University of Southern California Dissertations and Theses

Fabrication and application of plasmonic nanostructures: a story of nano-fingers

(USC Thesis Other)

Fabrication and application of plasmonic nanostructures: a story of nano-fingers

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

FABRICATION AND APPLICATION OF
PLASMONIC NANOSTRUCTURES: A STORY OF
NANO-FINGERS
By
Boxiang Song

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
(ELECTRICAL ENGINEERING)

May 2020
Copyright 2020 Boxiang Song

To my family for your endless love

iii

Acknowledgements
I would love to thank, first and foremost, my advisor, Prof. Wei Wu for his mentorship
and support throughout all my doctoral study. It is a great honor and bliss of my life to be
his Ph. D student. His passion to life, devotion to research, patience and modesty to people
are always inspiring me to be not only a better researcher but also a better person in general.
He is always there to help me, whenever I met a problem, no matter what problem it is. I
am grateful for his knowledgeable guidance and kind encouragement, which helped me go
through up and downs in research and everyday life. The knowledge, confidence and
philosophy about both research and life I learned from him are my lifelong treasure.

My sincere thanks to my dissertation and qualifying exam committee members, Prof.
Stephan Haas, Prof. Stephen Cronin, Prof. Han Wang and Prof. Aiichiro Nakano, for their
kind service and support.

I have collaborated with an amazing set of researchers, whose contributions are been
undoubtedly essential in the completion of my doctoral studies. I would like to express my
sincere gratitude to Prof. Stephan Haas, Dr. Roelof Groenewald, and Zhihao Jiang for the
support in theoretical analysis and great physical insights, Prof. Stephen Cronin, Dr.
Haotian Shi and Bo Wang for the help and knowledge in ALD deposition, fluorescence
and Raman spectra characterization, Dr. Adam Schwartzberg and Dr. Stefano Cabrini from
Lawrence Berkeley National Laboratory for the support in fabrication process and Prof.
Fanxin Liu from Zhejiang University of Technology for the help in imaging and
iv

characterization. Special thanks to Prof. Han Wang for valuable suggestions and help for
my academia career development, Prof. Armand Tanguay for advices on optical design and
how to present my work.

I would also like to thank all our research group members. I truly enjoyed working with
you in such a friendly, encouraging and helpful environment. I would like to thank Dr.
Yuhan Yao for his help in improving nano-finger fabrication process and all his supports
from beginning to end, Dr. He Liu for the tutorial on nanoimprint lithography and his
encourage in my darkest days, Dr. Yifei Wang for his help in fabrication mentoring and
his companion in my loneliest days, Dr. Yuanrui Li for sharing great ideas and fighting
through the early hard days together, Hao Yang for the collaboration in memristor project
and his help and sharing in my everyday life that really makes my every day, Deming Meng
for the ALD film characterization and his efforts to make my PhD life entertaining, Buyun
Chen for the collaboration in memristor project and his devoted attitude on science that
really encourages me, Yunxiang Wang for the demonstration of photocatalysis using nano-
fingers and his kind supports in both my research and daily life, Pan Hu for the realization
of inverse printing using nano-fingers and his self-discipline that always inspires me, Tse-
Hsien Ou for the SERS measurement on chemical detection and experience sharing on my
family life development, Zerui Liu for the sample and solution preparation on spectra
characterization and his hardworking that constantly motivates me. I am thankful for all
the friendship and supports from them. Wu’s group is not just a research group to me, but
a family I will treasure forever.

v

I am so lucky to have some fabulous friends outside of the group. Shanyuan, Yingjun,
Lurui, Jiachen, Haotian, Jihan, Bo, Sisi, Yu, Bofan, Zhi, Lang, Aoyang, Huan, Xiaodong,
Yongkui. Thank you all for being on my side all the time and making my life at USC
eventful and full of great memories. Special thanks to Dr. Shanyuan Niu, for the ten-year
friendship and supports on every aspect and to Dr. Yingjun Lyu as my best neighbor and
companion all through my PhD years.

Last, and most importantly, I would like to thank my family. I can never thank my parents
too much for all their endless love, support and sacrifice. They are the strongest shield in
my mind, and I would not accomplish this journey without them. Thank you, Fan, my wife,
for all your support all through many years’ oversea marriage and so many lonely
anniversaries. Thank you for being my best Fan during my prime time and my toughest
time. Thank you for helping me accomplish each step of my PhD Life. Fortunately, we
went through, together.

vi

Abstract
Gap plasmonic nanostructures are of great interest due to their ability to concentrate light
into small volumes, which enables the creation of a highly concentrated electromagnetic
(EM) field within nanoscale volumes that can enable the ultra-high sensitivity chemical
and biological detection. Theoretical studies, considering quantum mechanical effects,
have predicted the optimal spatial gap between adjacent nanoparticles to be in the sub-
nanometer regime in order to achieve the strongest possible field enhancement. While
many attempts have been made in recent years to fabricate gap plasmonic nanostructures
with sub-5 nm gaps, the limitation of traditional fabrication techniques poses serious
obstacles to reliably producing precisely controllable nano-gaps with high throughput.

The first part of this dissertation mainly focuses on the realization of the theoretically
predicted gap plasmonic structure. A technology is proposed to fabricate gap plasmonic
structures with sub-nanometer resolution, high reliability and high throughput using
collapsible nano-fingers. Different fabrication technological details are discussed and
compared for nano-fingers collapse in various pattern. The capillary force driven collapse
mechanism is studied for further fabrication designs. Atomic layer deposition process, as
the key role in determining the gap size, is optimized for the performances of nano-fingers.

This approach enables us to systematically investigate the quantum effects in gap
plasmon. We experimentally proved and observed optimal gap sizes for gap plasmonic
structures with different dielectric spacers. The effect of tunneling barrier height is studied
vii

by using different dielectric spacers, and the quantum nonlocal effect is also observed at
the meantime. A straightforward theoretical model is presented to analyze the quantum
effects in collapsible nano-fingers.

The application of nano-fingers in plasmon enhanced Raman spectroscopy is
demonstrated in the following. The multitude of benefits deriving from the unique nano-
finger nanostructure allows extremely high detection sensitivity at the single-molecular
level to be realized as demonstrated through bi-analyte surface-enhanced Raman scattering
measurement. The real-time Raman detection for big living yeast cells is also
demonstrated, which suggests a promising emerging technology for single-molecular
label-free sensing and opens the door to a wide range of opportunities disease diagnosis
and biological content identification.

In addition, plasmon enhanced fluorescence is studied using the collapsible nano-fingers.
In this part, we studied the mechanism of all contributing factors in plasmon enhanced
fluorescence (PEF) and propose a guiding rule for further PEF related design. The
quenching mechanism in PEF is analyzed experimentally and theoretically. The gap
plasmon induced PEF is studied in the sub-5 nm gap scale with sub-nanometer accuracy
experimentally. SERS signal is obtained simultaneous in PEF measurement, which offers
in-situ monitor on plasmonic field enhancement. By using collapsible nano-fingers, we
realized ultra-high enhancement factor over large area uniformly with stable and constant
fluorescence. That paves the way for the applications of PEF in chemical sensing and
biological labeling with high sensitivity requirements.
viii

By using large-area collapsible nanofingers, we can combine noble metal nanoparticles
and thin TiO2 films to create a novel plasmonic photocatalyst. The metal nanoparticle pairs
can couple the energy of incident light into ultra-small region inside the TiO2 film and then
the energy can be dissipated to accelerate chemical reaction by creating electron hole pairs.
Finite-difference time-domain simulations of this plasmonic photocatalyst shows that the
enhanced photocatalytic activity is due to the large enhancement of the local electric field
at the hot spots, which increases the electron–hole pair generation rate at the TiO2 surface,
and hence the photodecomposition rate of methyl orange. The mechanism of the
enhancement is discussed, and theoretical enhancement factor is calculated.

ix

Table of Contents

Acknowledgments……………………………………………………………………...iii
Abstract………………………………………………………………………………...vi
Chapter 1. Introduction for Gap Plasmonic Nanostructures............................................ 1
1.1 Gap Plasmon and Predicted Optimal Gap Size .......................................................1
1.2 Previous Methods to Fabricate Gap Plasmonic Nanostructures .............................2
1.3 Design of Collapsible Nano-Fingers .......................................................................7
Chapter 2. Fabrication of Nano-fingers ......................................................................... 10
2.1 Fabrication of Gold Nano-fingers by EBL mother molds ....................................10
2.2 Fabrication of Gold Nano-fingers by Interference Lithography ...........................15
2.3 Nano-finger Geometry Optimization for Successful Collapse .............................21
2.4 Capillary Force Induced Collapsing Mechanism ..................................................24
2.5 Atomic Layer Deposition for the Dielectric Spacer..............................................32
2.6 Crystallinity Characterization of ALD Film .........................................................35
Chapter 3. Investigation of Quantum Effects in Collapsible Nano-Fingers .................. 39
3.1 Absorption Spectrum Characterization .................................................................39
3.2 Effects of Different Tunneling Barrier Heights ....................................................43
x

3.3 Theoretical Analysis on Field Enhancement in Nano-Fingers .............................45
3.4 Theoretical Analysis on Resonant Frequency in Nano-Fingers ...........................49
Chapter 4. Application of Nano-Fingers in Plasmon Enhanced SERS ......................... 52
4.1 Introduction to Plasmon Enhanced SERS .............................................................52
4.2 Fabrication of Nano-Fingers for Plasmon Enhanced SERS .................................53
4.3 Characterizations of Gap-Plasmonic Nanofingers with Uniform 2 nm Gaps .......58
4.4 SERS Characterizations for ta-C Coated Gap Plasmonic Nanofingers ................61
4.5 Numerical Simulations ..........................................................................................63
4.6 Single-Molecule Detection ...................................................................................69
Chapter 5. Application of Nano-Fingers in Plasmon Enhanced Fluorescence.............. 71
5.1 Introduction to Plasmon Enhance Fluorescence ...................................................71
5.2 Capillary Force Induced Molecule Trapping in Nano-Fingers .............................73
5.3 Characterizations of Nano-fingers for PEF ...........................................................76
5.4 Simultaneous Monitor on Plasmon Enhanced Fluorescence and SERS ...............80
5.5 Dielectric Quenching ............................................................................................85
5.6 Minimize the Quenching Effect ............................................................................87
Chapter 6. Plasmon Enhanced Photocatalysis using Collapsible Nano-fingers ............ 94
xi

6.1 Introduction to Plasmon Enhanced Photocatalysis ...............................................94
6.2 Fabrication of Nano-fingers for Photocatalysis ....................................................96
6.3 Enhanced Methyl-Orange Decomposition ............................................................98
6.3 Numerical Calculations .......................................................................................101
Chapter 7. Conclusion and Future Works ................................................................... 104
REFERENCES ............................................................................................................ 109

1

Chapter 1. Introduction for Gap Plasmonic
Nanostructures

1.1 Gap Plasmon and Predicted Optimal Gap Size
Plasmonic nanostructures have many fascinating properties. For example, they
can focus light energy onto a small volume at nanometer scales.
1-6
This feature
makes them a promising platform for various applications, including optical
communication,
7
disease diagnosis,
8
and chemical sensing.
9
One class of the most
intriguing plasmonic nanostructures is that of nanoparticle pairs with small inter-
particle gaps, because it can create strong field enhancement at a hot spot between
the two particles.
6, 10-13
It has been theoretically predicted that the optimal hot spot
is a sub-nanometer gap between two metallic particles,
14-18
considering both
classical electromagnetic theory and quantum mechanical effects, as shown in
Figure 1.1. While many attempts have been made in recent years to fabricate gap
plasmonic nanostructures with sub-5 nm gaps,
19-24
the limitation of traditional
fabrication techniques poses serious obstacles to reliably producing precisely
controllable nano-gaps with high throughput. In order to experimentally investigate
these theoretical predictions of optimal gap size and design such structures for
practical applications, one needs to demonstrate accurately controlled gap size in
large periodic nano-gap arrays. Moreover, it is important to understand electron
2

tunneling across the gap in such structures, as this appears to be a key mechanism
in controlling plasmonic enhancement.

Figure 1.1 Field enhancement theoretically calculated in the middle of the junction
for an incident pulse resonant with the bonding dimer plasmon BDP.

1.2 Previous Methods to Fabricate Gap Plasmonic Nanostructures

Many previous works to fabricate sub-5 nm gap plasmonic structures have been
reported before.
19-24
Kevin J. Savage et al. demonstrated nanoscale plasmonic
cavity by using two gold-nanoparticle-terminated atomic force microscope (AFM)
3

tips oriented tip-to-tip,
19
as shown in Fig. 1.2. However, this method requires
ultrahigh precision in AFM tip alignment and is hard to handle at sub-5 nm gap
precision.

Figure 1.2 a) Scheme for simultaneous optical and electrical measurements of
plasmonic cavity formed between two Au-coated tips, shown in dark-field
microscope images (b) and false-colour scanning electron microscope image (c)
of a typical tip, end radius R=150 nm.

4

Wenqi Zhu et al. proposed a two-step EBL process to achieve metallic dimers
separated by angstrom-scale gaps
20
(Figure 1.3) by avoiding the limitation from the
resolution of the electron-beam resist and other factors. This is an elegant method
to obtain various gap sizes at sub-5 nm scale. However, the uncertainty of the angle
and alignment of two EBL process will pose significant randomness in the obtained
gap sizes. On the other hand, this method can not fabricate gap plasmonic structure
array with uniform gaps. Also, the EBL based process is high cost and low
efficiency in large area sample fabrication.

5

Figure 1.3 Lithographic nanofabrication of dimers with atomic length-scale gap-
width. (a) Schematic depiction of the two-step EBL process for the fabrication.
Solid objects represent the patterns in the first step. Dashed objects represent the
patterns in the second EBL, aligned with the first EBL using a procedure that
corrects offset position (error < 10 nm) and rotational angle (error < 1 mrad). This
two-step EBL process generates patterns from dimers with overlapping features
(overlapped by 40 nm) on the lower left corner, to dimers with large gap-widths
(40 nm) on the upper right corner. Although angstrom-scale gaps will not
necessarily be achieved at its center due to alignment error, they will occur at other
locations within the array. (b–e) Top-view: TEM images of four representative
Dimers I, II, III and IV with gradually increasing gap-width. The scale bars are 50
nm. (f–i) High resolution magnified-view TEM images of gap regions of dimers of
panels (b–e). In Dimer I, nanoparticles touch; while Dimers II-IV have gap-widths
of 2.0 Å, 6.7Å and 5.8 nm, respectively. The scale bars in (f–i) are 5, 2, 2 and 10
nm, respectively.

Vivek V. Thacker et al. DNA origami-based assembly of gold nanoparticle
dimers
25
, as shown in Fig 1.4. They employ self-assembly based on the DNA
origami technique for accurate positioning of individual gold nanoparticles. This
design leads to strong plasmonic coupling between two 40 nm gold nanoparticles
reproducibly held with gaps of 3.3± 1 nm. The nanogaps obtained from this method
6

depends strongly on the spacer DNA origami. However, the freedom of tuning the
spacer thickness of DNA origami is not arbitrary and limited by the biological
nature of selected DNA origami. In that case, the realization of sub-nanometer
precision on the gap size control cannot be realized. In addition, the patterning of
DNA origami is time consuming with high cost, which hinders its application in
large scale manufacture. Based on self-assembly technology, Longkun Yang et al.
proposed to use the diluted colloid of gold nanoparticles with the surface stabilized
by citrate and obtain individual dimer structures with subnanometer gap widths
formed randomly by capillary forces during solvent evaporation. Several wet
chemical synthesis methods like this were reported to fabricate subnanometer gap
nanostructures. Even though these methods are fast and low cost, the randomness
in both gap size and structure positioning limits their applications in studying the
fundamental science behind, especially the quantum effects that require
deterministic and precise gap size control at subnanometer scale.
7

Figure 1.4 A schematic of the NP dimers assembled on the DNA origami platform.
The NPs are coated with a ssDNA brush to prevent aggregation as well as facilitate
attachment to the origami platform.

1.3 Design of Collapsible Nano-Fingers
Here we present a technology to fabricate large-area gapped plasmonic structures
deterministically with atomic precision, high throughput and high reliability at low
cost. The technology is based on collapsible nano-fingers fabricated using
nanoimprint lithography (NIL), reactive-ion etch (RIE) and atomic-layer deposition
(ALD). Figure 1.5 shows a schematic illustration of such a gap plasmonic
nanostructure. A pair of metallic nanoparticles is placed on top of two nano-fingers
in flexible polymer (i.e. nanoimprint resist) with high aspect ratio. Atomic layer
deposition (ALD) is then used to coat a thin conformal dielectric layer. The ALD-
8

coated dielectric layer serves as the spacer to define the spatial gaps between the
metallic particles. By collapsing the pair of nano-fingers, two metallic nanoparticles
with dielectric coating contact each other. Therefore, the gap size between two
metallic nanoparticles is well defined by twice the thickness of the ALD-coated
dielectric layers. Since the ALD process deposits dielectric film atomic-layer-by-
atomic-layer with high conformity,
26-28
atomic precision on gap thickness control
is achieved. The small separation ensures strong coupling between each pair of
metallic nanoparticles, such that the hot spot with greatly enhanced electromagnetic
near-field is located within the dielectric nano-gap.

Figure 1.5 Schematic of controllable hot spots created using nano-fingers. A pair
of metallic nanoparticles with ALD-coated dielectric layers contact each other
when flexible nano-fingers beneath them collapse. The spatial gap size is defined
by twice the dielectric layer thickness, and the hot spot is at the gap center.

