Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Hot carriers in bare metals and photocatalytically active defect sites in dielectric/metal structures
(USC Thesis Other)
Hot carriers in bare metals and photocatalytically active defect sites in dielectric/metal structures
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
HOT CARRIERS IN BARE METALS
AND
PHOTOCATALYTICALLY ACTIVE DEFECT SITES IN
DIELECTRIC/METAL STRUCTURES
By
Bingya Hou
A Dissertation Presented to
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of
the Requirement for the Degree of
DOCTOR OF PHILOSOPHY
(ELECTRICAL ENGINEERING)
May 2019
i
Acknowledgement
First of all, I would like to thank my advisor, Dr. Stephen B. Cronin, for his invaluable
inspiration and guidance. I sincerely appreciate his dedication, hard work and trust. I become more
professional in the scientific research field because of him.
My sincere thanks also go to Dr. Wei Wu, Dr. Aiichiro Nakano, Dr. Han Wang and Dr.
Michelle Povinelli for being my dissertation committee and qualifying exam committee members.
I would like to thank Dr. Kian Kaviani for offering me a position as an EE504L teaching
assistant.
It is quite lucky for me to overlap and work with some senior Ph.Ds from Cronin research
group, and they are Dr. Jing Qiu, Dr. Rohan Dhall and Dr. Chun-Chung Chen. Their inspiring
advise and careful instructions have saved my Ph.D. life.
Mr. Haotian Shi and Mr. Jihan Chen joint Cronin research group almost at the same time as
me. I am grateful for their encouragement and discussion. They make my Ph.D. life more
interesting.
As a Ph.D. student, no one can achieve his/her accomplishment without support from fellows,
so I would like to thank all the graduated and current members from Cronin research group: Dr.
Nirakar Poudel, Dr. Lang Shen, Dr. Guangtong Zeng, Dr. Shun-Wen Chang, Dr. Wei-Hsuan Hung,
Dr. Mehmet Aykol, Dr. Zhen Li, Dr. Moh Amer, Ms. Fanqi Wu (Zhou group), Ms. Sisi Yang, Mr.
Bo Wang, Mr. Yu Wang, Mr. Bofan Zhao, Mr. Zhi Cai, Ms. Indu A A.
ii
I also feel very thankful for having lots of friends all over the world that we could share
happiness and sorrow, and they are Ms. Ming Liu, Ms. Xuanji Xue, Ms. Qi Luan, Ms. Shuying
Zhang, Ms. Yizhe Zhang, Ms. Yuxin Ji, Dr. Yong Zheng and Dr. Sha Luo.
I take pride in dedicating this dissertation to my beloved parents. I could not survive without
their endless love and constant support.
Last but not least, I would like to thank my husband, Dr. Anyi Zhang, for being my lifelong
companion and making me a better me.
iii
Abstract
Hot electrons have been reported in photocatalytic systems, introducing the exciting possibility
of overcoming high barrier reactions. Hot electron processes have also been reported in solid state
devices. However, there are several possible mechanisms capable of producing a photocurrent,
which are difficult to separate. In the thesis presented here, bare Au surfaces and TiO2-coated Au
surfaces were studied in order to demonstrate the existence of hot electrons and holes in bare Au
surfaces and photocatalytically active defect sites in TiO2.
Chapter 1 starts with a brief introduction of photocatalytic water splitting. Some significant
advances in this field are listed later, along with the requirements for efficient and robust
photocatalysts.
In Chapter 2, we provide a basic description of plasmons and hot carriers, and comparison the
near-field electromagnetic mechanism and charge injection mechanism for photocurrent
enhancement.
In Chapter 3, an AC photocurrent from bare Au surfaces is detected for both hydrogen
evolution reaction (HER) and oxygen evolution reaction (OER) using an AC lock-in technique,
which demonstrates the existence of hot electrons and holes unambiguously. For both HER and
OER, the AC photocurrent exhibits a relatively narrow peak, as the monoenergetic distribution of
hot electrons and holes photoexcited in the metal is swept through the resonance with the redox
potential of the desired half-reaction.
The photocurrent from TiO2-coated metal surfaces is discussed in both Chapters 3 and 4. From
the photocurrent spectra, we conclude that photocatalytically active sites in TiO2 can be selectively
detected no matter which metal substrate is chosen. These active defects sites are located at 2eV
iv
below the conduction band edge, which is evidenced by ultraviolet photoemission spectroscopy
(UPS).
In Chapter 5, a systematic study of photoluminescence (PL) spectroscopy of TiO 2-passivated
GaAs as a function of electrochemical potential in an ionic liquid solution is presented. We observe
a 7X increase in the PL intensity as the GaAs transitions from accumulation to depletion due to
the applied potential. We attribute this to the excellent control over the surface Fermi level enabled
by the high capacitance of the electrochemical double layer and TiO 2.
In Chapter 6, we provide a summary of future research directions and outlook for hot-electron
and defect-mediated photocatalysis.
v
TABLE OF CONTENTS
Chapter 1 Introduction to water splitting on semiconductor photocatalysts ........................ 1
1.1 Bases of photocatalytic water splitting ............................................................................ 1
1.2 Significant advances ....................................................................................................... 2
1.3 Requirements for efficient and robust photocatalysts ...................................................... 4
Chapter 2 Introduction to plasmonic metal/semiconductor photocatalysts .......................... 6
2.1 Bases of plasmons and hot carriers .................................................................................. 6
2.2 Energy transfer mechanisms of photocatalytic reaction rate enhancement by surface
plasmon resonance (SPR) ............................................................................................................. 8
Chapter 3 Hot electron-driven photocatalytic water splitting .............................................. 13
3.1 Abstract ........................................................................................................................ 13
3.2 Introduction ................................................................................................................... 14
3.3 Experimental section ..................................................................................................... 16
3.4 Results and discussion ................................................................................................... 20
3.5 Conclusion .............................................................................................................................. 25
Chapter 4 Resonant and selective excitation of photocatalytically active defect sites in TiO 2
................................................................................................................................................. 26
4.1 Abstract ......................................................................................................................... 26
4.2 Introduction ................................................................................................................... 27
4.3 Experimental section ..................................................................................................... 28
vi
4.3 Results and discussion ................................................................................................... 30
4.4 Conclusion .............................................................................................................................. 39
Chapter 5 Prevention of surface recombination by electrochemical tuning of TiO 2-passivated
photocatalysts .......................................................................................................................... 40
5.1 Abstract ......................................................................................................................... 40
5.2 Introduction ................................................................................................................... 40
5.3 Experimental section ..................................................................................................... 42
5.3 Results and discussion ................................................................................................... 44
5.4 Conclusion .............................................................................................................................. 51
Chapter 6 Future research directions and outlook for hot-electron and defect-mediated
photocatalysis .......................................................................................................................... 52
6.1 Hot electron-driven photocatalysis on plasmon resonant structures ................................ 52
6.2 Broad band plasmon resonant absorption ....................................................................... 52
6.3 Defect-mediated photocatalysis ..................................................................................... 53
6.4 Transient absorption spectroscopy ........................................................................................ 54
Bibliography ........................................................................................................................... 56
Appendix A: Additional materials characterization of thin TiO 2 films ............................... 62
Appendix B: Details of calculations of surface recombination rate ...................................... 67
vii
LIST OF FIGURES
Chapter 1 Introduction to water splitting on semiconductor photocatalysts ........................ 1
Figure 1.1 Principle of water splitting using semiconductor photocatalysts. ......................... 1
Figure 1.2 Band edge of semiconductors plotted with redox potentials of water splitting. .... 2
Figure 1.3 Electrochemical cell in which the TiO2 electrode is connected with a platinum
electrode. .............................................................................................................................. 3
Figure 1.4 Absorption spectrum of TiO2 superimposed over the solar spectrum. .................. 4
Figure 1.5 (a) Time dependence of the photocurrent density of bare GaP illuminated with 1
W/cm
2
in a 0.5M Na2SO4 solution at an applied overpotential of -0.7V. (b) Optical microscope
image, (c) atomic force image, and (d) surface topography of bare GaP surface after 5-hour
reaction. ................................................................................................................................ 5
Chapter 2 Introduction to plasmonic metal/semiconductor photocatalysts .......................... 6
Figure 2.1 Schematic diagram of optical excitations of surface plasmons followed by decay
of hot carriers. ....................................................................................................................... 6
Figure 2.2 Energy distributions of hot carriers, P(ω, ε), generated by the decay of surface
plasmons due to phonon-assisted and direct transitions, as a function of plasmon frequency (ω)
and carrier energy (ε) in (a) Al, (b) Ag, (c) Au, and (d) Cu. ................................................... 7
Figure 2.3 Schematic illustration of resonant photon scattering mechanism. ......................... 8
Figure 2.4 (a) High resolution transmission electron micrograph of a single AuNP supported
on SiO2 matrix showing darker contrast of AuNP as compared to SiO 2 support. (b) Schematic
representation of hot electron-induced dissociation of H2 on Au. (c) Real-time detection of rate
of HD formation with laser excitation (2.4 W/cm
2
, on) and without laser excitation (0.0 W/cm
2
,
off). Due to laser heating during 10 min of laser excitation, the temperature on the sample
changes reversibly by 8 °C, as shown in the figure, from 22 to 30 °C.. .................................. 9
Figure 2.5 (a) Photocurrent of anodic TiO2 with and without Au nanoparticles irradiated with
λ = 633nm light for 22s. (b) Photocurrent of anodic TiO 2 with and without Au nanoparticles at
zero bias voltage irradiated with UV (λ = 254nm) light for 22s. (c) SEM image of a 5nm Au
island film deposited on anodic TiO 2. (d) Electric field intensity at the interface of Au-TiO2
calculated using finite-difference time-domain (FDTD) method. ......................................... 10
Figure 2.6 Schematic illustration of charge injection mechanism. ....................................... 11
Figure 2.7(a) Volume of hydrogen evolved during the photocatalytic runs using Au/TiO2 and
TiO2 as catalysts under 532 nm laser irradiation and polychromatic light λ > 400 nm. (b)
Schematic illustration of photocatalytic activity of Au/TiO2 upon excitation of Au surface
plasmon. .............................................................................................................................. 12
viii
Chapter 3 Hot electron-driven photocatalytic water splitting .............................................. 13
Figure 3.1 Schematic diagrams of the (a, c) hot electron injection and (b, d) hot hole injection
of photoexcited carriers, illustrated for the hydrogen evolution reaction (HER) and oxygen
evolution reaction (OER). .................................................................................................... 16
Figure 3.2 UV-vis absorption spectra of a 10nm thick TiO 2 film taken before and after
annealing at 450°C for 30 min in an O2 gas environment. .................................................... 17
Figure 3.3 (a) Diagram illustrating the sample configuration. (b) Schematic diagram of the
three-terminal photoelectrochemical setup with the modulated laser and AC lock-in amplifier.
............................................................................................................................................ 18
Figure 3.4 Schematic circuit diagram of the AC lock-in technique. ..................................... 19
Figure 3.5 AC photocurrent plotted as a function of the reference potential for n- and p-type
bulk Si photocatalysts in a 0.5M H2SO4 solution. ................................................................ 20
Figure 3.6 (a) DC dark current and (b, c) AC photocurrent plotted as a function of the reference
potential for the hydrogen evolution reaction (HER). (c) is a zoomed-in version of (b). ....... 22
Figure 3.7 (a) DC dark current and (b) AC photocurrent plotted as a function of the reference
potential for the oxygen evolution reaction (OER). .............................................................. 24
Chapter 4 Resonant and selective excitation of photocatalytically active defect sites in TiO 2
................................................................................................................................................. 26
Figure 4.1 (a) Diagram illustrating the sample configuration. (b) Schematic diagram of the
three-terminal photoelectrochemical setup with the modulated light and AC lock-in amplifier.
............................................................................................................................................ 30
Figure 4.2 DC (a, c) and AC (b, d) photocurrent plotted as a function of the reference potential
for the hydrogen evolution reaction (HER) for (a, b) 5nm annealed TiO 2 deposited on Au under
532nm illumination. (c, d) 5nm annealed TiO 2 deposited on Cu under 532nm illumination.
