a great improvement of photoelectric
conversion efficiency. While, graphene is expected to overcome the shortcom‐
ings of carbon nanotubes, we will try to introduce it into photoanode, investi‐
gating whether it can effectively improve the photoelectric conversion effi‐
ciency of DSSC.
(2) More precise structural designation can be benefit for revealing the photoelectric
conversion behavior of TiO2-graphene system. In the third chapter, the stacking
structure is fabricated by using TiO2 nanosheet and graphene nanosheet. Then
electronic transport behavior is studied with a view to reveal its intrinsic rela‐
1.4 Novelty and Significance of This Thesis 35
tionship. After that, we extend the stacking structures to conducting polymer,

by using π–π stacking and electrostatic interaction to prepare polyaniline–
graphite oxide, where the graphite oxide act as a template and dopant at the
same time.
(3) Introducing the surface plasmon resonance effect into the counter electrode of
DSSC, enhance the efficiency of the battery by using of “standing wave” effect;
(4) Study of photocatalytic activity of TiO2 and 2D carbon nanomaterial composite.
To reduce interference, we composite graphene and graphdiyne with TiO2
nanosheet which exposes a specific crystal facet, revealing the intrinsic elec‐
tronic transport behavior from the theoretical and experimental analysis.
This thesis is started from how to enhance the utilization of solar energy, and
based on designing novel nanocomposites. The study involved multibranches
including nanotechnology, polymers, bionic, surface plasmon resonance and
photonic crystals, aiming to design and apply these technologies into solar cells and
photocatalysis, which shows important theoretical and practical value.
References
1. 张正华,李陵岚,叶楚平,杨平华,有机太阳电池与塑料太阳电池, (2006)
2. BP Statistical Review of World Energy, 2015
3. Li, J., Wang, S.: China Solar PV Report. China Environmental Science Press, (2007)
4. Zhao, Y.: Preparation of TiO2 Micro-/nano- Hierarchical Structures and Their Photoelectrical
Performance, Doctoral Dissertation of Graduate University of Chinese Academy of Sciences,
(2007)
5. Huang, C., Li, F., Huang, Y.: Ultrathin Films for Optics and Electronics, p. 305. Peking
University Press, 2001
6. 陈振兴,高分子电池材料,北京:化学工业出版社, (2006)
7. Zhong, D.: J. Appl. Opt. 22, 34 (2001)
8. Li, F., Xu, J., Zhang, D.: Energy Environ. 4, 52 (2007)
9. Peng, Y., Han, X., Wen, H.: J. Gannan Normal Univ. 6, 78 (2006)
10. He, J., Su, Z., Xiang, L., Wang, J., Wang, Y.: Chin. Polym. Bull. 4, 53 (2007)
11. Pei, L., Liu, C.: Electro-Optic Technol. Appl. 19, 16 (2004)
12. Zhang, X., Guan, Y., Wang, K., Xie, J.: Chin. J. Rare Metal 30, 529 (2006)
13. Li, J., Zou, Z., Long, F.: Energy Technol. 26, 164 (2005)
14. Goetzberger, A., Hebling, C.: Sol. Energy Mater. Sol. Cells 62, 1 (2000)
15. Qiao, J.: Preparing and Characterization of Polymer Solar Cells. Master Dissertation of
Graduate University of Chinese Academy of Sciences, (2003)
16. Zweibel, K., Ullal, H., Roedern, V.: 25th IEEE Photovoltaic Specialists Conference, p. 745,
(1996)
17. Kong, F.T., Dai, S.Y., Wang, K.J.: Adv. Optoelectron. 2007, 1 (2007)
18. O’Regan, B., Grätzel, M.: Nature 353, 737–740 (1991)
19. (a) Hagfeldt, A., Boschloo, G., Sun, L., Kloo, L., Pettersson, H.: Chem. Rev. 110, 6595 (2010),
(b) Burschka, J., Pellet, N.: Moon, S.-J., Humphry-Baker, R., Gao, P., Nazeeruddin, M. K.,
Grätzel, M.: Nature, 499, 316 (2013)
20. Kavan, L., Grätzel, M.: Electrochim. Acta 40, 643 (1995)
21. Cameron, P.J., Peter, L.M.: J. Phys. Chem. B 107, 14394 (2003)
22. O’Regan, B.C., Durrant, J.: Acc. Chem. Res. 42, 1799 (2009)
36 1 Introduction
23. Hagfeldt, A., Grätzel, M.: Acc. Chem. Res. 33, 269 (2000)
24. Koops, S.E., O’Regan, B.C., Barnes, P.R.F., Durrant, J.R.: J. Am. Chem. Soc. 131, 4808
(2009)
25. Wang, P., Wenger, B., Humphry-Baker, R., Moser, J.E., Teuscher, J., Kantlehner, W., Mezger,
J., Stoyanov, E.V., Zakeeruddin, S.M., Grätzel, M.: J. Am. Chem. Soc. 127, 6850 (2005)
26. Hamann, T.W., Jensen, R.A., Martinson, A.B.F., Van Ryswyk, H., Hupp, J.T.: Energy
Environ. Sci. 1, 66 (2008)
27. Pagliaro, M., Palmisano, G., Ciriminna, R., Loddo, V.: Energy Environ. Sci. 2, 838 (2009)
28. Jose, R., Thavasi, V., Ramakrishna, S.: J. Am. Ceram. Soc. 92, 289 (2009)
29. Zhang, Q.F., Deneau, C.S., Zhou, X.Y., Cao, G.Z.: Adv. Mater. 21, 4087 (2009)
30. Ito S., Liska P., Comte P., Charvet R. L., Pechy P., Bach U., Schmidt-Mende L., Zakeeruddin
S. M., Kay A., Nazeeruddin M. K., Grätzel M. Chem. Commun. 4351 (2005)
31. Zhang, Z.P., Ito, S., O’Regan, B., Kuang, D.B., Zakeeruddin, S.M., Liska, P., Charvet, R.,
Comte, P., Nazeeruddin, M.K., Pechy, P., Humphry-Baker, R., Koyanagi, T., Mizuno, T.,
Grätzel, M.: Z. Phys. Chem. 221, 319 (2007)
32. Hore, S., Nitz, P., Vetter, C., Prahl C., Niggemann, M., Kern, R.: Chem. Commun. 2011 (2005)
33. Sommeling, P.M., O’Regan, B.C., Haswell, R.R., Smit, H.J.P., Bakker, N.J., Smits, J.J.T.,
Kroon, J.M., van Roosmalen, J.A.M.: J. Phys. Chem. B 110, 19191 (2006)
34. Zukalova, M., Zukal, A., Kavan, L., Nazeeruddin, M.K., Liska, P., Grätzel, M.: Nano Lett.
5, 1789 (2005)
35. Shankar, K., Mor, G.K., Prakasam, H.E., Yoriya, S., Paulose, M., Varghese, O.K., Grimes,
C.A.: Nanotechnology 18, 065707 (2007)
36. Mor, G.K., Varghese, O.K., Paulose, M., Shankar, K., Grimes, C.A.: Sol. Energy Mater. Sol.
Cells 2006, 90 (2011)
37. Mor, G.K., Shankar, K., Paulose, M., Varghese, O.K., Grimes, C.A.: Nano Lett. 6, 215 (2005)
38. Shankar, K., Bara, J., Paulose, M., Wietasch, H., Varghese, O.K., Mor, G.K., LaTempa, T.J.,
Thelakkat, M., Grimes, C.A.: Nano Lett. 8, 1654 (2008)
39. Macak, J.M., Tsuchiya, H., Schmuki, P.: Angew. Chem. Int. Ed. 44, 2100 (2005)
40. Jennings, J.R., Ghicov, A., Peter, L.M., Schmuki, P., Walker, A.B.: J. Am. Chem. Soc. 130,
13364 (2008)
41. Bwana, N.N.: Nano Res. 1, 483 (2008)
42. Zhu, K., Vinzant, T.B., Neale, N.R., Frank, A.J.: Nano Lett. 7, 3739 (2007)
43. Hamann, T.W., Martinson, A.B.F., Elam, J.W., Pellin, M.J., Hupp, J.T.: J. Phys. Chem. C
112, 10303 (2008)
44. Saito, M., Fujihara, S.: Energy Environ. Sci. 1, 280 (2008)
45. Keis, K., Lindgren, J., Lindquist, S.E., Hagfeldt, A.: Langmuir 16, 4688 (2000)
46. Quintana, M., Marinado, T., Nonomura, K., Boschloo, G., Hagfeldt, A.: J. Photochem.
Photobiol. A 202, 159 (2009)
47. Nguyen, H.M., Mane, R.S., Ganesh, T., Han, S.H., Kim, N.: J. Phys. Chem. C 113, 9206
(2009)
48. Shin, Y.-J., Lee, J.-H., Park, J.-H., Park, N.-G.: Chem. Lett. 36, 1506 (2007)
49. Minoura, H., Yoshida, T.: Electrochemistry 76, 109 (2008)
50. Yoshida, T., Iwaya, M., Oekermann, T., Nonomura, K., Schlettwein, D., Wohrle, D., Minoura,
H.: Chem. Commun. p. 400 (2004)
51. Yoshida, T., Tochimoto, M., Schlettwein, D., Wohrle, D., Sugiura, T., Minoura, H.: Chem.
Mater. 11, 2657 (1999)
52. Galoppini, E.R.J., Chen, H., Saraf, G., Lu, Y., Hagfeldt, A., Boschloo, G.: J. Phys. Chem. B
110, 16159 (2006)
53. Martinson, A.B.F., McGarrah, J.E., Parpia, M.O.K., Hupp, J.T.: Phys. Chem. Chem. Phys.
8, 4655 (2006)
54. Tornow, J., Schwarzburg, K.: J. Phys. Chem. C 111, 8692 (2007)
55. Martinson, A.B.F., GoIes, M.S., Fabregat-Santiago, F., Bisquert, J., Pellin, M.J., Hupp, J.T.:
J. Phys. Chem. A 113, 4015 (2009)
56. Law, M., Greene, L.E., Johnson, J.C., Saykally, R., Yang, P.D.: Nat. Mater. 4, 455 (2005)
References 37
57. Law, M., Greene, L.E., Radenovic, A., Kuykendall, T., Liphardt, J., Yang, P.: J. Phys. Chem.
B 110, 22652 (2006)
58. Ferrere, S., Zaban, A., Gregg, B.A.: J. Phys. Chem. B 101, 4490 (1997)
59. Onwona-Agyeman, B., Kaneko, S., Kumara, A., Okuya, M., Murakami, K., Konno, A.,
Tennakone, K.: Jpn. J. Appl. Phys., Part 2 44, L731 (2005)
60. Tennakone K., Kumara G., Kottegoda I. R. M., Perera V. P. S. Chem. Commun. 15 (1999)
61. Kumara, G.R.R.A., Tennakone, K., Perera, V.P.S., Konno, A., Kaneko, S., Okuya, M.J.: Phys.
D. Appl. Phys. 34, 868 (2001)
62. Kay, A., Grätzel, M.: Chem. Mater. 14, 2930 (2002)
63. Ito, S., Makari, Y., Kitamura, T., Wada, Y., Yanagida, S.: J. Mater. Chem. 14, 385 (2004)
64. Tan, B., Toman, E., Li, Y.G., Wu, Y.Y.: J. Am. Chem. Soc. 129, 4162 (2007)
65. Lana-Villarreal, T., Boschloo, G., Hagfeldt, A.: J. Phys. Chem. C 111, 5549 (2007)
66. Guo, P., Aegerter, M.A.: Thin Solid Films 351, 290 (1999)
67. Wang, Z.S., Hara, K., Dan-Oh, Y., Kasada, C., Shinpo, A., Suga, S., Arakawa, H., Sugihara,
H.: J. Phys. Chem. B 109, 3907 (2005)
68. Galoppini, E.: Coord. Chem. Rev. 248, 1283 (2004)
69. Nazeeruddin, M.K., Humphry-Baker, R., Liska, P., Grätzel, M.: J. Phys. Chem. B 107, 8981
(2003)
70. Mann, J.R., Gannon, M.K., Fitzgibbons, T.C., Detty, M.R., Watson, D.F.: J. Phys. Chem. C
112, 13057 (2008)
71. Kuang, D., Klein, C., Ito, S., Moser, J.E., Humphry-Baker, R., Evans, N., Duriaux, F., Grätzel,
C., Zakeeruddin, S.M., Grätzel, M.: Adv. Mater. 19, 1133 (2007)
72. Kuang, D., Ito, S., Wenger, B., Klein, C., Moser, J.-E., Humphry-Baker, R., Zakeeruddin,
S.M., Grätzel, M.: J. Am. Chem. Soc. 128, 4146 (2006)
73. Haque, S.A., Palomares, E., Cho, B.M., Green, A.N.M., Hirata, N., Klug, D.R., Durrant, J.R.:
J. Am. Chem. Soc. 127, 3456 (2005)
74. Ardo, S., Meyer, G.J.: Chem. Soc. Rev. 38, 115 (2009)
75. Lammi, R.K., Wagner, R.W., Ambroise, A., Diers, J.R., Bocian, D.F., Holten, D., Lindsey,
J.S.: J. Phys. Chem. B 105, 5341 (2001)
76. Cherian, S., Wamser, C.C.: J. Phys. Chem. B 104, 3624 (2000)
77. Fungo, F., Otero, L., Durantini, E.N., Silber, J.J., Sereno, L.E.: J. Phys. Chem. B 104, 7644
(2000)
78. Fungo, F., Otero, L.A., Sereno, L., Silber, J.J., Durantini, E.N.: J. Mater. Chem. 10, 645 (2000)
79. Tachibana, Y., Haque, S.A., Mercer, I.P., Durrant, J.R., Klug, D.R.: J. Phys. Chem. B 104,
1198 (2000)
80. O’Regan, B.C., Lopez-Duarte, I., Martinez-Diaz, M.V., Forneli, A., Albero, J., Moreira, A.,
Palomares, E., Torres, T., Durrant, J.R.: J. Am. Chem. Soc. 130, 2906 (2008)
81. Nazeeruddin, M.K., Humphry-Baker, R., Grätzel, M., Murrer, B.A.: Chem. Commun. 719
(1998)
82. Nazeeruddin, M.K., Humphry-Baker, R., Grätzel, M., Wohrle, D., Schnurpfeil, G., Schneider,
G., Hirth, A., Trombach, N.: J. Porphyrins Phthalocyanines 3, 230 (1999)
83. Aranyos, V., Hjelm, J., Hagfeldt, A., Grennberg, H.: J. Porphyrins Phthalocyanines 5, 609
(2001)
84. He, J.J., Hagfeldt, A., Lindquist, S.E., Grennberg, H., Korodi, F., Sun, L.C., Akermark, B.:
Langmuir 17, 2743 (2001)
85. He, J., Benko, G., Korodi, F., Polivka, T., Lomoth, R., Akermark, B., Sun, L., Hagfeldt, A.,
Sundstrom, V.: J. Am. Chem. Soc. 124, 4922 (2002)
86. Komori, T., Amao, Y.: J. Porphyrins Phthalocyanines 6, 211 (2002)
87. Komori, T., Amao, Y.: J. Porphyrins Phthalocyanines 7, 131 (2003)
88. Giribabu, L., Kumar, C.V., Reddy, V.G., Reddy, P.Y., Rao, C.S., Jang, S.R., Yum, J.H.,
Nazeeruddin, M.K., Grätzel, M.: Sol. Energy Mater. Sol. Cells 91, 1611 (2007)
89. Reddy, P.Y., Giribabu, L., Lyness, C., Snaith, H.J., Vijaykumar, C., Chrasekharam, M.,
Lakshmikantam, M., Yum, J.H., Kalyanasundaram, K., Grätzel, M., Nazeeruddin, M.K.:
Angew. Chem. Int. Ed. 46, 373 (2007)
38 1 Introduction
90. Silvestri, F., Garcia-Iglesias, M., Yum, J.H., Vazqueza, P., Martinez-Diaza, M.V., Grätzel,
M., Nazeeruddin, M.K., Torres, T.: J. Porphyrins Phthalocyanines 13, 369 (2009)
91. Eu, S., Katoh, T., Umeyama, T., Matano, Y., Imahori, H.: Dalton Trans. 5476 (2008)
92. Palomares, E., Martinez-Diaz, M.V., Haque, S.A., Torres, T., Durrant, J.R.: Chem. Commun.
2112 (2004)
93. Moreira, A., Lopez-Duarte, I., Martinez-Diaz, M.V., O’Regan, B., Shuttle, C., Haji-
Zainulabidin, N.A., Torres, T., Palomares, E., Durrant, J.R.: J. Am. Chem. Soc. 129, 9250
(2007)
94. Macor, L., Fungo, F., Tempesti, T., Durantini, E.N., Otero, L., Barea, E.M., Fabregat-
Santiago, F., Bisquert, J.: Energy Environ. Sci. 2, 529 (2009)
95. Rawling, T., Austin, C., Buchholz, F., Colbran, S.B., McDonagh, A.M.: Inorg. Chem. 48,
3215 (2009)
96. Nazeeruddin, M.K., Kay, A., Rodicio, I., Baker, R.H., Müller, E., Liska, P., Vlachopoulos,
N., Grätzel, M.: J. Am. Chem. Soc. 115, 6382 (1993)
97. Tennakone, K., Kumara, G., Kumarasinghe, A., Wijayantha, K., Sirimanne, P.: Semicond.
Sci. Technol. 10, 1689 (1995)
98. Brian, O., Daniel, T.: Chem. Mater. 7, 1349 (1995)
99. Tennakone, K., Jumara, G., Kottegoda, I.: J. Phys. D Appl. Phys. 31, 2326 (1998)
100. Meng, Q., Takahashi, K., Zhang, X., Sutanto, I., Rao, T., Sato, O., Fujishima, A., Watanabe,
H., Nakamori, T., Uragami, M.: Langmuir 19, 3572 (2003)
101. Grätzel, M.: Nature 414, 338 (2001)
102. Snaith, H., Moule, A., Klein, C., Meerholz, K., Friend, R., Grätzel, M.: Nano Lett. 7, 3372
(2007)
103. Krüger, J., Plass, R., Cevey, L., Piccirelli, M., Grätzel, M., Bach, U.: Appl. Phys. Lett. 79,
2085 (2001)
104. Wang, P., Zakeeruddin, S., Moser, J., Humphry-Baker, R., Grätzel, M.: J. Am. Chem. Soc.
126, 7164 (2004)
105. Haque, S.A., Tachibana, Y., Willis, R.L., Moser, J.E., Grätzel, M., Klug, D.R., Durrant, J.R.:
J. Phys. Chem. B 104, 538 (2000)
106. Shi, C.W., Dai, S.Y., Wang, K.J., Pan, X., Kong, F.T., Hu, L.H.: Vib. Spectrosc. 39, 99 (2005)
107. Durr, M., Yasuda, A., Nelles, G.: Appl. Phys. Lett. 89, 3 (2006)
108. Yin, X., Zhao, H., Chen, L., Tan, W., Zhang, J., Weng, Y., Shuai, Z., Xiao, X., Zhou, X., Li,
X., Lin, Y.: Surf. Interface Anal. 39, 809 (2007)
109. Greijer, H., Lindgren, J., Hagfeldt, A.: J. Phys. Chem. B 105, 6314 (2001)
110. Murakami, T.N., Ito, S., Wang, Q., Nazeeruddin, M.K., Bessho, T., Cesar, I., Liska, P.,
Humphry-Baker, R., Comte, P., Pechy, P., Grätzel, M.: J. Electrochem. Soc. 153, A2255
(2006)
111. Ramasamy, E., Lee, W.J., Lee, D.Y., Song, J.S.: Appl. Phys. Lett. 90, 173103 (2007)
112. Imoto, K., Takahashi, K., Yamaguchi, T., Komura, T., Nakamura, J.-I., Murata, K.: Solar
Energy Mater Solar Cells 79, 459 (2003)
113. Suzuki, K., Yamaguchi, M., Kumagai, M., Yanagida, S.: Chem. Lett. 32, 28 (2003)
114. Bay, L., West, K., Winther-Jensen, B., Jacobsen, T.: Sol. Energy Mater. Sol. Cells 90, 341
(2006)
115. Saito Y., Kitamura T., Wada Y., Yanagida, S.: Chem. Lett. 1060 (2002)
116. Saito, Y., Kubo, W., Kitamura, T., Wada, Y., Yanagida, S.: J. Photochem. Photobiol. A 164,
153 (2004)
117. Wang, M., Anghel, A.M., Marsan, B., Cevey Ha, N.-L., Pootrakulchote, N., Zakeeruddin,
S.M., Grätzel, M.: J. Am. Chem. Soc. 131, 15976 (2009)
118. Martin, S.T., Herrmann, H., Choi, W., Hoffmann, M.R.: J. Chem. Soc., Faraday Trans. 90,
3315–3322 (1994)
119. Dawson, A., Kamat, P.: J. Phys. Chem. B 105, 960 (2001)
120. Franch, M., Ayllon, J., Peral, J., Domènech, X.: Catal. Today 101, 245 (2005)
121. Anpo, M., Takeuchi, M.: J. Catal. 216, 505 (2003)
122. Chen, X., Shen, S., Guo, L., Mao, S.: Chem. Rev. 110, 6503 (2010)
References 39
123. Ishibai, Y., Sato, J., Akita, S., Nishikawa, T., Miyagishi, S.J.: Photochem. Photobio. A: Chem.
188, 106 (2007)
124. Asahi, R., Morikawa, T., Ohwaki, T., Aoki, K., Taga, Y.: Science 293, 269 (2001)
125. Li, D., Haneda, H., Hishita, S., Ohashi, N., Labhsetwar, N.K.J.: Fluorine chem. 126, 69 (2005)
126. Khan, S.U.M., Al-Shahry, M., Ingler Jr., W.B.: Science 297, 2243 (2002)
127. Luan, Y., Fu, P., Dai, X., Du, Z.: Z. Prog. Chem. 16, 738 (2004)
128. Choi, W., Termin, A., Hoffmann, M.R.: J. Phys. Chem. 98, 13669 (1994)
129. Li, X., Xiong, Y., Li, Z., Xie, Y.: Inorg. Chem. 45, 3493 (2006)
130. Testino, A., Bellobono, I.R., Buscaglia, V., Canevali, C., D’Arienzo, M., Polizzi, S., Scotti,
R., Morazzoni, F.: J. Am. Chem. Soc. 129, 3564 (2007)
131. Wang, S.: Study on Preparation of TiO2 Nanomaterials and Their Application in Solar Energy
Conversion. Doctoral Dissertation of Graduate University of Chinese Academy of Sciences,
(2012)
132. Anpo, M., Kishiguchi, S., Ichihashi, Y., Takeuchi, M., Yamashita, H., Ikeue, K., Morin, B.,
Davidson, A., Che, M.: Res. Chem. Intermed. 27, 459 (2001)
133. Anpo, M.: Pure Appl. Chem. 72, 1265 (2000)
134. Anpo, M., Takeuchi, M.: Int. J. Photoenergy 3, 89 (2001)
135. Santz, P.A., Kamat, P.V.: Phys. Chem. Chem. Phys. 4, 198 (2002)
136. Baker, D.R., Kamat, P.V.: Adv. Funct. Mater. 19, 805 (2009)
137. Gopidas, K.R., Bohorquez, M., Kamat, P.V.: J. Phys. Chem. 94, 6435 (1990)
138. Evans, J.E., Springer, K.W., Zhang, J.Z.: J. Chem. Phys. 101, 6222 (1994)
139. Yamashita, Y., Ono, K., Tanaka, S., Imaeda, K., Inokuchi, H.: Adv. Mater. 6, 295 (1994)
140. Novoselov, K.S., Geim, A.K., Morozov, S.V., Jiang, D., Zhang, Y., Dubonos, S.V.,
Grigorieva, I.V., Firsov, A.A.: Science 306, 666 (2004)
141. Li, G., Li, Y., Liu, H., Guo, Y., Zhu, D.: Chem. Commun. 46, 3256 (2010)
142. Nelson J. Current Opinion in Solid State and Materials Science, p. 87 (2002)
143. Hong, H., Davidov, D., Kallinger, C., Lemmer, U., Feldmann, J., Harth, E., Gügel, A., Müllen,
K.: Synth. Met. 102, 1487 (1999)
144. Ma, W., Yang, C., Gong, X., Lee, K., Heeger, A.J.: Adv. Funct. Mater. 15, 1617 (2005)
145. Bottari, G., Torre, G., Guldi, D., Torres, T.: Chem. Rev. 110, 6768 (2010)
146. Clarke, T., Durrant, J.: Chem. Rev. 110, 6736 (2010)
147. Dillon, A.: Chem. Rev. 110, 6856 (2010)
148. Dresselhaus, M.S., Dresselhaus, G., Saito, R.: Solid State Commun. 84, 201 (1992)
149. Barazzouk, S., Hotchandani, S., Vinodgopal, K., Kamat, P.V.: J. Phys. Chem. B 108, 17015
(2004)
150. Hasobe, T., Fukuzumi, S., Kamat, P.V.: J. Phys. Chem. B 110, 25477 (2006)
151. Vietmeyer, F., Seger, B., Kamat, P.V.: Adv. Mater. 19, 2935 (2007)
152. Lomeda, J., Doyle, C., Kosynkin, D., Hwang, W.-F., Tour, J.: J. Am. Chem. Soc. 130, 16201
(2008)
153. Si, Y., Samulski, E.: Nano Lett. 8, 1679 (2008)
154. Salvio, R., Krabbenborg, S., Naber, W., Velders, A., Reinhoudt, D., van der Wiel, W.: Chem.-
Eur. J. 15, 8235 (2009)
155. Niyogi, S., Bekyarova, E., Itkis, M., McWilliams, J., Hamon, M., Haddon, R.: J. Am. Chem.
Soc. 128, 7720 (2006)
156. Veca, L., Lu, F., Meziani, M., Cao, L., Zhang, P., Qi, G., Qu, L., Shrestha, M., Sun, Y.: Chem.
Commun. 2565 (2009)
157. Salavagione, H., Gomez, M., Martinez, G.: Macromolecules 42, 6331 (2009)
158. Yang, H., Shan, C., Li, F., Han, D., Zhang, Q., Niu, L.: Chem. Commun. 3880 (2009)
159. Wang, S., Chia, P.-J., Chua, L.-L., Zhao, L.-H., Png, R.-Q., Sivaramakrishnan, S., Zhou, M.,
Goh, G., Friend, R., Wee, A., Ho, P.: Adv. Mater. 20, 3440 (2008)
160. Stankovich, S., Piner, R., Nguyen, S., Ruoff, R.: Carbon 44, 3342 (2006)
161. Wang, Y., Chen, X., Zhong, Y., Zhu, F., Loh, K.: Appl. Phys. Lett. 95, 063302 (2009)
162. Zu, S.-Z., Han, B.-H.: J. Phys. Chem. C 113, 13651 (2009)
163. Zhang, H., Lv, X.J., Li, Y.M., Wang, Y., Li, J.H.: ACS Nano 4, 380 (2010)
40 1 Introduction
164. Liu, J.C., Bai, H.W., Wang, Y.J., Liu, Z.Y., Zhang, X.W., Sun, D.D.: Adv. Funct. Mater. 20,
4175 (2010)
165. Du, J., Lai, X.Y., Yang, N.L., Zhai, J., Kisailus, D., Su, F.B., Wang, D., Jiang, L.: ACS Nano
5, 590 (2011)
166. Xiang, Q., Yu, J., Jaroniec, M.: Chem. Soc. Rev. 41, 782 (2012)
167. Zhang, J.T., Xiong, Z.G., Zhao, X.S.: J. Mater. Chem. 21, 3634 (2011)
168. Fan, W.Q., Lai, Q.H., Zhang, Q.H., Wang, Y.: J. Phys. Chem. C 115, 10694 (2011)
169. Zhang, N., Zhang, Y., Xu, Y.: Nanoscale 4, 5792 (2012)
170. Akhavan, O., Ghaderi, E.: J. Phys. Chem. C 113, 20214 (2009)
171. Sun, Y., Wu, Q., Shi, G.: Energy Environ. Sci. 4, 1113 (2011)
172. Kongkan, A., Martinez-Dominguez, R., Kamat, P.V.: Nano Lett. 7, 676 (2007)
173. Wang, H., Hu, Y.: Energy Environ. Sci. 5, 8182 (2012)
174. Long, M., Tang, L., Wang, D., Li, Y., Shuai, Z.: ACS Nano 5, 2593 (2011)
175. Ito, S., Murakami, T., Comte, P., Liska, P., Grätzel, C., Nazeeruddin, M., Grätzel, M.: Thin
Solid Films 516, 4613 (2008)
176. Musić, S., Gotić, M., Iva, M., Popović, S., Turković, A., Trojko, R., Sekulić, A., Furić, K.:
Mater. Sci. Eng. B 47, 33 (1997)
177. Jung, J.H., Kobayashi, H., van Bommel, K.J.C., Shinkai, S., Shimizu, T.: Chem. Mater. 14,
1445 (2002)
178. Cranford, S., Buehler, M.: Nanoscale 4, 4587 (2012)
179. Guo, Y., Jiang, K., Xu, B., Xia, Y., Yin, J., Liu, Z.: J. Phys. Chem. C 116, 13837 (2012)
180. Jiao, Y., Du, A., Hankel, M., Zhu, Z., Rudolph, V., Smith, S.: Chem. Commun. 47, 11843
(2011)
181. Srinivasu, K., Ghosh, S.: J. Phys. Chem. C 116, 5951 (2012)
182. Zhang, H., Zhao, M., He, X., Wang, Z., Zhang, X., Liu, X.: J. Phys. Chem. C 115, 8845 (2011)
183. Wang, S., Yi, L., Halpert, J.E., Lai, X., Liu, Y., Cao, H., Yu, R., Wang, D., Li, Y.: Small 8,
265 (2012)
184. Du, H., Deng, Z., Lu, Z., Yin, Y., Yu, L., Wu, H., Chen, Z., Zou, Y., Wang, Y., Liu, H., Li,
Y.: Synth. Met. 161, 2055 (2011)
185. Liu, C.-J., Burghaus, U., Besenbacher, F., Wang, Z.: ACS Nano 4, 5517 (2010)
186. Nishijima, Y., Ueno, K., Yokota, Y., Murakoshi, K., Misawa, H.: J. Phys. Chem. Lett. 1, 2031
(2010)
187. Schatz, G.C.: Acc. Chem. Res. 17, 370 (1984)
188. Diederich, F.: Nature 369, 199 (1994)
189. Malko, D., Neiss, C., Vines, F., Goerling, A.: Phys. Rev. Lett. 108, 086804 (2012)
190. Loh, K.P., Bao, Q., Ang, P.K., Yang, J.: J. Mater. Chem. 20, 2277 (2010)
Chapter 2
Two-Dimensional Graphene Bridges
Enhanced Photoinduced Charge Transport
in Dye-Sensitized Solar Cells
2.1 Introduction
Graphene, the 2D carbon nanomaterial has drawn much attention nowadays. It is a

