theoretical characterization of functional molecular materials › smash › get › diva2:525965...
Post on 09-Jun-2020
6 Views
Preview:
TRANSCRIPT
Theoretical Characterization of
Functional Molecular Materials
Xiuneng Song
(宋秀能)
Doctoral Thesis in Theoretical Chemistry and Biology
School of Biotechnology
Royal Institute of Technology
Stockholm, Sweden 2012
Theoretical Characterization of
Functional Molecular Materials
Thesis for Philosophy Doctor degree
Department of Theoretical Chemistry and Biology
School of Biotechnology
Royal Institute of Technology
Stockholm, Sweden 2012
c⃝ Xiuneng Song, 2012
pp i-xviii, 1-49
ISBN 978-91-7501-367-1
ISSN 1654-2312
TRITA-BIO Report 2012:16
Printed by Universitetsservice US-AB,
Stockholm, Sweden 2012
To my parents
Abstract
Nowadays, material, energy and information technologies are three pillar indus-
tries. The materials that have close relation with our life have also been the
foundation for the development of energy and information technologies. As the
new member of the material family, functional molecular materials have become
increasingly important for many applications, for which the design and character-
ization by the theoretical modeling have played the vital role. In this thesis, three
different categories of functional molecular materials, the endohedral fullerenes,
the fullerene derivatives and the self-assembled monolayers (SAMs), have been
studied by means of first principles methods.
The non-metal endohedral fullerene N@C60 is a special endohedral fullerene
that is believed to be relevant to the construction of future quantum computer.
The energy landscape inside the N@C60 has been carefully explored by density
functional theory (DFT) calculations. The most energy favorable potential energy
surfaces (PESs) for the N atom to move within the cavity have been identified.
The effect of the charging on the PESs has also been examined. It is found
that the inclusion of dispersion force is essential in determining the equilibrium
structure of N@C60. Furthermore, the performance of several commonly used
density functionals with or without dispersion correction has been verified for ten
different endohedral fullerenes A@C60 with the atom A being either reactive non-
metal or nobel gases elements. It shows that the inclusion of the dispersion force
does provide better description for the binding energy (BE), however, none of
them could correctly describe the energy landscape inside all the ten endohedral
fullerenes exclusively. It thus calls for the further improvement of current density
functionals for weak interacting systems.
Soft X-ray spectroscopy is a powerful tool for studying the chemical and elec-
tronic structures of functional molecular materials. Theoretical calculations have
been proven to be extremely useful for providing correct assignments for spectra
of large systems. In this thesis, we have performed first principles simulations for
the near-edge X-ray absorption fine structure (NEXAFS) and X-ray photoelectron
spectra (XPS) of fullerene derivatives and aminothiolates SAMs. Our calculated
spectra can accurately reproduce experimental results available for all the systems
under investigations, and identify the species or structures that are responsible for
vi
those unexpected spectral features observed in experiments. We have suggested a
modified building block (MBB) approach that allows to calculate NEXAFS spec-
tra of a large number of fullerene derivatives with very small computational cost,
and resolved the long standing puzzle around the experimental XPS and NEXAFS
spectra of SAMs with aminothiolates.
Preface
The works presented in this thesis have been carried out in the past three years,
2009.04 - 2012.06, at the Department of Theoretical Chemistry and Biology, School
of Biotechnology, Royal Institute of Technology, Stockholm, Sweden.
List of papers included in the thesis
Paper 1. Xiuneng Song, Yong Ma, Chuankui Wang and Yi Luo, Energy land-
scape inside the cage of neutral and charged N@C60, Chem. Phys. Lett.,
517 (2011) 119.
Paper 2. Xiuneng Song, Yuejie Ai, Yong Ma, Chuankui Wang and Yi Luo, The
equilibrium geometry of A@C60: a difficult case for conventional density
functional theory, ChemPhysChem, submitted.
Paper 3. Xiuneng Song, Weijie Hua, Yong Ma, Chuankui Wang and Yi Lu-
o,Theoretical study of core excitation spectroscopy of fullerene-based so-
lar cell acceptors, Phys. Chem. Chem. Phys. submitted.
Paper 4. Xiuneng Song, Yong Ma, Chuankui Wang, Paul M. Dietrich, Wolfgang
E. S. Unger and Yi Luo, Effects of protonation, hydrogen bonding and
photo-damaging on X-ray spectroscopy of the amine terminal group in
aminothiolate monolayers, J. Phys. Chem. C, 000 (2012) 000.
List of related papers not included in the thesis
Paper 1. Xiuneng Song, Yong Ma, Chuankui Wang and Yi Luo, Theoretical s-
tudy on Raman spectra of Rhodamine molecules in solutions, in manuscrip-
t.
Paper 2. Li-Li Lin, Xiuneng Song, Yi Luo and Chuankui Wang, Formation and
electronic transport properties of bimolecular junctions based on aromat-
ic coupling, J. Phys.: Condens. Matter, 22 (2010) 325102.
viii
Paper 3. Li-Li Lin, Xiuneng Song, Jiancai Leng, Zongliang Li, Yi Luo and
Chuan-Kui Wang, Determination of the Configuration of a Single Molecule
Junction by Inelastic Electron Tunneling Spectroscopy, J. Phys. Chem.
C, 114 (2010) 5199.
Paper 4. Chuankui Wang, Bin Zou, Xiuneng Song, Yingde Li, Zongliang Li
and Lili Lin, Simulations of inelastic electron tunneling spectroscopy of
semifluorinated hexadecanethiol junctions, Front. Phys. China, 4 (2009)
415.
Paper 5. Li-Li Lin, Jiancai Leng, Xiuneng Song, Zongliang Li, Yi Luo and
Chuan-Kui Wang, Effect of Aromatic Coupling on Electronic Transport
in Bimolecular Junctions, J. Phys. Chem. C, 113 (2009) 14474.
Paper 6. Jiancai Leng, Lili Lin, Xiuneng Song, Zongliang Li and Chuankui
Wang, Orientation of Decanethiol Molecules in Self-Assembled Monolay-
ers Determined by Inelastic Electron Tunneling Spectroscopy, J. Phys.
Chem. C, 113 (2009) 18353.
Comments on my contributions to the papers included
I have taken the major responsibility for calculations and writing manuscripts of
all the papers included in this thesis.
Acknowledgments
This thesis could not be completed without many people’s help. Herein, I should
say thanks to these people, for their great, selfless, kindly and generous help.
I sincerely appreciate my supervisor Prof. Yi Luo for his wonderful guide in
all the three years. Prof. Luo is not only a supervisor but also a good friend or
even family to me. In these three years, I learned so much from Prof. Luo. He
is a great supervisor, and I could always find the right direction from him when I
was in trouble during my study; he is a fantastic supervisor, and I could always
get amazing idea from him when I was racking my brain. The most important
thing that Prof. Luo has given me is the positive encouragement which has always
strengthened my confidence. Without that, I am very sure I could not finish this
thesis. And also, I won’t forget the good time spent together with Prof. Luo’s
family, I really enjoy the time being with them.
I would like to give my genuine thanks to Prof. Chuankui Wang of Shandong
Normal University (SDNU). It is him who introduced me to this research field,
and gave me the fundamental knowledge of my research. It is him who cares about
my life as my family. His attitude of life encouraged me a lot and helped me to
handle many hard situations. There are also a lot of thanks to Prof. Wang’s
family, for their care and love.
I am very grateful to Prof. Hans Agren, the head of the department of
Theoretical Chemistry & Biology, for giving me the opportunity to study in this
group. This group is a wonderful family under your leading, I get too much fun
and I enjoy the time being here.
Acknowledge all the members of our group for your help and suggestions.
Many thanks to Dr. Ying Fu and Dr. Yaoquan Tu for their help in my life. I
also got a lot of help from the lectures given by Prof. Faris Gel’mukhanov, Prof.
Kersti Hermansson, Prof. Boris Minaev, Prof. Margareta Blomberg and Prof.
Per E. M. Siegbahn. Thank our kindly secretaries Ms. Nina Bauer, Ms. Marlene
Johnsson and Ms. Lotta Rosenfeldt for their guide and help with many practical
matters.
There are many people of our group who gave the generous help to me all the
time. Yuping, I was so lucky to get your help when my first landing in Sweden,
x
I would remember the days we were together. Lili, thank you for all your help,
both in living and academic times. I would like to thank Dr. Bin Gao, Dr. Weijie
Hua and Dr. Jiayuan Qi for teaching me the X-ray technique. Many thanks to
Dr. Stafan Hellstrom, Guangjun Tian and Fuming Ying for their help about the
computer technique.
I want to thank my colleges Dr. Hui Cao, Dr. Shilv Chen, Dr. Xing Chen,
Dr. Peng Cui, Dr. Sai Duan, Dr. Qiu Fang, Dr. Kai Fu, Dr. Qiang Fu, Dr.
Junkuo Gao, Dr. Xiaofei Li, Dr. Xin Li, Dr. Rongzhen Liao, Dr. Na Lin, Dr.
Jicai Liu, Dr. Zhijun Ning, Dr. Hao Ren, Dr. Liqin Xue, Dr. Xifeng Yang,
Dr. Feng Zhang, Dr. Qiong Zhang, Dr. Wenhua Zhang, Dr. Ying Zhang, Dr.
Ke Zhao, Xinrui Cao, Zhihui Chen, Xiao Cheng, Wei Hu, Yongfei Ji, Yu Kang,
Hongbao Li, Junfeng Li, Li Li, Bonaman Xin Li, Lijun Liang, Xiangjun Shang,
Ce Song, Lu Sun, Xianqiang Sun, Yan Wang and ChunZe Yuan.
Thanks to Dr. Weijie Hua for sharing the LATEX template. Special thanks to
Dr. Yuejie Ai, Dr. Weijie Hua, Dr. Keyan Lian, Dr. Lili Lin, Li Gao, Quan Miao
and Ying Wang for their precious suggestions to improve the papers and thesis.
Finally, I would give my special thanks to my parents and my husband Yong
Ma for their endless love and support. I love them and I wish I could pay the love
back to them.