............................................................................................................................................ 31
Figure 4.3 Photocurrent obtained under (a) reduction and (b) oxidation conditions from a
TiO2/Au sample as a function of intensity of 532nm laser excitation. .................................. 32
Figure 4.4 (a) Normalized AC photocurrent spectra of 5nm TiO2 deposited on Au, Cu, and Al
films and corresponding photoluminescence spectrum of 5nm thick annealed TiO 2 deposited
on a Au film. (b) Energy band diagram of the photoexcitation mechanism of defect-mediated
photocatalysis. ..................................................................................................................... 33
Figure 4.5 Valence band spectra of a 5nm TiO2 film obtained by UPS at 21.2eV photon energy
plotted on (a) linear and (b) log scales. ................................................................................ 34
Figure 4.6 Ti 2p core level XPS spectrum of 5nm TiO2 deposited on Au. ........................... 35
ix
Figure 4.7 High resolution TEM images of TiO2 deposited with (a) 100 cycles and (b) 500
cycles on Au. (c) and (d) are zoom-in images of (a) and (b), respectively. ........................... 36
Figure 4.8 (a) The orthorhombic cell used to model the anatase surface. Nine different oxygen
vacancies were explored with the surface oxygen vacancies, VO1-O3, displaying visible spectral
signatures shifted away from the Fermi energy. (b) Projected density of states can depict the
localization of electrons in the band gap. These spectral signatures are similar for V O1 and VO2
since the electrons localize on surface Ti atoms. In contrast, electrons are more delocalized for
VO3 across farther neighboring Ti atoms. ............................................................................. 38
Chapter 5 Prevention of surface recombination by electrochemical tuning of TiO 2-passivated
photocatalysts .......................................................................................................................... 40
Figure 5.1 Schematic diagram of the three-terminal photoelectrochemical cell. The water
immersion lens is mounted on a microscope in the spectrometer for photoluminescence
measurements. ..................................................................................................................... 44
Figure 5.2 (a) The photoluminescence intensity and (b) Mott-Schottky (1/C
2
vs. V) plot of a
TiO2-passivated GaAs photocathode measured as a function of the reference potential. (c, d)
Simulated results for a GaAs photocathode over the same voltage range.............................. 45
Figure 5.3 Energy band diagrams and free carrier concentration of the semiconductor surface
under (a), (d) inversion, (b), (e) depletion, and (c), (f) accumulation conditions. .................. 47
Figure 5.4 (a) The normalized recombination rate of a GaAs semiconductor as a function of
surface potential under illumination. (b) The surface potential plotted as a function of NHE
extracted from the TCAD simulation. (c) The expected recombination rate as a function of
applied bias under illumination. ........................................................................................... 49
Chapter 6 Future research directions and outlook for hot-electron and defect-mediated
photocatalysis .......................................................................................................................... 52
Figure 6.1 Electron micrographs of (a) a silver bull’s eye and (b) a gold pyramid array. ..... 52
Figure 6.2 Schematic diagram of angle dependent photocatalytic measurement. ................. 53
Figure 6.3 (a) Ti 2p and (b) O 1s core level XPS spectra of various thicknesses of TiO 2 on
GaAs. ................................................................................................................................. 54
Figure 6.4 Transient absorption spectra of the Au grating structure taken with (a) TM-
polarized and (b) TE-polarized light. ................................................................................... 55
x
Appendix A: Additional materials characterization of thin TiO 2 films ................................ 62
Figure A.1 Frequency dependence of AC photocurrent for a TiO 2/Au sample. .................... 62
Figure A.2 Photocurrent obtained under (a) reduction and (b) oxidation conditions from a
TiO2/Au sample as a function of intensity of 532nm laser excitation. .................................. 63
Figure A.3 Photoluminescence (PL) spectra taken with (a) 532nm, (b) 633nm, and (c) 785nm
laser light. Here, we observe similar spectral features to those observed with 532nm light. In
Figure A.3(b), the PL above 1.96eV is cut off by the 633nm edge filter. In Figure A.3(c), the
PL is cut off by the 785nm edge filter. Several Raman peaks can be seen in the spectra, which
are laser wavelength dependent. .......................................................................................... 64
Figure A.4 (a) Transient absorption measurement of 5nm TiO 2 on 1mm CaF2 substrate with
285nm pump laser and white light probe. (b) Line scan of (a) integrated over all wavelengths
(440nm to 680nm). .............................................................................................................. 66
1
Chapter 1 Introduction to water splitting on semiconductor photocatalysts
1.1 Bases of photocatalytic water splitting
Photocatalysts for various chemical transformations induced by UV-vis light, including water
splitting, are almost exclusively semiconductors. In semiconductors, the energy difference
between the bottom of the conduction and the top of the valence band is referred as band gap.
When the energy of incident light is larger than band gap, electrons and holes are generated in the
conduction and valence bands, respectively. For water splitting, electrons can reduce H 2O into H2
and holes can oxidize water into O2, as shown in Figure 1.1.
Figure 1.1. Principle of water splitting using semiconductor photocatalysts.
The conduction and valence bands should align properly to efficiently transfer photo-generated
electrons and holes to water splitting reactions. The conduction band should be higher in energy
than the hydrogen evolution potential and the valence band should be lower in energy than the
oxygen evolution potential. Figure 1.2 shows the relationship between band structure of
Conduction Band
Valence Band
H
+
/H
2
O
2
/H
2
O
-
+
Potential
0V
+1.23V
2
semiconductors and redox potentials of water splitting. ZrO2, KTaO3, SrTiO3 and TiO2 are active
for water splitting when they are suitably modified with co-catalysts.
Figure 1.2. Band edge of semiconductors plotted with redox potentials of water splitting.
1
1.2 Significant advances
The electrochemical cell was initially introduced by Fujishima and Honda in 1972.
2
A
semiconducting n-type titanium dioxide (TiO2) was connected with a platinum electrode, as shown
in Figure 1.3. The cell was immersed in a pH = 4.7 aqueous electrolyte. Introduction of UV light
leads to the formation of electron-hole pairs at the TiO2 electrode. For n-type semiconductors,
energetic holes diffuse to the semiconductor/liquid interface where they participate in oxygen
evolution reaction (OER):
2𝐻 2
𝑂 + 4ℎ
+
→ 4𝐻 +
+ 𝑂 2
Energetic electrons move to the counter electrode, which is platinum in this case, to participate in
hydrogen evolution reaction (HER):
3
2𝑒 −
+ 2𝐻 +
→ 𝐻 2
The overall reaction is
2𝐻 2
𝑂 → 2𝐻 2
+ 𝑂 2
The solar energy is stored into the energy of chemical bonds.
Figure 1.3. Electrochemical cell in which the TiO2 electrode is connected with a platinum
electrode.
2
In 2012, Lee et al. fabricated nanostructured p-InP photocathodes with a TiO2 passivation layer
and a Ru cocatalyst for photochemical hydrogen production.
3
The H2 production efficiency for
these InP nanopillars reached its maximum of ~14% at an applied bias relative to a perfect anode
of ~0.5V.
An alternative design is to perform both reduction and oxidation half reactions on the interface
of the semiconductor and reacting environment. Kudo et al. demonstrated that in pure water,
4
Sr2Ta2O 7 and Sr2Ta2O 7 with similar layered perovskite structure split water into H2 and O2 under
UV irradiation without any additives.
4
These systems are usually particle-based photocatalysts and
the two half reactions occur at specially designed co-catalyst sites.
1, 5-7
1.3 Requirements for efficient and robust photocatalysts
Efficient photocatalysts need to absorb the UV-vis region of the solar spectrum. TiO2 is not a
good candidate because its wide band gap of 3.2eV limits the photo-absorption to only the UV
region of the solar spectrum.
Figure 1.4. Absorption spectrum of TiO2 superimposed over the solar spectrum.
8
Photocatalysts should be stable in the highly reactive environment. Many semiconductors,
including silicon (Si), gallium arsenide (GaAs) and gallium phosphide (GaP), are unstable when
operated under photoanodic conditions in aqueous electrolytes. Qiu et al. showed evidence of
corrosion on bare GaP surface.
9
An exponential decay was seen in the photocurrent and RMS
roughness of ±143nm was observed in atomic force microscope image.
5
Figure 1.5. (a) Time dependence of the photocurrent density of bare GaP illuminated with 1
W/cm
2
in a 0.5M Na2SO4 solution at an applied overpotential of -0.7V. (b) Optical microscope
image, (c) atomic force image, and (d) surface topography of bare GaP surface after 5-hour
reaction.
9
Hu et al. showed that TiO2 coatings (4-143nm thick) on Si, GaAs and GaP grown by atomic
layer deposition (ALD) could prevent corrosion.
10
Furthermore, TiO2 could promote hole
conduction through electronic defects.
6
Chapter 2 Introduction to plasmonic metal/semiconductor photocatalysts
Most single-component semiconductors do not satisfy all the requirements of band edge
alignment with redox potentials, efficiency, robustness and affordability discussed in last chapter.
Therefore, composite plasmonic metal/semiconductor photocatalysts have received significant
attention. In many studies, composite plasmonic metal/semiconductor photocatalysts have been
demonstrated to achieve significantly high photocatalytic reaction rate.
11-13
2.1 Bases of plasmons and hot carriers
Plasmons are collective oscillations of electrons coupling to electromagnetic fields. Electrons
on metal surfaces are excited and form propagating excitation waves when the metal surface
interacts with incident light. These electromagnetic waves are called surface plasmons.
Figure 2.1. Schematic diagram of optical excitations of surface plasmons followed by decay
to hot carriers.
14
Surface plasmons can decay non-radiatively via direct and phonon-assisted transitions by
generating excited carriers. These carriers are usually referred as hot carriers. The initial excited
carrier distribution from surface plasmon decay has been predicted by first-principles
calculations.
15
In Figure 2.2, the carrier distributions are plotted as a function of carrier energy
(horizontal axis) and plasmon/photon energy (axis normal to the page). In noble metals, direct
7
transitions have shown strong material dependence. For Au, electrons from d bands can be excited
to unoccupied states above the Fermi level. This direct transition results in “hot holes”, which are
more energetic than “hot electrons”.
Figure 2.2. Energy distributions of hot carriers, P(ω, ε), generated by the decay of surface
plasmons due to phonon-assisted and direct transitions, as a function of plasmon frequency (ω)
and carrier energy (ε) in (a) Al, (b) Ag, (c) Au, and (d) Cu.
15
Hot electrons can be used for diatomic molecular dissociation. In 2013, hot electron-induced
dissociation of H2 was reported on small Au nanoparticles (NPs) supported on SiO2.
16
It was
attributed to the hot electron generated by plasmonic decay. Hot electrons transferred to H 2
molecules and reduced barrier for H2 dissociation.
8
Figure 2.3. (a) (A) High resolution transmission electron micrograph of a single AuNP
supported on SiO2 matrix showing darker contrast of AuNP as compared to SiO 2 support. (B)
Schematic representation of hot electron-induced dissociation of H2 on Au. (C) Real-time
detection of rate of HD formation with laser excitation (2.4 W/cm
2
, on) and without laser
excitation (0.0 W/cm
2
, off). Due to laser heating during 10 min of laser excitation, the
temperature on the sample changes reversibly by 8 °C, as shown in the figure, from 22 to 30 °C.
2.2 Energy transfer mechanisms of photocatalytic reaction rate enhancement by surface
plasmon resonance (SPR)
By adding plasmonic metal structures, the rate of photocatalytic reactions on semiconductors
can be enhanced by SPR. This enhancement is due to the increase in charge carrier concentration.
Three non-mutually exclusive mechanisms have been proposed in the past.
7
H
2
(g) + D
2
(g) → 2HD(g)
9
Resonant photon-scattering mechanism
In this mechanism, plasmonic nanostructures are designed to scatter photons in the resonant
photon scattering field. Therefore, the electron-hole pair formation rate is enhanced in the
semiconductor.
7
Figure 2.4. Schematic illustration of resonant photon scattering mechanism.
7
Near-field electromagnetic mechanism
The plasmonic metal nanostructures can cause strong localized SPR-induced electromagnetic
field. The enhancement can be observed even the semiconductor and plasmonic metal
nanostructures are not in direct contact. For example, they can be separated by thin, non-
conductive spacers preventing any direct charge exchange between the two building blocks.
17-20
10
By adding Au nanoparticles on TiO2 films, Liu et al. observed an enhancement of up to 66X
in photocatalytic water splitting.
12
Figure 2.5. (a) Photocurrent of anodic TiO2 with and without Au nanoparticles irradiated with
λ = 633nm light for 22s. (b) Photocurrent of anodic TiO 2 with and without Au nanoparticles at
zero bias voltage irradiated with UV (λ = 254nm) light for 22s. (c) SEM image of a 5nm Au
island film deposited on anodic TiO 2. (d) Electric field intensity at the interface of Au-TiO2
calculated using finite-difference time-domain (FDTD) method.
12
Figure 2.5(d) shows the electric field intensity in the regions of nearly touched Au
nanoparticles on the TiO2 by FDTD simulation. The intensity is 1000X enhanced compared with
the incident light. This intense local fields produced by SPR couple lightly efficiently to the surface
of TiO2.
(a) ( )
(c) (d)
11
Charge injection mechanism
The metallic plasmonic nanostructures can absorb resonant photons and transfer the energetic
carriers to the nearby semiconductor.
Figure 2.6. Schematic illustration of charge injection mechanism.
7
In Figure 2.7 (a), the volumes of generated hydrogen by Au nanoparticles/TiO2 and TiO2 itself
under 532nm laser and polychromatic lamp light are plotted. As expected, TiO2 itself does not
exhibit any photocatalytic activity under these conditions. In contrast, the hydrogen evolution
using monochromatic 532 nm laser pulses reveals an increase in the formation rate. The proposed
mechanism is shown in Figure 2.7(b). Au nanoparticles are photo-excited, and electrons from Au
are injected into the TiO2 conduction band, which leads to the generation of holes in the Au
12
nanoparticles and electrons in the TiO2 conduction band. The electrons will participate in hydrogen
generation, and the holes will be quenched by EDTA in the solution as sacrificial electron donor.
Figure 2.7. (a) Volume of hydrogen evolved during the photocatalytic runs using Au/TiO2 and
TiO2 as catalysts under 532 nm laser irradiation and polychromatic light λ > 400 nm. (b)
Schematic illustration of photocatalytic activity of Au/TiO 2 upon excitation of Au surface
plasmon.
13
Au Ti
, nm
Au Ti
, nm
Ti
, nm
Ti
, nm
(a) ( )
13
Chapter 3 Hot electron-driven photocatalytic water splitting
3.1 Abstract
We report measurements of photocatalytic water splitting using Au films with and without
TiO2 coatings. In these structures, a thin (3-10nm) film of TiO2 is deposited using atomic layer
deposition (ALD) on top of a 100nm thick Au film. We utilize an AC lock-in technique, which
enables us to detect the relatively small photocurrents (~µA) produced by the short-lived hot
electrons that are photoexcited in the metal. Under illumination, the bare Au film produces a small
AC photocurrent (<1 µA) for both the hydrogen evolution reaction (HER) and the oxygen
evolution reaction (OER) due to hot electrons and hot holes, respectively, that are photoexcited in
the Au film. The samples with TiO2 produce a larger AC photocurrent indicating that hot electrons
are being injected from the metal into the TiO 2 semiconductor where they then reduce hydrogen
ions in solution forming H2 (i.e., 2H
+
+2e
-
→ H 2). The AC photocurrent exhibits a narrow peak
when plotted as a function of reference potential, which is a signature of hot electrons. Here, we
photoexcite a monoenergetic source of hot electrons, which produces a peak in the photocurrent,
as the electrode potential is swept through the resonance with the redox potential of the desired
half-reaction. This stands in contrast to conventional bulk semiconductor photocatalysts, whose
AC photocurrent saturates beyond a certain potential (i.e., light limited photocurrent). The
photocurrents produced at the metal-liquid interface are smaller than those of the metal-
semiconductor system, mainly because, in the metal-semiconductor system, there is a continuum
of energy and momentum states that each hot electron can be injected into, while for an ion in
solution, the number of energy and momentum states are very small.