9

Moreover, the tunneling barrier heights for electrons can also be controlled by
using appropriate spacer materials,
29
which strongly affect the plasmonic
enhancement. We have characterized the plasmonic enhancement characterized
with UV-VIS-NIR spectrophotometers by measuring the absorption spectra of
different samples. The evolution of the absorption peak as a function of gap size
for different dielectric gap materials was analyzed. We successfully detected
optimal gap sizes with strongest plasmonic enhancement experimentally. By using
different gap materials, we investigated the effect of tunneling barrier heights. It
was found that as the tunneling barrier height is decreased, the enhancement of
electron tunneling manifests itself in a wider optimal gap. Moreover, we observed
a redshift of the plasmon frequency with increasing gap size when the gaps
narrowed to sub-5 nm range,
21
which indicates the quantum effects at such small
gap scale.

10

Chapter 2. Fabrication of Nano-fingers

2.1 Fabrication of Gold Nano-fingers by EBL mother molds

The experimental procedure for the fabrication of the gold coated nanofingers is
described schematically in Figure 2.1. First, a silicon mold with pillar arrays of the
dimer symmetry in Figure 2.1a was formed by a combination of e-beam lithography
(EBL) and dry Si etching as described previously.
15
Then, as shown in Figure 2.1b,
UV-curable nanoimprint lithography (NIL) was utilized to transfer the Si pillar
pattern to a polymeric reverse tone mold using a custom-designed nanoimprint
machine.
18
The polymeric reverse-tone mold was used in a subsequent NIL step
(Figure 2.1c) to form the final polymer nanofingers shown in Figure 2.1d, with the
procedure similar to that previously reported.
19,20
Finally, Au with nominal
thickness of 50-80 nm was deposited on the sample by e-beam evaporation at
normal incidence to form the Au nanoparticles on the tips of the polymer fingers,
as shown in Figure 2.1e. The capillary force drives the coalescence of the
nanofingers; a similar phenomenon was observed in high aspect ratio microscale
structures.
21-23

11

Figure 2.1 Fabrication procedure for the nanofingers: (a) The fabrication of the
nanofinger silicon mold using e-beam lithography or interference lithography. (b)
The making of the daughter mold using nanoimprinting, (c, d) The fabrication of
the polymer nanofingers from the polymer daughter mold using nanoimprinting.
(e) e-beam deposition of 50-80 nm Au onto the nanofingers.

Figure 2.2 shows the scanning electron microscopy (SEM) images of nano-
fingers using mother molds made by EBL before and after closing. The diameter
and height of each individual gold nanofinger are 100 nm and 650 nm, respectively.
12

The center-to-center distance between two adjacent nano-fingers is 200 nm and the
pitch of the periodic nano-finger dimers is 600 nm. Even

13

Figure 2.2 (a) SEM image of dimer nano-fingers made using EBL mother molds
before collapse. (b) SEM images of corresponding nano-fingers after collapse. The
inset is magnified image of the same nano-fingers from a different viewing angle.
The diameter and height of each finger are 100 nm and 650 nm, respectively. Scale
bars in the SEM images are 1 µ m.
Even though it takes long time and high cost to fabricate EBL mother molds, the
direct writing property of EBL makes it feasible to fabricate more complicated gap
plasmonic structures including nano-finger patterns with different symmetries. Fig.
2.3 shows the SEM images of collapsed nano-fingers in triplets, quadruplets,
quintuplets and heptamers, respectively. All these structures pave the way for
accurate molecular detections, chemical sensors imaging and asymmetric
plasmonic catalytic antenna.
14

Figure 2.3 SEM images of collapsed nano-fingers in triplets, quadruplets,
quintuplets and heptamers, respectively.
However, due to the high cost and time consumption of EBL, the effective area
of nano-finger patterns is limited. Fig.2.4 shows the SEM image of typical EBL
defined nano-finger pattern regions, which is shown as the dark squares in this
image. The region sizes range from tens microns to several millimeters, which pose
obstacles in optical characterization and further integration into electronic devices.
15

Figure 2.4 SEM image of typical EBL defined nano-finger pattern regions.

2.2 Fabrication of Gold Nano-fingers by Interference Lithography

To avoid the limitation from EBL, we can choose mother molds made by
interference lithography. Even though interference lithography can produce only
periodic patterns like 1D gratings and 2D grids or holes array, it supports wafer size
(or larger) fabrication in one-time (1D gratings) or several exposures (2D patterns)
16

that offers much higher time efficiency than EBL. The height of patterns in as-
fabricated mother molds (2D grids) by interference lithography is only 100 nm due
to the limit of depth of focus (DOF) in interference lithography. So further pattern
transfer and deep RIE etching are required to obtain nano-fingers mother molds.
However, tapered sidewalls and limited heights in nano-fingers were obtained after
deep RIE etching (shown in Fig. 2.5), which prevent them from collapsing (the
collapse mechanism will be discussed in detail in the following chapter). This is
mainly caused by the impedance on mass transportation of reactive gases since the
pattern in interference lithography defined molds are hole patterns (reverse tone of
pillars). And the patterns are denser than that in EBL defined molds.

17

Figure 2.5 SEM image of tapered sidewalls and limited heights in nano-fingers
mold obtained from direct deep RIE etching after interference lithography. These
nano-fingers cannot collapse at all.

The renovative fabrication procedure of collapsible gold nano-fingers is
described schematically in Figure 2.6. First, a mold with grid hole arrays of 200 nm
pitch was formed by double interference lithography, as described in previous
publications.
30-32
Then, UV-curable nanoimprint lithography
33-36
(NIL) was utilized
to transfer the original hole pattern to a polymeric reverse-tone (pillar) daughter
mold using a custom-designed nanoimprint machine. UV nanoimprint resist and
lift-off underlayer were spin-coated onto glass substrates (Figure 2.6a). Then
subsequent NIL (Figure 2.6b), residual layer etching, metal evaporation and lift-off
steps were performed to create gold caps array (Figure 2.6c). As the pattern can be
defined precisely by the initial interference lithography and duplicated reliably by
NIL,
30
an array of nano-fingers can be fabricated uniformly over a large area with
high throughput after etching the uncovered UV nanoimprint resist (Figure 2.6d).
In our study, finger arrays over 1.4 inch by 1.4 inch areas were used to guarantee a
good signal-to-noise ratio. The samples were then covered by dielectric coatings of
variable thickness using ALD process (Figure 2.6e). Finally, after soaking into
ethanol and air-drying, the fingers closed together in a dimer configuration as
shown in Figure 2.6f. The capillary force drove the collapse of nano-fingers. A
18

similar mechanism was reported before.
37-41
We found that proper aspect ratio and
straight sidewalls are necessary for uniform dimer-like finger collapsing. Low
aspect ratio or tapered sidewall will prevent the fingers from collapsing, while
fingers with aspect ratio too high will result in a random collapse with large number
of fingers. Once the nanoparticles touch, they will not separate due to van der Waals
forces.
37
Figure 2.7 shows the scanning electron microscopy (SEM) images of
nano-fingers before and after closing. The diameter and height of each finger is 60
nm and 350 nm, respectively. Figure 2.8a shows the transmission microscopy
(TEM) image at the nano-gap after nano-fingers with 2 nm TiO2 ALD film coating
collapsed. In the picture, the gap size is defined as the distance between two closest
points of two gold nanoparticles, which is shown as twice the thickness of the ALD
film. Corresponding Energy-dispersive X-ray spectroscopy (EDS) mappings of Au,
Ti and O are shown in Figure 2.8b, 2.8c and 2.8d. The gap exists in Au EDS
mapping while no gap exists in Ti and O EDS mappings. That indicates the Au
nanoparticles are uniformly covered by continuous and conformal TiO2 film
deposition and the gap size is well defined by twice the thickness of the TiO2 film.

19

Figure 2.6 Fabrication procedure of nano-fingers: (a) Spin coating and UV
curing of 600 nm thick UV NIL resist, then spin coating100 nm lift-off underlayer
and 100 nm UV nanoimprint resist layers. (b) Nanoimprinting using daughter mold.
(c)After etching residual layer, metal evaporation and lift-off, gold caps array was
left on the thick UV nanoimprint resist. (d) Etching the underlying UV nanoimprint
resist without Au covering. (e) ALD deposition of dielectric films. (f) Soaking and
air-drying of the ethanol to induce the collapse of nano-fingers.
20

Figure 2.7 (a) SEM image of nano-fingers before collapse. (b) SEM images of
nano-fingers after collapse. The inset is magnified image of the same nano-fingers
from a different viewing angle. The diameter and height of each finger are 60 nm
and 350 nm, respectively. Scale bars in the SEM images are 200 nm.
21

Figure 2.8 (a) TEM image of the dielectric nano-gap in the nano-fingers with 2
nm TiO2 coating. (b) EDS mapping of Au in the TEM image. (c) EDS mapping of
Ti in the TEM image. (d) EDS mapping of O in the TEM image. Scale bars in the
images are 10 nm.

2.3 Nano-finger Geometry Optimization for Successful Collapse
Most importantly, the aspect ratio of nano-fingers dominates the plausible
randomness of the collapsing process. Low aspect ratio, as mentioned before in Fig.
2.5 and Fig. 2.9, will prevent the fingers from collapsing, while fingers with aspect
ratio too high will result in a random collapse with large number of fingers
(clusters). Typical SEM images of the collapse pattern of very high aspect ratio
22

nano-fingers are shown in Fig. 2.10. Based on the experiment results, an aspect
ratio ranging from 6 to 7 will perform optimally in the collapsing process.

Figure 2.9 SEM images of the low aspect ratio nano-fingers after soaking and
drying process. These nano-fingers cannot collapse as their aspect ratios are too
low, which prevent the collapsing process mechanically.

23

Figure 2.10 SEM images of the collapse pattern of very high aspect ratio nano-
fingers. Left figure is top view and right figure is side view.
On the other hand, the straight sidewalls of nano-fingers are preferred to facilitate
the collapsing process. As shown in Fig. 2.5, tapered sidewall will prevent the
fingers from collapsing. To obtain the straight sidewall, RIE recipes on imprint
resist etching should be optimized in the fabrication process. In RIE process, the
carbon-fluorine (C/F) ratio plays the most important role in determining the etching
24

profiles. High C/F ratio will lead to tapered sidewalls, while low C/F ratio results
in over-etch and undercuts. A precise combination of 20 sccm C4F8 and 25 sccm
SF6 etching gas was used to produce straight sidewalls in nano-fingers. In addition,
low gas pressure and high DC bias are preferred, because the RIE process is more
likely to etch the imprint resist physically and directionally rather than etch
chemically and isotropically in this case. However, the voltage should not be too
high as the increase DC power will cause mask erosion. Therefore, 10 mTorr gas
pressure and 30 W DC power were used in this process.

2.4 Capillary Force Induced Collapsing Mechanism
The as-fabricated nano-fingers array will be soaked into ethanol and dry in the
air. According to previous literature reports, the capillary meniscus interaction
force is mainly responsible for the pillar collapse rather than the often-cited Laplace
pressure difference due to isolated capillary bridges. At the end of liquid
evaporation process, the ethanol meniscus between adjacent nano-fingers still holds
and forms capillary meniscus, while the ethanol elsewhere has already evaporated
(shown in Fig. 2.11). Theoretical calculation given by Dinesh Chandra et al. of the
torque 𝜏 on an individual nano-finger suggests that,
𝜏 =
𝜋𝛾 𝑅 2
ℎ cos
2
𝜃 (
𝑥 2
⁄ )
2
− 𝑅 2

25

where 𝛾 is the surface tension of ethanol, 𝑅 is the radius of the nano-finger, ℎ is
the height of the nano-finger, 𝑥 is the center-to-center distance between two
adjacent nano-fingers, and 𝜃 is the contact angle between ethanol and nano-fingers.

Figure 2.11 Schematic of capillary force induced nano-finger collapsing
mechanism. Once the nanoparticles touch, they will not separate due to van der
Waals forces.
Realizing the mechanism behind the collapsing mechanism, manipulation of
result collapse pattern becomes possible. One direct method is the regulation of
spacing between individual nano-fingers. As shown in EBL defined collapsed
nano-finger pattern, the desired number (e.g. dimer, triplet, quadruplet, etc.) of
nano-fingers are placed in close vicinity as a “unit cell”, while the spacing between
unit cells is much larger than that between nano-fingers within unit cell. Therefore,
ethanol will be mostly left within the nano-finger unit cell at the end of evaporation
process and exert the torque to bend the nano-fingers within a unit cell to contact
each other (shown in Fig. 2.11).
26

However, the spacing between all nano-fingers is almost even due to the
limitation of periodicity in the pattern defined by interference lithography. Hence
geometric shapes of nanoparticles and pillars in nano-fingers should be modified to
generate desired collapse pattern. Based on the suggestion of collapsing mechanism,
the symmetric nano-finger array will collapse isotropically. That is, quadruplet unit
cell will be the dominant pattern, as shown in Fig. 2.12.

Figure 2.12 SEM image of symmetric nano-fingers with even spacing collapsing
in quadruplet pattern, which are defined by interference lithography.

27

Asymmetry is beneficial for collapse in nano-finger dimer pattern. The gold
nanoparticles on the tip of nano-fingers are symmetric shapes like squares or circles
if we follow the fabrication procedures described before. And they will shape the
nano-fingers in the same way as they also serve as the deep RIE etching mask. If
we modify the shape of these nanoparticles to be rectangle or ellipse, the capillary
force exerted on the long side of nano-fingers will exceed that on the short side.
Therefore, the nano-fingers will collapse anisotropically and be prone to form the
dimer-like patterns. The fabrication results before and after collapse are shown in
Fig. 2.13.
28

Figure 2.13 (a) SEM image of asymmetric nano-fingers with even spacing defined
by interference lithography before collapse. (b) SEM image of asymmetric nano-
fingers collapsing in dimer pattern. The inset is the zoon in image from another
viewing angle.

To fabricate such asymmetric patterns, angled metal evaporation should be
adopted. After nanoimprint process using 2D grid molds made by interference
29

lithography, periodic square holes pattern with even spacing are created. Then, 50
nm Gold on top of 2 nm Titanium (adhesive layer) was deposited onto the sample
by e-beam evaporation (Temescal BJD-1800 E-Beam Evaporator) at 10° off-
normal incidence. The schematic of this angled metal evaporation is shown in Fig.
2.14. Due to the shading effect, the gold masks will have reduced length on one
side, which makes them essentially rectangular shape.

Figure 2.14 Schematic of angled metal evaporation and asymmetric metallic
etching mask formation due to shading effect.

30

From Fig. 2.15a, we can observe that after a one-time soak-and-dry collapsing
process, the nano-fingers do not collapse thoroughly all over the sample area. This
phenomenon happens in both EBL defined nano-finger arrays and interference
lithography defined nano-finger arrays. Since entropy always wins, either the
fluctuation of ethanol evaporation and hence the liquid filling between nano-fingers
or the fluctuation of capillary force from geometric parameters will result in the
failure of collapse. To avoid the randomness brought by the fluctuation, we
repeated soak-and-dry process multiple times. Since the van der Waals forces
between touched nano-finger tips is much larger than the elastic restoring force of
bending nano-fingers, they will not separate once they touched. That is, the
collapsed nano-fingers exhibit excellent stability. In this case, we realized reliable
collapse and ensured collapse fidelity over large sample area. The result
comparison of conducting one-time collapsing process and five repeated collapsing
process is shown in Fig. 2.15b. Nevertheless, we should still be aware that the
collapse mismatch cannot be avoided in the interference lithography defined nano-
fingers given that even spacing between each nano-fingers. That means an
individual nano-finger can bend to either direction in an equal possibility. Therefore
the dimer arrays may not aligned in large area, even though the density of hotspots
will not change.
31

Figure 2.15 SEM images on comparison of conducting one-time collapsing process
and five repeated collapsing process. Some nano-fingers do not collapse after one-
time soak and dry process, but will finally collapse after multiple repeated
collapsing process.

32

2.5 Atomic Layer Deposition for the Dielectric Spacer
We used blank silicon wafers to deposit materials in order to optimize our ALD
recipes before we deposited on real samples. Ellipsometry was exploited to
accurately characterize ALD film’s thickness and refractive index, indicating film’s
quality and uniformity.
Table 2.1 is the summary of the recipes we adopted for the switching layer ALD
deposition. We deposited these films at USC labs and Lawrence Berkeley National
Labs. Several parameters are critical in this procedure. First of all, the precursors
(reactants) should be determined, usually metallic organic composite as metal
source and water or oxygen plasma as oxygen source. The logic behind it is similar
to CVD, considering the absorbability to substrate. The most common radical is
dimethylamido, and the corresponding metallic organic composites are preferred in
ALD system. They will provide good reactivity, uniform film formation one layer
by one layer and least damage to ALD system (another common selection is
metallic chloride, which is harmful to ALD system.
33

Table 2.1 The summary of recipes we developed for the dielectric spacer ALD
deposition onto nano-fingers.

After we decided the precursors, we optimized the growth temperature in order
to locate it at the ALD growth window, in which uniform deposition could be
realized with reasonable rate. Various growth temperature should be tested and
corresponding precursor pulse and wait time should be modified. (Lower growth
temperature requires longer pulse and wait time for precursors to uniformly adsorb
and react). For example, aluminum oxide (TMA and water) window is 50 to 250 ° C.
We tried the recipes from 80 ° C (0.1s water, wait 15s, 0.1s TMA, wait 15s, this is
one cycle) to 200 ° C (0.015s water, wait 3s, 0.015s TMA, wait 3s).
The key factor that determines the deposition temperature is that the nano-fingers
are not robust at high temperature. The SEM images of the nano-fingers before and
34

after TiO2 ALD deposition at 200 ° C are shown in Fig. 2.16. Obvious deformation
was observed at the top part of nano-fingers, which further prevent them from
collapsing. ALD deposits dielectric film with higher quality at higher temperature
as it provides enough energy for the material to diffuse and distribute uniformly on
the substrates. In addition, post-annealing process also contributes to the ALD film
quality. As a compromise, we adopted oxygen plasma instead of water vapor as the
oxygen source for ALD dielectric deposition. Since oxygen plasma provides
additional energy to the deposition process, we can use low temperature (80 to
100 ° C) for ALD deposition and obtain good film quality at the mean time. In this
case, the ellipsometry characterization on the monitor sample indicates film quality
as good as that obtained at 200 ° C using water vapor as oxygen source, while the
nano-fingers demonstrate reliable collapse after the standard soak-and-dry process.

