Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Plasmonic enhancement of catalysis and solar energy conversion
(USC Thesis Other)
Plasmonic enhancement of catalysis and solar energy conversion
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
PLASMONIC ENHANCEMENT OF CATALYSIS AND SOLAR ENERGY
CONVERSION
by
Wei Hsuan Hung
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(MATERIALS SCIENCE)
August 2011
Copyright 2011 Wei Hsuan Hung
ii
Dedication
This thesis is dedicated to my parents, brother and sister
for their endless support and encouragement.
iii
Acknowledgements
I would like to acknowledge my advisor, Prof. Stephen B. Cronin. With his guidance,
I have learned and developed independent thoughts and problem solving skills. During
my doctoral work in his lab, he taught me not only the methodologies of conducting
comprehensive experiments, but also to develop a positive attitude in performing difficult
research. In each project that I have accomplished, he has been a strong mentor that
deeply inspired my passion to further my research. I really appreciated and enjoyed my
time working with Prof. Cronin, for it was a truly wonderful and unforgettable journey in
my academic life.
Furthermore, I would also like to thank all members of the Cronin group, including
Rajay Kumar, Adam Bushmaker, who were there to lend a hand when I was a new
graduate student in this lab. My special thanks also goes out to all the current lovely
labmates, I-Kai Hsu, Chun-Chung Chen, Chia-Chi Chang, Jesse R. Theiss, Mehmet
Aykol, Prathamesh Pavaskar, Zu-Wei Liu, Wen-Bo Hou, Mohammed Amer, Rohan
Dhall, Shun-Wen Chang and Yun-Chiao Huang. My work could not have been carried
out smoothly without your support and wonderful collaboration. Chun-Chung Chen,
iv
Chia-Chi Chang, thanks for your continuous support when I needed it most. Jesse, my
lunch buddy, your life experiences are truly an inspiration. Mehmet (Memo), I appreciate
having you in the lab, because you make life so much easier when life gets hard in
experiment setup.
Additionally, I would also like to thank all my friends, Peter in Prof. Dapkus’s group,
Ray in Prof. Kim’s group, and all the members in Prof. Zhou’s and Prof. Thompson’s lab
at USC. I really appreciate all the support you have given me and wish you all the very
best in the near future.
v
Table of Contents
Dedication … ……… ………… ……… … ………… ……… … ………. …ii
Acknowledgements ………… ……… … ………… ……… … ………… iii
List of Figures … … ………… …….. …… ………… ……… … ………..viii
Abstrac t…… ……… ………… ……… … ………… ……… … ……. …..xii
Chapter 1: Introduction .......................................................................... 1
1.1 Background………………………...…………………...………………………1
1.2 Surface Plasmons in Metal Naoparticl …………………………………………6
1.3 Current Applications of Plasmonic Nanoparticles ............................................ 12
1.4 Contents of this Thesis ...................................................................................... 18
Chapter 2: Laser Directed Growth of Carbon-Based Nano-
structures by Plasmon Resonant Chemical Vapor Deposition ..........23
2.1 Introduction………………………………………………………..…………..23
2.2 Experimental Procedure .................................................................................... 25
2.2.1 Preparation of Plasmonic Substrate ............................................................ 25
2.2.2 Plasmon Resonant Chemical Vapor Deposition (PRCVD) System Setup.26
2.2.3 Sample Characterization ............................................................................. 26
2.3 Results and Discussion ..................................................................................... 27
2.3.1 PRCVD of Iron Oxide ................................................................................ 27
2.3.2 PRCVD Carbon Nanotubes ........................................................................ 30
2.3.3 Laser Writing of PRCVD Suspended Carbon Nanotubes .......................... 33
2.3.4 Material Characterization ........................................................................... 36
2.3.5 PRCVD Growth on SU-8 and Pattern Capability ...................................... 39
2.4 Conclusion ......................................................................................................... 41
Chapter 3: Plasmon Resonant Enhancement of Carbon
Monoxide Catalysis ………… ……… … ………….. .. .. .. .. .. .. .. .. . .. .. .. .. .. .. .42
3.1 Introduction ....................................................................................................... 42
3.2 Experimental Procedure .................................................................................... 44
3.2.1 Micro-reactor System Setup ....................................................................... 44
vi
3.2.2 Silicon-based Microchannel Fabrication .................................................... 45
3.3 Results and Discussion ...................................................................................... 47
3.3.1 Observation of in situ Temperature and Products of Plasmon Enhanced
CO Oxidation.............................................................................................. 47
3.3.2 Comparison of Au NPs, Iron Oxide and Au Nanoparticle/Fe
2
O
3
Composite ................................................................................................... 49
3.3.3 Thermally-driven Vs. Plasmonically-driven .............................................. 51
3.3.4 Quantification of the Plasmonic Enhancement .......................................... 53
3.4 Conclusion ......................................................................................................... 54
Chapter 4: Plasmonic Enhancement of Dye Sensitized Solar Cell ....56
4.1 Introduction ........................................................................................................ 56
4.2 Experimental Procedure .................................................................................... 59
4.2.1 Preparation of Anodic TiO
2
and Cell Assembling ..................................... 59
4.3 Result and Discussion ....................................................................................... 62
4.3.1 Electrical Characterization ......................................................................... 62
Table 1. Summary characteristics of dye sensitized solar cell with
and without plasmonic gold nanoparticles……………………………….66
4.3.2 FDTD Simulation of Plasmon Induced Electric Feilds ............................. 69
4.3.3 Photoluminescence Spectrum ..................................................................... 71
4.4 Conclusion ......................................................................................................... 74
Chapter 5: Plasmonic Enhancement of Methane Production............75
5.1 Introduction……………………………………………………………………75
5.2 Experimental Procedure .................................................................................... 78
5.2.1 Preparation of Plasmonic Substrate ............................................................ 78
5.3 Results and Discussion ...................................................................................... 79
5.3.1 Visible Light Illumination .......................................................................... 79
5.3.2 Ultraviolet Light Illumination ................................................................... .81
5.4 Conclusion……………………………………………………………………..83
Chapter 6: Rapid Prototyping of Three-Dimensional
Microstructures from Multi-Walled Carbon Nanotubes ...................85
6.1 Introduction ....................................................................................................... 85
6.2 Experimental Procedure .................................................................................... 86
6.2.1 Carbon Nanotubes Forest Sample Fabrication ........................................... 86
vii
6.2.2 Opitcal Measurement - Raman Spectroscopy ............................................ 87
6.3 Results and Discussion ...................................................................................... 88
6.3.1 Creating 3D Structures in Carbon Nanotubes Forest ................................. 88
6.3.2 Burnout Depth vs. Laser Power Exposure ................................................. 90
6.3.3 Surface after Burnout Treatment and Burnout Threshold Power ............... 91
6.4 Conclusion ......................................................................................................... 93
Chapter 7: Conclusions ..........................................................................95
References … ….. .....................................................................................97
viii
List of Figures
Figure 1.1 (a) Cathedral of Notre Dame (b) The Lycurgus Cup. 4
Figure1.2 (a) Bulk gold (b) Gold nanoparticles with various
diameters.
6
Figure1.3 (a) Surface charge resonance and plasmon-induce electric
field distribution at metal and dielectric interface. (b)
Dispersion relation of a thin film metal.
7
Figure 1.4 (a) A schematic of the localized surface plasmon.(b) Light
scattering and absorption of a spherical metal nanoparticle.
9
Figure 1.5
(a) Real, and (b) Imaginary part of the complex dielectric
function part of gold plotted with prediction of the Drude
model and Drude-Lorentz model.[14]
12
Figure 1.6 (a) Dispersion relation of light and surface plasmon
mode.(b) Subwavelength 90nm dot arrays fabricated using
365nm UV light with a 170nm period.
13
Figure 1.7 (a) SEM image of carbon nanotubes covered with a film of
silver nanoparticles. The white circle indicates the size and
location of the laser spot. (b) Raman spectra taken at the
laser spot position indicated in (a) before and after Ag
deposition.
15
Figure 1.8 Combined imaging and therapy of SKBr3 breast cancer
cells using HER2-targeted nanoshells.
16
Figure 1.9 (a) SEM image showing a FIB cross section of a fully
fabricated n-i-p a-Si:H solar cell grown on the patterned
back contact. (b) Best J-V measurements of the flat
reference and patterned n-i-p a-Si:H cells. Inset shows the
cell characteristics for each device. [30]
17
Figure 2.1 (a) SEM image of iron oxide nanocrystals deposited on Au
nanoparticles after laser irradiation in a CO environment.
(b) In situ Raman data taken during laser exposure. (c)
Raman spectra of amorphous carbon deposited during the
first 12s of irradiation and iron oxide deposited after 130s.
29
ix
Figure 2.2
(a,b) SEM images of carbon-based nanostructures
synthesized by irradiating Au nanoparticles in a CO
environment. (c) Raman spectrum of the material shown in
(a) and (b), exhibiting typical features of carbon nanotubes.
(d) In situ Raman data taken during laser exposure.
33
Figure 2.3 (a,b) Tilted SEM images of a suspended carbon nanotube
spanning two sequential growth spots. The white arrow
indicates the direction of laser movement. (c,d) High
magnification SEM images taken at the two ends of this
suspended carbon nanotube.
35
Figure 2.4 (a) The SEM image of PRCVD grown iron oxide. (b)
Energy dispersive X-ray (EDX) spectrum of
nanocrystalline material shown in (a). (c) Raman spectrum
of nanocrystaline Fe
2
O
3
grown by the PRCVD process.
37
Figure 2.5 TEM images of iron oxide nanoparticles interspersed
with CNTs.
38
Figure 2.6 (a) SEM image and (b) Raman spectrum of Fe
2
O
3
and
carbon nanotubes grown on SU-8 polymer.
40
Figure 3.1 Schematic diagram of the micro-reactor system. 45
Figure 3.2 Silicon micro-channels with glass cover and metal
eyelets.
47
Figure 3.3 (a) Quadropole mass spectrometer and (b) infrared
temperature data taken during laser irradiation in CO.
49
Figure 3.4 Quadropole mass spectrometer signals for CO
2
during 5sec
laser-induced reactions on various catalytic substrates.
51
Figure 3.5 Mass spectrometer signals of CO
2
taken after uniformly
heating the Au nanoparticle sample to 350
o
C.
52
Figure 3.6 (a) QMS signal for CO
2
during a series of 10 30-second
laser irradiations with a base substrate temperature of
300
o
C. (b) Comparison of the average CO
2
QMS signal for
laser irradiation and uniform heating.
54
Figure 4.1 Principle of operation and energy level diagram of the
plasmon-enhanced dye sensitized solar cell.
60
x
Figure 4.2
(a) Photograph of the plasmon-enhanced cell before
electric characterization (b) Raman spectrum of anodic
TiO
2
in anatase phase after annealing at 450
o
C.
62
Figure 4.3
(a) J–V characteristics of dye sensitized solar cells. (b)
Normalized incident photon to current conversion
efficiency spectrum with and without island-like gold
nanosturcutres.
64
Figure 4.4
I-V characteristics of dye sensitized solar cells with and
without island-like gold nanoparticles under (a) 532nm (b)
633nm monochromatic laser irradiation.
65
Figure 4.5
(a) Scanning electron microscope (SEM) image of 5nm
gold film on TiO
2
. (b) UV-Vis absorption spectrum of 5nm
gold film alone. (c) Short-circuit photo-current
enhancement factor of 5nm gold film on the dye loaded
TiO
2.
68
Figure 4.6
(a) SEM image of a 5 nm thick Au island film. (b-d)
simulated electric field intensity observed from the top
surface and cross section of the interface of Au–TiO
2
calculated using FDTD.
70
Figure 4.7
Absorption spectra of phosphorescence dye (black), 5nm
annealed gold (red), and dye with annealed 5nm gold
underneath(blue).
73
Figure 4.8
Emission spectra of (a) TPBP dye only (b) TPBP dye and
annealed 5nm gold on the solgel TiO
2.
73
Figure 5.1
(a) Absorption spectrum of TiO
2
and solar spectrum. (b)
Absorption spectrum of metal nanoparticles.
77
Figure 5.2
(a) UV-Vis absorption spectrum of TiO
2
with and without
gold nanoparticles. (b) Experimental setup showing the gas
phase photochemical cell.
79
Figure 5.3
(a) Photocatalytic product yields (after 15 h of visible
irradiation) from two different catalytic surfaces. (b)
Energy band alignment of anatase TiO
2
, Au, and the
relevant redox potentials of CO
2
and H
2
O under visible
illumination.
80
xi
Figure 5.4 (a) Photocatalytic product yields (after 15 h of 254 nm UV
irradiation) from two different catalytic surfaces. (b)
Energy band alignment of anatase TiO
2
, Au, and the
relevant redox potentials of CO
2
and H
2
O under 254 nm
UV illumination.
82
Figure 5.5 Energy band alignment of anatase TiO
2
, Au, and the
relevant redox potentials of CO
2
and H
2
O under 365 nm
UV illumination.
83
Figure 6.1 3D staircase structure fabricated in the MWNT surface. 89
Figure 6.2 (a) Concentric cylindrical structures patterned using the
laser burnout method. (b) Close-up image showing a slight
undercut profile.
89
Figure 6.3 Square arrays patterned with laser powers of (a) 730W
and (b) 1900W.
90
Figure 6.4 (a) Raman intensity vs. height profile before and after
1000μW laser exposure. (b) Relation between burnout
depth and laser power exposure.
91
Figure 6.5 (a) SEM image showing the change in density of nanotubes
after laser irradiation. (b) High magnification image
showing the nanoparticle substructure.
92
Figure 6.6 Intensity of the G band Raman mode as a function of laser
power.
93
xii
Abstract
This thesis is dedicated to exploring the potential applications of plasmonic metal
nanoparticles and understanding their fundamental enhancement mechanisms.
Photocatalysis and solar energy conversion are the two main topics investigated in this
work. In Chapter 2, we demonstrate the growth of a variety of carbonaceous materials
by plasmonic heating. When the metal nanoparticles are irradiated with a laser near their
plasmon resonant frequency, a localized region at high temperature is achieved due to the
ohmic losses at the plasmon resonance. In this plasmon resonant chemical vapor
deposition (PRCVD) process, high temperatures are created at the surface of the
plasmonic nanoparticles to trigger the dissociation of gaseous precursors, (e.g. carbon
monoxide), which results in the deposition of amorphous carbon, graphitic carbon, and
carbon nanotubes. The formation of iron oxide nanocrystals is also observed at the
beginning of the reaction due to a trace amount of iron pentacarbonyl in the CO feed gas.
The growth of iron oxide at the surface of gold nanoparticles forms a new type of
composite catalyst, Au nanoparticle/Fe
2
O
3
, which also catalyzes the growth of carbon
nanotubes through this plasmon excitation process. The real time temperature
dependence and sequential growth of different carbonaceous and metal oxide materials
xiii
are monitored and characterized by Raman spectroscopy and infrared spectroscopy.
Additionally, pre-defined microstructure geometries of crystalline iron oxide and carbon
nanotubes are demonstrated by rastering the focused laser spot during the growth process
in a controlled fashion.
In Chapter 3, the concentrations of gas phase reaction products are observed in real
time using mass spectrometry, which is used to evaluate the performance of Au
nanoparticle/Fe
2
O
3
composite catalysts. This new plasmonic composite catalyst exhibits
an excellent catalytic ability in the CO oxidation reaction, which exceeds that of the Au
nanoparticles and Fe
2
O
3
alone. This indicates that this reaction is not driven solely by
thermal (plasmonic) heating of the gold nanoparticles, but relies intimately on the
interaction of these two materials. This hybrid plasmonic nanoparticle catalyst and
PRCVD method open up new possibilities in the local chemistry, enabling new growth
pathways of materials, not possible using standard CVD methods with uniform heating.
In addition to photocatalysis, we also explored plasmonic enhancement of solar
energy conversion. In Chapter 4, plasmonic gold nanoparticles are incorporated into dye
sensitized solar cells (DSSCs) by electron beam deposition. Increased photocurrents are
observed due to the thin layer of plasmonic gold, which results in a 45 % increase in the
xiv
cell’s power conversion efficiency. This enhancement is attributed to the strong
plasmon-induced electric field from the presence of gold nanoparticles, as indicated by
electromagnetic finite-difference-time-domain (FDTD) simulations. Additionally, the
photoluminescence spectra of the TPBP dye molecule rule out the mechanism of plasmon
energy transfer through a Forester resonance process. Another potential application of
plasmonic nanoparticles that we have explored is solar fuel production. In Chapter 5, we
demonstrate an enhancement of solar methane production by the reduction of aqueous
carbon dioxide (CO
2
) in the visible wavelength range. The underlying enhancement
mechanisms of the Au nanoparticle/TiO
2
-catalyzed photoreduction of CO
2
are
investigated by irradiating several different sample configurations with a wide range of
wavelengths. Based on these results, we attribute the plasmonmic enhancement to the
local electric field enhancement. However, several questions remain open and further
studies will be required in order to obtain a deeper and more quantitative understanding
of the plasmonic enhancement process. Some future directions include exploring the
dependence of the doping concentration in the TiO
2
, the chemical state of the catalytic
surface, and how the plasmonic nanoparticles perform in co-catalyst systems (e.g., TiO
2
-
Fe
2
O
3
). Chapter 6 describes a side project carried out at the beginning of my graduate
xv
work. In this Chapter, we report a novel method for creating three-dimensional carbon
nanotube structures from dense, vertically-grown carbon nanotube forests. The minimum
power density for burning carbon nanotubes is also determined in this study.