14
3.2 Introduction
In 2011, Ingram
20
and Liu
12
reported plasmon-enhanced photocatalytic water splitting, and
attributed this to the local field enhancement of sub-band gap (i.e., defect) states in the TiO2. More
recently, hot electrons have been reported in photocatalytic systems, introducing the exciting
possibility of overcoming high barrier reactions. In 2013, Mukherjee et al. reported hot-electron-
induced photodissociation of H2 on small Au nanoparticles (AuNPs) supported on SiO2
16
and later
on TiO2.
11
DuChene et al. reported prolonged hot electron dynamics in plasmonic-
metal/semiconductor heterostructure photocatalysts.
21
Hot electron processes have also been
reported in solid state devices. In 1 , Brongersma’s g roup reported hot-electron photodetection
with a plasmonic nanostripe antenna.
22
In this work, they used a metal/oxide/metal stack, and
photocurrent was only generated when the incident photon energy was larger than the oxide barrier
energy. In 1 , Halas’ group reported similar measurements, which compared the polariza tion
dependence of plasmon-resonant devices with Ohmic and Schottky contacts with defect-rich
TiO2.
23
In the work of Robatjazi et al., plasmon resonant nanoparticles are deposited on top of
NiOx semiconducting films, and the resulting photocurrent generated was attributed to the direct
injection of electrons excited in the metal nanoparticles to ions in the solution.
24
However, in this
sample configuration, there are several possible mechanisms capable of producing a photocurrent,
which are difficult to separate. In our work presented here, we observe photocurrent directly on a
metal film with no semiconductor present, thus, unambiguously demonstrating direct injection of
photogenerated charge from a bulk metal. Several recent theoretical studies have concluded that
plasmon resonant excitations decay into hot electrons in metal nanostructures,
25-29
which enables
useful devices and structures to be engineered despite the extremely short lifetimes of hot electrons
in metals (~10fsec).
30, 31
15
The processes of hot electron injection and hot hole injection are illustrated in Figure 3.1. In
the case of hot electron injection (Figure 3.1(a)), the photon is absorbed in the metal, and the
excited electron is injected into the solution or an adsorbed ion to drive a reduction half-reaction
(e.g., H
+
/H2). Figure 3.1(b) shows the process of hot hole injection, in which a hot hole below the
Fermi level is used to drive oxidation half-reactions (e.g., OH
-
/O2). Figures 3.1(c) and (d) illustrate
the processes in which hot electrons and hot holes are injected into a thin TiO2 layer. The hot
electrons then propagate ballistically (i.e., without scattering) to the ions in solution. As such, these
charged carriers (electrons and holes) remain hot until they reach the ions in solution.
16
Figure 3.1. Schematic diagrams of the (a, c) hot electron injection and (b, d) hot hole injection
of photoexcited carriers, illustrated for the hydrogen evolution reaction (HER) and oxygen
evolution reaction (OER).
3.3 Experimental section
Photoelectrodes were fabricated by depositing 100nm Au on glass substrates using electron
beam deposition. Both bare Au films and TiO 2-coated Au films were studied here. For the TiO2-
coated films, a 1nm layer of Ti was deposited to serve as a seed layer for the atomic layer
deposition (ALD) process. A 3-10nm TiO2 film was then deposited by ALD at 250
°
C using
TDMAT as the Ti source and water vapor as the O source.
32-35
The TiO2 thickness was established
17
by ellipsometry for a 5nm film, which corresponds to 100 ALD cycles. 3 and 10nm thick films
correspond to 60 and 200 ALD cycles. The TiO 2 films were annealed in a quartz tube furnace at
450°C for 30 min while flowing O2 gas. UV-vis absorption spectra of a 10nm thick TiO2 film taken
before and after annealing are plotted in Figure 3.2. No significant differences are observed in the
UV-vis spectra after annealing indicating that the optical absorption of this material is not sensitive
to the long/short range order, which is increased during the annealing process. The photo-I-V
characteristics, however, are quite different after annealing (see Figures 3.6 and 3.7), reflecting the
fact that the transport properties are very sensitive to the long/short range order of this material.
Figure 3.2. UV-vis absorption spectra of a 10nm thick TiO 2 film taken before and after
annealing at 450°C for 30 min in an O2 gas environment.
Figure 3.3(a) shows an illustration of the sample geometry, where an insulated copper wire
was attached to the Au/TiO2 electrode using silver paint and the whole sample, excluding the top
surface, was encased in epoxy to insulate it from the electrolytic solution.
18
Figure 3.3. (a) Diagram illustrating the sample configuration. (b) Schematic diagram of the
three-terminal photoelectrochemical setup with the modulated laser and AC lock-in amplifier.
Photoelectrochemical measurements were performed using a three terminal potentiostat
(Gamry, Inc.), as illustrated in Figure 3.3(b). For the oxygen evolution reaction (OER), a Hg/HgO
reference electrode was used, and for the hydrogen evolution reaction (HER) a Ag/AgCl reference
electrode was used. For all reactions, a Pt wire was used as the counter electrode. The electrodes
were immersed in a pH=7, 0.5 M Na2SO4 solution. We used an AC lock-in technique, which
Gamry
Potentiostat
Pt
Counter
Electrode
B A
100nm Au
Glass
Slide
Substrate
Hg/HgO
Reference
Electrode
0.5 M Na
2
SO
4
solution
Lock-in
Amplifier
3nm TiO
2
ħ ω
Chopper
(b)
(a)
19
enables us to detect the potentially small photocurrents (~µA) produced by the short-lived hot
electrons and just 3nm of semiconductor material, as illustrated in Figure 3.3(b). The incident laser
was chopped at frequency ω chopper, which was 200Hz in our case. The chopper controller (Stanford
Research Systems, Inc., Model SR540) was connected to the “REF IN” terminal of the lock -in
amplifier (Standard Research Systems, Model SRS830 DSP), in order to synchronize the lock-in
amplifier with the light modulation, as illustrated in Figure 3.4.
Figure 3.4. Schematic circuit diagram of the AC lock-in technique.
A 1 Ω resistor wa s inserted between the working electrode terminal of the Gamry potentiostat
system and the sample, and the lock-in amplifier was used to measure the AC component of the
voltage drop across this 1 Ω resistor. Any photoresponse from the sample will produce an AC
20
voltage at the same frequency as the chopped laser. The AC photocurrent presented in Figures 3.6
and 3.7 was obtained by dividing the AC voltage by the resistance of the 1 Ω resistor ( I=V/R).
Figure 3.5. AC photocurrent plotted as a function of the reference potential for n- and p-type
bulk Si photocatalysts in a 0.5M H2SO4 solution.
3.4 Results and discussion
Figure 3.6 shows the DC dark current generated in the three basic sample types (i.e., a bare
100nm Au film and annealed and unannealed samples with 10nm TiO 2 deposited on 100nm Au
films). Under illumination, the bare Au film produces a very small AC photocurrent (<0.1 µA)
around 0.15V vs. NHE, as measured using the setup illustrated in Figure 3.3(a). The samples with
TiO2, on the other hand, produce a larger AC photocurrent indicating that hot electrons are being
injected from the metal into the TiO 2 semiconductor where they then reduce hydrogen ions in
21
solution forming H2 (i.e., 2H
+
+2e
-
→ H 2). The AC photocurrent exhibits a relatively narrow peak
at -0.15 V vs. NHE, which is a signature of hot electrons.
Here, we photoexcite a monoenergetic source of hot electrons, which produces a peak in the
photocurrent as the electrode potential is swept through the resonance with the reduction potential
of the hydrogen evolution reaction. This stands in contrast to conventional bulk semiconductor
photocatalysts, whose AC photocurrent saturates beyond a certain potential, as shown in Figure
3.5. Another key point supporting the hot electron hypothesis is that the unannealed sample
produces very little photocurrent. The unannealed TiO2 material is amorphous and conducts via a
hopping mechanism,
36
whereas the annealed TiO2 is crystalline, and can transport the hot electrons
ballistically without scattering. That is, the TiO 2 film is thinner than the mean free path of the
electrons and holes, so the hot electrons injected from the metal can propagate to the ions in
solution before relaxing back to equilibrium. Here, the photocurrents produced at the metal-liquid
interface are smaller than those of the metal-semiconductor system because, in the metal-
semiconductor system, there is a continuum of energy and momentum states that each hot electron
can be injected into, while for an ion in solution, the number of energy and momentum states are
very small (and possibly singular).
22
Figure 3.6. (a) DC dark current and (b,c) AC photocurrent plotted as a function of the reference
potential for the hydrogen evolution reaction (HER). (c) is a zoomed-in version of (b).
23
We have also measured photocatalytic oxygen evolution (i.e., OER) using the bare Au film
and Au/TiO2 electrodes, as shown in Figure 3.7. Here, we see a rather pronounced AC photocurrent
for the bare Au film peaked around 1.5V vs. NHE, which we attribute to the hot hole injection
mechanism illustrated in Figure 3.1(b). Again, the annealed TiO2-coated electrode, shows a higher
photocurrent, indicating that hot hole injection is the dominant photocurrent generation mechanism.
Here, the unannealed sample produces no detectable photocurrent because this material is
amorphous and conducts via a hopping mechanism. As such, the hot holes relax to equilibrium
even while traversing the relatively thin TiO 2 films, whereas the annealed TiO2 is crystalline,
which enables the hot holes to propagate ballistically without scattering and relaxing back to
equilibrium. It is somewhat surprising that the bare Au and Au/TiO 2 (annealed) samples only have
a small difference in AC current density, considering that the TiO 2 has much lower inherent
overpotential for driving OER compared to Au. Here, we believe that the hot electron process is
fundamentally different from equilibrium catalysis, and therefore is not governed by the same
kinetics (i.e., overpotentials) of standard reactions.
24
Figure 3.7. (a) DC dark current and (b) AC photocurrent plotted as a function of the reference
potential for the oxygen evolution reaction (OER).
Brown et al. calculated the initial excited carrier distributions generated by surface plasmon
decay. The plasmon-generated hot carrier distribution is extremely sensitive to details of the
electronic band structure, and for Au, the interband threshold is ~1.6-1.8eV. Above this threshold,
direct transitions from the d band to unoccupied states above the Fermi level dominate, and the
energy distribution of hot carriers exhibits one peak corresponding to hot electrons and one peak
corresponding to hot holes.
27
Several other groups have also reported narrow energy distributions
for photoexcited hot carriers in metals.
25, 26, 37
In the work of Avanesian et al., the electron-induced
adsorbate dynamics on metal surfaces were modeled using a nonadiabatic, first-principles inelastic
25
electron scattering model.
38
They predict narrowly distributed reaction probabilities, consistent
with the resonance behavior we observe experimentally.
3.5 Conclusion
In conclusion, we observe evidence of hot electron (and hot hole) driven photocatalysis on bare
Au films and TiO2-coated Au films. Here, an AC lock-in technique is used to detect the relatively
small photocurrents (~µA) produced by the short-lived hot electrons that are photoexcited in the
metal. For both the hydrogen evolution reaction (HER) and the oxygen evolution reaction (OER),
the AC photocurrent exhibits a relatively narrow peak, as the monoenergetic distribution of hot
electrons and holes photoexcited in the metal is swept through the resonance with the redox
potential of the desired half-reaction. The TiO2-coated Au films produce a higher photocurrent
than the bare Au films because of the large density of electronic states in the TiO 2 semiconductor
compared to the small number of states of the ions in solution. While these hot electron and hot
hole photocurrents are relatively small, plasmon resonant nanostructures can be utilized in the
future to engineer useful devices and structures despite the extremely short lifetimes of hot
electrons in metals (~10fsec).
26
Chapter 4 Resonant and selective excitation of photocatalytically active defect sites in TiO2
4.1 Abstract
It has been known for several decades that defects are largely responsible for the catalytically
active sites on metal and semiconductor surfaces. However, it is difficult to directly probe these
active sites because the defects associated with them are often relatively rare with respect to the
stoichiometric crystalline surface. In the work presented here, we demonstrate a method to
selectively probe defect-mediated photocatalysis, through differential AC photocurrent (PC)
measurements. In this approach, electrons are photoexcited from the valence band to a relatively
narrow distribution of sub-bandgap states in the TiO2, and then subsequently to the ions in solution.
Because of their limited number, these defect states fill up quickly resulting in Pauli blocking, and
are thereby undetectable under DC or CW excitation. In the method demonstrated here, the
incident light is modulated with an optical chopper while the photocurrent is measured with a lock-
in amplifier. Thin (5nm) films of TiO 2 deposited by atomic layer deposition (ALD) on various
metal films, including Au, Cu, and Al, exhibit the same wavelength-dependent photocurrent
spectra, with a broad peak centered around 2.0eV corresponding to the band-to-defect transition
associated with the hydrogen evolution reaction (HER). While the UV-vis absorption spectra of
these films show no features at 2.0eV, photoluminescence (PL) spectra of these photoelectrodes
show a similar wavelength dependence with a peak around 2.0eV, corresponding to the sub-band
gap emission associated with these defect sites. As a control, alumina (Al 2O3) films exhibit no PL
or PC over the visible wavelength range. The AC photocurrent plotted as a function of electrode
potential, shows a peak around -0.4 to -0.1V vs. NHE, as the monoenergetic defect states are tuned
through a resonance with the HER potential. This approach enables the direct photo-excitation of
catalytically active defect sites to be studied selectively without the interference of the continuum
27
interband transitions or the effects of Pauli blocking, which is limited by the slow turnover time of
the catalytically active sites, typically on the order of 1 µsec. We believe this general approach
provides an important new way to study the role of defects in catalysis in an area where selective
spectroscopic studies of these are few.
