UNIVERISTY OF TECHNOLOGY SYDNEY

THE EFFECT OF SURFACE DEFECTS ON

THE OPTICAL AND ELECTRICAL
PROPERTIES OF ZnO NANORODS


by

Suranan Anantachaisilp

A thesis submitted in partial fulfilment for

the degree of Doctor of Philosophy
in the
School of Mathematics and Physical Sciences
Faculty of Science

2015
 i


Declaration of Authorship

I, Suranan Anantachaisilp, declare that this thesis titled, “The Effect of Surface Defects
on the Optical and Electrical Properties of ZnO Nanorods” and the work presented in it
is my own. I confirm that: This work was done wholly or mainly while in candidature
for a co-tutelle research degree at this University and the Mahidol University.

ƒ Where any part of this thesis has previously been submitted for a degree or any
other qualification at this University and Mahidol University or any other
institution, this has been clearly stated.

ƒ Where I have consulted the published work of others, this is always clearly
attributed.

ƒ Where I have quoted from the work of others, the source is always given.

ƒ With the exception of such quotations, this thesis is entirely my own work.

ƒ I have acknowledged all main sources of help.

ƒ Where the thesis is based on work done by myself jointly with others, I have
made clear exactly what was done by others and what I have contributed
myself.

Signed:__________________________________________

Suranan Anantachaisilp

Date:____________________________________________

 ii


ABSTRACT

This thesis reports on the effect of native point defects on both the NH3 gas
sensing properties and the surface electronic structure of ZnO nanorods. Low
temperature hydrothermal growth at 90°C was utilised to synthesize ZnO nanorods
approximately 55 ± 5 nm in diameter. The type and density of intrinsic surface defects
on the ZnO nanorod was controlled using post growth annealing in O2 gas and Zn vapor
environments. Low voltage cathodoluminescence (CL) spectroscopy confirmed that the
heat treatment process produced different surface defect structures. The as-grown, O2,
and Zn annealed nanorods exhibited broad CL peaks centered at 1.90 eV (YL), 1.70 eV
(RL), and 2.44 eV (GL), which were attributed to O interstitials or LiZn deep acceptors,
acceptor-like VZn complexes, and donor-like VO related centers, respectively.
The first part of this thesis focuses on the influence of deep level surface
defects on NH3 gas response of ZnO nanorods. Electrical and gas sensing measurements
revealed that the NH3 gas sensitivity was 4.1 for the as-grown (YL), 22.6 for O2 anneal
(RL), and 1.4 for Zn vapor anneal (GL) samples. Hydrogen plasma treatment quenched
the RL and inverted the ammonia electrical response due to the incorporation of shallow
hydrogen donors. Changes to the gas sensing response were attributed to a shift in the
ZnO Fermi level position relative to the ammonia gas chemical potential due to the
formation near surface donor or acceptor centers.
In the second part of this thesis the effect of native point defects on the
surface electronic structure of all ZnO nanorod samples was studies utilizing X-ray
photoemission spectroscopy (XPS), X-ray Absorption Near-Edge Structure (XANES),
electron spin resonance (ESR), low voltage CL spectroscopy, sub band gap
photoluminescence (PL) spectroscopy and electrical I-V measurements. Correlative
characterization measurements established that the degree of surface band bending and
surface conductivity is controlled by the amount of adsorbed ambient gas species (O2,
H2O and OH), which is mediated by type and concentration of intrinsic surface point
defects.
The last part of this thesis concentrates on the chemical origin of RL peak
and its optical properties. A wide range of analysis techniques, including PL and CL
 iii


spectroscopy, XANES and ESR was used to comprehensively characterize the RL

emission. From these data, the RL has been assigned to highly lattice coupled VZn-
related acceptor-like center. No correlation was found between the RL and nitrogen
impurities.

KEY WORDS: ZnO NANORODS, HYDROTHERMAL GROWTH, SURFACE

ELECTRONIC PROPERTIES, GAS SENSING PREOPERTIES, RED
LUMINESCENCE
 iv


ACKNOWLEDGEMENTS

I would never have been able to start and finish my Ph.D. without the
guidance, help and support from Assoc. Prof. Siwaporn Meejoo Smith, Prof. Matthew
Phillips and Prof. Anthony Moon.
I would like to express my sincerest gratitude to my supervisors Prof.
Matthew Phillips and Assoc. Prof. Siwaporn Meejoo Smith for their valuable
suggestions and all important research skills.
I would like to give a special acknowledgements to my co-supervisor Prof.
Anthony Moon for his excellent and kind support. I would also like to extend my thanks
to my co-supervisor Asst. Prof. Tanakorn Osotchan for allowing me to work in his
laboratory and for his advice in this research. My gratitude is also extended to Assoc.
Prof. Prapin Wilairat, my co-supervisor, as well as Dr. Cuong Ton-That for his
comments and suggestions.
I would also like to thanks the external examiners Asst. Prof. Sukkaneste
Tungasmita and Dr. Jedsada Manyam for their valuable time.
I would also like to express my gratitude to Mark Lockrey, Sumin Choi,
Pakorn Pasitsuparoad, Toby Shanley, MAU staffs, my fellow research students and
colleagues in the MAU and SMJ Lab for their help with sample preparation and data
collection, acquisition and analysis, and for contributing a pleasant atmosphere for doing
research.
I would like to acknowledge The Royal Golden Jubilee Ph.D. Program and
Australian Research Council for scholarship and financial support.
Finally, I would like to express my deepest thanks to my parents and two
elder brothers for their understanding, support, and lots of useful advice.

Suranan Anantachaisilp
 v


CONTENTS

Page
Declaration of Authorship i
ABSTRACT ii
ACKNOWLEDGEMENTS iv
LIST OF PUBLICATIONS x
LIST OF TABLES xi
LIST OF FIGURES xii
LIST OF SCHEMES xx
LIST OF ACRONYMS xxi
CHAPTER I INTRODUCTION TO ZINC OXIDE 1
1.1 Motivation of the Research 2
1.2 Objectives of the Research 4
1.3 Thesis Outline 5
CHAPTER II OPTICAL, ELECTRICAL, AND GAS SENSING 6
PROPERTIES OF ZnO AND ITS SYNTHESIS: LITERATURE REVIEWS
2.1 Physical and Chemical Properties of ZnO 6
2.1.1 Crystal Structure 7
2.1.2 Electronic Band Structure 8
2.1.3 Defect Chemistry 9
2.1.3.1 Vacancies 12
2.1.3.2 Interstitials 14
2.1.3.3 Antisites 15
2.1.3.4 Impurities 15
2.1.3.5 Formation Energies and Processes of 16
Native Defects
2.2 Optical Properties and Processes in ZnO 19
2.2.1 Excitons 20
 vi


CONTENTS (cont.)

Page
2.2.2 Radiative Recombination 21
2.2.2.1 Near Band Edge Emission 21
2.2.2.2 Defect-Related Deep Level Emission 22
2.2.2.2.1 Green Luminescence 24
2.2.2.2.2 Yellow Luminescence 25
2.2.2.2.3 Red Luminescence 25
2.2.3 Temperature-Dependence Luminescence 26
2.3 Electrical Properties of ZnO 26
2.3.1 Ohmic Contacts to ZnO 27
2.3.2 Schottky Contacts to ZnO 28
2.4 Gas Sensing Properties of ZnO 28
2.4.1 Gas Sensor Characteristics 29
2.4.1.1 Sensitivity 29
2.4.1.2 Selectivity 29
2.4.1.3 Response Time and Recovery Time 30
2.4.2 Gas Sensing Mechanisms of ZnO 31
2.4.2.1 Oxidative Gas 32
2.4.2.2 Reductive Gas 33
2.4.2.2.1 Ammonia Gas 34
2.5 Synthesis of ZnO Nanostructures 34
2.5.1 High Temperature Approaches 35
2.5.2 Low Temperature Approaches 35
2.5.2.1 Hydrothermal Growth 35
CHAPTER III CHEMICALS AND EXPERIMENTAL DETAILS 38
3.1 Chemicals and Instruments 38
3.2 Sample Preparation 38
 vii


CONTENTS (cont.)

Page
3.2.1 ZnO Nanostructures Growth 39
3.2.1.1 Seed Layer Deposition 39
3.2.1.2 Hydrothermal Growth 40
3.2.2 Post Growth Processing 40
3.2.2.1 Annealing 40
3.2.2.2 Hydrogen Plasma Treatment 41
3.3 Scanning Electron Microscopy (SEM) 42
3.4 X-Ray Diffraction (XRD) 43
3.5 Cathodoluminescence Spectroscopy (CL) 44
3.6 Photoluminescence Spectroscopy (PL) 46
3.7 Synchrotron Light Experiments 46
3.7.1 X-Ray Photoelectron Spectroscopy (XPS) 47
3.7.2 X-Ray Absorption Near Edge Spectroscopy (XANES) 47
3.8 Electrode Deposition and Gas Sensor Device Assembly 48
3.9 Current Voltage Characteristic Measurement 49
3.10 Gas Sensing Measurement 49
3.10.1 Gas Measurement System 50
3.10.2 Designed experimental procedures 51
CHAPTER IV TAILORING DEEP LEVEL SURFACE DEFECTS IN ZnO 53
NANORODS FOR HIGH SENSITIVITY AMMONIA GAS
SENSING
4.1 Summary 53
4.2 Introduction 53
4.3 Experimental Section 55
4.3.1 Synthesis of ZnO Nanorods and Post-Growth 55
Processing
 viii


CONTENTS (cont.)

Page
4.3.2 Analysis Techniques 55
4.3.3 Fabrication of ZnO Nanorods Gas Sensor and Gas 56
Sensing Measurement
4.4 Results and Discussion 56
4.5 Conclusions 66
CHAPTER V THE SURFACE ELECTRONIC STRUCTURE OF ZnO 68
NANORODS: THE ROLE OF SURFACE DEFECTS
5.1 Summary 68
5.2 Introduction 68
5.3 Experimental Section 69
5.4 Results and Discussion 71
5.5 Conclusions 83
CHAPTER VI NATURE OF RED LUMINESCENCE IN OXYGEN 84
TREATED HYDROTHERMALLY GROWN ZINC OXIDE
NANORODS
6.1 Summary 84
6.2 Introduction 84
6.3 Experimental Section 85
6.3.1 Preparation and Post Growth Processing of ZnO 85
Nanorods
6.3.2 Characterization of Hydrothermally Grown ZnO 86
Nanorods
6.3.2.1 Cathodoluminescence 86
6.3.2.2 Photoluminescence 86
6.3.2.3 X-ray Absorption Near Edge 87
Spectroscopy (XANES)
6.3.2.4 Electron Spin Resonance 87
 ix


CONTENTS (cont.)


Page
6.4 Results and Discussion 87
6.5 Conclusions 99
CHAPTER VII CONCLUSIONS 100
REFERENCES 102



 x


LIST OF PUBLICATIONS

Refereed Journal Publications

S. Anantachaisilp, S.M. Smith, C. Ton-That, S. Pornsuwan, A.R. Moon, C. Nenstiel, A.
Hoffmann, M.R. Phillips, Nature of Red Luminescence in Oxygen Treated
Hydrothermally Grown Zinc Oxide Nanorods, Journal of Luminescence, 168 (2015) 20-
25.

S. Anantachaisilp, S.M. Smith, C. Ton-That, T. Osotchan, A.R. Moon, M.R. Phillips,

Tailoring Deep Level Surface Defects in ZnO Nanorods for High Sensitivity Ammonia
Gas Sensing, The Journal of Physical Chemistry C, 118 (2014) 27150-27156.

Matthew R. Phillips, Suranan Anantachaisilp, Liangchen Zhu, Laurent L. Cheong Lem,

Christian Nenstiel, Siwaporn Meejoo Smith, Cuong Ton-That, Axel Hoffmann, Visible
Luminescence in Bulk and Nanostructured ZnO (Invited Paper), 2014 SPIE Photonics
West, San Francisco, California, USA. February 1–6, 2014.

Conference Presentations
S. Anantachaisilp, S.M. Smith, C. Ton-That, T. Osotchan, A.R. Moon, M.R. Phillips,
ZnO Nanowires with Enhanced Sensitivity for Ammonia Gas Sensing at Room
temperature by Regulating Deep Level Surface Defects. 2014 MATERIALS
RESEARCH SOCIETY Fall Meeting & Exhibit, Boston, Massachusetts. November 30-
December 5, 2014.