zero band gap material [1], and electrons in it are just like massless relativistic parti‐
cles, so it has an excellent electrical conduction in two dimensions [2–6], even at
temperatures close to absolute zero and the Dirac point [1]. Based on its astonishing
advantages in electrics, graphene has been used in conductive glasses [7] organic
photovoltaic cells [8, 9], field effect transistors devices [10], and ultrasensitive
sensors [11, 12].
Dye-sensitized solar cells (DSSCs) have attracted considerable interests because
of their relatively low cost and high efficiency for the photoelectrical conversion of
solar cells, since Grätzel’s group introduced nanostructured TiO2 film into anode
electrodes [13]. Although ~11 % conversion efficiency was obtained [14, 15], further
improvements are still necessary. The major bottleneck is the transport of photo‐
generated electrons across the TiO2 nanoparticle network, which competes with the
charge recombination. To suppress the recombination and improve the transport,
there are several strategies including (1) using composite metal oxides as the semi‐
conductor with different band gaps [16]; (2) preparing the porous structure whose
direction is perpendicular to the conducting substrate [17–22]; and (3) introducing
charge carriers (1D nanomaterials) to direct photogenerated electron [23–25].
Considering their special charge transfer properties, the 1D nanomaterials, such
as the carbon nanotubes (CNTs), were introduced into the electrodes [23–25].
However, the efficiency was increased limitedly because of the point contact between
nanosphere TiO2 and columniform 1D nanomaterials (as shown in Scheme 2.1a).
On the contrary, the 2D graphene may benefit the charge separation not only due to
its excellent conductivity but also good contact with TiO2 nanoparticles. Graphene
is a 2D soft single molecular layered structure, and there are intermolecular forces
such as physisorption, electrostatic binding, or charge transfer interactions between

42 2 Two-Dimensional Graphene Bridges Enhanced Photoinduced Charge …
Scheme 2.1 The differences between (a, c) 1D and (b, d) 2D nanomaterial composite electrodes.
In 2D nanomaterial composite electrodes (graphene bridges), the TiO2 particles can anchor in the
graphene better, and the photoinduced electrons can be captured and transferred by the graphene;
but with the 1D nanomaterial compositing, there are less intermolecular forces and connection
between TiO2 and the 1D nanomaterial, therefore the transfer barrier is larger and the recombination
is much easier to happen
graphene and TiO2 [26], so the nanocrystalline TiO2 can anchor on the graphene
flake compactly (Scheme 2.1b) to form grapheme bridges, which will decrease the
TiO2–TiO2 contacts. Due to its excellent electrical conduction, the bridged graphene
behaves as an electron transfer medium to enhance the electron transport from the
conduction band (CB) of TiO2 at the anchor position quickly (Scheme 2.1d), and the
recombination is reduced as a result.
Herein, we incorporate the 2D graphene into TiO2 nanostructure photoanode to
form graphene bridges in DSSCs, and find that graphene can enhance the charge
transport rate so as to prevent the charge recombination and increase the light-
collection efficiency as well, so the photoelectrical conversion efficiency is
improved.
2.2 Results and Discussion
2.2.1 The Characterization of GO and Graphene
GO was synthesized from flake graphite by a modified Hummers method, and

graphene was produced by chemical reducing and annealing (see detailed methods)
[27]. The image of GO was analyzed through atomic force microscopy (AFM) on
new cleaved mica surface (Fig. 2.1). It can be seen that the lateral dimensions of
graphene are in the range of 0.2–2 μm and about 1 μm on average. Analysis of the
AFM images in Fig. 2.1b revealed the GO sheets with heights in about 1 nm from
the cross-sectional view in Fig. 2.1c, which is somewhat larger than the interlayer
2.2 Results and Discussion 43
Fig. 2.1 a Tapping-mode AFM image of graphene oxide. b The enlargement of a piece of graphene
oxide. c Height profile through the line shown in (b)
spacing of GO (0.73 nm) measured by X-ray powder diffraction (XRD) (Fig. 2.2),
however, the similar results can also be observed in other results [28–32]. In addition,
the X-ray photoelectron spectroscopy (XPS) (Fig. 2.3a) and Fourier transform IR
(FT-IR) (Fig. 2.3c, d) measurements also confirm that the graphite was well exfoli‐
ated into small pieces during the strong oxidation process, and the graphene was
synthesized [33].
Figure 2.4a, b are XPS survey spectra of photoanodes with and without graphene.
The C 1s peak of the electrode with graphene bridges is stronger than that without
graphene, confirming that the graphene was successfully introduced into the
photoanode.
2.2.2 Photocurrent–Voltage (I–V) Characteristics of Different

Electrodes
In order to study the effect of 2D graphene bridges in photovoltaic devices, a series

of DSSCs were fabricated with different content of graphene in photoanode (see
detailed methods). The traditional nanocrystalline electrode (Electrode 1) was also
fabricated to comparison. The thickness of all the different electrodes is the same
(7 μm), and the results are based on the average one among many electrodes for each
ratio.
The parameters and I–V characteristics for solar cells based on the different elec‐
trodes were exhibited in Table 2.1 and Fig. 2.5. The results indicate that the efficiency
(a) (b)
d=0.334nm 2-Theat=26.6 O
Intensity
Intensity
d=0.730nm 2-Theat=12.1O
20 40 60 20 40 60
2-Theta (degree) 2-Theta (degree)
(c)
O
d=0.356nm 2-Theat=25.0
Intensity
20 40 60
2-Theta (degree)
Fig. 2.2 XRD Patterns of a Graphite, b GO, and c graphene
will not be improved with the increase of the content of graphene all the while, and
0.6 wt% GO composite has got the optimization (we name this electrode as Electrode
2). Considering that the GO will lose weight after the reduction, 0.4 % graphene will
be left in the final electrode (about 70 % remaining, Tables 2.2 and 2.3). Moreover,
to compare the differences between the 1D and 2D composite, the CNT composite
photoanode was fabricated in the same way (Electrode CNT), and the content of
CNT is 0.4 %. Hence, Electrode 1, Electrode 2, and Electrode CNT were used in the
following experiments to discuss the effects of graphene.
As shown in Fig. 2.5a, a short-circuit current density (Isc) of 16.29 mA cm−2, an
open-circuit voltage (Voc) of 0.69 V, and the conversion efficiency (η) of 6.97 % for
Electrode 2 were obtained. As compared, Isc, Voc, and η were 11.26 mA cm−2, 0.69 V,
and 5.01 %, respectively, for Electrode 1. There are about 45 % increases in Isc and
39 % in η for Electrode 2, while nearly the same Voc compared with Electrode 1. In
this case, we improve the Isc significantly without sacrificing the Voc, indicating that
graphene is better than CNTs for DSSCs with the same technique [23–25]. The Fermi
level of CNT is between its CB and valence band, which means it is more negative
than its CB. Also, the CB of CNT is more negative (−4.5 eV vs. vacuum) than the
CB of TiO2 (−4.0 eV vs. vacuum). All these make the apparent Fermi level decrease
Fig. 2.3 The C 1s XPS spectrum of a GO and b Electrode 2. The C 1s XPS spectrum of GO
indicates the presence of four types of carbon bonds: C–C (284.8 eV), C–O (286.8 eV), C=O
(287.5 eV), and O–C=O (289.0 eV). Although Electrode 2 has the same C 1s species, the intensities
of oxide species are much weaker. Besides, there is an additional peak at 288.5 eV, corresponding
to the C in the C=N bonds of hydrazone, and FT-IR Spectrum of c GO and d graphene. The
characteristic vibrations are the broad O–H stretching peak at 3411 cm−1, the strong C=O peak at
1730 cm−1, the 3°-C–OH peak at 1460 cm−1, the O–H deformation peak at 1390 cm−1, the C–OH
stretching peak at 1240 cm−1, the C–O stretching peak at 1060 cm−1. The peak at 1622 cm−1 is due
to the skeletal vibrations of unoxidized graphitic domains. The spectrum of graphene confirms the
reduction of GO. The O–H peaks, the C–O peaks, and the C=O peak have been decreased. And a
new absorption band at 1550 cm−1 may be attributed to the skeletal vibration of the graphene sheets
in the CNT composite cell, and the Voc is sacrificed as a result [24]. However,
graphene is a zero band material, [1] and its calculated work function is a little more
positive (−4.42 eV vs. vacuum) than that of CNT, which is close to work function
of FTO (−4.4 eV vs. vacuum) [7]. Hence, the apparent Fermi level may not be
decreased in the graphene-bridged cell, and the Voc is not affected. By the way, the
introduced 2D graphene bridges increase the charge transport, and suppress the
electrons in TiO2 to recombine with the dye and redox species, and hence the Voc
will not be decreased.
Table 2.1 The parameters for solar cells based on the different electrodes
GOa or CNT No. of Voc (V) Isc (mA/cm2) FF (%) η (%)

loading (wt%) electrode
0 1 0.69 11.26 64.5 5.01
0.6 2 0.69 16.29 62.0 6.97
2.5 3 0.51 4.40 67.9 1.52
8.5 4 0.49 1.49 62.8 0.46
0.4 CNT 0.42 3.35 41.1 0.58
a
The amount of GO before being reduced. It has been reduced to graphene when it was fabricated
to cells. The weight loss is about 30 % after the reduction
Fig. 2.4 The XPS survey spectra of the photoanode a with and b without graphene bridges

Efficiency (IPCE) Performance of Different Electrodes
In addition, the improved Isc is also reflected in the increased IPCE performance
(Fig. 2.6). Comparing with Electrode 1, the IPCE response at all wavelengths is
enhanced by a factor of 1.56, while the maximum IPCE (IPCEmax) was 83 % at the
wavelength of 481 nm for Electrode 2. This increase in IPCE may be due to two
factors: (1) as graphene is a 2D nanomaterial, the TiO2 particles can anchor in it very
well, so the photoinduced electrons can be captured by the graphene easily. Because
graphene is an excellent electron carrier, the rapid photoinduced electron transport
is produced, and the recombination and back electron transfer are suppressed as a
result; (2) more porous structure of the anode has been formed due to the contribu‐
tions of graphene, which increase the light scattering as shown in the SEM (Fig. 2.7),
nitrogen adsorption experiments (Fig. 2.8a) and UV-vis spectra (Fig. 2.8b), respec‐
tively. Furthermore, there is a blue shift to the IPCEmax. The IPCEmax shift is common
in the former reports, but the reason is not very clear [34–36]. We presume that it
Table 2.2 The XPS data of GO

Name Start BE End BE Height FWHM Area (P) At.% SF
(eV) (eV) counts (eV) CPS. (ev)
C 1s 292.48 281.04 12417.87 2.1 54207.59 67.84 1
O 1s 537.73 527.05 24492.12 2.19 60343.76 31.59 2.85
N 1s 410.51 394.85 221.41 0.25 755.03 0.57 1.77
Table 2.3 The XPS data of graphene

Name Start BE End BE Height FWHM Area (P) At.% SF
(eV) (eV) counts (eV) CPS. (ev)
C 1s 294.2 281.05 46997.54 1.52 108204.5 87.05 1
O 1s 538.91 526.54 6680.79 3.46 27021.26 8.53 2.93
N 1s 407.49 395.2 2627.7 3.13 9327.07 4.43 1.8
Note The XPS data shows that the content of oxygen was decreased after the reduction. The weight
loss is oxygen in majority, and the weight losing is about 25 % (keep the content of C as a constant).
Also, the graphene will lose another 7 % weight after calcining at 450 °C in air. 2 So after the
treatment, the GO will have about 70 % left (30 % lose)
Fig. 2.5 a Photocurrent–voltage characteristics of different electrodes. The sensitizer was N3

(ruthenium dye). The cell active area was 0.20 cm2 and the light intensity was 100 mW cm−2; b
The back current of different electrodes (the insert of the figure is the enlarge plot of back current)
may be due to the introduction of many pores in the composite electrode (Fig. 2.7b).
And these pores extended the pore distribution to 500 nm (Fig. 2.8a), which includes
the wavelength around 480 nm, and the scattering is much easier happened. And
there is more reflectance effect around 480 nm (Fig. 2.8b), hence the effect in IPCE
is more remarkable, then the blue shift can take place. And it needs further analysis.
Figure 2.7 shows the SEM images of the electrodes after the reduction and
annealing, and there are more pores in Electrode 2 (Fig. 2.7b) than in Electrode 1
(Fig. 2.7a) and Electrode CNT (Fig. 2.7c). These pores are in the micrometer scale.
And the increased pore distributing was proved by the N2 adsorption and desorption
Fig. 2.6 Photocurrent action

spectra of conventional
electrode (Electrode 1), TiO2-
graphene composite electrode
(Electrode 2) and the CNT
composite electrode. IPCE
(%) = (1241Isc)/(λPin) × 100,
where Isc is short-circuit
current and Pin is the power of
the incident light
isotherms (Fig. 2.8a). The pore size distribution was calculated by the Barrett–
Joyner–Halenda (BJH) analysis from N2 desorption isotherms. The distribution
shows that Electrode 2 has large pores in the mesoporous (20–500 Å) and macro‐
porous (>500 Å) regions. Also, the pore size was extended to the micrometer scale
for the graphene composite electrode, which was formed when the H2O and N2H4
were released from the anode after the reduction and annealing [37]. This means that
the incorporating of graphene will introduce a hierarchical structure and increase the
roughness factor. But the introduced roughness did not increase the surface area. The
surface areas for Electrode 1 and Electrode 2 are 52.23 and 51.67 m2 g−1, respec‐
tively, calculated by the multipoint Brunauer–Emmett–Teller (BET) method, which
are nearly the same, indicating that the graphene composite electrode will not
improve the dye absorption, but the light scattering was increased based on the UV-
vis measurement (Fig. 2.8b). The diffuse spectrum shows that the introduced pore
performs as a light capture center, and the light scattering was increased by 7 % at
all wavelengths comparing with the traditional electrode. The increased light scat‐
tering will improve the photo-to-current conversion efficiency. But because the
increased light diffuse cannot transform to the photoexcited electrons absolutely, the
efficiency increased by light scattering must be less than 7 %. However, the IPCE
measurement shows that the efficiency was increased by 56 %. That means the major
improvement is not attributed to the light scattering.
With the higher graphene loading, there will be light-harvesting competition
between N3 dye and graphene, and the dye adsorption was reduced because the
TiO2 is surrounded by graphene. At the same time, the excessive graphene can act
as a kind of recombination center instead of providing electron pathway, and the
short circuit will happen easily. All these lead to the decrease of the total efficiencies.
Fig. 2.7 SEM micrographs of (a) Electrode 1, (b) Electrode 2 and (c, d) Electrode CNT. The
graphene composite electrode is more porous comparing with other electrodes. Figure (d) is the
enlargement of figure (c). Figure (d) indicated that the CNT cannot connect to the TiO2 particle
well, and it is the point of contact. The CNT also cannot disperse in the anode well, which may also
restrict the efficiency
Fig. 2.8 a BJH pore size distributions calculated from N2 desorption isotherms, and b Measured
diffuse reflectance of sensitized anodes
Fig. 2.9 a Nyquist and b Bode diagrams of the electrochemical impedance spectra of different
electrodes. c Is the equivalent circuit of the device. The spectra were measured under the
illumination of one sun at open-circuit potential
2.2.4 Electrochemical Impedance Spectra (EIS) Measurement

of Different Electrodes
To confirm the effect of 2D nanomaterial in transporting the electrons and restraining