4.2 Introduction
Defects in TiO2 have been studied extensively, providing an important mechanism in
photocatalytic energy conversion. In particular, oxygen vacancies (i.e., Ti
3+
states) have been
linked to catalytically active sites, particular in the water splitting and CO 2 reduction reactions
systems.
39
TiO2 films deposited by atomic layer deposition (ALD) are also known to have a high
concentration of defects and, hence, show enhanced water splitting and CO 2 reduction
efficiencies.
40-43
Qiu et al. have quantified these O-vacancies (i.e., Ti
3+
states) in ALD-deposited
TiO2 films using X-ray photoemission spectroscopy (XPS), and they have correlated these
vacancies with the photocatalytic activity of TiO2 films for both water splitting and CO2 reduction
reactions.
40, 42
Density functional theory calculations performed by Alexandrova’s group have
provided an atomistic picture of this enhancement mechanism, which show that both H2O and CO2
molecules bind stably to these non-stoichiometric Ti
3+
states.
41, 44
Furthermore, when they allow
their calculations to relax to their quantum mechanical ground state, they observe a spontaneous
transfer of one electron creating CO 2
-
. In these DFT calculations, this CO2
-
species is bent and
represents a high barrier intermediate species in this difficult reaction system.
In addition to oxygen vacancies, nitrogen defects can be created in TiO2 by annealing in NH3
gas, resulting in substantial sub-band gap absorption.
45-47
This is a well-studied system in which
the N-defect concentration can be controlled up to several percent by varying the annealing
28
temperature. Extensive surface science studies of N-doping have been performed in the research
groups of Rodriguez
48
and Yates.
49-51
Doping of metal oxides by ion implantation, followed by
calcinations in oxygen, has also been studied extensively as a means of dramatically increasing in
the photocatalytic activity in the visible wavelength range.
52-61
Among the elements studied, V, Cr,
Mn, Fe, and Ni were found to increase the photocatalytic activity of TiO 2 in the visible range
substantially.
62
In the work presented here, an AC lock-in measurement technique is employed to study
photocatalysis, revealing the behavior of sub-band gap states, that are resonant in both wavelength
and electrode potential, and give rise to a substantial increase in photocurrent in the hydrogen
evolution reaction (HER) process. In order to further characterize these sub-band gap states, we
collect photoluminescence (PL) spectra, which provide an independent measure of the energetics
of these important sub-band gap states. UV-Vis spectra are also obtained in order to provide a
complete picture of the band edge and sub-band gap absorption in the TiO2 thin film with and
without annealing.
4.3 Experimental section
Photoelectrodes were fabricated by depositing 100 nm thick films of Au, Cu and Al on glass
substrates using electron beam deposition. A 5 nm TiO2 film was then deposited by ALD at 250°C
using TDMAT as the Ti source and water vapor as the O source. The TiO2 thickness was
established by ellipsometry for a 5 nm film, which corresponds to 100 ALD cycles. The TiO 2/Au
films were annealed in a quartz tube furnace at 450°C for 30 min while O 2 gas was flowing. The
TiO2/Cu and TiO2/Al films were annealed in a quartz tube furnace at 450°C for 30 min while argon
gas was flowing. UV-vis absorption spectra of a 10nm thick TiO 2 film taken before and after
29
annealing are plotted in Figure 3.2. Figure 4.1(a) shows an illustration of the sample geometry, in
which an insulated copper wire was attached to the TiO2/metal electrode using silver paint and the
whole sample, excluding the top surface, was encased in epoxy to insulate it from the electrolytic
solution. Photoelectrochemical measurements were performed using a three terminal potentiostat
(Gamry, Inc.), as illustrated in Figure 4.1(b). For the HER, a Ag/AgCl reference electrode was
used and a Pt wire was used as the counter electrode. The electrodes were immersed in a pH=7,
Na2SO4 solution. We used an AC lock-in technique, which enables us to detect the relatively small
photocurrents (µA) produced by just 5nm of wide-band gap semiconductor material. In order to
vary the wavelength of the incident light, a 1000W xenon lamp was used in conjunction with a
monochromator to produce monochromatic light throughout the visible wavelength range. The
power reaching the sample surface was 2-3mW. Here, the light was filling the entire sample area.
The incident light was chopped at frequency ωchopper, which was 200 Hz in our case. The chopper
controller (Stanford Research Systems, Inc., Model SR ) was connected to the ‘‘REF IN’’
terminal of the lock-in amplifier (Standard Research Systems, Model SRS830 DSP), in order to
synchronize the lock-in amplifier with the light modulation. The AC voltage signal from the
auxiliary current monitor terminal of the potentiostat was used as the input signal of the lock-in
amplifier in order to collect the AC photocurrent generated by the TiO 2/metal films.
30
Figure 4.1. (a) Diagram illustrating the sample configuration. (b) Schematic diagram of the three-
terminal photoelectrochemical setup with the modulated light and AC lock-in amplifier.
4.4 Results and discussion
Figure 4.2 shows the AC and DC photocurrents plotted as a function of the reference potential
for the HER for 5nm TiO2 deposited on Au and Cu. Here, we observe a peak in the AC
photocurrent around -0.13V vs. NHE corresponding to the conditions under which we tune the
potential of the charge associated with the resonantly-excited defect states through the redox
potential of the HER half-reaction. This is considerably shifted from the DC onset potential of -
0.3V vs. NHE. For the TiO2/Cu electrode, the peak in AC photocurrent is observed at -0.4V vs.
NHE, which is also substantially shifted with respect to the DC onset potential of -0.6V vs. NHE.
31
Figure 4.2. DC (a, c) and AC (b, d) photocurrent plotted as a function of the reference potential
for the hydrogen evolution reaction (HER) for (a, b) 5nm annealed TiO 2 deposited on Au under
532nm illumination. (c, d) 5nm annealed TiO 2 deposited on Cu under 532nm illumination.
Figure 4.3 shows the light intensity dependence of the AC photocurrent from a TiO 2/Au sample
obtained for both reduction and oxidation reactions under 532nm illumination. Here, we observe
a linear response to the light intensity, as expected. This result indicates that we are successfully
converting photons to hot carriers under optical excitation.
11
-1.0 -0.5 0.0 0.5 1.0 1.5
-200
-150
-100
-50
0
50
5nmTiO
2
(annealed)/Au
DC Photocurrent ( A)
Potential vs. NHE (V)
-1.0 -0.5 0.0 0.5 1.0 1.5
-12
-10
-8
-6
-4
-2
0
5nmTiO
2
(annealed)/Au
AC Photocurrent ( A)
Potential vs. NHE (V)
-0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4
-0.7
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0.0
5nmTiO
2
(annealed)/Cu
AC Photocurrent ( A)
Potential vs. NHE (V)
(a)
(b) (d)
(c)
-0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
5nm TiO
2
(annealed)/Cu
DC Photocurrent (mA)
Potential vs. NHE (V)
32
Figure 4.3. Photocurrent obtained under (a) reduction and (b) oxidation conditions from a
TiO2/Au sample as a function of intensity of 532nm laser excitation.
Figure 4.4(a) shows the normalized AC photocurrent spectra of 5nm TiO2 deposited on Au,
Cu, and Al. Here, a clear peak can be seen around 2eV, which is well below the 3.2eV bandgap of
this material. This feature at 2eV corresponds to the photo-excitation of the sub-band gap defect
states that serve as catalytically-active sites for the HER, as illustrated in Figure 4.4(b). Further
evidence for this defect state is provided by the PL spectrum, also plotted in Figure 4.4(a), which
also exhibits a peak centered around 2.0eV. Interestingly, no such feature can be seen in the UV-
vis absorption spectra at 2.0eV. Figure 3.2 shows the UV-vis absorption spectra of both annealed
and unannealed TiO2 films. Here, both spectra show moderate sub-band gap absorption that is
monotonically decreasing with wavelength. However, no specific feature is present in these UV-
vis spectra at 2.0eV. This indicates that, while there are likely many different types of defect states
throughout the band gap, there is a specific defect located at 1.2eV above the valence band edge
that is catalytically active for the hydrogen evolution reaction.
33
Figure 4.4. (a) Normalized AC photocurrent spectra of 5nm TiO 2 deposited on Au, Cu, and
Al films and corresponding photoluminescence spectrum of 5nm thick annealed TiO 2
deposited on a Au film. (b) Energy band diagram of the photoexcitation mechanism of defect-
mediated photocatalysis.
(a)
(b)
+
- HER
Defect States
1.2eV
Conduction Band
Valence Band
2eV
1.8 2.0 2.2 2.4 2.6 2.8
0.0
0.2
0.4
0.6
0.8
1.0
5nmTiO
2
(annealed)/Au
5nmTiO
2
(annealed)/Cu
5nmTiO
2
(annealed)/Al
Photoluminescence Intensity (a.u.)
Normalized AC Photocurrent ( A/mW)
Photon Energy (eV)
34
Ultraviolet photoemission spectroscopy (UPS) is a common tool to study conduction and
valence band structure of materials.
63, 64
The valence band spectrum of an oxygen-annealed 5nm
TiO2 film was obtained by UPS at 21.2eV photon energy. The linear plot, Figure 4.5(a), shows the
position of valence band edge. The valence band edge is 3.2eV below the Fermi energy, which is
pinned at the conduction band edge. After annealing, a peak at approximately 2eV below the
conduction band edge was observed in the log plot, as shown in Figure 4.5(b), which agrees with
the peak position of photocurrent spectra.
Figure 4.5. Valence band spectra of a 5nm TiO2 film obtained by UPS at 21.2eV photon energy
plotted on (a) linear and (b) log scales.
20 15 10 5 0
Intensity (a.u.)
Binding Energy (eV)
valence
band edge
20 15 10 5 0
Intensity (a.u.)
Binding Energy (eV)
defect
states
(a) (b)
35
Figure 4.6 shows an X-ray photoemission spectroscopy (XPS) spectrum of a 5nm TiO 2 film
deposited on a Au film by atomic layer deposition. In addition to the main Ti
4+
peaks, which
correspond to stoichiometric TiO2, we observe a small shoulder peak corresponding to Ti
3+
states
(i.e., oxygen-vacancies), as reported previously by Qiu et al.
33
Figure 4.6. Ti 2p core level XPS spectrum of 5nm TiO2 deposited on Au.
468 466 464 462 460 458 456
Intensity (a.u.)
Binding Energy (eV)
XPS Data
Fitting
5nm TiO
2
Ti
3+
36
Figure 4.7. High resolution TEM images of TiO2 deposited with (a) 100 cycles and (b) 500
cycles on Au. (c) and (d) are zoom-in images of (a) and (b), respectively.
High resolution cross-sectional transmission electron microscope (TEM) images were taken of
5nm and 25nm thick TiO2 films deposited on Au films. Figure 4.7(c) shows amorphous TiO2
material, as expected. This material is non-stoichiometric and contains a high density of oxygen
vacancies, as evidenced by the UPS and XPS spectra described above. In contrast, Figure 4.7(d)
shows crystalline material. Qiu et al. have also demonstrated that above 10nm, TiO2 films become
crystalline.
32
These thicker TiO2 films are crystalline and have poor charge transfer characteristics
due to their insulating nature. Thus, we are more interested in thin and amorphous TiO 2 films.
5nm TiO
2
25nm TiO
2
1nm 1nm
(c)
(d)
(b) (a)
37
Plane-wave density functional theory (PW-DFT) calculations were performed on defective
(101) anatase in order to extract formation energies and spectral analyses of oxygen vacancies. A
single oxygen vacancy results in the presence of 2 excess electrons that may localize on
neighboring Ti atoms. Experimentally, spectral signatures were found ~2eV from the Fermi energy.
Due to the over-delocalization of electrons in DFT, we introduced a strong Hubbard U parameter
of 4.0eV into our calculations in order to recover a theoretical band gap is ~2.2eV. Utilizing Bader
charge analysis, we were able to determine the localization of electrons on surface Ti sites. For
surface oxygen vacancies VO1 and VO2, electrons localized on neighboring Ti atoms ~1.9Ǻ from
the vacancy site. For surface oxygen vacancy VO3, the two electrons localized on Ti atoms farther
away from the vacancy site, one on a Ti atom ~1.9Ǻ and the other on a Ti atom ~ .7Ǻ from the
vacancy site. The ramifications of this may be found in the theoretical density of states shown in
Figure 4.8(b). In comparison to VO1-O2, VO3’s spectral signatures of the localized electrons are
shifted farther away from the Fermi energy. We hypothesize that at the deeper oxygen vacancy
sites, the spectral signatures of the localized electrons also shift farther away from the Fermi energy.
However, the small theoretical band gap of ~2.2eV may obscure these spectral signatures and so,
we focus only on VO1-O3. Based on these DFT calculations, it seems that the specific catalytically
active defect sites that we observe in our photoelectrochemical measurements actually correspond
to sub-surface defects. While these results provide qualitative agreement with our experimental
results (i.e., sub-band gap states giving rise to visible light absorption), it is not possible to assign
a specific atomic defect to our observation, and it is likely that a large collective ensemble of these
defects contribute to the broad spectral distribution observed experimentally.
38
Figure 4.8. (a) The orthorhombic cell used to model the anatase surface. Nine different oxygen
vacancies were explored with the surface oxygen vacancies, V O1-O3, displaying visible spectral
signatures shifted away from the Fermi energy. (b) Projected density of states can depict the
localization of electrons in the band gap. These spectral signatures are similar for V O1 and VO2
since the electrons localize on surface Ti atoms. In contrast, electrons are more delocalized for V O3
across farther neighboring Ti atoms.
In our previous work, similar structures consisting of Au films with and without TiO2 coatings
were investigated.