Matthew Phillips, Suranan Anantachaisilp, Liangchen Zhu, Christian Nenstiel, Nadja

Jankowski, Siwaporn Meejoo Smith, Cuong Ton-That and Axel Hoffrnann, Chemical
Origin of Red Luminescence in Zn. 8th International Workshop on Zinc Oxide and
Related Materials, Niagara Falls, Ontario, Canada. September 8 - 11, 2014.
 xi


LIST OF TABLES

Table Page
1.1 Summary of ammonia gas sensor applications Taken from Timmer et al. 3
2.1 General physical and chemical properties of ZnO 6
2.2 Bound exciton peak energies (eV) in ZnO 24
2.3 Summary of hydrothermal growth of ZnO nanostructures and their morphology 36
3.1 List of Chemicals 38
3.2 List of Instruments 39
5.1 The relative percentage of the integrated intensity of the OI, OII and OIII peaks 77
from the O1s XPS emission for the as-grown, Zn annealed and O2 annealed
specimens.
 xii


LIST OF FIGURES

Figures Page
2.1 Unit cells of ZnO crystal structures: (a) cubic rocksalt (B1), (b) cubic 8
zinc blende (B3), and (c) hexagonal wurtzite (B4). Zn atoms are
represented in gray whereas O atoms are represented in black.
(a) The calculation of ZnO band structure using the HSE hybrid 9
2.2 functional. The valence-band maximum energy is at zero. (b) Hexagonal
ZnO band structure and symmetries. The valence band splits into three
sub-bands; A, B, and C resulting from the effects of crystal-field splitting
and spin-orbit coupling.
2.3 Illustration of point defects in ZnO crystal structures: (a) Zinc vacancy 12
ܸ௓௡ , (b) Oxygen vacancy ܸை , (c) Zinc interstitial ܼ݊௜ , (d) Oxygen
interstitial ܱ௜ , (e) Zinc antisite ܼ݊ை , (f) Oxygen antisite ܱ௓௡ and (g)
Substitutional impurity at a zinc site e.g. ‫݅ܮ‬௓௡ , ‫ݑܥ‬௓௡ or ‫݈ܣ‬௓௡ .
2.4 Calculated defect formation energy for ZnO native point defects in (a) 13
Zinc-rich conditions and (b) Oxygen-rich conditions.
2.5 Calculated ZnO native point defects formation energies as a function of 18
Fermi-level position in both Zn-rich (left) and O-rich (right) conditions.
Valence band maximum locates at the zero of Fermi level. Different slope
represents different charge state.
2.6 Thermodynamic transition levels for ZnO native point defects. 19
2.7 Characteristic ZnO luminescence spectrum presenting the near band edge 21
peak (NBE) and the broad deep level emission (DL).
2.8 Schematic diagram of the radiative recombination mechanisms. 23
2.9 Characteristic current-time sensing peak of ZnO nanorods annealed in O2 30
atmosphere response to ammonia gas.
2.10 Schematic diagram of oxygen adsorption on the ZnO surface; n = 1 or 2 32
and s = 1or 2. (Adapted from Kim et al.)
 xiii


LIST OF FIGURES (cont.)

Figures Page
2.11 Schematic diagram of the oxidative gas sensing mechanism on ZnO. 33
surface. (Green area is the depletion layer.)
2.12 Schematic diagram of the reductive gas sensing mechanism on ZnO 33
surface. (Green area is the depletion layer.)
2.13 Schematic diagram of ammonia gas sensing mechanism on the ZnO 34
surface; n = 1 or 2 and s = 1or 2.
3.1 Photograph of Teflon Lined Autoclave used as a reactor in hydrothermal 40
growth.
3.2 Photograph of Lindberg Blue Mini-mite Tube Furnace. 41
3.3 Photograph of plasma chamber showing the overall setup. 42
3.4 Penetration range of the primary electron beam with the origin of 43
electrons, x-rays and light.
3.5 Schematic of CL experimental setup. 45
3.6 (a) Twelve fingers of the interdigitated electrode (shown in gray). (b) Gas 50
sensor device comprised of ZnO sensing layer (light gray square),
aluminum electrodes (gray), polycarbonate board (outermost square) with
conduction line (black line), and connection wires (black dot).
3.7 Gas flow controller unit diagram; input channels are shown in dash lines 51
and output channels are shown in solid lines.
3.8 Homebuilt gas measurement system; dot line represents shortcut path 52
when the dry measurement starts.



 xiv


LIST OF FIGURES (cont.)

Figures Page
4.1 (a) SEM images of as-grown ZnO ‫ۦ‬0001ۧ nanorods on a sapphire substrate 57
and (b) PXRD patterns of as-grown, O2 and Zn atmosphere annealed ZnO
nanorods exhibiting the wurtzite crystal structure. The change in the
relative intensity of the diffraction peaks arises from a slight reorientation
of the nanorods relative to the substrate after thermal annealing.
4.2 SEM images of ZnO nanorods after annealing in (a) O2 and (b) Zn 58
atmosphere.
4.3 (a) Normalized CL spectra collected from Zn annealed, as-grown, and O2 59
atmosphere annealed ZnO nanorods at 5 kV, 80 K, scan area = 50 ȝm ×
43 ȝm, and beam current of 0.2 ± 0.02 nA. The ratio of raw intensities of
GL, YL, and RL centered at 2.44, 1.90, and 1.70 eV, respectively, is
0.057:0.204:1. All spectra were normalized to the DL peaks. (b) Room
temperature IíV characteristics of the Zn annealed, as-grown, and O2
atmosphere annealed ZnO nanorods measured under an ambient
atmosphere.
4.4 N K-edge XANES spectra of Nitrogen implanted ZnO crystal and ZnO 61
nanoparticles, and as grown, O2, and Zn atmosphere annealed ZnO
nanorods.
4.5 CL spectra collected from O2 atmosphere annealed and H plasma of O2 62
annealed ZnO nanorods at 5kV, 80K. The spectra were normalized to the
maximum intensity peaks.
4.6 Room temperature I-V characteristic of the Argon atmosphere annealed 64
ZnO nanorod gas sensor.
 
 xv


LIST OF FIGURES (cont.)

Figures Page
4.7 Sensitivity as a function of measurement time for (a) as-grown, (c) O2, (e) 65
Zn, and (g) H plasma after O2 annealed ZnO nanorod sensors and
sensitivity as a function of ammonia gas concentration for (b) as-grown,
(d) O2, (f) Zn, and (h) H plasma after O2 annealed ZnO nanorods sensors
at room temperature. Labels in the bottom right corner of (b), (d), (f), and
(h) indicate the signature luminescence (YL, RL, GL, and RL
suppression) exhibited by ZnO nanorods employed in the fabrication of
each gas sensor.
5.1 SEM images of the as-grown hydrothermal ZnO nanorod ensemble 71
containing a high density of uniform hexagonal <0001> nanorods
50 ± 5 nm in diameter. The oblique nanorod growth habit promotes
multiple contact of the sidewalls providing good electrical conductivity
across the ensemble.
5.2 Typical 5 kV CL spectra collected from as-grown, O2 atmosphere and Zn 72
vapor annealed ZnO nanorods at 80K, a scan area = 50 Pm x 43 Pm and
beam current of 0.2±0.02nA, showing yellow (YL) at 1.90 eV, green (YL)
at 2.44 eV and red (RL) at 1.70 eV.
5.3 presents a log I vs log V showing the 1000x and 100x increase in the 73
measured current after Zn vapor and O2 annealing respectively.
5.4 (a) ESR spectra of the as-grown and O2 annealed ZnO nanorods. Two 74
main paramagnetic ESR signals at g # 2.01 and 1.96 are assigned to singly
ionized surface oxygen vacancy defects and localized donors,
respectively. (b) ESR spectrum following Zn vapor annealing showing a
significant intensity increase is g # 2.00 line after the appearance of a
broad green CL emission centered at 2.44 eV and a pronounced shoulder
in the valence band photoemission spectrum ~ 3 eV below the Fermi level.
 xvi


LIST OF FIGURES (cont.)

Figures Page
5.5 The O1s XPS spectra for YL, RL and GL ZnO nanorod samples are 76
shown in Figures 5.5(a), 5.5(b) and 5.5(c) respectively. The XPS profiles
are all curve fitted into three peaks at 530.5 eV (Peak OI), 532.0 eV (Peak
OII) and 532.7 eV (Peak OIII). Peak OI results from O2- ions fully co-
ordinated by Zn and O atoms in the fully stoichiometric wurtzite crystal
structure, peak OII is due to either O2- ions in the vicinity of VO defects
and peak OIII arises from oxygen in strongly bound adsorbed gases,
specifically H2O and O2 molecules.
5.6 Valence band X-ray Photoemission Spectroscopy (VB-XPS) spectra 80
taken from the as-grown, O2 anneal and Zn vapor anneal ZnO nanorod
samples are presented in Figure. 5.6(a). The VB-XPS spectra for all three
samples consist of two dominant near edge peaks one centered at ~ 5 eV
ascribed to O 2p related states the other ~ 7 eV attributed to hybridized
Zn 4s + O 2p states. The leading edge of the valence band photoemission
showing the pronounced shoulder around 3 eV below the Fermi level that
forms after Zn vapour annealing is shown in Figure 5.6(b).
5.7 Downward band bending on the as-grown ZnO nanorod due to electron 81
accumulation and its decrease following O2 annealing using the valence
band photoemission spectroscopy results in Figure 5.6 (a) and 5.6 (b).
5.8 Comparison of continuous-wave 532 and 405 nm sub band gap laser 82
excitation PL spectra at 300 K for ZnO nanorods as-grown and annealed
in O2 gas. The sub band gap excited spectra reveal a common broad OL
emission at 1.95 eV except for the O2 annealed sample illuminated with
blue light which emits a broad RL around 1.83 eV indicated a different
defect structure.
 xvii


LIST OF FIGURES (cont.)

Figures Page
6.1 SEM image of the as-grown <0001> ZnO nanorod ensemble, showing 88
that the low temperature hydrothermal growth treatment produced a high
density of uniform hexagonal <0001> nanorods approximately 55 nm in
diameter orientated at different angles to the normal direction of the
substrate.
6.2 Typical cathodoluminescence spectra collected with 5 kV, IB = 0.2 ± 90
0.02 nA and scan area = 50 Pm u43 Pm. Fig 6.2a ZnO nanorods before
(“as-grown”) and after annealing at 650 qC in O2 gas for 30 minutes,
showing the creation of the red luminescence (RL) peak centered at
1.78 eV with a FWHM = 0.69 eV. Fig 6.2b illustrates rapid quenching of
the RL following exposure to a mild hydrogen plasma indicating that the
RL relates to an acceptor-like center as hydrogen is a donor in ZnO.
Fig 6.2c. RL spectrum at 15 K showing the peak red shifted to 1.69 eV
and narrowing of the FWHM to 0.57 eV. The assignment of the RL to
transition metal impurities is ruled out because of the absence of fine
structure arising from internal 3d transitions.
6.3 Normalized RL intensity versus excitation power (5 kV x IB) at 80 K (a) 91
and 300 K (b). The absence of a blue peak shift with increasing beam
current eliminates the assignment of the RL to donor acceptor pair
recombination transitions.
6.4 Plot of log (ICL) versus log (IB) at 80 K (a) and 300 K (b) with IB in nA. 92
The gradient, n, is the power law exponent in the relationship, ICL v IBn.
The nNBE ~ 1 is expected for excitonic recombination. The nRL ~ 0.8 is
higher than expected for a lattice coupled defect most likely due to a high
concentration of RL centers.
 xviii


LIST OF FIGURES (cont.)

Figures Page
6.5 CL intensity versus temperature at 80 K for the RL (a) and NBE (b) with 94
300 K with 5 kV, IB = 0.2 ± 0.02 nA and scan area = 50 Pm u 43 Pm.
The red shift and narrowing of the RL with decreasing temperature is
consistent with configuration coordinate model for radiative
recombination. The blue shift and appearance of phonon replicas as the
temperature approaches 93 K is expected for excitonic emission with the
bandgap shrinkage. The intensity of the RL and NBE exhibits “negative”
thermal quenching with increasing temperature.
6.6 The plot of the RL peak position and peak width as a function of 96
temperature are shown in Figure 6.6a and Figure 6.6b. Figure 5.6c
displays the RL intensity dependence on temperature. With decreasing
temperature from 300 K to 93 K a typical thermal response relationship
with an activation energy of EA = 78.8 meV is found. However,
increasing the temperature from 93 K to 300 K “negative” thermal
quenching is observed arising from the presence of shallow levels that
fill with carriers at low temperature during electron beam excitation and
depopulate with rising temperature increasing the luminescence
intensity.
6.7 CL spectra at 80 K of the NBE and RL before and after thermal annealing 97
in O2 are shown in Figure 6.7a with corresponding CL spectra in
Figure 6.7b. Figure 6.7a reveals emission due to donor bound exciton
around 3.36 eV and peaks at 3.32 and 3.26 eV ascribed to the presence
of acceptors in either DAP or free-to-bound transitions. No
interrelationship between the 3.32 and 3.26 eV NBE peaks and the RL
in the CL spectra is observed. Additionally there is no correlation
between the free-to-bound NBE and RL peaks and nitrogen as confirmed
in the N kD XANES results in Figure 6.9.
 xix


LIST OF FIGURES (cont.)

Figures Page
6.8 The temperature dependence of the DoX and peaks at 3.32 and 3.26 eV 98
are shown in Figure 6.8. The data show the usual red shift of the DoX
with increasing temperature as the band gap widens due to lattice
expansion and the electron-phonon effect. The absence of a
corresponding shift of the 3.32 and 3.26 eV peaks between 7 and 100 K
indicates that free carriers are involved in the recombination process
providing thermal energy (kbT) which offsets the band gap red shift. This
result suggests that these emission peaks arise from free-to-bound
transition rather than a DAP recombination process.
6.9 XANES spectra around the N energy from the “as-grown” and O2 99
annealed samples, as well as from ZnO nano-particles implanted with
nitrogen, are shown in Figure 6.9. The results show that no nitrogen is
present in both the “as-grown” and O2 annealed samples.

 xx


LIST OF SCHEMES

Schemes Page
4.1 The alteration of the ZnO Fermi level position owing to near surface acceptor 66
and donor


 xxi


LIST OF ACRONYMS


A0 Neutral acceptor

BB Band bending
BX Bound exciton

CB Conduction band
CC Configurational coordinate
CCD Charge-coupled device
CL Cathodoluminescence

D0X Neutral donor bound exciton

DAP Donor-acceptor pair
DL Deep level

EPR Electron paramagnetic resonance

ESR Electron spin resonance

FBT Free-to-bound transition

FWHM Full width at half maximum
FX Free exciton

GL Green luminescence

HMT Hexamethylene tetramine

I-V Current-Voltage
 xxii


LIST OF ACRONYMS (cont.)