1
Chapter 1: Introduction
1.1 Background
The development of nanotechnology and nanoscience is a key to accessing the other
side of beauty in nature. The influence of “Nano” has become widespread and has
resulted in revolutionary changes in many different disciplines, including chemistry,
physics, mathematics, biology, and engineering nanoscience, impacting not only our
fundamental understanding but also practical applications. After the birth of quantum
mechanics, the motion of electrons started to become well understood. However,
nanoscience (10
-9
meter) did not reveal its veil until the innovation of advanced
microscopies such as scanning tunneling microscopy (STM), atomic force microscopy
(AFM), and Transmission electron microscopy (TEM), which allowed researchers to
probe nanoscale systems/materials otherwise constrained by the optical diffraction limit.
Further development of advanced tools for synthesizing and characterizing
nanomaterials has lead to a more complete understanding of the vast differences between
nanoscale systems and their bulk material counterparts. Nanomaterials possess many
different properties from what we see on the macroscale. For example, high surface area
2
is one of the common and unique physical characteristics of nanomaterials. Carbon
nanotubes are one of the most well-known representatives of high surface area
nanomaterials, which can be depicted as a rolled-up sheet of graphite. The high surface-
to-volume ratio of carbon nanotubes results makes them very sensitive to the changes in
their ambient environment, which makes carbon nanotubes good candidates for gas
sensing applications[20]. Additionally, higher electron motilities can be achieved in
nanomaterials due to ballistic transport, commonly observed in nanotubes/nanowires
when the mean free path of the electron is larger than the length of the
nanotubes/nanowires.
In addition to carbon based materials, metallic materials have also been strongly
influenced by the development of nanotechnology in ways that hidden or absent in bulk
form. For example, significant catalytic activity can even be seen in noble metals when it
is on the scale of 1 nanometer[44]. Gold is considered to be a poor catalyst for most
applications, but when reduced to nanoscale dimensions (<5nm) and supported on metal
oxide semiconductors, it becomes a very reactive catalyst, particularly for oxidation of
carbon monoxide at low temperatures [11, 17, 52, 53, 96, 109]. This reduction in size to
the nanoscale opens up new degrees of freedom to design or redesign the nature of
3
catalysts by tuning the sizes of metal particles and the types of supporting metal oxide
materials.
In addition to changes in the chemical and catalytic properties of gold nanoparticles,
other interesting phenomena also occur when small metal nanoparticles interact with
incident light. Most notably is the surface plasmon resonance phenomenon, which will be
discussed in the following paragraphs in terms of the scattering cross section and
absorption coefficient of the nanoparticles.
Color changes in small metal nanoparticles due to plasmon absorption have been
utilized for over one thousand years. For example, the pigments used in the stained glass
in the Cathedral of Notre Dame shown in the Figure 1.1a are gold nanoparticles. Over
two thousand years ago, in Roman times, ancient people dyed the Lycurgus Cup
(fabricated in the 4
th
century AD) with gold and silver nanoparticles at a ratio of 3:7,
respectively. This cup appears as a different color when illuminated from behind and in
front (reflection: green, transmission: red) as shown in the Figure 1.1b
(
a)
(
b)
4
Figure 1.1 (a) Cathedral of Notre Dame. (b) The Lycurgus Cup.
These color changes in metal nanoparticles are due to the plasmon resonant
absorption from the strong interaction of incident light with the conduction electrons at
the surface of the metal nanoparticles That is also recognized as the surface plasmon
resonance (SPR) phenomenon. Figure 1.2 shows a color comparison of bulk gold and
nano-sized gold with different diameters. Bulk gold (on the left) shows its normal
yellowish color but the nano-sized gold (on the right) exhibits reddish color instead due
to plasmon resonant absorption. Additionally, gold nanoparticles in Figure 1.2b also
reveal a diameter dependence of the plasmon resonance absorption. Besides changes in
color, the collective oscillation of charges induced by the plasmon resonanance also bring
two other important properties such as localized plasmon-induced electric field and high
temperature. Local plasmon-induced electric fields, exist at and in close proximity to the
surface of plasmonic nanoparticles, can enhance incident electric filed intensity by
(a) (b)
5
several orders of magnitude. This enhancement from Plasmon-induced fields has been
used in many fields including surface enhanced Raman spectrospcopy (
SERS)[69]. Local high temperatures are also generated during the oscillation of
plasmonic charge producing plasmonic heating, which has been applied to photothermal
cancer therapy [33, 76].
Gold nanopartles (NPs) are the main plasmonic NPs utilized and discussed in this
thesis. There are several advantages of choosing gold nanoparticles in this study, For
example, the plasmon resonance absorption of gold lies in the visible wavelength range
and gold nanoparticles are more stable in air than other plasmonic metals (e.g. silver).
After this brief discussion of the background nano-size metal nanoparticles, the following
section will be dedicated to more fundamental description of surface plasmons in metalic
systems in terms of different modes of surface plasmon resonance, surface plasmon
polaritons (SPP), and localized surface plasmons (LSP)
(
a)
(
b)
6
Figure1.2 (a) Bulk gold (b) Gold nanoparticles with various diameters.
1.2 Surface Plasmons in Metal Naoparticles
Surface Plasmon Polaritons
When light is incident on a metal surface with appropriate incident angle and energy,
the energy of the photons will be converted and stored in the form of fluctuations of
density of conduction electrons at interface of planar metal (slab) and dielectrics, which is
as known as a surface plasmon. Figure 1.3a shows a surface Plasmon propagating at the
metal/dielectric interface resulting in the oscillation of charge density. Also, because the
surface plasmon generated and confined at the interface as a two dimensional wave,
which makes electric field (fringe pattern) decreases exponentially on both sides of
dielectric and metal.
By solving Maxwell’s equations, the maximum surface plasmon frequency is given
by
, where
is the plasmon frequency in bulk material and
is the
(a) (b)
7
dielectric constant of environment surrounding/covering the metal surface (e.g.
=1;
). The dispersion relation of the surface plasmon shown in the
Figure 1.3b describes that at small wave vectors region, the dispersion relation is very
similar to the light line, but with the increase of the wave vector, the frequency (energy)
of surface plasmon begins to drift way from the guide of light line, starting to bend over
and stay below the maximum frequency (
).
Figure1.3 (a) Surface charge resonance and plasmon-induce electric field distribution
at metal and dielectric interface. (b) Dispersion relation of a thin film metal.
Localized Surface Plasmon Resonance
Localized surface plasmon is another type of plasmon resonance. For small metal
nanoparticles, the incident electromagnetic field can be considered homogeneous over the
entire nanoparticles, which is known as a quasi-static approximation. Because the
(
a)
(
b)
(a)
(b)
8
nanoparitles are much smaller than the wavelength of light, conduction electrons in the
metal nanoparticles will experience uniform distribution of a static electric field from the
incident light. Therefore, instead of traveling and propagating at the metal/dielectric
interface, the conduction electrons in the metal nanoparticles will oscillate as a unit group
in to response to the electric field of incident light shown in the Figure 1.4a. There are
many theoretical works describing this localized surface plasmon phenomenon[101]. In
the early part of the 20
th
century, this collective movement of conduction electrons was
explained by Gustav Mie by solving the Maxwell’s equation for spherical nanoparticles.
To this study, the optical cross section of small spherical metal nanoparticles can be
understood by the Mie theory. Figure 1.4b shows a small metal nanoparticle incident with
a flux of electromagnetic radiation, which displays (under the plasmon
resonancecondition) a much larger optical cross section compared to its physical
geometric cross section. The scattering and absorption cross sections are be given by[51]
(1.1)
where
(1.2)
9
is the polarizability of the particle. Here, V is the particle volume,
is the dielectric
function of the metal particle and
is the dielectric function of the surroundings around
metal nanoparticles. When
, occursthe particle polarizability (α) becomes
very large, which triggers the formation of localized surface plasmon resonance. There
are many applications exploiting both of these phenomena (surface plasmon polariton
(SPP) and localized surface plasmon resonance (LSPR)), which will be discuss later in
this thesis in the section on current applications.
Figure 1.4 (a) A schematic of the localized surface plasmon. (b) Light scattering and
absorption of a spherical metal nanoparticle.
Oscillations of conduction electrons can be described by simple harmonic motion,
which can be written as
(
a)
(
b)
(a) (b)
10
(1.3)
where m and e are the mass and charge of an electron, respectively, c is the damping
coefficient associated with energy loss (e.g., the heat/phonon emission; known as
plasmonic heating). The constant k delineates a storing force in this system due to the
electrostatic force between the metal ion cores and conduction electrons, which is
analogous to the spring constant in the mechanical simple harmonic oscillation system.
However, in the metal system, electrons exist in Fermi Sea of electrons and are not bound
to any specific ion core. The Drude model is used to describe the behavior of free
electrons in metals (not including interband transitions). By calculating the magnitude
and phase of the oscillation movement, we can obtain a frequency dependent complex
dielectric constant of metal, which includes a real part (
) and an imaginary parts (
)
written as
(1.4)
where
is as known as the plasma frequency of bulk material, written as
, and the associated with the damping in the system is represented by
. Figure 1.5 shows the experimental data of the real and imaginary parts of the
dielectric function of gold together with fitting curves from two theoretical models; the
11
Drude model and the hybrid Drude-Lorentz model. The Drude model exhibits good
agreement with the experimental data (solid line) in the long wavelength region (>650nm)
but it breaks down in the short wavelength region due to the contribution of interband
transitions. With the incorporation of the Lorentz model, which is a well-known model
for visualizing the interaction of atoms and incident fields in dielectric systems, the
Drude-Lorentz model shows a better approximation in the both short and long
wavelength regions due to the consideration of the interband electronic transitions of gold.
The function of dielectric is a constant associated with the response of materials under
interaction with electric fields, for examples, the real part of the dielectric function is
mostly related to the energy storing in the material and the imaginary part of dielectric
function is related to the energy loss in the material. Metal systems are considered
relatively lossy at the resonance condition particularly when compared to dielectric
materials. This is because of the dissipations from the re-emission and heat generation on
resonance. As a result, the quality factor (Q) of these metals are quite small. In this thesis,
we exploit the high temperatures arising from plasmonic heating to drive and improve the
catalytic reactions, including carbon nanotubes (CNTs) synthesis and carbon monoxide
(CO) oxidation.
12
Figure1.5 (a) Real, and (b) Imaginary part of the complex dielectric function part of
gold plotted with prediction of the Drude model and Drude-Lorentz model.[81]
1.3 Current Applications of Plasmonic Nanoparticles
In addition to the utilization of metal nanoparticles as pigments, strongly plasmon
resonant nanoparticles have been applied to a wide range of applications across many
fields. In this section, we will briefly discuss a few major areas of interest utilizing these
plsamonic nanoparticles.
Sub-wavelength lithography
Nanolithography is always pushing the limits of the nanoscale due to the strong desire
to achieve higher densities devices (MOSFET) per chip. There are several developing
alternative lithographic methods aimed to avoid the diffraction limit of light, such as
electron-beam lithography, nano-imprint lithography[21, 72], scanning probe
lithography[22], and dip-pen lithography[49, 54]. Since photolithography is still the main
(
a)
(
b)
(a) (b)
13
fabrication throughout in the semiconductor industry, Many researchers are putting
efforts to find a feasible path toward overcoming the diffraction limit of light. Plasmon is
considered a good candidate for this purpose, by enabling the manipulation of light on the
scale of few nanometers. The basic mechanism can been understood by comparing the
dispersion relation of light (w=ck) with the surface plasmon dispersion, as shown in
Figure 1.6a. At the same energy (or frequency), the surface plasmon has a larger k
wavevector (smaller wavelength) and typically this plasmon mode can be excited through
a grating or a rough surface of metal to overcome the momentum mismatch of the
photons and plasmons. Figure 1.6b shows an array of dots 90nm in diameter fabricated
by plasmon lithography with an aluminum metal mask under 356nm UV irradiation.
Furthermore, according to the theoretical calculation, a 16.5nm line resolution can be
achieved with this method.
Figure 1.6 (a) Dispersion relation of light and surface plasmon mode. (b) Subwavelength
90nm dot arrays fabricated using 365nm UV light with a 170nm period.[102]
(
a)
(
b)
(a) (b)
14
Surface Enhance Raman Signal (SERS)
Surface plasmons have been studied by many research groups across many fields[3,
103, 117]. The intense electric fields produced near plasmon resonant metallic
nanoparticles have been utilized in surface enhanced Raman spectroscopy (SERS) for
over 30 years, yielding enhancement factors as high as 10
14
[65, 66, 69]. Numerical
simulations have predicted SERS enhancement factors up to 10
10
[88, 91]. Surface
plasmon resonance is used by biochemists to study the mechanisms and kinetics of
ligands binding to receptors.
Figure 1.7 shows one of the demonstration of surface enhance Raman spectroscopy
of carbon nanotubes using silver nanoparticles carried out Kumar et al.[69]. In this work,
a sample with 6nm silver shows enhanced Raman signal of carbon nanotubes by a factor
of 335 with respect to the sample before silver deposition. This enhancement is attributed
to the surface plasmon resonance of the Ag nanoparticles, which couples the incident
light into the nanotube very effectively. Besides carbon nanotubes, this surface enhance
Raman signal can also be used for signal molecular examination.
15
Figure 1.7 (a) SEM image of carbon nanotubes covered with a film of silver
nanoparticles. The white circle indicates the size and location of the laser spot. (b)
Raman spectra taken at the laser spot position indicated in (a) before and after Ag
deposition.
Photothermal Cancer Therapy
Because metals are lossy systemswhich tend to produce heat from ohmic loss,
especially within the resonance condition, several applications make use of the high
temperatures generated from plasmonic heating. One of interesting areas utilizing
plasmonic heating is in the field of biotechnology. Here, plasmonic nanoparticles with a
strong photothermal response have been exploited for the light-induced destruction of
cancer cells and tumors with conjugations of canner-specific antibodies. For example,
Loo et al. have demonstrated a novel nanoshell-based all-optical platform technology for
integrating cancer imaging and therapy applications. Their results are shown in Figure 1.7.
Top row shows scatter-based darkfield imaging of HER2 expression, middle row is cell
16
viability assessed via calcein staining, and bottom row presents silver stain assessment of
nanoshell binding in Figure 1.8. Cytotoxicity was observed in cells treated with a NIR-
emitting laser following exposure and imaging of cells targeted with anti-HER2
nanoshells only. Note increased contrast (top row, right column) and cytotoxicity (dark
spot) in cells treated with a NIR emitting laser following nanoshell exposure (middle row,
right column) compared to controls (left and middle columns). As a result,
immunotargeted gold nanoshells can provide better scattering contrast for imaging while
also exhibiting sufficient absorption to enable effective photothermal therapy.
Figure 1.8 Combined imaging and therapy of SKBr3 breast cancer cells using HER2-
targeted nanoshells. [76]
17
Plasmon Enhanced Thin Film Silicon Based Photovoltaic
Most recently, the other application of broad scientific and public interest is solar
energy conversion using these plasmonic nanoparticles. Here, light absorption of a host
material is strongly enhanced by the strong scattering and plasmon-induced electric fields
generated by the plasmon resonance phenomenon. Ferry et al. utilized the controlled
nanopatterning of a silver (Ag) back contact in improve the performance of a n-i-p
amorphous-Si:H solar cell, which exhibits an efficiency increase from 4.5% to 6.2%, with
a 26% increase in the short circuit current density as shown in the Figure 1.9.[31] In this
thesis, the enhancement of light absorption is utilized to improve the generation of
excitons to improve the photocurrent in a dye sensitized solar cell (DSSC) as well as the
yield of solar methane production.
Figure 1.9 (a) SEM image showing a FIB cross section of a fully fabricated n-i-p a-
Si:H solar cell grown on the patterned back contact. (b) Best J-V measurements of the
flat reference and patterned n-i-p a-Si:H cells. Inset shows the cell characteristics for
each device. [31]
(
b)
(
a)
(a) (b)
18
1.4 Contents of this Thesis
Chapter 2: Laser Directed Growth of Carbon-Based Nanostructures by Plasmon
Resonant Chemical Vapor Deposition
In this chapter, we exploit the strong plasmon resonance of gold nanoparticles in the
catalytic decomposition of CO to grow various forms of carbonaceous materials.
Irradiating gold nanoparticles in a CO environment at their plasmon resonant frequency
generates high temperatures and strong electric fields required to break the CO bond. By
varying the laser power, exposure time, and gas flow rate, we deposit amorphous carbon,
graphitic carbon, and carbon nanotubes. The formation of iron oxide nanocrystals
catalyzes the growth of carbon nanotubes. Predefined microstructure geometries are
patterned by moving the focused laser spot during the growth process, forming suspended
single-walled carbon nanotube structures. Raman spectroscopy, energy dispersive X-ray
spectroscopy, and transmission electron microscopy are used to characterize the resulting
material. The localized nature of the plasmonic heating enables growth of these materials,
while the underlying substrate remains at room temperature.