Xiuneng
Spring 2012, Stockholm
Abbreviations
ABT 4-aminobenzenethiol
AFM atomic force microscopy
APBT 4-aminophenylbutane-1-thiol
ATPT 3-(4”-amino-1,1’:4’,1”-terphenyl-4-yl)propane-1-thiol
AUDT 11-aminoundecane-1-thiol
BB building block
BE binding energy
BHJ bulk heterojunction
bisPC60BM bis[6,6]-phenyl-C61-butyric acid methyl ester
BO Born-Oppenheimer
CC Coupled Cluster
CDB cumulative double bond
CI Configuration Interaction
CL cross-linking
CV cyclic voltammetry
DCP dispersion correcting potential
DFT density functional theory
DPM bis(4-methoxyphenyl)methano[60]fullerene
ESR electronic spin resonance
FCH full core hole
GGA generalized gradient approximation
GIR-IR grazing incidence reflectance infrared spectroscopy
HBA hydrogen bonded amine
xi
xii Abbreviations
HF Hartree-Fock
HK Hohenberg-Kohn
HOMO highest occupied molecular orbital
HWHM half width at half-maximum
IA impedance analysis
IP ionization potential
KS Kohn-Sham
LDA Local density approximation
LDF London dispersion force
LSDA Local spin density approximation
LUMO lowest unoccupied molecular orbital
MBB modified building block
MDMO-PPV Poly[2-methoxy-5-(3’,7’-dimethyloctyloxy)-1,4-phenylene vinylene]
MEH-PPV Poly[2-methoxy-5-(2-ethylhexyloxy)-1,4-phenylenevinylene]
MO molecular orbital
MPn Møller-Plesset perturbation theory
MRI magnetic resonance imaging
NEXAFS near-edge X-ray absorption fine structure
NMR nuclear magnetic resonance
P3HT poly(3-hexylthiophene)
PAmi protonated amine
PAmm primary ammonium
PAC protonated amine complex
PC60BM [6,6]-Phenyl-C61-butyric acid methyl ester
PC70BM [6,6]-Phenyl-C71-butyric acid methyl ester
PC84BM [6,6]-Phenyl-C85-butyric acid methyl ester
PES potential energy surface
PSC polymer solar cell
PSCA polymer solar cell acceptor
SAM self-assembled monolayer
SC side chain
SCF self-consistent field
SEM scanning electronic microscopy
Abbreviations xiii
SERS surface-enhanced Raman scattering
STM scanning tunneling microscopy
SWNT single-wall carbon nanotube
TDS thermal desorption spectroscopy
ThC60BM [6,6]-Thienyl-C61-butyric acid methyl ester
XANES X-ray absorption near-edge structure
XAS X-ray absorption spectroscopy
XPS X-ray photoelectron spectroscopy
XES X-ray emission spectroscopy
∆KS ∆Kohn-Sham
List of Figures
1.1 Geometry structure of fullerene C60. . . . . . . . . . . . . . . . . . . . . 2
1.2 Geometry structures of endohedral fullerenes (a) N@C60, (b) Sc3N@C80
and (c) Sc2C2@C84. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Illustration of two Gd@C82 endohedral metallofullerenes in a (11,9) SWN-T with schematic representation of the modulation of conduction andvalence bands. One expects to see the minute elastic strain around themetallofullerenes. Reprinted with permission from ref. [17]. Copyright c⃝2003 Springer Berlin/Heidelberg. . . . . . . . . . . . . . . . . . . . . . . 3
1.4 Scheme for a solid-state spin quantum computer based on linear chains ofendohedral fullerenes. Reprinted with permission from ref. [22]. Copyrightc⃝ 2002 John Wiley and Sons. . . . . . . . . . . . . . . . . . . . . . . . . 4
1.5 Structures of (a) C60 and fullerene derivatives (b) PC60BM, (c) bisPC60BM,and (d) PC70BM. (e) Schematic draw of a bulk heterojunction polymersolar cell made of (PC60BM) fullerenes derivatives. Reprinted with per-mission from ref. [27]. Copyright c⃝ 2011 American Chemical Society. . . 5
1.6 Schematic diagram of an ideal, single-crystalline SAM of alkanethiolatessupported on a gold surface with a (111) texture. The anatomy andcharacteristics of the SAM are highlighted. Reprinted with permissionfrom ref. [2]. Copyright c⃝ 2005 American Chemical Society. . . . . . . . 6
2.1 Schematized London dispersion force interaction between two moleculeswith unsymmetrically distributed electron cloud. . . . . . . . . . . . . . 16
2.2 Parallel benzene dimer (D6h symmetry) PESs generated using differentmethods. Reprinted with permission from ref. [67]. Copyright c⃝ 2009John Wiley and Sons. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.1 Schematic representation of the absorption processes, a: XPS, b: NEX-AFS and emission processes, c: XES, d: RXES. The occupied and unoc-cupied orbitals are figured by solid and dashed lines. c, v and u denotethe core, valance, unoccupied orbitals, ~ω and ~ω′ are energies of incidentand emission photon, respectively. . . . . . . . . . . . . . . . . . . . . . 20
4.1 The energy contours of two σv planes of La@C82 (C2v) (energy in kcalmol−1), only the region of which the energy is less than 5 kcal mol−1
is shown. Reprinted with permission from ref. [79]. Copyright c⃝ 2007American Chemical Society. . . . . . . . . . . . . . . . . . . . . . . . . . 28
xiv
LIST OF FIGURES xv
4.2 Schematic view of the experimental set-up for the production of N@C60.(a) By simultaneous deposition of C60 on a substrate and irradiation withnitrogen ions. (b) By using the glow discharge for the production of theions. Reprinted with permission from ref. [95]. Copyright c⃝ 1998 SpringerBerlin/Heidelberg. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.3 (a) Five possible pathways for N atom to escape from N@C60. (b) Po-tential energy surfaces for S=3/2 state within the distance region of 0.0
A to 3.5 A from the center of the cage. . . . . . . . . . . . . . . . . . . . 31
4.4 (a) Potential energy surfaces of the N moves along the reaction pathwaystowards the center of the 6-6 Bond inside neutral N@C60 cage at differentspin states. (b) Potential energy surfaces obtained from B3LYP and M06-2X functionals both at S=3/2 around the center of the carbon cage. . . 32
4.5 Potential energy surfaces obtained at three different density functionalstowards the center of the 6-6 Bond from -0.5 A to 0.5 A for (a) Kr@C60
and (b) Ne@C60, respectively. . . . . . . . . . . . . . . . . . . . . . . . . 33
5.1 Energy level diagram of a donor/acceptor interface showing photo ex-citation of an electron into the LUMO of the donor, followed by elec-tron transfer into the LUMO of the acceptor, and migration of separatedcharges away from the interface. . . . . . . . . . . . . . . . . . . . . . . 36
5.2 Optimized structures of PC60BM, PC70BM, PC84BM, ThC60BM, DPMand bisPC60BM. The hydrogen, carbon, oxygen and sulfur atoms arerepresented by white, grey, red and yellow spheres, respectively. . . . . . 37
5.3 Calculated XPS of PC60BM, PC70BM and PC84BM, in comparison withthat of C60, C70 and C84. Inset show the carbon numbering. The energyshifts of the main peak with respect to corresponding naked fullerenes(denoted by δ) are -0.3, -0.2 and -0.3 eV for PC60BM, PC70BM andPC84BM, respectively. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
5.4 Comparison of C K-edge NEXAFS spectra from direct calculations (thickblue), the building block approach (thick red), and experiment (thickblack) [118]. The contributions from the side chain and fullerene backbonecomponents of the direct calculated spectra (thin blue) are also respec-tively compared with the spectra of isolated side chain and C60 (thinred). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
5.5 Theoretical C K-edge NEXAFS spectra of six PSCAmolecules from direc-t calculations (red), the conventional (blue) and modified (green) buildingblock approaches. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
6.1 Schematic representation of the modification of C6/Au (on the left) andC12/Au (on the right) by ionizing radiation. The branching of the in-dividual irradiation-induced processes is different for these two systems.Reprinted with permission from ref. [127]. Copyright c⃝ 2011 AmericanChemical Society . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
6.2 Structures of four aminothiolate molecules, AUDT, ABT, APBT andATPT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
6.3 Possible non-damaged and damaged species of AUDT under investigations. 45
6.4 Calculated N1s XPS of AUDT, ABT, APBT and ATPT (right) comparedwith the corresponding experimental results [134] (left). The bond lengths
marked on the molecules are given in A. . . . . . . . . . . . . . . . . . . 45
xvi LIST OF FIGURES
6.5 Calculated N 1s NEXAFS spectra of ATPT (right) and AUDT (left) andtheir derivatives compared with experiment [134]. . . . . . . . . . . . . . . 46
Contents
Abbreviations xi
List of Figures xiv
1 Introduction 1
2 Basic Quantum Chemical Methods 9
2.1 Born-Oppenheimer Approximation . . . . . . . . . . . . . . . . . 9
2.2 Hartree-Fock Method . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.3 Density Functional Theory . . . . . . . . . . . . . . . . . . . . . . 12
2.3.1 Thomas-Fermi model . . . . . . . . . . . . . . . . . . . . . 12
2.3.2 Hohenberg-Kohn Theorem . . . . . . . . . . . . . . . . . . 13
2.3.3 Kohn-Sham Equations . . . . . . . . . . . . . . . . . . . . 13
2.3.4 Exchange and Correlation Functionals . . . . . . . . . . . 14
2.4 London Dispersion Force . . . . . . . . . . . . . . . . . . . . . . . 15
3 Soft X-ray Spectroscopy 19
3.1 Basic theory and rules . . . . . . . . . . . . . . . . . . . . . . . . 20
3.1.1 Final state rule . . . . . . . . . . . . . . . . . . . . . . . . 20
3.1.2 Koopmans’ Theorem . . . . . . . . . . . . . . . . . . . . . 21
3.1.3 ∆ Kohn-Sham Approach . . . . . . . . . . . . . . . . . . . 21
3.1.4 Spectral Broadening . . . . . . . . . . . . . . . . . . . . . 22
3.2 X-ray Photoelectron Spectroscopy . . . . . . . . . . . . . . . . . . 22
3.3 Near-edge X-ray Absorption Fine Structure . . . . . . . . . . . . 23
xvii
xviii CONTENTS
4 Endohedral Fullerene 27
5 Fullerene-Based Organic Solar Cells 35
6 Self-Assembled Monolayer 41
7 Summary of included papers 47
7.1 Paper 1: Neutral and charged N@C60 . . . . . . . . . . . . . . . . 47
7.2 Paper 2: Structure of A@C60 . . . . . . . . . . . . . . . . . . . . . 47
7.3 Paper 3: Fullerene-based solar cell acceptors . . . . . . . . . . . . 48
7.4 Paper 4: Aminothiolates on SAMs . . . . . . . . . . . . . . . . . . 48
References 51
Chapter 1
Introduction
Functional molecular materials with specific electromagnetic, optical and biomed-
ical functionalities are the new synthetic materials that have made great contribu-
tions to the development of advanced technologies in the fields of energy, informa-
tion, health care, agriculture and military [1–3]. They have thus had a tremendous
impact on our daily life. There are huge numbers of functional molecular mate-
rials available nowadays and it is just impossible to cover all of them in a simple
thesis. Here I will focus on a few carbon based organic materials, in particularly
the fullerenes related systems.
In 1985, Kroto, Curl and Smalley discovered a new allotrope of carbon, the
spherical molecule C60 named as buckminsterfullerene or buckyball fullerene [4].
They were awarded the Nobel Prize in Chemistry in 1996 for this important
discovery [5]. As shown in Figure 1.1, the buckyball fullerene has a very high
symmetry of Ih and is composed of 12 pentagons and 20 hexagons. One can
immediately notice that the molecule looks just like a football, which is why the
name of football fullerene can occasionally be found in the literature. The special
structure of C60 leads to unique physical and chemical properties, such as good
solubility in aromatic solvents, higher reaction activity with metals and oxygen
and good electron accepting or donating ability [6], which have made them excellent
candidates for many applications [7]. Over the last decades, extensive studies on the
fullerene have been conducted and many remarkable new physical and chemical
properties have been revealed. The discovery of the fullerene arguably marks
the beginning of nano-science and nano-technology. It has certainly triggered
1
2 1 Introduction
decades long drama on hunting for new carbon based materials. The discoveries
of carbon nanotubes and graphene are two great highlights of the adventure. It
is worthy to mention that the Nobel Prize in Physics in 2010 went to Andre
Geim and Konstantin Novoselov “for groundbreaking experiments regarding the
two-dimensional material graphene”[5].
Figure 1.1 Geometry structure of fullerene C60.
Over the years, fullerenes of different size have been synthesized in the lab-
oratory. Very recently, they have even been found in the outer space [8]. The
vast space within the carbon cage of the fullerene has also provided opportuni-
ty to host another element, leading to the new class of synthetical molecules:
endohedral fullerene A@Cn. The first endohedral fullerene La@C60 was already
synthesized in 1985 [9] right after the discovery of C60. At present, a large number
of endohedral fullerenes have been made experimentally by different techniques
with the ion implantation, the arc-discharge of carbon rods, high pressure and high
temperature, respectively [10–14]. Nowadays, one has inserted metals, non-metals,
noble gases, ions and small molecules into the fullerene of different size. Figure
1.2 shows typical endohedral fullerenes inserted with atom and molecules.
The endohedral complexes have attracted broad attention not only for their
fascinating structures, but also excellent properties that are believed to be useful
for applications in biomedicine, energy, and electronics [15]. For instance, the com-
plex Gd@C82 has been successfully employed in drug delivery system and magnetic
resonance imaging (MRI) [16]. It has also been encapsulated into a single-wall car-
bon nanotube (SWNT) to form the (Gd@C82)n@SWNT complex that looks like a
peapod [17], see Figure 1.3. Inside the peapod, each Gd@C82 possesses a magnetic
3
(a) (b) (c)
Figure 1.2 Geometry structures of endohedral fullerenes (a) N@C60, (b) Sc3N@C80
and (c) Sc2C2@C84.
moment and acts like a tiny bar magnet. If one is able to change the orientation
of each magnet, it would be a perfect way to represent and store data as we do
now with a magnetic hard drive. In other ways, with such a peapod, it becomes
possible to manipulate data at the atomic level, something that is highly desir-
able for future Quantum Computing. In this context, the non-metal endohedral
fullerene N@C60 is also quite relevant, in which the encapsulated N atom remain-
s in its atomic spin state (S=3/2) [18] due to very weak interaction between the
carbon cage and the N atom. This guarantees a long electron spin relaxation for
the system, making the endohedral fullerene N@C60 a perfect element for future
quantum computer [19–21]. A possible quantum computer constructed by endohe-
dral fullerenes is schematically drawn in Figure 1.4. Maybe in the foreseeable
future, the quantum computer will replace the current computer systems and all
expensive calculations as presented in the thesis will be finished in seconds.
Figure 1.3 Illustration of two Gd@C82 endohedral metallofullerenes in a (11,9) SWNTwith schematic representation of the modulation of conduction and valence bands. Oneexpects to see the minute elastic strain around the metallofullerenes. Reprinted withpermission from ref. [17]. Copyright c⃝ 2003 Springer Berlin/Heidelberg.
4 1 Introduction
Figure 1.4 Scheme for a solid-state spin quantum computer based on linear chains of
endohedral fullerenes. Reprinted with permission from ref. [22]. Copyright c⃝ 2002 JohnWiley and Sons.
Another significant application of fullerenes is in the making of so-called poly-
mer solar cells (PSCs). The finding of new energy resources has become the major
international efforts in recent years. The simple fact that the annual energy input
of solar irradiation on Earth is several orders of magnitude higher than the yearly
energy consumption of the whole world and it is the cleanest energy resource avail-
able which have made the solar energy at the top of our wish list. Solar cell is the
basic element that can effectively covert the solar energy into the electricity. The
commercially available solar cells are made of inorganic materials, in particular
the highly purified silicon crystal. Compared to silicon-based devices, there are
several obvious advantages for using polymer based solar cells: inexpensive to fab-
ricate, lightweight, flexible, designable at the molecular level. On the social side,
polymer solar cells can have much less negative environmental impact than that
of inorganic devices. However, low efficiency and relative unstable under sunlight
are two major disadvantages of current polymer solar cells. Much of the studies
have been devoted to improve the performance of polymer solar cells, for which
the properties of organic materials play an important role.
Polymer solar cells are constructed by electron donor and electron accep-
tor materials, which mimic the semiconductor p-n junctions. The n-type organic
semiconductor fullerene and its derivatives are the widely used acceptor mate-
rials [23] due to the unique electronic structure of fullerenes. Figure 1.5 presents
the typical architecture of bulk heterojunction polymer solar cells (BHJ PSC-
s). In this device, three p-type conjugated polymers, such as Poly[2-methoxy-5-
(2-ethylhexyloxy)-1,4 -phenylenevinylene] (MEH-PPV), Poly[2-methoxy -5-(3’,7’
-dimethyloctyloxy)-1,4- phenylene vinylene] (MDMO-PPV) and poly(3- hexylth-
iophene) (P3HT), are commonly employed as the donor materials. In 1992, the
5
buckyball fullerene C60 was adopted as the acceptor and mixed with the MEH-
PPV to produce the polymer (organic) solar cell [24]. However, the poor solubility
of the C60 has severely limited its applicability in PSC. In 1995, the C60 derivative,
[6,6]-Phenyl-C61-butyric acid methyl ester (PC60BM), with excellent solubility was
used to blend with MEH-PPV to construct PSC with much better performance [25].
Since then, various PC60BM-like C60 derivatives and C70 or C84 derivatives have
been designed and synthesized to act as the acceptors [26–30]. Among all the donor
and acceptor materials, the P3HT blended with PC60BM-like C60 derivatives have
still been the most widely studied materials for PSCs in the last two decades. The
key to significantly improve the performance of PSCs is to carefully study the
electronic properties of the donor and acceptor materials with atomic resolutions.