the recombination in the nanostructure, EIS, a powerful tool to clarify the electronic
and ionic transport processes in DSSCs [38, 39] was measured under the illumination
of one sun at open-circuit potential (Fig. 2.9). In the Nyquist plots (Fig. 2.9a), the
biggest semicircle at medium frequencies is attributed to photoinjected electrons in
the TiO2 or back reaction from the injected electrons in TiO2 to the electrolyte (ca.
1–100 Hz), which is the most important in the device. As shown, the semi-cycle at
intermediate frequency regions were decreased for Electrode 2 indicating the fast
redox activity of the electrolyte in the working electrode interface. The equivalent
circuit of the devices (Fig. 2.9c) has been reported [40–42]. Using Z View software
and fitting the semicircle of medium frequencies gives R3 (resistance of charge
transfer) and C1 (capacity). Compared with Electrode 1, Electrode 2 has a smaller
R3 (21.66 Ω vs. 29.86 Ω), but a bigger C1 (1655 μF vs. 965 μF), showing that the
charge recombination in Electrode 2 has been reduced. These results were also
supported by the Bode plots (Fig. 2.9b). The medium frequency peak of Electrode
2 is shifted to lower frequency (from ~10.1 to ~6.9 Hz), comparing with Electrode
1. Therefore, incorporating graphene into the TiO2 nanocrystalline anode will
increase the electron transport rate restraining the charge recombination.
In addition, some 1D nanomaterials are also good electron carriers, to prove the
outstanding properties of the 2D nanomaterial, the CNT composite electrodes were
also fabricated in the same route, and the surface area of electrode CNT was almost
the same with other electrodes (53.08 m2 g−1). As shown in the former figures (elec‐
trode CNT), this electrode had a low efficiency (0.58 %), and the IPCE was low too.
The reasons were observed by the EIS measurement. The electrode CNT has the
largest R3 (140 Ω) and smallest C1 (28 μF), also the medium frequency peak is at
Fig. 2.10 The UV-vis

transmittance spectrum of
different electrodes
107 Hz (the highest). These results indicate the CNT is not an ideal charge carrier
in this system. The low dark current proves that CNT is a good electron capture [25]
(Fig. 2.5b), and it is a good conductor indubitably. But its performance in DSSCs is
not so good. That may be due to two factors: (1) CNT is a 1D nanomaterial. The
columniform structure cannot attach the TiO2 particles well, and it is isolated in the
anode (Fig. 2.7d). It will not perform as the electron capturer and carrier well; (2)
the CNT is not coherent, and the captured electrons cannot transport to the FTO (due
to the CNT is not coherent) or TiO2 (due to the energy level reason), the resistance
is high (R3) as a result. Also, many electrons are captured by CNT without trans‐
porting; hence the electrons are superfluous, and the recombination is easy to happen,
so the capacity (C1) is low and the frequency is high. Also, in the CNT composite
cell, the light scattering is the lowest, but the transmittance is the highest (Fig. 2.10),
and the efficiency decreased to some extent.
2.2.5 Operational Principle of Device
Due to its excellent electrical conduction, the 2D graphene bridges behave as an

electron transfer channel, which can transport the photo-induce electron quickly [23].
And the energy level of graphene is between the CB of TiO2 and FTO [43].
Scheme 2.2a shows the operational principle. Under illumination, the CB of semi‐
conductor TiO2 receives the electrons from photoexcited dye. Because the TiO2 is
anchored with 2D graphene, and the graphene is homogeneous in the system, the
excited electrons are captured by the graphene without any obstruction. The collected
electrons can transport from TiO2 to the conductive substrate quickly and effectively
through graphene bridges (Scheme 2.2b), and hence the adverse reactions (recom‐
bination and back reaction) are suppressed.
Scheme 2.2 The operational principle of the device: The introduced 2D graphene bridges perform
as an electron accepter, and transfer the electrons quickly. Hence, the recombination and back
reaction are suppressed
2.3 Conclusions
In summary, 2D graphene is a rising star in material science, and we introduced it

into the nanocrystalline anode of DSSC successfully. The short-circuit current
density was increased by 45 %, and the conversion efficiency was increased by 39 %.
This may be due to a lower recombination and a faster electron transport with the
introduction of graphene chiefly. Also the higher light scattering is benefit to a certain
extent. This study proved that the 2D charge carriers have more advantages than the
1D material such as the compactly anchor and the homogenous dispersion. The
incorporation of 2D nanomaterial with 0D nanoparticles can also be extended to
other electronic devices, especially using graphene as an electron accepter and
charge transfer medium.
2.4 Postscript
At the beginning of this work, we were not exactly sure that graphene is suitable or
not for DSSC, so a lot of attempts were taken by introducing graphene to each part
of the cell. Here, I think it is interesting to write the results, which is benefit to
understanding the interaction between graphene and TiO2.
We have tried to introduce graphene into the counter electrode and also the elec‐
trolyte, but with the further understanding of graphene, we think the energy level is
not match for this approach. Also, there are some difficulties in the film formation
for graphene on FTO, otherwise, one can try fabricating of TiO2-graphene hetero‐
junction, and more promising data may come out.
Please note, in this work, we did not introduce a compact TiO2 layer in DSSC,
neither we did not introduce hierarchical structure into the photoanode by the post‐
treatment with TiCl4. For the former, based on the result we obtained, we have tried
to build a compact TiO2 film, but as discussed in the text, for energy level reasons,
2.4 Postscript 53
the graphene captured electrons are difficult to transfer to the conduction band of
TiO2. Hence, the compact TiO2 film will affect the electron transport, thus reducing
the efficiency of the battery. Herein, it is another proof that proves that the electron
transfer direction is from TiO2 to graphene. The reason for abandoning treatment
with TiCl4 is because it requires annealing for second time. While graphene will be
removed to some extent during each annealing circle, it is not easy to precisely
control the amount of graphene. Then the experiment is difficult to repeat.
Meanwhile, we have noticed some papers give a hypothesis that the charge is
transferred from graphene to TiO2. Based on my knowledge, we need a further proof
to give such a conclusion. At least, based on our experiment result, it is more reason‐
able to speculate that the electrons transfer direction is from TiO2 to graphene. Also,
we have tried using graphene as dye (in the absent of N3) to excite photoelectrons,
but no obvious photocurrent is observed. That means no electrons are injected to the
conduction band of TiO2 from graphene, or the recombination is very fast. On the
other hand, even if the electrons can be transferred from graphene to TiO2, the main
body of the charge transport should still be TiO2. Considering that the bottle neck of
DSSC is the charge transport efficiency, there would be no great benefit to introduce
graphene into the cell.
However, a possible way is to design graphene quantum dots or chemically
modify graphene, which will adjust the energy level of carbon material. One may
get a proper size for the charge transport, but the position of valance band should be
also considered, which needs to match the energy level of electrolyte.
I have to say, what we used is not exactly graphene, but reduced graphene oxide
instead, which presents defects. These defects make it not a true zero band gap but
a semimetallic material. I tend to consider it as a semiconductor, then it should form
a heterojunction structure with TiO2, which would produce a certain bending at the
band edge. Hence, in the experiment, we can observe that with the addition of
graphene, the Voc is decreased to a certain extent. It may due to a reduction of the
apparent Fermi level, which is more clear in a dark voltage spectrum.
There are still a lot of mechanisms that are not clear enough in this work. Hence,
further modification and improvement are quite necessary. For example: (1) Because
it is not possible to add the compact layer, the graphene is easy to cause a short circuit
to some extent. How to choose a compact layer with suitable energy level may be
the key to further improve efficiency. (2) The real situation of charge transport is not
very clear. It still needs a deeper study to reveal whether graphene really possesses
advantages comparing with other carbon materials. Some more advantaged testing
tools or methods can be introduced to explore the mechanism.
2.5 Detailed Methods
Synthesis of GO: GO was synthesized from flake graphite (average particle diameter
of 4 μm, Qingdao Tianhe Graphite Co. Ltd., Qingdao, China) by a modified
Hummers method [27]. All of the other chemicals were of analytical reagent grade,
and used without further purification. Briefly, graphite powder (5 g) and NaNO3
(3.75 g) were placed in a flask. Then, concentrated H2SO4 (375 mL) was added slowly
while stirring in an ice water bath. KMnO4 (22.5 g) was added gradually under stir‐
ring over 1 h, and stirred for another 2 h. After stirring vigorously for 5 days at room
temperature, the mixture was stirred at 35 °C for 2 h, and then diluted with 5 wt%
H2SO4 aqueous solution (700 mL) over 1 h. With stirring at 98 °C for 2 h, when the
temperature was reduced to 60 °C, 30 wt% H2O2 (30 mL) was added, and the mixture
was stirred for 2 h at room temperature. The mixture was centrifuged and washed
with aqueous solution of 3 wt% H2SO4/0.5 wt% H2O2 (2 L) for 15 times. Then the
bottom solid was washed with 3 wt% HCl (2 L) aqueous solution with a similar
procedure, and one time using H2O (2 L). After adding another 2 L deionized water
and dispersing the solid using vigorous stirring and bath ultrasonication for 30 min,
the final water solution was treated with a weak basic ion-exchange resin to remove
the remaining HCl acid. The final solution was concentrated to 7.5 mg/mL.
Preparation of different DSSCs: To prepare graphene composite electrode, 3 g
poly-vinylalcohol (PVA, MW 22,000) water solution (30 %) was dissolved with the
mixed solvent containing 9 ml H2O and 12.5 ml ethanol. Then, 1.62 g P-25 TiO2 was
added to make the suspension of TiO2. After the suspension was dispersed, Electrode
1 was prepared by doctor blade technique on fluorine-doped tin oxide (FTO) glass
for comparison. As to the preparation of graphene composite electrodes, 1.25 ml GO
aqueous solution (7.5 mg/ml) was added into the suspension (0.6 wt%). After that,
it was stirred and dispersed in ultrasonic cleaner for about 30 min, respectively, so
as to let GO get dispersed in the suspension well, and it can anchor TiO2 perfectly,
and then Electrode 2 was prepared. Also the electrodes with 2.5 wt% and 8.5 wt%
content of GO were prepared (Electrode 3 and 4) in the same way. In addition, the
CNT composite electrode (0.4 wt%) was fabricated as 1.62 g P-25 TiO2, 3 g PVA
water solution (30 %), 9 ml H2O, 12.5 ml ethanol and 6.5 mg CNT (Shenzhen Nano‐
tech Port Co. Ltd., the diameter is 20–40 nm, and the purity is >95 %), and the
fabricated route is the same. And the thickness of all the different films is about 7
μm. At last, the electrodes were treated under hydrazine vapor at 40 °C for 24 h to
reduce GO. After the films were rinsed with deionized water and dried by heating
to 40 °C in vacuum for 3 h, the electrodes were annealing at 400 °C under Argon
flow for 3 h, and the GO was transformed to graphene totally. Then, they were
calcined at 450 °C for 1 h in air. The as-prepared films were sensitized by soaking
in ethanol solution containing 3 × 10−4 M the ruthenium dye, Ru(dcbpy)2(NCS)2
(N3) (dcbpy = 2,2′-bipyridine-4,4′-dicarboxylic acid) for 24 h. The DSSCs were
composed of a sensitized photoanode, a platinum counter electrode, and an electro‐
lyte. The electrolyte was composed of 0.5 M LiI, 0.05 M I2, 0.5 M tert-butylpyridine,
and 0.6 M 1-propyl-3-methylimidazolium iodide in 3-methoxypropionitrile.
Characterization: The AFM images were measured with SPA400 (Seiko Instru‐
ments Inc.) on new cleaved mica surface in tapping mode in air. The XRD patterns
were obtained by using an X’Pert PRO MDP with CuKα radiation (λ = 1.5405 Å)
with 30 mA and 40 kV. XPS data were obtained with an ESCALab220i-XL electron
spectrometer from VG Scientific using 300 W AlKα radiation. The FT-IR spectrum
was recorded by Bruker Equinox 55 FTIR spectrometer. The I–V characteristics of
2.5 Detailed Methods 55
the cell were measured by an electrochemical analyser (CHI630A, Chenhua Instru‐

ments Co., Shanghai) under solar simulator illumination (CMH-250, Aodite Photo‐
electronic Technology Ltd, Beijing) at room temperature. The IPCE was measured
by illumination with monochromatic light, which was obtained by a series of light
filters with different wavelengths. SEM images were obtained using a JEOL
JSM-6700F scanning electron microscope at 3.0 kV. UV-vis spectra were recorded
on a Hitachi Model U-4100 Spectrophotometer. The nitrogen adsorption and desorp‐
tion isotherms at the temperature of liquid nitrogen (77 K) were measured on a
Quantachrome Autosorb-1 sorption analyzer with prior degassing under vacuum at
200 °C overnight. Total pore volumes were determined using the adsorbed volume
at a relative pressure of 0.99. Multipoint BET surface area was estimated from the
relative pressure range from 0.05 to 0.2. The pore size distribution of the electrodes
was analyzed using the BJH algorithm. The EIS was carried out on a Zahner IM6e
impedance analyzer (Germany) in the frequency range 0.02 Hz–100 kHz with illu‐
mination of 100 mW/cm2.
References
1. Freitag, M.: Nat. Nano. 3, 455 (2008)

2. Geim, A.K., Novoselov, K.S.: Nat. Mater. 6, 183 (2007)
3. Novoselov, K.S., Geim, A.K., Morozov, S.V., Jiang, D., Zhang, Y., Dubonos, S.V., Grigorieva,
I.V., Firsov, A.A.: Science 306, 666 (2004)
4. Stankovich, S., Dikin, D.A., Dommett, G.H.B., Kohlhaas, K.M., Zimney, E.J., Stach, E.A.,
Piner, R.D., Nguyen, S.T., Ruoff, R.S.: Nature 442, 282 (2006)
5. Katsnelson, M.I.: Mater. Today 10, 20 (2007)
6. Li, D., Kaner, R.B.: Science 320, 1170 (2008)
7. Wang, X., Zhi, L., Müllen, K.: Nano Lett. 8, 323 (2008)
8. Liu, Z.F., Liu, Q., Huang, Y., Ma, Y.F., Yin, S.G., Zhang, X.Y., Sun, W., Chen, Y.S.: Adv.
Mater. 20, 3924 (2008)
9. Liu, Q., Liu, Z.F., Zhang, X.Y., Yang, L.Y., Zhang, N., Pan, G.L., Yin, S.G., Chen, Y.S., Wei,
J.: Adv. Funct. Mater. 19, 894 (2009)
10. Di, C.A., Wei, D.C., Yu, G., Liu, Y.Q., Guo, Y.L., Zhu, D.B.: Adv. Mater. 20, 3289 (2008)
11. Schedin, F., Geim, A.K., Morozov, S.V., Hill, E.W., Blake, P., Katsnelson, M.I., Novoselov,
K.S.: Nat. Mater. 6, 652 (2007)
12. Ang, P.K., Chen, W., Wee, A.T.S., Loh, K.P.: J. Am. Chem. Soc. 130, 14392 (2008)
13. O’Regan, B., Grätzel, M.: Nature 353, 737 (1991)
14. Nazeeruddin, M.K., De Angelis, F., Fantacci, S., Selloni, A., Viscardi, G., Liska, P., Ito, S.,
Takeru, B., Grätzel, M.: J. Am. Chem. Soc. 127, 16835 (2005)
15. Chiba, Y., Islam, A., Watanabe, Y., Komiya, R., Koide, N., Han, L.: Jpn. J. Appl. Phys. 45,
L638 (2006)
16. Kong, F.T., Dai, S.Y., Wang, K.J.: Adv. OptoElectron. 2007, 1 (2007)
17. Kong, F.T., Dai, S.Y.: Prog. Chem. 18, 1409 (2006)
18. Kuang, D.B., Brillet, J., Chen, P., Takata, M., Uchida, S., Miura, H., Sumioka, K., Zakeeruddin,
S.M., Grätzel, M.: ACS Nano 2, 1113 (2008)
19. Mor, G.K., Shankar, K., Paulose, M., Varghese, O.K., Grimes, C.A.: Nano Lett. 6, 215 (2006)
20. Grimes, C.A.: J. Mater. Chem. 17, 1451 (2007)
21. Zhu, K., Neale, N.R., Miedaner, A., Frank, A.J.: Nano Lett. 7, 69 (2007)
13364 (2008)
23. Kongkanand, A., MartinezDominguez, R., Kamat, P.V.: Nano Lett. 7, 676 (2007)
24. Brown, P., Takechi, K., Kamat, P.V.: J. Phys. Chem. C 112, 4776 (2008)
25. Yen, C.Y., Lin, Y.F., Liao, S.H., Weng, C.C., Huang, C.C., Hsiao, Y.H., Ma, C.C.M., Chang,
M.C., Shao, H., Tsai, M.C., et al.: Nanotechnology 19, 1 (2008)
26. Williams, G., Seger, B., Kamat, P.V.: ACS Nano 2, 1487 (2008)
27. Becerril, H.A., Mao, J., Liu, Z., Stoltenberg, R.M., Bao, Z., Chen, Y.: ACS Nano 2, 463 (2008)
28. Niyogi, S., Bekyarova, E., Itkis, M.E., McWilliams, J.L., Hamon, M.A., Haddon, R.C.: J. Am.
Chem. Soc. 128, 7720 (2006)
29. Xu, Y.X., Bai, H., Lu, G.W., Li, C., Shi, G.Q.: J. Am. Chem. Soc. 130, 5856 (2008)
30. Gomez-Navarro, C., Weitz, R.T., Bittner, A.M., Scolari, M., Mews, A., Burghard, M., Kern,
K.: Nano Lett. 7, 3499 (2007)
31. Nakashima, N., Tomonari, Y., Murakami, H.: Chem. Lett. 31, 638 (2002)
32. Nakayama-Ratchford, N., Bangsaruntip, S., Sun, X., Welsher, K., Dai, H.J.: J. Am. Chem. Soc.
129, 2448 (2007)
33. Nethravathi, C., Rajamathi, M.: Carbon 2008, 46 (1994)
34. Yum, J.H., Hagberg, D.P., Moon, S.J., Karlsson, K.M., Marinado, T., Sun, L.C., Hagfeldt, A.,
Nazeeruddin, M.K., Grätzel, M.: Angew. Chem. Int. Ed. 48, 1576 (2009)
35. Tian, H.N., Yang, X.C., Chen, R.K., Pan, Y.Z., Li, L., Hagfeldt, A., Sun L.C., Chem. Commun.
3741 (2007)
13364 (2008)
37. Stankovich, S., Dikin, D.A., Piner, R.D., Kohlhaas, K.A., Kleinhammes, A., Jia, Y., Wu, Y.,
Nguyen, S.T., Ruoff, R.S.: Carbon 45, 1558 (2007)
38. Kubo, W., Murakoshi, K., Kitamura, T., Yoshida, S., Haruki, M., Hanabusa, K., Shirai, H.,
Wada, Y., Yanagida, S.: J. Phys. Chem. B 105, 12809 (2001)
39. Zhao, Y., Zhai, J., Wei, T.X., Jiang, L., Zhu, D.B.: J. Mater. Chem. 17, 5084 (2007)
40. Park, N.G., Kim, K.M., Kang, M.G., Ryu, K.S., Chang, S.H., Shin, Y.: J. Adv. Mater. 17, 2349
(2005)
41. Wang, Q., Moser, J.E., Grätzel, M.: J. Phys. Chem. B 109, 14945 (2005)
42. van de Lagemaat, J., Park, N.G., Frank, A.J.: J. Phys. Chem. B 104, 2044 (2000)
43. Grätzel, M.: Nature 414, 338 (2001)
Chapter 3
Bioinspired Stacking Structures
for Photoelectric Conversion
3.1 Granum-Like Stacking Structures with TiO2–Graphene

Nanosheet for Improving Photoelectric Conversion
3.1.1 Introduction
Nature has developed an efficient and precise method to convert photo energy to
chemical energy in green plants. Previous investigations [1, 2] have reported the
mechanism of photosynthesis in which the granum is the photosynthetic reaction
centre in the primary reaction while photoelectric conversion is realized through
the first-step energy conversion. A granum is composed of 10–100 thylakoids in
stacking structure with antenna pigment molecules and cytochrome b6f complex
in each thylakoid. After light is captured via the antenna pigment molecules, elec‐
trons are excited and transferred to the b6f complex, which can collect and
transfer the electrons to the thylakoid lumen (Scheme 3.1a). There are four key
elements to carry out the primary reaction: (1) a stack structure; (2) the antenna
pigment molecules, which absorb the light; (3) the photosynthetic reaction centre
where photoelectric conversion takes place; and (4) a fast electron transport
process, whereby photo-induced electrons are collected by the complex in each
thylakoid and transported to the outside. However, because these thylakoids are
rather fragile and can function only when inserted in an organism, it is hard to
study them directly in man-made systems. Hence, it is useful to construct a similar
photoelectric system to imitate the chloroplast.
Graphene, a two-dimensional (2D) carbon structure, has excited considerable
interest as a new electron transport material in recent materials science and
condensed-matter studies [3–8]. Furthermore, negatively charged Ti1-δO2 nano‐
sheets have drawn much attention recently due to their significant optical and
electric properties [9, 10]. Meanwhile, graphene (GR) can perform as strong

58 3 Bioinspired Stacking Structures for Photoelectric Conversion
Scheme 3.1 The scheme of the (a) real and (b) imitated photosystem with stacking structure.
a The granum is stacked by thylakoids (right), and the enlarge figure (left) is the photoelectric
conversion in each thylakoid. The antenna pigment molecules absorbed the light and excited
electrons, these electrons were captured by electron acceptor, and then transport out of the system.
b Imitating the stacking structure in leaf (right), the blue layer is TiO2 nanosheet and the gray layer
is graphene. The Enlarge figure (left) shows TiO2 performs as the antenna pigment molecules and
graphene performs as the electron accepter. They packed together to form a unit and several units
stacked together to form a granum-like structure
electron-collector and carrier in a TiO2/graphene composite system because their

energy levels and physical properties are compatible [11, 12].
Inspired by nature, we packed TiO2 and graphene nanosheets into a 2D unit
(TiO2/graphene) similar to a thylakoid, where TiO2 performs as a photoelectric
conversion centre to absorb light and excite the electrons, and graphene is like the
b6f complex to capture electrons and transport them out of the circuit. Finally, several
units packed together to form a three-dimensional stacking structure similar to
granum (Scheme 3.1b).
3.1 Granum-Like Stacking Structures with TiO2–Graphene … 59
3.1.2 Results and Discussion
3.1.2.1 Characterization of Nanosheets
In this work, TiO2 nanosheets were synthesized according to the method reported by
Sasaki [9, 13]. Graphene oxide (GO) was synthesized from flake graphite by a modi‐
fied Hummers method [14]. The product was characterized by X-ray diffraction
(XRD). A series of sharp peaks of the fabricated TiO2 colloidal suspension (Fig. 3.1a)
were assigned as the 010, 020, 030 facets, respectively. The highly ordered and repet‐
itive peaks reflect the fact that the TiO2 swelled into nanosheets with layer spacing of
0.85 nm calculated by the Bragg equation, in agreement with previous reports [15]. The
transmission electron micrographs (TEM) show that the nanosheets of TiO2 were
fabricated (Fig. 3.1b). The XRD pattern of graphite had a peak centered at 2θ = 26.6°,
corresponding to a d-spacing of 0.334 nm, which was absent in the GO sample
(Fig. 3.1c). Once a new peak at 2θ = 12.1° was observed, corresponding to the (002)
inter-planar spacing of 0.73 nm, it was determined that graphite had been exfoliated
and pristine GO was formed [16]. The images of GO and TiO2 were analyzed through
atomic force microscopy (AFM) on new cleaved mica surface (Fig. 3.1e, f). It can be
seen that the lateral dimensions of GO is in the range of 0.2–2 μm and about 1 μm on
average. And TiO2 is smaller than GO in the lateral dimensions scale with a range of
several hundreds of nanometers on average. Analysis of the AFM images in Fig. 3.1g
revealed the thinnest GO sheets with heights in about 1 nm from the cross-sectional
view, and with a height of 1.1 nm for TiO2, which is somewhat larger than the inter‐
layer spacing measured by XRD, however, the similar results can also be observed in
other results [17–21]. According to the AFM images (Fig. 3.1e, f), it appears as though
TiO2 and GO display the nanosheet morphology directly.
(a) (b) (c) (d)
2- Theta (Degree) 2- Theta (Degree)
(e) (f) (g) (h)
Fig. 3.1 The XRD, TEM, and AFM images of TiO2 (a, b, e, g) and graphene oxide (c, d, f, h).
These images indicated the nanosheets were fabricated, and the similar nanosheets morphology
showed the benefit to be stacked forming a granum-like structure
Fig. 3.2 The fabrication of the stacked film with LBL method
3.1.2.2 Fabricating and Optical Spectra of the Stacking Film
Typically, a layer-by-layer (LBL) assembly technique was used in our experiment