19
Chapter 3: Plasmon Resonant Enhancement of Carbon Monoxide Catalysis
In this chapter, we irradiate gold nanoparticles at their plasmon resonance frequency
to create immense plasmonic charge and high temperatures, which are then used to drive
catalytic reactions. By integrating strongly plasmonic nanoparticles with strongly
catalytic metal oxides, significant enhancements in the catalytic activity are achieved.
Here, we study the plasmonically driven catalytic conversion of CO to CO
2
by irradiating
Au nanoparticle/Fe
2
O
3
composites. The reaction rate of this composite greatly exceeds
that of the Au nanoparticles or Fe
2
O
3
alone, indicating that this reaction is not driven
solely by the thermal (plasmonic) heating of the gold nanoparticles, but relies intimately
on the interaction of these two materials. A comparison of the plasmonically driven
catalytic reaction rate with that obtained under uniform heating shows an enhancement of
at least two orders of magnitude.
Chapter 4: Plasmon Enhancement of Dye Sensitized Solar Cell
In this Chapter, we have successfully demonstrated plasmonic enhancement of dye
sensitized solar cells with up to 45 % increase in the cell power conversion efficiency by
incorporating a layer of island-like plasmonic gold nanoparticles. The UV-Vis absorption
spectra of island-like plasmonic films show good agreement with the IPCE spectrum,
20
indicating a higher enhancement at longer wavelengths. Finite-difference-time-domain
(FDTD) simulations are also carried out to investigate the distribution of the plasmon-
induced electric field in the 5nm island-like gold film. Furthermore, we believe the
enhancement observed in the dye sensitized solar cell can be attributed to the plasmon-
induced electric fields, which increase the exciton generation rate and dissociations. The
excitation spectra rule out the mechanism of plasmon energy transfer through Forester
resonance energy transfer.
Chapter 5: Plasmonic Enhancement of Methane Production
A systematic study of the mechanisms of Au nanoparticle/TiO
2
-catalyzed
photoreduction of CO
2
and water vapor is carried out over a wide range of wavelengths.
When the photon energy matches the plasmon resonance of the Au nanoparticles (free
carrier absorption), which is in the visible range (532 nm), we observe a 17%
enhancement in the photocatalytic activity due to the intense local electromagnetic fields
created by the surface plasmons of the Au nanoparticles. These intense electromagnetic
fields enhance sub-bandgap absorption in the TiO
2
, thereby enhancing the photocatalytic
activity in the visible range. When the photon energy is high enough to excite d band
electronic transitions in the Au, in the UV range (254 nm), a different mechanism occurs
21
resulting in the production of additional reaction byproducts, including C
2
H
6
, CH
3
OH,
and HCHO. This occurs because the energy of the d band excited electrons lies above the
redox potentials of the additional reaction byproducts CO
2
/C
2
H
6
, CO
2
/CH
3
OH, and
CO
2
/HCHO. In visible wavelength range, the dominant catalytic mechanism involves
enhanced sub-bandgap transitions in TiO
2
and, thus, occurs at the TiO
2
surface. The
mechanism dominating in the UV range is attributed to the interband transitions of Au
and TiO
2
and, therefore, occurs at both Au and TiO
2
surfaces.
Chapter 6: Rapid Prototyping of Three-Dimensional Microstructures from Multi-
Walled Carbon Nanotubes
This is a side project carried out at the beginning of my PhD, which is about to create
a three-dimensional microstructure from carbon nanotube forest. In this chapter, we
report a novel method for creating three-dimensional carbon nanotube structures,
whereby a focused laser beam is used to selectively burn local regions of a dense forest of
multi-walled carbon nanotubes. Raman spectroscopy and scanning electron microscopy
are used to quantify the threshold for laser burnout and depth of burnout. The minimum
power density for burning carbon nanotubes in air is found to be 244W/m
2
. We create
various 3-dimensional patterns using this novel method, illustrating its potential use for
22
the rapid prototyping of carbon nanotube microstructures. Undercut profiles, changes in
nanotube density, and nanoparticle formation are observed after laser surface treatment
and provide insight into the dynamic process of the burnout mechanism.
23
Chapter 2: Laser Directed Growth of
Carbon-Based Nanostructures by Plasmon
Resonant Chemical Vapor Deposition
2.1 Introduction
Surface plasmon resonance in metal nanoparticles has been investigated for many
years [57]. Many applications, such as surface-enhanced Raman spectroscopy (SERS)
and surface plasmon-enhanced fluorescence spectroscopy (SPFS), for example, are made
possible by the strong electric fields generated by the plasmon resonance phenomenon [4,
61]. Sub-wavelength plasmon-assisted lithography can be achieved using nanoparticle
arrays in conjunction with conventional UV lithography [9, 100]. Nanoscale integrated
optics exploit the short wavelength of plasmonic waves to reduce device dimensions by
an order of magnitude relative to the optical wavelength of similar frequency [5].
Plasmon resonant nanoparticles have also created new applications in chemistry and
biology [75], including non-invasive cancer treatment therapy [93].
Nanoscale materials have opened up new degrees of freedom in the design of
catalysts with efficiencies and selectivities exceeding those of their bulk counterparts. For
24
example, gold is considered to be a poor catalyst for most applications. However, gold
nanoparticles on metal oxide supports have recently demonstrated high catalytic activity
[11, 42, 43]. This work and others have established the great potential for new chemical
pathways at the nanoscale.
Here, we present a method for the laser-directed growth of carbonaceous
nanomaterials that exploits the surface plasmon resonance phenomenon in metal
nanoparticles. In this plasmon resonant chemical vapor deposition (PRCVD) process, the
high temperatures reached through plasmon resonant excitation of nanoparticles, together
with their catalytic properties, enable deposition of amorphous carbon, graphitic carbon,
iron oxide, and carbon nanotubes. The large temperature gradients generated by the
nanoparticles enable two very different chemical processes to take place side-by-side, i.e.,
the CO dissociation and the formation of iron oxide and carbon nanotubes. This localized
chemistry enables new growth processes to occur that cannot be achieved by uniform
heating and standard CVD. Moving the position of the focused laser spot during the
growth process makes possible the growth of 3-dimensional microstructures, such as
suspended carbon nanotubes. The localized nature of the nanoplasmonic heating provides
a means for growing carbon nanotubes without heating the remainder of the substrate,
25
thus enabling the integration of carbon nanotubes with new classes of substrate materials
and devices.
2.2 Experimental Procedure
2.2.1 Preparation of Plasmonic Substrate
There are two methods we can use to prepare plasmonic substrates. One involves a
commercial gold nanoparticles colloid with diameter of 20min from Ted Pella, Inc.. The
surface of substrate (e.g. silicon) is first modified to be hydrophilic by O
2
plasma. A layer
of poly-L-lysine is then spun on the O
2
plasma treated substrate as an adhesion layer.
This substrate is then baked at 120
o
C for 30 minutes after a Deionized (DI) water rinse.
After the baking step, we can put a drop of the gold colloid solution on this prepared
substrate and the density of Au nanoparticles can be controlled by the length of time
before rinsing with DI water.
The second method we use to prepare plasmonic substrates uses electron-beam
deposition of Au thin films. This process is known to produce island-like growth of
strongly plasmonic nanoparticles separated by only a few nanometers [69]. Gold
nanoparticle arrays are deposited on glass and silicon substrates by electron-beam
evaporation of 5nm thick films of Au.
26
2.2.2 Plasmon Resonant Chemical Vapor Deposition (PRCVD) System Setup
In our experimental setup, a 532nm 5W Spectra Physics solid state laser is collimated
and focused through a Leica DMLM microscope and used to irradiate the plasmonic
substrates described above. A 50X long working distance objective lens with NA=0.5
and spot size=1.25μm is used to irradiate the sample while carbon monoxide (CO) is
flowed through the reaction chamber (Linkam stage). Before laser irradiation, the
reaction chamber is purged by flowing CO for 5 minutes. An automated XYZ translation
stage (Prior, Inc.) provides 0.1μm precision and is also under computer control.
2.2.3 Sample Characterization
Samples were characterized in a Renishaw inVia micro-Raman spectrometer. High
magnification images and electron diffraction patterns of micro- and nano-structures
grown in this way were taken in a JEOL JEM-2100 LaB6 transmission electron
microscope (TEM). The temperature dependence of Raman spectra were calibrated in a
Linkam THMS temperature controlled stage.
27
2.3 Results and Discussion
2.3.1 PRCVD of Iron Oxide
Figure 2.1 shows the catalytic decomposition of CO and the subsequent deposition of
solid (crystalline) material. In situ Raman spectroscopy, taken during laser exposure at a
power of 12.3mW, is used to identify the material deposited and reveals an interesting
transition that occurs during the deposition process. During laser exposure, broad Raman
peaks appear at 1344 and 1581 cm
-1
, which correspond to the D and G band Raman
modes of amorphous carbon [30]. The Raman spectra exhibit strong amorphous carbon
peaks, during the first few seconds of laser irradiation, that decrease in intensity with time
and then vanish after 25 seconds. After 60 seconds, a different peak is observed at 1289
cm
-1
, which we attribute to the formation of iron oxide. Raman spectra, taken post growth,
show peaks at 227, 246, 301, 292, 411, 500, 612, and 1320 cm
-1
, all of which are
consistent with the hematite phase of Fe
2
O
3
.[25, 80, 85] The iron and oxygen
composition of these crystallites was verified by energy dispersive X-ray spectroscopy
(EDX). The Raman, EDX, and high resolution transmission electron microscopy
(HRTEM) results corroborating the structure and composition of these Fe
2
O
3
nanocrystals are shown in the later section. Although ultra high purity CO was used in
28
these experiments, some trace amounts of Fe are present from the gas cylinder. It is
remarkable that such rapid growth of Fe
2
O
3
can be achieved without any intentional
feedstock. The 1289 cm
-1
Raman peak, observed during the PRCVD growth process, is
significantly downshifted with respect to the iron oxide peak taken after the growth
(Figure 2.1c), which appears at 1320 cm
-1
. This downshift indicates the high temperatures
reached during the PRCVD process, which results in the thermal expansion of the lattice
and, hence, softening of the Raman active phonon mode [25, 80]. The large Raman
signals of amorphous carbon observed in the first few seconds of the deposition are
attributed to the surface enhanced Raman spectroscopy (SERS) phenomenon and are not
indicative of the amount of material present in the focal volume [61]. Figure 2.1(c) shows
an in situ Raman spectrum of the amorphous carbon taken during the deposition (after the
first 12s) and a post-growth Raman spectrum of the faceted crystal grains after a 130s
deposition.
29
Figure 2.1 (a) SEM image of iron oxide nanocrystals deposited on Au nanoparticles after
laser irradiation in a CO environment. (b) In situ Raman data taken during laser exposure.
(c) Raman spectra of amorphous carbon deposited during the first 12s of irradiation and
iron oxide deposited after 130s.
(c)
(a)
laser spot
Iron oxide
(b)
0 20 40 60 80 100 120 140
0
100
200
300
400
500
600
700
800
Intensity (counts)
Amorphous Carbon
Iron Oxide
1581 (cm
-1
)
1289 (cm
-1
)
Time (second)
1200 1300 1400 1500 1600 1700 1800
Intensity (a.u.)
Raman Shift (cm
-1
)
Amorphous Carbon Peak
(first 12s)
Iron Oxide Peak (after growth)
30
2.3.2 PRCVD Carbon Nanotubes
By increasing the laser power to 23 mW, the exposure time to 10 minutes and the CO
flow rate to 4000 sc cm, we observe the formation of graphitic carbon and carbon
nanotubes. The SEM images in Figure. 2.2 show a porous microstructure grown on the
substrate in the region of the focused laser spot. The high magnification SEM image,
shown in Figure 3.2b, reveals a high density of carbon nanotubes interspersed with iron
oxide nanoparticles. The synthesis of carbon nanotubes was verified by Raman
spectroscopy, as shown in Figure. 2.2c. Several Raman signatures, unique to carbon
nanotubes, are observed, including radial breathing modes (RBMs) and splitting of the G
band, with G
+
occurring at 1590 cm
-1
and G
-
at 1566 cm
-1
. The RBM peaks shown in
Figure 2.2c appear at 140 and 166 cm
-1
, corresponding to nanotube diameters of 2.1 and
1.5nm, respectively, by the relation . 27 / 204
t RBM
d [82] The intensity of the D
band, occurring at 1335 cm
-1
, gives a measure of the amount of disorder and defects in
carbon nanotubes.[30] The small D/G band intensity ratio indicates that these are highly
crystalline nanotubes with few defects.
Figure 2.2d shows the time evolution of the Raman spectra taken during the growth
process. During the first 75 seconds, a broad G-band, characteristic of amorphous carbon,
31
is found to decrease in intensity over time and vanish. After 230 seconds of laser
exposure, sharp G-band Raman modes, characteristic of carbon nanotubes, are seen
increasing with exposure time. A consistent downshift of almost 30 cm
-1
(from 1578 cm
-1
to 1550 cm
-1
) is observed over the course of the deposition. These large downshifts
correspond to temperatures as high as 1180
o
C[18]. This indicates that the plasmonic
nanoparticles are increasing in temperature during the growth process due to
nanoparticles’ decoupling from the substrate. This is consistent with the image in Figure
3.2b, which shows that the nanoparticles are raised up off the substrate and are
intertwined in the resulting carbonaceous micro/nanostructure.
In order to prove that the results shown in Figures 2.1 and 2.2 are caused by a
resonant phenomenon, we performed the same experiment with 785nm light. Here, we
observed no decomposition of CO and no deposition of material on the substrate, even at
laser powers 10 times higher than those used with 532nm. Therefore, in order for
sufficient electric fields to be achieved and significant heating to occur, the laser energy
must match the plasmon resonance frequency of the nanoparticles. Standard chemical
vapor deposition (CVD) with uniform heating was also tested by flowing CO over the
same nanoparticles at 900
o
C. This resulted in no carbon deposition of any form. From
32
this, we conclude that the plasmonically driven catalytic process shown in Figures 2.1
and 2.2 is not possible using standard bulk catalysts and uniform heating and relies on the
large temperature gradients generated by the local plasmonic heating of the nanoparticles.
In the PRCVD growth process, the plasmonic nanoparticles decompose CO
catalytically to produce carbon and oxygen, by the reaction 2CO 2C + O
2
. This
reaction was verified by the detection of O
2
in the reactant byproducts using mass
spectrometry. From prior studies of CO oxidation[8, 14, 45, 46, 62, 74, 105, 113], the
catalytic dissociation of CO on Au nanoparticles is known to occur above 300
o
C. We
attribute this plasmon-induced catalytic dissociation of CO to the elevated temperatures
produced in the nanoparticles due to Joule-like heating associated with the surface
plasmon excitation. The free carbon atoms and molecular oxygen solidify to form
amorphous carbon, iron oxide, graphite, and carbon nanotubes, in the low temperature
regions of the substrate. In this process, the iron oxide catalyzes the growth of carbon
nanotubes.[67] We believe that the condensation of atomic carbon and molecular oxygen
removes any good reverse reaction pathways, enabling a very different chemical reaction
to be driven forward efficiently.
33
Figure 2.2 (a,b) SEM images of carbon-based nanostructures synthesized by irradiating
Au nanoparticles in a CO environment. (c) Raman spectrum of the material shown in (a)
and (b), exhibiting typical features of carbon nanotubes. (d) In situ Raman data taken
during laser exposure.
2.3.3 Laser Writing of PRCVD Suspended Carbon Nanotubes
One of the important aspects of this technique is that the position of the focused laser
spot can be manipulated during the growth process to pattern microstructures in a
controllable way. Figure 2.3 shows tilted SEM images between two sequential growth
locations. At each location, the PRCVD growth was performed at a laser power of 23mW
0 50 100 150 200 250 300 350 400
0
1000
2000
3000
4000
5000
Amorphous Carbon
Carbon Nanotubes
Raman Intensity (counts)
Time (second)
Amorphous Carbon
Carbon Nanotubes
(a)
(b)
(c) (d)
Iron Oxide
oxide
CNTs
nanotubes
RBM
D band
G
+
band
G
-
34
in a CO gas flow rate of 4000sc cm for an exposure time of 7 minutes. A suspended
carbon nanotube can be seen spanning the space between the two growth regions, which
are separated by 5m. The white arrow indicates the direction of laser movement at a
speed of 500m/s. The conformation of the suspended nanotube to the direction of laser
movement indicates that the nanotubes can be grown controllably in any direction. This
method was repeated between 20 different growth locations, ranging from 2-8m in
separation. Successful growth of suspended nanotubes was achieved in more than half of
these instances. It is remarkable that such taut suspended nanotubes can be grown rather
insensitively to the speed of laser movement. Large temperature gradients have been used
previously to achieve well-aligned nanotube growth[50]. It is likely that the large
temperature gradients generated in the vicinity of the focused laser spot are responsible
for this aligned growth. We also note that the iron oxide nanoparticles move during the
growth process, possibly due to optical trapping or thermal migration[94, 108].
Branching at the ends of the nanotube indicates that this is a bundle of nanotubes. The
Raman spectra of this nanotube bundle exhibit a radial breathing mode at 155 cm
-1
,
corresponding to a nanotube diameter of 1.59 nm. More complex geometries, such as
lines and predefined shapes, can be patterned with this technique, as shown in the
PRCVD iron oxide pattern section.