Figure 1.5 Structures of (a) C60 and fullerene derivatives (b) PC60BM, (c)bisPC60BM, and (d) PC70BM. (e) Schematic draw of a bulk heterojunction polymersolar cell made of (PC60BM) fullerenes derivatives. Reprinted with permission fromref. [27]. Copyright c⃝ 2011 American Chemical Society.
For many applications, the molecular materials need to be placed on certain
surfaces, either metallic or semi-conducting. How to organize the molecular mate-
rials at the interface with good order is an important issue that has been addressed
by many studies. One way to achieve it is to make the so-called self-assembled
monolayer (SAM), which itself qualifies as an functional molecular material. It is
a kind of organic membrane formed by the surfactant molecules that are sponta-
neously adsorbed on the surface [2,31]. In 1983, Nuzzo et al. successfully produced
6 1 Introduction
the SAMs, gold-alkylthiolate monolayer, for the first time [32]. Since then, the
SAMs have been widely investigated and applied in many fields, owing to their
excellent properties, such as simple preparation conditions, highly ordered struc-
tures, and good stability [31]. Figure 1.6 shows one of the most studied SAMs,
alkanethiolates on gold surface [2]. It can be seen that in a typical SAM, the sur-
factant molecule often contains the head group, the spacer chain and the terminal
functional group. The head group, such as thiolate group, is boned strongly to
the gold surface that ensures the order of the molecular layer. The length of the
spacer chain controls the thickness of the monolayer, and the terminal functional
group can effectively interact with other systems. The choice of the functional
group sometime determines the application of the SAMs. For instance, the amine
terminal group is one of the most versatile functional groups for tethering the
biomolecules onto the substrates [33]. The well ordered SAMs can thus be used as
a good substrate to fix other molecules in the space through properly designed
terminal groups. They are certainly one of important forms for organic electronic
and photonic devices [2].
Figure 1.6 Schematic diagram of an ideal, single-crystalline SAM of alkanethiolatessupported on a gold surface with a (111) texture. The anatomy and characteristics ofthe SAM are highlighted. Reprinted with permission from ref. [2]. Copyright c⃝ 2005American Chemical Society.
To understand and to control the properties of functional molecular materials
require detailed information about their chemical and electronic structures. The
most commonly used experimental techniques for characterizing the molecular
materials are the soft X-ray spectroscopy, such as X-ray absorption spectroscopy
(XAS), X-ray photoelectron spectroscopy (XPS), and X-ray emission spectroscopy
7
(XES) [34–36], which are also the central focuses of this thesis. Apparently the elec-
tronic spin resonance (ESR) [37] and nuclear magnetic resonance (NMR) [38,39] have
also been employed for certain cases. Due to the large size and complicated struc-
ture of functional molecular materials, it is difficult, if not impossible, to assign
the experimental spectra based on physical and chemical intuition. Highly corre-
lated theoretical methods need to be applied to provide reliable assignments for
the spectra and to reveal underlying microscopic mechanisms. Over the years, the
density functional theory (DFT) has proven to be quite useful for many systems
and properties. In this thesis, we will show the successful applications of DFT
methods, but will also reveal the cases where conventional density functionals fail
to provide physical meaningful results.
This thesis consists of six more chapters. In Chapter 2, some fundamental
theories and concepts that are relevant to the research work presented in the thesis
will be provided. In Chapter 3, we will briefly describe the basic processes involved
in the soft X-ray spectroscopy and corresponding computational methods. The
following chapters, i.e. Chapters 4 and 5, will describe the theoretical modeling
of endohedral fullerenes and fullerene derivatives. The effects of X-ray damage
on aminothiolate SAM will be thoroughly examined in Chapter 6. Finally, a
summary of included papers will be presented in Chapter 7.
Chapter 2
Basic Quantum Chemical
Methods
All the works I have done for this thesis are based on the quantum mechanics.
The studies of molecular structures or molecular properties on the basis of quan-
tum mechanics are generally named as quantum chemistry. The fundamental of
quantum chemistry is in a large extent to solve the Schrodinger equation. In the
strict sense, it is impossible to solve the Schrodinger equation directly for a multi-
electron system. We thus need to adopt a variety of approximations in order to
solve practical problems. In this chapter, I will briefly describe the so-called first
principles theoretical methods that are related to the works included in the thesis.
2.1 Born-Oppenheimer Approximation
The non-relativistic Hamiltonian operator for a molecule can be written as the
sum of the kinetic energy T and the potential energy V as follows:
Htot = TN + Te + VNe + Vee + VNN . (2.1)
In this expression, N and e represent the nuclei and electrons, respectively. TN
is the nuclear kinetic energy operator, Te is the electronic kinetic energy operator,
VNe is the attraction operator between the nuclei and electrons, Vee is the repulsion
9
10 2 Basic Quantum Chemical Methods
operator between the electrons and VNN is the repulsion operator between the
nuclei. The Schrodinger equation for the multi-electron system can be written as:
HtotΨ(r, R) = EΨ(r, R), (2.2)
where Ψ is the wave function, E is the eigenenergy, r and R are the coordinates
of electrons and nuclei, respectively.
As the mass of the nucleus is much heavier than the electron. Even for the
lightest nucleus, the proton H weighs roughly 1800 times more than an electron. In
other words, the nucleus should move much slower than the electron, which gives
the good reason to study the motion of nuclei and electrons separately. In this case,
when studying the motion of the nuclei, the nuclei are considered to move slowly in
the uniform electronic field. Whereas when studying the motion of electrons, the
nuclei can be assumed to be fixed in space and the electrons are moving in the field
generated by the nuclei. Such a consideration is the famous Born-Oppenheimer
(BO) approximation that is very important for effectively solving the Schrodinger
equation of a multi-electron system.
With this approximation, when the positions of nuclei are fixed, the nuclear
kinetic energy operator TN in Eq.(2.1) is zero and the repulsion between the
nuclei VNN in Eq.(2.1) can be treated as a constant. So the Schrodinger equation,
Eq.(2.2), becomes:
(He + VNN)Ψe = UΨe, (2.3)
where He is the electronic Hamiltonian operator, Ψe is the electronic wave func-
tion, U is the eigenenergy including the electronic energy and the repulsion be-
tween the nuclei. The constant VNN in the Hamiltonian expression will not influ-
ence the electronic wave function, so the electronic Schrodinger equation can be
simplified as follows:
HeΨe = EeΨe. (2.4)
Ee is the electronic eigenenergy, and U = Ee+VNN . As above mentioned, the
electronic Hamiltonian includes the VNe, so the nuclear configuration will affect
2.2 Hartree-Fock Method 11
the electronic wave functions. Furthermore, the electronic energy depends on the
nuclear configuration. The different electronic energies corresponding to different
nuclear configurations construct the so-called potential energy surface (PES).
2.2 Hartree-Fock Method
One of the simplest approach to solve the electronic Schrodinger Eq.(2.4) is the
so-called Hartree-Fock (HF) method by assuming that an electron is only affected
by the averaged field generated from the rest electrons. HF method uses a Slater
determinant as the N-electron wave function and the Slater determinant can be
written as [40]:
Ψ(r1, · · · , rN) =1√N !
∣∣∣∣∣∣∣∣∣∣∣
ψ1(r1) ψ2(r1) · · · ψN(r1)
ψ1(r2) ψ2(r2) · · · ψN(r2)...
.... . .
...
ψ1(rN) ψ2(rN) · · · ψN(rN)
∣∣∣∣∣∣∣∣∣∣∣(2.5)
here, ψk(ri) is an one-electron wave function, and ri is the ith electron’s coordinate.
Then we can solve the Eq.(2.4) in HF form:
Fiψk(ri) = ϵkψk(ri), (2.6)
where Fi is the Fock operator which can be expressed as follows:
Fi = hi +N∑j
(Jij − Kij), (2.7)
where hi is the single electron Hamiltonian operator of ith electron which includes
the single electron kinetic energy and the attraction energy between the nuclei
and electron, Jij is the coulomb operator which describes the repulsion between
the electrons i and j, Kij is the exchange operator which describes the exchange
energy between the electrons i and j. By using a so-called self-consistent field
(SCF) method, the solution of the Eq.(2.6) namely the energy of the molecular
orbital (MO) can be calculated:
12 2 Basic Quantum Chemical Methods
εk = ⟨ϕk|F |ϕk⟩ = ⟨ϕk|h|ϕk⟩+N∑j
(⟨ϕk|Jj|ϕk⟩ − ⟨ϕk|Kj|ϕk⟩). (2.8)
The total energy of the system can be given by [41]:
E0 =N∑k
εk −1
2
N∑k
N∑j
(⟨ϕk|Jj|ϕk⟩ − ⟨ϕk|Kj|ϕk⟩). (2.9)
Obviously, the HF method has its limitations as it does not contain the cor-
relation energy between the electrons [42]. To consider the electronic correlation
energy, many post-HF methods have been developed to get more accurate results.
Among them, one can mention a few, for instance, the Configuration Interaction
(CI) [43,44] method, Coupled Cluster (CC) [45–47] method, and Møller-Plesset pertur-
bation theory (MPn)[41]. One should be aware that these methods require much
more computational resources and are often limited for relatively small systems.
For practical use, in particular for the functional molecular materials studied in
this thesis, one has to find alternative ways than these high cost methods. In this
context, density functional theory (DFT) becomes really relevant.
2.3 Density Functional Theory
2.3.1 Thomas-Fermi model
In 1927, one year after the Schrodinger equation was published, Thomas and Fermi
[48,49] proposed the so-called Thomas-Fermi model that was based on the electron
density, rather than the wavefunction. It was suggested that the energy of the
system can be described as a function related to the electron density ρ(r). The
Thomas-Fermi model is a rough model with the analytic expression of the kinetic
energy without considering the electronic exchange and correlation energy. In or-
der to consider the electronic exchange and correlation effects, one of the simplest
solutions is to add corrections into the energy expression. In 1964, Hohenberg and
Kohn created the modern DFT on the basis of the Thomas-Fermi model [50].
2.3 Density Functional Theory 13
2.3.2 Hohenberg-Kohn Theorem
The two Hohenberg-Kohn (HK) theorems are the rigorous theoretical basis of
modern DFT. The first HK theorem points out that when the N electrons system
in the external potential Vext, the Vext is uniquely determined by the electron
density. The Hamiltonian of the system:
H = Te + VNe + Vee + VNN (2.10)
is determined only by ρ(r). The system’s energy is also the functional of the
density, which can be written as:
E = E[ρ]. (2.11)
The second HK theorem states that for a given external potential, the ground
state energy of a system can be obtained from the energy functional if and only
if using the true ground state density. Therefore, the ground state energy can
be obtained by a variational calculation of the energy functional to the density
function.
2.3.3 Kohn-Sham Equations
According to the HK theorems, the energy functional of an isolated molecular
system can be written as:
E[ρ] = T [ρ] + VNe[ρ] + Vee[ρ], (2.12)
where T [ρ] is the electronic kinetic energy, VNe[ρ] is the attraction between the
nuclei and electrons, Vee[ρ] is the repulsion between the electrons. Kohn and Sham
proposed: it can be assumed that one can use the non-interacting kinetic energy
functional Ts[ρ], which has the same electron density distribution ρs(r) with the
real molecular system ρ(r), to replace the electronic kinetic energy T [ρ]. Then the
energy functional of the non-interacting system can be described as:
Es[ρ] = Ts[ρ] + Veff [ρ], (2.13)
14 2 Basic Quantum Chemical Methods
where Ts[ρ] is the kinetic energy functional of the non-interacting system, Veff [ρ] is
an effective potential functional. Considering that ρs(r) = ρ(r) and Es[ρ] = E[ρ],
if combining the Eq.(2.12) and Eq.(2.13), we can get:
Veff [ρ] = VNe[ρ] + Vee[ρ] + (T [ρ]− Ts[ρ]). (2.14)
The effective potential could be written more specifically,
Veff (r) = VNe(r) +
∫ρ(r′)
|r − r′|dr′ + Vxc(r), (2.15)
where the second part is the electron-electron Coulomb repulsion term, the third
part is called exchange and correlation functional. By solving the Schrodinger
equation of this one-electron system,
[−1
2∇2
i + Veff (r)]ψi(r) = ϵiψi(r), (2.16)
we can get the orbital ψi(r) and the corresponding density ρ(r):
ρ(r) = ρs(r) =N∑i=1
|ψi(r)|2. (2.17)
These above equations are called Kohn-Sham (KS) equations. In 1998, Kohn
was awarded to share the Nobel Prize in Chemistry for his contributions of DFT [5].
2.3.4 Exchange and Correlation Functionals
The most important problem in DFT approach is that the exact exchange and
correlation functional of the system is unknown. The way to deal with this issue
is to develop some efficient exchange and correlation functionals by the approx-
imations. Local density approximation (LDA) is widely used in practice. The
exchange and correlation functional in LDA is written as
ELDAxc [ρ] =
∫ρ(r)εxc(ρ(r))dr, (2.18)
2.4 London Dispersion Force 15
where the εxc is the exchange-correlation energy density under the homogeneous
electron gas approximation. When the density functionals are applied in spin-
polarized systems, the exchange and correlation functional in the local spin density
approximation (LSDA) is
ELSDAxc [ρα, ρβ] =
∫ρ(r)εxc(ρα(r), ρβ(r))dr. (2.19)
The other widely applied approximation is the generalized gradient approxi-
mations (GGA), in which the gradient of the density is taken into account. The
exchange and correlation functional in GGA is
EGGAxc [ρα, ρβ] =
∫εxc(ρα(r), ρβ(r);∇ρα(r)∇ρβ(r))dr. (2.20)
There is a common way to construct the functional by combining the different
exchange and correlation terms together, called the hybrid functional. The most
popular hybrid functional is B3LPY [51] functional which combined B3 parameter
exchange functional and LYP correlation functional:
EB3LY Pxc = ELDA
xc +0.2(EHFx −ELDA
x )+0.72(EB88x −ELDA
x )+0.81(ELY Pc −ELDA
c ).