[9, 22, 23] (Fig. 3.2). The building blocks of Ti1-δO2 and GO are both negatively
charged. With a positively charged molecular poly(diallyldimethylammonium
chloride) (PDDA) as counter-ions, the stacking structure was fabricated. After
stacking, GO was reduced to graphene by hydrazine and annealed under argon flow
at 400 °C, and the PDDA was then removed by calcining at 450 °C in air. At the
same time, the TiO2 stacking films were treated with the same approach, including
the hydrazine and annealing in Ar and air.
The LBL assembly on quartz glass was monitored using UV–vis absorption
spectra. The (TiO2)n nanosheets had an absorption peak at 262 nm [9], and the inten‐
sity was observed to increase linearly with increasing numbers of TiO2 units
(Fig. 3.3a). Introducing graphene make the peak position a small blue shift. The
absorption of these films also increased linearly with the number of units (TiO2/
graphene) as they increased from 5 to 25 (Fig. 3.3b). The slope of the absorption
increased from 0.034 to 0.073 and a little movement of the peak position were both
due to the introduction of graphene, which was attributed to the graphene absorption
and the interactions between d-π electrons of titanium oxide and graphene [23].
3.1.2.3 Characterizations of the Stacking Films
We investigated the morphology of the stacking structure by AFM. The film of

(PDDA/TiO2/PDDA/graphene)25 was scratched first, and the height gradient was
observed (Fig. 3.4a). In the cross section shown in Fig. 3.3a, the step is clearly visible
and the height appears to be ~65 nm after stacking 25 bilayers of PDDA/TiO2/PDDA/
graphene. This corresponds to ~2.6 nm for one unit, which agrees well with the
previous report [13, 23]. The XPS measurement showed the additional C–O bond in
Fig. 3.3 The absorption of (a) (TiO2)n and (b) (TiO2/graphene)n with different numbers of units.
The insert figures indicated the unit of the film. In (a), the units were TiO2 nanosheets (the blue
block), while in (b) TiO2 (the blue block) and graphene (the black block) made up each unit. The
absorption peak kept increasing with the numbers of units
TiO2/graphene, which indicated the existence of graphene after reduction and

calcining (Fig. 3.4b). The O 1 s XPS spectrum of (TiO2)n only had a single peak
centered at 530.2 eV characteristic of the Ti–O bond (black line), while the (TiO2/
graphene)n spectrum had an additional peak at 532.3 eV corresponding to the typical
C–O bond in graphene (green line). Meanwhile, the low-angle XRD peak can be
indexed to a layered structure with a repeating unit of 0.97 nm indicating a separation
distance of TiO2−–TiO2 or TiO2–graphene layers after annealing (Fig. 3.4c) [13].
The XRD peak of TiO2–graphene was similar with TiO2−–TiO2, which might indi‐
cate the few-layers graphene exist and with a similar thickness with the TiO2 sheets.
TEM images showed the stacking structure directly. The image darkness of the
scraped film shows an increased contrast in the edge, demonstrating that the nano‐
sheets were stacked together, and the nanosheets retained this morphology even after
the annealing treatment (Fig. 3.4d) [13].
3.1.2.4 Photocurrent Response for Different Layers Stacking Films
The photocurrent response experiments were carried out under pulsed monochro‐
matic UV light (λ = 254 nm) irradiation without a bias voltage in the acetonitrile
solution with I3−/I− as the oxidation/reduction pair (Fig. 3.5). Figure 3.5 shows the
photocurrent response obtained from different numbers of stacking layers of
(TiO2)n and (TiO2/GR)n (n = 5, 10, 15, 20, 25). When the number of stack layers was
5, the signal was not steady, but we can see that the (TiO2/GR)5 film had a higher
cathodic current response of 60 nA versus 27 nA (Fig. 3.5a, b).
When the number of layers increased to 10, the cathodic photocurrents of both
(TiO2/GR)10 and (TiO2)10 stacking films were increasing to 108 and 44 nA, respec‐
tively. The photocurrent was prompt, steady, and reproducible during several on/off
Fig. 3.4 The characterizations of the stacking films. a The AFM image indicates the height gradient
of (PDDA/TiO2/PDDA/graphene)25 (above). The bottom one is the cross section corresponding to
the morphology image. The maximum of 65 nm height fall reflected 2.6 nm thickness of each unit.
b The XPS spectrum of the stacked films after reduction, the additional C–O bond (green line) in
O1s proved the existence of graphene. c The XRD pattern and d TEM image of the stacked films
after annealing. This phenomenon indicated the layered structure was kept after annealing
cycles with a linear increase relative to the change of the film thickness. This is due
to the fact that increasing the TiO2 content in the photoanode, increased the amount
of light absorbed, and thus increased the number of excited electrons. Notably, when
the numbers of unit were increased to 15, the growth rate of cathodic photocurrent
slowed down and even decreased. For (TiO2)15, adding five additional layers only
increased the current from 44 to 49 nA. For the (TiO2/GR)15, an additional five layers
make little change, with a current of 98 nA. When the numbers of layers increased
to 20, the photocurrent of (TiO2)20 film was reduced to 25 nA, while that of the
(TiO2/GR)20 film changed from cathodic to anodic photocurrent (−570 nA). Upon
further increasing the number of layers, the anodic current was observed in both of
the stacked films. The final photocurrent at 25 layers for the (TiO2)25 film was −46
nA, while that of the (TiO2/GR)25 was 20 times higher (−965 nA) (Fig. 3.5c, d).
Fig. 3.5 The photocurrent response obtained from different stack structures: (TiO2)n and
(TiO2/GR)n (n = 5, 10, 15, 20, 25). a The photocurrent of (TiO2)n (n = 5, 10, 15). The (TiO2)10 film
had a higher response than the (TiO2)5 film, while the current of (TiO2)15 film was nearly the same
with the (TiO2)10. b The change in current for the (TiO2/GR)n (n = 5, 10, 15) film, which increases
at first ((TiO2/GR)5 to (TiO2/GR)10) then later decreases a little ((TiO2/GR)10 to (TiO2/GR)15). c
When the unit (TiO2)n increased to 20, the current decreased, and (TiO2)25 showed an anodic current.
d The anodic current was observed in (TiO2/GR)20, and was much larger than the former test. The
current was still increasing when the units’ number increased to 25. e The comparison of the current
response of the whole different stacks. The blue and red columns indicted the current of (TiO2)n
and (TiO2/GR)n respectively. And the enlarged one was the current response of the mono-TiO2
stacked films. The graphene composite film had a much more sensitive current response, and the
current was much easier to change to the negative direction. (Here we defined the positive value as
the cathodic current and the negative one as the anodic current.) f Is the device used in the
photoelectrical measurement
(Here we defined the positive value as the cathodic current and the negative value
as the anodic current according to our testing system.) All of this data is shown
simplified in Fig. 3.5e. The (TiO2/GR)n stack had a much higher photocurrent
response, which we ascribe to the effect of graphene. More importantly, the direction
of the photocurrent changed when the film became very thick, an effect which is
usually not observed due to the film thickness limitation [23]. Anodic photocurrent
is beneficial in solar cells, however, the stacked film must be thick enough to achieve
this in a finished photovoltaic device. Therefore, better understanding of the mech‐
anism in this bioinspired photoelectric conversion system can enable novel strategies
in the design of solar cells.
3.1.2.5 Operational Principle of the Device
In general, there are two directions for electron transport in the stacking films
(Scheme 3.2a). TiO2 absorbs photons and electrons are excited from the valence band
to the conduction band (Route 1). The photo-induced electrons are then transported
to the FTO (the red arrows), while the remaining hole is reduced by I− in the elec‐
trolyte (Route 5). The electrons transported from the electrolyte to the FTO will yield
anodic current, thus efficiently outputting current to the external circuit and
improving device performance [24]. However, transport mechanisms moving elec‐
trons away from the anode cannot be ignored as these adversely affect the efficiency
[25]. The photoexcited electrons in Route 1 may be transported to the electrolyte
(Route 4), and the remaining holes are reduced by electrons from the external circuit
(Route 6), thus generating cathodic current. Current flow in this direction greatly
decreases the total current that can be supplied by the solar cell. Aside from cathodic
current, recombination (Route 2) can also decrease the number of excited electrons
without supplying current to the external circuit [26]. Graphene serves to capture
electrons, collecting the excited electrons and transporting them to the FTO
(Scheme 3.2a, shaded region) [12]. Because the work function of graphene
(−4.42 eV vs. vacuum) [27] is lower than that of the conduction band of TiO2
(−4.0 eV vs. vacuum), the electrons can only be transported from TiO2 to graphene
and not in the opposite direction (Route 3 in Scheme 3.2a) [12]. The TiO2/graphene
unit functions like a diode, promoting the electron transport in one direction, i.e.,
from TiO2 to graphene, and then to FTO, while the recombination route is
suppressed.
With regards to the stacking structure (Scheme 3.2b, c), there are two directions
for electron transport in the film. One is transporting to the substrate (labeled in red),
and the other is to the electrolyte or recombination (labeled in blue). In the pure
TiO2 stacking structure (Scheme 3.2b), there is a contact barrier between the TiO2
and the TiO2 stacking layers, and excited electrons must get over the barrier in order
to be transported. Holes left in the TiO2 nanosheet after excitation increase the like‐
lihood of recombination taking place in the stacking film. Photo-induced electrons
are also easily transferred to the electrolyte, since the electrolyte permeates the layers
of nanosheets. This has the effect of increasing the cathodic current [28] and thus
Scheme 3.2 Scheme of the electrons transport. In a, the photoexcited electrons are transported
from the valance band of TiO2 to the conductive band (Route 1), and through the grapheme film,
as depicted by the red arrows, to the substrate. The blue lines indicate the adverse transfer. The
introduced graphene (shaded region) can collect the electrons and suppress the Route 2 and 4 and
suppress the recombination. The benefit of graphene can be seen in b and c. The graphene directs
the charge transfer direction, and suppress the recombination and back reaction, which performs
as a charge transport bridge
the observed current in this device was low. To increase the photo current, we
inserted 2D graphene into the TiO2 layers (Scheme 3.2c). As graphene is an excellent
electron-collector [11], the excited electrons are captured by nearby graphene,
suppressing recombination in the titania. Meanwhile, graphene can transport the
electrons out of the circuit thus increasing the photo current. Because of the matching
of the energy levels, the existence of a graphene bridge allowed the electrons to be
more easily transported to FTO. Furthermore, by increasing the film thickness the
photocurrent continually improved. This indicates that the inclusion of graphene in
the device may allow us to break the titania “thickness limit” that has limited the
photon harvesting ability of previous devices [29].
A primary factor limiting the efficiency of solar cells is the competition between
the optical path length required for light absorption and the charge diffusion length
Ln [30]. For example, the Ln in dye sensitized solar cells can be attributed to Eq. (3.1)
[23, 31]:
√
Ln = Dn 𝜏n (3.1)
where Dn is the apparent electron diffusion coefficient and τn is the apparent lifetime
of photo-induced electrons. Typical values of Dn are in the range of 10−12–10−8
m2 s−1, and τn values are found in milliseconds [30, 32]. Therefore, the diffusion
length is calculated to be in the range from 7 to 30 μm at light intensities up to 0.1
sun. This implies that the film cannot be too thick or else recombination in the TiO2
becomes the preferred pathway. The carrier diffusion constant of graphene is
0.3 m2 s−1 [33], which is much larger. Moreover, graphene has a high electron mobi‐
lity, and the captured electrons can be transported out quickly. Therefore, the lifetime
of photo-induced electrons τn should be greater in (TiO2/GR)n films. Hence,
according to Eq. (3.1), we can deduce that the Ln of graphene should be much greater,
while the captured electrons can be transported more easily to the external circuit.
This should allow a great improvement in the anodic current. By this assumption,
we propose that if the granum-like stacking structure were introduced as the photo‐
anode in a DSSC device, we can exceed the existing thickness limitation of the
photoanode and further improve the photoelectric conversion efficiency.
3.1.2.6 Electrolyte Diffusion in the Stacking Film
Additionally, the photoelectric conversion is affected by the electrolyte diffusion.

The cyclic voltammograms of different layered films were measured in a propylene
carbonate solution containing 0.1 mol/L LiClO4. As depicted in Fig. 3.6a, the cyclic
voltammogram showed a well-defined anodic peak but a poor cathodic peak, as has
been reported previously. The peaks can be accounted for by the oxidation/reduction
of the TiO2, accompanying insertion/extraction of Li+ ions into/from the stacking
structure (Reaction Formula 3.1a).
( ) ( )
NH4 , H 4δ TiIV O + X(Li+ + e− ) ↔ LiX NH4 , H 4𝛿 TiIII
(1−δ) 2 X
TiIV O
1−δ− X 2 (3.1a)
The anodic peak current and the square root of the potential sweep rate can be
ascribed to Randle–Sevcik equation:
Ip/ = kn3∕2 D1∕2 cv1∕2

A (3.2)
where Ip is peak current, A is the surface area of the electrode, k is a constant

k = 0.4463[F3/(RT)]1/2 (F is the Farady constant, R is the molar gas constant, T(K)
is the temperature, k = 2.69 × 105 at room temperature), n is the transfer electron
Fig. 3.6 The cyclic voltammograms of the stack structures. a The current was improved with the
sweep rate increasing. b The linear relationship between the oxidation peak current and the square
root of the potential sweep rate. The (TiO2/GR)n film would bring a larger diffusion coefficient
number, D is the diffusion coefficient of the electrolyte ion, c is the concentration of

the electrolyte, and v is the sweep rate.
The linear relationship between the oxidation peak current and the square root of
the potential sweep rate suggests that the diffusion of Li+ is a rate-determining step
in oxidation of Ti3+ in the stacking structure (Fig. 3.6b). According to Eq. (3.2), the
slope is kn3/2D1/2c. As k, n, c are constants with the same experiment conduction, the
slope is proportional to D1/2. With the same TiO2 content, the (TiO2/GR)10 had a
larger slope (0.48) than (TiO2)10 (0.29). The same results also can be observed from
(TiO2/GR)25 and (TiO2)25 (0.81 vs. 0.61), indicating inserting graphene into the
TiO2 layers will increase the diffusion coefficient, which improved the cell because
the redox reaction was taken place faster. Meanwhile, with the same TiO2 content,
the (TiO2/GR)n film had a higher anodic peak current, indicating more electrons were
involved in redox reactions than in the pure titania film. Thus, the introduced
graphene suppresses recombination and prolongs the lifetime of photo-generated
electrons.
3.1.3 Conclusions
In summary, we have imitated the function and structure of granum in green plants
by fabricating TiO2/graphene stacked structures. With TiO2 performing the role of
the antenna pigment molecules and graphene replacing the b6f complex, we found
that the graphene would promote the electron–hole separation efficiency. The (TiO2/
GR)25 can bring a 20 times higher anodic current than (TiO2)25, and the current
increased with thicker films. Importantly, the anodic current, and the device function
overall, improved greatly with increasingly thicker films. These results express the
initial performance of the device and demonstrate the interface effect for titania
nanocrystal solar cells. With this improvement, titania solar cells can break the
thickness bottleneck and achieve greater efficiencies.
3.1.4 Detailed Methods
3.1.4.1 Synthesis of TiO2 Nanosheet
TiO2 nanosheet was synthesized by the Sasaki method. First, Cs-doped TiO2 was
obtained through calcining with TiO2 and CsCO3. TiO2 and CsCO3 were mixed
uniformly and calcined at 800 °C for one hour, and then heated at 800 °C for 20 h
to 2 cycles. The Cs in the product was exchanged by H in a 1 M HCl solution, and
the solution-to-solid ratio was 100 cm3/g. The exchange was taken for 3 cycles and
change the acid for a fresh one every 24 h. The acid-exchanged product was washed
with water and dried in air. Finally, the protonic titanate was stirred with a 0.017
mol/dm3 tetrabutylammonium hydroxide solution. The solution-to-solid ratio was
250 cm3/g. After stirring for 10 days, an opalescent solution was obtained.
3.1.4.2 Fabricating the Stacking Film
The stacked film was fabricated by LBL method. Substrates such as quartz and FTO
were deeply cleaned and dipped in polyethylenimine solution (Mw = 750000, 2.5
g/L, pH = 9) for 20 min.
After rinsing with water, TiO2 (0.08 g/L, pH = 9) nanosheet were assembled on
them by immersing for 10 min. Then the substrates were dipped in poly(diallyldi‐
methylammonium (PDDA, Mw = 100,000–200,000) chloride) (20 g/L, pH = 9) for
10 min. Again rinsed with water, the substrates were immersed in GO solution
(0.08 g/L, pH = 9) for another 10 min. Then the substrates were dipped in PDDA
for 10 min after washing. Here, one cycle was finished. Multilayered films with the
different number of layers were fabricated by repeating the cycle.
To compare with pure TiO2 film, GO was not introduced into the cycle, and the
operation is nearly the same. After the stacking, GO was reduced to graphene by
hydrazine and annealing under argon flow at 400 °C, and the PDDA was removed
by calcining at 450 °C in air. All of the measurements were carried out in atmosphere
at room temperature on several samples.
3.1.4.3 Characterization
The AFM images were measured with SPA400 (Seiko Instruments Inc.) and Veeco
MultiMode 8 Scanning Probe Microscope with SNL (Sharp Nitride Lever) probe on
FTO surface or new cleaved mica surface in tapping mode in air. TEM images
obtained using a JEOL 1011 transmission electron microscope at 100 kV and JEOL
2100 at 200 kV. The film XRD patterns were obtained using a Bruker D8 Advance
Instrument with Cu Kα radiation (λ = 1.5405 Å). XPS data were obtained with an
ESCALab220i-XL electron spectrometer from VG Scientific using 300 W Al Kα
radiation. The I–V characteristics were measured by an electrochemical analyzer
(CHI630A, Chenhua Instruments Co., Shanghai) with I2/I− as the electrolyte under
254 nm monochromatic light (13 mW/cm2) at room temperature. The cyclic voltam‐
metry measurement was on the Zahner IM6e impedance analyzer (Germany) with
propylene carbonate containing 0.1 mol/L LiClO4 as the supporting electrolyte. The
counter and reference electrodes were platinum sheet and Ag/Ag+/acetonitrile,
respectively.
3.2 Stacking Nanostructures of Polyaniline with Graphene

Oxide as the Dopant and Template
3.2.1 Introduction
Polyaniline (PANI) is one of the most promising conducting polymers because of

its environment stability, facile process ability, low cost, and reversible control of
its conductivity [34–36]. At the current time, PANI nanostructures such as hollow
spheres, nanotubes and nanofibers or nanowires even their junctions have received
great attention because of not only unique properties of their bulk materials, but also
showing characters of surface, size, and quantum effect of nanomaterials. These
unusual properties lead to promising applications of the nanostructured PANI in
nanomaterials and nanotechnology [37]. Up to date, various nanostructures of PANI
described above have been prepared by a hard-template or soft-template method
[38]. However, no paper dealing with layered nanostrutures of PANI has been
reported yet.
Graphene is a single layer of carbon and a promising candidate for next-generation
nano-electronic devices because of its high conductivity and mobility as well as
excellent tensile modulus [39]. For instance, it shows an excellent high mobility in
two dimensions even at temperatures close to absolute zero and the Dirac point [40].
However, pure graphene cannot be dissolved or dispersed even after the lasting
ultrasonic treatment [41, 42] that limits its application in technology. To solve this
problem, various modified graphene by chemical modifications or non-covalent
functionalizes have been reported [42, 43]. Among those modified graphene, GO is
the most easily synthesized and dispersed graphene derivatives, which have been
widely used in material chemistry and basic research [44]. For instance, various
PANI-GO composites have been prepared in the presence of acidic dopant [45]. As
well known, GO is characterized to be a lamellar solid with un-oxidized aromatic
regions and aliphatic regions containing phenol, carboxyl, and peroxide groups
induced by oxidation [46]. The carboxyl group on the surface of GO might make
itself as the dopant of PANI because of its proton doping mechanism, resulting in a
delocalized poly-semiquinone radical cation [47]. As our best knowledge, there is

no paper of PANI doped directly with GO as the dopant. The unique properties of
GO mentioned above therefore promised us to directly dope PANI with GO as the
dopant.
Herein, the layered nanostructures of PANI, for the first time, are self-assembled
by using GO directly doping the emeraldine base form of PANI (PANI-EB)
dissolved in m-cresol and the ethanol as the dopant. Typical preparation process of
the PANI-GO was as follows: GO aqueous solution (10 mg/ml) synthesized by a
modified Hummers method [12, 14] was slowly added under stirring into the solution
of PANI-EB dissolved in m-cresol and the ethanol. The resultant dark-green precip‐
itate was centrifuged, followed by washing with ethanol for several times, and finally
dried at 40 °C in vacuum for 24 h. Morphology of the PANI-GO is strongly affected
by the content of additional GO into PANI-EB solution, and the layered nanostruc‐
ture was only obtained at a high additional content of the GO aqueous solution (e.g.,
3600 or 4800 μl). Interestingly, the GO was served as dopant and template at the
same time, which was discussed based on the structural characterizations by Scan‐
ning electron microscope (SEM), UV–vis and FTIR spectra, X-ray photoelectron
spectroscopy (XPS), XRD, and room-temperature conductivity measured by a four-
probe method.
3.2.2 Results and Discussion
GO was synthesized from flake graphite by a modified Hummers method [12, 14],
and its structures was characterized by AFM, FTIR, XRD and XPS (Fig. 3.7). AFM
images released that the GO is a flake structure with about 1 μm on average of the
lateral dimensions of graphene, and about 1 nm in thickness (Fig. 3.7a). XRD pattern
of GO only shows a peak at 2θ = 12.1° corresponding to the (002) inter-planar
spacing of 7.30 Å. It differs from that of the flake graphite, which has a peak centered
at 2θ = 26.6°, corresponding to a d-spacing of 3.34 Å (Fig. 3.7b), indicating that
graphite has been exfoliated and pristine GO was formed [44]. FTIR of the GO
showed that the characteristic peaks of GO such as broad O–H stretching peak at
3405 cm−1, the strong C=O peak at 1730 cm−1, the O–H deformation peak at
1403 cm−1, the C–OH stretching peak at 1222 cm−1, and the C–O stretching peak at
1083 cm−1 are observed (Fig. 3.7c), which were consistent with previous reports
[42]. Additionally, it has been demonstrated by XPS that the GO has four types of
carbon bonds such as C–C (284.8 eV), C–O (286.8 eV), C=O (287.5 eV), and
O–C=O (289.0 eV) (Fig. 3.7d). All above characterizations are identical to standard
GO and there are carboxylic groups on the surface of GO, which can be used as
proton dopant of PANI.
PANI-GO was directly prepared by adding GO aqueous solution into the PANI-
EB solution dissolved in a mixture solution of m-cresol and ethanol. As compared
with other methods reported previously [45], the method provided by us is simpler
and more inexpensive because of emitting inorganic or organic acids as the dopant.
3.2 Stacking Nanostructures of Polyaniline with Graphene … 71
Fig. 3.7 a Tapping-mode AFM image of graphene oxide. b XRD Patterns of Graphite (short dash
dot) and GO (solid). c The FT-IR Spectrum of GO. d The C1s XPS spectrum of GO
Moreover, it is found that morphology of the PANI-GO is affected by the additional