LO Longitudinal optical

NBE Near band edge

OL Orange luminescence

PL Photoluminescence

RL Red luminescence

S Sensitivity
sccm standard cubic centimetres per minute
SEM Scanning electron microscopy
TEY Total electron yield
TFY Total fluorescence yield

UV Ultra-violet

VB Valence band

XANES X-ray absorption near edge spectroscopy

XPS X-ray photoelectron spectroscopy
XRD X-Ray diffraction

YL Yellow luminescence
University of Technology Sydney Introduction / 1

CHAPTER I
INTRODUCTION TO ZnO

Metal-oxide semiconductors are widely acknowledged as crucial

technological materials for the development of novel electronic and photonic devices.1
ZnO is one of the metal oxide semiconductors which has attracted great attention in this
field because of its wide band gap (3.44 eV at low temperature and 3.37 eV at room
temperature), large exciton binding energy (60 meV) and the ease of growth of its
nanostructures on a variety of substrates. Consequently, ZnO has emerged as an
important potential candidate as a functional material applicable in variety of novel nano-
devices.2-4 Furthermore, ZnO nanomaterials of different morphologies; such as
nanorods, nanowires, nanotubes, nanobelts and nanoparticles, show great potential as
enabling materials in broad range of applications, such as field effect transistors,
optically pumped lasers, photodetectors, optical switches and sensors due to their unique
and exciting electronic, optical and piezoelectric properties.5 In addition, ZnO
nanostructures have also received much attention for chemical detection applications
owing to the strong sensitivity of its surface conductivity to the presence of adsorbed
species.6 Gas sensors, pressure sensors and biosensors are based on the change in
conductance with the reversible chemisorption process of reactive gases on the ZnO
surface, piezoelectric property, biocompatibility and nontoxicity of ZnO, respectively.7
However, numerous promising applications of ZnO critically depend upon
the development of growth and processing techniques to control the concentration and
distribution of the bulk and surface point defects 8-9 that mediate the optical and electrical
properties.10 Of equal importance is the establishment of methods to regulate the surface
morphology of ZnO nanostructures because it is well known that the surface-to-volume
ratio also has a significant effect on the material’s physical properties.
Suranan Anantachaisilp Introduction/ 2


1.1 Motivation of the Research

One of the most widely emitted gases in our atmosphere is ammonia (NH3)
gas. It is emitted from both its natural origin as well as direct and indirect human
activities. There are variety of ammonia sources including the chemical industry. It is
commonly utilized in the production of petrochemicals, fertilizer and paper as well as
used as a refrigerant in commercial refrigeration, food and beverage industry, air-
conditioning, pharmaceuticals production, and in power plants. However, it has a
substantial effect on both environment and our health which need to be concern. It is
one cause of global warming. In addition, it has a high toxicity when anhydrous
ammonia mixes with water. The human respiratory system has a high moisture content
and therefore anhydrous ammonia is easily mixed with water and irritates or burns the
respiratory tract depending on the concentration. It also causes chronic lung disease.
The level of ammonia in all industries needs to be detected for which an ammonia gas
sensor has been used. An ammonia gas sensor is also used in medical applications. For
example, some diseases such as kidney disorders or stomach ulcers caused by
Helicobacter pylori bacterial stomach infection, are diagnosed by measuring the
ammonia levels in exhaled air.11-13 Table 1.1 shows a summary of ammonia gas sensor
applications.
Recently, semiconducting metal oxide materials have been extensively
utilized as the gas sensors due to their low cost. ZnO is one of the most promising
ammonia gas sensors since it is chemically and thermally stable. It operates on the
change of conductance principle after gas molecules are adsorbed on the sensing layers.
However, use of ZnO as an ammonia gas sensor still need development. In some
ammonia gas sensor applications, high sensitivity, short response time and ability to
operate at low temperatures is required, as shown in Table 1.1. In order to improve the
electrical response of ZnO nanostructures, the nanostructural properties (morphology,
particle size and surface area) as well as the carrier concentration need to be
controlled.14-18 In addition, the surface point defects have a significant effect on the gas
sensitivity. However, the exact role of the surface point defects on the surface
conductivity remains controversial. This research therefore focuses mainly on a study
of the relationship between the surface point defects and the ammonia gas sensing
properties.
University of Technology Sydney Introduction / 3

Table 1.1: Summary of ammonia gas sensor applications Taken from Timmer et al.19
Require Temperature
response range Remarks
Application Detection limit time

Environmental
Monitoring ambient condition 0.1 ppb to >200ppm20 Minutes 0-40 °C Reduce environmental pollution
Measure in stables 1 to >25 ppm21 ~ 1 min 10-40 °C Protect livestock animals and farmers
Automotive
Measure NH3 emission from 4->200 g/min22 Seconds Up to 300 °C NH3 emission is not regulated at this
vehicles (concentration time
Passenger cabin air control unknown) ~1 s 0-40 °C Automotive air quality sensor mainly
50 ppm23 aim on NOX and CO levels21
Detect ammonia spill Seconds Up to 600 °C Control Urea injection in SCR NOX
1-100 ppm24 reduction
Chemical
Leakage alarm 20->1000 ppm23, 25 Minutes Up to 500 °C Concentrations can be very high at
Medical NH3 plants and can even be explosive
Breath analysis 50-2000 ppb11, 13 ~ 1 min 20-40 °C Diagnosis of peptic ulcer cause by
bacteria, small gas volumes

The surface electronic structure of all samples with different types of defects
are also studied in order to understand the mechanisms of gas adsorption as well as the
interactions between native point defects and ambient gas species. These provide more
information about the role of surface point defects which are useful for gas sensor
applications.
The O2 annealed ZnO sample emitted red luminescence (RL) provides the
highest NH3 gas sensitivity and the chemical origin of RL is still under debate, the
chemical origin of this peak and its optical properties were comprehensively studied.

Suranan Anantachaisilp Introduction/ 4


1.2 Objectives of the Research

This research mainly focuses on the systematic introduction of different
types of surface point defects in ZnO nanorods through controlled growth and thermal
processing in oxidizing (O2) and reducing (Zn vapor) atmosphere. ZnO nanorods having
surface defects of various types were further employed for fabrication of gas sensing
devices and their electrical responses to gas exposure probed and compared with the
native point defect on the surface. In this work, ammonia gas was employed as a model
of a reductive gas in order to illustrate the inter-relationships between deep level surface
defects, electrical properties, and sensing performance in ZnO nanorods.
The following are the summary of the main objectives of this research:
x To introduce various types of surface point defects into low temperature
hydrothermally grown ZnO nanorods systematically
x To investigate the influence of the deep level surface point defects on
electrical and gas sensing properties of ZnO nanorods
x To investigate the effect of native point defects on the surface electronic
structure of ZnO nanorods
x To investigate the chemical origin of the RL peak and its optical
properties
University of Technology Sydney Introduction / 5

1.3 Thesis Outline

This thesis mainly focuses on the role of surface defects on the physical
properties in ZnO nanorods. The thesis is structured as follow:
Chapter 1 gives a general introduction to ZnO and provides a brief overview
of utilizing ZnO as a semiconductor gas sensor. In addition, it provides the motivation
as well as the objectives of the research. The final section of this chapter presents the
outline of the thesis.
A literature review is provided in chapter 2. It presents some fundamental
properties of ZnO, including physical, chemical, optical, electrical, and gas sensing
properties. The relationship between defects and optical transitions in ZnO is
introduced. It also includes the characteristics of gas sensor and the gas sensing
mechanisms of ZnO. Finally, some of ZnO nanostructures synthesis methods are
presented.
Chapter 3 provides a full description of chemical and experimental details
utilized in this work: details of chemicals and instruments, sample preparation and post
growth processing, and details of each characterization technique.
Chapter 4, 5, and 6 present the experimental results and discussions. Chapter
4 focuses on the influence of deep level surface defects on the electrical and ammonia
gas sensing properties of ZnO nanorods. Chapter 5 studies the effect of native point
defects on the surface electronic structure of ZnO nanorods. Chapter 6 mainly
investigates the nature of red luminescence in O2 annealed ZnO nanorods.
Chapter 7 provides the general conclusions of the research.



Suranan Anantachaisilp Literature Review/ 6

CHAPTER II
OPTICAL, ELECTRICAL, AND GAS SENSING PROPERTIES OF
ZnO AND ITS SYNTHESIS: LITERATURE REVIEWS

In this chapter, an overview of ZnO properties including physical, chemical,

optical, electrical, and gas sensing properties are provided. In the first section, the details
of ZnO defects and their formation processes are given. The effect of defects on the
optical properties is also discussed. A history of ZnO gas sensors and their sensing
mechanism are introduced as well. In the last section, some of the ZnO nanostructures
synthesis methods are presented.

2.1 Physical and Chemical Properties of ZnO

ZnO is a binary II-VI direct wide band gap semiconductor with large exciton
binding energy. It mostly crystallizes in hexagonal wurtzite structure under ambient
conditions. General physical and chemical properties of ZnO are demonstrated in Table
2.1.

Table 2.1: General physical and chemical properties of ZnO

Physical and chemical properties Values
Molar mass 81.408 g/mol
Density 5.606 g/cm-3
Melting point 1975°C (decomposes)
Solubility 0.16 mg/100 ml (at 300K)
Crystal system Wurtzite (at 300K)
Space group P63mc
Lattice constant: a 3.2496 Å
c 5.2065 Å
University of Technology Sydney Literature Review / 7

Energy gap 3.37 eV (at 300K)

Exciton binding energy 60 meV (at 300K)
Intrinsic carrier concentration <106 cm-3 (max n-type doping
>1020 cm-3) electrons
Electron effective mass ݉௘‫ = כ‬0.24 ݉௘
Electron Hall mobility for low n-type 200 cm2/V·s (at 300K)
conductivity ݉௛‫ = כ‬0.59 ݉௘
Hole effective mass 5.50 cm2/V·s (at 300K)
Hole Hall mobility for low p-type conductivity

2.1.1 Crystal Structure

In particular, binary II-VI compounds have strong ionic bonding and wide
band gaps. By contrast, ZnO has stronger covalent bonding than ionic bonding leading
to strong crystal strength. ZnO has three main crystal structures which are rocksalt (B1),
zinc blende (B3), and wurtzite (B4) as displayed in Figure 2.1. B1, B3, and B4 are
strukturbericht symbols utilized to identify crystal structures. The crystal structure of
ZnO depends on its environment. Among all crystal structures, wurtzite structure is the
most thermodynamically stable phase of ZnO compared with other possible crystal
structures. This also confirms that covalent bonding is dominating because tetrahedral
coordination in wurtzite structure is representative of sp3 covalent bonding.
In ZnO wurtzite crystal structure, there are two atoms per unit cell. It
consists of two hexagonal close packed sub-lattices; HCP Zn2+ and HCP O2- stacking
along the c-axis. The lattice parameters are a = 3.2495 Å and c = 5.2069 Å.26 The
hexagonal wurtzite ZnO has an open structure where Zn atoms occupy half of the
tetrahedral sites. While the octahedral sites are empty; thus, there are many sites for ZnO
to accommodate intrinsic defects and extrinsic dopants.27
In wurtzite ZnO, there are four interested face terminations; the nonpolar m
(10ͳ0) face, the semipolar a (11ʹ0) face as well as the polar c (0001)-Zn and (000ͳ)-O
faces. Between the two polar surfaces, the chemical and physical properties are
different. The polar c surfaces are flat and stable since there is no reconstruction.28 By
comparison, the semipolar a face is less stable and has rougher surface than the polar c
Suranan Anantachaisilp Literature Review/ 8

and the nonpolar m faces. These different properties have significant effects on the ZnO
nanostructures growth processes.