(2.21)
2.4 London Dispersion Force
London dispersion force [52,53] (LDF, also named as dispersion force, London force,
temporary dipole-induced dipole force) is named after the German-American physi-
cist Fritz London. LDF is a kind of force between the atoms and molecules and it
is a weak force belonging to the more general van der Waals force. LDF is induced
by the instantaneous polarization multipoles of molecules. The constant motion
of the electrons will cause the unsymmetrical electronic distribution around the
nucleus, as the system A (an atom or a molecule) shown in Figure 2.1. If the
system B is close to the system A, the electronic distribution will be distorted by
the appearance of the dipole in the system A, because of the repelling between
the electrons. Therefore, the instantaneous dipoles will attract each other and it
is the formation of London dispersion force presented between these two systems.
16 2 Basic Quantum Chemical Methods
δ+
δ-
δ+
δ-
nucleus electron
A B
Figure 2.1 Schematized London dispersion force interaction between two moleculeswith unsymmetrically distributed electron cloud.
London forces are present between all molecules, including the polar and
non-polar molecules. It is the only intermolecular force existing between neutral
atoms. There will be no attractive force between the noble gas atoms without the
London dispersion force. It should be stressed that such an interaction between
the nobel gas atoms has made it possible for the nobel gas to exist in the liquid
form.
The statement of the attractive interaction between noble gas atoms was
clarified by Fritz London in 1930 [54–56], and it was dealt with as a kind of repre-
sentative of the dispersion interaction in several decades. Nowadays, the DFT is
widely used as the most popular methodology, many developers of the functionals
are trying to solve the long-range London dispersion interaction by means of DFT
method. It is known that most of the DFT methods have difficulties in describing
this kind of interaction [57–60] because they can not provide −C6/R6 dependence
of the dispersion energy at the interatomic distance R. In the past decade, many
new density functionals were developed in order to resolve this issue. The dis-
persion correction terms are added into the density functionals to estimate this
weak interaction, for instance, the SSB-D functional developed by Swart et. al [61],
the DFT-D developed by Grimme et. al [62–64], M06 class developed by Truhlar et.
al [65] and B3LYP-dispersion correcting potential (DCP) of DiLabio’s group [66]. So
far, a lot of functionals have successfully described the dispersion interaction in
many systems. However, for some special cases, the availability and the reliability
of the functionals are still under debate.
One representative system with the non-covalent interaction is the parallel
benzene dimer. The potential energy surface of the parallel benzene dimer ob-
tained by the different methods, including HF, CCSD(T) and some DFT methods,
2.4 London Dispersion Force 17
are shown in Figure 2.2. The CCSD(T) method could predict the accurate dis-
persion binding between two benzene molecules. One can see that HF and some
conventional DFT methods do not predict the correct dispersion interaction, and
the dispersion corrected density functionals work well.
Figure 2.2 Parallel benzene dimer (D6h symmetry) PESs generated using different
methods. Reprinted with permission from ref. [67]. Copyright c⃝ 2009 John Wiley andSons.
Chapter 3
Soft X-ray Spectroscopy
X-rays, or called Roentgen rays were discovered in 1895 byWilhelm Conrad Roent-
gen, which are part of electromagnetic spectrum whose energies are in the region
of one hundred to ten thousands of eV. Commonly, the X-ray is divided into two
parts based on its energy, namely the soft X-ray with lower energy (102 − 104
eV) and the hard X-ray with higher energy (104 − 105 eV). The X-ray in different
energy region can find its own application areas. The soft X-ray, whose energy
coincides with the electronic transition energies of most elements, is widely used
to study the electronic structure and the chemical compositions of a variety of
materials, particularly the organic functional molecular materials. The reason is
simple since the basic elements of a functional molecular material are carbon, ni-
trogen and oxygen atoms, for which the binding energies (BEs) of the core orbital
are in the soft X-ray energy region around 290eV, 400eV and 540eV, respectively.
When the X-ray interacts with the matter, several excitation and emission
processes can take place, which correspond to different spectroscopies. In this the-
sis, we focus on three commonly used spectroscopies, namely X-ray photoelectron
spectroscopy (XPS), X-ray absorption spectroscopy (XAS) and X-ray emission
spectroscopy (XES). Among them, XAS and XPS are resulted from the excita-
tion processes. In this case, the photon with the energy ~ω is absorbed by the
molecule, and the electron in the core orbital will be excited to the unoccupied
orbitals or even ejected to the vacuum level, i.e. ionized. After the photon absorp-
tion and the core electron excitation, the molecule is in an intermediate excited
state that has a short lifetime and ready to decay. The main decay channel is
19
20 3 Soft X-ray Spectroscopy
a non-radiative one which causes the Auger electron emission. The other decay
channel is radiative one that gives the XES, in which the electron in valence orbital
goes down into the core hole and one photon with the energy ~ω′ is simultaneously
emitted. The X-ray processes related are illustrated in Figure 3.1.
In this chapter, the computational methods for XPS and XAS will be briefly
summarized. The basic theory and rules associated with those computational
approaches are also presented.
(c)
(d)
e-
c
v
u
ħω
XPS
c
v
u
ħω
NEXAFS
c
v
u
XES
c
v
u
RXES
(a)
(b)
ħω'
ħω'
Figure 3.1 Schematic representation of the absorption processes, a: XPS, b: NEXAFSand emission processes, c: XES, d: RXES. The occupied and unoccupied orbitals arefigured by solid and dashed lines. c, v and u denote the core, valance, unoccupiedorbitals, ~ω and ~ω′ are energies of incident and emission photon, respectively.
3.1 Basic theory and rules
3.1.1 Final state rule
There is an important rule that has often been considered in the calculations of the
K-edge (1s orbital) X-ray absorption and emission spectra, namely the final state
rule, which was developed for calculating the X-ray emission spectra of metals by
3.1 Basic theory and rules 21
von Barth and Grossman [68]. According to this rule, the accurate absorption and
emission spectra could be obtained by the final state wave functions without the
knowledge of the initial state of the corresponding process. In other words, the
core excited state and the ground state are adequate for computing the absorption
and emission spectra, respectively [68,69]. This rule is practically applicable because
a 1s orbital is often well separated from the other molecular orbitals and hence,
there is no much difference in their behaviors between the initial and final states.
3.1.2 Koopmans’ Theorem
Within the HF framework (see Section 2.2), the total energies of a N -electron
system in its ground state as well as the state with one electron removed from
orbital m are given respectively by,
EN =N∑i
εi −1
2
N∑ij
(Jij −Kij) + Vnn,
EmN−1 =
N−1∑i
εi −1
2
N−1∑ij
(Jij −Kij) + Vnn.
(3.1)
Corresponding total energy difference, i.e., the ionization potential (IP), is simply
estimated as the negative value of an orbital energy εm:
IPm = EmN−1 − EN = −hm − 1
2
N∑i
(Jmi −Kmi) = −εm. (3.2)
This is often known as the Koopmans’ Theorem, which provides a simple way
to calculate the IP. Obviously this theorem assumes no decay after removing an
electron,i.e., all passive orbitals are kept frozen during the ionization process.
3.1.3 ∆ Kohn-Sham Approach
In reality, when a core-electron is ionized, a hole is left in the core orbital. This
results in an excited state that is very unstable and the energy needs to be lowered
through electronic relaxation. The so-called ∆ Kohn-Sham (∆KS) approach [70,71]
22 3 Soft X-ray Spectroscopy
considers the relation which therefore predicts more accurate result than the Koop-
mans’ Theorem. Mathematically, the IP of a 1s electron is given by [70],
IP1s = Eopt|n1s=0 − Eopt|n1s=1, (3.3)
where n1s denotes the occupation number of the 1s. And the state with n1s = 0
is often termed as the full core hole (FCH) state.
3.1.4 Spectral Broadening
Actually, what we obtain directly from computer programs are stick spectra con-
sisting of a set of transition energies and a corresponding set of intensities. For
many thermal, optical or mechanical reasons, the calculated stick spectrum should
be broadened. Various broadening techniques are often employed to simply reflect
these above-mentioned effects. Two kinds of distribution functions that are widely
used in literatures are the Gaussian and Lorentzian broadening functions given
by,
ΛGaussian(ω, ωif ,Γ) =1
η√2πexp[−(ω − ωif )
2
2η2], η =
Γ√2ln2
, (3.4)
ΛLorentzian(ω, ωif ,Γ) =1
π
Γ
(ω − ωif )2 − Γ2. (3.5)
Here Γ represents the half-width at half-maximum (HWHM). By using the broad-
ening functions, each transition i → f is broadened centering at ωif . Then,
each transition intensity is multiplied to the distribution function, and summa-
tion at the same energy leads to the resolved spectrum. Finally, summation over
the atom-specific spectrum at each non-equivalent center (of the same element),
weighted by their relative abundance, leads to the final total spectrum which can
be compared with experiment. In all included works of this thesis, the Gaussian
line shape is chosen since it is more widely used in literatures.
3.2 X-ray Photoelectron Spectroscopy
X-ray photoelectron spectroscopy is a useful technique used in the field of the
surface science. When the X-ray is shined on matter, a core-level electron is
3.3 Near-edge X-ray Absorption Fine Structure 23
emitted and one gets the XPS. The XPS records the energy distribution of the
emitted core electrons. The energy conservation for the photoemission process is
expressed as
E~ω = Ekin + Eϕ + EB(i), (3.6)
where E~ω denotes the X-ray photon energy, Ekin represents the kinetic energy
of the emitted photoelectron, Eϕ stands for the work function, and EB(i) is the
binding energy of the ith orbital (i.e. IPi). Experimentally, the photon energy of
the X-ray source is known, and the work function is also known and often dealt as
a constant, the kinetic energy of the emitted electrons is recorded by the electron-
energy analyzer and hence, the BE of ith orbital can be calculated from Eq.(3.6).
On the other hand, the theoretical BE of the ith orbital can be approximately cal-
culated by Koopmans’ Theorem (Section 3.1.2) or more accurately from the ∆KS
approach (Section 3.1.3). The XPS spectra are finally evaluated by broadening
the different IP bars.
One of the important features of XPS is to get the chemical shift of the
binding energy, which reveals the chemical environment of the atom. From the
chemical shift, the chemical composition of sample, oxidation state of element,
etc., can be figured out. The practical usage of XPS can be dated to 1960s when
Kai Siegbahn and coworkers [72] showed the chemical-structure dependence of the
C K edge binding energy. Although it is an old-fashioned technique, it is still
widely applied nowadays in many research fields. Since XPS often focuses on the
core orbitals, it reflects more of the atomic feature; while the X-ray absorption or
emission spectroscopy considers more molecular electronic structure respectively
in the virtual or valence levels.
3.3 Near-edge X-ray Absorption Fine Structure
With respect to the X-ray absorption, in this thesis, only the near-edge X-ray ab-
sorption fine structure (NEXAFS) [also named as the X-ray absorption near-edge
structure (XANES)] spectrum is studied. The NEXAFS spectroscopic technique
was developed since the last century in detecting the structures of molecules on
surfaces [73]. The NEXAFS spectrum mainly probes the electronic structure of
24 3 Soft X-ray Spectroscopy
the unoccupied orbitals. Experimentally, the NEXAFS spectrum is obtained via
the relaxation process. As is mentioned in the beginning of this chapter, there
is a non-radiative relaxation channel after the core excitation, which is the ma-
jor relaxation channel and results in the Auger-electron emission. The NEXAFS
spectrum can thus be measured via the Auger-electron yield. Alternatively, one
can also measure the spectra via the fluorescence channel.
On the other hand, when the NEXAFS spectrum is computed theoretically,
the interaction between the X-ray and the electron could be handled as a pertur-
bation. By applying the time-dependent perturbation theory and Fermi’s golden
rule, the transition rate could be calculated, and consequently the absorption cross
section and the oscillator strength could be obtained. The final working expression
for the oscillator strength of X-ray absorption is given by,
Iif =2mωif
~|ϵ · ⟨ψf |r|ψi⟩|2, (3.7)
where the ⟨ψf | and |ψi⟩ denote the final and initial MOs involved in the transition,
ωif is the energy difference between two MOs, and ϵ is the electronic polarization
vector. Here the dipole approximation, single-particle approximation and final-
state rule are taken into account and detailed derivation process can be found
elsewhere [74].
One feature of the NEXAFS spectroscopy is the polarization dependence,
which is usually employed to detect the orientation of ordered molecules. From
Eq.(3.7), one can see that the dipole transition matrix is dependent on the orien-
tation with respect to ϵ. When the molecules are randomly oriented, for instance,
in the gas phase or in bulk solution, the statistical effect should be taken into
account. As a simple estimation, one can average over the x, y, and z directions
and obtains,
Iif =2mωif
~1
3(|⟨ψf |x|ψi⟩|2 + |⟨ψf |y|ψi⟩|2 + |⟨ψf |z|ψi⟩|2). (3.8)
When molecules are well oriented, the polarization dependence of the NEXAFS
spectroscopy could show its effect from the angular-dependent detections. For
example, for a graphene positioned in the x − y plane, the π∗ orbital is parallel
with the z direction, while the σ∗ orbital is parallel to the x − y plane. Hence,
3.3 Near-edge X-ray Absorption Fine Structure 25
the oscillator strength of the 1s to π∗ and σ∗ transitions could be respectively
calculated as [75]
Iσ∗
if =2mωif
~1
2(|⟨ψf |x|ψi⟩|2 + |⟨ψf |y|ψi⟩|2), (3.9)
Iπ∗
if =2mωif
~|⟨ψf |z|ψi⟩|2. (3.10)
Here for the σ∗ transition, an average over the x and y directions is employed to
simulate the random orientation in the x− y plane.