content of GO into PANI-EB solutions. As shown in Fig. 3.8a, the PANI-EB powder
is granular morphology in shape. However, the layered structures became clearer
when the content of the additional GO is increased (Fig. 3.8b–d, and also in
supporting information). Interestingly, uniform and smooth layered structures of the
PANI-GO are formed when a high content of the additional GO (e.g. 3.6 ml) was
used as shown in Fig. 3.8c. Also, the low-angle XRD prove the performance of the
lamellar structure (Fig. 3.9a). However, when the amount of GO is over-abundant
(Fig. 3.8d), the layered structure is still kept, but the surface is more smooth, which
is due to the GO coating. The average thickness of the layers and the average distance
of the inter-layers were measured to be about 25 and 35 nm (Fig. 3.8c), respectively.
These results indicated that GO is served as a template in the formation of the layered
structures via a self-assembly process.
The molecular structure of the layered structures of PANI-GO was characterized
by FTIR and UV–Vis. spectrum, XPS and XRD to prove doping function of GO in
the layered structures. First, the FTIR spectral of the PANI-GO, PANI-EB and GO
were measured as shown in Fig. 3.9b. All character bands of PANI chains, such as
C=C stretching deformation of quinoid at 1576 cm−1, benzene rings at 1497 cm−1,
Fig. 3.8 The SEM images of (a–d) doped PANI with different additional content of GO
(a) (b)
(1)
(2)
(3)
2-Theta (Degree)
Fig. 3.9 a The low-angle XRD of GO doped PANI. b The FT-IR spectra of (1) GO, (2) PANI-EB
and (3) GO doped PANI
the C–N stretching of secondary aromatic amine at 1302 cm−1 [48] was observed in
the PANI-GO. This is similar to that of the PANI-EB (Fig. 3.9b), indicating the
PANI-GO has same polymeric chain structure of the PANI. However, some differ‐
ences from either PANI-EB or GO are observed. For instance, the PANI-EB reveals
a band at 827 cm−1, which attributed to the out-of-plane bending vibrations of C–H
band in the aromatic ring. However, it became weak after doping, because the PANI
chain was anchored by the GO plane, and aromatic structures were conjugated with
the GO plane via π-stacking. The out-of-plane bending vibrations were restricted
naturally. Moreover, a strong and broad absorption at 3400 cm−1 assigned as O–H
stretching vibration of GO, and the absorption at 1726 cm−1 and 1622 cm−1 corre‐
sponded to the C=O stretching of COOH groups and the O–H bending vibration are
disappeared in the PANI-GO. These results suggested that the PANI-GO not only
has main chain of the PANI, but also the PANI is doped with GO as the dopant.
To further provide more positive evidences for supporting the doping function
of GO in the PANI-GO, UV–Vis. spectrum of the PANI-GO dissolved in m-cresol
solvent were measured as compared with PANI-EB. As shown in Fig. 3.10, a
strong absorption at 630 nm, which is assigned as the excitation of the quinoid
ring of PANI [49], is observed in the PANI-EB, indicating that the PANI-EB is
insulator. On the other hand, the intensity of this peak in the PANI-GO is signif‐
icantly decreased, indicating that GO doping took place on the quinoid ring of
PANI to form conductive PANI-GO. In general, this peak in the emeraldine salt
form (ES), which is also called as conducting state of PANI, is disappeared
instead of new peak at 950 nm appeared [49]. In case of the PANI-GO, however,
the peak at 630 nm is still present and no new peak at long wavelength is
observed, predicting that the conductivity of the PANI-GO might be not high due
to low doping degree. The prediction is consistent with room temperature—
conductivity of 1.7 × 10−6 S/cm as measured by a four-probe method. The poor
room-temperature conductivity might be due to lower doping degree. In order to
prove it, XPS of the PANI-GO was measured and compared with that of PANI-
EB. As shown in Fig. 3.11a, b, the N1s spectra of the PANI-GO was de-convo‐
luted into quinonoid imine (=N– structure), benzenoid amine (–NH– structure),
and positively charged nitrogens (N+ structure) as reported [50]. As compared
with XPS of the PANI-EB, the intensity of the –N= and =N– structure in the
PANI-GO was decreased, and the N+ was formed, indicating that GO doping took
place on the quinoid segment of the poly-emeraldine chain. The doping degree
assigned as the ratio of N+/N was increased fast when the doping amount below
1.0 ml in this system, but it became a constant when the GO kept adding
(Fig. 3.11c), that is may be due to the PANI chain was packed in the GO layers,
and the over-abundant GO cannot connect with the chain because of the space
resistance. And the doping degree was calculated to be 0.3 that is lower than that
of 0.5 for the full doped PANI-ES [49]. As a result, the lower doping degree
might be one reason for the low room-temperature conductivity of the PANI-GO.
Additionally, although the carboxylic groups on the surface of GO can be used as
the dopant of PANI, the GO-itself actually is insulating, which might be another
reason for the low room-temperature conductivity. In order to improve the
conductivity at room temperature, the PANI-GO was re-doped with 1.0 M HCl
as the dopant. The room-temperature conductivity of the re-doped PANI-GO
achieved as high as 5.8 × 10−4 S/cm, which is enhanced by 341 times, indicating
Fig. 3.10 The UV–vis spectra of a PANI-EB, b PANI-GO and c GO in m-cresol
Fig. 3.11 The N1s spectra of a PANI-EB and b PANI-GO, c is the doping degree to the amount
of GO adding
that facile re-doping process is an efficient approach to improve the electrical

properties of the PANI-GO.
According to SEM imagines and structural characterizations described above, it
is proposed that the planar π–π structure and the carboxyl groups on the surface of
GO are served as the template and dopant, respectively, resulting in novel layered
nanostructures of the PANI-GO. The proposed molecular structure of the layered
structures is shown in Scheme 3.1. It assumes that the PANI-EB is doped with the
carboxyl groups on the surface of GO as the dopant, leading to the polymeric chain
of the PANI along the planar π–π structure of GO as the template to form layered
nanostructures. In other words, the formation of the layered nanostructures of the
PANI-GO is caused by a cooperation effect of doping and template function of GO.
3.2.3 Conclusion
In summary, the layered structures of the PANI-GO were directly prepared by adding
GO aqueous solution into the PANI-EB solution dissolved in a mixture solution of
m-cresol and ethanol. As compared with methods reported previously, the method
provided is simpler and more inexpensive because of saving inorganic or organic
acids as the dopant that opens a new way to prepare hybrid materials of PANI with
GO. It was proposed that the π–π planar structure of GO and the carboxyl groups on
the surface of GO are served as the template and dopant, respectively, that leads to
formation of the novel layered structures. The doping function of GO in the PANI-
GO has been demonstrated by structural characterizations and electrical properties
measured by a four-probe method. Although the conductivity of the layered struc‐
tures can be enhanced by a post-doping process, further improvement of the electrical
properties of the PANI-GO is desired.
3.2.4 Detailed Methods
The synthesis of GO doped PANI The emeraldine base form of PANI (PANI-EB)
was bought from Aldrich. The PANI-GO was directly prepared by adding the GO
aqueous solution into the PANI-EB solution dissolved in m-cresol and the ethanol
solvent. Typical preparation process was as follows: GO aqueous solution
(10 mg/ml) was added slowly under stirring into the PANI-EB solution dissolved in
m-cresol and the ethanol solvent. A dark-green precipitate was formed in the reactor
immediately. Keep stirring for 10 h, and the precipitate was centrifuged, washed
carefully with ethanol for several times until the supernatant became colorless, and
followed by drying at 40 °C in vacuum for 24 h to product PANI-GO. A series of
PANI-GO samples were prepared at different content of additional GO into the
PANI-EB solution in order to study the effect of additional content of GO into PANI-
EB solution on morphology and molecular structure of the PANI-GO.
Characterization Morphology of the PANI-GO was measured by Atom Force
Microscope (AFM) and SEM, respectively. The AFM images were measured with
SPA400 (Seiko Instruments Inc.) on new cleaved mica surface in tapping mode in
air, whereas SEM images were recorded using a JEOL JSM-6700F at 3.0 kV.
Molecular structure of the PANI-GO was measured by Fourier transform IR (FT-
IR) spectrum (Bruker Equinox 55 FTIR spectrometer), UV-visible spectrum (Hitachi
U4100 spectrophotometer), XRD (X’Pert PRO MDP with Cu Kα radiation
(λ = 1.5405 Å) with 30 mA and 40 kV) and XPS (ESCALab220i-XL electron spec‐
trometer from VG Scientific using 300 W AlKα radiation). The conductivity at room
temperature was measured by a four-probe method with Keithley 4200.
3.3 Postscript
3.3.1 Discussion About Experiment Details
(1) Reduction method for GO. In the experiment of fabrication stacking structure
of (TiO2/GO)n, there are many methods to reduce GO to graphene. Considering
the experimental conditions, UV irradiation and hydrazine reduction are suit‐
able for our system. Honestly speaking, hydrazine reduction is not the best
method, one of the reasons is the toxic of hydrazine hydrate. Second, the film
reduced by hydrazine tends to fall off. But the subsequent experiments found
that the UV light cannot totally remove PDDA, so we have to take the hydrazine
method in subsequent experiments. For hydrazine method, we can either use its
vapor or soak the film in hydrazine solution. Based on the experiment, we have
found the vapor reduction was insufficient, comparing with the soak way. But
it is important to control the exposure period, the film is easy to fall off if the
time is long, and too short GO will not be completely reduced.
(2) Phenomenon observed when using UV irradiation to reduce (TiO2/GO)n film.
When using an ultraviolet light (Xe lamp) to reduce (TiO2/GO)n, we found with
increasing of irradiation time, the film gradually darken, while the hydrophilic
surface became hydrophobic, which means GO is likely to be reduced.
However, after two hours, with the increase of exposure time, no substantial
change in the UV-visible absorption spectrum. The XPS data also showed that
GO was not sufficiently reduced even the time is long enough. At this moment,
we asked ourselves, is the reduction caused by photoexcited electrons of TiO2
or direct by UV irradiation. Hence, we fabricated the pure GO stacking film and
irradiated with UV light. We also found the similar phenomenon as what
observed in the present of TiO2. Hence, we cannot give the conclusion if the
TiO2 can reduce GO in this system, because between GO and TiO2 there is still
a layer of PDDA. Also, as we known from SAXRD, this irradiation method
cannot completely degrade PDDA. Are the photoexcited electrons can over‐
come the barrier of polymer to reduce GO? It is still not clear.
(3) Ordering. In the experiment, we also examined which will show a better
response by adjusting the last layer between graphene and TiO2. Preliminary
results showed that depositing graphene on the outside will bring an obvious
enhancement of current response. Hence, we adopted this condition throughout
the experiment. Also we noticed the height of GO should be higher than TiO2
during fabrication of the stacked structure. If not, there is no advantages in the
current response, and sometimes even worse than pure TiO2 stacked film. We
speculated that if graphene is incoherent, or if it cannot contact with the conduc‐
tive glass, graphene will become a recombination center, trapped electrons
cannot be exported, bringing an adversely impact.
3.3 Postscript 77
3.3.2 Perspective
(1) As described in the article, this stacking method, is expected to exceed the
thickness limit of solar cells, which will lead to higher efficiency, but there are
many difficulties to be overcome, and the most important is the purity and area
of graphene. On the one hand we need to obtain high purity graphene, that the
electrons can transmit in a high-speed, but if so, the pure graphene is lack of
oxygen, which has a low interaction with TO2. Hence, it is difficult to form a
stable stacking film with metal oxide, as the film is not strong enough. Also, the
commonly used TiO2 is electric neutrality, how can make it closely assemble
with graphene is very important.
(2) This stacking structure brings new inspiration, which is very promising: that is
composite two materials (conductor–semiconductors, semiconductor–semi‐
conductor, conductor–insulator) with this layer-by-layer method to form a
stacking structure. With this structure, first, we can build in series or parallel
heterojunction; second, we can design a parallel circuit. Considering the nano‐
sheet itself is only a nanometer thickness, if we can well control the electron
transport, and avoid tunneling, it can be fabricated to supercapacitors. And a
more attractive direction is this method may be adopted in integrated chip tech‐
nology and bring a new revolution.
(3) In the work of PANI-GO, have to say, though we get a new morphology of
polyaniline, and find a new dopant, the conductivity of final product is still not
high. We believe that one of the reasons is that doping degree is not high; and
the second reason is that the poor conductivity of GO-itself. If there are ways
to reduce GO to graphene without destroying the molecular chains and morphol‐
ogies of polyaniline, it may bring a huge change in the conductivity and bring
a possibility in controlling current directional transmission.
References
1. Li, X., Fan, T., Zhou, H., Chow, S.-K., Zhang, W., Zhang, D., Guo, Q., Ogawa, H.: Adv. Funct.
Mater. 19, 45 (2009)
2. Shimoni, E., Rav-Hon, O., Ohad, I., Brumfeld, V., Reich, Z.: Plant Cell 17, 2580 (2005)
4. Geim, A.K., Novoselov, K.S.: Nat. Mater. 6, 183 (2007)
6. Stankovich, S., Dikin, D.A., Dommett, G.H.B., Kohlhaas, K.M., Zimney, E.J., Stach, E.A.,
Piner, R.D., Nguyen, S.T., Ruoff, R.S.: Nature 442, 282 (2006)
7. Katsnelson, M.I.: Mater. Today 10, 20 (2007)
8. Song, J., Yin, Z., Yang, Z., Amaladass, P., Wu, S., Ye, J., Zhao, Y., Deng, W.-Q., Zhang, H.,
Liu, X.-W.: Chem. Eur. J. 17, 10832 (2011)
9. Sakai, N., Ebina, Y., Takada, K., Sasaki, T.: J. Am. Chem. Soc. 126, 5851 (2004)
10. Sasaki, T., Watanabe, M.: J. Phys. Chem. B 101, 10159 (1997)
11. Williams, G., Seger, B., Kamat, P.V.: ACS Nano 2, 1487 (2008)
12. Yang, N., Zhai, J., Wang, D., Chen, Y., Jiang, L.: ACS Nano 4, 887 (2010)
13. Sasaki, T., Ebina, Y., Fukuda, K., Tanaka, T., Harada, M., Watanabe, M.: Chem. Mat. 14, 3524
(2002)
14. Hummers, W.S., Offeman, R.E.: J. Am. Chem. Soc. 80, 1339 (1958)
15. Sasaki, T., Watanabe, M.: J. Am. Chem. Soc. 120, 4682 (1998)
17. Niyogi, S., Bekyarova, E., Itkis, M.E., McWilliams, J.L., Hamon, M.A., Haddon, R.C.: J. Am.
Chem. Soc. 128, 7720 (2006)
18. Xu, Y.X., Bai, H., Lu, G.W., Li, C., Shi, G.Q.: J. Am. Chem. Soc. 130, 5856 (2008)
19. Gomez-Navarro, C., Weitz, R.T., Bittner, A.M., Scolari, M., Mews, A., Burghard, M., Kern,
K.: Nano Lett. 7, 3499 (2007)
20. Nakashima, N., Tomonari, Y., Murakami, H.: Chem. Lett. 31, 638 (2002)
21. Nakayama-Ratchford, N., Bangsaruntip, S., Sun, X., Welsher, K., Dai, H.J.: J. Am. Chem. Soc.
129, 2448 (2007)
22. Yao, H.-B., Wu, L.-H., Cui, C.-H., Fang, H.-Y., Yu, S.-H.: J. Mater. Chem. 20, 5190 (2010)
23. Manga, K.K., Zhou, Y., Yan, Y., Loh, K.P.: Adv. Funct. Mater. 19, 3638 (2009)
24. Grätzel, M.: Nature 414, 338 (2001)
25. Yen, C.Y., Lin, Y.F., Liao, S.H., Weng, C.C., Huang, C.C., Hsiao, Y.H., Ma, C.C.M., Chang,
M.C., Shao, H., Tsai, M.C., Hsieh, C.K., Tsai, C.H., Weng, F.B.: Nanotechnology 19, 1 (2008)
26. Kongkanand, A., MartinezDominguez, R., Kamat, P.V.: Nano Lett. 7, 676 (2007)
27. Wang, X., Zhi, L.J., Mullen, K.: Nano Lett. 8, 323 (2008)
28. Peter, L.M., Wijayantha, K.G.U.: Electrochim. Acta 45, 4543 (2000)
29. Law, M., Greene, L.E., Johnson, J.C., Saykally, R., Yang, P.: Nat. Mater. 4, 455 (2005)
30. Liu, C.-J., Burghaus, U., Besenbacher, F., Wang, Z.L.: ACS Nano 4, 5517 (2010)
31. Oekermann, T., Zhang, D., Yoshida, T., Minoura, H.: J. Phys. Chem. B 108, 2227 (2004)
32. Archana, P.S., Jose, R., Vijila, C., Ramakrishna, S.: J. Phys. Chem. C 113, 21538 (2009)
33. Berger, C.: Science 312, 1191 (2006)
34. Skotheim, T.A., Elsenbaumer, R.L., Reynolds, J.R.: Handbook of Conducting Polymers.
Marcel Dekker, New York (1997); [b] Premamoy, G., Samir, K.S., Amit, C.: Eur. Polym. J.
35, 699 (1999)
35. Wu, T.M., Lin, Y.W., Liao, C.S.: Carbon 43, 734–740 (2005); [b] Wu, T.M., Lin, Y.W.:
Polymer 47, 3576 (2006)
36. Zengin, H., Zhou, W.S., Jin, J.Y., Czerw, R., Smith, D.W., Echegoyen, L., Carroll, D.L.,
Foulger, S.H., Ballato, J.: Adv. Mater. 14, 1480 (2002)
37. Wan, M.X.: In: Li, Q. (ed.), Conducting Polymers with Micro or Nanometer Structure.
Tsinghua University Press, Beijing and Springer, Berlin, Heidelberg (2008)
38. Wan, M.X.: Macromol. Rapid Commun. 30, 963–975 (2009)
39. Novoselov, K.S., Jiang, Z., Zhang, Y., Morozov, S.V., Stormer, H.L., Zeitler, U., Maan, J.C.,
Boebinger, G.S., Kim, P., Geim, A.K.: Science 315, 1379 (2007). [b] Bunch, J.S., van der
Zande, A.M., Verbridge, S.S., Frank, I.W., Tanenbaum, D.M., Parpia, J.M., Craighead, H.G.,
McEuen, P.L.: Science 315, 490 (2007). [c] Li, D., Muller, M.B., Gilje, S., Kaner, R.B.,
Wallace, G.G.: Nat. Nano. 3 101 (2008). [d] Gilje, S., Han, S., Wang, M.S., Wang, K.L., Kaner,
R.B.: Nano Lett. 7, 3394 (2007)
41. Moore, V.C., Strano, M.S., Haroz, E.H., Hauge, R.H., Smalley, R.E., Schmidt, J., Talmon, Y.:
Nano Lett. 3, 1379 (2003)
42. Stankovich, S., Piner, R.D., Chen, X., Wu, N., Nguyen, S.T., Ruoff, R.S.: J. Mater. Chem. 16,
155 (2006)
43. Bai H, Xu YX, Zhao L, Li C, Shi GQ, Chem. Commun. 2009, 1667. [b] Niyogi S, Bekyarova
E, Itkis ME, McWilliams JL, Hamon MA, Haddon RC, J. Am. Chem. Soc. 2006, 128, 7720.
[c] Xu YX, Bai H, Lu GW, Li C, Shi GQ, J. Am. Chem. Soc. 2008, 130, 5856
References 79
44. Wang, X., Zhi, L., Müllen, K.: Nano Lett. 8, 323 (2008). [b] Liu, Z.F., Liu, Q., Huang, Y., Ma,
Y.F., Yin, S.G., Zhang, X.Y., Sun, W., Chen, Y.S.: Adv. Mater. 20, 3924 (2008). [c] Liu, Q.,
Liu, Z.F., Zhang, X.Y., Yang, L.Y., Zhang, N., Pan, G.L., Yin, S.G., Chen, Y.S., Wei, J.: Adv.
Funct. Mater. 19, 894 (2009). [d] Nethravathi, C., Rajamathi, M.: Carbon 46, 1994 (2008)
45. Bissessur, R., Liu, P.K.Y., White, W., Scully, S.F.: Langmuir 22, 1729 (2006). [b] Matsuo, Y.,
Higashika, S., Kimura, K., Miyamoto, Y., Fukutsuka, T., Sugie, Y.: J. Mater. Chem. 12, 1592
(2002). [c] Wang, H.L., Hao, Q.L., Yang, X.J., Lua, L., Wang, X.: Electrochem. Commun.
11, 1158 (2009)
46. Cassagneau, T., Fendler, J.H., Johnson, S.A,, Mallouk. T.E.: Adv. Mater. 12, 1363 (2000). [b]
Cassagneau, T., Guerin, F., Fendler, J.H.: Langmuir 16, 7318 (2000)
47. Huang, W.S., Humphrey, B.D., MacDiarmid, A.G.: J. Chem. Soc. Faraday Trans. 82, 2385
(1986)
48. Chiang, J.C., MacDiarmid, A.G.: Synth. Met. 13, 193 (1986). [b] MacDiamid, A.G., Chiang,
J.C., Richter, A.F., Epstein, A.J. Synth.Met. 18, 285 (1987)
49. Wan, M.X.: J. Polym. Sci. Part A 30, 543 (1992)
50. Gu, H., Su, X., Loh, K.P.: J. Phys. Chem. B 109, 13611 (2005)
Chapter 4
Enhanced Light Harvesting in Plasmonic
Dye-Sensitized Solar Cells Using Gold
Topological Light Trapping Layer
4.1 Introduction
Dye-sensitized solar cells (DSSCs) have drawn much attention as its promising low
cost and high efficiency for the photoelectrical conversion since Grätzel’s group
introduced nanocrystalline TiO2 into anode electrode [1]. Although over 11 %
conversion efficiency was achieved [2, 3], further improvement is still necessary. In
this multi-component system, light collection efficiency is important because it
directly related to the charge separation and transport [4], then the photocurrent and
photoelectrical conversion efficiency was affected. The light harvesting efficiency is
determined by the amount of dye attached, the extinction coefficient of the sensitizers
and the optical path length in the nanocrystalline film [5]. Many works have been
done in enhancing the light collection, such as coupling the photocrystal layer [6],
introducing light scattering particles [7, 8] or voids [9], incorporating large solid
particles [10, 11], introducing hierarchical porous or branched inner channels [5,
12], using core-shell metal-insulator nanoparticles [13] and well-controlled TiO2
nanostructures [14, 15].
However, a primary factor limiting the efficiency of the cells is the competi‐
tion between the optical path length for light absorption and charge diffusion [16].
That is, increasing the film thickness can enhance the dye absorption and light
harvesting undoubtedly, but only 10 μm thickness of the photoanode can bring the
best cell performance [17], because the excessive enhancing the film thickness
will increase the electron transport length and the photo carrier recombination
rates.
Surface plasmon resonance (SPR) of gold is widely used in the biochemistry
and physical fields [18–22], shows the ability to localize the irradiate light and
prolong the optical path length, which is of interest for applications in DSSCs
(Fig. 4.1). Here, we introduce a two dimensional gold topological ordered light
reflection layer with monodisperse polystyrene sphere as template, which will