65
For the bare metal electrodes, the mechanism of photocatalysis was attributed
to hot electrons photoexcited in the metal. The mechanism of photocurrent generation in TiO2-
coated Au electrodes was attributed to hot electron injected by the metal film because of our
inability to detect these defect states in the UV-vis absorption process. However, these defect states
are somewhat elusive because of the Pauli blocking associated with their finite density and slow
39
turnover time, and we now have a better understanding of this system. It should be noted that the
photoexcited charge generated at these O-vacancy defect sites is bound to the defect site and is not
free to propagate throughout the TiO2 crystal.
4.5 Conclusion
In conclusion, we have demonstrated a method to selectively probe defect-mediated
photocatalysis through differential AC photocurrent measurements. Here, we drive the
photoexcitation of electrons (or holes) from the valence band to relatively narrow distribution of
sub-band gap states, and then to the ions in solution. Because of their limited number, these defect
states fill up quickly resulting in Pauli blocking and are undetectable under DC or CW conditions.
In the method demonstrated here, the incident light is modulated with an optical chopper and the
photocurrent is measured with a lock-in amplifier. Thin (5nm) films of TiO2 deposited on different
metals (Au, Cu, and Al) using ALD exhibit the same wavelength-dependent photocurrent spectra,
with a broad peak centered around 2.0eV. PL spectra also show a peak at 2.0eV, corresponding to
the sub-band gap emission associated with these defect sites. In addition, the AC photocurrent
shows a peak around -0.4 to -0.1V vs. NHE, as the monoenergetic defect states are tuned through
a resonance with the HER potential. This approach enables the photoexcitation of catalytically
active defect sites to be studied selectively without the interference from the continuum of
interband transitions or the effects of Pauli blocking, which is quite pronounced because of the
slow turnover time (~1µsec) of the catalytically active sites.
40
Chapter 5 Prevention of surface recombination by electrochemical tuning of TiO2-passivated
photocatalysts
5.1 Abstract
We present a systematic study of photoluminescence (PL) spectroscopy of TiO 2-passivated
GaAs as a function of electrochemical potential in an ionic liquid solution. We observe a 7X
increase in the PL intensity as the GaAs transitions from accumulation to depletion due to the
applied potential. We attribute this to the excellent control over the surface Fermi level enabled by
the high capacitance of the electrochemical double layer and TiO 2. This allows us to control the
surface carrier concentration and corresponding non-radiative recombination rate. In addition to
photoluminescence spectroscopy, we also measured the capacitance-potential (i.e., C-V)
characteristics of these samples, which indicate flat band potentials that are consistent with these
regimes of ion accumulation observed in the photoluminescence measurements. We have also
performed electrostatic simulations of these C-V characteristics, which provide a detailed and
quantitative picture of the conduction and valance band profiles, and charge distribution at the
surface of the semiconductor. These simulations also enable us to determine the range of potentials
over which the semiconductor surface experiences depletion, inversion, and accumulation of free
carriers. Based on these simulations, we can calculate the Shockley-Read-Hall (SRH)
recombination rate and model the PL intensity as a function of voltage. We show that this approach
allows us to explain our experimental data well.
5.2 Introduction
At typical photocatalytic semiconductor-liquid interfaces, there is usually considered to be an
equilibrium between the Fermi energy in the semiconductor and the redox potential of the ions in
41
solution. This gives rise to band bending and a built-in potential at the surface of the semiconductor,
which causes separation of the photoexcited electron-holes pairs, when illuminated.
6
In
photocatalysis literature, these bands are often only sketched qualitatively, even though
commercially available tools for rigorous modeling are commonly used for semiconductor devices.
In addition, there are electrostatic charges trapped at the surface of most semiconductors that
further shift this equilibrium. As such, it is difficult (if not impossible) to predict the position of
the flat band potential at a given semiconductor-liquid interface without careful calibration to
experiment. Capacitance-voltage measurements (i.e., Mott-Schottky measurements) are
commonly used to determine the carrier type
66-69
and establish the position of the flat band
potential at the semiconductor-liquid interface experimentally.
70
The flat band potential of TiO2-
passivated InP photocathodes have been measured by Lin et al. using the Mott-Schottky method,
by performing a linear fit of the 1/C
2
-V data and extrapolating the voltage axis intercept.
71
Hu et
al. also used this method to acquire the values of the built-in voltage produced in n-type GaAs
nanowire/CH3CN-FeCp2
+/0
junctions and planar GaAs/CH3CN-FeCp2
+/0
junctions.
72
In addition to modulating the band bending, as mentioned above, the ions in the solution can
screen trapped surface charges in the semiconductor, which typically give rise to surface
recombination and represent a major loss mechanism in semiconductor photocatalysis. Over the
past few years, several research groups have utilized ionic liquids to passivate the surface states of
semiconductors.
73, 74
This approach has been used to mitigate the effects of surface depletion
66, 75,
76
and non-radiative surface recombination due to the dangling bonds and surface states in the
semiconductor.
73
Arab et al. observed a 12-fold enhancement in the PL intensity in GaAs
nanowires with no applied potential, and an up to 42X increase in the PL intensity of GaAs
42
nanosheets with AlGaAs passivation.
73, 77
In monolayer MoS2, both the photocurrent and PL
intensity are enhanced by a factor of two to three by ionic liquid gating.
74
In the work presented here, we measure direct evidence of the reduction of surface
recombination in a photocatalytic semiconductor under applied electrochemical potentials. We
correlate the onset of this PL enhancement with flat band potentials obtained by C-V Mott-
Schottky measurements, which enables us to determine the range of electrochemical potentials
over which the semiconductor undergoes depletion, inversion, and accumulation of free charge.
Electrostatic modeling of the C-V characteristics using Technology Computer Aided Design
(TCAD) Sentaurus provides a detailed picture of the band profiles and charge distributions under
these applied electrochemical potentials. A detailed mechanism of this gate-induced modulation
of the PL efficiency is developed, within the context of this electrostatic model.
5.3 Experimental section
TiO2-passivated semiconductors have become an important new class of photocatalyst. Here,
a thin layer of TiO2 deposited by atomic layer deposition (ALD) is able to prevent photocorrosion
of most known semiconductors without sacrificing photocatalytic performance.
10, 40-43, 78-81
This
represents a major breakthrough in photocatalysis, which was previously relegated to
electrochemically robust materials like metal oxides, which typically have poor carrier mobilities
and lifetimes, as well as large band gaps. For example, in the work presented here, we use
commercially available GaAs with a mobility of µ=250cm
2
/V·s, as compared to TiO2, which has
typical carrier mobilities of µ=1cm
2
/V·s.
82
Furthermore, the carrier concentration of the GaAs is
known and is highly uniform, which enables precise modeling of the band bending at the
43
photocatalytic interface. Here, we use p-type (111) oriented GaAs substrates with doping
concentrations of 5.2-7.7×10
17
cm
-3
. Back ohmic contacts were made to the p-GaAs by evaporating
5nm thick Ti followed by 50nm of Au. Atomic layer deposition (ALD) of 5nm TiO 2 was performed
at 250
o
C on the p-GaAs wafers, using Tetrakis(dimethylamido)titamium(IV) (TDMAT) as the
titanium source and water vapor as the oxygen source. The carrier gas during the deposition was
argon with a flow rate of 90 sccm, and TDMAT is used for the second half-cycle. An insulated
copper wire was attached to the back contact of the p-GaAs sample using silver paint, and the
entire sample, excluding the TiO2-passivated surface, was encased in epoxy to insulate it from the
solution. The solution is a non-aqueous ionic liquid solution consisting of 0.1M 1-ethyl-3-
methylimidazolium tetrafluoroborate ([EMIM][BF4]) in acetonitrile. A three-terminal potentiostat
(Gamry Reference 600) was used to maintain a potential between the TiO 2-passivated GaAs
working electrode and the Ag/AgNO3 reference electrode. Under certain potentials,
photoluminescence spectra were taken using a spectrometer (Reinshaw inVia Raman Microscope)
with a water immersion lens covered by a 13µm thick Teflon sheet (American Durafilm) to protect
it from the acetonitrile solution, as illustrated in Figure 5.1. Mott-Schottky measurements were
also performed by the Gamry Reference 600 potentiostat.
44
Figure 5.1. Schematic diagram of the three-terminal photoelectrochemical cell. The water
immersion lens is mounted on a microscope in the spectrometer for photoluminescence
measurements.
5.4 Results and discussion
Figure 5.2 shows the photoluminescence intensity of a TiO 2-passivated GaAs photocathode,
measured using the configuration illustrated in Figure 5.1. Here, we see an increase in the
photoluminescence intensity for potentials below -0.5V vs. NHE, reaching a 7-fold increase
around -1.5V vs. NHE. The 1/C
2
-V plot shown in Figure 5.2(b) indicates a flat band potential of -
0.5V vs. NHE. This increase in the PL intensity is caused by the accumulation of cations on the
surface of the semiconductor. These cations screen the surface states in the GaAs that normally
cause non-radiative recombination, thus increasing the PL intensity. The surface states in the GaAs
are associated with native defects, e.g. arsenic vacancies, gallium vacancies, and antisite defects.
83
Based on the 1/C
2
vs. V curve plotted in Figure 5.2(b), we are able to determine the range of
45
potentials over which we have inversion, depletion, and accumulation of free carriers in the
semiconductor, as labeled in Figure 5.2(a).
Figure 5.2. (a) The photoluminescence intensity and (b) Mott-Schottky (1/C
2
vs. V) plot of a
TiO2-passivated GaAs photocathode measured as a function of the reference potential. (c,d)
Simulated results for a GaAs photocathode over the same voltage range.
In order to model the C-V behavior of our semiconductor surface, we performed electrostatic
simulations of this device using the TCAD Sentaurus software package, which solves Poisson’s
equation iteratively with the electron and hole continuity equations and provides the self-consistent
charge density profile in the semiconductor. Here, we have used a p-type GaAs substrate and 5 nm
dielectric layer on top to model the device. We have included the effect of interface states on the
C-V behavior by including interface traps at the semiconductor-dielectric (TiO2) interface. The
46
series capacitance of the TiO2 layer plus the double layer is modeled by an effective dielectric
constant of the TiO2 layer. The voltage is applied to the back contact of the GaAs substrate, as in
the experimental configuration. The simulated Mott-Schottky (i.e., 1/C
2
vs. V) plot is shown in
Figure 5.2(d), which agrees well with the experimental data. By fitting the position and magnitude
of the C-V plot, we can extract the band diagrams and charge distributions in the dark, as plotted
in Figure 5.3. For each of the Figure 5.3(a)-(c), the energy scale has been set relative to the Fermi
Level. As such, in each figure the 0eV reference point corresponds to E F. In this case, the
significant difference in band diagrams between accumulation and inversion/depletion occurs
because of the fundamentally different physics that occurs at the surface. In p-GaAs, during
depletion and inversion, an increased concentration of negative charge at the surface causes the
surface Fermi level to move away from the valence band and closer to the conduction band. During
accumulation, the increased concentration of positive charge at the surface causes the surface
Fermi level to move in the opposite direction as compared to depletion/inversion.
47
Figure 5.3. Energy band diagrams and free carrier concentration of the semiconductor surface
under (a), (d) inversion, (b), (e) depletion, and (c), (f) accumulation conditions.
The photoluminescence intensity in semiconductors can be written as the product of two terms,
(Internal Radiative Efficiency) × (Escape Probability), where the escape probability is the
probability that a photon emitted in the semiconductor will escape into free space, and is related
to the geometry of the device and the carrier generation profiles. In our case, since the carrier
generation profiles and geometry are fixed across the entire measurement range, we can assume
that the change in luminescence efficiency can be modeled by a change in internal radiative
efficiency. Thus, by tracking the filling fraction of these traps as a function of applied voltage, we
can predict the change in radiative efficiency of our GaAs semiconductor by the following model:
𝐼𝑛𝑡𝑒𝑟𝑛𝑎𝑙 𝑅𝑎𝑑𝑖𝑎𝑡𝑖𝑣𝑒 𝐸𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑐𝑦 =
𝐵 𝑛 2
𝐴𝑛 + 𝑆𝑛 + 𝐵 𝑛 2
0 20 40 60 80 100
0
1x10
15
2x10
15
3x10
15
4x10
15
Electron Density (cm
-3
)
Position (nm)
Inversion
0 20 40 60 80 100
0
1x10
17
2x10
17
3x10
17
4x10
17
5x10
17
Hole Density (cm
-3
)
Position (nm)
Depletion
0 20 40 60 80 100
0
1x10
19
2x10
19
3x10
19
4x10
19
5x10
19
6x10
19
7x10
19
Hole Density (cm
-3
)
Position (nm)
Accumulation
(a)
(d) (e)
(b)
(f)
(c)
10 20 30 40 50
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
E
F
E
C
Energy (eV)
Position (nm)
Inversion
E
V
10 20 30 40 50
-1.0
-0.5
0.0
0.5
1.0
1.5
E
V
E
F
Energy (eV)
Position (nm)
E
C
Depletion
10 20 30 40 50
-0.5
0.0
0.5
1.0
1.5
2.0
E
F
E
V
Energy (eV)
Position (nm)
Accumulation
E
C
48
where A is the bulk Shockley-Read-Hall (SRH) recombination parameter, B is the radiative
recombination parameter, S is a fitting parameter corresponding to SRH recombination due to the
surface, and n is the carrier concentration. We have shown in Figure 5.2 that we can quantitatively
extract the position of the surface Fermi level in the dark through a combination of experimental
measurements and simulations. We use this to calculate the change in the surface recombination
parameter S as a function of applied voltage under optical illumination. To carry this out, we use
the full equation for SRH recombination
84
:
𝑈 =
𝜎 𝑛 𝜎 𝑝 𝑣 𝑡 ℎ
𝑁 𝑡 (𝑛𝑝 − 𝑛 𝑖 2
)
𝜎 𝑛 (𝑛 + 𝑛 𝑖 exp (
𝐸 𝑡 − 𝐸 𝑖 𝑘𝑇
)) + 𝜎 𝑛 (𝑝 + 𝑛 𝑖 exp (
𝐸 𝑖 − 𝐸 𝑡 𝑘𝑇
))
where n and p are the electron and hole capture cross sections, vth is the thermal velocity, Nt is the
trap density, n and p are the electron and hole concentrations, ni is the intrinsic carrier concentration,
Et is the trap energy level, and Ei is the intrinsic Fermi level. We can then evaluate this equation
as a function of surface Fermi level and illumination. Here, we assume a trap distribution that
exponentially decays in energy starting at the valence band edge, as previously measured for GaAs
surfaces,
85
and integrate the recombination rate over the entire bandgap for each condition. To
evaluate the carrier concentration as a function of band bending, we use the functions:
𝑛 = 𝑁𝑐 exp (−
𝐸 𝐶 −𝐸 𝐹 𝑘𝑇
) + ∆𝑛 = 𝑁𝑐 exp (−
𝐸 𝐶 −𝐸 𝐹𝑛
𝑘𝑇
)
𝑝 = 𝑁𝑣 exp (−
𝐸 𝐹 − 𝐸 𝑉 𝑘𝑇
) + ∆𝑝 = 𝑁𝑣 exp (−
𝐸 𝐹𝑝
− 𝐸 𝑉 𝑘𝑇
)
to calculate the carrier concentration, where Nc and Nv are the effective density of states in the
conduction and valence band, and n and p are the excess carrier concentrations due to optical
generation, which can be calculated by multiplying the optical generation rate by the carrier
49
lifetime. EF is the Fermi level of the system in the dark, and E Fn and E Fp are the electron and hole
quasi-Fermi levels, respectively.