The accurate excited energies could be calculated by the ∆KS scheme. Such
a procedure is evidently time-consuming for moderate or large systems with lots
of unoccupied orbitals. It is quite common that a particular transition, the one
from 1s to the lowest unoccupied molecular orbital (LUMO), is computed. The
corresponding transition energy is calculated as
ε∆KS1s→LUMO = NE1s→LUMO − NEGS. (3.11)
The shift value for calibration could be calculated by ϖ = εcalc.1s→LUMO− ε∆KS1s→LUMO.
Consequently, the computed spectra are uniformly shifted by subtracting this
value ϖ. An alternative choice is to calibrate the theoretical spectrum directly to
experiment. This often is used in cases that an accurate energy is too expensive
to achieve.
Chapter 4
Endohedral Fullerene
A fullerene molecule is a carbon cage consisting of pentagons and hexagons. Its
internal cavity structure thus makes it possible to accommodate other elements,
such as ions, atoms and small molecules in its inner sphere. This kind of structure
is named as endohedral fullerene, denoted as A@Cn, where A represents the en-
dohedral atom or small molecules, and n stands for the number of carbons of the
cage. At present, many different types of endohedral fullerenes with noble gases,
ions, non-metals, metals, metal nitrides, hydrides and small molecules have been
synthesized and characterized both experimentally and theoretically. It seems like
that there is no limit to the fascinating development of the field and one can al-
ways expect more beautiful structures to come in the future. In Chapter 1, the
nice structures of three endohedral fullerenes were shown. In this chapter, we will
discuss more about the history and characters of the endohedral fullerenes.
Right after the discovery of the buckyball fullerene C60 in 1985, Curl, Kroto
and Smalley immediately realized that it should be possible to place elements
into the cavity of the carbon cage. Just a few days later, the first endohedral
fullerene, La@C60, was successfully synthesized [9]. However, the extraction of
the metallofullerene La@C82 was not possible until 1991 [76]. With arc-discharge
method, one can now synthesize many different endohedral metallofullerenes with
alkaline earths and transition metals, for instance, Sc@C82, Ce@C82, Pr@C82 and
Gd@C82[77]. It is worth noting that most metals in endohedral metallofullerenes
are concentrated on the second and third family elements of Periodic Table of El-
ements, or some lanthanides and actinide elements. The commonly used fullerene
27
28 4 Endohedral Fullerene
cages are C60, C80, C82 and C84. The endohedral metallofullerenes combine the
properties of both endohedral metal atoms and the buckyball fullerene, which
make them more versatile for applications in even broader areas [15].
The exact location of the endohedral metal atom inside a metallofullerene is
not easy to determine even experimentally due to the lack of proper techniques.
However, the general consensus is that the metal atom does not occupy the center
position of the buckyball fullerene and the carbon cage appears to be slightly
deformed [78]. In other words, the interaction between the metal atom and the
carbon cage can be fairly strong. Theoretical modeling is of course a valuable
tool in determining the equilibrium geometry of the metallofullerene. Figure 4.1
gives the theoretical PESs of the endohedral La atom moving inside the fullerene
C82, which clearly demonstrates that the location of the La atom is on the off-
center position, with a relatively shallow potential over a visible region [79]. One
can expect that under the room temperature, the La atom could oscillate around
its equilibrium. It is noted that for certain metallofullerenes, the location of the
endohedral metal atom is much dependent on the density functional used for the
calculations [80].
Figure 4.1 The energy contours of two σv planes of La@C82 (C2v) (energy in kcalmol−1), only the region of which the energy is less than 5 kcal mol−1 is shown. Reprintedwith permission from ref. [79]. Copyright c⃝ 2007 American Chemical Society.
It is really impressive to see that experimentalists can now put not only one
metal atom, but also multi-metal atoms inside the carbon cage. We can mention
here a few of examples, like the di-metals endohedral fullerenes (Sc2@C82, Y2@C82,
La2@C82 and Sc2@C84), three metals endohedral fullerenes (Sc3@C82), and four
metals endohedral fullerene Sc4@C82[77]. In 1999, the discovery of the endohe-
dral nitride cluster fullerene Sc3N@C80 opens the new chapter of the endohedral
fullerenes [81]. Similar compounds with different metal atoms, such as Y3N@C80
29
and ErSc3N@C80, have also been synthesized [82,83]. In general, there are about five
different kinds of clusters that have been inserted into the carbon cage, namely
metal nitrides (M3N), metal carbide (M2C2), metal oxide (M4O2), metal sulfide
(M3S) and metal hydrocarbon (M3CH)[81,83–88]. These new types of endohedral
mixed metal cluster fullerenes are of high yield and high thermal stability. They
are considered to be good functional materials for biomedical applications [89].
It seems like that it is more difficult to insert non-metal elements into the
carbon cages than the metal atoms. The first non-metal endohedral fullerenes
synthesized in 1993 were noble gas endohedral fullerenes He@C60 and Ne@C60[90].
Up to now, there are wide range of noble gas endohedral fullerenes available,
such as Hem@Cn, Nem@Cn, Arm@Cn, Krm@Cn and Xem@Cn, by high pressure
experimental method [91–93]. The existence of the noble gas in the carbon cage
has often been verified by means of the NMR of the corresponding noble gases
atom [94]. As shown in the Paper 2 of this thesis, it is quite surprise to see that
none of conventional density functional theory is able to correctly describe the
equilibrium structure of all these endohedral fullerenes. It seems to be a great
challenge for current density functionals to accurately describe the very weak
interaction between the nobel gas atom and the carbon cage.
The synthesis of N@C60 in 1996 led to the creation of a new type of endohedral
fullerenes with the reactive non-metal elements, such as N, O, P and S atoms [96,97].
Among them, the endohedral fullerene N@C60 has been one of the most studied
system. It is mainly due to its unique magnetic property. It was found that
the N atom keeps its atomic spin state (S=3/2) after enclosed into the fullerene
C60[18,95]. Such a property makes it a perfect candidate for quantum computer
working with the electron spin [19]. The future prospect with quantum computer
is extremely bright. A simple estimation has shown that a task that will take
one million years to finish with current computer can be completed in seconds by
quantum computer [89]. Figure 4.2 illustrates the ion implantation technique that
has been used to produce the endohedral fullerene N@C60 in the experimental
laboratory. This gives a flavor of how complicated the experimental setup is.
It has also been found experimentally in recent years that under the high
temperature, around about 500 K, the endohedral fullerene N@C60 becomes un-
stable and the N atom can escape out from the carbon cage [98,99]. The underlying
30 4 Endohedral Fullerene
(a)
(b)
Figure 4.2 Schematic view of the experimental set-up for the production of N@C60.(a) By simultaneous deposition of C60 on a substrate and irradiation with nitrogenions. (b) By using the glow discharge for the production of the ions. Reprinted withpermission from ref. [95]. Copyright c⃝ 1998 Springer Berlin/Heidelberg.
mechanism has also been examined by different theoretical calculations with low
accuracy [100–103]. One of my projects was to obtain the most accurate PESs to
describe the escaping pathways of the N atom in and out of the carbon cage. In
the following, some important results from this study will be recaptured.
From geometrical point of view, there are five possible escaping pathways for
the N atom to get out of the C60 cage, as shown in Figure 4.3(a). It starts from
the center of the cavity toward (1) the center of the five-membered ring (5MR),
(2) the center of the six-membered ring (6MR), (3) the center of the C=C bond
that connects two six-membered rings (6-6 Bond), (4) the center of the C-C bond
connecting the five and six membered rings (5-6 Bond) and (5) one of the carbon
atoms (C atom). It can be clearly seen from the calculated PESs illustrated in
Figure 4.3(b) that the N atom prefers to move along the pathways (6-6 Bond)
and (5-6 Bond). Or in other words, the pathways toward the center of C=C/C-C
bond are energetically favorable.
It is also very important to understand how different spin states are evolved
31
(Å)
(a) (b)
A: 5MR B: 6MR C: 6-6 Bond
D: 5-6 Bond E: C Atom
Figure 4.3 (a) Five possible pathways for N atom to escape from N@C60. (b) Potential
energy surfaces for S=3/2 state within the distance region of 0.0 A to 3.5 A from thecenter of the cage.
along the PES. For the favorable (6-6 Bond) pathway, the PESs for the spin states
S=3/2 and 1/2 are plotted in Figure 4.4(a). It can be seen that within the region of
0 to 1.7 A, the total energy of the spin state S=3/2 is much lower than that of the
spin state S=1/2, which indicates that the N atom keeps its atomic state around
the center of the cavity and its interaction with the carbon cage is quite weak.
As expected when the N atom moves close to the carbon cage, the interaction
between them gets stronger and the spin state S=1/2 becomes the energetically
favorable. The energy barrier for the N atom to move out of the cage is found to be
as high as 2.85 eV, which can hardly be overcome by thermal effects alone. Based
on our calculated energy profiles, it is just impossible for the N atom to escape
from the carbon cage even at 500 K. It should be noted that the calculations were
done for a single molecule, whereas the experiment was conducted for molecules
in condensed phases. The intermolecular interaction might drastically change the
picture, however, with current computational capacity, it is still not possible for
us to calculate the consequence of such an effect on the PES.
One of the most surprising observations from the calculations is the shape of
the PES around the center of the cavity. Figure 4.4 (b) shows the PESs in the
region of -0.5 A to 0.5 A for the spin state S=3/2 obtained from two different
density functionals, B3LYP and M06-2X, respectively. The PES from B3LYP cal-
32 4 Endohedral Fullerene
(Å)(Å)
(a) (b)
Figure 4.4 (a) Potential energy surfaces of the N moves along the reaction pathwaystowards the center of the 6-6 Bond inside neutral N@C60 cage at different spin states.(b) Potential energy surfaces obtained from B3LYP and M06-2X functionals both atS=3/2 around the center of the carbon cage.
culations shows a very peculiar behavior, the total energy for the N atom at the
center of the carbon cage is not a globe minimum, but a local maximum. When
the M06-2X functional is applied, the PES becomes normal and the complex with
the N atom at the center is a true minimum. The dispersion corrected density
functional M06-2X thus gives more reasonable result for this system. It highlights
the importance of the dispersion force in determining the equilibrium structure
of N@C60 due to the weak interaction between the N atom and the carbon cage.
One can expect that such a situation should be common for all the endohedral
fullerenes with the reactive non-metal and nobel gases atoms. We have studied
a series of endohedral fullerenes, A@C60 (A=H, C, N, O, S, P, He, Ne, Ar, Kr),
by applying the commonly used functionals, such as BLYP [51,104,105], BP86 [105,106]
and B3LYP [51], as well as the dispersion corrected functionals, B3LYP-DCP [66],
B97-D [63] and M06-2X [65]. It is quick shock to see that none of these functionals
can provide physically meaningful results for the PESs of all endohedral fullerenes.
The inclusion of the dispersion force does give negative binding energy that sta-
bilizes the system, but it does not automatically solve the problem related to the
PES.
Figure 4.5 displays the PESs along the 6-6 Bond pathway in the region of -0.5
to 0.5 A inside of the A@C60 (A = Kr, Ne) calculated by B3LYP-DCP, B97-D and
M06-2X, respectively for the singlet spin state (S=0). It is reasonable to observe
that all functionals give consistent PES for the endohedral complex Kr@C60 in
33
(Å)(Å)
Kr@C60
(a) (b)
Figure 4.5 Potential energy surfaces obtained at three different density functionals
towards the center of the 6-6 Bond from -0.5 A to 0.5 A for (a) Kr@C60 and (b) Ne@C60,respectively.
Figure 4.5(a). In this case, the PES is smooth and the center point corresponds
to the energy minimum. Whereas for the PESs of the Ne@C60 in Figure 4.5
(b), the situation seems to be out of control. Three functionals give completely
different PESs and none of them makes any physical sense. At this stage, we do
not know what is the main reason behind such difference. It is also difficult to
understand why Kr@C60 should differ so much from Ne@C60 in the eyes of these
functionals. Normally the parameterizations of dispersion corrected functionals
are based on the study of π-π interacting systems, like benzene dimers. In the
case of the endohedral fullerenes A@C60, the interaction is between a single point
and a spherical π conjugated system, which is very different from that of benzene
dimers. Maybe the endohedral fullerenes A@C60 are the perfect candidates for
the fine tuning of the dispersion corrected functionals.
Chapter 5
Fullerene-Based Organic Solar
Cells
Solar energy is a kind of clean and renewable energy source. To resolve the energy
crisis and environment problems, in recent years, much research and industrial
efforts have been paid on how to harvest the solar energy, usually by converting it
into the electric energy. Historically, the Bell institute successfully produced the
silicon solar cells in 1954, which opened a new chapter in photoelectric conver-
sion [107]. In the past few decades, this kind of traditional, inorganic solar cell has
been well developed and the power conversion efficiency has reached over 20%.
Moreover, commercialization leads to their extensive use in our daily life, aviation
and aerospace industries, and so on. However, there are still many disadvantages
for the traditional solar cells which limit their further applications, for instance,
high cost, heavy weight, high energy consumption in the process of production,
and accompanying environmental pollution, just name a few. It is therefore neces-
sary to look for inexpensive, new solar cell materials which can extend applications
of the solar energy. Organic semiconductor materials represent a kind of cheap
candidate. In comparison with the traditional silicon-based solar cells, organic
ones have more advantages, including the low cost, easy fabrication and soft fea-
tures, which lead to a trend that they will replace the inorganic solar cells in
the future [108]. The earliest report on organic solar cells dated back to 1959 by
Kallmann and coworkers [109]. In 1986, the concept of electronic donor (p-type) -
electronic acceptor (n-type) was first proposed by Tang et al. [110]. Organic solar
35
36 5 Fullerene-Based Organic Solar Cells
cells have been developed rapidly since the 1990s. In 1992, the photo-induced elec-
tron transfer from a conducting polymer to C60 was firstly reported by Heegar’s
group [24]. They invented the BHJ formed by MEH-PPV (as a donor material) and
fullerene derivative (as an acceptor material) [25]. Since then, the polymer solar
cell materials have attracted broad attentions by researchers [23,111–115].