Figure 2.1: Unit cells of ZnO crystal structures: (a) cubic rocksalt (B1), (b) cubic zinc
blende (B3), and (c) hexagonal wurtzite (B4). Zn atoms are represented in gray whereas
O atoms are represented in black.29

2.1.2 Electronic Band Structure

The electrical properties of semiconductors can be defined by the electronic
band structure since they relate to crystal translational order. Understanding the band
structure, consequently, is necessary to incorporate materials in device applications. A
number of theoretical methods with varying complexity have been utilized to calculate
the electronic band structure of ZnO.
ZnO is a direct and wide band-gap semiconductor with a high exciton
binding energy (‫ܧ‬௕ ൌ ͸Ͳܸ݉݁ሻ. The ZnO band gap is 3.37 eV at room temperature and
3.44 eV at low temperature. The calculation of its band structure is shown in Figure
2.2a.The band structure is located along high symmetry lines in hexagonal Brillouin
zone. Since ZnO is a direct band gap semiconductor, its highest valence band maximum
and lowest conduction band minimum sit at the ī-point of the Brillouin zone. The
conduction band is s-like two-fold degenerate containing the empty zinc 4s orbital
whereas the valence band is p-like consisting of the fully occupied oxygen 2p orbitals.30
University of Technology Sydney Literature Review / 9

The benefits of its wide band gap are high breakdown voltages, ability to sustain large
electric fields, low noise enervation, and high temperature and high power operation.29

Figure 2.2: (a) The calculation of ZnO band structure using the HSE hybrid functional.
The valence-band maximum energy is at zero. (b) Hexagonal ZnO band structure and
symmetries. The valence band splits into three sub-bands; A, B, and C resulting from
the effects of crystal-field splitting and spin-orbit coupling.31-32
Due to the effects of crystal-field splitting and spin-orbit coupling, the
valence band of ZnO is split into three sub-band states; A, B, and C (Figure 2.2b). The
A and C sub-bands have ߁଻ symmetry while the B posses ߁ଽ symmetry.32 The valence
band order of these three sub-bands is ߁଻ -߁ଽ -߁଻ which is consistent with߂௦௢ .33 The
energetic differences of these sub-bands are ߂஺஻ ൌ ͶǤͻܸ݉݁ and ߂஻஼ ൌ Ͷ͵Ǥ͹ܸ݉݁.32
The electronic band structure depends on temperature and pressure. The
band gap at ī point expands with decrease in temperature. The band gap temperature
dependence can be described by the following empirical relationship.34
D் మ
‫ܧ‬௚ ሺܶሻ ൌ  ‫ܧ‬௚ ሺܶ ൌ Ͳሻ െ (2.1)
்ାE

where ‫ܧ‬௚ ሺܶሻ is the energy gap at temperature ܶ, ‫ܧ‬௚ ሺܶ ൌ Ͳሻ is the energy gap at 0 K, D
and E are the temperature coefficients for which D = ͷǤͲͷ ൈ  ͳͲିସ eV K-1 and E = -900 K
for temperature up to 300 K.35

2.1.3 Defect Chemistry

Defects play an important role in the performance and reliability of
semiconductors; therefore, understanding the role of defects on the physical properties
Suranan Anantachaisilp Literature Review/ 10

of these materials is essential for their development and improvement.36 Like other
semiconductors, the applications of ZnO also depend upon the control of defects as well
as their associated charge carriers. Before studying ZnO defects and the effect of the
defects on its properties in depth, it is important to know the basic of crystal lattice
defects or imperfections. Defects in materials can be classified by their dimension.37
3D-Defects
The 3-dimensional defects; volume or bulk defects, are the aggregation of
atoms or vacancies. This type of defect can be categorized into 4 classes by their size
and effect on the materials properties. First are precipitates, small particles distributed
in the lattice through solid state reaction increasing the strength of materials. Dispersants
are a second type of volume defects. They are larger particles which can be large
precipitates, grains, or polygranular particles distributed in the microstructure.
Inclusions are another type of 3-dimensional defects which are extraneous particles or
large precipitate particles with undesirable components in microstructure. The last type
of bulk defects are voids or pores which are gases bubbles or clustering of vacancies in
solid state, decreasing the materials strength.
2D-Defects
Interfaces or planar defects are called 2-dimensional defects. This type of
defects are caused by the distortion in a crystal across the plane divided into 2 main
types; stacking fault and grain boundary.
A stacking fault is usually found in close packed structures - face centre
cubic (FCC) and hexagonal closed pack (HCP). It occurs due to the change in stacking
sequence over a few atomic planes. In the FCC structure, the normal arrangement
ABCABCABC can be changed to ABCAB_ABC causing by stacking fault. The pattern
ABABABAB of HCP arrangement is broken into ABABA_AB which also known as a
stacking fault.
A grain boundary is another type of planar defects. The interface of small
crystallites or grains are called as grain boundaries. The atoms in the boundary regions
are often in poor arrangement causing an irregular growth to the solid during
crystallization. They are generally found in polycrystalline materials and have a
significant effect on the mechanical properties of materials.
University of Technology Sydney Literature Review / 11

1D-Defects
The 1-dimensional defects; dislocations or line defects, are the deviation of
an entire row of lattice points row from a perfect pattern; they are usually undesirable.
and always act as electrical defects in semiconductors and optical materials. Line defects
consist of two major types; edge dislocation and screw dislocation.
An edge dislocation happens when an extra half-plane of atoms terminates
in the midway of a crystal causing the distortion in the adjacent planes of atoms. In order
to preserve the arrangement of crystal planes on both sides, the extra plane breaks and
joins bonds around the edge of the terminating plane. In this type of dislocation, the
Burgers vector; the magnitude and direction of distortion in the crystal lattice, is
perpendicular to the dislocation line.
A screw dislocation is more difficult to visualize. It is formed by a similar
method as the edge dislocation but in a helical path. Unlike an edge dislocation, the
Burgers vector is parallel to the dislocation line.
0D-Defects
Point defects are defined as 0-dimensional defects since they disturb the
crystal structure at only one point or around one lattice point. Point defects can be
divided into two groups which are intrinsic defects and extrinsic defects. Intrinsic
defects in a pure material consist of vacancies, interstitials as well as antisites. On the
other hand, extrinsic defects are initiated by another host atom or impurity atom called
substitutional or impurities. In semiconductors, point defects play an important role on
their various electronic and optical properties. Defects have been investigated for
several decades in bulk semiconductors; however, their role in nanostructured materials
is the subject of recent research. In particular, the impact of surface defects on the
physical properties of nanostructures requires further investigation owing to the small
length scales and large surface-to-volume ratio.27 Despite its simple formula, ZnO has
very rich defect chemistry38 with native vacancy and interstitial point defects and
inadvertently included impurities such as hydrogen and lithium being the major centres.
Moreover in nanostructured ZnO, as established in this thesis, surface point defects have
a major effect on its electronic and optical properties.
Suranan Anantachaisilp Literature Review/ 12

Figure 2.3: Illustration of point defects in ZnO crystal structures: (a) Zinc vacancy ܸ௓௡ ,
(b) Oxygen vacancy ܸை , (c) Zinc interstitial ܼ݊௜ , (d) Oxygen interstitial ܱ௜ , (e) Zinc
antisite ܼ݊ை , (f) Oxygen antisite ܱ௓௡ and (g) Substitutional impurity at a zinc site e.g.
‫݅ܮ‬௓௡ , ‫ݑܥ‬௓௡ or ‫݈ܣ‬௓௡ .
2.1.3.1 Vacancies
A vacancy is an intrinsic defect which is formed when some
atoms are missing from their normal lattice sites in crystal. In all crystalline materials,
vacancies mostly appear in high concentration. They also have an effect on the atoms
migration in the crystal lattice.30 They can diffuse through the lattice after filling by
neighbouring atoms. The surrounding crystal structure needs to be stabilized to avoid
vacancy collapse. The common vacancies in ZnO are zinc (ܸ௓௡ ) and oxygen (ܸை )
vacancies where zinc and oxygen atoms are absent from the ZnO lattice, respectively
(Figure 2.3a and b). Furthermore, they are the most prevalent defects in ZnO.
University of Technology Sydney Literature Review / 13

Figure 2.4: Calculated defect formation energy for ZnO native point defects in (a)
Zinc-rich conditions and (b) Oxygen-rich conditions.39
Zinc vacancies (ܸ௓௡ ) electronic structure in ZnO can be
modelled by utilizing molecular orbital theory. ܸ௓௡ involves four O dangling bonds and
six electrons. The absence of a Zn atom in ZnO lattice makes ܸ௓௡ an acceptor. It can
receive up to two additional electrons. ܸ௓௡ is much easier to form in n-type than p-type
ZnO due to the very high formation energies in p-type ZnO.34 In an oxygen-rich
condition, ܸ௓௡ is the dominant species among other point defects in n-type ZnO.
଴ ି ଶି
Between the three charge state configurations;ܸ௓௡ , ܸ௓௡ , and ܸ௓௡ , the most favourable
ଶି
ܸ௓௡ species is ܸ௓௡ because of its lowest formation energy as shown in Figure 2.4(b).
According to a first principle calculation, 0/1- and 1-/2- charge states transitions of ܸ௓௡
deep acceptors exist at 0.1-0.2 eV and 0.8-1.2 eV above the valence band maximum,
respectively.29, 33, 40-41 Consequently, a transition from the conduction band or a shallow
donor to the ܸ௓௡ deep acceptor would emit luminescence in the range of 2.2-2.6 eV,
which is green. Thus the green luminescence is typically attributed to ܸ௓௡ .33, 42-43
Suranan Anantachaisilp Literature Review/ 14

An Oxygen vacancy (ܸை ) is a deep donor defects. It is the most

dominant species in zinc-rich conditions owing to its lowest formation energies even if
comparing with interstitial zinc (ܼ݊௜ ) as presented in Figure 2.4(a). The electronic
structure of ܸை is also able to be explained by molecular orbital theory similar to ܸ௓௡ . It
contains four Zn dangling bonds and two electrons. According to this simple model,
there are three possible charge states which are doubly occupied (neutral; ܸை଴ ), singly
occupied (+1; ܸைା ), and empty (+2; ܸைଶା ). The stability of charge states depends upon
the Fermi level position in the bandgap. The formation energy of ܸைା is much higher
than that of ܸை଴ and ܸைଶା in all Fermi levels; hence, ܸை is recognized as a negative-U
defect and ܸைା is always thermodynamically unstable.44 According to a first principle
calculation, 0/2+ charge states transition of ܸை deep donors occurs at approximately 0.5-
0.8 eV above the valence band maximum.45-46 The 0/2+ transition level has been
estimated to be around 1-2 eV below the conduction band minimum by band gap
estimation error of local density approximation.47-49 The broad green luminescence band
centred in the range of 2.4-2.5 eV is usually assigned to 2+/0 state transition of ܸை .39, 50-
52

2.1.3.2 Interstitials
An interstitial is also an intrinsic defect where an atom occupies
an interstitial site; a lattice site on which an atom does not normally appear. Interstitials
are moderately unusual defects since they are high energy defects. In most crystalline
materials, the interstitial sites are quite small in size and rarely have a bonding
configuration.30 Zinc (ܼ݊௜ ) and oxygen (ܱ௜ ) interstitials are regular interstitials in ZnO
which are shown in Figure 2.3c and d.
A Zinc interstitial (ܼ݊௜ ) is a shallow donor defect which can
occupy in both tetrahedral and octahedral sites in the wurzite ZnO. However, it is more
stable at the octahedral site because of fewer geometrical constraints.34 In n-type ZnO,
ܼ݊௜ has a high formation energy even under zinc-rich conditions (Figure 2.4a). By
comparison, the formation energy of ܼ݊௜ଶା is much lower in p-type ZnO resulting in a
possible source of charge compensation in p-type ZnO.53 ܼ݊௜ also has a very low
migration barrier of 0.57 eV making it is unstable and highly mobile at room
temperature.35, 42 In addition, ܼ݊௜ generally form complexes with impurities due to its
high formation energy and mobility.
University of Technology Sydney Literature Review / 15

An Oxygen interstitial (ܱ௜ ) is a deep acceptor defect. It can

occupy the octahedral or tetrahedral interstitial sites or form split interstitials.34
Nevertheless, the tetrahedral site is unstable. ܱ௜ in this site also freely relaxes into a
split-interstitial configuration. The ܱ௜ atom stays in a lattice site with one of the nearest-
neighbour O atoms. ܱ௜ which occupy the octahedral site (ܱ௜ ሺ‫ݐܿ݋‬ሻሻ are electrical active
interstitials. ܱ௜ଶି ሺ‫ݐܿ݋‬ሻ which can accept two electrons has the lowest formation energy
compared with ܱ௜ ሺ‫ݐܿ݋‬ሻ and ܱ௜ଵି ሺ‫ݐܿ݋‬ሻ as presented in Figure 2.4b. The deep acceptor
transition levels for the 0/1- and 1-/2- charge states are at 0.72 and 1.59 eV above the
valence band maximum.35
2.1.3.3 Antisites
An Antisite is one type of intrinsic defect which is found in
compounds containing weakly ionized atoms. This energetically allows atoms to
exchange sites; therefore, an A-atom can sit on the lattice spot of B-atom or vice versa,
where A and B are atoms in host material. The antisites in ZnO are zinc (ܼ݊ை ) and
oxygen (ܱ௓௡ ) antisites, as displayed in Figure 2.3e and f. Both ܼ݊ை and ܱ௓௡ have high
formation energies (Figure 2.4a and b); thus, they have less influence than other native
defects.
A Zinc antisite (ܼ݊ை ) is a double donor defect in n-type ZnO.
The transitions of ܼ݊ை are located in the conduction band. This type of defect can split
into ܼ݊௜ and ܸை resulting from its migration. In n-type ZnO, the dissociation barrier of
ܼ݊ைଶା into ܼ݊௜ଶା and ܸை଴ is around 1.3 eV.35
An Oxygen antisite (ܱ௓௡ ) is a deep acceptor defect which is
rarely found under equilibrium conditions. Nonetheless, it can be formed under non-
equilibrium conditions e.g. under irradiation and ion implantation. The transition levels
0/1- and 1-/2- of ܱ௓௡ are at 1.52 and 1.77 eV above the valence band maximum,
respectively.35 The splitting of ܱ௓௡ into ܸ௓௡ and ܱ௜ requires higher migration barrier
than that of vacancies or interstitials.
2.1.3.4 Impurities
An impurity is an extrinsic point defect which forms by
introducing external atoms into the host material. It can be either a substitutional or an
interstitial defect depending on its occupational site. Substitutional impurities are
created when external atoms occupy a lattice site while an external atom occupies an
Suranan Anantachaisilp Literature Review/ 16