Figure 5.4. (a) The normalized recombination rate of a GaAs semiconductor as a function of
surface potential under illumination. (b) The surface potential plotted as a function of NHE
extracted from the TCAD simulation. (c) The expected recombination rate as a function of
applied bias under illumination.
Figure 5.4(a) shows the normalized recombination rate plotted as a function of surface
potential under light condition. The details of the calculations, including the constants used are
50
given in the Appendix B. Here, we see that, when illuminated, the recombination rate peaks when
the bands are flat (i.e., surface potential ~0 V) and when the bands are bent to near inversion (i.e.,
surface potential ~EG/e). This behavior can be understood by considering that the SRH
recombination rate is dependent on the degree to which the semiconductor is out of equilibrium,
quantified as np-ni
2
. Thus, when the surface is depleted, the primary source of electrons and holes
will be optical generation. Since that rate is fixed by the external light source, the np product will
remain fixed, and the recombination rate will remain fixed. However, when the surface is in the
inversion or accumulation regions, the np product will increase, thus increasing the recombination
rate. In our case, the recombination rate will increase when the electron or hole carrier
concentration becomes large compared to the level set by optical generation. In order to model
this behavior, we can use the optical generation rate in conjunction with the surface potential vs.
NHE relation extracted from the TCAD simulation, as shown in Figure 5.4(b), to calculate the
expected recombination rate as a function of applied bias under illumination. With this, we obtain
the relative change in the recombination rate near the surface, which we can then use as the
parameter S to calculate the relative change in the internal radiative efficiency, which we have
plotted in Figure 5.2(c). Importantly, we see that this gives us good agreement with the
experimental results.
The key physical reason behind the PL behavior exhibited here is the modulation of surface
state recombination activity due to the relative location of the Fermi level in the bandgap. In both
TiO2-passivated GaAs and bare GaAs, there exist some distribution of surface states in the bandgap.
However, the specific energetic distributions and density may vary. Thus, this same overall
behavior is expected from both bare and TiO 2-passivated surfaces. However, there will be
51
differences in the magnitude of the effect, depending on both the distribution and density of surface
traps.
Photocatalysis at the semiconductor-liquid interface represents a complex process, which
includes band bending, built-in electric fields, surface recombination of photoexcited carriers, and
charge transfer to the ions in solution. There are loss mechanisms associated with each of these
key components to the overall photoconversion efficiency. This complex process is often
oversimplified in order to provide a basic interpretation of the data. The method and results
demonstrated here represent an important step towards obtaining a more rigorous understanding
the overall “photocatalytic” or photoelectrochemically system, in which these key components can
be decoupled experimentally.
5.5 Conclusion
In conclusion, we observe a large increase in the photoluminescence intensity under large
negative applied potentials due to the electrochemically-induced filling of surface states (i.e.,
surface recombination centers). GaAs, which is known to suffer from severe surface states, shows
a 7X increase in photoluminescence intensity. By modeling the capacitance-voltage (i.e., 1/C
2
-V)
data measured over the same range of applied electrochemical potentials, we obtain a detailed
picture of the conduction and valence band profiles, as well as the charge density profiles, thus
establishing the range of potentials over which the semiconductor undergoes depletion, inversion,
and accumulation of free carriers. By calculating the relative surface recombination rates as a
function of applied bias, we are able to accurately predict the modulation of the PL intensity that
is observed experimentally.
52
Chapter 6 Future research directions and outlook for hot-electron and defect-mediated
photocatalysis
6.1 Hot-electron driven photocatalysis on plasmon resonant structures
Hot electrons excited in plasmonic metal nanostructures have been used in solid-state devices
and photocatalysis recently. Chalabi et al. have studied the voltage-dependence, spectral-
dependence, strip width-dependence, and polarization-dependence of the photocurrent from a
metal-insulator-metal (MIM) device.
22
One of the metallic contacts was reshaped into a plasmonic
strip antenna. It has been demonstrated that surface plasmon excitations can result in a favorable
redistribution in the electric fields in the stripe that enhances the photocurrent. Lang et al. have
reported plasmon resonant excitation of hot electrons in plasmonic grating structures of Au.
86
Therefore, wavelength-scale nanometallic structures, such as nanowires or strips, can be
considered for hot-electron photocatalysis.
Figure 6.1. Electron micrographs of (a) a silver bull’s eye and (b) a gold pyramid array .
87
6.2 Broad band plasmon resonant absorption
In plasmonic grating structures of Au, Lang et al. has observed peaks in the photocurrent at incident
angles of ±9° from normal when the incident light (633nm HeNe laser) is polarized parallel to the
(a) (b)
53
incident plane and perpendicular to the lines on the grating.
86
This angle dependence is due the
wavevector matching between the incident light and the plasmon resonant mode of the grating.
Instead of 633nm HeNe laser, if the plasmonic grating structures of Au is illuminated with a solar
simulator, light with different wavelengths will be resonant with various plasmon modes, therefore
the plasmonic grating structures of Au can absorb photons more efficiently. By varying the
incident angle of the sunlight using the same setup, the maximum photon absorption efficiency
can be reached.
Figure 6.2. Schematic diagram of angle dependent photocatalytic measurement.
6.3 Defect-mediated photocatalysis
Qiu et al. observed a larger Ti
3+
state density in thinner TiO2 films.
33
The density of Ti
3+
states
in ALD-deposited films can be manipulated by the number of cycles. By annealing TiO2 films in
NH3 gas, nitrogen defect states can be created in TiO 2.
45-47
Therefore, we can use the approach
mentioned in Chapter 4 to probe photocurrent spectra of TiO2 films with different thicknesses. We
can also compare the photocurrent spectra of O 2-annealed and NH3-annealed TiO2 films.
Sun simulator
Chopper
Rotational stage
Grating electrode Electrochemical cell
54
Figure 6.3. (a) Ti 2p and (b) O 1s core level XPS spectra of various thicknesses of TiO 2 on
GaAs.
33
6.4 Transient absorption spectroscopy
Transient absorption spectroscopy (TAS) is a pump-probe technique, in which a
femtosecond pump pulse is first used to optically excite hot electrons and hot holes in the
material, and then broadband absorption spectra are collected as a function of time. The
features in transient absorption spectra correspond to specific absorption mechanisms. TAS
can be used to study the interband transitions in bare metals and defect states in
dielectric/metal structures.
55
Figure 6.4. Transient absorption spectra of the Au grating structure
86
taken with (a) TM-
polarized and (b) TE-polarized light.
56
Bibliography
1. Kudo, A.; Miseki, Y., Heterogeneous photocatalyst materials for water splitting. Chemical
Society Reviews 2009, 38, 253-278.
2. Fujishima, A.; Honda, K., Electrochemical photolysis of water at a semiconductor
electrode. nature 1972, 238, 37.
3. Lee, M. H.; Takei, K.; Zhang, J.; Kapadia, R.; Zheng, M.; Chen, Y. Z.; Nah, J.; Matthews,
T. S.; Chueh, Y. L.; Ager, J. W., p‐Type InP nanopillar photocathodes for efficient solar‐driven
hydrogen production. Angewandte Chemie International Edition 2012, 51, 10760-10764.
4. Kudo, A.; Kato, H.; Nakagawa, S., Water splitting into H 2 and O2 on new Sr2M2O7 (M=
Nb and Ta) photocatalysts with layered perovskite structures: factors affecting the
photocatalytic activity. The Journal of Physical Chemistry B 2000, 104, 571-575.
5. Bard, A. J.; Fox, M. A., Artificial photosynthesis: solar splitting of water to hydrogen and
oxygen. Accounts of chemical research 1995, 28, 141-145.
6. Walter, M. G.; Warren, E. L.; McKone, J. R.; Boettcher, S. W.; Mi, Q.; Santori, E. A.;
Lewis, N. S., Solar Water Splitting Cells. Chemical Reviews 2010, 110, 6446-6473.
7. Linic, S.; Christopher, P.; Ingram, D. B., Plasmonic-metal nanostructures for efficient
conversion of solar to chemical energy. Nature materials 2011, 10, 911.
8. Xuan, Y.; Duan, H.; Li, Q., Enhancement of solar energy absorption using a plasmonic
nanofluid based on TiO2/Ag composite nanoparticles. RSC Advances 2014, 4, 16206-16213.
9. Qiu, J.; Zeng, G.; Pavaskar, P.; Li, Z.; Cronin, S. B., Plasmon-enhanced water splitting on
TiO2-passivated GaP photocatalysts. Physical Chemistry Chemical Physics 2014, 16, 3115-
3121.
10. Hu, S.; Shaner, M. R.; Beardslee, J. A.; Lichterman, M.; Brunschwig, B. S.; Lewis, N. S.,
Amorphous TiO2 coatings stabilize Si, GaAs, and GaP photoanodes for efficient water
oxidation. Science 2014, 344, 1005-1009.
11. Mukherjee, S.; Zhou, L.; Goodman, A. M.; Large, N.; Ayala-Orozco, C.; Zhang, Y.;
Nordlander, P.; Halas, N. J., Hot-electron-induced dissociation of H2 on gold nanoparticles
supported on SiO2. Journal of the American Chemical Society 2014, 136, 64-7.
12. Liu, Z.; Hou, W.; Pavaskar, P.; Aykol, M.; Cronin, S. B., Plasmon resonant enhancement
of photocatalytic water splitting under visible illumination. Nano letters 2011, 11, 1111-6.
13. Gomes Silva, C. u.; Juárez, R.; Marino, T.; Molinari, R.; Garcí a, H., Influence of excitation
wavelength (UV or visible light) on the photocatalytic activity of titania containing gold
nanoparticles for the generation of hydrogen or oxygen from water. Journal of the American
Chemical Society 2010, 133, 595-602.
14. Sundararaman, R.; Narang, P.; Jermyn, A. S.; Goddard III, W. A.; Atwater, H. A.,
Theoretical predictions for hot-carrier generation from surface plasmon decay. Nature
communications 2014, 5, 5788.
15. Brown, A. M.; Sundararaman, R.; Narang, P.; Goddard III, W. A.; Atwater, H. A.,
Nonradiative plasmon decay and hot carrier dynamics: effects of phonons, surfaces, and
geometry. ACS nano 2015, 10, 957-966.
16. Mukherjee, S.; Libisch, F.; Large, N.; Neumann, O.; Brown, L. V.; Cheng, J.; Lassiter, J.
B.; Carter, E. A.; Nordlander, P.; Halas, N. J., Hot electrons do the impossible: plasmon-
induced dissociation of H2 on Au. Nano letters 2013, 13, 240-7.
57
17. Awazu, K.; Fujimaki, M.; Rockstuhl, C.; Tominaga, J.; Murakami, H.; Ohki, Y.; Yoshida,
N.; Watanabe, T., A plasmonic photocatalyst consisting of silver nanoparticles embedded in
titanium dioxide. Journal of the American Chemical Society 2008, 130, 1676-1680.
18. Christopher, P.; Ingram, D. B.; Linic, S., Enhancing photochemical activity of
semiconductor nanoparticles with optically active Ag nanostructures: photochemistry
mediated by Ag surface plasmons. The Journal of Physical Chemistry C 2010, 114, 9173-9177.
19. Kumar, M. K.; Krishnamoorthy, S.; Tan, L. K.; Chiam, S. Y.; Tripathy, S.; Gao, H., Field
effects in plasmonic photocatalyst by precise SiO 2 thickness control using atomic layer
deposition. ACS Catalysis 2011, 1, 300-308.
20. Ingram, D. B.; Linic, S., Water splitting on composite plasmonic-metal/semiconductor
photoelectrodes: evidence for selective plasmon-induced formation of charge carriers near the
semiconductor surface. Journal of the American Chemical Society 2011, 133, 5202-5.
21. DuChene, J. S.; Sweeny, B. C.; Johnston-Peck, A. C.; Su, D.; Stach, E. A.; Wei, W. D.,
Prolonged hot electron dynamics in plasmonic-metal/semiconductor heterostructures with
implications for solar photocatalysis. Angewandte Chemie 2014, 53, 7887-91.