82 4 Enhanced Light Harvesting in Plasmonic Dye-Sensitized …
(a) (b)
ITO TiO2 Au Electrolyte Electrode ITO Au TiO2 Electrolyte Electrode
(c)
Fig. 4.1 The Au-TiO2 composite system. The introduced Au can bring a bend of the band edge,
exciting the hot electrons and giving a current. The anode or cathodic photocurrent will be generated
from (a) or (b) system [23]. Copyright © 2009 WILEY-VCH Verlag GmbH & Co. KGaA,
Weinheim. c Is the Scheme of SPR effect. The incident light would bring a resonance and change
the transport direction when reach the interface of noble metal [24]. Copyright © 2010, Rights
Managed by Nature Publishing Group
improve the light harvesting efficiency significantly based on the plasmonic effect
while not affecting the electron transport length in DSSCs.
In our previous work, micro/nanocomposite anodes were fabricated by electro‐
hydrodynamic (EHD) method, showing a better light harvesting and conversion
efficiency [4, 5]. Taking this advantage, the TiO2 hierarchical photoanode was
prepared by EHD method so that to keep the large light harvesting efficiency in this
work (Scheme 4.1d). The structure of Au-modified DSSC is shown in Scheme 4.1a.
The light transmits through the photoanode with micro/nanocomposite structures
and reaches to the counter electrode. With the topological ordered gold layer, the
transmission light is omnidirectional reflected back to the photoanode again, and the
light utilization is increased (Scheme 4.1b). While for the common counter electrode
(Scheme 4.1c), the transparent FTO will reflect only a little amount of the light, most
of the solar energy is transmitted out of the system. Hence, due to the omnidirectional
light collection, more electrons can be excited and the photocurrent and photoelec‐
trical conversion efficiency is increased.
Scheme 4.1 The operational scheme of the Au topological layer: a is the structure of the DSSC.
The solar irradiates from the photoanode, and transmits through the TiO2 layer to the counter
electrode. b When there is an Au topological layer coated the counter electrode, the transmission
light is reflected back to the photoanode, hence more light is used and more electrons are excited.
While for the common electrode (c) most of the transmission light will go through the transparent
FTO, the light utilization is limited. d Is SEM image of the photoanode fabricated by the EHD
technique. This hierarchical structured can enhance the light harvesting reported in the previous
work
4.2.1 Characterization of the Topological Ordered Au Film
The polymer template film was prepared by means of vertical deposition on clean
FTO at first [25, 26]. The clean FTO was vertically positioned in a vial containing
the P(St-MMA-AA) colloidal suspensions with concentrations of 0.2 wt% at a
temperature of 60 °C and a humidity of 60 % for ca. 48 h [26]. The surface
morphology of the film was examined using scanning electron microscopy (SEM)
and given in Fig. 4.2b. It can be seen clearly that the polymer sphere has a uniform
diameter of ~100 nm, and a perfect hexagonally ordered and long range well-ordered
structure was formed after packing. Figure 4.2a shows the surface of the electrode
after the vacuum evaporation of gold. The surface looks rougher than the PS sphere
which means that the gold is coated on the PS surface. Figure 4.2c shows the side
view of the gold coating counter electrode (named as Electrode-Au). The gold layer
is deposited on the PS sphere and stacked tightly. Figure 4.2d shows the tilt view of
the inside surface of gold layer after annealing. The convex–concave structure can
be observed in this image, which proves the formation of gold topological light
trapping layer in DSSC device. As shown, there is a bowl-like structure on the PS-
Au interface, indicating that this layer will reflect and scatter the light efficiently with
Fig. 4.2 The morphology of electrodes, a is the surface after gold evaporation and c is the side
view of the electrode. The interface between PS and gold is clearly observed, and the gold layer is
about 230 nm thick, b is the SEM image of the surface of FTO with a perfect hexagonally ordered
and long range well-ordered structure after positioned in a ~100 nm polymer sphere colloidal
suspensions, d is the tilt view of inside surface of the gold layer after annealing
a low transmission. To compare with the gold coated electrode, a common electrode
without gold coating was treated with the same method at the same time, which
named as Electrode-FTO in this paper.
4.2.2 Photovoltaic Behavior of DSSCs
The photovoltaic performance of different electrodes was measured at condition

chosen to simulate sunlight (AM1.5, 100 mW/cm2). Table 4.1 summarizes the
parameters of ~10 samples used in our experiment. And a current density versus
voltage (I–V) curve on the average value is exhibited in Fig. 4.3a. A short-circuit
photocurrent density (Jsc) of 14.6 mA/cm2, an open-circuit voltage (Voc) of 0.69 V,
and the conversion efficiency (η) of 6.64 % for Electrode-Au were obtained. As
compared, a Jsc, Voc, and η were 12.6 mA/cm2, 0.69 V, and 5.61 % respectively for
Electrode-FTO. There is about 16 % increase in Jsc and 18 % in η for Electrode-Au
Table 4.1 Parameters for solar cells based on the different electrodesa
Device Jsc (mA/cm2) Voc (V) ff (%) η (%)
Electrode-FTO 12.6 ± 0.3 0.69 ± 0.02 65 ± 2 5.6 ± 0.4
Electrode-Au 14.6 ± 0.4 0.69 ± 0.01 64 ± 2 6.6 ± 0.4
a
Photovoltage parameters of different electrodes. The sensitizer was N3 (ruthenium dye). The cell
active area was 0.20 cm2, and the light intensity was 100 mW/cm2. The values show here is the cell
with an average performance based on about ten samples
Fig. 4.3 a Photocurrent voltage characteristics of different electrodes. The Electrode-Au can
increase photocurrent notably, and the efficiency is improved. b–c The UV-vis spectrum of different
electrodes. The diffuse reflectance spectra (b) shows the Electrode-Au has a strong reflectance than
Electrode-FTO in the entire wavelength, and the light absorption is enhanced by the gold layer (c)
and nearly the same Voc compared with Electrode-FTO. This improvement can be
attributed to more light collection and more electrons inducing because of the gold
topological structure, which can be proved by the optical measurement.
4.2.3 Light Harvesting Properties of Two Electrodes
The light utilization is dependent on many factors. By introducing gold topological

structure into the counter electrode, the optical length is extended and the light
collection efficiency is enhanced. Figure 4.3b shows the reflectance spectra of
different counter electrodes. Compare with Electrode-FTO, the topological gold
structure brings a 77.6 % reflectance increased at entire wavelength. This strong
reflectance can make the transmission light reused in the photoanode, and more light
absorption can improve the cell performance [27]. Especially, there is a strong
reflectance around 600 nm and a stronger enhancement over 600 nm, which is due
to reflection of the localized plasmon resonance. The absorption spectra of the sealed
solar cells were shown in Fig. 4.3b. The Electrode-FTO sealed cell shows typical
spectra of the dye, N3 [28]. While the introduced topological layer enhanced the
light absorption by 33.3 %. Notably, there is an addition absorption peak around
620 nm and the wavelength range over 600 nm shows a much stronger absorption
increase, which can be attributed to the plasmonic resonance absorption by Au [29].
Hence more solar energy is restricted in the solar cells, so as to increase the light
intensity and result a better photocurrent generation.

Efficiency (IPCE) Performance of Different Electrodes
In addition, the improved light harvesting is directly reflected in the increased IPCE
performance (Fig. 4.4). Comparing with Electrode-FTO, the IPCE response at all
wavelengths is enhanced by a factor of 1.25, while the peak value of IPCE
(IPCEwavelength) was increased by 19 % for IPCE436 with Electrode-Au. This improve‐
ment can be attributed to the enhanced light harvesting, which is observed in the
UV-vis measurement. Some of the incident light transmits through the photoanode,
and the plasmon resonance and reflectance is taken place at the gold layer, then the
transmission light is reflecting back to the photoanode and reused by dye to exciting
more electrons, hence, the photocurrent is increased. Notably, this introduced gold
layer increased the photocurrent both in the wavelength range lower than 550 nm
(19 % increase), and the longer wavelengths (over 550 nm, 37 % increase).
Comparing with the UV-vis absorption spectra in Fig. 4.3c, there is 40 % increase
of the absorption for the Electrode-Au sealed cell in the wavelength range between
550 and 800 nm. However, the 40 % improvement of light absorption brings 38 %
increase of IPCE, which means that the reflected energy improvement does not
transport to the photocurrent totally because of the uncompleted dye absorption
efficiency in this range [28], the recombination and resistance in the photoanode.
Moreover, the IPCE shows a larger improvement at the longer wavelength range,
because the IPCE in the short wavelength range is mainly corresponding to the strong
absorption of dye and the gold layer performs better at the longer wavelength.
Fig. 4.4 The IPCE measurement of different electrodes. The Electrode-Au shows a higher Photo-
to-Current Conversion Efficiency. IPCE (%) = (1241Jsc)/(λPin) × 100, where Jsc is short-circuit
current and Pin is the power of the incident light
Fig. 4.5 The photovoltaic behavior under different light intensity. The photocurrent density is
increased direct proportion to the light intensity, but the efficiency is reduced (a) due to the decrease
of fill factor (b)
4.2.5 Photovoltaic Behavior Under Different Light Intensity
Furthermore, the Jsc−Pin and η-Pin relationship can exhibit the effect of light clearly,
and the curves are shown in Fig. 4.5a. It can be seen that short-circuit photocurrent
density of these two electrodes is linear to the light intensity. The slope is 0.123 mA/mW
for Electrode-Au and 0.134 mA/mW for Electrode-FTO. This direct proportion rela‐
tionship between Jsc and incident light intensity indicated that neither electron diffu‐
sion in TiO2 film nor ion transport in electrolyte limit the cell performance, and the
electricity process in our samples is only controlled by the step of photon absorption
and electron injection [30]. Therefore, the incorporation of gold layer had a negligible
influence on the nature of the charge transport process in the device [31]. This means
increasing optical path length with our method will improve the cell performance
directly.
Meanwhile, the efficiency was decreased with the light intensity increasing, but
the Electrode-Au always shows a higher efficiency than Electrode-FTO, especially
in a strong light intensity. This means the gold topological structure can enhance the
cell performance even at a wide solar irradiation, which will widen the cell’s appli‐
cation. When the incident light intensity increased to about 150 mW/cm2, the η was
decreased to 6.1 and 5.1 % for Electrode-Au and Electrode-FTO, respectively. This
decrease of η is attributed to the reduction of fill factor, which is decreased from 74.8
to 61.5 % for Electrode-Au and 75.8 to 63.9 % for Electrode-FTO (Fig. 4.5b). The
fill factor of Electrode-Au is always lower than the reference, that is also because
the gold layers increase the light reflection, which enhance the light intensity in other
words.
4.3 Conclusion
In summary, a gold topological ordered layer was introduced into the DSSC as a
counter electrode. Accompany with the plasmon resonance of gold, this structure
can enhance the light scattering and reflection and improve the cell performance.
Using other improvement in photoanode, this method can keep increasing the light
collection. Also, without enhancing the thickness of photoanode, the optical path
length is increased a lot. The efficiency and photocurrent was increased by 18 and
16 %, respectively, and the IPCE was enhanced by 25 % at all wavelengths. Besides,
this counter electrode performs well at both strong and low solar irradiation. This
idea can also improve other solar cells performance by increasing the light utilization
using plasmon resonance.
The synthesis of Poly(St-MMA-AA): The latex particles were synthesized by one-

step emulsion polymerization based on our previous method. In details, styrene
(9.5 g), methyl methacrylate (0.5 g), and acrylic acid (0.5 g) were dispersed in 50 mL
of aqueous solution dissolving of emulsifier of buffer agent of NH4HCO3 (0.5 g) and
sodium dodecylbenzenesulfonate (NaDBS) (9–15 mg). After the system was heated
at 70 °C, the initiator (NH4)2S2O4 (APS) (0.5 g) was charged into above system, and
the polymerization reaction began. Followed by the polymerization running at 70 °C
for 1.5 h, the system’s temperature was raised to 80 °C, The reaction was finished
after another 3 h.
Preparation of the counter electrode: The template was fabricated on a hydro‐

philic fluorine doped tin oxide (FTO) substrate by a vertical deposition method using
colloidal suspensions of poly(St-MMA-AA) of 0.2 wt% at a constant temperature of
60 °C and a relative humidity of 60 % for about 48 h. Gold was vacuum evaporation
on the back side of FTO at 5 mA, then the substrate was calcined at 450 °C for 1 h
to remove the polymer. After these, the Pt was deposited on the conductive side at
2 mA for 6 min. To compare with the Electrode-Au, Electrode-FTO was operated
with the same method only without gold evaporation.
Preparation of photoanode: To prepare the photoanode, 3.00 g of poly(vinyl‐
alcohol) (PVA, MW = 20,000 ~ 30,000) water solution (30 %) was dissolved with
the mixed solvent containing 9.0 mL of H2O and 12.5 mL of ethanol. Then, 1.62 g
of P-25 TiO2 was added to make the suspension of TiO2. Then suspension was
dispersed in ultrasonic cleaner. FTO glass substrate was spin coating by a solution
containing 8.5 mL titanium(IV) n-butoxide, Ti(O(CH2)3CH3)4, 0.9 mL deionized
water, 1.6 mL acetylacetone and 30 mL ethanol. After keeping the thin layer in air
for two hours, it could be used in the EHD procedure. In an EHD technique, 1.5 mL
suspension of TiO2 was placed into a 5 mL syringe equipped with a needle made of
stainless steel. The distance between the needle tip and collector was 12 cm, and the
voltage was set at 30 kV. Once the photoanode dried, it was calcined at 450 C for
1 h. The calcined film was put into a 0.2 M TiCl4 aqueous solution for further depo‐
sition of TiO2 to improve the performance. After being left overnight, the treated
electrode was washed with distilled water and fired again at 450 °C for another 1 h.
(The film is about 7 μm thick) Then the prepared film was sensitized by immersing
it into a solution containing 3 × 10−4 M Ru(dcbpy)2(NCS)2 (N3) (dcbpy = 2,2′-
bipyridine-4,4′-dicarboxylic acid) for 24 h.
Measurements: SEM images were obtained using a JEOL JSM-6700F scanning
electron microscope at 3.0 kV. The I–V characteristics of the cell were measured by
an electrochemical analyser (CHI660A, Chenhua Instruments Co., Shanghai) under
solar simulator illumination (CMH-250, Aodite Photoelectronic Technology Ltd.,
Beijing) at room temperature with an electrolyte composing of 0.5 M LiI, 0.05 M I2,
0.5 M tertbutylpyridine, and 0.6 M 1-propyl-3-methylimidazolium iodide in
3-methoxypropionitrile. UV-vis spectra were recorded on a Hitachi Model U-4100
spectrophotometer. The IPCE was measured by illumination with monochromatic
light, which was obtained by a series of light filters with different wavelengths.
4.5 Postscript
Besides this work, we also considered testing the photo-electric performance by

tuning the plasmonic response. Because the SPR effect is quite related to the aggre‐
gate numbers of nanoparticles. By referring Chen’s work [32], we synthesized gold
assemblies with different nanoparticle numbers, and then we introduce it to the
photoanode, aiming to see if the different SPR response will affect the efficiency of
DSSC. As a primary result, we found at a certain assembly degree, the efficiency is
Fig. 4.6 a–c The assembly of Au with different amount of NaCl addition (a: 0 μL; b: 4 μL; c:
8 μL). d is the UV-vis absorption spectrum of different aggregation. A new peak at the longer
wavelength appeared, accompany with the Au assembly. e–f The IPCE of different Au aggregation
increased to some extent, but the difference is not obvious, further work is necessary
(Fig. 4.6). As SPR effect has been reported for tens of years, whether or not we can
apply it into DSSC is still an open question. If we can use a theoretical analysis to
guide the experimental design, it will bring a much brighter future.
References
1. O’Regan, B., Grätzel, M.: Nature 353, 737 (1991)

2. Nazeeruddin, M.K., De Angelis, F., Fantacci, S., Selloni, A., Viscardi, G., Liska, P., Ito, S.,
Takeru, B., Grätzel, M.: J. Am. Chem. Soc. 127, 16835 (2005)
3. Chiba, Y., Islam, A., Watanabe, Y., Komiya, R., Koide, N., Han, L.: Jpn. J. Appl. Phys. 45,
L638 (2006)
4. Zhao, Y., Sheng, X., Zhai, J., Jiang, L., Yang, C., Sun, Z., Li, Y., Zhu, D.: ChemPhysChem
8, 856 (2007)
5. Zhao, Y., Zhai, J., Wei, T., Jiang, L., Zhu, D.: J. Mater. Chem. 17, 5084 (2007)
6. Nishimura, S., Abrams, N., Lewis, B.A., Halaoui, L.I., Mallouk, T.E., Benkstein, K.D., van de
Lagemaat, J., Frank, A.J.: J. Am. Chem. Soc. 125, 6306 (2003)
7. Tebby, Z., Babot, O., Michau, D., Hirsch, L., Carlos, L., Toupance, T.: J. Photochem. Photobiol.
A-Chem. 205, 70 (2009)
8. Tebby, Z., Uddin, T., Nicolas, Y., Olivier, C. l., Toupance, T., Labrugère, C., Hirsch, L.: ACS
Appl. Mater. Interfaces 3, 1485 (2011)
9. Hore, S., Nitz, P., Vetter, C., Prahl, C., Niggemann, M., Kern, R.: Chem. Commun. 2011 (2005)
10. Nusbaumer, H., Zakeeruddin, S.M., Moser, J.E., Grätzel, M.: Chem. Eur. J. 9, 3756 (2003)
11. Vargas, W.E., Niklasson, G.A.: Sol. Energy Mater. Sol. Cells 69, 147 (2001)
12. Zhao, Y., Zhai, J., Tan, S., Wang, L., Jiang, L., Zhu, D.: Nanotechnology 17, 2090 (2006)
References 91
13. Brown, M.D., Suteewong, T., Kumar, R.S.S., D’Innocenzo, V., Petrozza, A., Lee, M.M.,
Wiesner, U., Snaith, H.J.: Nano Lett. 11, 438 (2010)
14. Crossland, E.J.W., Kamperman, M., Nedelcu, M., Ducati, C., Wiesner, U., Smilgies, D.M.,
Toombes, G.E.S., Hillmyer, M.A., Ludwigs, S., Steiner, U., Snaith, H.J.: Nano Lett. 9, 2807
(2008)
15. Crossland, E.J.W., Nedelcu, M., Ducati, C., Ludwigs, S., Hillmyer, M.A., Steiner, U., Snaith,
H.J.: Nano Lett. 9, 2813 (2008)
16. Liu, C.-J., Burghaus, U., Besenbacher, F., Wang, Z.: ACS Nano 4, 5517 (2010)
17. Law, M., Greene, L.E., Johnson, J.C., Saykally, R., Yang, P.: Nat. Mater. 4, 455 (2005)
18. Homola, J.: Anal. Bioanal. Chem. 377, 528 (2003)
19. Qi, Z.-M., Honma, I., Zhou, H.: Opt. Lett. 2006, 31 (1854)
20. El-Sayed, I.H., Huang, X., El-Sayed, M.A.: Nano Lett. 5, 829 (2005)
21. Schatz, G.C.: Acc. Chem. Res. 17, 370 (1984)
22. Szamocki, R., Reculusa, S., Ravaine, S., Bartlett, P.N., Kuhn, A., Hempelmann, R.: Angew.
Chem. Int. Ed. 45, 1317 (2006)
23. Sakai, N., Fujiwara, Y., Takahashi, Y., Tatsuma, T.: ChemPhysChem 10, 766 (2009)
24. Atwater, H.A., Polman, A.: Nat. Mater. 9, 865 (2010)
25. Wang, J., Wen, Y., Hu, J., Song, Y., Jiang, L.: Adv. Funct. Mater. 17, 219 (2007)
26. Li, H., Chang, L., Wang, J., Yang, L., Song, Y.: J. Mater. Chem. 18, 5098 (2008)
28. Nazeeruddin, M.K., Kay, A., Rodicio, I., Humphry-Baker, R., Mueller, E., Liska, P.,
Vlachopoulos, N., Grätzel, M.: J. Am. Chem. Soc. 115, 6382 (1993)
29. Sakai, N., Tatsuma, T.: Adv. Mater. 22, 3185 (2010)
30. Guo, X.-Z., Luo, Y.-H., Li, C.-H., Qin, D., Li, D.-M., Meng, Q.-B.: Curr. Appl. Phys. 12, e54
(2012)
31. Wu, J.-L., Chen, F.-C., Hsiao, Y.-S., Chien, F.-C., Chen, P., Kuo, C.-H., Huang, M.H., Hsu,
C.-S.: ACS Nano 5, 959 (2011)
32. Chen, G., Wang, Y., Yang, M., Xu, J., Goh, S.J., Pan, M., Chen, H.: J. Am. Chem. Soc. 132,
3644 (2010)
Chapter 5
Photocatalytic Properties of Graphdiyne
and Graphene Modified TiO2: From Theory
to Experiment
5.1 Introduction
Heterogeneous photocatalytic oxidation of organic compounds by titanium dioxide

(TiO2) has received much attention as a promising technology for pollution reme‐
diation due to strong UV light absorption, strong oxidation property, and excellent
photostability of TiO2. The efficiency of the photocatalytic process is measured by
quantum yield (ϕ), which is defined as the number of events occurring per photon
absorbed [1]. In an ideal system, ϕ is proportional to a simple relationship:
/( )
𝜙 ∝ kCT kCT + kR
kCT is the rate of the charge transfer processes and kR indicates the electron-hole
recombination rate (both bulk and surface). The above equation reveals that
preventing the electron-hole recombination of TiO2 under illumination would be
critical for improvement of the quantum yield, and a plausible way is to combine the
semiconductor with some electron acceptors [2]. Among them, graphene (GR) has
been confirmed to be a good candidate to form the composites with TiO2 as the
photocatalyst [3].
Since isolated in 2004 [4], graphene had captured the attention of scientists,
because it is one of the thinnest and strongest materials in the world. Furthermore, GR
is highly transparent, possesses high electrical and thermal conductivity, and has a
large-specific surface area [5–7]. Owning to these superior properties, GR, and its
composites have been extensively investigated for variety of applications including
supercapacitors [8–10], fuel cells [11], batteries [12], catalysis [13, 14], photovoltaics
[15, 16], chemical and biosensors [17–21], photonics [22], and optoelectronics [23].
Graphdiyne (GD), another 2D carbon allotrope that contains both sp and sp2 carbon
atoms was predicted by Baughman in 1997 [24]. Li and coworkers successfully
synthesized large area graphdiyne films on the surface of copper via a cross-coupling