interstitial site for interstitial impurities. Due to the small size of interstitial sites, large
atoms are regularly substitutional whereas small atoms frequently fill in interstitial
sites.30 If the impurities have different charges from that of host lattice ion, the excess
charge is compensated by a paired vacancy or interstitial via a charge compensation
mechanism to maintain overall charge neutrality. In semiconductors, impurities play an
important role on most of engineering properties especially electrical properties by
introducing donor or acceptor levels in the band gap. The substitutional atoms with a
different charge from the host semiconductor serve as electron donors or acceptors. In
order to control ZnO conductivity, donor and acceptor impurities have been doped into
ZnO. The common shallow donor impurities which have been doped into ZnO are
aluminium (Al), gallium (Ga), indium (In), fluorine (F) and hydrogen (H). Lithium (Li),
sodium (Na), copper (Cu) and nitrogen (N) are some examples of acceptor impurities
usually doped in ZnO. Figure 2.3g shows the diagram of a substitutional impurity at a
zinc site in ZnO lattice e.g. ‫݅ܮ‬௓௡ , ‫ݑܥ‬௓௡ or ‫݈ܣ‬௓௡ .
The impurities introduced into ZnO create new energy levels in
the band gap which associated with excitons bound to impurities. The new energy levels
can be observed by the additional luminescence peaks. For instance, the fine structure
luminescence centred at around 2.4 eV is introduced by copper impurities; the yellow-
orange luminescence centred at 2.1 eV is related to excitons bound to lithium impurities
at a zinc site (‫݅ܮ‬௓௡ ሻ.54-55
2.1.3.5 Formation Energies and Processes of Native Defects
Native point defects have a strong effect on electrical and optical
properties of a semiconductor and, of course, on ZnO. In order to succeed in the use of
ZnO devices applications, it is important to understand how the defects are incorporated
into ZnO. The formation energies and their processes, transition between charge states
as well as migration, allow us to predict point defect concentrations and know their
incorporation during growth and processing. The formation energies and electronic
structure of native point defects in ZnO; ܸ௓௡ , ܸை , ܼ݊௜ , ܱ௜ , ܼ݊ை , and ܱ௓௡ , were calculated
by Janotti et al.35 and Kohan et al.33 using a first principle, plane-wave pseudopotential
technique with a supercell approach. Under thermodynamic equilibrium and no defect-
defect interaction, the concentration of point defects (ܿ) depends upon its formation
energy (‫ ܧ‬௙ ) which can be calculated by the following equation:
University of Technology Sydney Literature Review / 17

ா೑
ܿ ൌ ܰ௦௜௧௘௦ ‡š’ቀെ ቁ (2.2)
௞ಳ ்

where ܰ௦௜௧௘௦ is the number of sites in the crystal where the defect can take place, ݇஻ is
the Boltzmann constant, and ܶ is the temperature. According to equation 2.2, the defects
with high formation energies rarely form while the defects with the low formation
energies form in high equilibrium concentration. For a point defect in charge state (q),
the formation energy is computed as
‫ ܧ‬௙ ሺ‫ݍ‬ሻ ൌ  ‫ ܧ‬௧௢௧ ሺ‫ݍ‬ሻ െ ݊௓௡ Ɋ௓௡ െ ݊ை Ɋை െ ‫ܧݍ‬ி (2.3)
where ‫ ܧ‬௧௢௧ ሺ‫ݍ‬ሻ is a total energy of a system, ݊௓௡ and ݊ை are the number of zinc and
oxygen atoms, Ɋ௓௡ and Ɋை are zinc and oxygen chemical potentials, and ‫ܧ‬ி is the Fermi
energy.
The chemical potential is related to the growth conditions; Zn-
rich, O-rich or in between. Ɋ௓௡ is equal to Ɋ௓௡ሺ௕௨௟௞ሻ in a Zn-rich condition and Ɋை is also
equal to its bulk chemical potential (Ɋைሺ௕௨௟௞ሻ ) in an O-rich condition. In intermediate
ratios between zinc and oxygen, both Ɋ௓௡ and Ɋை are less than their bulk chemical
potential (Ɋ௓௡ሺ௕௨௟௞ሻ and Ɋைሺ௕௨௟௞ሻ ). Note that Zn and O chemical potentials are in
equilibrium with ZnO: Ɋ௓௡ +Ɋை < Ɋ௓௡ை .
The intrinsic point defects, in general, have low formation
energies; therefore, they usually form during growth processes. The formation energies
of ZnO native points defects; ܸ௓௡ , ܸை , ܼ݊௜ , ܱ௜ , ܼ݊ை , and ܱ௓௡ , are presented in Figure
2.5. These formation energies and charge transition levels were calculated from first
principles.
Most of defects are electrically active. They typically produce
levels in the semiconductors band gap including transitions between different charge
states in the same defect.56 The energy levels of defects in different charge states can
also be calculated from first principles derived from the formation energy calculation.
The defect transition levels ᖡሺ‫ݍ‬Ȁ‫ݍ‬ǯሻin the calculations are defined as the Fermi-level
positions for which the formation energies of q and q’charge states are equal. ᖡሺ‫ݍ‬Ȁ‫ݍ‬ǯሻ
can be calculated by equation 2.4.
ƍ
ா ೑ ሺ஽೜ Ǣாಷ ୀ଴ሻିா ೑ ሺ஽೜ Ǣாಷ ୀ଴ሻ
ᖡሺ‫ݍ‬Ȁ‫ݍ‬ǯሻ ൌ  (2.4)
௤ǯି௤
Suranan Anantachaisilp Literature Review/ 18

where ‫ ܧ‬௙ ሺ‫ܦ‬௤ Ǣ ‫ܧ‬ி ൌ Ͳሻ is the defect ‫ ܦ‬in the charge state ‫ ݍ‬formation energy when ‫ܧ‬ி ൌ
Ͳ.

Figure 2.5: Calculated ZnO native point defect formation energies as a function of
Fermi-level position in both Zn-rich (left) and O-rich (right) conditions. Valence band
maximum is located at the zero of Fermi level. Different slopes represent different
charge states.35
The thermodynamic transition levels for native point defects in ZnO are
demonstrated in Figure 2.6. These thermodynamic transition levels relate to thermal
ionization energies. A shallow level is the transition level to which the defect is
thermally ionized at room temperature; by contrast, a deep level is the transition level
that it is not ionized at room temperature. For shallow centres, they can happen in two
cases. First, they happen when the transition level in the band gap is located near the
band edges; valence-band maximum or conduction-band minimum. Another occurs
when the transition level is a resonance in the conduction or valence band. Notably,
some of transition levels ᖡሺ‫ݍ‬Ȁ‫ݍ‬ǯሻcannot be observed in optical spectroscopy
experiments.
Migration of native point defects in the crystal lattice is also important to
study since it provides a key to understanding how point defects are incorporated during
growth and processing as well as providing self-diffusion and impurity diffusion
University of Technology Sydney Literature Review / 19

models. These diffusions are often moderated by native defects. Equation 2.5 show the
calculation of the activation energy for self-diffusion (ܳሻ which is the summation of the
formation energy of the defect that mediates the self-diffusion (‫ ܧ‬௙ ) and its migration
energy barrier (‫ܧ‬௕ ).57
ܳ ൌ  ‫ ܧ‬௙ ൅  ‫ܧ‬௕ (2.5)

Figure 2.6: Thermodynamic transition levels for ZnO native point defects.35

2.2 Optical Properties and Processes in ZnO

Optical properties and processes in ZnO have been studied for several
decades. Like other semiconductors, ZnO optical properties have their origin in both
intrinsic and extrinsic effects. Intrinsic transitions occur between the electrons in the
conduction band and the holes in the valence bands as well as between free and bound
excitons while extrinsic transitions are introduced by dopants, impurities, point defects
or complexes, generating electronic states in the band gap.
Luminescence is light emission from a material after external excitation
through any process except blackbody radiation. The emission spectrum is related to
intrinsic processes (without any defect and impurity) as well as extrinsic processes (with
simple or complex defects and impurities). Consequently, the luminescence properties
provide valuable information for studying defects and defect levels in materials. Both
intrinsic and extrinsic processes can be investigated by numerous luminescence
techniques. Different techniques have different types of excitation energy sources such
as heat (thermoluminescence), voltage (electroluminescence), photons
(photoluminescence), and electrons (cathodoluminescence).
Suranan Anantachaisilp Literature Review/ 20

Photoluminescence (PL) and Cathodoluminescence (CL) are the two main

characterization tools for studying ZnO light emission processes. Photoluminescence
spectroscopy is a powerful tool to characterize ZnO optical processes and is highly
sensitive to point defects; therefore, it reveals characteristic radiative transition
channels. The cathodoluminescence of ZnO has received great attention as the signal
originates from a recombination of electrons and holes at defects and chemical
impurities produced in the fabrication process. In addition, the ZnO
cathodoluminescence is very sensitive to the surface state of the ZnO and impurities
under low voltage excitation (1 – 5 kV).58
The characteristic emission spectra of ZnO consist of two major peaks
which are near band edge (NBE) emission and deep level (DL) emission related to
defects as shown in Figure 2.7. The NBE emission peak located at approximately 3.3
eV is dominated by free and bound exciton emission (FX and D0X) along with LO
phonon replicas. The DL emission band usually appears in the range of 1.7 eV to 2.4
eV depending on defect types or impurities.

2.2.1 Excitons
The exciton is formed by the electrostatic Coulomb attraction between a free
excited electron and a hole. The attractive force between an electron and a hole is
defined as the exciton binding energy (‫ܧ‬௑ ). The exciton binding energy or ionization
energy can be calculated by equation 2.6 utilizing a hydrogen-like model.59
௠ೝ‫ כ‬௘ ర
‫ܧ‬௑ ൌ (2.6)
ଶ¾మ ఌమ ௡మ
where ݊ is the quantum number (݊ = 1 corresponding to exciton at ground state), ݉௥‫ כ‬is
the reduced mass of the electron-hole pair, ݁ is the elementary charge, ¾ is the reduced
Plank’s constant, and ߝ is the dielectric constant.
ZnO has a large exciton binding energy of 60 meV; hence it has a
significantly high percentage of excitons at room and higher temperatures. This also
leads to the high emission efficiency of ZnO based devices.
University of Technology Sydney Literature Review / 21

Figure 2.7: Characteristic ZnO luminescence spectrum presenting the near band edge
peak (NBE) and the broad deep level emission (DL).

2.2.2 Radiative Recombination

After electrons are excited from the valence band, an electron-hole pair is
generated (Figure 2.8a). An electron in the conduction band can then recombined with
a hole in the valence band. In radiative recombination, the excess energy is emitted in
the form of photon. The relative intensity depends on the concentration of a particular
defect and its capture cross-section.58 Due to self-absorption effects, ZnO radiative
emission from excitons reaches the material surface when the recombination happens
near the surface only. As shown in Figure 2.8, there are several possible radiative
recombination mechanisms.
2.2.2.1 Near Band Edge Emission
In ZnO, Near Band Edge (NBE) emission is a radiation emitted
in the UV region at approximately 3.3 eV. The most common recombination mechanism
is band to band recombination in which the excited electron in the conduction band falls
back to an empty free hole in the valence band (Figure 2.8b). Then it emits excess energy
in the form of photon with an energy equal to the band gap energy (Eg). However, the
Suranan Anantachaisilp Literature Review/ 22

NBE emission in ZnO is dominated by free and bound exciton emission (FX, D0X, A0X)
along with LO phonon replicas because of the very large and stable exciton binding
enrgy of 60 meV. This leads to a decrease in the recombination energy gap (E = Eg-EX)
as presented in Figure 2.8c. The bound exciton recombination exhibits a variety of
different sharp exciton lines, around 40 μeV in width, and usually observed at low
temperature. At higher temperature, these peaks are broader and merge at room
temperature. Their photon energies are characteristic for each defect with a narrow
energy range between 3.348 and 3.374 eV. The sharp lines are labeled as I lines. Since
there are 12 transitions, the lines are labeled from I0 to I11 previously used by Renolds
et al.60 There is a number of the assignment of the I lines as shown in Table 2.2 but their
assignment is still controversial.
In addition, the energy recombination gap can be more reduced
by free to bound (Figure 2.8d and e), donor-acceptor pair (Figure 2.8f), and some point
defects or transition metal impurity (Figure 2.8g).
2.2.2.2 Defect-Related Deep Level Emission
The point defects with deep levels in ZnO typically produce a
broad luminescence band in the visible range located between 2.9 eV and 1.5 eV in the
emission spectra. These deep level luminescence emissions are commonly observed as
green, yellow as well as red luminescence. There are a number of assignments of the
defect-related luminescence bands including native defects and impurities with point
defects such as oxygen and zinc vacancies (VO and VZn), zinc interstitial (Zni),
substitutional lithium on zinc site (LiZn) and etc. For example, the GL band in ZnO was
assigned as a transition from oxygen vacancy (VO) to the valence band and the YL band
at 1.96 eV is attributed to substitutional lithium on a zinc site (LiZn) in ZnO.61 Although
the defect-related bands have been assigned in articles, a conclusion on the emission
origin is still under debate.
University of Technology Sydney Literature Review / 23