Two major components of a PSC material are the donor and acceptor mate-
rials. Figure 5.1 schematically depicts the basic principle on how a donor-acceptor
solar cell works. When the light irradiates on the active layer, electron-hole pairs
will be generated in the polymer donor and then move to the donor-acceptor inter-
face where they become separated. Specifically, the electron goes to the LUMO
of the acceptor, while the hole stays in the highest occupied molecular orbital
(HOMO) of the donor. As a consequence, the hole and electron are transported
to the positive and negative poles, respectively, which forms the electricity and
completes the photoelectric conversion.
Donor Acceptor
LUMO
LUMO
HOMOHOMO
Interface
++
e-
e-
e-
ħω
Figure 5.1 Energy level diagram of a donor/acceptor interface showing photo excita-tion of an electron into the LUMO of the donor, followed by electron transfer into theLUMO of the acceptor, and migration of separated charges away from the interface.
At present, P3HT serves as the most representative polymer donor material.
Well-ordered P3HT has strong chain-chain interaction in the solid film. While for
the acceptors, the soluble fullerene derivative PC60BM is the most widely used
material. From Figure 5.1 one can also see that the energy levels of the donor
and acceptor are very important factors that influence the property of a solar cell
material. For instance, the energy gap between HOMO of the donor and LUMO
of the acceptor is proportional to the open circuit voltage of the solar cells [116].
37
XPS and NEXAFS spectroscopies are powerful tools for identifying elec-
tronic structures of molecules, surfaces and bulk materials. NEXAFS spectra
experiments were performed for only a few polymer solar cell acceptor (PSCA)
molecules, including PC60BM[117–119], and [6,6]-Phenyl-C71-butyric acid methyl es-
ter (PC70BM) [118]. High quality theoretical NEXAFS calculation is even rare [119].
We therefore performed a systematic theoretical study on the C K-edge NEXAFS
and XPS spectra of six representative PSCA materials: PC60BM, PC70BM, [6,6]-
Phenyl-C85-butyric acid methyl ester (PC84BM), [6,6]-Thienyl-C61-butyric acid
methyl ester (ThC60BM), bis(4-methoxyphenyl)methano[60]fullerene (DPM) and
bis[6,6]-phenyl-C61-butyric acid methyl ester (bisPC60BM). The optimized chem-
ical structures of these molecules are illustrated in Figure 5.2.
Figure 5.2 Optimized structures of PC60BM, PC70BM, PC84BM, ThC60BM, DPMand bisPC60BM. The hydrogen, carbon, oxygen and sulfur atoms are represented bywhite, grey, red and yellow spheres, respectively.
Figure 5.3 shows the XPS spectra of PCnBM and the fullerene cages Cn (n
= 60, 70, 84). One can see that the main peak of PCnBM has a weak red shift of
ca. 0.2–0.3 eV in comparison with that of Cn. This is because of the weak elec-
tronic charge transfer from the side chain (SC) to the fullerene and consequently
increased electron screening [120]. We find that the phenyl-ring carbons exhibit
similar IPs with the fullerene carbons (except the two which are directly bonded
38 5 Fullerene-Based Organic Solar Cells
288 289 290 291 292 293 294Energy(eV)
Inte
nsity(a
rb. unit)
O
O6
35 21 1'
4
7
PC84BM
PC70BM
PC60BM
C70
C60
C84
n=60,70,842,5,6,7
11'
δ
δ
δ
1' 1
11'
2,5,6,7
2,5,6,7
4
Cn
Figure 5.3 Calculated XPS of PC60BM, PC70BM and PC84BM, in comparison withthat of C60, C70 and C84. Inset show the carbon numbering. The energy shifts of themain peak with respect to corresponding naked fullerenes (denoted by δ) are -0.3, -0.2and -0.3 eV for PC60BM, PC70BM and PC84BM, respectively.
to the side chain). Our results show that the chemical environments have strong
influences on the IPs of carbons in a PSCA molecule.
Figure 5.4 compares the C 1s NEXAFS spectra of PC60BM by direct calcu-
lations, the building block (BB) approach, and experiment [118]. One can find that
the NEXAFS spectrum from direct calculations (tot. in Figure 5.4) of PC60BM
matches well with the experiment. The main difference is the double-peak at
around 286.2 eV, where theoretical result predicts a different intensity ratio than
experiment. The discrepancy maybe because the experiment is measured in the
condensed phase (thin film) while the calculation is done in the gas phase. It
suggests that the direct calculation overestimates the deformation of the fullerene
backbone introduced by the attached side chain. This is verified by a better pre-
diction of this feature by the building block approach, since it completely inherits
the Ih symmetry possessed by C60. Comparing the directly calculated spectra of
PC60BM with C60, one can find an additional feature emerges at 284.7 eV. This
is contributed by the transitions from the 1s orbitals of phenyl carbons in the
side chain to corresponding LUMOs of the atom-specific calculations. Our results
39
282 284 286 288 290 292 294
C60
SC
C60
+SC
C60
component
SC component
tot.
exp.PC
60BM
284.7
285.0
Energy(eV)
Inte
nsity
(arb
. uni
t)
Figure 5.4 Comparison of C K-edge NEXAFS spectra from direct calculations (thick
blue), the building block approach (thick red), and experiment (thick black) [118]. Thecontributions from the side chain and fullerene backbone components of the directcalculated spectra (thin blue) are also respectively compared with the spectra of isolatedside chain and C60 (thin red).
show a 0.3 eV red shift of this feature when the side chain is connected to the C60
cage. Moreover, a red shift of 0.4 and 0.5 eV is found for PC70BM and PC84BM,
respectively.
Hence, we suggest a simple computational method to effectively predict the
spectra of PSCA molecules. Since the rest three ThC60BM, DPM and bisPC60BM
are all C60 derivatives, we “borrow” the shift value of -0.3 eV obtained from
PC60BM while doing a BB calculation. In other words, we manually shift the
spectra of the isolated side chain by -0.3 eV, before we perform a conventional
BB calculation. We name such a procedure as the modified building block (MB-
B) approach. As shown in Figure 5.5, such a modification effectively improves
calculated spectra than the conventional BB method in the region around peak
b (ca. 284.7 eV). The peaks c and d are not improved since they are mainly
contributed by transitions from fullerene carbons. As a result, one can see that
for all the four C60-based PSCA molecules, the spectra in this region look quite
similar. Since the side chain has more variations than the backbone within the
large family of the PSCA molecules, this region therefore is of less importance in
40 5 Fullerene-Based Organic Solar Cells
Figure 5.5 Theoretical C K-edge NEXAFS spectra of six PSCA molecules from di-rect calculations (red), the conventional (blue) and modified (green) building blockapproaches.
characterizing new PSCA materials. By employing such a treatment, one can fully
take advantage of the pre-calculated result of the corresponding fullerene. With
only a little computational effort for each side chain carbons, a total spectroscopy
with good accuracy can be achieved. The efficiency is more evident for large sys-
tems, for instance, bis- or mutiadducts (i.e., with two or more side chains). In a
word, our MBB approach provides a simplified way for theoretical prediction of
the NEXAFS spectra of fullerene-based PSCA molecules.
Chapter 6
Self-Assembled Monolayer
Self-Assembled Monolayer (SAM) is one kind of organic membrane that plays an
important role in the surface science. An early report in 1946 described that
surfactant molecules could be absorbed on the clean metal surface to form the
monolayer [121]. However, it was not until 1983 that the first SAMs of disulfides
on gold surface came about [32]. From then on, the preparation of various SAMs
has gained lots of interests due to their superior features of highly-ordered struc-
ture, well-defined chemical bonding, flexible designing and so on [2]. It offers an
important means to design the surface structure with desirable properties.
Generally, molecules in solution can adsorb on the surface of a solid through
the chemical absorption. However, it takes specific control to let the molecules
form the self-assembled monolayer. In other words, not all molecules can be used
to generate SAMs. Normally a molecule in SAM consists of three components or
groups, namely, the head group, the spacer chain and the functional group. The
flexible properties of the SAM thus depends heavily on the characteristics of these
three parts. For instance, the modification of the head group which adsorbs on
the surface through chemical adsorption, will affect chemical interaction between
molecules and solid surface and furthermore the stability of the molecules on the
surface. Meanwhile, different chemical properties can be obtained by adjusting
the length of the middle-located spacer chain. In many applications, such as
the luminescence of molecules, the spacer chain is used to adjust the interaction
between the metal surface and the dye molecules placed on the top of SAM, in turn
the luminescence strength and color. In addition, as exposed to the air, alternative
41
42 6 Self-Assembled Monolayer
functional groups may display distinctive adsorption and reaction properties of
the SAMs surface [2,122]. Many chemical groups, such as methyl group, hydroxyl
group, and amino group have been applied to be the functional group of the SAMs.
Among these functional groups, for instance, the amino group is widely introduced
to immobilize the biomolecules and adhere cells in the bioscience field [123].
For different molecular materials, the SAMs can nevertheless be divided into
a few categories, for instance the organic sulfide absorbed on the metal or metal
oxides surface, the organosilane absorbed on the hydroxy surface, and the mixed
system with a variety of molecules [2]. At present, one of the broadly investigated
class is the SAMs based on the alkanethiols absorbed on the gold surface [2]. This
kind of SAMs have numerous advantages in application of molecular devices, like
the stable gold surface, the proper bond strength between the S and Au atoms,
the controllable reaction conditions, and the highly ordered property [2].
Nowadays, a variety of techniques were employed to characterize the SAM-
s, including the cyclic voltammetry (CV), the impedance analysis (IA) and the
chronocoulometry in electrochemical characterization; spectroscopy measurements
including the XPS, XAS, grazing incidence reflectance infrared spectroscopy (GIR-
IR), thermal desorption spectroscopy (TDS) and surface-enhanced Raman scat-
tering (SERS); microscope measurements including the scanning tunneling mi-
croscopy (STM), atomic force microscopy (AFM) and scanning electronic mi-
croscopy (SEM) [124–126]. Among of these techniques, the XAS is one of the most
useful and common tools to determine the structure properties of SAMs [33]. It can
give the information about the bonding between the head group and the metal
surface, the ordering or preferential orientation of the SAM, and the electronic
properties of the functional group. However, one has also to be aware that the
X-ray is a sword with two blades, that is, the high energy of the X-ray brings
serious irradiation damage of the samples.
The X-ray damage on alkanethiolate monolayers has attracted considerable
attentions both experimentally and theoretically [127–132]. Figure 6.1 shows the
schematic representation of the possible structure modifications of the SAM by
the X-ray damage [127]. It can be seen that the molecules in the SAM can be
cross linked together through the bonding of two neighboring molecules by the
desorption of the head group from the surface. It has come to our attention when
43
Figure 6.1 Schematic representation of the modification of C6/Au (on the left)and C12/Au (on the right) by ionizing radiation. The branching of the individualirradiation-induced processes is different for these two systems. Reprinted with permis-sion from ref. [127]. Copyright c⃝ 2011 American Chemical Society
we got the chances to study the aminothiolates SAMs with amino functional group
that the damage of the samples during the X-ray spectroscopy measurements
might be unavoidable. This concern was resulted from the fact that both the
experimental XPS and NEXAFS spectra of aminothiolates SAMs possess extra
peaks with unknown origins. The assignment of these extra spectral peaks is still
under debate.
H
C
N
S
AUDT ABT APBT ATPT
Figure 6.2 Structures of four aminothiolate molecules, AUDT, ABT, APBT andATPT.
We have carried out systematic studies on the XPS and XAS of four aminoth-
iolate molecules, which are 11-aminoundecane-1-thiol (AUDT), aromatic 4-aminob-
44 6 Self-Assembled Monolayer
enzenethiol (ABT), araliphatic 4-aminophenylbutane-1-thiol (APBT), 3-(4”-amino-
1,1’:4’,1”-terphenyl-4-yl)propane-1-thiol (ATPT) with structures depicted in Fig-
ure 6.2. These molecules have been experimentally studied by several experi-
mental groups with different spectral assignments. For the N 1s XPS of AUDT
sample, both Baio [133] and Dietrich [134] have observed the peak at the binding
energy of 399.5 eV corresponding to the free amine head group. An observed
new peak at BE = 401.4 eV was attributed to the protonated or hydrogen bond-
ed amine group [133,134]. For the N K-edge NEXAFS spectrum of AUDT sam-
ples, the resonances near 401 and 407 eV were clearly assigned to the primary
amine moieties and N-C bond, respectively [133,134]. However, the assignment of
a low-intensity pre-edge feature at 398 eV is still under debate. For this fea-
ture assignment, a hybridized or dissociated state of ammonia [135], a positively
charged form of amine [136], as well as a σ∗ resonance from the protonated amine
group [134], have been hypothesized. Similar situations were also found in the
experimental N K-edge NEXAFS spectra of aromatic SAMs with the amino func-
tional group [123,134,137]. A specific feature at 398 eV was tentatively assigned to be
attributed from the mesomeric structures, i.e. the σ−bonded free amine and the
π−bonded iminium species.
In order to reveal the origin of the extra peaks in the experimental XPS
and NEXAFS spectra, both the non-damaged and damaged species of these four
molecules were considered. On the one hand, the non-damaged species include
amine, protonated amine (PAmi), protonated amine complex (PAC), hydrogen
bonded amine (HBA) and primary ammonium (PAmm). On the other hand, the
irradiation damage can generate imine, nitrile, cumulative double bond (CDB),
azo and cross-linking (CL) species. Herein, we take AUDT as an example, all
possible species, including the non-damaged and damaged species of AUDT are
shown in Figure 6.3.