35

Figure 2.16 SEM images of the nano-fingers (a) before and (b) after TiO2 ALD
deposition at 200 ° C. Deformation emerges at the top part of nano-fingers after
ALD deposition at high temperature.

2.6 Crystallinity Characterization of ALD Film
ALD is a chemical vapor deposition technique based on successive, separated,
and self-terminating gas-solid reactions of typically two gaseous reactants. Like
other thin film deposition techniques, the ALD temperature affects the crystallinity
of deposited material. The deposited material undergoes transition from amorphous
to polycrystalline at characteristic temperature. Al2O3 can be deposited by ALD
using water (H2O) and trimethylaluminum (TMA) as precursors. To compare the
crystallinity of Al2O3 films at different deposition temperatures, 200 cycles Al2O3
(around 16 nm) were deposited on Si substrate at four different temperatures (80
° C, 120 ° C, 160 ° C, and 200 ° C), separately.
The refractive index of these four Al2O3 films were measured by ellipsometer.
With the increase of deposition temperature, the Al2O3 refractive index is increased,
as shown in Table 2.2. The data indicates that the Al2O3 deposited at higher
temperature is denser. The corresponding X-ray diffraction (XRD) spectra of Al2O3
films at these four different deposition temperatures are shown in Figure 2.17a The
Al2O3 films deposited at 80 °C and 120 °C don’t have characteristic peaks in the
XRD spectra, while the Al2O3 films deposited at 160 ° C and 200 ° C have clear
36

characteristic peaks in the XRD spectra. These indicate that the Al2O3 films
deposited at 80 ° C and 120 ° C are amorphous, while the Al2O3 films deposited at
160 ° C and 200 ° C are polycrystalline. According to the Scherrer equation 𝜏 =
𝐾𝜆
𝛽𝑐𝑜𝑠𝜃 (where 𝜏 is the average grain size, 𝐾 is a dimensionless shape factor, 𝜆 is the
X-ray wavelength, 𝛽 is the line broadening at half the maximum intensity, and 𝜃 is
the Bragg angle), the estimated average grain size of polycrystalline Al 2O3 can be
calculated, as shown in Table 2.2. The average grain size of polycrystalline Al2O3
is also increased with the increase of deposition temperature (Table 2.2), which is
consistent with the change of refractive index.
Figure 2.17b, c, d and e show the TEM images of Al2O3 films deposited at 80
° C and 200 ° C, and the corresponding FFT patterns, respectively. These figures
clearly indicate that the Al2O3 film deposited at 80 ° C is amorphous, while the
Al2O3 film deposited at 200 ° C is polycrystalline. It is the direct evidence to
demonstrate that the deposited Al2O3 film has transition from amorphous to
polycrystalline with the increase of deposition temperature. Above all, the
crystallinity of deposited Al2O3 can be controlled through the ALD temperature.
37

Figure 2.17 Controlling the crystallinity of Al2O3 by changing ALD temperature.
(a) XRD spectra of Al2O3 films at different deposition temperatures (80 ° C, 120 ° C,
160 ° C, and 200 ° C). (b) TEM image of 80 ℃ deposited Al2O3 film, which is
amorphous. The SiO2 layer is the native oxide layer on top of the Si substrate. (c)
38

TEM image of 200 ℃ deposited Al2O3 film, which is polycrystalline. The SiO2
layer is the native oxide layer on top of the Si substrate. (d) Corresponding FFT
pattern of TEM image in (b) (Frequency range DC ~ 1/0.020 [1/nm]). (e)
Corresponding FFT pattern of TEM image in (c) (Frequency range DC ~ 1/0.020
[1/nm]).

Al2O3
Deposition
Temperature
Refractive
Index
(𝝀 = 𝟓𝟑𝟐𝒏𝒎 )
Crystallinity
Average Grain
Size
𝟖𝟎 ℃ 1.630 Amorphous N/A
𝟏𝟐𝟎 ℃ 1.647 Amorphous N/A
𝟏𝟔𝟎 ℃ 1.652 Polycrystalline ~ 150.5nm
𝟐𝟎𝟎 ℃ 1.656 Polycrystalline ~ 207.0nm

Table 2.2 Refractive index, crystallinity and average grain size of Al2O3
deposited at different temperatures.

39

Chapter 3. Investigation of Quantum Effects in Collapsible
Nano-Fingers

3.1 Absorption Spectrum Characterization
Surface plasmons are excited in these collapsed nano-fingers upon light
incidence. By coupling with incident photons, one observes strong enhancement
while reradiating the energy.
42-44
The gap size plays a critical role in tuning this
plasmonic enhancement.
45-49
Based on classical electromagnetic theory, stronger
near-field enhancement
50
and a diverging red-shift
51
of bonding plasmon resonance
emerges as two nanoparticles approach each other. However, extensive theoretical
studies have shown that for these very narrow gaps at sub-5 nm, several factors can
significantly reduce the field enhancement.
14-16, 52-54
Quantum mechanical effects
become relevant in this regime. They are basically (i) electrons tunneling from one
nanostructure to another through the gap, and (ii) the finite spatial profile of the
plasmon-induced screening charge (nonlocality). To properly take these effects into
consideration, a fully quantum mechanical treatment based on linear quantum
calculations predicts that shrinking the gap size will result in stronger tunneling and
therefore limited field enhancement for ultra-small gaps.
55-58
In the dimer-like gap
plasmonic structures, bonding dimer plasmon (BDP) originating from the
hybridization of the dipolar plasmon modes of individual nanoparticles and charge
40

transfer plasmon (CTP) referring to the electron tunneling between nanoparticles,
are the two dominant competing plasmon modes.
14, 17
Based on these modes,
optimal gap sizes which can provide the strongest enhancement have been predicted
theoretically.
14-18

In order to study the plasmonic enhancement of the closed nano-fingers, we chose
the absorption spectrum as an indicator, which is convenient and straightforward.
Figure 3.1 shows the absorption spectra of the collapsed nano-fingers with TiO2 as
the gap material. The gap sizes were chosen as 2, 3, 4, 5, 6 nm, respectively. As the
gap size is shrunk from 6 nm to 4 nm, the peak absorption strength at resonance
frequency was observed to increase. This is consistent with the classical
electromagnetic prediction. However, when the gap size was reduced from 4 nm to
2 nm, the peak absorption strength at resonance frequency weakened. This indicates
that plasmonic quantum effects deviate the spectral response of the surface plasmon
modes from the classical model for TiO2 gap at this scale. It is clear that the 4 nm
TiO2 gap system has the strongest plasmon light absorption. This means the most
energy is stored in the surface plasmon, and the strongest near-field enhancement
is realized by reradiating the energy. We successfully observed and determined the
optimal gap size systematically in experiment. In addition, the absorption peaks
exhibit a red shift as the gap size is increased, which is also opposite to classical
prediction but agrees with the quantum mechanical calculation. For comparison,
41

perpendicular polarized light was illuminated on samples in the same setup. Figure
3.2 shows a comparison result for absorption spectra of nano-fingers with 4 nm
TiO2 gap. No absorption peak is observed if the polarization was perpendicular.
Based on previous work reported,
21
parallel polarized light will primarily excite the
longitudinal bonding dipole plasmon (LBDP) mode, which is generally called BDP
in the literature. The absorption peaks of the previous results mainly originate from
the combination of LBDP and CTP modes. However, upon the perpendicularly
polarized light’s illumination, the transverse dipole plasmon (TDP) mode is the
dominant mode excited, which originates from the uncoupled dipole in individual
nanoparticles. In that case, the nano-finger pairs behave more like isolated
monomers rather than dimers, since TDP is perpendicularly polarized with the
resonant position and weakly dependent on the gap size. As a result, the absorption
strength from TDP is too weak to be observed after the compensation of
background signal.
42

Figure 3.1 Absorption spectra of collapsed nano-fingers with TiO2 as gap
material. The gap sizes are 2, 3, 4, 5, 6 nm, respectively.

43

Figure 3.2 Absorption spectra of collapsed nano-fingers by using parallel and
perpendicular polarized incident light.

3.2 Effects of Different Tunneling Barrier Heights
In order to investigate the dependence of plasmonic enhancement on the electron
tunneling effect, we used different ALD dielectric materials as the gap material. As
the electrons tunnel from one gold nanoparticle to another, the tunneling barrier
height, based on energy band diagrams, is defined as the difference between the
Fermi energy of gold (5.1 eV) and the electron affinity (EA) of corresponding
dielectric materials.
29
We chose TiO2, WO3, and SiO2, with electron affinities equal
to 4.21 eV, 3.45 eV, and 0.90 eV, respectively,
59-61
so that the corresponding
44

tunneling barrier heights are 0.89 eV, 1.65 eV, and 4.20 eV. With reduced tunneling
barrier height, the tunneling strength is enhanced, which is manifested by a wider
optimal gap. This is why the observed optimal gap size for TiO2 gap is much larger
than the optimal vacuum gap size theoretically predicted in most of the literature,
considering that vacuum holds a 5.1 eV tunneling barrier height. In order to analyze
the underlying physics of observed enhancements and redshifts, we plotted the peak
absorption strength and the resonance frequency versus the wavelength of incident
light for different material, as shown in Figure 3.3. From the Figure 3.3a, we
identified the optimal gap sizes with strongest plasmonic absorption for different
gap materials. We also investigated the red shift of resonance frequency with
increasing gap size in Figure 3.3b. The uncertainties of the peak absorption strength
and peak location are caused by the differences between individual samples,
including defect level and geometric variance. Triple standard deviation was
applied as the error for a confidence interval of 99.7%. They are all within
permissible error range.

45

Figure 3.3 (a) The measured peak absorption for different values of gap size from
three different gap materials. The data is fit according to Equation 1. (b) The
measured red shift of absorption peaks with increasing gap size from three different
gap materials. The data is fit according to Equation 2.

3.3 Theoretical Analysis on Field Enhancement in Nano-Fingers
In order to understand and correlate the experimental observations with quantum
based model calculations incorporating electron tunneling and nonlocality, we need
a model which can ultimately guide the effective design of potential applications.
There have been extensive studies of the optical properties of spherical
nanoparticles separated by relatively large distances.
62, 63
However, these
calculations become more complicated when the nanoparticles are close to each
other, such that light scattered from them exhibits phase coherence.
64
Based on the
observation that coherence is responsible for the electro-magnetic hotspot
65

46

observed between nanoparticles in this experiment we propose the simple,
phenomenological model discussed below to describe the electric field
enhancement at the center point between two nanoparticles for different separation
distances (D). This model also provides an argument for the observed redshift of
the plasmon frequency with increasing separation. It is known that the dominant
peak from the electron energy loss spectrum (EELS) of a pair of metallic
nanoparticles separated by a small distance is due to the lowest order longitudinal
dipole mode
66
referred to as the bonding dimer plasmon (BDP).
67
Classical
calculations show the onset of another strong plasmon mode (known as the charge
transfer plasmon, CTP) when the nanoparticles touch.
51, 68
Quantum calculation
show a weakening of the BDP mode with a strengthening of CTP mode at finite
separation distances (i.e. D > 0).
14
This is due to quantum mechanical tunneling
69

allowing charge transfer even without a direct conduction path.

We believe the observations discussed in this work can primarily be attributed to
three physical processes. Firstly, the field enhancement seen midway between
nanoparticles can be attributed to the strong Coulomb field caused by separated
charges on both nanoparticles. The strength of this field depends on the amplitude
of the incident field, the dielectric constant of the coating material and various other
properties of the system. We call this parameter 𝑞 𝑒𝑓𝑓 to stress that it originates from
opposite charges coming close to each other due to the longitudinal dipole
47

oscillations on each nanoparticle. Secondly, quantum mechanical tunneling allows
charge transfer between the particles even with a finite D, decreasing the charge
separation and consequently also the strength of the field enhancement. In the
simplest example, that of 2 quantum wells separated by a barrier of width D, it is a
well-known result that the probability of tunneling goes as 𝑒 −𝜆𝐷
where λ is the
decay length of the wave function which depends on the barrier height and the
particle energy. Applying this simplified approach here we let 𝑞 𝑒𝑓𝑓 → 𝑞 𝑒𝑓𝑓 (1 −
𝑒 −𝜆𝐷
) to account for the loss of charge separation and the consequent weakening
of the bonding dimer plasmon due to tunneling. Finally, we have to account for
energy loss in the system which occurs through many channels such as absorption
in the dielectric coating and decay of the surface plasmon. For simplicity we
modeled this as a simple damping term that depends on the separation between the
particles. This is justified since the further the separation, the more coating each
particle has, which in turn increases absorption. Thus we arrived at the following
expression to describe the most important processes affecting the strength of the
field enhancement (F.E. in arbitrary units) seen between nanoparticles
𝐹 . 𝐸 . ~ [𝐸 0
+ 𝑞 𝑒𝑓𝑓 (1 − 𝑒 −𝜆𝐷
)]𝑒 −𝛼𝐷
(1)
where 𝐸 0
is a parameter related to the strength of the incident/driving field and 𝛼 is
the damping coefficient.

48

We presented the results in arbitrary units to stress that this expression is meant
solely as a phenomenological description of the relevant physics. Since the
geometry and physical dimensions of the experimental system greatly affects the
parameters used, we did not attempt to derive the values of these parameters from
first principles but rather fitted them according to the experimental results in order
to see if a qualitative analysis of the fit matches our intuition of the parameters. The
experimental data from three different nanoparticle composites (TiO2, WO3 and
SiO2) were fit according to this model and the results are shown in Figure. 3.3a.
The plot fits show the different physical processes going on. For example, consider
the plot for WO3; at D = 1nm tunneling dominates, which reduces the measured
field enhancement compared to D = 3nm where tunneling is less probable. The
same can be seen for TiO2. At large separation tunneling no longer has a big effect,
but energy loss to the surrounding medium lowers the peak absorption. We thus
identified an optimal separation distance for maximum field strength between the
particles. In the case of SiO2 the tunneling barrier is so high that we did not see an
optimal separation. The fit for TiO2 is not as exact as the others but it still
qualitatively agrees with the experiment showing an optimal separation between 3
and 4nm. Indeed, the fits show a proper ordering of the tunneling barrier height for
the three materials.

49

3.4 Theoretical Analysis on Resonant Frequency in Nano-Fingers
It has been shown that the plasmon frequency of gold nanoparticles redshifts with
an increase in particle size.
62
The plasmon frequency for metal nanoparticles coated
with a dielectric also exhibit redshifts with increased coating thickness.
70-71
The
redshift is believed to be due to polarization of the surrounding medium which
decreases the surface charge on the nanoparticle, which in turn reduces the restoring
force of the plasmon oscillations.
72
Our observed redshift of the single nanoparticle
plasmon mode due to the proximity of the other nanoparticle is also consistent with
this explanation. The further apart the nanoparticles are from each other the more
dielectric material separates them i.e. the more screening both nanoparticles feel.
Consequently, we expect the plasmon frequency to be dependent on the inverse of
the dielectric constant (ϵ) of the coating material. Furthermore, redshift is also
caused by ‘phase retardation’ which works as follows: the separation between the
nanoparticles leads to a phase shift of the electromagnetic wave at the neighboring
nanoparticle compared to the plasmon which emitted the wave. The charge
oscillation on each nanoparticle will consequently be ahead in phase from the
electromagnetic wave coming from the other nanoparticle leading to a decrease in
phase velocity and therefore a redshift in the plasmon frequency. This redshift is
expected to be linearly dependent on the time it takes a signal to travel from one
particle to the other which is proportional to 𝑛𝐷 where 𝑛 is the index of refraction
of the dielectric coating material (recall the speed of electromagnetic radiation in a
50

medium is 𝑣 = 𝑐 /𝑛 ). The index of refraction of the dielectric coating material were
obtained by measuring control samples with ellipsometry after ALD deposition.
The values are 2.59, 2.20, 1.45 for TiO2, WO3, and SiO2, respectively. We therefore
expect the redshift to be characterized by the following relationship:
𝜆 𝑚𝑎𝑥
= 𝐴 𝑛 𝜖 𝐷 + 𝜆 0

where A is some proportionality constant that depends on the geometry of the
system and not on the material used for the coating. Finally, recall that for non-
magnetic materials 𝜖 = 𝑛 2
, thus
𝜆 𝑚𝑎𝑥
= 𝐴 𝐷 𝑛 + 𝜆 0
(2)
The measured red shift is shown in Figure 3.3b as a function of 𝐷 /𝑛 . We see that
the geometric parameter (slope of lines of best fit) matches closely for all three
systems as expected.

We should mention that several experiments have found blue shift for increasing
particle separation but at larger separations than shown here.
21
This stands to reason
since the coupling between nanoparticles will decay with increased distance and
therefore we should expect the rate of redshift to decrease at some point and
eventually turn to blue shift in order to restore the single nanoparticle plasmon
frequency (λ0). We saw a linear redshift up to the maximum separation considered
in the experiment for all three materials but judging by the rate of decrease in
51

measured field strength between the particles we would not expect the redshift to
continue according to the above model for much more than 10nm separation.