35
Figure 2.3 (a,b) Tilted SEM images of a suspended carbon nanotube spanning two
sequential growth spots. The white arrow indicates the direction of laser movement. (c,d)
High magnification SEM images taken at the two ends of this suspended carbon nanotube.
36
2.3.4 Material Characterization
Iron oxide
Energy dispersive X-ray spectroscopy (EDX) and micro-Raman spectroscopy
Figure 2.4 shows the energy dispersive X-ray (EDX) spectrum of the
nanocrystalline material grown by the plasmon resonant chemical vapor deposition
(PRCVD) process. The area shown in the SEM image on the right side of this figure was
irradiated to obtain this EDX spectrum. The spectrum shows a predominance of Fe and
O, indicating a crystalline phase of iron oxide. Figure 2.4(c) shows the low energy
Raman spectrum for the nanoscrystalline material produced by the PRCVD process.
These spectra were taken with a 100X objective lens focusing the laser beam to a
diffraction limited spot size. All of the peaks observed, 227, 246, 293, 301, 441, 500, and
612 cm
-1
, correspond to the Raman active vibrational modes of the hematite phase of
Fe
2
O
3
.
37
Figure 2.4 (a) The SEM image of PRCVD grown iron oxide. (b) Energy dispersive
X-ray (EDX) spectrum of nanocrystalline material shown in (a). (c) Raman spectrum
of nanocrystaline Fe
2
O
3
grown by the PRCVD process.
0 1 2 3 4 5 6 7 8
Counts (a.u.)
Fe K
Energy (KeV)
Fe K
Si K
O K
Ck
Fe L
(c)
(b)
(a)
612
500
411( cm
-1
)
293
246
227
301
38
Carborn Nanotubes
The TEM images in Figure. 2.5 show a porous microstructure, similar to that shown
in Fig 3.2, grown by the PRCVD process on a SiN TEM membrane. These images show
carbon nanotubes grown from iron oxide nanoparticles, which provide a seed for the
nanotube growth[67]. From SEM images taken at a tilted angle, a layer of iron oxide
nanocrystals is seen lining the bottom of the coral-like porous microstructure coated with
graphitic carbon and carbon nanotubes. These graphite-coated nanoparticles aggregate to
form clusters during the plasmon resonant CVD process, creating the coral-like porous
structure shown in Figure 2.2(a). Despite its rather messy appearance, the porous
microstructure shown in Figure 2.2(a) is actually a highly active, light-driven catalyst that
integrates strongly catalytic nanostructures (Au nanoparticles) with strongly catalytic
materials (iron oxide).
Figure 2.5 TEM images of iron oxide nanoparticles interspersed with CNTs
39
2.3.5 PRCVD Growth on SU-8 and Pattern Capability
Figure 2.6 shows the PRCVD growth of Fe
2
O
3
and carbon nanotubes on top of
100m wide pillars patterned lithographically in an SU-8 polymer, using a laser power of
0.2W and CO gas flow rate of 4000sc cm. Figure 2.6b shows the micro-Raman
spectrum taken from this growth region. The sharp G band lineshape and the 1320 cm
-1
peak indicate the presence of carbon nanotubes and Fe
2
O
3
, respectively. These results
exemplify the localized nature of the plasmonic heating, which enables the growth of
carbon nanotubes and iron oxide on new classes of substrate materials, with negligible
heating of the underlying substrate.
In order to further demonstrate the ability to patterned pre-defined geometries, we
continuously move the laser spot during the PRCVD growth process. Figure 2.7 shows
SEM images of PRCVD material grown in this way. Figure 2.7a shows the deposition of
a circular pattern 40m in diameter. The structure is observed to delaminate due to the
thermal contraction upon cooling. Similarly, Figure 2.7b shows an SEM image of a right
angle shape, which is also observed to delaminate slightly at the corner. In these growths,
the focused laser spot was moved at a rate of 0.2m/s while irradiating with a laser power
of 170mW in a CO environment. This high power is needed to compensate for the short
40
exposure time. However, deposition can be achieved under low power as mentioned in
the manuscript after a longer laser exposure time.
Figure 2.6 (a) SEM image and (b) Raman spectrum of Fe
2
O
3
and carbon nanotubes
grown on SU-8 polymer.
Figure 2.7 SEM image of (a) circular and (b) right angle PRCVD line growths.
G band
G’ band
1320 cm
-1
RBM
(b)
(a)
(a) (b)
41
2.4 Conclusion
In summary, we have demonstrated synthesis of single-walled carbon nanotubes and
iron oxide nanocrystals using a plasmon resonant chemical vapor deposition process.
Real-time, in situ Raman spectroscopy reveals the sequential growth of different
carbonaceous and metal oxide materials during the plasmon resonant CVD process. A
feasible method for controlling the growth and direction of suspended carbon nanotubes
is demonstrated using the laser as a heat source movable in three dimensions. The
localized heating of plasmonic nanoparticles opens up new possibilities in the local
chemistry, enabling new growth pathways of materials, not possible using standard CVD
methods with uniform heating. In addition, the local heating of plasmonic nanoparticles
enables growth of carbon nanotubes and iron oxide without heating of the entire substrate,
thus enabling the growth of carbon nanotubes on new classes of substrate materials and
devices.
42
Chapter 3: Plasmon Resonant Enhancement
of Carbon Monoxide Catalysis
3.1 Introduction
Macroscopic catalysts are based on traditional chemical pathways. At the nanoscale,
the catalytic properties of materials are often quite different from their bulk counterparts.
For example, gold is considered to be a poor catalyst for most applications. However,
gold nanoparticles on metal oxide supports demonstrate high catalytic activity even at
room temperature[42, 43]. This work and others have established the great potential for
new chemical pathways at the nanoscale. In this work, we investigate the additional
catalytic enhancement produced by the surface plasmon resonance phenomenon.
Surface plasmons have been studied by many research groups across many fields[3,
103, 117]. The intense electric fields produced near plasmon resonant metallic
nanoparticles have been utilized in surface enhanced Raman spectroscopy (SERS) for
over 30 years, yielding enhancement factors as high as 10
14
[65, 66, 69]. Numerical
simulations have predicted SERS enhancement factors up to 10
10
[88, 91]. Surface
plasmon resonance is used by biochemists to study the mechanisms and kinetics of
43
ligands binding to receptors. Plasmonic phenomena have also been explored for
electromagnetic energy transport below the diffraction limit[78], subwavelength optical
devices[77], and lasers[23, 92]. Because the imaginary part of the dielectric function is
dominant, much of the energy in the plasmon resonance is dissipated as heat. This has
been utilized for non-invasive cancer treatment therapy[93]. Both plasmonic heating and
plasmon-induced charge are advantageous when coupled to catalysts.
Metal oxides (e.g., TiO
2
, Fe
2
O
3
, PbO) are very promising photocatalysts for a number
of applications, including solar fuel production, oxidation of pollutants, and anti-
fogging/self-cleaning coatings for windows and lenses. When gold is dispersed as fine
particles (2-5nm) over select metal oxides, it has been found to exhibit exceptionally high
catalytic activity[42, 43, 109], far exceeding that of the metal oxide and gold catalysts
separately[39, 99]. The catalytic oxidation of carbon monoxide has become an important
field of study in and of itself. For gold nanoparticles on transition metal oxides, the
oxidation of CO is an exothermic reaction with extremely low catalytic activation
barriers[45]. In this work, we explore a fundamentally different mechanism of catalytic
enhancement, achieved with plasmonic excitation. By monitoring the temperature of the
reaction (using infrared and Raman spectroscopies), we separate the effects of plasmonic
44
heating from exothermic chemical heating. To determine the catalytic enhancement
factor, we compare the CO reaction rate achieved under laser irradiation with that
observed under uniform heating. We also compare the reaction rate of the Au
nanoparticle/Fe
2
O
3
composite with that of its constituent materials, in order to further
establish the uniqueness of this catalytic enhancement process.
3.2 Experimental Procedure
3.2.1 Micro-reactor System Setup
In order to measure concentration of reactants and product and in-situ temperature
changes during plasmon enhancement CO oxidation. We designed a micro-reactor
system consisting of a gas delivery system, an optical microscope, and an excitation laser
source, as illustrated schematically in Figure 3.1. A 532nm 5W Spectra Physics solid
state laser is collimated and focused through a Leica DMLM microscope. A 50X long-
working distance microscope objective lens with NA=0.5 and spot size=1.25μm is used
to irradiate the sample while carbon monoxide (CO) is flowed through the micro-reactor
system. The byproducts of these plasmon-induced reactions are monitored using a
Pfieffer Omnistar quadrupole mass spectrometer (QMS).
45
Figure 3.1 Schematic diagram of the micro-reactor system.
3.2.2 Silicon-based Microchannel Fabrication
Since the sensitivity of the mass spectrometer is limited by the background vapors,
namely the unreacted reactants, micron-sized channels (similar to microfluidic channels)
etched in silicon[10] are used to reduce the volume of gas flowing through the system to
a 20 x 20m
2
cross-sectional area. The Bosch process, a deep reactive-ion etching, is
applied to excavate the trenches in the PR patterned silicon wafer. Sulphur hexafluoride
(SF
6
) is flown and chemically reacted with Si during an etching step, which is followed
46
by a deposition step with C
4
F
8
feedstock. The trench depth is proportional to the number
of loops of etching/deposition steps which is with an etching rate of 0.5m/loop.
A photograph of these micro-channels is shown in Figure 3.2. This is a critical step
in realizing the full sensitivity of the mass spectrometer for measuring small reactions on
the order of the focused laser spot size (~1m
2
). Gold nanoparticles are deposited in the
bottom of the micro-channels by electron-beam evaporation of 5nm thick films of Au.
This is known to produce island-like growth of strongly plasmonic nanoparticles
separated by only a few nanometers[69]. Raman and near infrared spectra are collected
with the same objective lens used to irradiate the sample in a Renishaw inVia micro-
Raman spectrometer. The catalyst temperatures are determined from the downshifts
observed in the Fe
2
O
3
Raman modes. The calibration of these temperature-induced
downshifts were performed in a Linkam THMS temperature controlled stage. The
catalyst temperatures are also determined from the blackbody radiation intensity observed
at a wavelength of 980nm, which increases with the temperature to the fourth power
(I=T
4
)[29].
47
Figure 3.2 Silicon micro-channels with glass cover and metal eyelets.
3.3 Results and Discussion
3.3.1 Observation of in situ Temperature and Products of Plasmon Enhanced CO
Oxidation
When Au nanoparticles are irradiated with light at their plasmon resonance frequency
in a dilute Fe(CO)
5
/CO gas environment, crystalline Fe
2
O
3
(in the hematite phase) is
deposited. This process is called plasmon resonant chemical vapor deposition (PRCVD),
and is described in a previous publication[53]. Figure 3.3a shows the mass spectrometer
signals of CO
2
and O
2
plotted during a 1300 second laser exposure (48mW at 532nm).
Figure 3.3b shows the corresponding temperature data (as determined from the calibrated
48
Raman and infrared emission spectra) taken during the same 1300 second laser exposure.
During the first 300 seconds, the temperature remains constant at a value of 330
o
C, due to
the plasmonic heating of the nanoparticles. During this time, Fe
2
O
3
nanocrystals are
deposited. Throughout the first 300 seconds of laser exposure, the CO
2
and O
2
concentrations (Figure 3.3a) remain unchanged. After 300 seconds, the CO
2
concentration increases and the O
2
concentration decreases as they are produced and
consumed in the reaction 2CO + O
2
2CO
2
, respectively. The sudden increase in
temperature observed after 300 seconds is caused by the exothermic CO oxidation, which
becomes strongly catalyzed by the newly formed Fe
2
O
3
-Au nanoparticle composite. This
highly exothermic reaction (H=-532 kJ/mol) causes the temperature to increase rapidly,
exceeding 1000
o
C after 550 seconds. The drop in temperature observed after 770 seconds
is caused by blockage of the micro-channel, which restricts the flow of reactants and is,
therefore, not related to the reaction kinetics. If, at any time, the laser is turned off, the
reaction stops and the temperature drops to room temperature. As described in our
previous work, we observe the formation of carbon nanotubes at these elevated
temperatures (above 700
o
C). The precipitation of carbonaceous material occurs by the
reaction 2CO C + CO
2
. While this reaction has a lower negative heat of formation
49
(H=-173 kJ/mol) than the oxygenation reaction described above, it becomes
thermodynamically (entropically) favorable at these high temperatures.
Figure 3.3 (a) Quadropole mass spectrometer and (b) infrared temperature data taken
during laser irradiation in CO.
3.3.2 Comparison of Au NPs, Iron Oxide and Au Nanoparticle/Fe
2
O
3
Composite
The catalytic activity of this Au nanoparticle/Fe
2
O
3
composite is then evaluated in
purified CO, in which the Fe(CO)
5
has been removed, to prevent further deposition of
Fe
2
O
3
. Figure 3.4 shows the mass spectrometer signals of CO
2
plotted during a 5 second
50
laser exposure (48mW at a wavelength of 532nm). Here, we observe a rapid increase in
the CO
2
concentration when the laser is incident on the Fe
2
O
3
-Au nanoparticle composite.
However, when Au nanoparticles are irradiated in a region with no Fe
2
O
3
, no change in
the CO
2
signal is observed, as shown in Figure 3.4. Similarly, when we irradiate Fe
2
O
3
alone with no Au nanoparticles, we observe no production of CO
2
. We, therefore,
conclude that this reaction is not driven solely by the thermal (plasmonic) heating of the
gold nanoparticles, but relies intimately on the interaction of these two materials. Metal
nanoclusters on metal oxide supports are known to produce catalytic efficiencies that
exceed those of the two constituent materials by themselves[42, 43]. However, in these
prior studies, nanocluster sizes less than 5nm were required in order for significant
catalytic enhancement to be achieved[43]. The nanoparticle sizes used in this plasmonic
work are one order of magnitude larger than this. Furthermore, these previous studies
were carried out under uniform heating, rather than laser irradiation. As shown below,
these Au nanoparticle/Fe
2
O
3
composites show no detectible catalysis under uniform
heating. Therefore, the CO gas reaction kinetics we observe in this work originates from
a fundamentally different catalytic mechanism.
51
Figure 3.4 Quadropole mass spectrometer signals for CO
2
during 5sec laser-induced
reactions on various catalytic substrates
3.3.3 Thermally-driven Vs. Plasmonically-driven
In order to distinguish this plasmonically-driven catalysis from standard thermally-
driven catalysis, we uniformly heated the same gold nanoparticles in the same gas
environment to 350
o
C without laser irradiation. This is slightly higher than the
temperature achieved by plasmonic heating (see Figure 3.3). The CO
2
mass spectrometer
signal of this reaction is plotted in Figure 3.5. Once the CO gas flow is turned on, there is
an increase in the CO
2
signal, since the Au nanoparticles themselves are slightly catalytic.
The CO
2
production rate begins to decrease after 3 minutes, as the catalytic surface
52
becomes coked with amorphous carbon and Fe
2
O
3
. This is a well-known problem in
catalysis, limiting catalytic performance and lifetime. The same uniform heating reaction
was repeated at temperatures of 550 and 900
o
C, resulting in more severe coking of the
sample surface that could be clearly seen by SEM. These results are very different from
those observed under laser irradiation, where the CO
2
signal, and hence reaction rate,
increased monotonically for 15 minutes. Here, we observe a reaction rate that is
comparable to that achieved under laser irradiation (~pA); however, the reaction area is
2-3 orders of magnitude larger than that of the plasmonically-driven reactions in Figures
4.3 and 4.4, indicating that superior catalytic performance is obtained with plasmonic
excitation.
Figure 3.5 Mass spectrometer signals of CO
2
taken after uniformly heating the Au
nanoparticle sample to 350
o
C.
0 2 4 6 8 10 12 14 16
1.3
1.4
1.5
1.6
CO
2
QMS Signal (pA)
Time (min)
53
3.3.4 Quantification of the Plasmonic Enhancement
In order to quantify the plasmonic enhancement in the catalytic activity, we compared
the CO
2
production rate of these Au nanoparticle/Fe
2
O
3
composites under laser
irradiation with that under uniform heating. Figure 3.6a shows the mass spectrometer
signal for CO
2
plotted versus time at a base substrate temperature of 300
o
C. The laser is
pulsed on and off 10 times with 30-second pulse durations. With the laser on, the CO
2
signal rises significantly above the baseline. This was repeated at several different base
substrate temperatures. Figure 3.6b shows the average CO
2
quadropole mass
spectrometer (QMS) signal under laser irradiation compared with that under uniform
heating over a wide range of base substrate temperatures. The data shows that the laser-
induced catalysis is at least 2-3 orders of magnitude higher than that of uniform heating.
We see no detectible change in the baseline CO
2
signal, even after 25 depositions of
Fe
2
O
3
on the Au nanoparticles. This 2-3 order-of-magnitude enhancement is a lower
limit for the plasmonic enhancement, since the uniform heating data is limited by the
noise in our system.
As a final control experiment, we performed the same experiment with 785nm light,
which is below the plasmon resonance energy of the Au nanoparticles. Here, we observed
54
no decomposition of Fe(CO)
5
or CO and no deposition of material on the substrate, even
at laser powers 10 times higher than those used with 532nm. Therefore, in order for
sufficient electric fields to be achieved and significant heating to occur, the laser energy
must match the plasmon resonance frequency of the nanoparticles. Also, regions on the
substrate without nanoparticles were irradiated, and no decomposition of CO or
deposition of Fe
2
O
3
was observed.