22. Chalabi, H.; Schoen, D.; Brongersma, M. L., Hot-electron photodetection with a plasmonic
nanostripe antenna. Nano letters 2014, 14, 1374-80.
23. Zheng, B. Y.; Zhao, H.; Manjavacas, A.; McClain, M.; Nordlander, P.; Halas, N. J.,
Distinguishing between plasmon-induced and photoexcited carriers in a device geometry.
Nature communications 2015, 6, 7797.
24. Robatjazi, H.; Bahauddin, S. M.; Doiron, C.; Thomann, I., Direct Plasmon-Driven
Photoelectrocatalysis. Nano letters 2015, 15, 6155-61.
25. Govorov, A. .; Zhang, H.; Gun’ko, Y. K., Theory of Photoinjection of Hot Plasmonic
Carriers from Metal Nanostructures into Semiconductors and Surface Molecules. The Journal
of Physical Chemistry C 2013, 117, 16616-16631.
26. Sundararaman, R.; Narang, P.; Jermyn, A. S.; Goddard, W. A., 3rd; Atwater, H. A.,
Theoretical predictions for hot-carrier generation from surface plasmon decay. Nature
communications 2014, 5, 5788.
27. Brown, A. M.; Sundararaman, R.; Narang, P.; Goddard, W. A., 3rd; Atwater, H. A.,
Nonradiative Plasmon Decay and Hot Carrier Dynamics: Effects of Phonons, Surfaces, and
Geometry. ACS nano 2016, 10, 957-66.
28. Narang, P.; Sundararaman, R.; Atwater, H. A., Plasmonic hot carrier dynamics in solid-
state and chemical systems for energy conversion. Nanophotonics 2016, 5.
29. Sakurai, H.; Haruta, M., Synergism in methanol synthesis from carbon dioxide over gold
catalysts supported on metal oxides. Catalysis Today 1996, 29, 361-365.
30. Brongersma, M. L.; Halas, N. J.; Nordlander, P., Plasmon-induced hot carrier science and
technology. Nature nanotechnology 2015, 10, 25-34.
31. Clavero, C., Plasmon-induced hot-electron generation at nanoparticle/metal-oxide
interfaces for photovoltaic and photocatalytic devices. Nature Photonics 2014, 8, 95-103.
32. Qiu, J.; Zeng, G.; Ha, M.-A.; Hou, B.; Mecklenburg, M.; Shi, H.; Alexandrova, A. N.;
Cronin, S. B., Microscopic Study of Atomic Layer Deposition of TiO2 on GaAs and Its
Photocatalytic Application. Chemistry of Materials 2015, 27, 7977-7981.
33. Qiu, J.; Zeng, G.; Ge, M.; Arab, S.; Mecklenburg, M.; Hou, B.; Shen, C.; Benderskii, A.
V.; Cronin, S. B., Correlation of Ti
3+
states with photocatalytic enhancement in TiO 2-
passivated p-GaAs. Journal of Catalysis 2016, 337, 133-137.
58
34. Xie, Q.; Jiang, Y.-L.; Detavernier, C.; Deduytsche, D.; Van Meirhaeghe, R. L.; Ru, G.-P.;
Li, B.-Z.; Qu, X.-P., Atomic layer deposition of TiO2 from tetrakis-dimethyl-amido titanium
or Ti isopropoxide precursors and H2O. Journal of Applied Physics 2007, 102, 083521.
35. Lim, G. T.; Kim, D.-H., Characteristics of TiOx films prepared by chemical vapor
deposition using tetrakis-dimethyl-amido-titanium and water. Thin Solid Films 2006, 498, 254-
258.
36. Pham, H. H.; Wang, L. W., Electronic structures and current conductivities of B, C, N and
F defects in amorphous titanium dioxide. Physical chemistry chemical physics : PCCP 2015,
17, 11908-13.
37. Manjavacas, A.; Liu, J. G.; Kulkarni, V.; Nordlander, P., Plasmon-Induced Hot Carriers in
Metallic Nanoparticles. Acs Nano 2014, 8, 7630-7638.
38. Avanesian, T.; Christopher, P., Adsorbate Specificity in Hot Electron Driven
Photochemistry on Catalytic Metal Surfaces. The Journal of Physical Chemistry C 2014, 118,
28017-28031.
39. Lu, G., Amy Linsebigler, and John T. Yates Jr., Ti
3+
Defect Sites on TiO2(110): Production
and Chemical Detection of Active Sites. The Journal of Physical Chemistry C 1994, 98, 11733-
11738.
40. Qiu, J.; Zeng, G. T.; Ge, M. Y.; Arab, S.; Mecklenburg, M.; Hou, B. Y.; Shen, C. F.;
Benderskii, A. V.; Cronin, S. B., Correlation of Ti
3+
states with photocatalytic enhancement in
TiO2-passivated p-GaAs. Journal of Catalysis 2016, 337, 133-137.
41. Qiu, J.; Zeng, G. T.; Ha, M. A.; Hou, B. Y.; Mecklenburg, M.; Shi, H. T.; Alexandrova, A.
N.; Cronin, S. B., Microscopic Study of Atomic Layer Deposition of TiO 2 on GaAs and Its
Photocatalytic Application. Chemistry of Materials 2015, 27, 7977-7981.
42. Zeng, G. T.; Qiu, J.; Hou, B. Y.; Shi, H. T.; Lin, Y. J.; Hettick, M.; Javey, A.; Cronin, S.
B., Enhanced Photocatalytic Reduction of CO2 to CO through TiO2 Passivation of InP in Ionic
Liquids. Chemistry-a European Journal 2015, 21, 13502.
43. Zeng, G.; Qiu, J.; Pavaskar, P.; Li, Z.; Cronin, S. B., CO 2 Reduction to Methanol on TiO2-
Passivated GaP Photocatalysts. ACS Catalysis 2014, 4, 3512.
44. Qiu, J.; Zeng, G.; Ha, M. A.; Ge, M.; Lin, Y.; Hettick, M.; Hou, B.; Alexandrova, A. N.;
Javey, A.; Cronin, S. B., Artificial Photosynthesis on TiO 2-Passivated InP Nanopillars. Nano
letters 2015, 15, 6177-81.
45. Linic, S.; Christopher, P.; Ingram, D. B., Plasmonic-metal nanostructures for efficient
conversion of solar to chemical energy. Nature Materials 2011, 10, 911-921.
46. Irie, H.; Watanabe, Y.; Hashimoto, K., Nitrogen-concentration dependence on
photocatalytic activity of TiO2-xNx powders. Journal of Physical Chemistry B 2003, 107, 5483-
5486.
47. Asahi, R.; Morikawa, T.; Ohwaki, T.; Aoki, K.; Taga, Y., Visible-light photocatalysis in
nitrogen-doped titanium oxides. Science 2001, 293, 269-271.
48. Nambu, A.; Graciani, J.; Rodriguez, J. A.; Wu, Q.; Fujita, E.; Sanz, J. F., N doping of
TiO2(110): Photoemission and density-functional studies. J Chem Phys 2006, 125, 8.
49. Thompson, T. L.; Yates, J. T., Control of a surface photochemical process by fractal
electron transport across the surface: O
-2
photodesorption from TiO2(110). J Phys Chem B
2006, 110, 7431-7435.
50. Diwald, O.; Thompson, T. L.; Zubkov, T.; Goralski, E. G.; Walck, S. D.; Yates, J. T.,
Photochemical activity of nitrogen-doped rutile TiO2(111) in visible light. J Phys Chem B 2004,
108, 6004-6008.
59
51. Diwald, O.; Thompson, T. L.; Goralski, E. G.; Walck, S. D.; Yates, J. T., The effect of
nitrogen ion implantation on the photoactivity of TiO 2 rutile single crystals. J Phys Chem B
2004, 108, 52-57.
52. Anpo, M.; Dohshi, S.; Kitano, M.; Hu, Y.; Takeuchi, M.; Matsuoka, M., The preparation
and characterization of highly efficient titanium oxide-based photofunctional materials.
Annual Review of Materials Research 2005, 35, 1-27.
53. Yamashita, H.; Harada, M.; Misaka, J.; Takeuchi, M.; Neppolian, B.; Anpo, M.,
Photocatalytic degradation of organic compounds diluted in water using visible light-
responsive metal ion-implanted TiO2 catalysts: Fe ion-implanted TiO2. Catalysis Today 2003,
84, 191-196.
54. Yamashita, H.; Harada, M.; Misaka, J.; Takeuchi, M.; Ikeue, K.; Anpo, M., Degradation
of propanol diluted in water under visible light irradiation using metal ion-implanted titanium
dioxide photocatalysts. Journal of Photochemistry and Photobiology A-Chemistry 2002, 148,
257-261.
55. Takeuchi, M.; Yamashita, H.; Matsuoka, M.; Anpo, M.; Hirao, T.; Itoh, N.; Iwamoto, N.,
Photocatalytic decomposition of NO on titanium oxide thin film photocatalysts prepared by an
ionized cluster beam technique. Catal Lett 2000, 66, 185-187.
56. Yamashita, H.; Harada, M.; Misaka, J.; Takeuchi, M.; Ichihashi, Y.; Goto, F.; Ishida, M.;
Sasaki, T.; Anpo, M., Application of ion beam techniques for preparation of metal ion-
implanted TiO2 thin film photocatalyst available under visible light irradiation: Metal ion-
implantation and ionized cluster beam method. Journal of Synchrotron Radiation 2001, 8, 569-
571.
57. Anpo, M., Utilization of TiO2 photocatalysts in green chemistry. Pure and Applied
Chemistry 2000, 72, 1265-1270.
58. Anpo, M., Use of visible light. Second-generation titanium oxide photocatalysts prepared
by the application of an advanced metal ion-implantation method. Pure and Applied Chemistry
2000, 72, 1787-1792.
59. Iino, K.; Kitano, M.; Takeuchi, M.; Matsuoka, M.; Anpo, M., Design and development of
second-generation titanium oxide photocatalyst materials operating under visible light
irradiation by applying advanced ion-engineering techniques. Current and Applied Physics
2006, 6, 982-986.
60. Takeuchi, M.; Yamashita, H.; Matsuoka, M.; Anpo, M.; Hirao, T.; Itoh, N.; Iwamoto, N.,
Photocatalytic decomposition of NO under visible light irradiation on the Cr-ion-implanted
TiO2 thin film photocatalyst. Catal Lett 2000, 67, 135-137.
61. Anpo, M.; Che, M., Applications of photoluminescence techniques to the characterization
of solid surfaces in relation to adsorption, catalysis, and photocatalysis. Advances in Catalysis
2000, 44, 119-257.
62. Choi, W. Y.; Termin, A.; Hoffmann, M. R., The role of metal-ion dopants in quantum-
sized TiO2 - Correlation between photoreactivity and charge-carrier recombination dynamics.
J Phys Chem-Us 1994, 98, 13669-13679.
63. Fleming, L.; Fulton, C.; Lucovsky, G.; Rowe, J.; Ulrich, M.; Lüning, J., Local bonding
analysis of the valence and conduction band features of TiO 2. Journal of Applied Physics
2007, 102, 033707.
64. Tao, J.; Luttrell, T.; Batzill, M., A two-dimensional phase of TiO2 with a reduced bandgap.
Nature chemistry 2011, 3, 296.
60
65. Hou, B.; Shen, L.; Shi, H.; Kapadia, R.; Cronin, S. B., Hot electron-driven photocatalytic
water splitting. Physical chemistry chemical physics : PCCP 2017, 19, 2877-2881.
66.Ager, J. W.; Jones, R. E.; Yamaguchi, D. M.; Yu, K. M.; Walukiewicz, W.; Li, S. X.; Haller,
E. E.; Lu, H.; Schaff, W. J., p-type InN and In-rich InGaN. physica status solidi (b) 2007, 244,
1820-1824.
67. Ager, J. W.; Miller, N.; Jones, R. E.; Yu, K. M.; Wu, J.; Schaff, W. J.; Walukiewicz, W.,
Mg-doped InN and InGaN – Photoluminescence, capacitance–voltage and thermopower
measurements. physica status solidi (b) 2008, 245, 873-877.
68. Wang, K.; Miller, N.; Iwamoto, R.; Yamaguchi, T.; Mayer, M. A.; Araki, T.; Nanishi, Y.;
Yu, K. M.; Haller, E. E.; Walukiewicz, W.; Ager, J. W., Mg doped InN and confirmation of
free holes in InN. Applied Physics Letters 2011, 98, 042104.
69. Miller, N.; Ager, J. W.; Smith, H. M.; Mayer, M. A.; Yu, K. M.; Haller, E. E.; Walukiewicz,
W.; Schaff, W. J.; Gallinat, C.; Koblmüller, G.; Speck, J. S., Hole transport and
photoluminescence in Mg-doped InN. Journal of Applied Physics 2010, 107, 113712.
70.Sato, N., Electrochemistry at Metal and Semiconductor Electrodes. Elsevier Science: 1998.
71. Lin, Y.; Kapadia, R.; Yang, J.; Zheng, M.; Chen, K.; Hettick, M.; Yin, X.; Battaglia, C.;
Sharp, I. D.; Ager, J. W.; Javey, A., Role of TiO2 Surface Passivation on Improving the
Performance of p-InP Photocathodes. Journal Of Physical Chemistry C 2015, 119, 2308-2313.
72. Hu, S.; Chi, C.-Y.; Fountaine, K. T.; Yao, M.; Atwater, H. A.; Dapkus, P. D.; Lewis, N. S.;
Zhou, C., Optical, electrical, and solar energy-conversion properties of gallium arsenide
nanowire-array photoanodes. Energy & Environmental Science 2013, 6, 1879.
73. Arab, S.; Anderson, P. D.; Yao, M.; Zhou, C.; Dapkus, P. D.; Povinelli, M. L.; Cronin, S.
B., Enhanced Fabry-Perot resonance in GaAs nanowires through local field enhancement and
surface passivation. Nano Research 2014, 7, 1146-1153.