52

Chapter 4. Application of Nano-Fingers in Plasmon
Enhanced SERS

4.1 Introduction to Plasmon Enhanced SERS
Surface-enhanced Raman scattering (SERS), providing fingerprint identification
and potential single molecular applications in life science, is one of the most
desirable analytical technique
1
. The SERS mechanism mainly stems from the
electromagnetic (EM) enhancement under surface plasmons resonant excitation in
metallic nanostructures, while the chemical effect could also play a role
2-4
. It has
been widely accepted that the greatest degree of EM field enhancement can be
obtained within a sub-nanometer gap between metal nanoparticles
5-14
, in which the
ultimate gap limit in vacuum is roughly 0.5 nm, before charge recombination occurs
through the quantum mechanical electron tunneling
15-22
. Single-molecular SERS
was first demonstrated using aggregates of Ag nanoparticles to create nanometer
junctions with maximum enhancement of up to 10
12
-fold
23, 24
, but this method was
usually not stable due to molecules jumping, chemical instability on the metallic
surface, and reproducibility in fabrication over a large area
25
. In recent years, many
attempts have been made to fabricate ordered gap plasmonic nanostructures with
sub-5 nm gaps
26-30
, but limitations of traditional techniques, including electron-
beam lithography and physical mold etching, have shown reproducibility obstacles
53

in producing highly uniform nanogaps
31-35
. For applications in single-molecular
SERS, one of important methods is to create the uniform gap plasmonic
nanostructures with sub-nanometer gaps. As the ideal platforms, they should also
have the features of high reproducibility, chemical stability, biological
compatibility, and ease of fabrication at the manufacturing scale
25
. In addition,
scientists have recently started to apply dielectric coated metallic particles as a way
to improve their chemical stability and biocompatibility
36-39
. Xu et al have reported
in a theoretical study
25
that SERS enhancement can reach up to 10
14
-fold inside the
dielectric cavity, between physically touched SiO2-coated Ag nanospheres, that
allows for single-molecular SERS detection. Their study also predicted a further
one to two order SERS enhancement increase if a larger refractive index dielectric
is used
25
. However, in most cases it is impractical to make analyte molecules to be
embedded just inside the dielectric cavity. Thus the knowledge in making the
outwards spillage of the coupled EM field from the dielectric cavity is essential,
especially in the case of the touched gap plasmonic nanostructures.

4.2 Fabrication of Nano-Fingers for Plasmon Enhanced SERS
For application of nano-fingers in plasmon enhanced Raman spectrum (SERS),
especially in chemical and biological detection, high resolution and accuracy
instead of large effective area are the first priority during the fabrication of
plasmonic nanostructures. In this case, electron beam lithography (EBL) made
54

mother molds are preferred rather than double interference lithography made
mother molds for fabricating nano-fingers substrate. EBL mother molds enable
precise design and control on the size and spacing of each individual nano-fingers,
which facilitates accurate hot spots array distribution and faithful collapse rate.

In our method, Ag nanoparticles are placed on the top of nanofingers instead of
Au for better SERS performances, shown schematically in Fig. 4.1a. If there is no
dielectric films deposition, then the touched pair of nanofingers (as shown in Fig.
4.1b) will induce strong energy dissipation and lead to dramatic decrease of
localized EM field
40
. However, we can deposit an ultrathin (1 nm) ta-C film (as
shown in Fig. 4.1c) on the outside surface of the nanofingers by FCVA (filtered
cathodic vacuum arc) in which the samples are tilted and spanned for better
coverage on the side wall of each finger, as shown schematically in Fig. 4.1d. As a
result, the physically touched gap plasmonic nanostructures with highly uniform
gaps can be obtained after a collapsing process, in which 2-nm-gaps are defined by
twice the ta-C film’s thickness, as shown schematically in Fig. 4.1e. The ta-C film
is especially critical to have the following properties. Firstly, the 1 nm ta-C film is
pinhole-free that has been proved by a previous oxalic acid corrosion test
41, 42
, and
also highly uniform with atomic precision in thickness of ~ 1 Å due to the control
of the Faraday cup during the deposition process
41
. In the work above, we
systematically investigated the effects of gap size and tunneling barrier height to
55

understand the local field enhancement mechanism of gap plasmonic
nanostructures and found that the optimal gap size for field enhancement strongly
depends on the selected dielectric materials embedded in gap (here, we chose TiO2,
SiO2, and WO3) and their tunneling barrier heights. This is due to the quantum
effect allowing charge transfer even without a direct conduction path, which is
different from the classical electromagnetic prediction. In this case, ta-C has a low
electron affinity (EA) with 1.5 eV and the Fermi level of Ag is 4.26 eV
44, 45
. The
barrier height, defined by the difference between them, is 2.76 eV
46
, and thus the
classical bonding dipole plasmon mode is dominant in our dimer-like gap
plasmonic structures composed of 2-nm-ta-C gaps
43
. When the thickness of ta-C is
below 1 nm, it is not pinhole-free, so we selected the 1 nm ta-C as the limitation to
protect the Ag
41
. Secondly, ta-C film is a high-κ dielectric material
47,48
. Its
refractive index of 1 nm-thickness ta-C film is measured to be ~2.44, which is larger
than that of ultrathin Al2O3 and SiO2 films
42
. For an isolated nanofinger, the electric
field at resonance would be spatially re-distributed across the boundary from the
inside to outside of the ta-C and abrupt enhancement in the interface of ta-C/air
rises due to the existence of high-κ dielectric ta-C coating, as shown schematically
in Fig.4.1f. After the collapsible process, in a ta-C coated gap plasmonic
nanostructure, the coupled EM field will be spilled outwards from the top of fingers
rather than being localized within the dielectric gap, shown schematically in
Fig.4.1g. Finally, the composition of ta-C film contains only carbon atoms, and
56

over 90% sites in ta-C is sp
3
(diamond structure) hybridization, which means that
the ta-C film has excellent mechanical properties and biocompatibility
47
. Especially
important, the non-bonded carbon (red bonds, shown in Fig.4.1c) at the outer
surface of ta-C film can easily attract bioactive molecules, which is useful to SERS
detection. Owing to the relatively small size of carbon atoms and their short inter-
atomic distance, ta-C is much superior to other dielectric materials, such as, Al2O3,
TiO2 and SiO2
47
. When ta-C coated gap plasmonic nanofingers are used in SERS,
the small analyte molecules are easily trapped into the location of the spilling-out
EM field as shown in Fig.4.1h, and the large analytes such as alive cells, can also
be situated on the top surface of fingers that can be analogous to SHINERS (shell-
isolated nanoparticle-enhanced Raman spectroscopy) technology as shown in
Fig.4.1i
36
. In addition, the slightly saddle-shape of the collapsible fingers can
prevent molecules from jumping and traps them on the bottom of the saddle. Due
to the 10
12
-fold enhancement from the spilling-out EM field, single-molecular
SERS can be easily realized via our proposed nanostructure.
57

Figure 4.1 Schematic of the ultrathin ta-C coated gap plasmonic Ag-fingers
for SERS. (a) & (b) Metal polymer fingers collapse when they are exposed to
ethanol and dried in air. (c) Schematic atomic structure of ta-C film. (d) & (e) ta-C
coated gap plasmonic Ag-nanofingers are formed by the collapsible process and the
gap size is defined by twice the ta-C thickness. (f) & (g) Schematic of EM field for
an isolated nanofinger and a dimer nanofingers. (h) & (i) ta-C coated gap plasmonic
Ag-nanofingers application in SERS for micro-molecules and living cells on
macroscale sizes.

58

4.3 Characterizations of Gap-Plasmonic Nanofingers with Uniform
2 nm Gaps
Fig.4.2a shows the scanning electron microscopy (SEM) image of nanofingers with
dimer symmetry before collapsing process. The inset image in Fig.4.2a is the
transmission electron microscope (TEM) cross-sectional view of single nanofinger.
The results show that the diameter of nanofinger and the periodicity are 70 nm and
500 nm, respectively, which is consistent with the design. In order to demonstrate
the coverage of the 1 nm ta-C on the Ag tip of nanofingers, the high resolution-
TEM (HR-TEM) cross-sectional analysis for single ta-C coated Ag nanofinger was
performed, shown in Fig.4.2b, and the result indicated that the Ag tips were
uniformly covered by continuous and conformal 1-nm ta-C films. It is noted that
for Fig.4.2b, the C and Cr coating layers are firstly pre-coated to protect the
interface of the samples and provide better cross-sectional view. After the
collapsing process, the nanofingers will form a completely touched dimer
nanostructure, shown in SEM image of Fig.4.2c. If no dielectric coating, metal
nanofingers will be touched each other, as shown in the TEM cross-sectional image
of Fig.4.2d, and thus there is a direct conduct path that decreases the localized EM
field
15,16
. In our proposed gap plasmonic nanofingers, the nanogap with twice the
ta-C thickness will be formed when the ta-C coated nanofingers collapse, as shown
in Fig.4.2e. Corresponding energy-dispersive X-ray spectroscopy (EDS) mapping
of Ag also confirms the existence of 2 nm nanogap, as shown in Fig.4.2f. Since the
59

precise gap size is precisely controlled by the thickness of ta-C film, the critical
dimension of the nanofingers is not critical for reliable formation of the gaps
40
. In
order to further study the nanogap obtained by twice the thickness of ta-C films,
the electron energy loss spectrum (EELS) were performed, as shown in Fig.4.2g-i.
We note that EDS is not sensitive to the light elements such as carbon, and thus
EELS is used here. Across the gap areas shown in the yellow arrow in Fig.4.2g, the
line scan of EELS analysis for carbon and silver was firstly done as shown in
Fig.4.2h. The result shows that the gap size is finely controlled to be ~ 2 nm, which
is consistent to the twice of the experimental ta-C thickness. Furthermore, the EELS
mappings for the areas (shown in Fig.4.2g) were performed and the result was
shown in the Fig.4.2i. In Fig.4.2i, red area was the Cr mapping, blue area was the
carbon mapping, and green area was Ti mapping. The white area represented the
Ag tips. The Ti deposition was to improve the adhesion between the polymer and
metal. The EELS mapping showed that Ag nanoparticles were uniformly covered
by continuous and conformal ta-C film deposition. In addition, the carbon EELS
signals above the dash line shown in Fig.4.2i were also observed, which may be
induced by the trapped molecules and carbon contamination during the FIB process.
It also proved that the molecules were easily trapped on the bottom of the saddle
area, and thus the spilling-out EM field from the nanogap could enormously
enhance their Raman signals.
60

Figure 4.2 SEM and HR-TEM characterizations for gap plasmonic
nanofingers. (a) SEM image of nanofingers before collapse. The inset is the TEM
cross-sectional view of single nanofinger. (b) HR-TEM cross-sectional view of 1
nm ta-C coated single Ag nanofinger. (c) Top-view SEM image of 1 nm ta-C
coated Ag nanofingers array after collapse. (d) TEM cross-sectional view of Au
nanofingers after collapse. (e) TEM cross-sectional view of 1 nm ta-C coated Ag
fingers after collapse and its EDS mapping of Ag (f). (g) HR-TEM cross-sectional
view of 1 nm ta-C coated Ag nanofingers after collapse. (h) EELS line scan of C
and Ag in the HR-TEM image of 1 nm ta-C coated Ag fingers after collapse. (i)
EELS mapping of 1 nm ta-C coated Ag fingers after collapse in the HR-TEM
image (red color area is Cr; blue color area is carbon; white color area is Ag; green
61

area is Ti.). Scale bars: 600 nm for (a) & (c); 30 nm for inset of (a), (d), (e) & (f);
5nm for (b), (g) & (i)

4.4 SERS Characterizations for ta-C Coated Gap Plasmonic
Nanofingers
After careful characterizations of ta-C coated gap plasmonic nanofingers, we
investigated their potential in SERS by choosing a model molecule of Rhodamine
6G (R6G) in our experiments. As a comparison, both Au nanofingers and ta-C
coated Ag nanofingers are simultaneously incubated in a 10
-5
M R6G ethanol
solution for 30 min, taken out and air-dried. They are subsequently rinsed
thoroughly with ethanol, which ensures that the coverages of R6G molecules are
similar on both the Au and ta-C surface. After this incubation process, ta-C coated
Ag nanofingers arrays form the gap plasmonic nanostructures that produce strong
local electric fields for Raman detection, while the Au nanofingers will touch each
other. Fig.4.3a shows the Raman spectra of R6G on these two kinds of SERS-active
substrates, in which typical Raman peaks of R6G are consistent with previous
reports
50
. Although the results show that the Raman intensity for both substrates
are similar, it is worthwhile to note that the laser power output for ta-C coated Ag
nanofingers is only 10
-3
of that for Au nanofingers. By the evaluation of their SERS
enhancement factors (EF) through a roughly simplified formula
(EF=(Isurf*Nvol)/(Ivol*Nsurf), where Ivol and Isurf are the conventional Raman and
62

SERS intensities, Nvol and Nsurf represent the number of molecules probed in a bulk
sample and on the SERS substrate, respectively.)
41
, it indicates that the EF on the
ta-C coated Ag nanofingers is ~10
11
while the EF on the Au nanofingers is ~10
8
.
To characterize the uniformity of SERS-active substrates, the Raman mapping with
a step of 1 µm was performed, as plotted in Fig.4.3b, which confirms that the
variation of Raman intensity is limited no more than 10%, showing the excellent
uniformity of ta-C coated ordered-Ag nanofingers array for a SERS substrate.
At present, in SERS application for big analytes such as cells, it is beneficial to
provide a 2-dimentional distribution of hotspots. Our closed ta-C coated Ag
nanofingers can provide uniformly ordered hotspots for these large analytes
situated upon the surface of the finger tips. In addition, the saddle shape region
formed by the pairs of the collapsed and touched nanofingers could greatly
contribute to limit the moving of living cells in solution. Fig.4.3c shows the SERS
spectrum of a living yeast cell situated on the top of the nanofingers. The Raman
peaks of 1330cm
-1
, 1380cm
-1
, 1435cm
-1
, 1625cm
-1
are similar to the Raman spectra
of proteins
36
. In our experiment, the laser output power was only 2.5µ W, which is
essential for Raman measurement of living substances. We also evaluated the
chemical stability of the ta-C coated Ag nanofingers array with different storage
time under ambient conditions. It is in consequence of the chemical protection from
the 1 nm-thick ta-C film that the Ag nanofingers array shows excellent SERS
activity even after one year, shown in Fig.4.3d.
63

Figure 4.3 SERS characterizations. (a) Raman spectra of R6G on ta-C coated Ag
nanofingers array (red, laser output power is 2.5µ W) and uncoated Au nanofingers
array (blue, laser output power is 2.5mW). (b) Raman mapping intensity at 1650
cm
-1
on ta-C coated Ag nanofingers array after closing. (c) Raman spectrum of a
living yeast cell on ta-C coated Ag nanofingers after closing. (d) Raman spectra of
R6G on ta-C coated Ag nanofingers array after closing with different storage times.
All Raman spectra were obtained at the excitation wavelength of 532 nm.

4.5 Numerical Simulations
It was previously reported that due to the excitation of localized surface plasmon
resonances, the strongest field enhancements in the gaps between closely
64

interacting metal nanoparticles could be obtained
25
. However, for physically
touched gap plasmonic nanostructures which means the gap was defined by the
dielectric region between the metallic nanoparticles, the greatly enhanced EM fields
should be located within the dielectric gap region where the analyte molecules
cannot access
25,39
. In order to explore the mechanism of enhanced EM field in our
proposed microstructure, the simulations based on classical electromagnetic theory
were performed. Due to the relatively high tunneling barrier height for our 2-nm-
ta-C gap plasmonic structures, we have not considered the quantum mechanical
effects in our calculations
43
. We start the numerical simulation from the single ta-
C coated Ag nanofinger and then consider the coupling for touched nanofingers by
using a commercial finite-element method based software package (COMSOL
Multiphysics) in the following discussion.
Firstly, the numerical simulations of the scattering spectra for the single Ag
nanofinger coated with various dielectric layers were performed. The numerical
model was approximated by a polymer cylinder of 650 nm in height and 70 nm in
diameter with its axis normal to the substrate. This polymer cylinder was capped
on top by a 50 nm-thick Ag nanodisk with the same diameter forming a metal-
dielectric nanofinger structure, which was then integrally coated with an ultrathin
ta-C film of 1 nm thickness. A 50 nm-thick continuous Ag film was also modeled
on the semi-infinite silicon substrate corresponding to the actual experimental
process of Ag deposition. The permittivity of Ag was interpolated from the
65

experimental data
51
, and the refractive index of the polymer was set as 1.41
49
. As
shown in Fig. 4.4a, there is a dominant scattering peak at ~ 475 nm with an incident
electric field, which corresponds to a plasmon-induced dipole resonance identified
by the typical surface charge distribution in the inset. The scattering performance
of the single metal-dielectric nanofinger is not significantly affected by increasing
the refractive index (nd) of the dielectric coating layer, except for a slight spectral
red-shift of the resonant peaks. For the sample without dielectric coating (i.e., nd =
1), the electric field would strongly bound at the Ag/air interface under the
excitation of this dipole resonance, and then decay rapidly along the radial direction
shown as Fig. 4.4b. However, the electric field at resonance would be spatially re-
distributed due to the existence of the dielectric coating around the bare Ag
nanodisk. Discontinuity of the electric field at a boundary between the two different
media could be described as n·( ε1E1- ε2E2)=0 (with n being the unit vector
perpendicular to the interface, and ε1, ε2 being the permittivity of media on both
sides of the boundary), which leads to an abrupt increase of the electric field across
the boundary from the dielectric to air, followed by Eair_b=( εd/ εair)Em_b. Here, Eair_b
and Em_b are defined as electric field at the air-side and dielectric-side of the
boundary, respectively. On the other hand, the amplitude of Eair_b is also determined
by the decay of the electric field from the silver/dielectric interface. In our work,
we chose the high-index (nd = 2.44 for 1 nm ta-C film, measured by J. A. Woollam
ellipsometer in which the model was set up by the Veeco company) dielectric of ta-
66

C as coating layer. Based on the above principle, in the case with a sufficiently thin
coating layer, the electric field at ta-C/air interface might be much stronger than
that at the silver/ta-C interface, as show in Fig. 4.4c.