Figure 3.6 (a) QMS signal for CO
2
during a series of 10 30-second laser irradiations with
a base substrate temperature of 300
o
C. (b) Comparison of the average CO
2
QMS signal
for laser irradiation and uniform heating.
3.4 Conclusion
In conclusion, we demonstrate plasmonically driven catalysis of CO on Au
nanoparticle/Fe
2
O
3
composites. The catalytic performance of these composites greatly
0 100 200 300 400 500 600 700
0.15
0.20
0.25
0.30
0.35
0.40
CO
2
QMS Signal (pA)
Time (sec)
0 100 200 300 400 500
10
-13
10
-12
10
-11
10
-10
Laser Irradiation
Uniform Heating
CO
2
QMS Signal (A)
Temperature (
o
C)
(b) (a)
55
exceeds that of the Au nanoparticles and Fe
2
O
3
alone, when irradiated with visible light.
Control experiments indicate that the plasmonically driven catalytic reaction rate is
several orders of magnitude higher than that obtained under uniform heating, without
irradiation. We conclude that this enhanced catalytic reaction is not driven solely by the
thermal (plasmonic) heating of the gold nanoparticles, but relies on the interaction
between the plasmonic nanoparticles and the metal oxide catalyst. The improved catalytic
processes described in this work will likely be extended to other applications such as
solar energy conversion and storage.
56
Chapter 4: Plasmonic Enhancement of Dye
Sensitized Solar Cell
4.1 Introduction
Dye-sensitized solar cells (DSSCs) have drawn much attention in the worldwide
energy research due to its promising power conversion efficiency and its low cost
fabrication process. However, after the highest efficiency (11%) reported in middle of
1990’s[86, 90], the increase of the cell performance is not very slow and not pronounced
comparing its theoretical maximum value [34, 87]. The state of art of solar cell generally
is consisted of four procedures which are light absorption, exciton separation, carrier
transportation and collection. Many efforts have focused/invented on each of these four
steps to improve the cell overall performance. Examples of this include higher efficiency
of adsorbed molecular[40], broader spectrum of light absorption [114], generation of
multiple exitons by a single photon [27].
Besides modification of the absorbing dye molecular, another approach is to
incorporate the plasmonic metal nanoparticles to improve the light harvesting by
exploiting their unique optical property, surface plasmon resonance. As discussed in
57
previous chapters, surface plasmon resonance is produced by the collective oscillation of
electrons confined to the surface of metals, when the incident light matches the plasmon
frequency of irradiated metals.[84] These collective oscillations of charges bring new
optical properties to these plasmonic nanoparticles such as surface plasmon resonance
(SPR) which resulted in the strongly enhanced electric field around surface of NPs. This
enhanced electric field from the surface plasmon resonance can be several thousand times
larger than that of incident light and able to assist the absorption of light and separation of
charges[116].
Plasmonic enhancement has been demonstrated in inorganic/organic thin film solar
cells[6, 71, 95]. Some prior works have been achieved by utilizing an electron beam
lithography to pattern plasmonic nanostructures on the indium tin oxide (ITO)[68] or
blending nanoparticles within organics[26]. However, even though electron beam
lithography provides high controllability of nanostructures’ shape, size and separation,
this scheme is still thought not practical in large scale fabrication. For the blending
method, the solubility and non-uniform distribution present potential issues in
manufacturability. Therefore, people are still looking for a simpler and more effective
fashion to incorporate plasmonic nanostructures into solar cells.
58
Here, we present a simple and low cost method to integrate plasmonic nanostructures
into dye sensitized solar cells by using electron beam deposition of a 5nm film of gold.
Since 5nm is not enough to form a continuous film, this scheme is known to produce
island-like growth of strongly plasmonic nanoparticles with few-nanometer separations.
One drawback of this plasmonic enhancement mechanism is the short lifetime of
dipoles/excitons caused by the damping effect [56, 64]. Therefore, a fast charge
separation system has become a requirement to achieve pronounced plasmonic
enhancement. Nevertheless, this requirement makes a dye-sensitized solar cell a good
candidate due to its ultra fast charge transfer process[98].
In this study, plasmonic enhancement is observed by the increase in photocurrent and
power conversion efficiency of DSSCs. In addition to I-V characterization, the
ultraviolet-visible (UV-Vis) and spectral response of photocurrent are measured to
determine the enhanced region of light absorption and photocurrent generation and
provide a fundamental understanding of the energy transfer from the plasmonic
nanoparticles to the absorbing dye molecules. Finite-difference-time-domain (FDTD)
simulations are carried out to further understand the basic mechanism of plasmon
enhancement.
59
4.2 Experimental Procedure
4.2.1 Preparation of Anodic TiO
2
and Cell Assembling
Anodic TiO
2
(ATO) is prepared by electrochemically oxidizing titanium foils in an
ethylene glycol electrolyte containing 0.25wt% NH
4
F and 2wt% H
2
O with an anodization
potential of 30 V applied for one hour[35]. Crystalline TiO
2
is achieved by performing
an annealing treatment at 450
o
C for 5 hours after the anodization. Ruthenium-based dye
(N719) purchased from Sigma-Aldrich, Inc is used as a sensitizer in this study. The dye
loading process is completed by soaking the ATO anode in the N719 dye solution for 24
hours and then rinsing with IPA to remove the unattached dye on the surface of ATO. A
plasmonic layer of 5nm gold is deposited on the annealed TiO
2
before and after the dye
loading process by electron-beam evaporation at a rate of 1Å/sec. The counter electrode
consists of indium tin oxide (ITO) with an additional layer of 1nm Pt on top of the
surface, which accelerates the charge transfer process in the reduction of triodide (I
3
-
)[32].
Figure 4.1 shows the basic principle of operation and energy level diagram of our
plasmon-enhanced dye sensitized solar cell. When light is absorbed by the dye, it excites
electrons from the HOMO to LUMO of the dye. The electrons will then be transferred to
60
the titania conduction band, which are then conducted around the external circuit. The
holes in the HOMO then accept electrons from the ions in solution. The electrons in the
external circuit come back into the solution at the ITO counter electrode and recover the
ions in solution, completing the electrochemical circuit. The details of the plasmon-
enhancement mechanism will be discussed in the later section.
Figure 4.1 Principle of operation and energy level diagram of the plasmon-enhanced dye
sensitized solar cell.
The dye-loaded TiO
2
electrode and Pt-coated ITO counter electrode were assembled
into a sandwich type cell shown in Figure 4.2a which is sealed by a hot melting seal foil
61
of 100μm thickness purchased from Solarnix, Inc.. The area of open window in the
middle of the seal foil is 0.25cm
2
. Two drilled holes with area less than 0.02 cm
2
on the
top of counter electrode are open for the injection of I
-
/I
3
-
electrolyte. Figure 4.2a shows
an image of the plasmon-enhanced dye sensitized solar cell after the assembling process.
Raman spectroscopy is employed to characterize the quality of the anodic titanium
dioxide film after annealing treatment. The Raman spectrum shown in Figure 4.2b was
taken with a 100X objective lens focusing the laser beam to a diffraction limited spot size.
All of the peaks observed, 150, 396, 517 and 633cm
-1
, correspond to the Raman active
vibrational modes of anatase crystalline phase TiO
2
[115].
62
0 300 600 900 1200 1500
0
500
1000
1500
2000
2500
3000
3500
4000
Anatase TiO
2
Raman Intensity (counts.)
Raman Shift (cm
-1
)
Figure 4.2 (a) Photograph of the plasmon-enhanced cell before electric characterization.
(b) Raman spectrum of anodic TiO
2
in anatase phase after annealing at 450
o
C.
4.3 Result and Discussion
4.3.1 Electrical Characterization
The J-V characterization of the dye sensitized cells with and without incorporation of
plasmon-enhanced Au layer is performed under AM 1.5 solar simulator equipped with a
(a)
(b)
63
450 W xenon lamp with a power output of 100 mW/cm
2
calibrated by a reference Si
photodiode. Figure 4.3a exhibits a comparison of the photovoltaic measurement
with/without island-like plasmonic nanostructures. The plasmon-enhanced cell shows
around a 45% enhancement in the photocurrent (1.81mA/cm
2
) relative to that of the
control cell (1.25 mA/cm
2
) with a similar value of open circuit voltage (V
oc
) and fill
factor (FF), as shown in Table 1. Photocurrent spectra are also measured in this study.
Figure 4.3b shows the normalized incident photon-to-current conversion efficiency (IPCE)
with and without 5nm gold plotted as a function of wavelength. The plasmon-enhanced
cell exhibits an enhancement in photocurrent over the entire dye absorption range from
450nm to 700nm as well as the NIR tail above 700nm, which is due to extra charges from
TiO
2
and has been verified by absorption spectrum of TiO
2
.
64
-0.8 -0.4 0.0 0.4 0.8
-3
-2
-1
0
1
2
3
Current Density (mA/cm
2
)
Voltage(V)
with Au
without Au
400 500 600 700 800
0.0
0.2
0.4
0.6
0.8
1.0
Normalized IPCE
Wavelength (nm)
with gold
without gold
Figure 4.3 (a) J–V characteristics of dye sensitized solar cells. (b) Normalized incident
photon to current conversion efficiency spectrum with and without island-like gold
nanostructures.
(a)
(b)
65
We have also characterized the performance of cells under monochromatic laser
irradiation (532nm and633nm), as shown in Figure 4.4. A summary of the data taken
under the monochromatic laser irradiation is also listed in the Table 1. Briefly, the
plasmonic enhancements of the cell performance under 532nm and 633nm light are 9.5%
and 13.4%, respectively. In the NIR wavelength range, enhancement factors of 7.5 are
observed, however, since this is far from the absorption bands of the dye molecule, the
overall photoconversion efficiencies are quite low in this wavelength range.
Figure 4.4 I-V characteristics of dye sensitized solar cells with and without island-like
gold nanoparticles under (a) 532nm (b) 633nm monochromatic laser irradiation.
(a) (b)
-0.4 0.0 0.4 0.8
-1.0x10
1
-5.0x10
0
0.0
5.0x10
0
1.0x10
1
Current (A)
Voltage(V)
Without Au
With Au
=633
-0.8 -0.4 0.0 0.4 0.8
-6.0x10
1
-4.0x10
1
-2.0x10
1
0.0
2.0x10
1
Current (A)
Voltage(V)
Without Au
With Au
nm
66
Table 1. Summary characteristics of dye sensitized solar cell with and without plasmonic
gold nanoparticles.
Light
source
Jsc(mA/cm2) Voc(V) FF Eff1(%) Eff2(%)
Enhancement
(%)
white
light
w/o Au 1.277 0.67 0.61 0.52 0.6
45
w/Au 1.806 0.66 0.617 0.74 0.9
Isc(mA) Voc(V) FF Eff1(%) Eff2(%)
532nm w/o Au 0.041 0.73 0.593 1.7 2.1
9.52
w/Au 0.05 0.6 0.632 1.9 2.3
Isc(mA) Voc(V) FF Eff1(%) Eff2(%)
633nm w/o Au 0.004 0.54 0.434 0.347 0.4164
13.35
w/Au 0.006 0.44 0.482 0.393 0.472
In order to understand the preferential enhancement in the infrared wavelength region,
we performed scanning electron microscopy (SEM) on these island-like gold films to
investigate the shapes, gaps, and sizes of the nanoparticles. Figure 4.5a shows a 5 nm
nominal thickness gold film deposited on anodic TiO
2
, which exhibits formation of
island-like nanostructures with few-nanometer separations. The same 5nm gold film is
also deposited on glass slides for UV-Vis absorption measurements. In Figure 4.5b, the
absorption of this island-like gold film shows an asymmetric V-shape curve with a valley
located around 520nm and a higher absorption tail in the longer wavelength region. This
is consistent with our previous results under the monochromatic laser irradiation, in
67
which a higher improvement of cell efficiency was observed at longer wavelengths. In
order to quantify plasmon enhancement factors as a function of wavelength, we
normalized the plasmon-enhanced IPCE spectrum with that of dye only (control cell),
and the result shown in the Figure 4.5c. Interestingly, this plot presents a wavelength
independence-like enhancement from 450nm to 700nm, which is different from the UV-
Vis spectrum of 5nm gold. However, this constant enhanced region coincides with the
absorption range of dye (N719). Notably, beyond the 700nm the enhancement factor
begins to increase rapidly, indicating a higher enhancement in the photocurrent at longer
wavelengths.
68
Figure 4.5 (a) Scanning electron microscope (SEM) image of 5nm gold film on TiO
2
. (b)
UV-Vis absorption spectrum of 5nm gold film alone. (c) Short-circuit photocurrent
enhancement factor of 5nm gold film on the dye loaded TiO
2
.
(a) (b)
(c)
450 500 550 600 650 700 750 800
Absorption
Wavelength (nm)
5nm Gold film
450 500 550 600 650 700 750 800
0
1
2
3
4
5
6
7
8
9
Enhacement factor
Wavelength (nm)
enhanced ratio with/ without gold
69
4.3.2 FDTD Simulation of Plasmon Induced Electric Feilds
There are several potential mechanisms causing the plasmon enhancement. Besides
the light scattering to increase the effective optical thickness, we believe that the strong
electric fields, generated from localized surface plasmon resonance (LSPR) and
propagating surface plasmon resonance (SPR), play an important role associated with the
generation and dissociation of excitons. In order to explore further insight of the
plasmon-induced electric field, we carry out the finite-difference-time-domain (FDTD) to
simulate the plasmon-induced filed distribution from this island-like gold film.
The electromagnetic response of this gold island film is calculated using the finite-
difference time domain (FDTD) method as shown in Figures 4(b-d). Here, we use the
actual SEM image shown in Figure 4a to define the spatial extent of the gold islands in
this simulation. Local “hot spots” can be seen between nearly touching Au nanoparticles.
Figure 4.6d shows a cross-sectional plot of the electric field distribution of one of these
hot spot regions in the z-dimension. In this hot spot region, the electric field intensity at
the TiO
2
surface reaches 1000 times that of the incident electric field intensity. This
means that the photon absorption (and hence electron-hole pair generation) rate is 1000
times higher than that of the incident electromagnetic radiation.
70
Figure 4.6 (a) SEM image of a 5 nm thick Au island film. (b-d) simulated electric field
intensity observed from the top surface and cross section of the interface of Au–TiO
2
calculated using FDTD.
We can calculate the enhancement factor based on the results of the FDTD simulation.
Since the photon absorption rate is proportional to the electric field squared (|E|
2
), we
integrate |E|
2
over the whole film, and divide by the integral of the incident
electromagnetic field squared (|E
o
|
2
), as follows:
() (4.1)
71
In the z-dimension, we only integrate from the TiO
2
surface (z = 0) to the one exciton
diffusion length below the surface (z = -10nm). The value for the EF when integrating
over the whole simulation area (400 nm × 300 nm) is 12X. It should be noted, however,
that this value is for a random distribution of gold islands that are not optimized
geometrically. If, instead, we integrate only over the area of one hot spot, as shown in
Figure 4.6c, this would yield an EF of 190X.
According to this simulation result, we believe the enhancement of solar energy
conversion is based to the plasmon-induced electric field as shown in the simulation
result, which assists the processes of exciton generation and dissociation. Additionally,
the plasmon-induced “hot excitons”, high energy excitions, can be understand from the
mechanism of the plasmon-exciton coupling[89, 116], which facilitates excitons to
escape from Coulombic binding.
4.3.3 Photoluminescence Spectrum
In the last part of this plasmonic enhancement study, we performed photo-excitation
measurements on the samples made of solgel TiO
2
, rather than ATO, loaded with
phosphorescence dye, platinum tetra (1, 3-di-tert-butylphenyl) tetrabenzoporphyrin (Pt
tbu (TPBP)), with and without 5nm gold films. Here, an additional annealing step was
72
performed after the electron-beam deposition, causing changes of morphology from
island-like to a spherical-like. This change in shape of the Au nanoparticles shifts the
plasmonic frequencies, as shown in the UV-Vis absorption spectra of Figure 4.7. The
purpose of shifting the plasmonic frequency of the Au film is to create a scenario of non-
overlap with the absorption spectrum of the phosphorescence dye. The non-overlap of
absorption with the resonance of the Au NPs is facilitated to investigate Forester energy
transfer between the plasmons and excitons in the phosphorescent dye. Figure 4.8 shows
the emission of phosphorescence from the dye at 760nm as a function of the incident of
scanning wavelengths. There is no pronounced peak observed at the excited wavelength
of 532nm, which is the corresponding plasmon resonant spot of annealed 5nm gold film.
The absence of a peak at 532nm in the excitation spectrum indicates that the plasmon
energy in the nanostructures is not transferred to phosphorescence dye through the
Forester resonance energy transfer mechanism. These results indicate that the
improvement of cell performance originates from another path, which we assume is the
plasmon-induced electric field.
73
300 400 500 600 700 800 900 1000
Absorption
Wavelength (nm)
Pt organics
annealed 5nm gold
organics and annealed 5nm gold
Figure 4.7 Absorption spectra of phosphorescence dye (black), 5nm annealed gold (red),
and dye with annealed 5nm gold underneath(blue).