94 5 Photocatalytic Properties of Graphdiyne and Graphene …
reaction using hexaethynylbenzene precursors in 2010 [25]. The resultant GD showed

semiconducting property with a measured conductivity of 2.516 × 10−4 S/m, and was
predicted to be the most stable structure amongst various diacetylenic nonnatural
carbon allotropes because of the dialkyne between the benzene rings [26]. Also Malko
predicted that some GD allotropes, which do not have hexagonal symmetry and two
self-doped nonequivalent distorted Dirac cones, could possess electronic properties
superior than that of graphene [27]. However, these reports are purely based on the
theoretical prediction without sufficient experimental supports.
Density functional theory (DFT) calculations predicted that graphdiyne is a
narrow direct band gap material. Remarkably, the calculated intrinsic charge mobi‐
lity of graphdiyne sheet is up to 105 cm2 V−1 s−1 at room temperature in some allo‐
tropes [28], which is similar to the value of graphene [29, 30]. Such a high change
mobility suggests that GD should have the excellent electron transport property,
which may be utilized in the composites to improve the photocatalytic performance
of TiO2. This hypothesis was partially confirmed by our previous work that the P25-
GD nanocomposite gave a good catalysis performance [31]. Nevertheless, under‐
standing the charge transfer mechanism at the interface of TiO2 and graphdiyne was
found to be difficult due to the complicate phases and facets of the loaded P25
samples.
In this manuscript, DFT calculation is employed to investigate charge transfer
actions between 2D carbon materials and anatase TiO2 of different crystal facets.
The surface free energies of anatase TiO2 crystals with different facets can vary
significantly [32, 33] because of the distinctive surface atomic arrangements. In
general, {001} and {110} facets of anatase TiO2 have high surface energy [34, 35]
compared with the most commonly obtained {101} facets [36]. Herein, we chose
the aforementioned three anatase TiO2 facets to composite with GD and GR,
respectively (Scheme 5.1). For DFT calculation, the supercells of six composites
were built and geometrically optimized. The results of calculation and experi‐
ment show that the TiO2(001)–GD has the most effective charge separation
capacity, the most abundant impurity levels and the highest oxidation ability,
exhibiting an oxidative degradation rate constant of 1.63 times of pure TiO2(001)
and 1.27 times of TiO2(001)–GR composites.
5.2.1 Theoretical Analysis and Structural Characterizations
The density functional theory (DFT) employing a Dmol3 code was used to calculate
the electronic structure of different types of 2D TiO2–carbon composite materials.
The supercells of six composites (defined as TiO2(001)–GD, TiO2(101)–GD,
TiO2(110)–GD, TiO2(001)–GR, TiO2(101)–GR, and TiO2(110)–GR) were built, and
the optimized structures were shown in Fig. 5.1. For TiO2(101)–GR and TiO2(110)–
GR composites, the equilibrium distances between graphene and TiO2 are so large
Scheme 5.1 Scheme of

TiO2–GD (top) or TiO2–GR
(bottom) nanosheets stacking
materials
that even the nearest C atoms were far as 3.185 and 3.34 Å to the TiO2 planes,
indicating the weak interaction between GR and TiO2 surface. For TiO2(101)–GD,
TiO2(110)–GD, and TiO2(001)–GR composites, thanks to the strong adsorption
between carbon atoms and the TiO2 support [37], numbers of new C–O bonds are
formed from O in TiO2 and atop C in GD or GR. Especially for TiO2(001)–GD
composite, beside C–O bonds, the new Ti–C bonds are also obtained. Those new
bonds demonstrated the chemisorbed situation between TiO2 and GD or GR [38]. In
case of GR, only chemical bond formation with {001} facets could occur; However,
more active GD could be chemisorbed with every facets of TiO2, giving rise to strong
electron hybridization. The calculated results also indicate that the shortest C–O
bond (ca. 1.354–1.433 Å) is formed in TiO2(001)–GD system among the entire
composites, suggesting the strongest C–O bond formation comparing with other five
composites (Table 5.1).
In detail, the formation of C–O–Ti and C–Ti bonds can be observed by the
electron density differences (Fig. 5.2a, b). For the TiO2(001)–GR composite,
C–O σ bond is formed by O and atop C atoms, which plays a bridge role in the
charge transfer at the interface between TiO2 and GR. Moreover, for the
TiO2(001)–GD composite, besides the C–O bond, another Ti–C π bond is also
formed (top right pictures in Fig. 5.2a). This particular bond would make TiO2
anchor to the GD layer tightly, which is beneficial for the charge transfer. The
bonding formation can also be observed from the experimental results of X-ray
Table 5.1 Supercell

a (Å) b (Å) c (Å) Mismatch (%)
parameters and mismatches of
the composites TiO2(101)–GD 9.85 16.71 25 3.82
TiO2(110)–GD 9.49 16.22 25 0
TiO2(001)–GD 18.93 15.77 25 4.02
TiO2(101)–GR 10.02 21.98 25 3.17
TiO2(110)–GR 10.36 10.36 25 5.31
TiO2(001)–GR 21.98 11.80 25 4.04
Fig. 5.1 Structure of TiO2–GD and TiO2–GR composites
photoelectron spectroscopy (XPS) and Fourier transform infrared (FTIR) spectra

in the following discussion (Fig. 5.3d, e).
To further demonstrate the charge transfer ability, we calculated the Mulliken
charge for the surface of TiO2(001)–GD and TiO2(001)–GR (Fig. 5.2c). It can be
seen that the GD or GR surface has a positive Mulliken charge, forming large oppo‐
site interface dipole at the interface [39, 40], leading to a strong built-in electric field
throughout the superlattice. In this case, with light excitation, the tendency of
Fig. 5.2 Plots of electron density difference at the composites interfaces: a TiO2(001)–GD, b
TiO2(001)–GR. Isosurface values for electron density difference plots are −0.08 and 0.08 e/Å3,
where the accumulation and depletion of electrons are represented in blue and yellow, respectively.
c The Mulliken charge of GD or GR surface in the composites
electrons transport to carbon would be high, prolonging the holes’ lifetime at the
valence band of TiO2. This feature suppresses the charge recombination, and thus
improves the photocatalytic activity. Since the magnitude of charge accumulation
in GD is larger than GR surface, stronger electrons capture ability of TiO2(001)–GD
composites could be expected compared to TiO2(001)–GR composites.
Because the {001} facets possess the strongest bond formation with carbon mate‐
rials, we chose TiO2 with exposed {001} facets in the following experiments to verify
the calculation results. In a typical experiment, the anatase TiO2 nanosheets with
exposed {001} facets were synthesized by hydrothermal method according to Lou’s
report [41]. Scanning electron microscopy (SEM) images confirm that the sample
contains very large carpet-like structures with a side length of several micrometers
and a thickness of ca. 30 nm (Fig. 5.4a). These 2D structures are formed via the
lateral aggregation of smaller TiO2 nanosheets. The transmission electron micro‐
scopy (TEM) images further confirm the morphology characteristics of nanosheets,
and the highly symmetries selected area electron diffraction (SAED) pattern can be
indexed as [001] zone (Fig. 5.4b, d) [42]. In addition, the high-resolution TEM (HR-
TEM) image clearly shows the continuous (100) atomic planes with a lattice spacing
of ca. 3.7 Å, corresponding to the {100} planes of anatase TiO2 single crystals
(Fig. 5.4c) [43]. The X-ray diffraction (XRD) patterns of the products are presented
in Fig. 5.3c. All the identified peaks can be unambiguously assigned to anatase
TiO2 (JCPDS card no. 21-1272; S. G. I41/amd; a0 = 3.7852 Å, c0 = 9.5139 Å). The
Raman spectrum also shows the typical peaks of anatase phase without any rutile
phase, and the typical vibration of GD and GR can be also distinguished from their
composites (Fig. 5.5). In detail, the Raman spectra of TiO2(001), TiO2(001)–GR,
and TiO2(001)–GD. The TiO2(001) all shows the distinct Raman bands at 143, 196,
395, 516, and 637 cm−1. These bands can be attributed to the Raman-active modes
of the anatase phase TiO2 with the symmetries of Eg, Eg, B1g, A1g, and Eg,
Fig. 5.3 The HR-TEM images of a TiO2(001)–GD and b TiO2(001)–GR composites. c XRD
pattern of different materials. d C1s XPS spectra of TiO2(001)–GD and TiO2(001)–GR. e is FTIR
spectroscopy of TiO2(001), TiO2(001)–GD, and TiO2(001)–GR
Fig. 5.4 Characterization of the anatase TiO2 nanosheets with exposed (001) facets. SEM (a), TEM
(b), and HR-TEM (c) images of the nanosheets, and d is the SAED pattern of (001) TiO2
respectively. It is obvious that there are no Raman bands of rutile phase TiO2
(Fig. 5.5a) are observed, indicating all the TiO2 obtained were in the same crystal
phase, anatase. The Raman peaks of TiO2(001)–GD shown at 1356 and 1593 cm−1
can be assigned to the vibration of aromatic rings of the GD substrates. The Raman
spectrum of TiO2(001)–GR shows the presence of D, G, and 2D bands at 1350, 1604,
and 2700 cm−1, respectively. G band is common to all the sp2 carbon forms and
provides information on the in-plane vibration of sp2 bonded carbon atoms. The D
band suggests the presence of sp3 defects, which indicates the successful chemical
modification of the graphite sheets. Moreover, the weak and broad 2D band also
suggests the existence of disorder. The above results reveal the successfully combi‐
nation of the carbon substrates with TiO2(001) catalysts.
A series of TiO2(001)–GD and TiO2(001)–GR composites with different ratios
of GD or GR to TiO2 are obtained by a simple hydrothermal treatment of the 2D-
TiO2 and 2D-GD or 2D-GR in a mixture of ethanol and water. No typical XRD
diffraction peaks of GR and GD could be observed from the composite samples
(Fig. 5.3c), which is reasonable because the sharp peak of anatase TiO2 shielded the
main characteristic peak of GR and GD [31]. Figure 5.3a, b are the HR-TEM images
of the TiO2(001)–GD and TiO2(001)–GR composites, respectively. The revealed
Fig. 5.5 Raman spectra of a TiO2(001), b TiO2(001)–GR, and c TiO2(001)–GD
lattice fringes are assigned to the TiO2 nanosheets and the amorphous morphology
is attributed to graphdiyne (Fig. 5.3a). Also, the few curving fringes are corresponded
to the few layers’ graphene (Fig. 5.3b). Both Fig. 5.3a, b indicate a perfectly connec‐
tion between TiO2 and 2D carbon sheets. It should be noted that such connections
are not only the physical attachment but also the chemical bonding. The high-
resolution C 1s XPS spectra of the composites (Fig. 5.3d) reveal that the carbon
materials are successfully bound with TiO2. The C 1s XPS spectrum of TiO2(001)–
GR indicates the presence of three types of carbon bonds at 284.9, 286.4, and
288.8 eV that can be assigned to a C 1s orbital of C–C, C–O, and O–C=O, respec‐
tively [44]. Notably, an additional peak at 284.1 eV is discerned from the spectrum
of TiO2(001)–GD, corresponding to the presence of Ti–C bond in the TiO2(001)–
GD sample [45]. Besides, the C–C bond could be fitted to C–C (sp) and C–C (sp2)
located at 285.2 and 284.5 eV, respectively, which is ascribed to the benzene rings
alike structure by diacetylenic linkages in GD [25]. These results are also supported
by the DFT calculations results (Fig. 5.2).
The C–O–Ti bond formation is also explored by the FTIR spectroscopy
(Fig. 5.3e). For both TiO2(001)–GD and TiO2(001)–GR, the broad absorption at
ca. 500 cm−1 represents the Ti–O–Ti stretching vibrations [46], while the new bands
Fig. 5.6 a Photocatalytic degradation of MB over TiO2(001), TiO2(001)–GD, TiO2(001)–GR

composites and the blank experiment (without any photocatalyst). b Plot of the remaining
methylene blue (MB) in solution treated with different samples: Initial and equilibrated with
TiO2(001), TiO2(001)–GD, TiO2(001)–GR composites in the dark after 30 min stirring. These
composites have similar BET surface areas, which are 13.18, 13.49, and 14.21, respectively. The
increase absorption of MB is mainly due to the π–π stacking between 2D carbon sheets and MB
molecules
near 1150 cm−1 are assigned to the Ti–O–C stretching modes [47]. The DFT simu‐
lation also matches the characterization results perfectly (Fig. 5.2a, b).
5.2.2 Performance in Photocatalytic Degradation
The above theoretical calculations and structural characterizations suggest that

TiO2(001)–carbon bonds are of the superior charge transport properties that would
enhance the photocatalytic activity. The photocatalytic degradation of methylene
blue (MB) is used as a model system to validate this prediction (Fig. 5.6a). The
normalized temporal concentration change (C/C0) of MB during the photocatalytic
degradation experiment is found to be proportional to the normalized maximum
absorbance (A/A0), which is derived from the changes in the dye’s absorption profile
(λ = 663 nm) at a given time interval.
It is found that the degradation efficiency is dependent of GD or GR loading in
the composite (Table 5.2). According to a previous report [48], the decomposition
of dye could be assigned to a pseudo-first-order kinetics reaction with a simplified
Langmuir-Hinshelwood model, when C0 is low [49]. That is:
( / )
ln C0 C = kt,
where, k is the apparent first-order rate constant. For the pure TiO2(001), the obtained
k value is estimated to be 0.0152 min−1. While for a TiO2(001)–GD composite of
0.4 wt% GD, the k is highest and up to 0.0247 min−1, which is 1.62 times of that
obtained from the pure TiO2(001). For the TiO2(001)–GR composites, a weight ratio
Table 5.2 Photocatalytic degradation reaction kinetics over TiO2 nanosheets with different GD
and GR loading content
Loading (wt%) 0 0.2 0.4 0.6 0.8
−2 −1 GD 1.52 2.24 2.47 2.34 2.30
k/10 min
GR 1.52 1.58 1.66 1.95 1.65
of 0.6 % GR loading gives the highest rate constant of 0.0195 min−1, 1.28 times of
the pure TiO2(001). For all the cases investigated, the 2D carbon-based TiO2 compo‐
site show higher photocatalytic efficiencies than that of the pure TiO2 nanosheets,
which should be attributed to the synergetic charge transfer effect of the composites.
The effect of GD and GR loading on the performance of the resultant composites
would be explained based on the balance between the benefit of synergetic charge
transport and reduced light utilization efficiency. Because both GD and GR of strong
light absorbance reduce the light utilization efficiency, excess amount of GD, or GR
loading would overwhelm the benefit of synergetic charge transport, resulting a
decreased efficiency. The lower GD loading to reach such a critical point is because
the GD is relatively thicker and difficult to be dispersed than that of GR [25]. Mean‐
while, considering that all the three samples have the similar specific surface areas
(Fig. 5.6b), the higher dye absorption amount of the TiO2(001)–GD and TiO2(001)–
GR composites is attributed to the π–π stacking between dye molecules and 2D
carbon sheets. Interestingly, in comparison of TiO2(001)–GD and TiO2(001)–GR
composites, GR composite shows the better adsorptivity because of the large
π-conjugation structures (Fig. 5.6b), whereas the GD composite performs better in
the following photodegradation process, which should be attributed to its electronic
properties. In details, the excellent photocatalytic activity of the TiO2(001)–GD
composites could be ascribed to the following reasons.
5.2.3 Principle Analysis
(a) TiO2–GD composites possess improved charge separation ability. In the

TiO2–carbon composite materials, the photogenerated electrons from TiO2 can be
captured by the π-conjugated structure of carbon via a percolation mechanism.
Hence, both GD and GR act as the electron acceptors in the TiO2-carbon system,
effectively suppressing the charge recombination and leaving more holes to form
reactive species that promote the degradation of dyes. On the basis of analysis with
DFT method and XPS (Figs. 5.2 and 5.3), the GD has the strongest bonding and the
largest dipole with {001} facet of TiO2, which is able to promote the charge sepa‐
ration, retard the recombination and increase the photocatalytic ability.
(b) TiO2–GD composites possess abundant impurity levels. To gain insight into
the electronic structure of composites, the band structures, and partial density of states
(PDOS) of the six TiO2-carbon composites are illustrated in Figs. 5.7 and 5.8.
Fig. 5.7 Band structure and partial density of states (PDOS) for the composites, a and c are the
band structures of TiO2(001)–GD and TiO2(001)–GR, respectively. b and d Are the PDOS of
TiO2(001)–GD and TiO2(001)–GR, respectively. The Fermi level is set to zero. The red lines
represent the CBM and VBM position of the composites
Figure 5.8 shows that there are no isolated energy levels in the band gap of
TiO2(101)–GR or TiO2(110)–GR composite. While due to the strong hybridization and
bond formation, many isolated energy levels localized between the valence band
maximum (VBM) and conduction band minimum (CBM) can be discerned for both
TiO2(001)–GR and TiO2(001)–GD composites (Figs. 5.7 and 5.8) [50]. According to
the electron energy band structure theory, the impurity levels near the VBM belong to
p-type doping, while those close to the CBM is considered as n-type doping. Owing
to hole and electron is the donor for the p-type and n-type doping, respectively, we can
deduce from Figs. 5.1 and 5.7 that both positive and negative charges are introduced
into the composites. Carbon atoms are reported to show different valences of the (−2,
+4) prices [51, 52], when GD or GR is mixed with TiO2. In this case, the impurity
level is independent and it is not recombined, resulting in the electrons transfer from
VBM to isolated energy level or from isolated energy level to CBM more easily.
Meanwhile, with these isolated levels the electrons are more difficult to fall back from
CB to VBM, which retard the corresponding recombination [53]. The electronic exci‐
tation energy from the occupied isolated energy levels to the CBM is also smaller than
Fig. 5.8 The band structure and partial density of states for the composites, the Fermi energy is
set as zero, a TiO2(101)–GD, b TiO2(110)–GD, c TiO2(101)–GR, d TiO2(110)–GR. The red lines
represent CBM and VBM position of the composites
the energy from the VBM to CBM, which makes the electrons be excited more easily.
Here, we conclude that because GD introduce the impurity levels to TiO2 more easily,
while GR can only react with TiO2(001) facet, GD can be considered as a better mate‐
rial for property modification compared with GR. Furthermore, (001) facet of TiO2 is
the active facet so that it is easier to be modified by GR. Also, the TiO2(001)–GD
composites have the most impurity levels among all the composites (Figs. 5.7 and 5.8),
so there are more donor and accepter impurity levels for the electrons and the recom‐
bination will be prevented more efficiently.
It is noted that, for all TiO2–GD composites, the valence band is formed by hybrid
of C-2p and O-2p orbitals. The shallow impurity level that is close to the VBM of
the composites is obtained by the O-2p orbital and C-2p orbital, while the deep
impurity level is mainly constituted by the C-2p orbital. As comparison, for the
TiO2(001)–GR composites, the valence band is also formed by hybrid of C-2p and
O-2p orbitals, but both the shallow and deep impurity states are only originated from
C-2p electrons (Figs. 5.7 and 5.8). In both TiO2(001)–GD and TiO2(001)–GR
composites, continuum states near the VBM are formed rather than isolated state,
which can benefit enhancing the lifetime of the photoexcited carriers [54]. The
magnitude of PDOS of the continuum states of TiO2(001)–GD is stronger than that
of TiO2(001)–GR, revealing that the lifetime of the photoexcited carriers in
TiO2(001)–GD is longer than that in TiO2(001)–GR.
(c) TiO2–GD composites possess higher oxidation ability. X-ray photoelectron
valence band spectra (Fig. 5a) show that VB maxima of all three materials have some
differences. The VB values of pure TiO2(001) is ca. 1.95 eV lower than the Fermi
level, which is consistent with the former report [55, 56]. And for the TiO2(001)–
GD and TiO2(001)–GR, the VB values are ca. 1.62 and 1.88 eV lower than their
Fermi levels, respectively, which are similar with the simulation results (Table 5.3).
We also identify the valence band position of the composites by analyzing their band
structure and the partial density of states in Figs. 5.7, 5.8 and 5.9. The calculation
results based on Figs. 5.7 and 5.9 show that accompanying with the shift of the Fermi
level, the valence band positions for the TiO2(101)–GD, TiO2(110)–GD, TiO2(001)–
GD, TiO2(101)–GR, TiO2(110)–GR, and TiO2(001)–GR composites are −7.64,
−6.75, −6.85, −6.04, −5.62, and −6.63 eV, respectively. It proves that the TiO2–
GD composites always have the lower VB positions in respect to the corresponding
TiO2–GR composites. The more negative VB values would have the higher oxidation
ability [57]. In another word, comparing with TiO2–GR composites, TiO2–GD
composites always show higher oxidation ability, which leads to a higher photode‐
gradation performance.
5.3 Conclusion
In conclusion, the chemical structure and electronic properties of TiO2–GD and