Figure 4.4 Simulations for the uncoated and 1 nm ta-C coated single Ag
nanofinger. (a) Scattering spectra of uncoated and 1 nm dielectric coated single Ag
nanofinger, and the inset is the charge distribution. (b) Plasmonic electric field of
Emax/Einc for the uncoated single Ag nanofinger, and the inset is the electric field
distribution. (c) Plasmonic electric field of Emax/Einc for the 1 nm ta-C coated single
Ag nanofinger, and the inset is the electric field distribution.

In our case, by tilting the pairs of nanofingers oppositely as described above, the
top Ag nanodisk with 1 nm-thick ta-C coating would gradually approach and
67

contact each other eventually. The simulated evolution of the spectral scattering as
a function of gap size (distance between the tips of the dimer-like ta-C coated silver
nanodisks, dgap) which is determined by the tilted angle ( θ) of the ta-C coated
nanofinger, is plotted in Fig. 4.5a. The incident electric field polarized along the
dimer axis. For this dimer with the maximum separation (i.e., not being tilted), the
scattering spectrum is dominated by the single-nanofinger dipole mode, since the
interaction between Ag nanodisks dimer is fairly weak. With decreasing dgap, the
stronger near-field interactions of the two approaching ta-C coated Ag nanodisks
hybridize the dipole resonances into bonding-dimer plasmom modes, which present
an obvious redshift
52,53
. It is worthwhile to note that at extremely small dgap, a new
resonant band appeared at the shorter wavelength, being attributed to a quadrupole
interaction of the dimer system.
Furthermore, the calculated electric field enhancement (E/Einc, where Einc is the
electric field intensity at the 532 nm excitation, aiming at the actual Raman
excitation wavelength) versus dgap is depicted as shown in Fig. 4.5b. To avoid the
singularity of the field value, the selected probe-point of electric field (marked as
point A in the inset of Fig. 4.5b) is set at 0.25 nm away from the midpoint of the
connected-line of tips of dimer. The EA/Einc increased exponentially with
decreasing dgap, and the highest enhancement of the localized electric field is
obtained when the ta-C coated Ag nanodisks dimer is perfectly touched. In this case,
the SERS enhancement factor which is generally proportional to the fourth power
68

of the local electric field enhancement ((E/Einc)
4
), would reach up to ~10
12
as shown
in Fig. 4.5b. The ultrathin high-index dielectric coating layer around the metal could
modify the plasmonic fields and enable a relatively larger electric field to be
localized at the outer surface of the dielectric layer rather than in the dielectric gap
between the metal counterparts as shown in Fig. 4.5c. The volume, with
enhancement over 10
8
, is roughly 51 nm
3
and the maximum enhancement is 10
12
.
Such a huge field enhancement combined with a considerable large “hot” volume
exposure compared to free space is very beneficial for single-molecular SERS
detection
23
. Thus, due to the modification of ta-C to surface plasmons of Ag
nanofinger, the coupled EM field can spill out from the dielectric gap, which is
responsible for the SERS enhancement in single-molecular level.

69

Figure 4.5 Simulations for 1 nm ta-C coated Ag nanofingers after closing. (a)
Scattering spectra of 1 nm ta-C coated Ag nanofingers after closing as a function
of the gap sizes. (b) Plasmonic electric field of EA/Einc for 1 nm ta-C coated Ag
nanofingers after closing with the gap sizes, and the inset is the charge distribution
with different gap sizes. (c) Plasmonic electric field of Emax/Einc and charge
distribution for 1 nm ta-C coated Ag nanofingers with physical touch.

4.6 Single-Molecule Detection
We used a bi-analyte SERS method
50
to demonstrate the single-molecule
detection in our suggested platform just as we illustrated by the enhancement factor
via experiments and theoretical calculation. This method is a contrast based
spectroscopy technique using two different molecules at the same time, and
facilitates reliable statistics based on a large spectral sample size for single-
molecules SERS detection
50
. In our case, an aqueous solution of R6G and lysozyme
of the same concentration was dispensed on the touched gap plasmonic nanofingers
with 2-nm-ta-C gap sizes for single-molecules SERS at 532 nm excitation. The
significant difference between R6G and lysozyme’s Raman fingerprints make them
highly distinguishable. Raman intensity mappings of selected molecule vibration
modes (R6G band at 613 cm
-1
and lysozyme band at 821 cm
-1
)

at 10
-10
M and 10
-12

M concentrations are shown in Fig. 4.6a-d. At extremely low concentration of 10
-
12
M, percentage of overlapping R6G-lysozyme signal decreases and the single
70

molecule regime has occurred, as shown in Fig. 4.6c-e. Based on the argument of
Le Ru and Etchegion
54
, the bi-analyte nature of target molecules serves as strong
evidence that the spectrum of one certain analyte from an individual pixel is
attributed to single molecule
54
. This single-molecular SERS experiments
demonstrate that our purposed gap plasmonic structure can realize single-molecular
label-free detection, and are consistent with our theoretical simulation.

Figure 4.6 Spatial resolved Raman intensity mapping at different
concentrations. (a) Mapping of R6G peak at 613 cm
-1
(red), (b) Mapping of
lysozyme peak at 821 cm
-1
(green) intensity mappings of the same area at 10
-10
M.
(c) Mapping of R6G peak at 613 cm
-1
(red), (d) Mapping of lysozyme peak at 821
cm
-1
(green) intensity mappings of the same area at 10
-12
M. (e) Selected spectra
from the mapping area. Scale bar for mappings is 5 µ m.
71

Chapter 5. Application of Nano-Fingers in Plasmon
Enhanced Fluorescence

5.1 Introduction to Plasmon Enhance Fluorescence
Fluorescence has diverse applications, such as single molecule detection,
biological labelling and optoelectronic devices. Since the fluorescent emission of
the molecules can be enhanced by increased incident electromagnetic (EM) field
intensity on the molecules, plasmon-enhanced fluorescence (PEF) is widely
adopted thanks to the enhanced EM field in the vicinity of metal nanostructures.
The cartoon shown in Fig. 5.1 is an attempt to provide a simplified picture of
plasmon enhanced fluorescence. P0 is the excitation power to be absorbed by both
the fluorophore and the nanoparticle leading to an excited electronic state of the
molecule and a LSPR in the nanostructure. LSPRs are collective oscillations of
conduction band electrons that generate intense EM local fields in their vicinity
with exponential spatial variation on the nanometer scale, and their coupling to the
fluorophore molecule could improve the absorption and emission quantum
efficiency.

While many attempts have been made in recent years to improve the
performances of PEF, its development is hindered in applications that require high
72

sensitivity and precision due to limited enhancement factor (EF). There are two
major factors affecting the enhancement factor in PEF, EM field intensity on the
molecules and energy transferred from the excited molecule to the vicinal metal
that quenches molecular emission. It has been widely accepted that the greatest
degree of EM field can be obtained in gap plasmonic nanostructures, which are
formed by metallic nanoparticles with sub-nanometer inter-particle gaps. A precise
control on the gap size at sub-nanometer scale is critical to gain the strongest
plasmonic hotspots. On the other hand, the key point that determines the molecule-
to-metal energy transfer is the distance between the molecule and the metal. In
order to experimentally investigate the continuous transition from fluorescence
enhancement to fluorescence quenching, one needs to vary molecule-metal distance
continuously with sub-nanometer precision at optimally designed plasmonic
hotspots. However, the limitation of traditional fabrication techniques poses serious
obstacles at such a small length scales. Moreover, it is important to understand the
correlation between EM field enhancement and energy transfer, since that plays the
key role in the optimization of PEF nanostructures.
73

Figure 5.1 Cartoon of simplified plasmon enhanced fluorescence.

5.2 Capillary Force Induced Molecule Trapping in Nano-Fingers
Here we present an approach to fabricate large-area gap plasmonic nanostructures
with atomic precision control of gap size and molecule-metal distance
deterministically in high reliability and high throughput at low cost. That facilitates
a detailed study on the effects of both field enhancement and energy transfer on
plasmon-enhanced fluorescence when the sub-5 nm molecule-metal distance and
the gap size are varied with sub-nanometer precision. This approach can address
the challenges mentioned above in PEF, snamely, charge recombination through
quantum mechanical electron tunneling at small gap size and fluorescence
quenching through energy transfer at small molecule-metal distance. Therefore,
optimization of PEF substrates can be realized for strongest EF with large effective
74

area uniformly in future design of PEF related applications. Our method is based
on collapsible nanofingers fabricated using nanoimprint lithography (NIL),
reactive-ion etching (RIE) and atomic-layer deposition (ALD). Figure 5.2a shows
a schematic illustration of such a gap plasmonic nanostructure. Au nanoparticles
are placed on top of nanofinger pairs in flexible polymer (i.e., nanoimprint resist)
with high aspect ratio. A thin conformal dielectric layer is deposited uniformly by
ALD onto the nanofingers before we collapse them using capillary force. The
dielectric layer acts as the spacer to prevent Au nanoparticles from directly
contacting each other, which induces strong electron transfer and leads to a
dramatic decrease of the localized EM fields. Since the ALD process deposits the
dielectric films with high conformity and atomic precision, the gap size between
two Au nanoparticles is accurately defined by twice the dielectric film thickness.
In order to obtain the strongest EM field enhancement at the plasmonic hotspots,
the gap size can be optimized based on different dielectric materials that poses
different tunneling barrier heights for electrons. When the collapsed nanofingers
are used in plasmon-enhanced fluorescence, the small fluorophores are easily
trapped into the spilling-out EM filed around the hotspot at the gap center due to
capillary force, as shown in Figure 5.2b. As the dielectric film covers the
nanofingers conformally and uniformly, the molecule-metal distance is well
defined around the Au nanoparticles by the ALD film thickness and can be varied
continuously with sub-nanometer steps in the sub-5 nm range. In that sense, the
75

dielectric gap size is twice the molecule-metal distance. Based on the absorption
spectra of collapsed nanofinger, plasmon resonance frequency of this gap
plasmonic nanostructure can be determined. Correspondingly, Nile blue was
deposited as the fluorophore to the collapsed nanofingers, as its absorption
frequency coincides with the plasmon frequency. Here, a 633 nm-wavelength laser
is used to excite the Nile blue deposited on the collapsed nanofingers accordingly
for fluorescence intensity characterization. The evolution of the Nile blue
fluorescence signature peak at 667 nm as a function of gap size for different
dielectric gap materials was analyzed. We successfully detected the optimal gap
size for ultra-strong fluorescence intensity up to 1631-fold over large area
uniformly. In addition, the 592 cm
-1
Raman fingerprint of Nile blue was also
examined as the gap size varies. The Raman scattering intensity scales to the fourth
order of the EM field, which is a direct indicator of the localized field intensity at
the plasmonic hotspots. A comparison between fluorescence intensity and Raman
intensity versus dielectric gap size reveals the contributions of both field intensity
and energy transfer mechanism to plasmonic fluorescence enhancement and
quenching in sub-5 nm range. The quantum mechanical electron tunneling
dominates at this small scale, while the dielectric material guides the tunneling
effect by changing the tunneling barrier height.

76

Figure 5.2 Schematic diagram of the controllable plasmonic hot spots created by
collapsible nano-fingers for fluorescence enhancement. (a) A pair of metallic
nanoparticles with ALD-coated dielectric layers contact each other when flexible
nano-fingers beneath them collapse. The spatial gap size is defined by twice the
dielectric layer thickness, and the hot spot is at the gap center. (b) The fluorescent
dye molecules are deposited onto collapsed nano-fingers, and the fluorophores
adhere to nanofingers after solvent evaporation and are excited by the laser.

5.3 Characterizations of Nano-fingers for PEF
Figure 5.3a shows the scanning electron microscopy (SEM) image of 2.5 nm
TiO2 coated nanofingers before the collapsing process. The fabrication procedure
of collapsible the nanofingers is based on nanoimprint lithography (NIL) and
reactive-ion etching (for details, see Methods). The nanofingers are formed by Au
nanoparticles on top of UV nanoimprint resist pillars. Nanofinger arrays over large
77

area (1.4 inch by 1.4 inch) can be fabricated uniformly with high throughput by the
advantage of NIL, which also guarantees a good signal-to-noise ratio. It has been
reported that the appropriate aspect ratio and geometry of nanofingers is the key to
achieving a successful collapsing process. The diameter and height of each
nanofinger is 60 nm and 350 nm respectively, and the pitch is 200 nm. The
collapsing process is driven by the capillary force, while the nanofingers are soaked
into ethanol and air-dried. Similar mechanisms were reported before. After the
collapsing process, the nanofingers will close together to form a completely
touched dimer nanostructure, as shown in Figure 5.3b. The van der Waals force
keeps the nanofingers from separating once they touch. In order to demonstrate the
coverage of the TiO2 film on the Au tip of the nanofingers, transmission electron
microscopy (TEM) cross sectional analysis was performed at the nanogap after
collapsing process. The TEM image in Figure 5.3c shows two Au nanoparticles that
are separated by a 5 nm gap, which is twice the thickness of the ALD coated TiO 2
film. Corresponding energy-dispersive X-ray spectroscopy (EDS) maps of Au, Ti,
and O shown in Figures 5.3d, 2e and 2f also confirm the 5 nm TiO2 gap between
the two Au nanoparticles. This indicates that the dielectric film covers the Au
nanoparticles uniformly and conformally, which further acts as the spacer between
the adsorbed fluorophores and the Au nanoparticles and defines the molecule-metal
distance by the dielectric film thickness. Since the precise gap size and molecule-
metal distance are controlled by the ALD deposited film thickness, the critical
78

dimensions can reach sub-nanometer, atomic precision. The collapsed nanofingers
were then soaked into 1 µ M Nile blue ethanol solution and air-dried for further
fluorescence characterization. It has been proven that, due capillary forces, the
fluorescent Nile blue molecules are easily trapped on the bottom of the saddle area
at the touch point, which is the hottest plasmonic hotspot exhibiting ultra-strong
EM field enhancement.
79

80

Figure 5.3 (a) SEM image of nano-fingers before collapse. (b) SEM images of
nano-fingers after collapse. The inset is the image of the same nano-fingers from a
different viewing angle. The diameter and height of each finger are 60 nm and 350
nm, respectively. The ccale bars in the SEM images are 200 nm. (c) TEM image of
the dielectric nano-gap in the collapsed nano-fingers. (d) EDS mapping of Au in
the TEM image. (e) EDS mapping of Ti in the TEM image. (f) EDS mapping of O
in the TEM image. Scale bars in the TEM images are 10 nm.

5.4 Simultaneous Monitor on Plasmon Enhanced Fluorescence and
SERS
The fluorescence intensity of fluorophores is determined by two processes: (i)
excitation by the incident field affected by local environment and (ii) emission of
radiation influenced by the competition between radiative and nonradiative decay.
Plasmon-enhanced fluorescence mainly benefits from the enhancement of the
excitation process, which scales as the square of local EM field enhancement at the
plasmonic hotspots. The gap size plays a critical role in the plasmonic field
enhancement of gap plasmonic structures, including our collapsed nanofingers.
While the shrinking gap size is expected to generate increased field enhancement,
the quantum mechanical electron tunneling effect becomes significant in the sub-5
nm gap regime. Due to the balance between the bonding dimer plasmon and charge
transfer plasmon, the optimal gap sizes for strongest field enhancement have been
81

demonstrated experimentally for different gap materials. However, the optimal gap
size for field enhancement is not necessarily optimal for fluorescence enhancement.
The emission of excited fluorophores is significantly influenced by the nonradiative
energy transfer to Au nanoparticles they adsorb, which leads to a decrease of
quantum yield (fluorescence quenching). Nonradiative energy transfer is strongly
dependent on the molecule-metal distance and significantly quenches the
fluorescence at sub-5 nm range. The distance between adsorbed Nile blue
molecules and Au nanoparticles is defined by the dielectric spacer thickness, which
is half of the gap size in the collapsed nanofingers. Therefore, the optimal gap size
for PEF is expected to be larger than the one providing strongest field enhancement
in compensation for this energy transfer mechanism.