300 400 500 600 700
0
1x10
4
2x10
4
3x10
4
4x10
4
5x10
4
6x10
4
7x10
4
Emission(counts)
Wavelength (nm)
Dye on solgel TiO
2
300 400 500 600 700
0.0
2.0x10
4
4.0x10
4
6.0x10
4
8.0x10
4
1.0x10
5
1.2x10
5
1.4x10
5
1.6x10
5
1.8x10
5
Wavelength (nm)
Emission (a.u.)
Dye_gold_solgel TiO
2
Figure 4.8 Emission spectra of (a) TPBP dye only (b) TPBP dye and annealed 5nm gold
on the solgel TiO
2
.
(a)
(b)
74
4.4 Conclusion
In conclusion, we have successfully demonstrated plasmonic enhancement of dye
sensitized solar cells with up to 45 % increase in the cell power conversion efficiency by
incorporating a layer of island-like plasmonic gold nanoparticles. UV-Vis absorption
spectrum of the island-like plasmonic film shows good agreement with the IPCE
spectrum, indicating a higher enhancement at longer wavelengths. Finite-difference-
time-domain (FDTD) simulations are also carried out to investigate the distribution of the
plasmon-induced electric field in the 5nm island-like gold film. Furthermore, we believe
the enhancement observed in the dye sensitized cell can be attributed to the plasmon-
induced electric fields, which increase the exciton generation rate and dissociations. The
photoluminescence spectra rule out the mechanism of plasmon energy transfer through
Forester resonance energy transfer.
75
Chapter 5: Plasmonic Enhancement of
Methane Production
5.1 Introduction
In addition to the solar-to-electrical energy conversion in dye sensitized solar cells
(DSSCs), we extend the use of plasmonic enhancement to solar-to-chemical conversion
application in this chapter. Here, we explore the photocatalytic reduction of carbon
dioxide (CO
2
) with water to form methane and other hydrocarbons. This has several
advantages over direct solar-to-electric conversion. As an intermittent energy source
(night and day), there is no way to store this enormous amount of electricity each day to
be used during the night. It is well known that many households with solar panels do not
have an adequate method of storing their unused power. Batteries give the option of
storing limited amounts of energy, but suffer from finite lifetimes and contain
significantly toxic chemicals that are difficult to dispose of. Also, there are huge losses
associated with transporting electricity over large distances. Therefore, a method of
storing the sun’s energy in chemical bonds and then releasing it without harmful
byproducts is indeed the “Holy Grail” in solar energy conversion.
76
Photocatalytic conversion of carbon dioxide into hydrocarbons is of particular interest
for its potential to convert an abundant greenhouse gas to useful hydrocarbon fuels. This
was demonstrated by Honda and co-workers in 1979 through the reduction of carbon
dioxide into organic compounds, formic acid, formaldehyde, methyl alcohol, and
methane under a Xe lamp irradiation.[55] In addition, Halmann reported formic acid
production from aqueous CO
2
at the p-type GaP electrode of in an electrochemical
photocell[37]. The same result was observed with oxide semiconductors in a
photochemical solar cell.[38] Fujiwara et al. utilized ZnS nanocrystallites to reduce CO
2
to formic acid. Furthermore, Thampi et al. reported selective conversion of CO
2
to
methane at room temperature and atmospheric pressure, using highly dispersed Ru/RuO
x
loaded onto TiO
2
as a photocatalyst.[107]
TiO
2
is one of the most promising photocatalysts for carbon dioxide reduction due to
its long-term photostability. However, it does not absorb light in the visible region of the
electromagnetic spectrum. Because of TiO
2
’s short wavelength cutoff, only a small
fraction of solar photons (~4%) can be used to drive this photocatalyst, as shown in the
Figure 5.1a. Several attempts have been made to overcome this main disadvantage for the
photocatalytic conversion of carbon dioxide into hydrocarbons under visible
illumination[2]. For example, Varghese et al. doped nitrogen into TiO
2
and extend its
77
absorption edge from 400nm to around 520nm.[110] Wang et al. added CdSe (Cadmium
Selenide) quantum dots to the TiO
2
as an absorber to enhance visible light
absorption.[111] Wu et al. used a TiO
2
coated optical fiber as the reaction cell to improve
TiO
2
optical depth. They also deposited silver and copper on the TiO
2
surface, which
assists gas molecule attachment on the surface and as electron trap. However, from their
absorption spectra they claimed no changes in the TiO
2
absorption.[112]
Here, we report a new mechanism of incorporating plasmon resonant nanoparticles on
the top of TiO
2
catalysts by electron beam deposition to enhance the visible light
absorption, as shown in The Figure 5.1b.
Figure 5.1 (a) Absorption spectrum of TiO
2
and solar spectrum. (b) Absorption spectrum
of metal nanoparticles.
(b) (a)
78
5.2 Experimental Procedure
5.2.1 Preparation of Plasmonic Substrate
Anodic TiO
2
(ATO) is prepared by electrochemically oxidizing titanium foils in an
ethylene glycol electrolyte containing 0.25wt% NH
4
F and 2wt% H
2
O with an anodization
potential of 30 V applied of one hour[35]. Crystalline TiO
2
is achieved by annealing at
450
o
C for 5 hours after the anodization. A 5nm plasmonic gold film is then deposited by
electron beam deposition, which is known to form island like gold nanoparticles. Figure
5.2a shows the absorption spectra of TiO
2
with and without presence of plasmonic gold
nanoparticles.
The experiment setup for hydrocarbon generation is shown in Figure 5.2b, the
photocatalytic reduction of CO
2
and H
2
O are carried out in a sealed 51.6 ml stainless steel
reactor with a quartz window. The reactor is first purged with CO
2
saturated water vapor
for 1 h before closing the system. The reactor is then illuminated with UV light (254 nm
20 mW/cm
2
) and visible light (532 nm 350 mW/cm
2
, Spectra Physics solid state laser
green laser) for 15 h at 75
o
C. The reaction products are analyzed using a Varian gas
chromatograph (GC) equipped with TCD (with a detection limit of 10 nmol/mol for CO
2
)
and FID (with a detection limit of 9 nmol/mol for hydrocarbons) detectors.
79
Figure 5.2 (a) UV-Vis absorption spectrum of TiO
2
with and without gold nanoparticles.
(b) Experimental setup showing the gas phase photochemical cell.
5.3 Results and Discussion
5.3.1 Visible Light Illumination
Figure 5.3a shows the GC data observed after 15 h visible (532 nm laser) illumination.
Here, methane is the only product detected by the GC for two samples, bare TiO
2
and Au
nanoparticle deposited on TiO
2
. These reactions can be understood by comparing the
conduction and valence band energies of TiO
2
with the reduction potentials of CO
2
for
the three potential reduction products CH
4
, HCHO, and CH
3
OH, as shown in Figure 5.3b.
Since the conduction band of TiO
2
lies above the reduction potential of CO
2
/CH
4
, it is
energetically favorable for electrons from the conduction band of TiO
2
to transfer to CO
2
(downhill transfer) to initiate the reduction of CO
2
with H
2
O producing CH
4
. Methane is
(a) (b)
(Reaction cell)
(532nm laser beam)
80
the only favorable product since the reduction potentials of CO
2
/HCHO and CO
2
/CH
3
OH
lie above the conduction band of TiO
2
. Because the photon energy of 532nm wavelength
light (2.41eV) is smaller than the band gap energy of TiO
2,
the electrons involved in the
reduction of CO
2
under the 532nm laser irradiation are excited from defect states in the
TiO
2
(e.g. surface defect states/doping states). Integration of the peak areas in Figure 5.3a
gives a 17% enhancement for the sample with gold nanoparticles with respect to the bare
TiO
2
sample. Again, this enhancement is attributed to the strong plasmon induced electric
fields generated from gold nanoparticles to improve light absorption in the TiO
2
, when
the incident photon energy matches the plasmon resonance of gold.
Figure 5.3 (a) Photocatalytic product yields (after 15 h of visible irradiation) from two
different catalytic surfaces. (b) Energy band alignment of anatase TiO
2
, Au, and the
relevant redox potentials of CO
2
and H
2
O under visible illumination.
(a)
(b)
CH
4
2.4 2.5 2.6 2.7 2.8 2.9 3.0
Counts (a.u.)
Time(mins)
Without gold
With gold
=532nm
81
5.3.2 Ultraviolet Light Illumination
In order to understand the mechanism of enhanced hydrocarbon production by the
photoreduction of CO
2
and H
2
O, we also characterized this reaction under 254nm
wavelength UV irradiation. Figure 5.4a shows the product yields after 15 h UV (254 nm
mercury lamp) illumination. For the bare TiO
2
catalyst, methane is the only product
detected by GC. However, for the Au nanoparticle/TiO
2
-catalyzed reactions, we observe
additional reaction byproducts, including ethane, formaldehyde, and methanol. Again, we
are able to understand these results by comparing the conduction and valence band
energies of TiO
2
and Au with the reduction potentials of CO
2
for the reduction products
CH
4
, HCHO, and CH
3
OH, as shown in Figure 5.4b. While the conduction band of TiO
2
lies above the reduction potential of CO
2
/CH
4
, it is below the reduction potentials of
CO
2
/HCHO and CO
2
/CH
3
OH. As a result, methane is the only product for the bare TiO
2
-
catalytzed reaction. The energy of the 254 nm wavelength light (4.88 eV) exceeds the
minimum energy required for interband transitions in Au (2.5 eV) and therefore is able to
excite electrons from its d band to the conduction band, which lies above the conduction
band of TiO
2
and the reduction potentials of CO
2
/CH
4
, CO
2
/CH
3
OH, and CO
2
/HCHO, as
shown in Figure 5.4b. These highly energetic electrons are then able to drive all of these
products of the reduction of CO
2
and H
2
O. Thus, methane, ethane, formaldehyde, and
82
methanol are observed in the Au nanoparticle/TiO
2
. Again, with integration the methane
peaks area from Figure 5.4a, we obtain 5 fold enhancements with respect to the bare TiO
2
sample due to the additional high energy electrons from d band transition of gold.
Figure 5.4 (a) Photocatalytic product yields (after 15 h of 254 nm UV irradiation) from
two different catalytic surfaces. (b) Energy band alignment of anatase TiO
2
, Au, and the
relevant redox potentials of CO
2
and H
2
O under 254 nm UV illumination.
In addition to samples with plasmonic gold deposited on TiO
2
supports, we also
performed photocatalytic reduction CO
2
using bare gold nanoparticles supported on an
inactive glass substrate under 532nm laser and 254nm UV irradiation. Not surprisingly,
under 532nm irradiation, this bare gold sample shows a negligible yield; however, with
254nm irradiation due to the high energy electron from the d band transition we obtained
all of the hydrocarbons.
2.0 2.4 2.8 3.2 3.6 4.0
Counts (a.u.)
Time (sec)
ATO_gold_254nm
ATO_254nm
=532nm
(a)
(b)
HCHO
CH 4
CH 3CH 3
CH 3OH
83
The last control experiment we performed in this study was to irradiate these three
samples, bare TiO
2
; Au nanoparticles on TiO
2
and bare gold nanoparticles on the glass
under 365 nm UV illumination. Under 365 nm UV illumination, no hydrocarbons were
formed for the bare 5 nm Au thin film deposited on glass and only methane was formed
for both the TiO
2
and Au nanoparticle/TiO
2
-catalyzed reactions. The reason no products
are formed on the bare Au nanoparticle surface is that the energy of the 365 nm light (3.4
eV) is not high enough to excite electrons beyond the redox potentials of the CO
2
/CH
4
,
CO
2
/HCHO, CO
2
/CH
3
OH, as shown in Figure 5.5.
Figure 5.5 Energy band alignment of anatase TiO
2
, Au, and the relevant redox potentials
of CO
2
and H
2
O under 365 nm UV illumination.
5.4 Conclusion
In summary, we observe a 17% plasmonic enhancement in the photocatalytic
reduction of CO
2
with H
2
O under visible illumination with the addition of plasmonic Au
nanoparticles deposited on top of the TiO
2
surface. This enhancement is attributed to the
84
strong electric fields created by the surface plasmon resonance of the Au nanoparticles,
which excites electron-hole pairs locally at the interface surface of gold and TiO
2
and
drives these reactions at an accelerated rate.
85
Chapter 6: Rapid Prototyping of Three-
Dimensional Microstructures from Multi-
Walled Carbon Nanotubes
6.1 Introduction
Carbon nanotubes have attracted a lot of attention over the past 15 years due to their
exceptional properties which far exceed those of most known bulk materials. These
properties include high mechanical strength, high surface area, and high thermal and
electrical conductivity[7, 28, 58]. Many potential applications have been proposed that
exploit these exceptional properties[83, 97, 106]. The ability to pattern carbon nanotube
microstructures is an important step in realizing these applications. Conventional
lithography techniques are limited to patterning 2D microstructures and require a
sequence of fabrication steps that introduce chemical residues that are incompatible with
specialized biological applications[41]. It is therefore important to investigate alternatives
to the controlled fabrication of nanotube microstructures to enable a broader set of
applications. In this chapter, we describe a non-chemical, local patterning method that
86
leaves the patterned nanotubes unperturbed as grown. It is therefore suitable for
chemically sensitive applications.
Several techniques for patterning 3D carbon nanotube structures have been explored
previously[16, 60, 73]. These approaches are based on bottom up growth of MWNTs
from a patterned catalyst, which is limited to 2D-like geometries. Complex 3D
microstructures have been fabricated in systems other than nanotubes using an optical 2-
photon photopolymerization process[24, 63, 79]. While this technique provides high
spatial resolution for creating 3D structures, it is limited to polymer-based resin materials
that are electrically insulating, which severely limits their potential applications. Also,
these materials do not possess the desirable surface properties of carbon nanotubes,
which can be easily modified with the vast repertoire of carbon based chemistry[36, 104].
6.2 Experimental Procedure
6.2.1 Carbon Nanotubes Forest Sample Fabrication
Carbon nanotubes forest samples were received from our collaborator Michael J
Bronikowski in the Jet Propulsion Laboratory. Carbon nanotube forests are grown by
chemical vapor deposition (CVD) by passing ethylene (C
2
H
4
) over a pre-deposited iron
catalyst on Si wafers. The iron catalyst is prepared by evaporating a 2.5nm film of Fe on
87
silicon substrates with 400nm of thermal oxide. The nanotube growth takes place in a
heated tube furnace at 650
o
C[12, 13].
6.2.2 Opitcal Measurement - Raman Spectroscopy
A 532nm 5W Spectra Physics solid state laser is collimated and focused through a
Leica DMLM microscope, and used to irradiate these samples. The samples are
manipulated spatially on a PRIOR ProScan II high precision microscope stage. Raman
spectra are taken with a Renishaw inVia Raman Spectrometer from the scattered light
collected by the same objective lens. With this optical setup, we determine the minimum
threshold laser power for burning carbon nanotubes in air by observing changes in the
intensity of nanotube Raman spectra before and after laser exposure. The Raman
intensity gives a measure of the density of carbon nanotubes within the focal volume. We
use a 50X long working distance objective lens with NA=0.5 and spot size=1.25μm.
Nanotubes were exposed at laser powers between 50μW and 9000μW for 1 second. The
results were not sensitive to exposure time, indicating that the burnout occurs on a much
shorter timescale. To determine the relative change in nanotube density, Raman spectra
are taken at low sub-threshold laser powers (50μW) with an accumulation time of 120
seconds before and after high power laser exposure.
88
6.3 Results and Discussion
6.3.1 Creating 3D Structures in Carbon Nanotubes Forest
Here, we create three-dimensional microstructures using a focused laser beam to
selectively burn local regions of a dense forest of multi-walled carbon nanotubes. Raman
spectroscopy is used to systematically quantify this process in a controlled fashion to
determine the laser power threshold for burning carbon nanotubes and also the depth of
burnout at different laser powers. Figure 1 shows a scanning electron microscope (SEM)
image of a 3D staircase structure patterned using this technique. The clear depth change
can be seen in the side view of this image. The dimensions of the steps are 5μm high,
30μm wide and 7μm deep. A square volume was burned adjacent to this staircase
microstructure to allow for easier viewing. This technique can be used to precisely
control carbon nanotube forests to create well-defined channel geometries for gas and
liquid transport through carbon nanotube membranes[1, 47]. Curved surfaces can also be
easily created using this technique. Figure 6.2 shows cylindrical structures fabricated by
burning concentric rings in the MWNT forest. These images exemplify the high aspect
ratios that can be achieved with this technique. Again, a square box was removed
adjacent to this microstructure for better viewing. These `deep trenches may be suitable
for superhydrophobic microfluidic channels with complex geometries[59].
We have also patterned square arrays using this non-contact method, as shown in
Figure 6.3 These arrays may be suitable for field emission applications[19]. The large
undercut seen in Figure 6.3(b) shows the effect of using high laser powers to pattern these
microstructures. The high laser power causes the units of the matrix to twist and rotate
due to their mechanical instability and the high temperatures reached during burnout.
89
From the work of Cataldo, et. al, the burnout process in carbon nanotubes is expected to
occur at 800
o
C in air[15]. In fabricating these fragile structures, the sequence of write
steps and laser power must be taken into consideration to avoid collapse and distortion of
delicate microstructures with high aspect ratios, such as those shown in Figure 6.3(b).