Figure 2.8: Schematic diagram of the radiative recombination mechanisms;

a Excitation
b Band to band E = Eg
c Free and bound exciton E = Eg - EX [FX, (D0X), (A0X)]
d Free hole to donor bound electron E = Eg - ED [D0, h]
e Free electron to acceptor bound hole E = Eg - EA [e ,A0]
௤మ
f Donor-acceptor pair E = Eg - ED - EA - [D0, A0]
ସగఌ೚ ఌ௥

g internal excitation 1s levels point defects, 3d and 4f levels transition metals

where Eg is the bandgap of the material.
EX is the exciton bonding energy.
ED is the energy of the donor levels as measured from the bottom of the EC.
EA is the energy of the acceptor levels as measured from the top of the EV.
௤మ
is the Coulombic interaction between the electron and the hole.
ସగఌ೚ ఌ௥
Suranan Anantachaisilp Literature Review/ 24

Table 2.2: Bound exciton peak energies (eV) in ZnO

Line Energy (eV)
3.370762, 3.372463, 3.372553, 3.374162, 3.375462, 3.377262
I0
3.367953, 3.370262, 64, 3.371463, 3.371853, 3.374162, 3.375462
I1
3.366263, 3.367063, 3.367453, 3.368662, 3.369362, 3.370762
I2
3.364065-66, 3.365062, 3.365263, 3.365366, 3.366053, 3.366462, 3.366553,
I3
3.369366
I4
3.362266, 3.362463, 3.362853, 65, 3.363266, 3.363462-63, 3.364362
I5
3.361063, 3.361165, 3.361453, 3.361862, 3.362064
I6, I7
3.360053, 3.360263, 3.360453, 3.360562, 3.360665, 3.360853, 3.361064
I8
3.359266, 3.359353, 3.359463, 3.359766, 3.359853, 62
I9
3.356263, 3.356462, 3.356665, 3.356753, 3.358064
I10
3.353062, 3.353153
I11
3.348162, 3.348453

2.2.2.2.1 Green Luminescence

The green luminescence in ZnO is emitted in the range of 2.4 to
2.6 eV. In earlier studies, the green luminescence was attributed to copper impurities.54
However, later evidence suggests that the intrinsic defects oxygen and zinc vacancies
are more favourable for the broad GL band. From the defect formation energy
calculation of native point defects in ZnO, VO appears to be the most favourable in zinc
rich growth conditions whereas VZn is the most common in oxygen rich conditions.33
Copper impurities give a fine structure GL band centred at 2.4 eV while native point
defects (VO and VZn) show a structureless band at the same energy with an equal peak
width.
The fine structure GL band is mainly correlated with copper
impurities. According to the Dingle study,49 the broad GL band at approximately 2.45
eV with two sharp lines on zero-phonon line is related to a divalent copper ion Cu2+ in
a ZnO zinc lattice site (CuZn). The EPR results are closer to the EPR observation of CuZn
acceptor ground state in ZnO (g// = 0.7383 ± 0.0003, gŏ = 1.5237 ± 0.003) studied by
Dietz et al.67
University of Technology Sydney Literature Review / 25

The structureless GL band in ZnO was ascribed to native point

defects and complex defects such as VZn acceptor,33, 68 Zni,69 OZn,70 and VO45, 71-74. The
GL band was explained by various types of electron transitions. Vanheusden et al.72
explained GL by the transition of electron from VO donor level near conduction band to
the valence band (D௅h-type recombination). Guo et al.68 described it by a DAP
transition in which electrons move from VO or other donor levels to deep VZn acceptor
levels. Kohan et al.33 proposed that GL is the electron transition from the conduction
band to the VZn acceptor (e௅A type). In addition, GL was elucidated by a transition of
an electron between two states of VO (intracenter transition).74 Leiter et al.74 have shown
in their model that the two-electron ground state of the neutral VO is a diamagnetic
singlet state.
2.2.2.2.2 Yellow Luminescence
ZnO also emits yellow luminescence centred at 2.0 – 2.2 eV.
According to PL studies of undoped ZnO, the YL has been attributed to excess oxygen
in samples75 and oxygen interstitials (Oi) occurring during hydrothermal processes.76-78
The production of YL from the hydrothermally grown ZnO nanorods was also observed
and systematically studied by PL.79 It was also suggested that the PL yellow emission
was due to the presence of OH groups or Zn(OH)2 on the surface of ZnO nanorods rather
than oxygen interstitials. Some studies also reveal that YL is also related to lithium
impurities. It has been shown that the YL in hydrothermally grown ZnO could be related
to deep LiZn acceptors.55, 80 YL is believed to result from a donor-acceptor pair transition
from a shallow donor to the LiZn acceptors.81 Notably, adsorption of OH groups on the
ZnO surface is favourable under hydrothermal treatments, and the chemical origin of
YL produced by hydrothermally grown ZnO could be different from those obtained by
other preparation methods.
2.2.2.2.3 Red Luminescence
The red luminescence in ZnO has a peak centred at 1.7 -1.9 eV.
The red and orange-red emissions were previously assigned as originating from oxygen
interstitials (Oi),80, 82
zinc vacancies (VZn),83 and substitutional nitrogen defects on
oxygen sites.84 The RL in ZnO is also frequently related to excess zinc since it appears
after annealing ZnO in zinc rich atmosphere.85-86 Annealing ZnO in zinc vapour with
hydrogen in the environment leads to the RL generation as well. This process is able to
Suranan Anantachaisilp Literature Review/ 26

be reversed by removing hydrogen; therefore, the RL may be associated with the

interaction between oxygen vacancies and hydrogen.87 In the most recent work, Chen et
al. have studied the nature of RL in ZnO single crystals. They proposed that the RL is
related to the optical transition between a shallow donor and Zni-VO complex or VZnVO
di-vacancies defect centres.88

2.2.3 Temperature-Dependence Luminescence

In order to understand the physical processes of optical materials like ZnO
in depth, temperature dependence measurements are utilized. These allow us to
recognize the interaction of energy levels with phonons. The levels over the interaction
provide information of energy shifts,89 broadening,90 and energy levels dissociation.53
From low temperature to room temperature, the spectroscopic features overlap into one
due to the broadening and energy shift. At low temperature, the separation of peaks
reveals the dominant processes.
The temperature dependence measurement is a very interesting method for
the free excitons temperature dependence investigation; however, there is still limited
knowledge about the temperature dependence of ZnO spectral reflectivity properties.91
One difficulty is precise determination of the ZnO electronic structure at room
temperature since the features broaden and overlap into one broad peak at temperatures
above 150K. The temperature intervals have to be small enough between each scan in
order to see the effect of the increase in phonon interaction on energy levels. The
spectroscopic feature gradually changes in both energy and intensity when the
temperature changes. Additionally, the intensity of fine structure resulting from energy
levels with tiny localization or binding energies significantly decreases at the
temperature equals to its dissociation energy.

2.3 Electrical Properties of ZnO

ZnO is undoubtedly attractive for a variety of electronic and optoelectronic
applications at room temperature owing to its direct and large bandgap as well as high
exciton binding energy (60 meV at 300K). The large bandgap leads to numerous
advantages such as high temperature and power operation, lower noise generation,
University of Technology Sydney Literature Review / 27

higher breakdown voltages as well as large electric fields sustainability.92 However,

electrical properties of ZnO are difficult to measure since there is wide range of sample
qualities.
The electron effective mass and the hole effective mass are 0.24 m0 and 0.59
m0, respectively. For a low n-type conductivity, the corresponding electron Hall
mobility at 300K is μ = 200 cm2/V•s while μ = 5-50 cm2/V•s for a low p-type
conductivity. 93 Due to the relatively high electron mobility, ZnO overpowers other wide
bandgap materials for photo-electro-chemical applications.
The point defects also have a substantial effect on the conductivity of ZnO.
According to Van de Walle’s results, hydrogen is a dominant background donor in ZnO
since it is effortlessly ionized with quite low formation energy.94 Furthermore, other
results confirmed that hydrogen shallow donors in ZnO play a significant role in the
conductivity of ZnO.95-97 On the other hand, VO and Zni are deep donors with high
ionization energy leading to difficult free electrons production.33 Consequently, they do
not make a significant contribution to conductivity of ZnO. In addition, Zni has a limited
ability to contribute to n-type conductivity because of its high formation energy
although some theoretical studies suggested that it is a shallow donor.98-99 However, the
Zni complexes (Zni-NO) play a considerably role in ZnO conduction.100
In order to study the electrical properties of ZnO, two electrodes are needed.
Therefore, ohmic and schottky contacts to ZnO must be considered.

2.3.1 Ohmic Contacts to ZnO

The ohmic contact is a low resistance junction carrying electrical current
from metal to semiconductor as well as in the opposite direction. The current-voltage
relationship is a linear and symmetric in both positive and negative voltages with a fast
response for a small voltage and no parasitic resistance. The parasitic resistance is one
of the main problems for long-lifetime optical and electrical devices operations.
Moreover, the high-resistance metal-semiconductor ohmic contacts resulting from
thermal stress and poor contact reduce the device performance.
To achieve a high performance of ZnO-based optical and electrical devices,
ohmic contacts need to be thermally stable and reliable as well as low in resistance.
There are two methods to attain this high performance: first, decrease the metal
Suranan Anantachaisilp Literature Review/ 28

semiconductor barrier height by carrying out surface preparation leading to an increase

in carrier tunnelling probability; the other is enhancing the effective carrier
concentration near the surface.101-103

2.3.2 Schottky Contacts to ZnO

A Schottky contact is a metal-semiconductor contact having a large barrier
height. In addition, it has low doping concentration which is less than the density of
states in the conduction band or valence band. There are numerous difficulties in the
formation of Schottky contacts; for example, the chemical reactions between the metal
and the semiconductor, the surface states, the contaminants, the surface layer defects as
well as the metal diffusion into the semiconductor.94
For ZnO device applications, high quality Schottky contacts are required. It
is still a challenge to understand and control ZnO Schottky barriers according to the
wide and variable range of barrier heights from the same metal on a ZnO surface.104
Additionally, ZnO barrier heights are strongly affected by extrinsic factors i.e. the
quality of crystal and surface treatment.104 However, a high work function can be
applied to a ZnO crystal surface so that a useful ZnO Schottky barrier is created.

2.4 Gas Sensing Properties of ZnO

ZnO is one of the most interesting semiconducting metal oxides for the high
performance gas sensors utilized in the ppm-sensitivity range. Moreover, one-
dimensional (1D) ZnO nanostructures provide higher gas sensing performance
according to their high surface area to volume ratio.105-107 They can detect various kinds
of gases including both reductive gases and oxidative gases such as H2,108-111 NH3,112-
114
H2S,115-116 CO,117-118 C2H5OH,119 H2O,120-121 O2,122 O3,111, 123 and NO2.124-125 The
detection of gas molecules can be attained by measuring the change in resistivity or
conductivity with the reversible chemisorption process of reactive gases on the ZnO
surface.94, 111
However, high sensitivity and selectivity as well as low operating
temperature are still a big challenge in making commercial ZnO gas sensors.7, 126
University of Technology Sydney Literature Review / 29

2.4.1 Gas Sensor Characteristics

The measured principal parameter of the semiconducting oxide sensors is
ଵ
resistance (R) or conductance (I, ). All sensor operating characteristics are resulting
ோ

from this simple measurement. Resistance or conductance is a simple and easily

measured parameter but it is a second-order parameter. Hence, the elementary steps of
the sensing processes are still undeveloped.
The fundamental characteristics of gas sensor performance are the
sensitivity, selectivity, and the sensing response reversibility. These three characteristics
are measurable parameters.
2.4.1.1 Sensitivity
The sensitivity is one of the key characteristics of gas sensor
revealing the lowest analyte gas concentration that gas sensor can detect. To consider
the donor gases adsorption on n-type semiconducting oxides, sensitivity is the easiest
parameter to study. At equilibrium states in a gas sensor, the sensitivity of sensor is the
ratio of the change in signal amplitude to the original signal. Since the resistance or
current of sensors is recorded, the sensitivity can be calculated by equation 2.7 for n-
type semiconductor.
ோೌ ூ೒
ܵ ൌ ൌ (2.7)
ோ೒ ூೌ

where ܵ is a sensitivity of gas sensor at a particular concentration of gas, ܴ௔ and ‫ܫ‬௔ are
the resistance and current of the sensor when exposed to air (baseline resistance and
current), where ܴ௚ and ‫ܫ‬௚ are the resistance and current when exposed to the analyte gas.
‫ܫ‬௔ and ‫ܫ‬௚ are shown in Figure 2.9. A high value of ܵ is high in sensitivity.
2.4.1.2 Selectivity
Selectivity demonstrates how well sensor recognizes one gas
amongst others. In general, one semiconducting gas sensor can detect more than one
gas; therefore, it exhibits poor selectivity. Under the same operating conditions, analyte
gas is always desired to provide higher sensitivity than other interfering gases. With the
coexistence of interfering gases in the atmosphere, the selectivity of the sensor can be
described by equation 2.8.
ௌ௘௡௦௜௧௜௩௜௧௬௢௙௜௡௧௘௥௙௘௥௜௡௚௚௔௦
݈ܵ݁݁ܿ‫ ݕݐ݅ݒ݅ݐ‬ൌ  (2.8)
ௌ௘௡௦௜௧௜௩௜௧௬௢௙௔௡௔௟௬௧௘௚௔௦
Suranan Anantachaisilp Literature Review/ 30

Similar to other semiconducting gas sensors, pure ZnO also has

poor selectivity. ZnO exhibits the same responses to various types of reductive gases
and organic vapour. Nonetheless, the ZnO sensor with certain functional groups
affecting the adsorption processes can overcome this problem.127-128
2.4.1.3 Response Time and Recovery Time
Response time and recovery time are significant parameters
displaying how fast gas sensors are able to respond to analyte gases and how fast they
recover. Response time is the time required by sensor to reach 90% of the saturation
value of resistance or current when exposed to an analyte gas. Recovery time is the time
required by the sensor to achieve 90% of its original signal state after changing back to
exposure to clean air.
To be a good sensor, both response time and recovery time need
to be small. ZnO nanorods gas sensors exhibit a long response time as well as long
recovery time as shown in Figure 2.9.