In order to investigate the plasmon-enhanced fluorescence in the collapsed
nanofingers, we need to adjust the gap size in fine steps reliably. Figure 5.4a shows
the fluorescence spectra of Nile blue on collapsed nanofingers with TiO2 as the gap
material. The gap sizes were chosen as 2, 3, 4, 5, 6, 7nm respectively based on
previously reported 4 nm optimal size for strongest EM field enhancement. The 667
nm Nile blue fluorescence peaks (broad peaks) and the 592 cm
-1
Raman fingerprints
(sharp peaks) are examined simultaneously as the gap size varies for comparison.
The measured fluorescence intensity and Raman intensity of Nile blue for different
values of gap sizes are plotted in Figure 5.4b. The fluorescence intensity reaches
82

the maximum at the 5 nm gap size, while the Raman intensity maximizes at 4 nm
gap. As the direct indicator of the field enhancement at plasmonic hotspot, Raman
intensity scales as E
4
, where E is the local field. Hence the gap size for strongest
Raman enhancement coincides with that which exhibits the strongest plasmon light
absorption, as reported before. However, the fluorescence intensity increases
significantly when the gap expands from 4 nm to 5 nm, and gradually decreases
when the gap size exceeds 5 nm. This successfully demonstrates the balance
between field enhancement and energy transfer in plasmon-enhanced fluorescence,
which deviates the fluorescence response from a pure field driven process like
Raman signals at this sub-5 nm scale. The uncertainties of the fluorescence and
Raman intensities originate from the fluctuation across the large area sample,
including defects and geometry variance. That indicates uniform fluorescence over
the whole area of large nanofinger arrays. To simulate the fluorophore excitation
process, which is purely field driven, finite-difference time-domain (FDTD)
calculation of the localized EM field distribution was performed (Figure 5.4c). The
633 nm plain waves polarized along the x-direction were incident from the top
perpendicularly to the collapsed nanofingers with 5 nm TiO2 gap. The settings are
the same as that in the experiment, which ensures the strongest plasmonic resonance
and longitudinal bonding dipole plasmon mode. The maximum field enhancement
occurs around the contact point where the most fluorophores tend to adsorb. To
evaluate the enhancement factor (EF) of PEF, the same fluorescence measurement
83

was repeated on two control samples. One is plain glass, the other one is non-
collapsible nanofingers with same dielectric coating. The non-collapsible
nanofingers were fabricated similarly as the collapsible ones with shorter RIE
etching time, which resulted in reduced height. They will not bend and contact
during the soak-dry process. In this case, they behave essentially as a core-shell
nanoparticles array, which is widely used in previous PEF study. As shown in
Figure 5.4d, the fluorescence intensity of Nile blue on collapsed nanofingers with
5 nm TiO2 gap is enhanced by 620-fold compared to that on plain glass, while the
enhancement factor on non-collapsible nanofingers with same the TiO2 coating is
only 20.5-fold. Compared to strongly coupled longitudinal bonding dipole plasmon
mode in the plasmonic dimers of collapsed nanofingers, the isolated core-shell
monomers provide much weaker local EM field enhancement and thus significantly
reduced fluorescence intensity.
84

Figure 5.4 (a) Fluorescence intensity of Nile blue on collapsed nanofingers with
different TiO2 gap sizes. (b) The measured fluorescence intensity and the measured
592 cm
-1
signature Raman intensity of Nile blue for different values of gap size. (c)
FDTD calculation of local field enhancement in collapsed nanofingers with 5 nm
TiO2 gap. (d) Fluorescence intensity of Nile blue on collapsed nanofingers with 5
nm TiO2 gap, non-collapsible nanofingers with same TiO2 coating and plain glass.

85

5.5 Dielectric Quenching
When molecules get closer to the metallic nanoparticles, the nonradiative decay
rate increases significantly, which strongly quenches the emission power.
Fortunately, the nanofingers substrates have a dielectric spacers layer between
fluorescence molecules and metallic nanoparticles to avoid direct contact between
them. However, instead of contact with metallic nanoparticles, molecules directly
touch the dielectric film. We must modify our dielectric spacer material selection,
considering the quenching effect from its contact. While TiO2 quenches
fluorescence significantly, Al2O3 does not quenches fluorescence significantly.
Both materials exhibit increased quenching as thickness increases, however that
saturates around 1.5 to 2 nm. The result of Nile blue fluorescence spectra on plain
TiO2 and Al2O3 substrate are shown in Fig. 5.5.
(a)
86

(b)

87

Figure 5.5 Dielectric spacer quenching. (a) TiO2 quenches fluorescence
significantly, (b) Al2O3 does not quenches fluorescence significantly. Both
materials exhibit increased quenching as thickness increases, however that saturates
around 1.5 to 2 nm.

5.6 Minimize the Quenching Effect
As the key role in collapsed nanofingers, the dielectric material of the spacer
contributes to the plasmon-enhanced fluorescence from both field enhancement and
energy transfer aspects. Since the electrons tunnel between Au nanoparticles
through the dielectric spacer, a higher tunneling barrier that is defined as the
difference between the Fermi energy of gold (5.1 eV) and the electron affinity (EA)
of the dielectric material will result in reduced tunneling effect and therefore
increased field enhancement at optimized gap size. The reduced tunneling effect is
manifested by a narrower optimal gap for strongest field enhancement. On the other
hand, the dielectric material influences the energy transfer rate from fluorophore to
metal even though it has limited energy dissipation compared to Au nanoparticles.
From these field enhancement considerations, we applied Al2O3 as the gap material
in the collapsed nanofingers in comparison to TiO2. Since the electron affinities of
TiO2 and Al2O3 are 4.2 eV and 1.0 eV respectively, Al2O3 retains a much larger
tunneling barrier height (4.1 eV) than TiO2 (0.9 eV). We repeated the fluorescence
and Raman characterization of the Al2O3 coated nanofingers with different gap
88

sizes and the results are shown in Figures 5.6a and 5.6b. The Raman intensity is
enhanced as the gap size reduces from 7 nm all the way to 2 nm, which implies the
strongest fluorophore excitation occurs at 2 nm Al2O3 gap. However, the
fluorescence intensity is weaker than that at wider gaps. As the result of balancing
with energy transfer process, the 5 nm Al2O3 gap exhibits largest fluorescence
enhancement. Figure 5.6c shows FDTD calculation of the localized EM field
distribution in collapsed nanofingers with 5 nm Al2O3 gap. As indicated by the
Raman intensity as well, the field enhancement factor is slightly smaller than that
in 5 nm TiO2 gapped nanofingers. Nevertheless, the fluorescence enhancement of
5 nm Al2O3 gapped nanofingers enhance the Nile blue fluorescence by 1631-fold
uniformly across the large area, which is larger than the largest fluorescence
enhancement in TiO2 gapped nanofingers (Figure 5.6d). This can be attributed to
the reduction of energy transfer from molecule to metal, which therefore enhance
the quantum efficiency and overall fluorescence intensity.
89

Figure 5.6 (a) Fluorescence intensity of Nile blue on collapsed nanofingers with
different Al2O3 gap sizes. (b) The measured fluorescence intensity and the
measured 592 cm
-1
signature Raman intensity of Nile blue for different values of
gap size. (c) FDTD calculation of local field enhancement in collapsed nanofingers
with 5 nm Al2O3 gap. (d) Fluorescence intensity of Nile blue on collapsed
nanofingers with 5 nm Al2O3 gap, non-collapsible nanofingers with same Al2O3
coating and plain glass.

90

To get a deeper understanding of fluorescence quenching mechanism on the
quantum-mechanical level, here we provide an analytical model to describe the
emission process of the excited fluorophores, which can ultimately guide the
effective design of potential applications. There have been extensive studies of the
fluorescence quenching phenomena in gap plasmonic structures with relatively
large distances and variation steps. However, the calculations and mechanism of
fluorescence quenching become more complicated when the gap sizes shrink, such
that quantum tunneling effects between fluorophores and nanoparticles play the
important role. To analyze the fluorescence quenching mechanism which
contribute in emission process, we could first eliminate the field enhancement
effect that contributes to fluorescence intensity in the fluorophore molecule
excitation process. Since the simultaneously measured Raman intensity is a direct
indicator of the localized field intensity at the plasmonic hotspots which is not
affected by the fluorescence quenching effect, we could divide the fluorescence
intensity by the square root of the simultaneously measured Raman intensity and
define the ratio as the fluorescence efficiency (FLE) which purely depicts the
emission process contributing to the fluorescence intensity. Figure 5.7a shows FLE
of Nile blue in the collapsed nanofingers with two kinds of gap materials -- TiO2
and Al2O3 for different values of the gap size 𝑑 . Here we address three main
physical processes to explain the variation of FLE with 𝑑 . First, the excited Nile
blue molecules radiate to the Au-nanoparticle dimer with emission rate Γ which can
91

be derived from Weisskopf–Wigner approach
73-74
. Since the molecules are closely
surrounded by Au nanoparticles, direct radiation from the molecules into free space
is negligible
75
. It turns out that Γ depends on the electromagnetic local density of
states (LDOS) 𝜌 (𝜔 𝑟 ) at the emission frequency 𝜔 𝑟 (=667 nm) of the molecule. This
LDOS which depends on the geometric structure of the dimer-cavity, for example,
the dimer-gap 𝑑 , describes the number of available photon modes in dimer-cavity
at the emission frequency of Nile blue and furthermore determines energy-transfer
strength from molecules to Au nanoparticles. Such strength is proportional to
experimentally measured absorption strength of the Au-nanoparticle dimer-cavity
at the frequency 𝜔 𝑟 , labeled as Abs (𝜔 𝑟 ), for different gap sizes. Secondly, the
energy emitted to the Au nanoparticles can either radiate into free space (to be
detected as the fluorescence signal) or lose energy to heat (Ohmic loss)
75
. Since
the radiation wavelength is much larger than the distance between molecules and
the surface of Au nanoparticles, by using electrostatic approximation and the
image-charge method
76
, we find that the Ohmic loss is proportional to 1 𝑑 4
⁄ (put
the detailed derivation in APPENDIX if needed). Furthermore, at such a small gap
dimension, electron-tunneling between the molecules and the Au nanoparticles
leads to a weakening of the radiative fluorophore dipole 𝑝 𝑟 → 𝑝 𝑟 (1 − 𝑒 −𝜆𝑑 /2
),
where 𝑒 −𝜆𝑑 /2
accounts for the tunneling probability
77
. Noticing that Ohmic loss
and tunneling effect are more significant at smaller gap size, we finally end up with
a phenomenological formula for FLE (in arbitrary units) as
92

FLE= 𝐴 ∙ A bs (𝜔 𝑟 )(1 −
𝐵 𝑑 4
) (1 − 𝑒 −
𝜆𝑑
2
)
2
.
From here we can see FLE, which has already excluded the field enhancement
effect, is now determined by the energy emitted from the molecules to the Au
nanoparticles (~Abs (𝜔 𝑟 )), subjected to tunneling effect and Ohmic loss. We can
experimentally obtain the ratio
FLE
Abs(𝜔 𝑟 )
and then fit it to the function 𝐴 (1 −
𝐵 𝑑 4
) (1 − 𝑒 −
𝜆𝑑
2
)
2
, for two gap materials, shown in Figures 5.7b and 5.7c,
respectively. The uncertainties are caused by the fabrication fluctuation between
individual measuring spot among different samples, including defects and
geometric variance. Triple standard deviation was applied as the error to the fitting
for a confidence interval of 99.7%, which indicates a permissible error range. The
inverse of the fitting parameter 𝜆 measures characteristic tunneling-length of the
gap material. By fitting the experimental data, we find that 1 𝜆 TiO
2
⁄ ≈ 7.14 𝑛𝑚 and
1 𝜆 Al
2
O
3
⁄ ≈ 3.34 𝑛𝑚 , indicating that TiO2 has a larger characteristic length of
tunneling than Al2O3. This qualitatively agrees with the fact that TiO2 has a
smaller tunneling barrier than Al2O3. Simple estimation from our fitting
parameters indicate that the tunneling barrier of Al2O3 is about (
𝜆 Al
2
O
3
𝜆 TiO
2
)
2
≈ 4.57
times as of TiO2 , which is pretty close to the realistic ratio
4.2 eV
0.9 eV
≈ 4.55. To this
point, we could observe that both tunneling effects and Ohmic loss to metallic
nanoparticles reduce as the gap size increases. On the other hand, a close match
93

between the inherent emission frequency of fluorophore molecules and the strong
absorption frequency of gap plasmonic structure will be beneficial to increase the
fluorescence intensity. We should mention that this model stands within small gap
sizes (< 10 nm), since the coupling between the fluorophore molecules and metallic
nanoparticles will decay with increased distances. Since the strongest field
enhancement and therefore excitation stay at this small gap range, we could figure
out this model for further design in potential applications with the subnanometer
precision step on the spacer thickness change.

Figure 5.7 (a) Fluorescence efficiency of Nile blue on collapsed nanofingers with
different TiO2/Al2O3 gap sizes. (b) The measured FLE /Abs (𝜔 𝑟 ) of the molecule
fluorophore for different values of gap size, with the gap material being Al2O3.
The data is fitted according to the equation-(label). (c) The measured FLE /Abs (𝜔 𝑟 )
of the molecule fluorophore for different values of gap size, with the gap material
being TiO2. The data is fitted according to the equation-(label).
94

Chapter 6. Plasmon Enhanced Photocatalysis using
Collapsible Nano-fingers

6.1 Introduction to Plasmon Enhanced Photocatalysis
Photocatalysis, a process that can convert solar energy into chemical energy, has
drew increasing attention for decades due to its great potential in the environmental
and energy field. Among various photocatalytic materials, semiconductor
photocatalysts are a type of promising materials that has been widely used in the
treatment of pollutants, water splitting and energy conversion. However, the large
band gap of these materials has limited the photocatalytic efficiency under solar
light. For instance, the band gap of TiO2 is 3.2eV, which means that only ultraviolet
light, accounting for ~4% of the solar energy, can be harvested to promote chemical
reaction. The poor ability to utilize visible light energy obstacle the further
improvement of their performance. To build visible-light-active photocatalyst, one
way is combining noble metal nanoparticles with semiconductor photocatalysts.
The surface plasmon resonances (SPRs) that produced by metal nanoparticles can
concentrate energy into ultra-small region near the surface, which can then be
dissipated into chemically active materials by generating hot carriers, resulting in
higher reaction rate.
95

An ideal plasmonic photocatalyst should interact strongly with visible light and
have high transfer efficiency for energy stored in the SPR modes. Recent studies
have shown that noble metal structures with nanogap features can provide high field
enhancement at the hot spot between two metals. The gap plasmon produced by the
two metals is characterized by intense EF enhancement. A straightforward way to
utilize this enhancement is to place semiconductor materials exactly at the hot spots,
providing a pathway for stored energy be dissipated by forming electron-hole pairs.
Here, we demonstrate a technique to fabricate gap-plasmonic photocatalyst by
combing collapsible nanofingers with thin dielectric film deposition. Au
nanoparticles are placed on the top of flexible polymer nanofingers. A thin layer of
TiO2 are deposited on the surface of nanofingers as the gap material. By
introducing capillary force using ethanol solution, four nearby nanofingers can
approach and touch each other, forming gap-plasmonic nanostructures. Compared
with chemically synthesized hybrid nanostructures, our gap-plasmonic
nanostructures are patterned with precise control over the size, shape and
interparticle distance. Since the plasmonic resonance of metal nanoparticles can be
tuned by controlling their size and shape, we can achieve high visible light response.
Also, the TiO2 thin film are placed exactly at the hot spots, ensuring most of the
concentrated energy can be utilized. The photocatalytic activities of our gap-
plasmonic nanostructures under visible light exposure were evaluated by photo-
degradation of methyl orange (MO). The experiment result shows our gap-
96

plasmonic nanostructure can achieve high catalyst activity under visible light.
Finite-difference time-domain (FDTD) calculations of the electric response of the
TiO2 coated gap-plasmonic nanostructures provide a quantitative prediction of the
photocatalytic enhancement factor and further calculation of the enhancement
factor of this nanostructure exactly correspond with our experiment values.

6.2 Fabrication of Nano-fingers for Photocatalysis
The large-area TiO2 coated gap-plasmonic nanostructures were prepared using a
previous reported method. First, one square inch NIL mold with two-dimensional
hole array (200nm pitch) was fabricated using double interference lithography (IL).
Then, a polymeric reverse-tone daughter mold was duplicated for further process
using UV-curable nanoimprint lithography. UV nanoimprint resist and lift-off
underlayer were spin-coated on a silicon substrate (Figure 6.1a), following a NIL
process to transfer pattern on the top layer (Figure 6.1b). A two-dimensional gold
nanoparticle array can be formed after residual layer etching, e-beam evaporation
and lift-off (Figure 6.1c). Preforming RIE step by using gold array as etching mask,
we can get high-aspect-ratio nanofingers on the substrate (Figure 6.1d). An ultrathin
TiO2 shell was then deposited on the surface of nanoparticles using atomic layer
deposition (ALD), as it can control the thickness of material in sub-nanometer level
(Figure 6.1e). Capillary force was introduced on each nanofinger when the sample
was soaked into ethanol solution and dried in air (Figure 6.1g). Due to the
97

symmetric geometry of this pattern, each four nearby nanofingers would approach
and contact each other to form a gap-plasmonic nanostructure. Figure 6.1f and
Figure 6.1i exhibit the EF enhancement of isolated and collapsed nanofingers
respectively. As we can see in the following discussion, the latter one can produce
much stronger local field.

Figure 6.1. a) Spin coating and UV curing 600nm UV NIL resist and then spin
coating 100nm lift-off layer and 100nm UV NIL resist. b) Performing UV curable
nanoimprint followed by residual layer etching. c) 50nm gold nanoparticles are
placed on the bottom layer after metal evaporation and lift-off process. d, g) High-
aspect-ratio nanofingers collapse when they are exposed to ethanol and dried in air.
e, h) TiO2 coated gap-plasmonic nanostructures are formed by ALD deposition and
98

collapse process. f, i) Schematic of EF enhancement of isolated and collapsed
nanofingers.