Figure 6.1 3D staircase structure fabricated in the MWNT surface.
Figure 6.2 (a) Concentric cylindrical structures patterned using the laser burnout method.
(b) Close-up image showing a slight undercut profile.
90
Figure 6.3 Square arrays patterned with laser powers of (a) 730W and (b) 1900W.
6.3.2 Burnout Depth vs. Laser Power Exposure
In this 3D patterning technique, it is important to determine the depth of burnout, and
hence resolution in the z-direction. In the x-y plane, the resolution is limited by the spot
size of the objective lens. It is difficult to measure the depth of burnout because of the
soft surface of the MWNT forest, which prohibits the use of a profilometer or atomic
force microscope. Raman spectroscopy provides a non-contact method for measuring the
surface height and is therefore suitable for working with this delicate system. In order to
measure the burnout depth using Raman spectroscopy, we measure spectra at various
heights with respect to the sample surface. Figure 6.4 shows the Raman intensity vs.
depth profile taken before and after laser exposure at 1000μW and exhibits a Gaussian
intensity-height profile. The relative shift of these Gaussian peaks corresponds to the
depth of burnout. The resulting depth-laser power data is plotted in Figure 6.4(b) and
exhibits a linear relation between the laser power and depth of burnout. In the limit of
low laser power, the minimum burnout depth is found to be 5μm. This is limited mainly
by the relatively low numerical aperture of the objective lens (NA=0.5). Using a higher
numerical aperture lens would provide more tight confinement of the laser light and
allow more precise patterning of the MWNT surface. The drop in intensity from 2200 to
91
1450 photon counts after laser exposure reflects the change in nanotube density of the
laser treated surface. This change in nanotube density has been seen in SEM images in
the later section, and may provide a way of varying the wetability of the hydrophobic
surface of MWNT forests[48, 59, 70].
Figure 6.4 (a) Raman intensity vs. height profile before and after 1000μW laser exposure.
(b) Relation between burnout depth and laser power exposure.
6.3.3 Surface after Burnout Treatment and Burnout Threshold Power
We observe several interesting phenomenon on the surface of the MWNT forests
after laser treatment. SEM images reveal white spots ranging from 100nm to 200nm on
top of the burned MWNT surface Figure 6.5. At higher magnification, these white spots
can be resolved as nanotube bundles that aggregate during the exothermic burnout
process. This aggregation demonstrates the dynamic nature of the burnout process of
these MWNTs.
We determine the threshold for laser burnout to occur at 300μW, which corresponds
to a power density of 244μW/μm
2
for a 1.25μm spot size. To determine the relative
92
change in nanotube density, Raman spectra are taken at low sub-threshold laser powers
(50μW) with an accumulation time of 120 seconds before and after high power laser
exposure. Figure 6.6 shows the relative changes in the Raman intensity of carbon
nanotubes after laser exposure. A decreasing exponential dependence can be seen in the
intensity of the G band Raman mode with increasing laser power. The semi-log plot
shown in the inset of the figure, exhibits a turning point at ~300μW, indicating the
threshold for laser burnout. This dose test was repeated several times on different samples
and consistently resulted in the same exponential dependence. From this data and the
1.25μm spot size, we determine the threshold of the burnout process to be 244 μW/μm
2
.
Figure 6.5 (a) SEM image showing the change in density of nanotubes after laser
irradiation. (b) High magnification image showing the nanoparticle substructure.
93
Figure 6.6 Intensity of the G band Raman mode as a function of laser power
6.4 Conclusion
In conclusion, a novel method for creating three-dimensional microstructures from
carbon nanotubes is presented, using a focused laser beam to selectively burn local
regions of a dense forest of multi-walled carbon nanotubes. Raman spectroscopy is used
to quantify the threshold for laser burnout and depth of burnout. Several 3D patterns have
been created with this novel patterning method, illustrating its potential use for the rapid
prototyping of carbon nanotube microstructures. Several interesting phenomena were
observed after laser surface treatment, including the formation of nanoparticles and a
lowering of the nanotube surface density. This patterning method can be used widely to
expand the application of MWNTs and serve as a basis for developing similar patterning
methods in other material systems.
94
This research was supported in part by the James H. Zumberge Fund, the Powell
Foundation and the National Science Foundation Graduate Research Fellowship Program.
Research carried out at the Jet Propulsion Laboratory (JPL), California Institute of
Technology, was supported under a contract with the National Aeronautics and Space
Administration (NASA).
95
Chapter 7: Conclusions
In this thesis, plasmonic enhancement of metal nanoparticles is demonstrated and
discussed in two main aspects, photocatalysis and solar energy conversion. We have
achieved synthesis of single-walled carbon nanotubes and iron oxide nanocrystals by a
plasmon resonant chemical vapor deposition (PRCVD). Also, by analyzing Raman
spectra taken with the same laser used to irradiate samples, we characterize the micron-
scale reaction during the PRCVD process in real-time revealing the sequential growth of
different carbonaceous and metal oxide materials. Furthermore, these in situ Raman
spectra also provide time resolved temperatures of the reaction in the focal volume of
irradiating laser. Additionally, we have also synthesized a novel plasmonic composite
catalyst consisting of Au nanoparticle embedded in Fe
2
O
3,
which exhibits strong catalytic
activity in the oxidation of CO. We believe this hybrid plasmonic nanoparticle catalyst
can open up new possibilities in the local chemistry, enabling new growth pathways of
materials, not possible using standard CVD methods with uniform heating.
In addition to photocatalysis, plasmonic enhancement is also observed in dye
sensitized solar cells (DSSCs) with the incorporation of a thin layer of plasmonic gold
nanoparticles deposited by electron beam deposition. These cells exhibits a 45 % increase
in power conversion efficiency. In this study, finite-difference-time-domain (FDTD)
simulations are also carried out to investigate the distribution of the plasmon-induced
electric fields, which suggest that the enhanced photocurrents arise from the strong
plasmon-induced electric fields due to the presence of gold nanoparticles. Additionally,
96
the photoluminescence spectra of dye molecule TPBP rule out the mechanism of plasmon
energy transfer through Forester resonance energy transfer. Enhancement of solar
methane production is also achieved using the same strong plasmon-induced fields
generated by gold nanoparticles. As a preliminary study, the mechanisms of Au
nanoparticle/TiO
2
-catalyzed photoreduction of CO
2
are distinguished by conducting over
a wide range of wavelengths with different sample configurations. Based on these results,
there are several further studies of interest to obtain a deeper and more quantitative
understanding. Some future directions include exploring the dependence of the doping
concentration in the TiO
2
, the chemical state of the catalytic surface, and how the
plasmonic nanoparticles perform in co-catalyst systems (e.g. TiO
2
-Fe
2
O
3
).
97
References
1. Ackerman, D.M., et al., Diffusivities of Ar and Ne in carbon nanotubes.
Molecular Simulation, 2003. 29(10-11): p. 677-684.
2. Adachi, K., K. Ohta, and T. Mizuno, Photocatalytic Reduction of Carbon-Dioxide
to Hydrocarbon Using Copper-Loaded Titanium-Dioxide. Solar Energy, 1994.
53(2): p. 187-190.
3. Atay, T., J.H. Song, and A.V. Nurmikko, Strongly interacting plasmon
nanoparticle pairs: From dipole-dipole interaction to conductively coupled
regime. Nano Letters, 2004. 4(9): p. 1627-1631.
4. Attridge, J.W., et al., Sensitivity Enhancement of Optical Immunosensors by the
Use of a Surface-Plasmon Resonance Fluoroimmunoassay. Biosensors &
Bioelectronics, 1991. 6(3): p. 201-214.
5. Atwater, H.A., The promise of plasmonics. Scientific American, 2007. 296(4): p.
56-63.
6. Atwater, H.A. and A. Polman, Plasmonics for improved photovoltaic devices.
Nature Materials, 2010. 9(3): p. 205-213.
7. Bachtold, A., et al., Aharonov-Bohm oscillations in carbon nanotubes. Nature,
1999. 397(6721): p. 673-675.
8. Baiker, A., et al., Carbon-Monoxide Oxidation over Catalysts Prepared by in-Situ
Activation of Amorphous Gold-Silver-Zirconium and Gold-Iron-Zirconium Alloys.
Journal of Catalysis, 1995. 151(2): p. 407-419.
9. Barnes, W.L., A. Dereux, and T.W. Ebbesen, Surface plasmon subwavelength
optics. Nature, 2003. 424(6950): p. 824-830.
10. Barsic, D.N., Small-Scale Liquid-State Dynamics in Nanometer Size Devices.
2004, California Institute of Technology: Pasadena, California. p. 103.
11. Boyd, D.A., et al., Plasmon-assisted chemical vapor deposition. Nano Letters,
2006. 6(11): p. 2592-2597.
12. Bronikowski, M.J., CVD growth of carbon nanotube bundle arrays. Carbon, 2006.
44(13): p. 2822-2832.
13. Bronikowski, M.J., H.M. Manohara, and B.D. Hunt, Growth of carbon nanotube
bundle arrays on silicon surfaces. Journal of Vacuum Science & Technology A,
2006. 24(4): p. 1318-1322.
98
14. Cant, N.W. and N.J. Ossipoff, Cobalt promotion of Au/TiO
2
catalysts for the
reaction of carbon monoxide with oxygen and nitrogen oxides. Catalysis Today,
1997. 36(1): p. 125-133.
15. Cataldo, F., An investigation on the optical properties of carbon black, fullerite,
and other carbonaceous materials in relation to the spectrum of interstellar
extinction of light. Fullerenes Nanotubes and Carbon Nanostructures, 2002. 10(2):
p. 155-170.
16. Chakrapani, N., et al., Capillarity-driven assembly of two-dimensional cellular
carbon nanotube foams. Proceedings of the National Academy of Sciences of the
United States of America, 2004. 101(12): p. 4009-4012.
17. Chen, M.S. and D.W. Goodman, The structure of catalytically active gold on
titania. Science, 2004. 306(5694): p. 252-255.
18. Chiashi, S., et al., Temperature dependence of Raman scattering from single-
walled carbon nanotubes: Undefined radial breathing mode peaks at high
temperatures. Japanese Journal of Applied Physics, 2008. 47(4): p. 2010-2015.
19. Choi, W.B., et al., Fully sealed, high-brightness carbon-nanotube field-emission
display. Applied Physics Letters, 1999. 75(20): p. 3129-3131.
20. Chopra, S., et al., Selective gas detection using a carbon nanotube sensor.
Applied Physics Letters, 2003. 83(11): p. 2280-2282.
21. Chou, S.Y., P.R. Krauss, and P.J. Renstrom, Imprint lithography with 25-
nanometer resolution. Science, 1996. 272(5258): p. 85-87.
22. Cooper, E.B., et al., Terabit-per-square-inch data storage with the atomic force
microscope. Applied Physics Letters, 1999. 75(22): p. 3566-3568.
23. Cubukcu, E., et al., Plasmonic laser antenna. Applied Physics Letters, 2006.
89(9): p. 093120.
24. Cumpston, B.H., et al., Two-photon polymerization initiators for three-
dimensional optical data storage and microfabrication. Nature, 1999. 398(6722):
p. 51-54.
25. deFaria, D.L.A., S.V. Silva, and M.T. deOliveira, Raman microspectroscopy of
some iron oxides and oxyhydroxides. Journal of Raman Spectroscopy, 1997.
28(11): p. 873-878.
26. Duche, D., et al., Improving light absorption in organic solar cells by plasmonic
contribution. Solar Energy Materials and Solar Cells, 2009. 93(8): p. 1377-1382.
27. Ellingson, R.J., et al., Highly efficient multiple exciton generation in colloidal
PbSe and PbS quantum dots. Nano Letters, 2005. 5(5): p. 865-871.
99
28. Falvo, M.R., et al., Bending and buckling of carbon nanotubes under large strain.
Nature, 1997. 389(6651): p. 582-584.
29. Fan, Y.W., et al., Probing Planck's Law with Incandescent Light Emission from a
Single Carbon Nanotube. Physical Review Letters, 2009. 102(18): p. 4.
30. Ferrari, A.C. and J. Robertson, Interpretation of Raman spectra of disordered and
amorphous carbon. Physical Review B, 2000. 61(20): p. 14095-14107.
31. Ferry, V.E., et al., Improved red-response in thin film a-Si:H solar cells with soft-
imprinted plasmonic back reflectors. Applied Physics Letters, 2009. 95(18): p
183503.
32. Fu, D.C., et al., Dye-Sensitized Back-Contact Solar Cells. Advanced Materials,
2010. 22(38): p. 4270.
33. Gobin, A.M., et al., Near-infrared resonant nanoshells for combined optical
imaging and photothermal cancer therapy. Nano Letters, 2007. 7(7): p. 1929-
1934.
34. Gratzel, M., Photoelectrochemical cells. Nature, 2001. 414(6861): p. 338-344.
35. Grimes, C.A., Synthesis and application of highly ordered arrays of TiO
2
nanotubes. Journal of Materials Chemistry, 2007. 17(15): p. 1451-1457.
36. Guo, X.F., et al., Covalently bridging gaps in single-walled carbon nanotubes
with conducting molecules. Science, 2006. 311(5759): p. 356-359.
37. Halmann, M., Photoelectrochemical Reduction of Aqueous Carbon-Dioxide on P-
Type Gallium-Phosphide in Liquid Junction Solar-Cells. Nature, 1978. 275(5676):
p. 115-116.
38. Halmann, M., M. Ulman, and B. Aurianblajeni, Photochemical Solar Collector
for the Photoassisted Reduction of Aqueous Carbon-Dioxide. Solar Energy, 1983.
31(4): p. 429-431.
39. Hammer, B. and J.K. Norskov, Why Gold is the Noblest of All the Metals. Nature,
1995. 376(6537): p. 238-240.
40. Hara, K., et al., Molecular design of coumarin dyes for efficient dye-sensitized
solar cells. Journal of Physical Chemistry B, 2003. 107(2): p. 597-606.
41. Harrison, B.S., et al., Oxygen producing biomaterials for tissue regeneration.
Biomaterials, 2007. 28(31): p. 4628-4634.
42. Haruta, M., Novel Catalysis of Gold Deposited on Metal Oxides. Catal. Surveys
Jpn., 1997. 1: p. 61.
100
43. Haruta, M., Size- and support-dependency in the catalysis of gold. Catalysis
Today, 1997. 36(1): p. 153-166.
44. Haruta, M., et al., Low-Temperature Oxidation of Co over Gold Supported on
Tio2, Alpha-Fe2o3, and Co3o4. Journal of Catalysis, 1993. 144(1): p. 175-192.
45. Haruta, M., et al., Gold Catalysts Prepared by Coprecipitation for Low-
Temperature Oxidation of Hydrogen and of Carbon Monoxide. Journal of
Catalysis, 1989. 115(2): p. 301-309.
46. Hoflund, G.B., et al., Au/MnO
x
Catalytic Performance Characteristics for Low-
Temperature Carbon Monoxide Oxidation. Applied Catalysis B-Environmental,
1995. 6(2): p. 117-126.
47. Holt, J.K., et al., Fast mass transport through sub-2-nanometer carbon nanotubes.
Science, 2006. 312(5776): p. 1034-1037.
48. Hong, Y.C. and H.S. Uhm, Superhydrophobicity of a material made from
multiwalled carbon nanotubes. Applied Physics Letters, 2006. 88(24): p.244101.
49. Huang, L., et al., Matrix-Assisted Dip-Pen Nanolithography and Polymer Pen
Lithography. Small, 2010. 6(10): p. 1077-1081.
50. Huang, S.M., X.Y. Cai, and J. Liu, Growth of millimeter-long and horizontally
aligned single-walled carbon nanotubes on flat substrates. Journal of the
American Chemical Society, 2003. 125(19): p. 5636-5637.
51. Huffman, C.F.B.a.D.R., Absorption and scattering of light by small particles.
1983, New York: Wiley-Interscience.
52. Hung, W.H., et al., Plasmon Resonant Enhancement of Carbon Monoxide
Catalysis. Nano Letters, 2010. 10(4): p. 1314-1318.
53. Hung, W.H., et al., Laser Directed Growth of Carbon-Based Nanostructures by
Plasmon Resonant Chemical Vapor Deposition. Nano Letters, 2008. 8(10): p.
3278-3282.
54. Hyun, J., et al., Molecular recognition-mediated fabrication of protein
nanostructures by dip-pen lithography. Nano Letters, 2002. 2(11): p. 1203-1207.
55. Inoue, T., et al., Photoelectrocatalytic Reduction of Carbon-Dioxide in Aqueous
Suspensions of Semiconductor Powders. Nature, 1979. 277(5698): p. 637-638.
56. Jain, P.K., W. Qian, and M.A. El-Sayed, Ultrafast electron relaxation dynamics
in coupled metal nanoparticles in aggregates. Journal of Physical Chemistry B,
2006. 110(1): p. 136-142.
101
57. Jeanmaire, D.L. and R.P. Vanduyne, Surface Raman Spectroelectrochemistry .1.
Heterocyclic, Aromatic, and Aliphatic-Amines Adsorbed on Anodized Silver
Electrode. Journal of Electroanalytical Chemistry, 1977. 84(1): p. 1-20.