TiO2–GR composites with different TiO2 facets are calculated by first-principles
density functional theory. The results disclose that the TiO2(001)–GD composite
Table 5.3 The calculated Fermi energy and VB and CB positions of the TiO2(001)–GD or
TiO2(001)–GR composites
EFermi (eV) ΔECalc./Expre. (eV) VBa (eV) CB (eV)
TiO2(001)–GD −5.19 1.66/1.62 −6.85 −5.08
TiO2(001)–GR −4.82 1.81/1.88 −6.63 −5.03
a
VB = EFermi − ECalc./Expre, EExpre is the value between VB and Fermi level from the XPS data
exhibits the most outstanding performance in rich electronic structure, charge sepa‐
ration, and the oxidation ability compared with pure TiO2(001) or TiO2(001)–GR
composite, which make itself become an excellent photocatalyst candidate of the
high efficiency. Our experiment further confirms the theoretical prediction that the
TiO2(001)–GD composite shows the highest photocatalysis performance compared
with other 2D carbon-based TiO2 composites even including TiO2(001)–GR. As a
result, we can expect that graphdiyne will become a superb competitor among the
different types of 2D carbon materials in the applications of photocatalysis and
photovoltaics.
Reagents: Titanium isopropoxide was purchased from J & K Company. HF (>48 %)

was purchased from Aladdin Reagent Company. Flake graphite of average particle
diameter 4 μm was purchased from Qingdao Tianhe Graphite Co. Ltd., China. Unless
otherwise specified, MB, and other reagents and materials were obtained commer‐
cially from the Beijing Chemical Reagent Plant (Beijing, China) and used as received
without further purification. The experiments were carried out at room temperature
and humidity.
Synthesis of TiO2 with (001) facets exposed: TiO2 was synthesized by following
Lou’s method [41]. Briefly, 3 ml of 48 % HF was slowly added into 10 ml of titanium
isopropoxide in 100 mL Teflon-lined stainless steel autoclave under mild stirring,
and then it was heated at 180 °C for 24 h. After cooled naturally to room temperature,
the white product was harvested and washed thoroughly with DI water and ethanol
by filtration, which was then dried at 60 °C overnight.
Synthesis of GD: [25] GD was synthesized on the surface of copper via a cross-
coupling reaction using hexaethynylbenzene as the precursors. In brief, the monomer
of hexaethynylbenzene was synthesized in good yield (62 %) by addition of tetra‐
butylammonium fluoride (TBAF) to a THF solution of hexakis[(trimethyl‐
silyl)ethynyl]benzene for 10 min at 8 °C. The GD was successfully grown on the
surface of copper foil in the presence of pyridine by a cross-coupling reaction of the
monomer of hexaethynylbenzene for 72 h at 60 °C under a nitrogen atmosphere. In
the process of forming GD, the copper foil was not the only a catalyst for the cross-
coupling reaction but also the substrate for the growing GD film.
Fig. 5.9 a XPS VB spectra and b CB, VB position of different composites. TiO2–GD composites
always have lower VB positions in respect to TiO2–GR composites
Synthesis of Composite Photocatalysts: The TiO2(001)–GD and TiO2(001)–GR

composites were obtained via a hydrothermal method. Briefly, GD or GO was
dissolved in a solution of distilled H2O (20 mL) and ethanol (40 mL) by ultrasonic
treatment for 1 h, and a proper weight of TiO2 was added to the obtained solution
which was stirred for another 2 h to achieve a homogeneous suspension. The suspen‐
sion was then placed in a 100 mL Teflon-sealed autoclave and maintained at 120 °C
for 3 h to achieve the composite of TiO2 and carbon. Finally, the resulting composite
was recovered by filtration, rinsed with ethanol for several times, and dried at room
temperature. Based on the XPS data (Fig. 2.3a and related analysis), the content of
oxygen in GO samples was about 33 %, while after the reduction, there was only
8 % oxygen left in the sample of the reduced graphene oxide. If we supposed that
the amount of C element could not be lost during the reduction process, the weight
loss of graphene oxide after the reduction was estimated to be around 27 % based on
the simple calculation. Note: The graphene amount in the composite was calculated
on the reduced graphene oxide (0.6 % loading represented that the composite
contained 0.6 mg reduced graphene oxide and 99.4 mg TiO2).
Photocatalytic Experiments: Photodegradation of MB was observed by using
UV-vis absorption spectroscopy. In a typical process, 30 mg photocatalyst was added
into an aqueous solution of MB (0.01 g L−1, 2.7 × 10−5 M, 40 mL) placed in a 50 mL
cylindrical quartz vessel, ultrasonic dispersion before stirred in dark for 30 min. At
given time intervals, the photoreacted solution was analyzed by recording variations
of the absorption band maximum (663 nm) in the UV-vis spectra of MB. Under
ambient conditions, 300 K recirculated water and stirring, the photoreaction vessel
was exposed to the AM 1.5 solar simulator illumination (Xe lamp, 100 mW/cm2).
Characterization: The XRD patterns were obtained by using an X’Pert PRO MDP
instrument with CuKα radiation (λ = 1.5405 Å) at 30 mA and 40 kV. SEM images
were obtained using a JEOL JSM-6700F scanning electron microscope at 3.0 kV.
TEM and SAED images were obtained on JEOL-1011 at 100 kV. HR-TEM images
were obtained on an FEI Tecnai F20 instrument at an acceleration voltage of 200 kV.
FTIR spectra were recorded on a Bruker EQUINOX55 FTIR spectrophotometer.

UV-vis spectra were recorded on a Hitachi Model U-4100 spectrophotometer. XPS
VB data were obtained with an ESCALab220i-XL electron spectrometer from VG
Scientific using 300 W AlKα radiation. The binding energies obtained in the XPS
analysis were corrected with reference to C 1s (284.8 eV). Raman spectra were
measured on a Renishaw-2000 Raman spectrometer at a resolution of 1 cm−1 by
using an argon-ion laser of the 514.5 nm line as the excitation source. The nitrogen
adsorption and desorption isotherms at the temperature of liquid nitrogen (77 K)
were measured on a Quantachrome Autosorb-1 sorption analyzer. Multipoint BET
surface area was estimated at the relative pressure range from 0.05 to 0.2.
Computation details: We performed DFT calculations using Dmol3 code with the
generalized gradient approximation (GGA) and PW91. Double Numerical plus
polarization (DNP) with a real-space cutoff of 5.2 Å had been employed during the
geometry optimization and single point energy calculation. DFT Semi-core Pseu‐
dopots were implemented for relativistic effects, which replaced core electrons by
effective potentials of Ti elements; all-electron was used as the core treatment for O
and C elements. The maximum energy change convergence threshold for self-
consistent field (SCF) was 1 × 10−6 Ha/atom; atomic relaxation was carried out until
all components of the residual forces were less than 0.002 Ha/Å, and the maximum
displacement was within 0.005 Å. Monkhorst-Pack k-point mesh was used for
geometry optimization and electronic property calculations
Because of the different lattice parameters of GD, GR, and TiO2, we chose
different supercells to make sure that the mismatch of the TiO2–GR or GD surfaces
was less than 5.31 %. The A vacuum layer larger than 15 Å was used to avoid the
spurious interactions between two adjacent layers, and the optimized structure of the
composite was shown in Fig. 5.1.
5.5 Postscript
Graphyne, as a new member in carbon family, a lot of physical and chemical prop‐
erties are still not very clear, although scientists have made a lot of theoretical anal‐
ysis. This material, composite with benzene and carbon triple bond may have
different permutations mode. The versatile structures will appear different intrinsic
properties. The big difficulty now is how to get the high-quality graphyne. Also, for
the applications, another problem is due to the inertness of its surface, which is
difficult to disperse in solvent or modify for application. But the benefit of it is quite
obvious, because of the triangular void in the molecular structure of graphyne, it will
be benefit for the diffusion of electron and even small molecules. If we can get a
single layer of graphyne, it will be interesting to study the ion/molecule transfer
through the film, and will bring a broader space for its application.
References 109
References
1. Linsebigler, A.L., Lu, G., Yates, J.T.: Chem. Rev. 95, 735 (1995)
2. Pelaez, M., Nolan, N.T., Pillai, S.C., Seery, M.K., Falaras, P., Kontos, A.G., Dunlop, P.S.M.,
Hamilton, J.W.J., Byrne, J.A., O’Shea, K., et al.: Appl. Catal. B: Environ. 125, 331 (2012)
3. Zhang, H., Lv, X., Li, Y., Wang, Y., Li, J.: ACS Nano 4, 380 (2009)
5. Wee, A.T.S.: ACS Nano 6, 5739 (2012)
6. Huang, X., Qi, X., Boey, F., Zhang, H.: Chem. Soc. Rev. 41, 666 (2012)
7. Huang, X., Yin, Z., Wu, S., Qi, X., He, Q., Zhang, Q., Yan, Q., Boey, F., Zhang, H.: Small
2011, 7 (1876)
8. Wu, Q., Xu, Y.X., Yao, Z.Y., Liu, A.R., Shi, G.Q.: ACS Nano 2010, 4 (1963)
9. Huang, X., Zeng, Z., Fan, Z., Liu, J., Zhang, H.: Adv. Mater. 24, 5979 (2012)
10. Cao, X., Shi, Y., Shi, W., Lu, G., Huang, X., Yan, Q., Zhang, Q., Zhang, H.: Small 7, 3163
(2011)
11. Qu, L., Liu, Y., Baek, J.-B., Dai, L.: ACS Nano 4, 1321 (2010)
12. Liu, H., Wang, G.X., Liu, J., Qiao, S.Z., Ahn, H.J.: J. Mater. Chem. 21, 3046 (2011)
13. Huang, C., Li, C., Shi, G.: Energy Environ. Sci. 5, 8848 (2012)
14. Wu, S., He, Q., Zhou, C., Qi, X., Huang, X., Yin, Z., Yang, Y., Zhang, H.: Nanoscale 4, 2478
(2012)
16. Chen, C., Cai, W., Long, M., Zhou, B., Wu, Y., Wu, D., Feng, Y.: ACS Nano 4, 6425 (2010)
17. He, Q.Y., Zeng, Z.Y., Yin, Z.Y., Li, H., Wu, S.X., Huang, X., Zhang, H.: Small 8, 2994 (2012)
18. He, Q., Wu, S., Gao, S., Cao, X., Yin, Z., Li, H., Chen, P., Zhang, H.: ACS Nano 5, 5038 (2011)
19. Sudibya, H.G., He, Q., Zhang, H., Chen, P.: ACS Nano 2011, 5 (1990)
20. Cao, X., He, Q., Shi, W., Li, B., Zeng, Z., Shi, Y., Yan, Q., Zhang, H.: Small 7, 1199 (2011)
21. He, Q., Wu, S., Yin, Z., Zhang, H.: Chem. Sci. 3, 1764 (2012)
22. Bao, Q., Loh, K.P.: ACS Nano 6, 3677 (2012)
23. Sun, Z.P., Hasan, T., Torrisi, F., Popa, D., Privitera, G., Wang, F.Q., Bonaccorso, F., Basko,
D.M., Ferrari, A.C.: ACS Nano 4, 803 (2010)
24. Haley, M.M., Brand, S.C., Pak, J.J.: Angew. Chem. Int. Ed. 36, 836 (1997)
25. Li, G., Li, Y., Liu, H., Guo, Y., Li, Y., Zhu, D.: Chem. Commun. 46, 3256 (2010)
26. Diederich, F.: Nature 369, 199–207 (1994)
27. Malko, D., Neiss, C., Vines, F., Goerling, A.: Phys. Rev. Lett. 108, 086804 (2012)
28. Long, M., Tang, L., Wang, D., Li, Y., Shuai, Z.: ACS Nano 5, 2593 (2011)
29. Berger, C., Song, Z., Li, X., Wu, X., Brown, N., Naud, C., Mayou, D., Li, T., Hass, J.,
Marchenkov, A.N., et al.: Science 312, 1191 (2006)
30. Tan, Y.W., Zhang, Y., Bolotin, K., Zhao, Y., Adam, S., Hwang, E.H., Das Sarma, S., Stormer,
H.L., Kim, P.: Phys. Rev. Lett. 99, 246803 (2007)
31. Wang, S., Yi, L., Halpert, J.E., Lai, X., Liu, Y., Cao, H., Yu, R., Wang, D., Li, Y.: Small 8,
265 (2012)
32. Lazzeri, M., Vittadini, A., Selloni, A.: Phys. Rev. B 63, 155409 (2001)
33. Diebold, U.: Surf. Sci. Rep. 48, 53 (2003)
34. Yang, H.G., Liu, G., Qiao, S.Z., Sun, C.H., Jin, Y.G., Smith, S.C., Zou, J., Cheng, H.M., Lu,
G.Q.: J. Am. Chem. Soc. 131, 4078 (2009)
35. Lazzeri, M., Vittadini, A., Selloni, A.: Phys. Rev. B 63, 155409 (2001)
36. Yang, H.G., Sun, C.H., Qiao, S.Z., Zou, J., Liu, G., Smith, S.C., Cheng, H.M., Lu, G.Q.: Nature
453, 638 (2008)
37. Jayasekera, T., Xu, S., Kim, K.W., Nardelli, M.B.: Phys. Rev. B 84, 035442 (2011)
38. Wang, Q.J., Che, J.G.: Phys. Rev. Lett. 103, 066802 (2009)
39. Wong, K., Zeng, Q.H., Yu, A.B.: J. Phys. Chem. C 115, 4656 (2011)
40. Khomyakov, P.A., Giovannetti, G., Rusu, P.C., Brocks, G., van den Brink, J., Kelly, P.J.: Phys.
Rev. B 79, 195429 (2009)
41. Chen, J.S., Liu, J., Qiao, S.Z., Xu, R., Lou, X.W.: Chem. Commun. 47, 10443 (2011)
42. Liu, G., Yang, H.G., Wang, X., Cheng, L., Pan, J., Lu, G.Q., Cheng, H.-M.: J. Am. Chem. Soc.
131, 12868 (2009)
43. Xiang, Q., Yu, J., Wang, W., Jaroniec, M.: Chem. Commun. 47, 6906 (2011)
45. Akhavan, O., Abdolahad, M., Abdi, Y., Mohajerzadeh, S.: Carbon 47, 3280 (2009)
46. Oliveira, C., Gonçalves, L., Almeida, B.G., Tavares, C.J., Carvalho, S., Vaz, F., Escobar
Galindo, R., Henriques, M., Susano, M., Oliveira, R.: Surf. Coat. Tech. 203, 490 (2008)
47. Pillai, S.C., Periyat, P., George, R., McCormack, D.E., Seery, M.K., Hayden, H., Colreavy, J.,
Corr, D., Hinder, S.J.: J. Phys. Chem. C 111, 1605 (2007)
48. Sun, L., Zhao, Z., Zhou, Y., Liu, L.: Nanoscale 4, 613 (2012)
49. Wang, X.H., Li, J.G., Kamiyama, H., Moriyoshi, Y., Ishigaki, T.: J. Phys. Chem. B 110, 6804
(2006)
50. Shi, W.-J., Xiong, S.-J.: Surf. Sci. 2010, 604 (1987)
51. Zhu, W., Qiu, X., Iancu, V., Chen, X.-Q., Pan, H., Wang, W., Dimitrijevic, N.M., Rajh, T.,
Meyer III, H.M., Paranthaman, M.P., et al.: Phys. Rev. Lett. 103, 226401 (2009)
52. Yang, K., Dai, Y., Huang, B., Whangbo, M.-H.: J. Phys. Chem. C 113, 2624 (2009)
53. Zhou, Z., Li, M., Guo, L.: J. Phys. Chem. Solids 71, 1707 (2010)
54. Ma, X., Wu, Y., Lu, Y., Xu, J., Wang, Y., Zhu, Y.: J. Phys. Chem. C 115, 16963 (2011)
55. Pan, J., Liu, G., Lu, G.Q.M., Cheng, H.-M.: Angew. Chem. Int. Ed. 50, 2133 (2011)
56. Pan, J., Wu, X., Wang, L., Liu, G., Lu, G.Q., Cheng, H.-M.: Chem. Commun. 47, 8361 (2011)
57. Chen, X., Shen, S., Guo, L., Mao, S.S.: Chem. Rev. 110, 6503 (2010)
Chapter 6
Conclusions and Perspectives
In order to improve the utilization of solar energy, this thesis focuses on designing
and developing new nanocomposites, and studying their applications in the field of
dye-sensitized solar cell and photocatalysis. The overall design origins from two
strategies, enhancement of electron transport and light utilization efficiency. To
improve the electron transport, we first introduced graphene into photoanode of dye-
sensitized solar cell and found an appropriate graphene composite that can well
enhance the photoelectric conversion efficiency. After knowing the charge transport
mechanism between TiO2 and graphene, we prepared a more precise structure,
forming some biomimetic stacking structures with graphene and titanium oxide
nanosheets through the layer-by-layer process, and investigated the photoelectric
conversion and electron transport properties in this kind of structure. Besides the
inorganic stacking structure, we also extend this method to organic stacked structure
with graphite oxide and conductive polymer.
In terms of increasing light utilization efficiency, we introduced surface plasmon
resonance effect into the counter electrode of dye-sensitized solar cells, and the
photoelectric conversion efficiency is improved as expected.
After that, we moved the study to photocatalysis of 2D–2D composites. Titanium
oxide nanosheet is combined with another two kinds of 2D carbon materials respec‐
tively, graphdiyne and graphene. Simulation first predicted the bonding and electron
transport situation in the composites. And later the theoretical result was verified by
photocatalytic degradation experiment.
The main conclusions of this thesis are as follows:
1. Graphene was introduced into the titanium oxide photoanode as a bridge for
electron transport in DSSC. The introduced graphene can extend the life of
photoexcited electrons and resist recombination of electrons and holes.

112 6 Conclusions and Perspectives
The resulting photogenerated current was improved by 45 %, accompany with

the photoelectric conversion efficiency improved by 39 %.
2. Inspired by nature, the stacking structure of titanium oxide and graphene were
fabricated by layer-by-layer method to imitate the structure and function of
granum. Through photoelectric behavioral testing to the stacked structure, we
found that the graphene composite film exhibited higher photoelectric conversion
capability. The light-generated current is 20 times higher than the ordinary tita‐
nium oxide film under the same conditions with 25 structural units. More impor‐
tantly, we found that the current direction reversion occurred with increasing
film thickness. That is when the film is thin, the electrodes often exhibit cathode
photocurrent and only when the film reaches a certain thickness, anodic photo‐
current will be produced. After that, then we stacked graphite oxide and
conducting polymers (polyaniline) using π–π interaction. In this case, we found
the graphene oxide performs as both template and dopant, which stabilized the
stacked structure.
3. By the aspect of increasing the light utilization, a gold topological ordered layer
was introduced into the counter electrode of DSSC. Accompany with the surface
plasmon resonance of gold, this structure can enhance the light scattering and
reflection, then improve the cell performance. The efficiency and photocurrent
was increased by 18 and 16 %, respectively, and the IPCE was enhanced by 25 %
at all wavelengths. Besides, this counter electrode performs well at both strong
and low solar irradiation.
4. At last, we explored the 2D TiO2–2D carbon composite to photocatalysis. The
results obtained from theoretical and experimental analysis disclose that the
TiO2(001)–GD composite exhibits the most outstanding performance in rich
electronic structure, charge separation, and the oxidation ability compared with
pure TiO2(001) or TiO2(001)–GR composite, which makes itself become an
excellent photocatalyst candidate of the high efficiency.
This thesis focuses on two aspects as enhancing the charge transport and
increasing the light utilization, investigating the performance of 2D carbon material
and SPR effects on both DSSC and photocatalysis. Meanwhile, this thesis reveals
the charge transport performance in nanocomposite, realizing the high efficient
conversion of solar energy. Based on the work obtained here, this field can be devel‐
oped in the following directions:
1. In the work of introducing graphene into photoanode, because of the lacking of
compact layer, the short circuit is easy to take place. How to choose a suitable
material for the compact layer is a key point for the further enhancement of
efficiency.
2. Until now, no direct proof to reveal the mechanism of charge transport in TiO2–
Carbon composite. The similar and different properties among of graphene,
graphyne, and carbon nanotube are still an open question. What is the intrinsic
property for different carbon materials?
3. Further effort is still necessary to get a high purity and large scale fabrication of
graphene. The mostly used graphene nowadays is still reduced graphene oxide,
6 Conclusions and Perspectives 113
which has a lot of defects. The electric property will be affected by these defects.
Moreover, the relationship between the chemical modification and electric prop‐
erty is still unclear. If we can control the electric performance by modifying
graphene, a lot of work can be developed.
4. To graphdiyne, the urgency is to make it well dispersed. This whole carbon
material is very easy to aggregate and its performance is affected. Moreover, the
properties of graphyne are related to their structures. Because the conductivity
of graphyne is lower than graphene, one needs to consider the balance between
the charge capture, and transport, avoid forming the recombination center.
Importantly, we can pay more attention to the triangle void in its plane. This
void/modified void can be used for the ions or electrons transport channel, which
will bring some new ideas for research.
5. The stacking structure brings new inspiration, which is very promising, i.e,
composite two materials (conductor–semiconductors, semiconductor–semicon‐
ductor, conductor–insulator) with lay-by-layer method to form a stacking struc‐
ture. With this structure, first, we can build in series or parallel heterojunction;
second, we can design a parallel circuit. Considering the nanosheet itself is only
a nanometer thickness, if we can well control the electron transport, and avoid
tunneling, it can be fabricated to supercapacitors. And a more attractive direction
is this method may be adopted in integrated chip technology and bring a new
revolution.

The Preparation of Nano Composites and Their Applications in Solar Energy Conversion

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

The Preparation of Nano Composites and Their Applications in Solar Energy Conversion

Transféré par

Droits d'auteur :

Formats disponibles

Springer Theses

Recognizing Outstanding Ph.D. Research

Recognizing Outstanding Ph.D. Research

Theses are accepted into the series by invited nomination only

• They must be written in good English.

More information about this series at http://www.springer.com/series/8790

The Preparation of Nano

ISSN 2190-5053 ISSN 2190-5061 (electronic)

© Springer-Verlag GmbH Germany 2017

Printed on acid-free paper

This Springer imprint is published by Springer Nature

Beijing, China Prof. Dan Wang

After graduation of 3 years, it is time to reread the original acknowledgment, and I

Systems, Ph.D. Programs Foundation of the Ministry of Education of China and

Yunnan Garden, Singapore

5.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

AFM Atomic Force Microscopy

© Springer-Verlag GmbH Germany 2017 1

1.1 Dye-Sensitized Solar Cell (DSSC)

1.1.1 Development of Solar Cells

1.1.1.1 Classiﬁcation of Solar Cells

1.1.1.2 Development of Solar Cells

Fig. 1.2 The structure of a

1.1.2 Basic Principle of DSSC

1.1.2.1 Structure of DSSC

1.1.2.2 Principle of DSSC

electrolyte through TiO2 network is very important. Hence, it is necessary to opti‐

1.1.2.3 Parameter of DSSC

In order to evaluate the performance of DSSC, a series of characterization parameters

1.1.3 Introduction of the Constitution of DSSC

1.1.3.1 Nanocrystal Photoanode

Fig. 1.4 A schematic Energy

Fig. 1.5 Schematic drawing

phthalocyanine molecule, it needs a further modiﬁcation in subsequent experiments.

is often speculated through phenomenon. Most of the improvements in performance

1.1.3.4 Counter Electrode

Counter electrode is also known as photocathode. Besides performing as the electron

Another signiﬁcant role of titanium dioxide is photocatalysis. As known, semicon‐

TiO2 + h𝜈 → TiO2 (e−CB + h+VB ) (1.3)

In the picosecond period, photogenerated electrons migrate to the surface of

Ti4+ + e−CB → Ti3+ (1.4)

The photogenerated electrons can be eﬀectively captured by oxygen molecules

h+VB + OH− → OH⋅ (1.7)

h+VB + H2 O → OH⋅ + H+ (1.8)

Organic are attack by the radicals, degradation reactions take place

e−CB + h+VB → E (1.10)

Ti3+ + h+VB → Ti4+ (1.11)

(a) Metal (b)

1.2.2 Modiﬁcation of Catalyst

1.2.2.1 Suppress Recombination of Photogenerated Electrons and Holes

1.2.2.2 Modiﬁcation to Enhance the Visible Light Response

(a) V/NHE (b) V/NHE

Conduction band -1.0 Conduction band -1.0

1.3 Carbon Materials in Photoelectric Conversion System

Carbon material is progressing accompany with the development of human being.

1.3.1 Donor–Acceptor Photovoltaic Material

1.3.1.1 Structure of Fullerenes

1.3.1.2 Application of Fullerenes in Photovoltaic Field

The investigation of fullerenes in the photovoltaic area is mainly based on light-

above is not conducive to the carrier transport or collection. Hence, he proposed an

1.3.2 Application of One-Dimensional Carbon Nanotubes

1.3.2.1 Structure of Carbon Nanotubes

1.3.2.2 Application of Carbon Nanotubes in the Photovoltaic Field