6.3 Enhanced Methyl-Orange Decomposition
The photocatalytic activity of our gap-plasmonic nanostructures was evaluated
by the photo-degradation of methyl orange (MO) under visible light irradiation.
Figure 6.2a shows our experiment set up. One square centimeter sample and one
milliliter 20mg/L MO aqueous solution were added into to a transparent cuvette
sealed by a plastic film. A green laser (532 nm, 3 W) beam was expanded and
reflected to the container vertically. The whole set up was covered by a black
curtain to avoid environment light. The UV-Vis absorbance spectra of the MO
solution were monitored by a Varian Cary 50 UV–Vis spectrophotometer before
and after exposure to green laser irradiation. Figure 6.2b, c, d show the MO
absorption spectra at different irradiation times by using ultra-thin TiO2 film,
isolated nanofingers and collapsed nanofingers as photocatalyst. First, we
performed a control experiment using one square centimeter silicon substrate
coated 2nm TiO2 film by ALD as photocatalyst. Only 1.5% of MO degraded after
irradiating the sample for 9h. Given that the photon energy is too weak to promote
TiO2 produce electron-hole pairs, this decrease can be contributed to the heating
effect of the strong laser power. Then, isolated nanofingers replaced the control
sample as photocatalyst. The plasmon produced by gold nanoparticles capped on
99

the isolated nanofingers can promote the TiO2 shell produce more electron-hole
pairs. As a result, the MO absorbance is observed to drop by 6%, a four-fold photo-
degradation rate compared with the former one. As the photocatalyst was changed
to collapsed nanofingers, a 35% reduction of the MO absorbance is observed. This
over twenty-fold improvement is contributed to the much stronger gap plasmon
produced by the gap-plasmonic nanostructure. The slight distortion of the
absorbance curve after exposure in the last experiment is believed due to the
formation of some byproduct.
The TiO2, whose band gap is 3.2eV, cannot absorb a photon with a wavelength
of 532nm (E=2.33eV) directly. One possible theory to explain the observed
enhancement is that two-photon process is happened inside the TiO2 due to the
intense field enhancement. In this case, the enhancement factor should be
proportional to the fourth power of the electric field. As only the energy
concentrated inside the dielectric can be transferred into chemical reaction, the
electric field enhancement in the volume of the catalyst is what we should focus on.
Furthermore, we can consider the photocatalytic enhancement as an average effect
of electric field enhancement, meaning that the enhancement factor should be equal
to the integral of the fourth power of the electric field intensity over the whole
volume of dielectric material. Thus, the photocatalytic enhancement is given by:
enhancement factor =
∮
|𝐸 |
4

𝑣𝑜𝑙𝑢𝑚𝑒 𝑜𝑓 𝑇𝑖 𝑂 2
∮
|𝐸 0
|
4

𝑣𝑜𝑙𝑢𝑚𝑒 𝑜𝑓 𝑇𝑖 𝑂 2

100

Here, E is the electric filed inside TiO2, E0 is the electric field of incident light.

Figure 6.2 a) Schematic of photocatalysis experiment set up. b, c, d) UV–Vis
spectra of MO aqueous solution before (black) and after (dash) visible light
exposure using 2nm TiO2 film, isolated nanofingers and collapsed nanofingers as
photocatalyst.

101

6.3 Numerical Calculations
In order to calculate the electromagnetic field enhancement, Finite-difference
time-domain (FDTD) are utilized to simulate the electric field distribution under
532nm of isolated and collapsed nanofingers respectively. Considering that
electrons has low possibility to tunnel through a 4nm TiO2 film, we do not consider
quantum effect in our calculation.
First, we perform FDTD numerical simulation of the extinction spectra of the
single nanofingers coated with 2nm TiO2 shell. A polymer cylinder capped with a
50nm gold nanoparticle stand vertically on an infinite SiO2 mold. The diameter and
height of the cylinder is 70nm and 300nm respectively. This nanofinger is coated
with an ultrathin TiO2 film of 2nm thickness. The refractive index of TiO2 is
interpolated from the data measured by an ellipsometer, and the refractive index of
UV NIL resist is set as 1.35. Figure 6.3a shows that the extinction spectrum of
isolated nanoparticle coated 2nm TiO2 film has a peak at ~530nm. The optical
response of the single nanoparticle is not significantly affected by the dielectric
coating layer, except for a slight spectral redshift of the resonant peaks, which
corresponds to the refractive index variation introduced by TiO2. Since the distance
between two isolated nanofingers is relative large, the interaction between different
nanofingers is fairly weak. The charge distribution is dominated by dipolar plasmon
mode, which can be observed in the insert image in Figure 6.3a. The electric filed
intensity produced by the isolated nanofingers can only reach up to 5 leads to a low
102

enhancement factor as expected. As the four nearby nanofingers approach and
touch each other, the single dipole mode will strongly interact with each other to
form bonding dimer plasmon (BDP), which can concentrate energy into extremely
small volume at the gap. The electric field distribution around the gap is shown in
Figure 6.3b. Compared with isolated gold nanoparticle, gap nanostructure can
provide much stronger local field enhancement at the hot spot, up to 10
6
. Although
the field intensity at the hot spot is extremely large, the average enhancement factor
is limited by the small volume of the hot spot compared with whole TiO2 film. We
can calculate the enhancement factor using the equation we mentioned. They are
50X and for isolated and collapsed nanofingers

respectively.

103

Figure 6.3 a) Extinction spectra of isolated nanofingers without coating and with
2nm TiO2 coating, the inset is the electric field distribution. b) Electric field
distribution at the gap.

104

Chapter 7. Conclusion and Future Works
In this thesis, the story of nano-fingers from fabrication to application is
summarized and covered with technological details. We created a technology to
fabricate down to sub-nanometer gap plasmonic structures with high precision,
high reliability and high throughput deterministically. Precisely controlled nano-
gap arrays down to sub-nanometer in large area can be produced using collapsible
nano-fingers. We experimentally proved and observed optimal gap sizes for gap
plasmonic structures with different dielectric spacers. Meanwhile, we also
demonstrated direct control of tunneling strength by simply changing the material
between gold nanoparticles and showed that lower tunneling barrier heights result
in larger optimal gap sizes. This technology enables us to put active materials at the
hottest part of optimally designed gap plasmonic hot spot, which is a great platform
for various applications including plasmon enhanced SERS and fluorescence.

The application of nano-fingers in plasmon enhanced Raman spectroscopy is
demonstrated. The multitude of benefits deriving from the unique nano-finger
nanostructure allows extremely high detection sensitivity at the single-molecular
level to be realized as demonstrated through bi-analyte surface-enhanced Raman
scattering measurement. The real-time Raman detection for big living yeast cells is
also demonstrated, which suggests a promising emerging technology for single-
105

molecular label-free sensing and opens the door to a wide range of opportunities
disease diagnosis and biological content identification.

In addition, plasmon enhanced fluorescence is studied using the collapsible nano-
fingers. In this part, we studied the mechanism of all contributing factors in plasmon
enhanced fluorescence (PEF) and propose a guiding rule for further PEF related
design. The quenching mechanism in PEF is analyzed experimentally and
theoretically. The gap plasmon induced PEF is studied in the sub-5 nm gap scale
with sub-nanometer accuracy experimentally. SERS signal is obtained
simultaneous in PEF measurement, which offers in-situ monitor on plasmonic field
enhancement. By using collapsible nano-fingers, we realized ultra-high
enhancement factor over large area uniformly with stable and constant fluorescence.
That paves the way for the applications of PEF in chemical sensing and biological
labeling with high sensitivity requirements.

In addition, by using large-area collapsible nanofingers, we can combine noble
metal nanoparticles and thin TiO2 films to create a novel plasmonic photocatalyst.
The metal nanoparticle pairs can couple the energy of incident light into ultra-small
region inside the TiO2 film and then the energy can be dissipated to accelerate
chemical reaction by creating electron hole pairs. Finite-difference time-domain
simulations of this plasmonic photocatalyst shows that the enhanced photocatalytic
106

activity is due to the large enhancement of the local electric field at the hot spots,
which increases the electron–hole pair generation rate at the TiO2 surface, and
hence the photodecomposition rate of methyl orange. The mechanism of the
enhancement is discussed and theoretical enhancement factor is calculated.

There are still many fascinating properties of nano-fingers, which could enable
many further photonic and electronic applications. In the meantime, as a novel
plasmonic nanostructure, nano-fingers present many subtle physics behind the
experimental results. Single molecule detection (SMD) and mapping, one-step
fluorescence spectra detection (time resolved fluorescence spectra) is an example.
To demonstrate the superiority of our optimized nanofingers substrates in PEF,
similar fluorescence spectrum mapping at ultra-low concentration fluorophore
should be performed to realize its single molecule level detection with high
accuracy and sensitivity. Also, the single molecule enhancement factor in PEF
could also be characterized rather than an averaged value over a macro-dimension.
In addition, photobleaching effect can act as a helpful tool for single molecule
detection in PEF. Usually, an ultra-low analyte concentration down to the nanomole
level is necessary to perform SMD experiments. However, it is difficult to detect at
this condition in practice. But with enhancement large enough, the one-step
photobleaching in fluorescence spectrum should be detected in time-resolved
107

fluorescence spectrometry. This is also a necessary characterization of
photostability for PEF in nanofinger substrates.

Understanding the photobleaching mechanism of PEF in nanofingers is also
important. The physical approach to harvest more photons from fluorescent
molecules is based on enhancing the radiative decay rate of the fluorophore kr. As
the molecule spends shorter times in the excited state before the emission of a
photon, it can perform more excitation−emission cycles before undergoing
photobleaching. The way to physically enhance kr is based on Fermi’s “golden rule”
𝑘 𝑟 = [
〈𝑖 |𝐻 |𝑓 〉
2
ħ
2
] 𝜌 (𝜐 ). The state ⟨i| corresponds to the excited molecule in the absence
of any photon, ⟨f | is the final state of the relaxed molecule and a single photon, H
is the molecule-field interaction Hamiltonian. The key factor is ρ(ν), the density of
photon states at the emission frequency ν. A higher photonic mode density (PMD)
of the right frequency and polarization facilitates a faster radiative decay of the
excited fluorophore. Remarkably, most investigations on plasmonic enhancement
of molecular fluorescence have focused on enhancing intensity and the emission
directionality. Despite its fundamental practical significance, the possibility of
increasing photostability of fluorophores by means of an enhanced PMD has
remained rather unexplored. While a shorter distance between a fluorophore and
metallic nanoparticles offers stronger electromagnetic coupling thus higher PMD,
it also facilitates better energy transfer that further reduce photobleaching. However,
108

quenching effect increases significantly as the distance between fluorophore and
metallic nanoparticles. To solve this dilemma, both simulation and experiments are
required. The geometry, such as the diameter of an individual nanofinger, is the
parameter we can tune that contributes to reduce photobleaching without quenching
the fluorescence intensity.

All in all, I believe there are many more fundamental questions to be answered
regarding the property of nano-fingers, just like there are many exciting discoveries
to be made in this class of novel nanostructures. Joint expertise from different
research communities, such as ultrafast spectroscopy, epitaxial thin film growth,
biological characterization and circuit design could further promote the
understanding and applications in these novel plasmonic nanostructures.

109

REFERENCES
1. Linic, S.; Christopher, P.; Ingram, D. B., Plasmonic-Metal Nanostructures
for Efficient Conversion of Solar to Chemical Energy. Nat. Mater. 2011, 10, 911-
921.
2. Atwater, H. A.; Polman, A., Plasmonics for Improved Photovoltaic Devices.
Nat. Mater. 2010, 9, 205-213.
3. Mü hlschlegel, P.; Eisler, H.-J.; Martin, O.; Hecht, B.; Pohl, D., Resonant
Optical Antennas. Science 2005, 308, 1607-1609.
4. Novotny, L.; Van Hulst, N., Antennas for Light. Nat. Photonics 2011, 5,
83-90.
5. Giannini, V.; Ferná ndez-Domí nguez, A. I.; Heck, S. C.; Maier, S. A.,
Plasmonic Nanoantennas: Fundamentals and Their Use in Controlling the
Radiative Properties of Nanoemitters. Chem. Rev. 2011, 111, 3888-3912.
6. Alvarez-Puebla, R.; Liz-Marzá n, L. M.; Garcí a de Abajo, F. J., Light
Concentration at the Nanometer Scale. J. Phys. Chem. Lett. 2010, 1, 2428-2434.
7. Ansell, D.; Radko, I.; Han, Z.; Rodriguez, F.; Bozhevolnyi, S.; Grigorenko,
A., Hybrid Graphene Plasmonic Waveguide Modulators. Nat. Commun. 2015, 6.
8. Song, J.; Huang, P.; Duan, H.; Chen, X., Plasmonic Vesicles of Amphiphilic
Nanocrystals: Optically Active Multifunctional Platform for Cancer Diagnosis and
Therapy. Acc. Chem. Res. 2015, 48, 2506-2515.
110

9. Stewart, M. E.; Anderton, C. R.; Thompson, L. B.; Maria, J.; Gray, S. K.;
Rogers, J. A.; Nuzzo, R. G., Nanostructured Plasmonic Sensors. Chem. Rev. 2008,
108, 494-521.
10. Schuller, J. A.; Barnard, E. S.; Cai, W.; Jun, Y. C.; White, J. S.; Brongersma,
M. L., Plasmonics for Extreme Light Concentration and Manipulation. Nat. Mater.
2010, 9, 193-204.
11. Halas, N. J.; Lal, S.; Chang, W.-S.; Link, S.; Nordlander, P., Plasmons in
Strongly Coupled Metallic Nanostructures. Chem. Rev. 2011, 111, 3913-3961.
12. Pasquale, A. J.; Reinhard, B. r. M.; Dal Negro, L., Engineering Photonic–
Plasmonic Coupling in Metal Nanoparticle Necklaces. Acs Nano 2011, 5, 6578-
6585.
13. Kelly, K. L.; Coronado, E.; Zhao, L. L.; Schatz, G. C., The Optical
Properties of Metal Nanoparticles: The Influence of Size, Shape, and Dielectric
Environment. J. Phys. Chem. B 2003, 107, 668-677.
14. Marinica, D. C.; Kazansky, A. K.; Nordlander, P.; Aizpurua, J.; Borisov, A.
G., Quantum Plasmonics: Nonlinear Effects in the Field Enhancement of a
Plasmonic Nanoparticle Dimer. Nano Lett. 2012, 12, 1333-1339.
15. Tame, M. S.; McEnery, K.; Özdemir, Ş.; Lee, J.; Maier, S.; Kim, M.,
Quantum Plasmonics. Nat. Phys. 2013, 9, 329-340.
111

16. Esteban, R.; Borisov, A. G.; Nordlander, P.; Aizpurua, J., Bridging
Quantum and Classical Plasmonics with a Quantum-Corrected Model. Nat.
Commun. 2012, 3, 825.
17. Nordlander, P.; Oubre, C.; Prodan, E.; Li, K.; Stockman, M., Plasmon
Hybridization in Nanoparticle Dimers. Nano Lett. 2004, 4, 899-903.
18. Garcí a-Martí n, A.; Ward, D. R.; Natelson, D.; Cuevas, J. C., Field
Enhancement in Subnanometer Metallic Gaps. Phys. Rev. B 2011, 83, 193404.
19. Savage, K. J.; Hawkeye, M. M.; Esteban, R.; Borisov, A. G.; Aizpurua, J.;
Baumberg, J. J., Revealing the Quantum Regime in Tunnelling Plasmonics. Nature
2012, 491, 574-577.
20. Zhu, W.; Crozier, K. B., Quantum Mechanical Limit to Plasmonic
Enhancement as Observed by Surface-Enhanced Raman Scattering. Nat. Commun.
2014, 5.
21. Yang, L.; Wang, H.; Fang, Y.; Li, Z., Polarization State of Light Scattered
from Quantum Plasmonic Dimer Antennas. ACS Nano 2016, 10, 1580-1588.
22. Chen, Y.; Bi, K.; Wang, Q.; Zheng, M.; Liu, Q.; Han, Y.; Yang, J.; Chang,
S.; Zhang, G.; Duan, H., Rapid Focused Ion Beam Milling Based Fabrication of
Plasmonic Nanoparticles and Assemblies Via “Sketch and Peel” Strategy. ACS
nano 2016, 10, 11228-11236.
23. Forati, E.; Dill, T. J.; Tao, A. R.; Sievenpiper, D., Photoemission-Based
Microelectronic Devices. Nat. Commun. 2016, 7.
112

24. Duan, H.; Hu, H.; Kumar, K.; Shen, Z.; Yang, J. K., Direct and Reliable
Patterning of Plasmonic Nanostructures with Sub-10-Nm Gaps. ACS nano 2011, 5,
7593-7600.
25. Thacker, V. V.; Herrmann, L. O.; Sigle, D. O.; Zhang, T.; Liedl, T.;
Baumberg, J. J.; Keyser, U. F., DNA Origami Based Assembly of Gold
Nanoparticle Dimers for Surface-Enhanced Raman Scattering. Nat. Commun. 2014,
5, 3448.
26. George, S. M., Atomic Layer Deposition: An Overview. Chem. Rev. 2009,
110, 111-131.
27. Elam, J.; Routkevitch, D.; Mardilovich, P.; George, S., Conformal Coating
on Ultrahigh-Aspect-Ratio Nanopores of Anodic Alumina by Atomic Layer
Deposition. Chem. Mater. 2003, 15, 3507-3517.
28. Standridge, S. D.; Schatz, G. C.; Hupp, J. T., Toward Plasmonic Solar Cells:
Protection of Silver Nanoparticles Via Atomic Layer Deposition of Tio2. Langmuir
2009, 25, 2596-2600.
29. Tan, S. F.; Wu, L.; Yang, J. K.; Bai, P.; Bosman, M.; Nijhuis, C. A.,
Quantum Plasmon Resonances Controlled by Molecular Tunnel Junctions. Science
2014, 343, 1496-1499.
30. Liu, H.; Yao, Y.; Wang, Y.; Wu, W., Full-Color Reflective Display System
Based on High Contrast Gratings. J. Vac. Sci. Technol., B: Nanotechnol.
Microelectron.: Mater., Process., Meas., Phenom. 2014, 32, 06FE04.
113