Figure 6.4 shows one set of experimental XPS for N 1s of the four molecules
listed in Figure 6.2, in which one can always observe two broad peaks, although
there is only one nitrogen atom presented in amino group. We have calculated the
binding energy or ionization potential of N 1s for all four aminothiolate molecules
and the results fit well with the peak around 399-400 eV, i.e. the one assigned
45
SH
NH3
SH
NH2
SH
NH3
SH
NH2
SH
NH3
HX X Y
Amine PAmi HBA PAmm PAC
Non-damaged species
X=OH, Cl, SH, AC
Y=H2O, AC
SH
NH
SH
N
SH
NH
SH
N N
HSSH
NH
SH
NH2
SH
N
SH
NH2
Azo CDB Imine Nitrile
Damaged species
CL1 CL2
Figure 6.3 Possible non-damaged and damaged species of AUDT under investigations.
to -NH2 group. It is found that all the species suggested in the experimental
studies, like protonated and hydrogen bonded ones, can not reproduce the extra
high energy peak. A new structure, namely PAmm in between the HBA and
PAmi, is found to be the best possible candidate that can contribute to this
spectral peak. By assuming the bond length of O· · ·H and O· · ·H· · ·N are 1.37A
and around 2.80 A, respectively, all experimental spectra are well reproduced, as
nicely demonstrated in Figure 6.4.
O
N
H
2.79
1.37
1.37
1.37
2.79
2.81
2.68
H OH
1.37
-H2N
-H2N
-H2N
-H2N
H OH
H OH
H OH
ATPT
APBT
ABT
AUDT
-NH2
-NH2
-NH2
-NH2
404 403 402 401 400 399 398 397 404 403 402 401 400 399 398 397
Inte
nsity/a
rb. u
nits
Inte
nsity/a
rb. u
nits
Energy/eV Energy/eV
Figure 6.4 Calculated N1s XPS of AUDT, ABT, APBT and ATPT (right) compared
with the corresponding experimental results [134] (left). The bond lengths marked on
the molecules are given in A.
46 6 Self-Assembled Monolayer
The N atom in the amine group is normally sp3 hybridized and its K-edge
absorption spectrum should exhibit a broad σ peak. However, the experiments
have observed pre-edge features in the NEXAFS spectra of AUDT and ATPT [134],
as shown in Figure 6.5. It can be seen that all non-damaged species do not change
the bonding situation of the N atom and can naturally not produce these extra
spectral features. We have then turned our attention to the damaged species.
The calculated NEXAFS spectra of the possible damaged products, for instance,
imine, nitrile, azo, CDB and CL species, do give the π∗ features at the right
energy position, as highlighted in Figure 6.5. For the AUDT compound, the π∗
structure observed in the experiment is quite broad, which could be associated
with the fact that there are several possible products can all contribute to this
energy region. We have only considered three photo-damaged products, azo, CL1
and CL2 species for the ATPT. Among them, the azo and CL2 can give π∗ peaks
located around 398.1 eV and 396.8 eV, respectively. It seems that the former
gives better agreement with the experimental spectrum. Our studies demonstrate
that the theoretical calculations can play a decisive role in identifying the photo-
damage products.
395 400 405 410 415 420 425
CL1
CL2
CDB
Azo
Nitrile
Imine
PAmm-OH
PAC−H O2
PAmi
Amine
exp
AUDT
Energy(eV)
Inte
nsity(a
rb.
un
it)
395 400 405 410 415 420 425
CL1
CL2
Azo
PAmm-OH
PAC-H2O
PAmi
Amine
exp
ATPT
Energy(eV)
Inte
nsity(a
rb.
un
it)
Figure 6.5 Calculated N 1s NEXAFS spectra of ATPT (right) and AUDT (left) and
their derivatives compared with experiment [134].
Chapter 7
Summary of included papers
7.1 Paper 1: Neutral and charged N@C60
We have calculated potential energy surfaces that describe the motion of the N
atom inside the cavity of the endohedral fullerene N@C60. It aims to explain the
experimental observation that under the high temperature, the N atom can escape
from the carbon cage. The most energetic favorable escaping pathway for the N
atom is identified, which goes from the center toward the C-C/C=C bond which
connects two neighboring rings. Around the center of the cavity, the complex
prefers to remains the same spin state as the isolated atom (S=3/2), implying
that the interaction between the N atom and the carbon cage is indeed very small
and the N@C60 can be used as a spin system for quantum computing. One of the
important findings of this study is the important role played by the dispersion
force in determining the equilibrium geometry of the N@C60. With the popular
B3LYP functional, the complex with the N atom at the center is actually not
a stable minimum, but a local maximum. In contrast, the dispersion corrected
functionals, B3LYP-DCP and M06-2X, are found to provide correct description
for the system.
7.2 Paper 2: Structure of A@C60
Inspired by the findings in Paper 1, we decided to systematically study the equi-
librium structure of the family of endohedral fullerene A@C60 and to verify the
47
48 7 Summary of included papers
importance of the dispersion force for such systems. Ten endohedral fullerenes
A@C60, A = H, C, O, S, N, P, He, Ne, Ar, and Kr, have been chosen that
cover both reactive non-metal and nobel gases atoms. Several commonly used
functionals, BLYP, BP86, and B3LYP, together with three dispersion corrected
functionals, B3LYP-DCP, B97-D and M06-2X, have been employed. It is found
that the inclusion of the dispersion force can give correct negative binding energy
and stabilize the system. However, to our surprise, none of the functionals is ca-
pable of correctly describe the structure of all ten endohedral fullerenes, not only
the equilibrium geometry, but also the potential energy surface for the motion of
the N atom. How to improve the density functional for such systems is certain
something that should be explored.
7.3 Paper 3: Fullerene-based solar cell acceptors
The basic idea of this study is to examine how accurate the theoretical modeling
can be for XPS and NEXAFS spectra of fullerene-based solar cell acceptors and
to provide correct spectral assignments for these spectra. Six typical molecules,
PCnBM (n=60, 70, 84), ThC60BM, DPM and bisPC60BM have been calculated,
which reveal the specific structure-spectrum relationship. The calculated results
are in perfect agreement with the available experimental spectra. We have also
proposed a modified building block approach to quickly resemble the total spec-
trum of the entire system by summing up those from two small species. It will
allow to obtain the NEXFAS spectra of the large number of fullerene-based solar
cell acceptors available with very little computational effort.
7.4 Paper 4: Aminothiolates on SAMs
Amine-terminated SAMs are one kind of versatile functional materials. However,
their experimental XPS and NEXFAS spectra have long been known to be difficult
to explain because of the appearance of unexpected spectral features. Differen-
t structures or hypotheses have been put forward over the years from different
groups. In this study, we have shown from theoretical calculations that the extra
spectral peak in XPS spectra is resulted from a new specie, named as primary
7.4 Paper 4: Aminothiolates on SAMs 49
ammonium that is in between protonated and hydrogen bonded structures. The
X-ray damage is found to be quite common for SAMs and its products can lead to
extra spectral features in NEXFAS spectra. The possible photo-damaged prod-
ucts have been identified by comparing the experimental and calculated spectra.
This study resolves a long standing controversy on structure determination of this
useful system that has had extensive applications in biotechnology.
References
[1] Wrobel, D.; Dudkowiak, A. Mol. Cryst. Liq. Cryst. 2006, 448, 15.
[2] Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G. M. Chem.Rev. 2005, 105, 1103.
[3] Ratera, I.; Veciana, J. Chem. Soc. Rev. 2012, 41, 303.
[4] Kroto, H. W.; Heath, J. R.; O’brien, S. C.; Curl, R. F.; Smalley, R. E. Nature1985, 318, 162.
[5] http://nobelprize.org.
[6] Kong, Q.; Zhuang, J.; Li, X.; Cai, R.; Zhao, L.; Qian, S.; Li, Y. Appl. Phys. A2002, 75, 367.
[7] Jensen, A. W.; Wilson, S. R.; Schuster, D. I. Bioorg. Med. Chem. 1996, 4, 767.
[8] Cami, J.; Bernard-Salas, J.; Peeters, E.; Malek, S. E. Science 2010, 329, 1180.
[9] Heath, J. R.; O’brien, S. C.; Zhang, Q.; Liu, Y.; Curl, R. F.; Kroto, H. W.;Tittel, F. K.; Smalley, R. E. J. Am. Chem. Soc. 1985, 107, 7779.
[10] Buhl, M.; Patchkovskii, S.; Thiel, W. Chem. Phys. Lett. 1997, 275, 14.
[11] Hu, Y. H.; Ruckenstein, E. Chem. Phys. Lett. 2004, 390, 472.
[12] Brown, R. M.; Ito, Y.; Warner, J. H.; Ardavan, A.; Shinohara, H.; Briggs, G.A. D.; Morton, J. J. L. Phys. Rev. B 2010, 82, 033410.
[13] Morscher, M.; Seitsonen, A. P.; Ito, S.; Takagi, H.; Dragoe, N.; Greber, T. Phys.Rev. A 2010, 82, 051201.
[14] Whitener, K. E. J. Phys. Chem. A 2010, 114, 12075.
[15] Chaur, M. N.; Melin, F.; Ortiz, A. L.; Echegoyen, L. Angew. Chem. Int. Ed. 2009,48, 7514.
[16] Xing, G.; Yuan, H.; He, R.; Gao, X.; Jing, L.; Zhao, F.; Chai, Z.; Zhao, Y. J.Phys. Chem. B 2008, 112, 6288.
[17] Okazaki, T.; Shimada, T.; Suenaga, K.; Ohno, Y.; Mizutani, T.; Lee, J.; Kuk, Y.;Shinohara, H. Appl. Phys. A 2003, 76, 475.
51
52 REFERENCES
[18] Murphy, T. A.; Pawlik, T.; Weidinger, A.; Hohne, M.; Alcala, R.; Spaeth, J. M.Phys. Rev. Lett. 1996, 77, 1075.
[19] Meyer, C.; Harneit, W.; Naydenov, B.; Lips, K.; Weidinger, A. Appl. Magn. Reson.2004, 27, 123.
[20] Morton, J. J. L.; Tyryshkin, A. M.; Ardavan, A.; Porfyrakis, K.; Lyon, S. A.;Briggs, G. A. D. J. Chem. Phys. 2006, 124, 014508.
[21] Nikawa, H.; Araki, Y.; Slanina, Z.; Tsuchiya, T.; Akasaka, T.; Wada, T.; Ito, O.;Dinse, K. P.; Ata, M.; Kato, T.; Nagase, S. Chem. Commun. 2010, 46, 631.
[22] Harneit, W.; Meyer, C.; Weidinger, A.; Suter, D.; Twamley, J. Phys. Status SolidiB 2002, 233, 453.
[23] He, Y.; Li, Y. Phys. Chem. Chem. Phys. 2011, 13, 1970.
[24] Sariciftci, N. S.; Smilowitz, L.; Heeger, a. J.; Wudl, F. Science 1992, 258, 1474.
[25] Yu, G.; Gao, J.; Hummelen, J. C.; Wudl, F.; Heeger, A. J. Science 1995, 270,1789.
[26] Andersen, T. R.; Larsen-Olsen, T. T.; Andreasen, B.; Bottiger, A. P. L.; Car-le, J. E.; Helgesen, M.; Bundgaard, E.; Norrman, K.; Andreasen, J. W.; Jør-gensen, M.; Krebs, F. C. ACS Nano 2011, 5, 4188.
[27] Anctil, A.; Babbitt, C. W.; Raffaelle, R. P.; Landi, B. J. Environ. Sci. Technol.2011, 45, 2353.
[28] Kooistra, F. B.; Mihailetchi, V. D.; Popescu, L. M.; Kronholm, D.; Blom, P.W. M.; Hummelen, J. C. Chem. Mater. 2006, 18 .
[29] Wang, X.; Perzon, E.; Delgado, J. L.; de la Cruz, P.; Zhang, F.; Langa, F.;Andersson, M.; Inganas, O. Appl. Phys. Lett. 2004, 85, 5081.
[30] Sternfeld, T.; Thilgen, C.; Hoffman, R. E.; del Rosario Colorado Heras, M.;Diederich, F.; Wudl, F.; Scott, L. T.; Mack, J.; Rabinovitz, M. J. Am. Chem.Soc. 2002, 124, 5734.
[31] Ulman, A. Chem. Rev. 1996, 96, 1533.
[32] Nuzzo, R.; Allara, D. J. Am. Chem. Soc. 1983, 105, 4481.
[33] Hermanson, G. T. Bioconjugate Techniques, 2nd revised ed.; Academic Press (El-sevier), London, 2008.
[34] Becerril, H. A.; Mao, J.; Liu, Z.; Stoltenberg, R. M.; Bao, Z.; Chen, Y. ACS Nano2008, 2, 463.
[35] Fan, X.; Peng, W.; Li, Y.; Li, X.; Wang, S.; Zhang, G.; Zhang, F. Adv. Mater.2008, 20, 4490.
REFERENCES 53
[36] Jeong, H.-K.; Lee, Y. P.; Lahaye, R. J. W. E.; Park, M.-H.; An, K. H.; Kim, I. J.;Yang, C.-W.; Park, C. Y.; Ruoff, R. S.; Lee, Y. H. J. Am. Chem. Soc. 2008, 130,1362.
[37] Szabo, T.; Berkesi, O.; Forgo, P.; Josepovits, K.; Sanakis, Y.; Petridis, D.;Dekany, I. Chem. Mater. 2006, 18, 2740.
[38] Si, Y.; Samulski, E. T. Nano Lett. 2008, 8, 1679.
[39] Cai, W.; Piner, R. D.; Stadermann, F. J.; Park, S.; Shaibat, M. A.; Ishii, Y.;Yang, D.; Velamakanni, A.; An, S. J.; Stoller, M.; An, J.; Chen, D.; Ruoff, R. S.Science 2008, 321, 1815.
[40] Slater, J. C. Phys. Rev. 1929, 34, 1293.
[41] Szabo, A.; Ostlund, N. S. Modern Quantum Chemistry: Introduction to AdvancedElectronic Structure Theory ; Dover Publications Inc., 2000.