Figure 2.9: Characteristic current-time sensing peak of ZnO nanorods annealed in O2

atmosphere response to ammonia gas.
University of Technology Sydney Literature Review / 31

2.4.2 Gas Sensing Mechanisms of ZnO

The gas sensing mechanism is a change of sensor’s properties when the
environment changes, allowing the monitoring of the varying properties of the sensor.
The interaction of the sensor surface with the analyte gas can be explained by two key
functions; receptor and transducer functions. The receptor function is an electronic
change of the oxide surface through a gas-solid interaction with a target gas while the
transducer function is an electrical resistance change of the sensor through a surface
phenomenon. According to receptor and transducer functions, the various electronic
states are created by the interaction of the sensor surface with the analyte gas. The
electronic properties of oxide semiconductors are modified in both the oxidative and
reductive gases but in the opposite direction for n-type semiconductor and p-type
semiconductors. The charge carriers are electrons and holes in an n-type semiconductor
and p-type semiconductor, respectively. Consequently, the electron concentration is low
in the for n-type semiconductor leading to an increase in the resistance whereas the
resistance decreases in air for p-type semiconductor.
In this study, we mainly focus on n-type ZnO. First, consider the adsorption
and desorption phenomena on the ZnO surface exposed to normal air atmosphere. Since
oxygen is one of the most active components in the air, surface reactions in ZnO always
involve adsorbed oxygen species (ܱ௦ି௡ where n and s = 1 or 2). The oxygen gas
molecules can bind to the ZnO surface through both physisorption and chemisorption
processes depending upon the temperature. The following reactions represent the
oxygen adsorption processes in different temperature,14, 111, 129-130
At low temperature:
ܱଶሺ௚௔௦ሻ  ՞ ܱଶሺ௔ௗ௦ሻ (2.9)
ି
ܱଶሺ௚௔௦௢௥௔ௗ௦ሻ  ՞  ܱଶሺ௔ௗ௦ሻ (at T < 150 °C) (2.10)
At high temperature:
ି
ܱଶሺ௚௔௦ሻ ൅ ʹ݁ ି  ՞ ʹܱሺ௔ௗ௦ሻ (at 150 °C < T < 250 °C) (2.11)
ି ି
ܱଶሺ௔ௗ௦ሻ ൅ ݁ ି  ՞ ʹܱሺ௔ௗ௦ሻ (at 150 °C < T < 450 °C) (2.12)
ଶି
ܱଶሺ௚௔௦ሻ ൅ Ͷ݁ ି  ՞ ʹܱሺ௔ௗ௦ሻ (at T > 250 °C) (2.13)
ି ଶି
ʹܱሺ௔ௗ௦ሻ ൅ ݁ ି  ՞ ܱሺ௔ௗ௦ሻ (at T > 450 °C) (2.14)
Suranan Anantachaisilp Literature Review/ 32

where (gas) are free gases, (ads) are species adsorbed on the ZnO surface and ݁ ି are
electrons contributed by ZnO.
When oxygen gas molecules adsorb onto ZnO surface, they trap electrons
from the ZnO conduction band leading to an increase in resistance (a decrease in
conductivity). Meanwhile, the depletion layer (space-charge layer) is formed as shown
in Figure 2.10. The depletion layer is a deficient carriers region because of the electron
trapping by adsorbed oxygen species. The general depletion thickness (space-charge
thickness or Debye length) is approximately 30 nm for undoped ZnO.131
After the depletion layer is formed, ZnO gas sensor is subsequently exposed
to an analyte gas which can be either oxidative or reductive. The response to oxidative
or reductive gases can be detected by measuring the resistance or conductivity of ZnO
gas sensing materials. Oxidative and reductive gases provide different contributions.

Figure 2.10: Schematic diagram of oxygen adsorption on the ZnO surface; n = 1 or 2

and s = 1or 2. (Adapted from Kim et al.)111
2.4.2.1 Oxidative Gas
The oxidative gases; for example, O2,122 O3,111, 123 and NO2,124-
125
can be detected by ZnO. After the adsorption of adsorbed oxygen species (ܱ௦ି௡ ) on
the ZnO surface, the oxidative gas molecules react with them and remove more electrons
from the conduction band, resulting in thicker depletion layer as shown in Figure 2.11.
The thicker depletion layer is due to the lower carrier concentration. This result also
leads to a reduced conductivity.
University of Technology Sydney Literature Review / 33

2.4.2.2 Reductive Gas

ZnO can also detect reductive gases such as H2,108-111 NH3,112-
114
H2S,115-116 CO,117-118 C2H5OH,119 and H2O120-121 but the mechanism is opposite to
that of the oxidative gases. Moreover, it provides relatively high sensitivity. These gases
react with surface adsorbed oxygen (ܱ௦ି௡ ) and subsequently remove it and release
electrons back to the ZnO conduction band, leading to a decrease of the depletion layer
thickness as shown in Figure 2.12. In addition, it leads to an increase of conductivity.

Figure 2.11: Schematic diagram of the oxidative gas sensing mechanism on ZnO
surface.107 (Green area is the depletion layer.)

Figure 2.12: Schematic diagram of the reductive gas sensing mechanism on ZnO
surface.107 (Green area is the depletion layer.)
Suranan Anantachaisilp Literature Review/ 34

2.4.2.2.1 Ammonia Gas

In this work, ammonia (NH3) gas was employed as a model
reductive gas. Possible surface reactions related to the NH3 gas response sensitivity of
sensors fabricated by the ZnO are given as follows.
ି
ܰ‫ܪ‬ଷሺ௚௔௦ሻ ൅  ͵ܱଶሺ௔ௗ௦ሻ  ՞ ͸‫ܪ‬ଶ ܱሺ௚௔௦ሻ ൅  ʹܰଶሺ௚௔௦ሻ ൅ ͵݁ ି (at T < 150 °C) (2.15)
Figure 2.13 shows how NH3 gas reacts with adsorbed oxygen species on ZnO surface.

Figure 2.13: Schematic diagram of ammonia gas sensing mechanism on the ZnO
surface; n = 1 or 2 and s = 1or 2.

2.5 Synthesis of ZnO Nanostructures

Several methods have been developed over the previous decade to prepare
nanomaterials with unique structural and physical properties; for example, template-
assisted, vapour-liquid-solid (VLS), colloidal micelle and electrochemical processes.1
The high-temperature vapor-phase approaches; physical vapor deposition 132-133 as well
134-135
as chemical vapor deposition, have been widely utilized for controlled aligned
ZnO nanoarray growth. However, these approaches are typically expensive and energy-
consuming. Fortunately, there are various low temperature solution-phase approaches;
such as aqueous chemical growth (ACG) and hydrothermal growth (HTG). These
methods have been developed to fabricate controllable ZnO nanostructures on substrates
University of Technology Sydney Literature Review / 35

in recent decades due to their potential for scale-up of production, commercial feasibility
and low energy-consumption.133
Growth of well-controlled ZnO nanostructures can be carried out by a
variety of methods, including both high-temperature and low-temperature approaches.

2.5.1 High Temperature Approaches

The high temperature synthesis of ZnO nanostructures is typically carried
out at temperatures in the range of 500°C to 1500°C. It is usually a gas phase synthesis
performed under gaseous environment in closed chambers. There are several common
gas phase methods; for instance, vapour phase transport, including vapour solid (VS)
and vapour liquid solid (VLS) growth,136-139 physical vapour deposition,140 chemical
vapour deposition,141 metal organic chemical vapour deposition (MOCVD),142 thermal
oxidation of pure Zn and condensation,143 and microwave assisted thermal
decomposition.144 These methods can control the formation of ZnO nanostructures in
designed patterns; however, they require complicated equipment and consume high
energy.

2.5.2 Low Temperature Approaches

The low temperature approaches for ZnO nanostructures growth are
generally carried out in the solution phase. The growth of ZnO in the solution phase
synthesis occurs in liquid phase. The processes are regularly carried out in an aqueous
solution which is called hydrothermal growth. Other examples of the ZnO
nanostructures solution phase synthesis are Zinc Acetate Hydrate derived nano-colloidal
sol-gel route,145 Zinc Acetate Hydrate in alcoholic solutions with sodium hydroxide or
tetra methyl ammonium hydroxide,146-148 template assisted growth,149 spray pyrolysis
for growth of thin films,150-151 and electrophoresis.152
2.5.2.1 Hydrothermal Growth
In this study, hydrothermal growth (HTG) is employed owing
to its controllable ZnO nanostructures on various substrates, potential for scale-up
production, commercial feasibility and low energy-consumption.
In basic wet chemical methods, the crystallization of ZnO
happens via the hydrolysis of Zn salts in a basic solution. The ZnO nanostructures
Suranan Anantachaisilp Literature Review/ 36

formation needs an alkaline solution since divalent metal ions do not hydrolyse in acidic
solutions. Both KOH and NaOH are the most common alkali compound utilized in the
formation of ZnO nanostructures. For hydrothermal synthesis of ZnO nanostructures,
Zinc nitrate [Zn(NO3)2] and hexamethylenetetramine (C6H12N4) are generally used.
Zn2+ ions are produced by Zn(NO3)2. HMT is believed to act as a weak base but the
exact function of HMT during the growth is still unclear. It should slowly hydrolyze in
the water solution and gradually produce OH-.
The following is the possible formation process of ZnO nanostructures by Zn(NO3)2 and
HMT:153
C6H12N4 + 6H2O ļ 4NH3 + 6HCHO (2.16)
H2O + NH3 ļ OH- + Nସା (2.17)
Zn2+ + 4NH3 ļ [Zn(NH3)4]2+ (2.18)
Zn2+ + 2OH- ļ Zn(OH)2 or Zn(NH3)42+ + 2OH- ļ Zn(OH)2(NH3)4 (2.19)
Zn(OH)2 / Zn(OH)2(NH3)4 ĺ ZnO + H2O (2.20)

All of the five reactions can be controlled by adjusting the

reaction parameters; precursor concentration, growth temperature and growth time.
The following table (Table 2.3) represents a summary of some
hydrothermal growth and the resulting morphology of the synthesis of ZnO
nanostructures.

Table 2.3: Summary of hydrothermal growth of ZnO nanostructures and their

morphology
Synthesis conditions Resultingmorphology
x Deposition of ZnO seed layer on
substrate.
x Growing ZnO nanosturcture in seed
solution (equimolar of
Zn(NO3)2.6H2O and HMT, C6H12N4)
at different temperature, time and pH

in sealable container. Representative SEM image of the ZnO nanorods:
(a) top view and (b) side view.154
University of Technology Sydney Literature Review / 37

SEM images of ZnO NSs on Si substrate with

different precursor concentrations of the growth
aqueous solution (a) at 25mM; (b)50 mM; (c) 100
mM; (d) 300 mM. Inset shows the magnified view
of the ZnO NSs (scale bar = 100 nm).155


SEM images of ZnO nanorods grown on Si
substrates with different ZnO seed layers: (a)
plane view of ZnO nanorod arrays grown on ZnO-
A seed layer; (b) plane view of ZnO nanorod
arrays grown on ZnO-B seed layer; and (c) plane
view of ZnO nanorod arrays grown on ZnO-C
seed layer.156
Suranan Anantachaisilp Experimental Details/ 38

CHAPTER III
CHEMICALS AND EXPERIMENTAL DETAILS

In this chapter, the synthesis of ZnO nanorods and post growth processes
are described. The experimental techniques utilized in this research to characterize
morphology, crystal structure, optical, electrical, and gas sensing properties of ZnO
nanorods will also be outlined.

3.1 Chemicals and Instruments

Chemicals utilized in this research are shown in Table 3.1. All chemicals
are analytical reagent grade and used as received without further purification. MilliQ
water and ethanol were used as the solvent in all the experiments.

Table 3.1: List of Chemicals

Chemicals Supplier
Zinc nitrate hexahydrate, 98%, Zn(NO3)2˜6H2O SIGMA-ALDRICH
Zinc acetate dihydrate, 98+%, (C2H3O2)2Zn˜2H2O SIGMA-ALDRICH
Hexamethylenetetramine, HMT, •99.0%, C6H12N4 SIGMA-ALDRICH
The synthesized ZnO samples were subsequently modified and
characterized by various techniques to study general, optical, electrical, and gas sensing
properties. Details of instruments used in this research are shown in Table 3.2.