74. Li, Z.; Chang, S.-W.; Chen, C.-C.; Cronin, S. B., Enhanced photocurrent and
photoluminescence spectra in MoS2 under ionic liquid gating. Nano Research 2014, 7, 973-
980.
75.Ager, J. W.; Jones, R. E.; Yamaguchi, D. M.; Yu, K. M.; Walukiewicz, W.; Li, S. X.; Haller,
E. E.; Lu, H.; Schaff, W. J., p-type InN and In-rich InGaN. Physica Status Solidi B-Basic Solid
State Physics 2007, 244, 1820-1824.
76. Brown, G. F.; Ager, J. W.; Walukiewicz, W.; Schaff, W. J.; Wu, J., Probing and modulating
surface electron accumulation in InN by the electrolyte gated Hall effect. Applied Physics
Letters 2008, 93.
77. Arab, S.; Chi, C.; Shi, T.; Wang, Y.; Dapkus, D. P.; Jackson, H. E.; Smith, L. M.; Cronin,
S. B., Effects of Surface Passivation on Twin-Free GaAs Nanosheets. ACS Nano 2015,
10.1021/nn505227q.
78. Chen, Y. W.; Prange, J. D.; Duhnen, S.; Park, Y.; Gunji, M.; Chidsey, C. E. D.; McIntyre,
P. C., Atomic layer-deposited tunnel oxide stabilizes silicon photoanodes for water oxidation.
Nature Materials 2011, 10, 539-544.
79. Qiu, J.; Zeng, G.; Pavaskar, P.; Li, Z.; Cronin, S. B., Plasmon-Enhanced Water Splitting
on TiO2-Passivated GaP Photocatalysts. Physical Chemistry Chemical Physics 2014, 16, 3115.
80. Lee, M. H.; Takei, K.; Zhang, J. J.; Kapadia, R.; Zheng, M.; Chen, Y. Z.; Nah, J.; Matthews,
T. S.; Chueh, Y. L.; Ager, J. W.; Javey, A., p-Type InP Nanopillar Photocathodes for Efficient
Solar-Driven Hydrogen Production. Angewandte Chemie-International Edition 2012, 51,
10760-10764.
61
81. Qiu, J.; Zeng, G.; Ge, M.; Lin, Y.; Javey, A.; Cronin, S. B., Artificial Photosynthesis on
TiO2-Passivated InP Nanopillars. 2015, submitted.
82. Tiwani, P.; Docampo, P.; Johnston, M. B.; Snaith, H. J.; Herz, L. M., Electron Mobility
and Injection Dynamics in Mesoporous ZnO, SnO 2, and TiO2 Films Used in Dye-Sensitized
Solar Cells. ACS nano 2011, 5, 5158-5166.
83. Woodall, J. M.; Freeouf, J. L., GaAs metallization: Some problems and trends. Journal of
Vacuum Science and Technology 1981, 19, 794-798.
84. Sze, S., Physics of Semiconductor Devices. 2006, Wiley, 3
rd
Edition.
85. Passlack, M.; Hong, M.; Mannaerts, J. P., Quasistatic and high frequency capacitance–
voltage characterization of Ga2O3–GaAs structures fabricated by in situ molecular beam
epitaxy. Applied Physics Letters 1996, 68, 1099-1101.
86. Shen, L.; Gibson, G. N.; Poudel, N.; Hou, B.; Chen, J.; Shi, H.; Guignon, E.; Cady, N. C.;
Page, W. D.; Pilar, A., Plasmon resonant amplification of hot electron-driven photocatalysis.
Applied Physics Letters 2018, 113, 113104.
87. Nagpal, P.; Lindquist, N. C.; Oh, S.-H.; Norris, D. J., Ultrasmooth patterned metals for
plasmonics and metamaterials. Science 2009, 325, 594-597.
88. Tornow, J.; Schwarzburg, K., Transient electrical response of dye-sensitized ZnO nanorod
solar cells. The Journal of Physical Chemistry C 2007, 111, 8692-8698.
62
Appendix A: Additional materials characterization of thin TiO 2 films
A.1 Frequency dependence
The frequency dependence of the AC photocurrent is plotted in Figure A.1 for a TiO2/Au
sample. The 5nm TiO2 film was annealed at 450°C for 30min while O 2 is flowing. Here, the
photocurrent originates from the sub-band gap defects within the TiO2, as described in the main
manuscript, and as evidenced by the UPS and XPS spectra. This data shows a very slow response
of these photoactive states, which is similar to what is observed in dye-sensitized solar cells.
88
Figure A.1. Frequency dependence of AC photocurrent for a TiO 2/Au sample.
63
A.2 Light intensity dependence
Figure A.2. Photocurrent obtained under (a) reduction and (b) oxidation conditions from a
TiO2/Au sample as a function of intensity of 532nm laser excitation.
64
A.3 Photoluminescence
Figure A.3. Photoluminescence (PL) spectra taken with (a) 532nm, (b) 633nm, and (c) 785nm
laser light. Here, we observe similar spectral features to those observed with 532nm light. In Figure
A.3(b), the PL above 1.96eV is cut off by the 633nm edge filter. In Figure A.3(c), the PL is cut off
by the 785nm edge filter. Several Raman peaks can be seen in the spectra, which are laser
wavelength dependent.
65
A.4 Transient absorption spectroscopy
Figure A.4 shows a time-resolved transient absorption (TA) spectrum on a 5nm TiO 2 film
deposited on 1mm thick CaF2 substrate. The measurement was performed with two laser beams.
The pump laser beam (285nm) was converted and optimized by optical parametric amplifier. The
white light (400-700nm) probe laser beam was produced by a CaF2 crystal. An excited-state
absorption is observed across all probe wavelengths. The band gap of TiO 2 is 3.2eV, which is
below the energy of the pump laser. Therefore, most of the pumped electrons were promoted to
the conduction band of TiO2. Due to the continuity of energy levels in the conduction band of TiO 2,
the TA signal shows no dependence on probe wavelength. Figure A.4(b) shows a line scan of the
TA signal. To produce a better signal to noise ratio, the signal was integrated from probe
wavelengths of 440nm to 680nm, where 550-585nm were excluded due to the pump scattering.
The scan shows two stages of decay. The first stage is from 0ps to 2ps, where most pumped
electrons decay rapidly from excited state of conduction band to the ground state, or conduction
band edge. The second stage is from 2ps and later, during which the conduction band electrons
recombine with holes in valence band. Since the band gap is relatively large (3.2eV), the decay is
much slower than the first stage.
66
Figure A.4. (a) Transient absorption measurement of 5nm TiO 2 on 1mm CaF2 substrate with
285nm pump laser and white light probe. (b) Line scan of (a) integrated over all wavelengths
(440nm to 680nm).
67
Appendix B: Details of calculations of surface recombination rate
The recombination rate is calculated by using:
𝑈 =
𝜎 𝑛 𝜎 𝑝 𝑣 𝑡 ℎ
𝑁 𝑡 (𝑛𝑝 − 𝑛 𝑖 2
)
𝜎 𝑛 (𝑛 + 𝑛 𝑖 exp (
𝐸 𝑡 − 𝐸 𝑖 𝑘𝑇
)) + 𝜎 𝑛 (𝑝 + 𝑛 𝑖 exp (
𝐸 𝑖 − 𝐸 𝑡 𝑘𝑇
))
where, n and p are given by:
𝑛 = 𝑁𝑐 exp (−
𝐸 𝐶 − 𝐸 𝐹 𝑘𝑇
) + ∆𝑛 = 𝑁𝑐 exp (−
𝐸 𝐶 − 𝐸 𝐹𝑛
𝑘𝑇
)
𝑝 = 𝑁𝑣 exp (−
𝐸 𝐹 − 𝐸 𝑉 𝑘𝑇
) + ∆𝑝 = 𝑁𝑣 exp (−
𝐸 𝐹𝑝
− 𝐸 𝑉 𝑘𝑇
)
Figure 5.4(a) in the manuscript was generated by using equations for U, n, and p as above with the
following parameters:
𝜎 𝑛 = 10
−13
cm
2
𝜎 𝑝 = 10
−13
cm
2
𝑣 𝑡 ℎ
= 4 × 10
7
cm/s
𝑁 𝑡 = 10
11
cm
-3
E
i
= 0.7 eV
68
𝐸 𝑔 = 1.424 eV
𝑇 = 300 K
𝑛 𝑖 = 2.1 × 10
6
cm
-3
𝑁 𝑎 = 5 × 10
17
cm
-3
𝛿 𝑝 = 𝛿 𝑛 = 0.9 × 10
13
cm
-3
𝑁 𝑐 = 4.7 × 10
17
cm
-3
𝑁 𝑣 = 9 × 10
18
cm
-3
.
Figure 5.4(c) was then generated by using the simulated values for band bending as a function of
applied bias.
The luminescence efficiency was calculated using the following equation:
Internal Radiative Efficiency =
Bpn
An+Sn+Bpn
in which:
A = 10
10
s
-1
S = U×10
-9
s
-1
B = 7.2×10
-10
cm
3
/s.
This was then plotted as a function of applied voltage using the same approach as before. The
parameters were chosen here to fit the absolute experimentally observed change in
photoluminescence.
Abstract (if available)
Abstract
Hot electrons have been reported in photocatalytic systems, introducing the exciting possibility of overcoming high barrier reactions. Hot electron processes have also been reported in solid state devices. However, there are several possible mechanisms capable of producing a photocurrent, which are difficult to separate. In the thesis presented here, bare Au surfaces and TiO₂-coated Au surfaces were studied in order to demonstrate the existence of hot electrons and holes in bare Au surfaces and photocatalytically active defect sites in TiO₂. ❧ Chapter 1 starts with a brief introduction of photocatalytic water splitting. Some significant advances in this field are listed later, along with the requirements for efficient and robust photocatalysts. ❧ In Chapter 2, we provide a basic description of plasmons and hot carriers, and comparison the near-field electromagnetic mechanism and charge injection mechanism for photocurrent enhancement. ❧ In Chapter 3, an AC photocurrent from bare Au surfaces is detected for both hydrogen evolution reaction (HER) and oxygen evolution reaction (OER) using an AC lock-in technique, which demonstrates the existence of hot electrons and holes unambiguously. For both HER and OER, the AC photocurrent exhibits a relatively narrow peak, as the monoenergetic distribution of hot electrons and holes photoexcited in the metal is swept through the resonance with the redox potential of the desired half-reaction. The photocurrent from TiO₂-coated metal surfaces is discussed in both Chapters 3 and 4. From the photocurrent spectra, we conclude that photocatalytically active sites in TiO₂ can be selectively detected no matter which metal substrate is chosen. These active defects sites are located at 2eV below the conduction band edge, which is evidenced by ultraviolet photoemission spectroscopy (UPS). ❧ In Chapter 5, a systematic study of photoluminescence (PL) spectroscopy of TiO₂-passivated GaAs as a function of electrochemical potential in an ionic liquid solution is presented. We observe a 7X increase in the PL intensity as the GaAs transitions from accumulation to depletion due to the applied potential. We attribute this to the excellent control over the surface Fermi level enabled by the high capacitance of the electrochemical double layer and TiO₂. ❧ In Chapter 6, we provide a summary of future research directions and outlook for hot-electron and defect-mediated photocatalysis.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Enhanced photocatalysis on titanium oxide passivated III-V semiconductors
PDF
Surface plasmon resonant enhancement of photocatalysis
PDF
Plasma emission across metal nanoparticle surfaces and semiconductor -liquid interface characterizations
PDF
Hot carrier enhanced photocatalysis in plasmon resonant metal grating systems
PDF
Artificial photosynthesis on titanium oxide passivated III-V semiconductors
PDF
Electromagnetic modeling of plasmonic nanostructures
PDF
Optoelectronic, thermoelectric, and photocatalytic properties of low dimensional materials
PDF
Hot electron driven photocatalysis and plasmonic sensing using metallic gratings
PDF
Fabrication and application of plasmonic nanostructures: a story of nano-fingers
PDF
Integrated photonics assisted electron emission devices
PDF
Plasmonic enhancement of catalysis and solar energy conversion
PDF
GaAs nanowire optoelectronic and carbon nanotube electronic device applications
PDF
Electronic and optoelectronic devices based on quasi-metallic carbon nanotubes
PDF
Light Emission from Carbon Nanotubes and Two-Dimensional Materials
PDF
Optical, mechanical, and electrical properties of nano-structured materials
PDF
Thermoelectric and transport studies of low dimensional materials & hot electron-driven photocatalysis on plasmon-resonant grating nanostructures
PDF
Monitoring local electric fields, charge densities at electrode surfaces using in-situ graphene/surface enhanced Raman spectroscopy (GERS/SERS) based Stark-shifts
PDF
Nano-engineered devices for display and analog computing
PDF
In-situ characterization of nanoscale opto-electronic devices through optical spectroscopy and electron microscopy
PDF
Targeted cell imaging by in-situ assembly and activation of hot spot SERS nanoprobes using split-fluorescent protein scaffolds
Asset Metadata
Creator
Hou, Bingya
(author)
Core Title
Hot carriers in bare metals and photocatalytically active defect sites in dielectric/metal structures
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Electrical Engineering
Publication Date
02/04/2019
Defense Date
11/30/2018
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
defect sites,hot electrons,OAI-PMH Harvest,photocatalysis,titanium dioxide,water splitting
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Cronin, Stephen Burke (
committee chair
), Nakano, Aiichiro (
committee member
), Wu, Wei (
committee member
)
Creator Email
bhou@usc.edu,natsu.akira8485@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-116392
Unique identifier
UC11675527
Identifier
etd-HouBingya-7040.pdf (filename),usctheses-c89-116392 (legacy record id)
Legacy Identifier
etd-HouBingya-7040.pdf
Dmrecord
116392
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Hou, Bingya
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
Tags
defect sites
hot electrons
photocatalysis
titanium dioxide
water splitting