58. Jones, D.E.H., Science of fullerenes and carbon nanotubes - Dresselhaus,MS,
Dresselhaus,G, Eklund,PC. Nature, 1996. 381(6581): p. 384-384.
59. Joseph, P., et al., Slippage of water past superhydrophobic carbon nanotube
forests in microchannels. Physical Review Letters, 2006. 97(15): p156104.
60. Jung, Y., et al., Controlling the aligned growth of carbon nanotubes by substrate
selection and patterning. Making Functional Materials with Nanotubes, 2002. 706:
p. 55-60
61. K. Kneipp, Y.W., H. Kneipp, L. T. Perelman, I. Itzkan, R. R. Dasari, and M. S.
Feld, Phys. Rev. Lett., 1997. 78: p. 1667
62. Kang, Y.M. and B.Z. Wan, Preparation of Gold in Y-type Zeolite for Carbon
Monoxide Oxidation. Applied Catalysis A-General, 1995. 128(1): p. 53-60.
63. Kawata, S., et al., Finer features for functional microdevices - Micromachines can
be created with higher resolution using two-photon absorption. Nature, 2001.
412(6848): p. 697-698.
64. Klar, T., et al., Surface-plasmon resonances in single metallic nanoparticles.
Physical Review Letters, 1998. 80(19): p. 4249-4252.
65. Kneipp, K., et al., Surface-enhanced and normal Stokes and anti-Stokes Raman
spectroscopy of single-walled carbon nanotubes. Physical Review Letters, 2000.
84(15): p. 3470-3473.
66. Kneipp, K., et al., Single molecule detection using surface-enhanced Raman
scattering (SERS). Physical Review Letters, 1997. 78(9): p. 1667-1670.
67. Kong, J., et al., Synthesis of individual single-walled carbon nanotubes on
patterned silicon wafers. Nature, 1998. 395(6705): p. 878-881.
68. Kulkarni, A.P., et al., Plasmon-Enhanced Charge Carrier Generation in Organic
Photovoltaic Films Using Silver Nanoprisms. Nano Letters, 2010. 10(4): p. 1501-
1505.
69. Kumar, R., H. Zhou, and S.B. Cronin, Surface-enhanced Raman spectroscopy and
correlated scanning electron microscopy of individual carbon nanotubes. Applied
Physics Letters, 2007. 91(22): p. 223105
70. Lau, K.K.S., et al., Superhydrophobic carbon nanotube forests. Nano Letters,
2003. 3(12): p. 1701-1705.
102
71. Lee, J.H., et al., High efficiency polymer solar cells with wet deposited plasmonic
gold nanodots. Organic Electronics, 2009. 10(3): p. 416-420.
72. Li, N.H., W. Wu, and S.Y. Chou, Sub-20-nm alignment in nanoimprint
lithography using Moire fringe. Nano Letters, 2006. 6(11): p. 2626-2629.
73. Li, X.S., et al., Bottom-up growth of carbon nanotube multilayers: Unprecedented
growth. Nano Letters, 2005. 5(10): p. 1997-2000.
74. Lin, S.D., M. Bollinger, and M.A. Vannice, Low Temperature CO Oxidation over
Au/TiO
2
and Au/SiO
2
Catalysts. Catalysis Letters, 1993. 17(3-4): p. 245-262.
75. Lofas, S., et al., Bioanalysis with Surface-Plasmon Resonance. Sensors and
Actuators B-Chemical, 1991. 5(1-4): p. 79-84.
76. Loo, C., et al., Immunotargeted nanoshells for integrated cancer imaging and
therapy. Nano Letters, 2005. 5(4): p. 709-711.
77. Maier, S.A., Plasmonics - Towards subwavelength optical devices. Current
Nanoscience, 2005. 1(1): p. 17-23.
78. Maier, S.A., et al., Local detection of electromagnetic energy transport below the
diffraction limit in metal nanoparticle plasmon waveguides. Nature Materials,
2003. 2(4): p. 229-232.
79. Maruo, S., O. Nakamura, and S. Kawata, Three-dimensional microfabrication
with two-photon-absorbed photopolymerization. Optics Letters, 1997. 22(2): p.
132-134.
80. McCarty, K.F., Inelastic Light-Scattering in -Fe
2
O
3
: Phonon vs Magnon
Scattering. Solid State Communications, 1988. 68(8): p. 799-802.
81. Mertens, H., Ellipsometry measurements performed with Hans Mertens
at the AMOLF institut.
82. Meyer, J.C., et al., Raman Modes of Index-Identified Freestanding Single-Walled
Carbon Nanotubes. Phys. Rev. Lett., 2005. 95(21): p. 217401-4.
83. Misewich, J.A., et al., Electrically induced optical emission from a carbon
nanotube FET. Science, 2003. 300(5620): p. 783-786.
84. Moskovits, M., Surface-Enhanced Spectroscopy. Reviews of Modern Physics,,
1985. 57(3): p. p. 783-826.
85. N. J. Foster, M.J.B., J. A. Creighton, M. J. Cole, Magnetite or Hematite? The
Iron-Oxides Returned by Stardust. Lunar and Planetary Science XXXIX, 2008: p.
1759.
103
86. Nazeeruddin, M.K., et al., Combined experimental and DFT-TDDFT
computational study of photoelectrochemical cell ruthenium sensitizers. Journal
of the American Chemical Society, 2005. 127(48): p. 16835-16847.
87. Nelson, J., The physics of solar cells. 2003: London, Imperial College Press.
88. Nordlander, P., et al., Plasmon Hybridization in Nanoparticle Dimers. Nano
Letters, 2004. 5: p. 899.
89. Ohkita, H., et al., Charge carrier formation in polythiophene/fullerene blend films
studied by transient absorption spectroscopy. Journal of the American Chemical
Society, 2008. 130(10): p. 3030-3042.
90. Oregan, B. and M. Gratzel, A Low-Cost, High-Efficiency Solar-Cell Based on
Dye-Sensitized Colloidal Tio2 Films. Nature, 1991. 353(6346): p. 737-740.
91. Oubre, C. and P. Nordlander, Finite-difference Time-domain Studies of the
Optical Properties of Nanoshell Dimers. J. Phys. Chem. B, 2005. 109: p. 10042.
92. Oulton, R.F., et al., Plasmon lasers at deep subwavelength scale. Nature, 2009.
93. Pissuwan, D., S.M. Valenzuela, and M.B. Cortie, Therapeutic possibilities of
plasmonically heated gold nanoparticles. Trends in Biotechnology, 2006. 24(2): p.
62-67.
94. Prikulis, J., et al., Optical spectroscopy of single trapped metal nanoparticles in
solution. Nano Letters, 2004. 4(1): p. 115-118.
95. Rand, B.P., P. Peumans, and S.R. Forrest, Long-range absorption enhancement in
organic tandem thin-film solar cells containing silver nanoclusters. Journal of
Applied Physics, 2004. 96(12): p. 7519-7526.
96. Sanchez, A., et al., When gold is not noble: Nanoscale gold catalysts. Journal of
Physical Chemistry A, 1999. 103(48): p. 9573-9578.
97. Sazonova, V., et al., A tunable carbon nanotube electromechanical oscillator.
Nature, 2004. 431(7006): p. 284-287.
98. Schnadt, J., et al., Experimental evidence for sub-3-fs charge transfer from an
aromatic adsorbate to a semiconductor. Nature, 2002. 418(6898): p. 620-623.
99. Schwank, J., Catalytic Gold: Applications of Elemental Gold in Heterogeneous
Catalysis. Gold Bull., 1983. 16: p. 103.
100. Shao, D.B. and S.C. Chen, Numerical simulation of surface-plasmon-assisted
nanolithography. Optics Express, 2005. 13(18): p. 6964-6973.
101. Sherry, L.J., et al., Localized surface plasmon resonance spectroscopy of single
silver nanocubes. Nano Letters, 2005. 5(10): p. 2034-2038.
104
102. Srituravanich, W., et al., Plasmonic nanolithography. Nano Letters, 2004. 4(6): p.
1085-1088.
103. Su, K.H., et al., Interparticle coupling effects on plasmon resonances of nanogold
particles. Nano Letters, 2003. 3(8): p. 1087-1090.
104. Tang, T., et al., Complementary response of In
2
O
3
nanowires and carbon
nanotubes to low-density lipoprotein chemical gating. Applied Physics Letters,
2005. 86(10): p. 103903.
105. Tanielyan, S.K. and R.L. Augustine, Effect of Catalyst Pretreatment on the
Oxidation of Carbon Monoxide over Coprecipitated Gold Catalysts. Applied
Catalysis A: General, 1992. 85(1): p. 73-87.
106. Tans, S.J., A.R.M. Verschueren, and C. Dekker, Room-temperature transistor
based on a single carbon nanotube. Nature, 1998. 393(6680): p. 49-52.
107. Thampi, K.R., J. Kiwi, and M. Gratzel, Methanation and Photo-Methanation of
Carbon-Dioxide at Room-Temperature and Atmospheric-Pressure. Nature, 1987.
327(6122): p. 506-508.
108. Toussaint, K.C., et al., Plasmon resonance-based optical trapping of single and
multiple Au nanoparticles. Optics Express, 2007. 15(19): p. 12017-12029.
109. Valden, M., X. Lai, and D.W. Goodman, Onset of catalytic activity of gold
clusters on titania with the appearance of nonmetallic properties. Science, 1998.
281(5383): p. 1647-1650.
110. Varghese, O.K., et al., High-Rate Solar Photocatalytic Conversion of CO
2
and
Water Vapor to Hydrocarbon Fuels. Nano Letters, 2009. 9(2): p. 731-737.
111. Wang, C.J., et al., Visible Light Photoreduction of CO2 Using CdSe/Pt/TiO
2
Heterostructured Catalysts. Journal of Physical Chemistry Letters, 2010. 1(1): p.
48-53.
112. Wu, J.C.S., et al., Application of optical-fiber photoreactor for CO
2
photocatalytic
reduction. Topics in Catalysis, 2008. 47(3-4): p. 131-136.
113. Yuan, Y.Z., et al., Preparation of supported gold catalysts from gold complexes
and their catalytic activities for CO oxidation. Catalysis Letters, 1996. 42(1-2): p.
15-20.
114. Zakeeruddin, S.M., et al., Molecular engineering of photosensitizers for
nanocrystalline solar cells: Synthesis and characterization of Ru dyes based on
phosphonated terpyridines. Inorganic Chemistry, 1997. 36(25): p. 5937-5946.
115. Zhang, W.F., et al., Raman scattering study on anatase TiO
2
nanocrystals. Journal
of Physics D-Applied Physics, 2000. 33(8): p. 912-916.
105
116. Zhu, X.Y., Q. Yang, and M. Muntwiler, Charge-Transfer Excitons at Organic
Semiconductor Surfaces and Interfaces. Accounts of Chemical Research, 2009.
42(11): p. 1779-1787.
117. Zou, S.L. and G.C. Schatz, Silver nanoparticle array structures that produce
giant enhancements in electromagnetic fields. Chemical Physics Letters, 2005.
403(1-3): p. 62-67.
Abstract (if available)
Abstract
This thesis is dedicated to exploring the potential applications of plasmonic metal nanoparticles and understanding their fundamental enhancement mechanisms. Photocatalysis and solar energy conversion are the two main topics investigated in this work. In Chapter 2, we demonstrate the growth of a variety of carbonaceous materials by plasmonic heating. When the metal nanoparticles are irradiated with a laser near their plasmon resonant frequency, a localized region at high temperature is achieved due to the ohmic losses at the plasmon resonance. In this plasmon resonant chemical vapor deposition (PRCVD) process, high temperatures are created at the surface of the plasmonic nanoparticles to trigger the dissociation of gaseous precursors, (e.g. carbon monoxide), which results in the deposition of amorphous carbon, graphitic carbon, and carbon nanotubes. The formation of iron oxide nanocrystals is also observed at the beginning of the reaction due to a trace amount of iron pentacarbonyl in the CO feed gas. The growth of iron oxide at the surface of gold nanoparticles forms a new type of composite catalyst, Au nanoparticle/Fe₂O₃, which also catalyzes the growth of carbon nanotubes through this plasmon excitation process. The real time temperature dependence and sequential growth of different carbonaceous and metal oxide materials are monitored and characterized by Raman spectroscopy and infrared spectroscopy. Additionally, pre-defined microstructure geometries of crystalline iron oxide and carbon nanotubes are demonstrated by rastering the focused laser spot during the growth process in a controlled fashion. ❧ In Chapter 3, the concentrations of gas phase reaction products are observed in real time using mass spectrometry, which is used to evaluate the performance of Au nanoparticle/Fe₂O₃ composite catalysts. This new plasmonic composite catalyst exhibits an excellent catalytic ability in the CO oxidation reaction, which exceeds that of the Au nanoparticles and Fe₂O₃ alone. This indicates that this reaction is not driven solely by thermal (plasmonic) heating of the gold nanoparticles, but relies intimately on the interaction of these two materials. This hybrid plasmonic nanoparticle catalyst and PRCVD method open up new possibilities in the local chemistry, enabling new growth pathways of materials, not possible using standard CVD methods with uniform heating. ❧ In addition to photocatalysis, we also explored plasmonic enhancement of solar energy conversion. In Chapter 4, plasmonic gold nanoparticles are incorporated into dye sensitized solar cells (DSSCs) by electron beam deposition. Increased photocurrents are observed due to the thin layer of plasmonic gold, which results in a 45% increase in the cell’s power conversion efficiency. This enhancement is attributed to the strong plasmon-induced electric field from the presence of gold nanoparticles, as indicated by electromagnetic finite-difference-time-domain (FDTD) simulations. Additionally, the photoluminescence spectra of the TPBP dye molecule rule out the mechanism of plasmon energy transfer through a Forester resonance process. Another potential application of plasmonic nanoparticles that we have explored is solar fuel production. In Chapter 5, we demonstrate an enhancement of solar methane production by the reduction of aqueous carbon dioxide (CO₂) in the visible wavelength range. The underlying enhancement mechanisms of the Au nanoparticle/TiO₂-catalyzed photoreduction of CO₂ are investigated by irradiating several different sample configurations with a wide range of wavelengths. Based on these results, we attribute the plasmonmic enhancement to the local electric field enhancement. However, several questions remain open and further studies will be required in order to obtain a deeper and more quantitative understanding of the plasmonic enhancement process. Some future directions include exploring the dependence of the doping concentration in the TiO₂, the chemical state of the catalytic surface, and how the plasmonic nanoparticles perform in co-catalyst systems (e.g., TiO₂-Fe₂O₃). Chapter 6 describes a side project carried out at the beginning of my graduate work. In this Chapter, we report a novel method for creating three-dimensional carbon nanotube structures from dense, vertically-grown carbon nanotube forests. The minimum power density for burning carbon nanotubes is also determined in this study.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Surface plasmon resonant enhancement of photocatalysis
PDF
Electromagnetic modeling of plasmonic nanostructures
PDF
Hot carrier enhanced photocatalysis in plasmon resonant metal grating systems
PDF
Enhanced photocatalysis on titanium oxide passivated III-V semiconductors
PDF
Carbon nanotube nanoelectronics and macroelectronics
PDF
Design and characterization of metal and semiconducting nanostructures and nanodevices
PDF
Raman spectroscopy of carbon nanotubes under axial strain and surface-enhanced Raman spectroscopy of individual carbon nanotubes
PDF
Nanomaterials for macroelectronics and energy storage device
PDF
Controlled synthesis, characterization and applications of carbon nanotubes
PDF
Carbon material-based nanoelectronics
PDF
Fabrication and application of plasmonic nanostructures: a story of nano-fingers
PDF
Electronic and optoelectronic devices based on quasi-metallic carbon nanotubes
PDF
GaAs nanowire optoelectronic and carbon nanotube electronic device applications
PDF
Carbon nanotube macroelectronics
PDF
Artificial photosynthesis on titanium oxide passivated III-V semiconductors
PDF
Printed and flexible carbon nanotube macroelectronics
PDF
Printed electronics based on carbon nanotubes and two-dimensional transition metal dichalcogenides
PDF
Plasma emission across metal nanoparticle surfaces and semiconductor -liquid interface characterizations
PDF
Hot carriers in bare metals and photocatalytically active defect sites in dielectric/metal structures
PDF
Optoelectronic, thermoelectric, and photocatalytic properties of low dimensional materials
Asset Metadata
Creator
Hung, Wei Hsuan
(author)
Core Title
Plasmonic enhancement of catalysis and solar energy conversion
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Materials Science
Publication Date
06/14/2011
Defense Date
05/23/2011
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
carbon nanotubes,gold nanoparticles,OAI-PMH Harvest,photocatalysis,plasmonic,solar energy conversion
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Cronin, Stephen B. (
committee chair
), Goo, Edward K. (
committee member
), Zhou, Chongwu (
committee member
)
Creator Email
wave0626@gmail.com,weihsuah@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c127-615627
Unique identifier
UC1356825
Identifier
usctheses-c127-615627 (legacy record id)
Legacy Identifier
etd-HungWeiHsu-23.pdf
Dmrecord
615627
Document Type
Dissertation
Rights
Hung, Wei Hsuan
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Repository Email
cisadmin@lib.usc.edu
Tags
carbon nanotubes
gold nanoparticles
photocatalysis
plasmonic
solar energy conversion