31. Savas, T.; Schattenburg, M.; Carter, J.; Smith, H. I., Large ‐ Area
Achromatic Interferometric Lithography for 100 Nm Period Gratings and Grids. J.
Vac. Sci. Technol., B: Microelectron. Nanometer Struct.--Process., Meas., Phenom.
1996, 14, 4167-4170.
32. Yao, Y.; Wang, Y.; Liu, H.; Li, Y.; Song, B.; Wu, W., Line Width Tuning
and Smoothening for Periodical Grating Fabrication in Nanoimprint Lithography.
Appl. Phys. A 2015, 121, 399-403.
33. Chou, S. Y.; Krauss, P. R.; Renstrom, P. J., Nanoimprint Lithography. J.
Vac. Sci. Technol., B: Microelectron. Nanometer Struct.--Process., Meas., Phenom.
1996, 14, 4129-4133.
34. Yao, Y.; Liu, H.; Wang, Y.; Li, Y.; Song, B.; Bratkovsk, A.; Wang, S.-Y.;
Wu, W., Nanoimprint Lithography: An Enabling Technology for Nanophotonics.
Appl. Phys. A 2015, 121, 327-333.
35. Gates, B. D.; Xu, Q.; Stewart, M.; Ryan, D.; Willson, C. G.; Whitesides, G.
M., New Approaches to Nanofabrication: Molding, Printing, and Other Techniques.
Chem. Rev. 2005, 105, 1171-1196.
36. Colburn, M.; Johnson, S. C.; Stewart, M. D.; Damle, S.; Bailey, T. C.; Choi,
B.; Wedlake, M.; Michaelson, T. B.; Sreenivasan, S.; Ekerdt, J. G. In Step and
Flash Imprint Lithography: A New Approach to High-Resolution Patterning,
Microlithography'99, International Society for Optics and Photonics: 1999; pp 379-
389.
114

37. Ou, F. S.; Hu, M.; Naumov, I.; Kim, A.; Wu, W.; Bratkovsky, A. M.; Li,
X.; Williams, R. S.; Li, Z., Hot-Spot Engineering in Polygonal Nanofinger
Assemblies for Surface Enhanced Raman Spectroscopy. Nano Lett. 2011, 11, 2538-
2542.
38. Chandra, D.; Yang, S., Capillary-Force-Induced Clustering of Micropillar
Arrays: Is It Caused by Isolated Capillary Bridges or by the Lateral Capillary
Meniscus Interaction Force? Langmuir 2009, 25, 10430-10434.
39. Kotera, M.; Ochiai, N., Three-Dimensional Simulation of Resist Pattern
Deformation by Surface Tension at the Drying Process. Microelectron. Eng. 2005,
78, 515-520.
40. Chini, S. F.; Amirfazli, A., Understanding Pattern Collapse in
Photolithography Process Due to Capillary Forces. Langmuir 2010, 26, 13707-
13714.
41. Duan, H.; Berggren, K. K., Directed Self-Assembly at the 10 Nm Scale by
Using Capillary Force-Induced Nanocohesion. Nano Lett. 2010, 10, 3710-3716.
42. Zhang, R.; Zhang, Y.; Dong, Z.; Jiang, S.; Zhang, C.; Chen, L.; Zhang, L.;
Liao, Y.; Aizpurua, J.; Luo, Y. e., Chemical Mapping of a Single Molecule by
Plasmon-Enhanced Raman Scattering. Nature 2013, 498, 82-86.
43. Li, J. F.; Huang, Y. F.; Ding, Y.; Yang, Z. L.; Li, S. B.; Zhou, X. S.; Fan, F.
R.; Zhang, W.; Zhou, Z. Y.; Ren, B., Shell-Isolated Nanoparticle-Enhanced Raman
Spectroscopy. nature 2010, 464, 392-395.
115

44. Zhu, Z.; Yan, Z.; Zhan, P.; Wang, Z., Large-Area Surface-Enhanced Raman
Scattering-Active Substrates Fabricated by Femtosecond Laser Ablation. Sci.
China: Phys., Mech. Astron. 2013, 56, 1806-1809.
45. Schnell, M.; Garcia-Etxarri, A.; Huber, A.; Crozier, K.; Aizpurua, J.;
Hillenbrand, R., Controlling the near-Field Oscillations of Loaded Plasmonic
Nanoantennas. Nat. Photonics 2009, 3, 287-291.
46. Su, K.-H.; Wei, Q.-H.; Zhang, X.; Mock, J.; Smith, D. R.; Schultz, S.,
Interparticle Coupling Effects on Plasmon Resonances of Nanogold Particles. Nano
Lett. 2003, 3, 1087-1090.
47. Lim, D.-K.; Jeon, K.-S.; Kim, H. M.; Nam, J.-M.; Suh, Y. D., Nanogap-
Engineerable Raman-Active Nanodumbbells for Single-Molecule Detection. Nat.
Mater. 2010, 9, 60-67.
48. Lim, D.-K.; Jeon, K.-S.; Hwang, J.-H.; Kim, H.; Kwon, S.; Suh, Y. D.; Nam,
J.-M., Highly Uniform and Reproducible Surface-Enhanced Raman Scattering
from DNA-Tailorable Nanoparticles with 1-Nm Interior Gap. Nat. Nanotechnol.
2011, 6, 452-460.
49. Oh, J.-W.; Lim, D.-K.; Kim, G.-H.; Suh, Y. D.; Nam, J.-M., Thiolated
DNA-Based Chemistry and Control in the Structure and Optical Properties of
Plasmonic Nanoparticles with Ultrasmall Interior Nanogap. J. Am. Chem. Soc.
2014, 136, 14052-14059.
116

50. Jain, P. K.; Huang, W.; El-Sayed, M. A., On the Universal Scaling Behavior
of the Distance Decay of Plasmon Coupling in Metal Nanoparticle Pairs: A
Plasmon Ruler Equation. Nano Lett. 2007, 7, 2080-2088.
51. Romero, I.; Aizpurua, J.; Bryant, G. W.; De Abajo, F. J. G., Plasmons in
Nearly Touching Metallic Nanoparticles: Singular Response in the Limit of
Touching Dimers. Opt. Express 2006, 14, 9988-9999.
52. McMahon, J. M.; Gray, S. K.; Schatz, G. C., Optical Properties of Nanowire
Dimers with a Spatially Nonlocal Dielectric Function. Nano Lett. 2010, 10, 3473-
3481.
53. David, C.; Garcí a de Abajo, F. J., Spatial Nonlocality in the Optical
Response of Metal Nanoparticles. J. Phys. Chem. C 2011, 115, 19470-19475.
54. Teperik, T. V.; Nordlander, P.; Aizpurua, J.; Borisov, A. G., Quantum
Effects and Nonlocality in Strongly Coupled Plasmonic Nanowire Dimers. Opt.
Express 2013, 21, 27306-27325.
55. Garcí a de Abajo, F. J., Nonlocal Effects in the Plasmons of Strongly
Interacting Nanoparticles, Dimers, and Waveguides. J. Phys. Chem. C 2008, 112,
17983-17987.
56. Dong, T.; Ma, X.; Mittra, R., Optical Response in Subnanometer Gaps Due
to Nonlocal Response and Quantum Tunneling. Appl. Phys. Lett. 2012, 101,
233111.
117

57. Toscano, G.; Raza, S.; Xiao, S.; Wubs, M.; Jauho, A.-P.; Bozhevolnyi, S.
I.; Mortensen, N. A., Surface-Enhanced Raman Spectroscopy: Nonlocal
Limitations. Opt. Lett. 2012, 37, 2538-2540.
58. Teperik, T. V.; Nordlander, P.; Aizpurua, J.; Borisov, A. G., Robust
Subnanometric Plasmon Ruler by Rescaling of the Nonlocal Optical Response.
Phys. Rev. Lett. 2013, 110, 263901.
59. Wu, H.; Wang, L. S., Electronic Structure of Titanium Oxide Clusters: Tio
Y (Y= 1− 3) and (Tio 2) N (N= 1− 4). J. Chem. Phys. 1997, 107, 8221-8228.
60. Walter, C.; Hertzler, C.; Devynck, P.; Smith, G.; Peterson, J.,
Photodetachment of Wo− 3: The Electron Affinity of Wo3. J. Chem. Phys. 1991,
95, 824-827.
61. Pantelides, S. T., The Physics of Sio2 and Its Interfaces: Proceedings of the
International Topical Conference on the Physics of Sio2 and Its Interfaces Held at
the Ibm Thomas J. Waston Research Center, Yorktown Heights, New York, March
22-24, 1978. Elsevier: 2013.
62. Link, S.; El-Sayed, M. A., Size and Temperature Dependence of the
Plasmon Absorption of Colloidal Gold Nanoparticles. J. Phys. Chem. B 1999, 103,
4212-4217.
63. Link, S.; El-Sayed, M. A., Shape and Size Dependence of Radiative, Non-
Radiative and Photothermal Properties of Gold Nanocrystals. Int. Rev. Phys. Chem.
2000, 19, 409-453.
118

64. Born, M.; Wolf, E., Principles of Optics: Electromagnetic Theory of
Propagation, Interference and Diffaction of Light. Pergamon Press: 1959.
65. Fan, X.; Zheng, W.; Singh, D. J., Light Scattering and Surface Plasmons on
Small Spherical Particles. Light Sci Appl 2014, 3, e179.
66. Garcí a de Abajo, F. J., Optical Excitations in Electron Microscopy. Rev.
Mod. Phys. 2010, 82, 209-275.
67. Nordlander, P.; Oubre, C.; Prodan, E.; Li, K.; Stockman, M. I., Plasmon
Hybridization in Nanoparticle Dimers. Nano Lett. 2004, 4, 899-903.
68. Lassiter, J. B.; Aizpurua, J.; Hernandez, L. I.; Brandl, D. W.; Romero, I.;
Lal, S.; Hafner, J. H.; Nordlander, P.; Halas, N. J., Close Encounters between Two
Nanoshells. Nano Lett. 2008, 8, 1212-1218.
69. Zuloaga, J.; Prodan, E.; Nordlander, P., Quantum Description of the
Plasmon Resonances of a Nanoparticle Dimer. Nano Lett. 2009, 9, 887-891.
70. Lebedev, V.; Vitukhnovsky, A.; Yoshida, A.; Kometani, N.; Yonezawa, Y.,
Absorption Properties of the Composite Silver/Dye Nanoparticles in Colloidal
Solutions. Colloids Surf., A 2008, 326, 204-209.
71. Quinsaat, J. E. Q.; Nü esch, F. A.; Hofmann, H.; Opris, D. M., Dielectric
Properties of Silver Nanoparticles Coated with Silica Shells of Different
Thicknesses. RSC Adv. 2013, 3, 6964-6971.
119

72. Cheng, P.; Bao, J.; Wu, L.; Li, X.; Zhao, H.; Zhu, R.; Wang, J.; Li, D.,
Surface Plasmon Response of Metal Spherical Nanoshells Coated with Dielectric
Overlayer. Chem. Phys. Lett. 2013, 587, 40-44.
73. Weisskopf, V.; Wigner, E. P., Calculation of the Natural Brightness of
Spectral Lines on the Basis of Dirac's Theory. Z. Phys. 1930, 63, 54-73.
74. Scully, M. O.; Zubairy, M. S., Quantum Optics. American Association of
Physics Teachers: 1999.
75. Pelton, M., Modified Spontaneous Emission in Nanophotonic Structures.
Nat. Photonics 2015, 9, 427.
76. Jackson, J. D., Classical Electrodynamics. American Association of Physics
Teachers: 1999.
77. Song, B.; Yao, Y.; Groenewald, R. E.; Wang, Y.; Liu, H.; Wang, Y.; Li, Y.;
Liu, F.; Cronin, S. B.; Schwartzberg, A. M., Probing Gap Plasmons Down to
Subnanometer Scales Using Collapsible Nanofingers. ACS nano 2017, 11, 5836-
5843.

Abstract (if available)

Abstract Gap plasmonic nanostructures are of great interest due to their ability to concentrate light into small volumes, which enables the creation of a highly concentrated electromagnetic (EM) field within nanoscale volumes that can enable the ultra-high sensitivity chemical and biological detection. Theoretical studies, considering quantum mechanical effects, have predicted the optimal spatial gap between adjacent nanoparticles to be in the sub-nanometer regime in order to achieve the strongest possible field enhancement. While many attempts have been made in recent years to fabricate gap plasmonic nanostructures with sub-5 nm gaps, the limitation of traditional fabrication techniques poses serious obstacles to reliably producing precisely controllable nano-gaps with high throughput. ❧ The first part of this dissertation mainly focuses on the realization of the theoretically predicted gap plasmonic structure. A technology is proposed to fabricate gap plasmonic structures with sub-nanometer resolution, high reliability and high throughput using collapsible nano-fingers. Different fabrication technological details are discussed and compared for nano-fingers collapse in various pattern. The capillary force driven collapse mechanism is studied for further fabrication designs. Atomic layer deposition process, as the key role in determining the gap size, is optimized for the performances of nano-fingers. ❧ This approach enables us to systematically investigate the quantum effects in gap plasmon. We experimentally proved and observed optimal gap sizes for gap plasmonic structures with different dielectric spacers. The effect of tunneling barrier height is studied by using different dielectric spacers, and the quantum nonlocal effect is also observed at the meantime. A straightforward theoretical model is presented to analyze the quantum effects in collapsible nano-fingers. ❧ The application of nano-fingers in plasmon enhanced Raman spectroscopy is demonstrated in the following. The multitude of benefits deriving from the unique nano-finger nanostructure allows extremely high detection sensitivity at the single-molecular level to be realized as demonstrated through bi-analyte surface-enhanced Raman scattering measurement. The real-time Raman detection for big living yeast cells is also demonstrated, which suggests a promising emerging technology for single-molecular label-free sensing and opens the door to a wide range of opportunities disease diagnosis and biological content identification. ❧ In addition, plasmon enhanced fluorescence is studied using the collapsible nano-fingers. In this part, we studied the mechanism of all contributing factors in plasmon enhanced fluorescence (PEF) and propose a guiding rule for further PEF related design. The quenching mechanism in PEF is analyzed experimentally and theoretically. The gap plasmon induced PEF is studied in the sub-5 nm gap scale with sub-nanometer accuracy experimentally. SERS signal is obtained simultaneous in PEF measurement, which offers in-situ monitor on plasmonic field enhancement. By using collapsible nano-fingers, we realized ultra-high enhancement factor over large area uniformly with stable and constant fluorescence. That paves the way for the applications of PEF in chemical sensing and biological labeling with high sensitivity requirements. ❧ By using large-area collapsible nanofingers, we can combine noble metal nanoparticles and thin TiO₂ films to create a novel plasmonic photocatalyst. The metal nanoparticle pairs can couple the energy of incident light into ultra-small region inside the TiO₂ film and then the energy can be dissipated to accelerate chemical reaction by creating electron hole pairs. Finite-difference time-domain simulations of this plasmonic photocatalyst shows that the enhanced photocatalytic activity is due to the large enhancement of the local electric field at the hot spots, which increases the electron–hole pair generation rate at the TiO₂ surface, and hence the photodecomposition rate of methyl orange. The mechanism of the enhancement is discussed, and theoretical enhancement factor is calculated.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Electromagnetic modeling of plasmonic nanostructures

PDF

Applications enabled by plasmonic nano-finger and analog computing based on memristive devices

PDF

Design and characterization of metal and semiconducting nanostructures and nanodevices

PDF

Surface plasmon resonant enhancement of photocatalysis

PDF

Plasmonic enhancement of catalysis and solar energy conversion

PDF

Hot carriers in bare metals and photocatalytically active defect sites in dielectric/metal structures

PDF

Nano-fabricated devices in electrochemistry and cancer therapy

PDF

Integrated photonics assisted electron emission devices

PDF

Hot electron driven photocatalysis and plasmonic sensing using metallic gratings

PDF

Optimizing nanoemitters using quasi-aperiodicity and photoelectrochemical etching

PDF

Electronic and optoelectronic devices based on quasi-metallic carbon nanotubes

PDF

Transient pulsed plasma for pollution remediation and energy conversion applications

PDF

Plasmonic excitations in nanostructures

PDF

Light Emission from Carbon Nanotubes and Two-Dimensional Materials

PDF

Hot carrier enhanced photocatalysis in plasmon resonant metal grating systems

PDF

Artificial photosynthesis on titanium oxide passivated III-V semiconductors

PDF

Designable nonlinear power shaping photonic surfaces

PDF

Nano-engineered devices for optical application, biomedical detection and circuit accelerator

PDF

Nanophotonic light management in thin film silicon photovoltaics

PDF

High-throughput nanoparticle fabrication and nano-biomembrane interactions

Asset Metadata

Creator Song, Boxiang (author)

Core Title Fabrication and application of plasmonic nanostructures: a story of nano-fingers

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Electrical Engineering

Publication Date 03/30/2020

Defense Date 12/03/2019

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag gap plasmon,nanofabrication,nano-fingers,nanoimprint lithography,OAI-PMH Harvest,plasmonics,quantum effects

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Wu, Wei (committee chair), Cronin, Stephen (committee member), Haas, Stephan (committee member)

Creator Email boxiangs@usc.edu,songboxiang89@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-277686

Unique identifier UC11673841

Identifier etd-SongBoxian-8233.pdf (filename),usctheses-c89-277686 (legacy record id)

Legacy Identifier etd-SongBoxian-8233.pdf

Dmrecord 277686

Document Type Dissertation

Rights Song, Boxiang

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