[42] Pople, J. A. Rev. Mod. Phys. 1999, 71, 1267.
[43] Kutzelnigg, W. J. Mol. Struc. (Theochem.) 1988, 181, 33.
[44] Jensen, F. Introduction to computational chemistry ; John Wiley and Sons, 2007.
[45] Purvis, G. D.; Bartlett, R. J. J. Chem. Phys. 1982, 76, 1910.
[46] Raghavachari, K.; Trucks, G. W.; Pople, J. A.; Head-Gordon, M. Chem. Phys.Lett. 1989, 157, 479.
[47] Voorhis, T. V.; Head-Gordon, M. J. Chem. Phys. 2001, 115, 5033.
[48] Thomas, L. H. Math. Proc. Cambridge Phil. Soc. 1927, 23, 542.
[49] Fermi, E. Rend. Acad. Nat. Lincei 1927, 6, 602.
[50] Hohenberg, P.; Kohn, W. Phys. Rev. 1964, 136, B864.
[51] Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
[52] Stone, A. J. The theory of intermolecular forces; Oxford:Oxford University Press,1997.
[53] Kaplan, I. G. Intermolecular Interactions; Chichester:John Wiley & Sons, 2006.
[54] London, F. Z. Phys. 1930, 63, 245.
[55] London, F. Trans. Faraday Soc. 1937, 33, 8b–26.
[56] Hettema, H. Quantum Chemistry, Classic Scientific Papers; World Scientific, Sin-gapore, 2000.
[57] Perez-Jorda, J. M.; San-Fabian, E.; Perez-Jimenez, A. J. J. Chem. Phys. 1999,110, 1916.
54 REFERENCES
[58] Hobza, P.; Sponer, J.; Reschel, T. J. Comput. Chem. 1995, 16, 1315.
[59] Kristyan, S.; Pulay, P. Chem. Phys. Lett. 1994, 229, 175.
[60] Perez-Jorda, J. M.; Becke, A. Chem. Phys. Lett. 1995, 233, 134.
[61] Swart, M.; Sola, M.; Bickelhaupt, F. M. J. Comput. Chem. 2011, 32, 1117.
[62] Grimme, S. J. Comp. Chem. 2004, 25, 1463.
[63] Grimme, S. J. Chem. Phys. 2006, 124, 034108.
[64] Grimme, S. WIREs: Comput. Mol. Sci. 2011, 1, 211.
[65] Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157.
[66] Mackie, I. D.; DiLabio, G. A. J. Phys. Chem. A 2008, 112, 10968.
[67] Johnson, E. R.; Mackie, I. D.; DiLabio, G. A. J. Phys. Org. Chem. 2009, 22,1127.
[68] von Barth, U.; Grossman, G. Solid State Commun. 1979, 32, 645.
[69] von Barth, U.; Grossman, G. Phys. Rev. B 1982, 25, 5150.
[70] Triguero, L.; Plashkevych, O.; Pettersson, L. G. M.; Agren, H. J. Electron Spec-trosc. Relat. Phenom. 1999, 104, 195.
[71] Bagus, P. S. Phys. Rev. 1965, 139, 619.
[72] Siegbahn, K.; Nordling, C.; Fahlman, A.; Hamrin, K.; Hedman, J.; R, R. N.;Johansson, C.; Bergmark, T.; Karlsson, S. E.; Lindgren, I.; Lindberg, B. ESCA;atomic, molecular and solid state structure studied by means of electron spec-troscopy ; Almqvist and Wiksells, Uppsala, 1967.
[73] Stohr, J. NEXAFS Spectroscopy ; Springer Verlag: Berlin, Heidelberg, New York,1992.
[74] Hua, W. Structure and spectroscopy of bio- and nano-materials from first-principles simulations; Doctoral Thesis: KTH, Stockholm, 2011.
[75] Hua, W.; Gao, B.; Li, S.; Agren, H.; Luo, Y. Phys. Rev. B 2010, 82, 155433.
[76] Chai, Y.; Guo, T.; Jin, C.; Haufler, R. E.; Chibante, L. P. F.; Fure, J.; Wang, L.;Alford, J. M.; Smalley, R. E. J. Phys. Chem. 1991, 95, 7564.
[77] Shinohara, H. Rep. Prog. Phys. 2000, 63, 843.
[78] Ito, Y.; Fujita, W.; Okazaki, T.; Sugai, T.; Awaga, K.; Nishibori, E.; Takata, M.;Sakata, M.; Shinohara, H. ChemPhysChem 2007, 8, 1019.
[79] Jin, P.; Hao, C.; Li, S.; Mi, W.; Sun, Z.; Zhang, J.; Hou, Q. J. Phys. Chem. A2007, 111, 167.
REFERENCES 55
[80] Liu, L.; Gao, B.; Chu, W.; Chen, D.; Hu, T.; Wang, C.; Dunsch, L.; Marcelli, A.;Luo, Y.; Wu, Z. Chem. Commun. 2008, 474.
[81] Stevenson, S.; Rice, G.; Glass, T.; Harich, K.; Cromer, F.; Jordan, M.; Craft, J.;Hadju, E.; Bible, R.; Olmstead, M.; Others, Nature 1999, 401, 55.
[82] Olmstead, M. M.; de Bettencourt-Dias, A.; Duchamp, J. C.; Stevenson, S.;Dorn, H. C.; Balch, A. L. J. Am. Chem. Soc. 2000, 122, 12220.
[83] Dunsch, L.; Yang, S. Phys. Chem. Chem. Phys. 2007, 9, 3067.
[84] Wang, C.-R.; Kai, T.; Tomiyama, T.; Yoshida, T.; Kobayashi, Y.; Nishibori, E.;Takata, M.; Sakata, M.; Shinohara, H. Angew. Chem. Int. Ed. 2001, 40, 397.
[85] Stevenson, S.; Mackey, M. A.; Stuart, M. A.; Phillips, J. P.; Easterling, M. L.;Chancellor, C. J.; Olmstead, M. M.; Balch, A. L. J. Am. Chem. Soc. 2008, 130,11844.
[86] Dunsch, L.; Yang, S.; Zhang, L.; Svitova, A.; Oswald, S.; Popov, A. A. J. Am.Chem. Soc. 2010, 132, 5413.
[87] Krause, M.; Ziegs, F.; Popov, A. A.; Dunsch, L. ChemPhysChem 2007, 8, 537.
[88] Dunsch, L.; Yang, S. Small 2007, 3, 1298.
[89] Iezzi, E. B.; Duchamp, J. C.; Fletcher, K. R.; Glass, T. E.; Dorn, H. C. NanoLett. 2002, 2, 1187.
[90] Saunders, M.; Jimenez-Vazquez, H. A.; Cross, R. J.; Poreda, R. J. Science 1993,259, 1428.
[91] Cimpoesu, F.; Ito, S.; Shimotani, H.; Takagi, H.; Dragoe, N. Phys. Chem. Chem.Phys. 2011, 13, 9609.
[92] Yamamoto, K.; Saunders, M.; Khong, A.; Cross, R. J.; Grayson, M.; Gross, M. L.;Benedetto, A. F.; Weisman, R. B. J. Am. Chem. Soc. 1999, 121, 1591.
[93] Syamala, M. S.; Cross, R. J.; Saunders, M. J. Am. Chem. Soc. 2002, 124, 6216.
[94] Frunzi, M.; Cross, R. J.; Saunders, M. J. Am. Chem. Soc. 2007, 129, 13343.
[95] Weidinger, A.; Waiblinger, M.; Pietzak, B.; Almeida Murphy, T. Appl. Phys. A1998, 66, 287.
[96] Iwasiewicz-Wabnig, A.; Porfyrakis, K.; Briggs, G. A. D.; Sundqvist, B. Phys.Status Solidi B 2009, 246, 2767.
[97] Naydenov, B.; Mende, J.; Harneit, W.; Mehring, M. Phys. Status Solidi B 2008,245, 2002.
[98] Waiblinger, M.; Lips, K.; Harneit, W.; Weidinger, A.; Dietel, E.; Hirsch, A. Phys.Rev. B 2001, 64, 159901.
56 REFERENCES
[99] Toth, S.; Quintavalle, D.; Nafradi, B.; Korecz, L.; Forro, L.; Simon, F. Phys. Rev.B 2008, 77, 214409.
[100] Mauser, H.; Hirsch, A.; van Eikema Hommes, N.; Clark, T.; Pietzak, B.; Wei-dinger, A.; Dunsch, L. Angew. Chem. Int. Ed. 1997, 36, 2835.
[101] Sanville, E.; BelBruno, J. J. Phys. Chem. B 2003, 107, 8884.
[102] Yue, X.; Zhao, J. J.; Hao, C.; Ren, W. X.; Qiu, J. S. J. Comput. Theor. Nanosci.2008, 5, 1306.
[103] Song, X.; Ma, Y.; Wang, C.; Luo, Y. Chem. Phys. Lett. 2011, 517, 199.
[104] Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200.
[105] Becke, A. D. Phys. Rev. A 1988, 38, 3098.
[106] Perdew, J. P. Phys. Rev. B 1986, 33, 8822.
[107] Chapin, D.; Fuller, C.; Pearson, G. J. Appl. Phys. 1954, 25, 676.
[108] Wohrle, D.; Meissner, D. Adv. Mater. 1991, 3, 129.
[109] Kallmann, H.; Pope, M. J. Chem. Phys. 1959, 30, 585.
[110] Tang, C. W. Appl. Phys. Lett. 1986, 48, 183.
[111] Li, G.; Shrotriya, V.; Huang, J.; Yao, Y.; Moriarty, T.; Emery, K.; Yang, Y. Nat.Mater. 2005, 4, 864.
[112] Kim, Y.; Cook, S.; Tuladhar, S. M.; Choulis, S. a.; Nelson, J.; Durrant, J. R.;Bradley, D. D. C.; Giles, M.; McCulloch, I.; Ha, C.-S.; Ree, M. Nat. Mater. 2006,5, 197.
[113] Kim, J. Y.; Lee, K.; Coates, N. E.; Moses, D.; Nguyen, T.-Q.; Dante, M.;Heeger, A. J. Science 2007, 317, 222.
[114] Li, Y.; Zou, Y. Adv. Mater. 2008, 20, 2952.
[115] He, Y.; Chen, H.-Y.; Hou, J.; Li, Y. J. Am. Chem. Soc. 2010, 132, 1377.
[116] Brabec, C. J.; Cravino, A.; Meissner, D.; Sariciftci, N. S.; Fromherz, T.; Rispen-s, M. T.; Sanchez, L.; Hummelen, J. C. Adv. Funct. Mater. 2001, 11, 374.
[117] Guan, Z.-L.; Kim, J. B.; Wang, H.; Jaye, C.; Fischer, D. A.; Loo, Y.-L.; Kahn, A.Org. Electron. 2010, 11, 1779.
[118] Watts, B.; Swaraj, S.; Nordlund, D.; Luning, J.; Ade, H. J. Chem. Phys. 2011,134, 024702.
[119] Bazylewski, P. F.; Kim, K. H.; Forrest, J. L.; Tada, H.; Choi, D. H.; Chang, G. S.Chem. Phys. Lett. 2011, 508, 90.
[120] Hagstrom, S.; Nording, C.; Sieghbahn, K. Phys. Lett. 1964, 9, 235.
REFERENCES 57
[121] Bigelow, W.; Pickett, D.; Zisman, W. J. Colloid Sci. 1946, 1, 513.
[122] Smith, R. K.; Lewis, P. A.; Weiss, P. S. Prog. Surf. Sci. 2004, 75, 1.
[123] Graf, N.; Yegen, E.; Gross, T.; Lippitz, A.; Weigel, W.; Krakert, S.; Terfort, A.;Unger, W. E. Surf. Sci. 2009, 603, 2849.
[124] Kind, M.; Woll, C. Prog. Surf. Sci. 2009, 84, 230.
[125] Huang, X.; Huang, H.; Wu, N.; Hu, R.; Zhu, T.; Liu, Z. Surf. Sci. 2000, 459, 183.
[126] Klein, H.; Battaglini, N.; Bellini, B.; Dumas, P. Mater. Sci. Eng., C 2002, 19,279.
[127] Hamoudi, H.; Chesneau, F.; Patze, C.; Zharnikov, M. J. Phys. Chem. C 2011,115, 534.
[128] Huels, M. a.; Dugal, P.-C.; Sanche, L. J. Chem. Phys. 2003, 118, 11168.
[129] Shaporenko, A.; Zharnikov, M.; Feulner, P.; Menzel, D. J. Phys.: Condens. Matter2006, 18, S1677.
[130] Zharnikov, M. J. Electron. Spectrosc. Relat. Phenom. 2010, 178, 380.
[131] Heister, K.; Zharnikov, M.; Grunze, M. Langmuir 2001, 17, 8.
[132] Hamoudi, H.; Doring, K.; Chesneau, F.; Lang, H.; Zharnikov, M. J. Phys. Chem.C 2012, 116, 861.
[133] Baio, J. E.; Weidner, T.; Brison, J.; Graham, D. J.; Gamble, L. J.; Castner, D. G.J. Electron Spectrosc. Relat. Phenom. 2009, 172, 2.
[134] Dietrich, P. M.; Graf, N.; Gross, T.; Lippitz, A.; Krakert, S.; Schupbach, B.;Terfort, A.; Unger, W. E. S. Surf. Interface Anal. 2010, 42, 1184.
[135] Ozawa, K.; Hasegawa, T.; Edamoto, K.; Takahashi, K.; Kamada, M. J. Phys.Chem. B 2002, 106, 9380.
[136] Zubavichus, Y.; Zharnikov, M.; Schaporenko, A.; Grunze, M. J. Electron Spec-trosc. Relat. Phenom. 2004, 134, 25.
[137] Dietrich, P. M.; Graf, N.; Gross, T.; Lippitz, A.; Schupbach, B.; Bashir, A.;Woll, C.; Terfort, A.; Unger, W. E. S. Langmuir 2010, 26, 3949.
top related