3.2 Sample preparation

ZnO nanorods samples were prepared in two main steps. First, ZnO
nanostructures were synthesized via a hydrothermal method. Then the synthesized ZnO
samples were modified by post growth processing; annealing and plasma treatment.
University of Technology Sydney Experimental Details / 39

Table 3.2: List of Instruments

Instruments Model Location
Field Emission Scanning Electron Microscope Zeiss Supra 55VP SEM MAU, UTS
(FESEM)
X-Ray Diffraction (XRD) Siemens D5000 MAU, UTS
Environmental Scanning Electron Microscope FEI Quanta 200 ESEM with MAU, UTS
(ESEM) with Cathodoluminescence System Gatan MONOCL3
Photoluminescence Spectroscopy (PL) HeCd laser with Spex-1404 TU Berlin
double monochromator
ELECTROMETER/HIGH RESISTANCE METER Model 6157A, Keithley Physics, MU
Gas Measurement System Homebuilt Physics, MU
Photoemission Yield Spectroscopy (PYS) Homebuilt MAU, UTS
Note: MAU is Microstructural Analysis Unit, UTS is University of Technology Sydney, TU Berlin is
Technical University of Berlin and MU is Mahidol University.

3.2.1 ZnO nanostructures growth

In this work, silicon and sapphire wafers were chosen as the substrates. The
pieces of substrates were cleaned by a standard cleaning process:
x 15 minutes ultrasonic bath in ethanol
x 15 minutes ultrasonic bath in isopropanol
x 15 minutes ultrasonic bath in MilliQ water
x Dried in air at room temperature
Subsequently, the cleaned substrates were deposited with a ZnO seed layer
and followed by ZnO growth.
3.2.1.1 Seed layer deposition
The cleaned substrates were coated with a seed layer, as in the following steps:
x Dropped Zinc acetate solution (5mM in ethanol) on the substrate until it
fully covered the surface
x Rinsed it with ethanol after 10 seconds
x Dried in air at room temperature
These steps were repeated 5 times to obtain complete coverage.
x Put the seed layer coated substrate into the oven at 250oC for 20
minutes
All seed layer coating steps were repeated twice.
Suranan Anantachaisilp Experimental Details/ 40

3.2.1.2 Hydrothermal growth

After the deposition of a seed layer, ZnO nanostructures were
grown by a low temperature approach; hydrothermal growth (HTG). The pre-coated
substrate was suspended in a seed solution containing 0.025M zinc nitrate hexahydrate
and 0.025M hexamethylenetetramine at 90oC for 3 hours.157-158
The following shows the HTG steps:
x Put the pre-coated substrate in the seed solution in the Teflon
lined autoclave
Figure 3.1 shows the photograph of Teflon lined autoclave.
x Seal and put the autoclave into the oven at 90oC for 3 hours
After the growth
x Cool Teflon lined autoclave naturally
x Take out and wash the resulting samples with MilliQ water
to eliminate residual salts or amino complex D
x Dry in air at room temperature

Figure 3.1: Photograph of Teflon Lined Autoclave used as a reactor in hydrothermal

growth.

3.2.2 Post growth processing

In order to modify the surface defect structure of ZnO, the resulting samples
have been annealed in both oxidative and reductive environments after the growth.
Some of samples were then treated in hydrogen plasma.
3.2.2.1 Annealing
The resulting samples have been annealed at annealing
temperatures of 650°C in different gases; Inert gas (Ar), Oxidative gas (O2) and
Reductive gas (5%H2 in Ar and zinc vapor) at approximately 2 Torr pressure and
University of Technology Sydney Experimental Details / 41

100sccm flow rate for 30 minutes. All samples were annealed in the Lindberg Blue
Mini-mite Tube Furnace (Figure 3.2). The maximum annealing temperature of the
furnace is 1100oC. Annealing process can be done under vacuum or controlled gaseous
environments.

Figure 3.2: Photograph of Lindberg Blue Mini-mite Tube Furnace.

3.2.2.2 Hydrogen Plasma Treatment

Hydrogen plasma treatment was accomplished utilizing mild
hydrogen radio-frequency plasma for 20 minutes. Typically, hydrogen can be
incorporated in ZnO in the temperature range of 25oC to 450oC. The hydrogen
incorporation is as fast as 10 seconds. The energetic hydrogen radicals were ionised
from hydrogen gas. Subsequently, they penetrated into the sample and diffused in. In
this work, hydrogen radicals were generated by Radio Frequency power of 10W with
10 sccm H2 flow rate. An AG 0202-HV OS high voltage plasma generator from T&C
Power Conversion, Inc. was used to produce plasma. An Apex mass flow controller was
used to flow hydrogen gas into the evacuated chamber and pressure was examined by
Lam research series capacitance diaphragm gauge. ZnO samples were also heated at
100°C during plasma cycle to activate hydrogen diffusion. The samples were then
naturally cooled in a hydrogen atmosphere chamber after plasma treatments. The overall
setup of plasma chamber is demonstrated in Figure 3.3.
Suranan Anantachaisilp Experimental Details/ 42

Figure 3.3: Photograph of plasma chamber showing the overall setup.

3.3 Scanning Electron Microscopy (SEM)

The morphology and the average diameter of the ZnO nanorods were
investigated at 20 kV using a Zeiss Supra 55VP SEM. The Zeiss Supra 55VP SEM is a
high resolution "Field Emission Scanning Electron Microscopes" (FESEM). The spatial
resolution of the microscope is 1.0 nm, 1.7 nm, 3.5 nm, 4.0 nm, and 2.0 nm at 15 kV, 1
kV, 0.2kV, 1kV, and 30 kV (VP mode), respectively. Its high efficiency in-lens detector
produces exceptional quality scanning electron images down to 0.1 kV. The samples
were attached onto the specimen holders using carbon conductive tape.
During SEM examination, high energetic electrons from the incident
electron beam interact with the sample and generate two types of electrons; secondary
electron and backscattered electrons, as well as X-rays and light (Figure 3.4). Each
emission is used for difference purposes. Low energy secondary electrons are ejected
University of Technology Sydney Experimental Details / 43

during inelastic scattering of the primary electron beam providing topology information
of the sample surface while high energy backscattered electrons bounce out of the
sample after beam collides with sample nucleus giving the composition information. X-
rays and light assign element concentration and defect information, respectively.

Figure 3.4: Penetration range of the primary electron beam with the origin of
electrons, x-rays and light.159

3.4 X-Ray Diffraction (XRD)

Crystal structure and lattice parameters of the ZnO nanorods before and after
annealing were characterized by a Siemens D5000 X-ray Diffractometer with Cu-KD
radiation and a wavelength of 1.5418Å. X-ray diffraction was performed in T -2T
scanning mode in the range of 30°-65°. The crystalline phase and lattice constants were
calculated by ICSD using POWD-12++.
Suranan Anantachaisilp Experimental Details/ 44

In typical XRD, an incident beam of X-rays interacts with sample generating

constructive interference from lattice planes when conditions satisfy Bragg’s Law only
(Equation 3.1). Bragg’s Law reveals the relationship between electromagnetic radiation
wavelength (Ȝ), diffraction angle (T) and lattice spacing (d) in samples.

nȜ = 2d · sinT (3.1)

3.5Cathodoluminescence Spectroscopy (CL)

Cathodoluminescence is an emission of light from a material during
irradiation by an energetic beam of electrons. An emission of cathodoluminescence
light is generated via electron-hole pair formation mechanism. A characteristic
luminescence is emitted from the crystal under electron beam irradiation where the
relative intensity depends on the concentration of a particular defect and its capture
cross-section.58 In addition, the ZnO cathodoluminescence is very sensitive to the
surface state of the ZnO crystal and impurities under low voltage excitation (1 – 5 kV).58
Accordingly, in order to understand the characteristics of the ZnO surface defects, the
cathodoluminescence of ZnO under low voltage excitation (5kV) was studied in this
project.
All CL spectra were carried out inside an FEI Quanta 200 ESEM. The
spatial resolution of the microscope is 3.0 nm and 10 nm at 30 keV and 3 keV,
respectively. The microscope chamber was operated at pressure lower than 10-5 Torr. A
Faraday cup was used to measure beam current. The layout of the cathodoluminescence
setup is presented in Figure 3.5. The samples were placed on a cold stage, which allowed
fast cooling from room temperature (300K) down to liquid nitrogen temperature (80K)
and liquid helium temperature (5K). The samples were located at the focus point of a
parabolic mirror, photons generated from the sample was collected by the mirror and
subsequently directed through optical fibers to a detector. In this work, two detectors;
an Ocean Optics QE65000 detector and a Hamamatsu S7011-1007 CCD sensor, were
utilized on the CL system to collect full range CL spectra and high resolution CL of the
near band edge region of ZnO, respectively. The QE65000 consists of a 1024 ȋ 54
University of Technology Sydney Experimental Details / 45

active pixel area with acquisition range of 200 to 1100 nm with 5 nm in a spectral
resolution. The active pixel area of Hamamatsu S7011-1007 CCD sensor is 1044 ȋ 124.
The sensor was placed inside a head of a Hamamatsu C7021 detector. The light was
dispersed by an Oriel MS 257 1»4m monochromator with a 1200 lines/mm grating. The
spatial resolution of this set up is 0.1 nm over a 53 nm range.

Figure 3.5: Schematic of CL experimental setup.

Both intensity and wavelength of all CL spectra were calibrated by an Oriel
63355 200W QTH calibrated lamp and mercury lamp, respectively. Equation 3.2 was
used to calculate the intensity correction curve.
ே௢௥௠௔௟௜௭௘ௗ௥௘௙௘௥௘௡௖௘௜௡௧௘௡௦௜௧௬
‘””‡…–‹‘…—”˜‡ ൌ  (3.2)
ே௢௥௠௔௟௜௭௘ௗ௠௘௔௦௨௥௘ௗ௜௡௧௘௡௦௜௧௬

The wavelength calibration curve of high resolution CL spectra was done

by measuring atomic spectral lines of mercury after the ZnO spectra acquisition. The
calibrated wavelength was subsequently converted into electron volts (eV) by the
following equation:
ଵଶଷଽǤ଼ସଵ଼ହ଺
Š‘–‘‡‡”‰›ሺ‡ሻ ൌ  (3.3)
ఒሺ௡௠ሻ

To study temperature dependent CL, a Gatan C1002 liquid nitrogen cold

stage was used for cooling samples down to 80K. A Gatan CF302 continuous flow liquid
Suranan Anantachaisilp Experimental Details/ 46

helium cold stage was also used for cooling to 5K for collecting high resolution CL
spectra. Finally, Peak-O-Mat program was utilized to do curve fitting and data analysis.

3.6 Photoluminescence Spectroscopy (PL)

Photoluminescence spectroscopy (PL) is an important tool for investigating
a semiconductor’s optical processes. It is sensitive to point defects and the signal
generally originates near material surfaces. Since the characteristic radiative transition
channels are regularly induced by point defects, PL is often used to determine electronic
energy levels, binding energies as well as the chemical origin of point defects. It
provides a large number of defect related luminescence information i.e. excitons, donor-
acceptor pairs (DAP), phonon replica, thermal activation and quenching behaviour of
transition lines etc. The non-radiative channels for carrier recombinations related to
deep levels formation in the band gap can also be recognized from the PL intensity. In
addition, PL is attractive for using with materials with poor conductivity since the
electrical excitation or detection can be ignored.
In order to study the behaviour of surface defects of ZnO nanorods samples,
PL measurements were done in a liquid helium bath cryostat allowing measurements
from 7K to 300 K. A HeCd Laser with the 325 nm emission line and an output power
of 36 mW was used to excite the ZnO samples above the band gap. The emitted light
was dispersed by a Spex-1404 double monochromator (spectral resolution 50 μeV) and
detected by a bi-alkali photodetector.

3.7 Synchrotron Light Experiments

Accelerated electrons moving with 99.9997% speed of light under magnetic
field action emit the electromagnetic radiation known as Synchrotron light. Synchrotron
light is a particularly high intensity light with a wide range of energies in the
electromagnetic spectrum from infrared light to hard x-rays. It has several unique
properties including high brightness, wide energy spectrum, tunable, highly polarised,
and emitted in very short pulses. Due to a large number of useful properties of
synchrotron light, it was used in an extensive range of advanced research in various
University of Technology Sydney Experimental Details / 47

fields such as biosciences, medical research, environmental sciences, agriculture,

minerals exploration, engineering, forensics as well as advanced materials even at the
nanostructure scale. All synchrotron light experiments; X-ray photoemission
spectroscopy (XPS) and x-ray absorption near-edge structure (XANES) were conducted
on the Soft X-ray Spectroscopy beamline, Australian Synchrotron.160

3.7.1 X-Ray Photoelectron Spectroscopy (XPS)

X-Ray Photoelectron Spectroscopy (XPS) measures the energy of the
detected electrons emitted from the sample after irradiation by monochromatic soft x-
ray photons. XPS is used to determine the chemical identity and element concentrations
in the sample by the following equation.
‫ܧ‬௕ ൌ ݄߱ െ ‫ܧ‬௞ െ ‫׎‬ௗ (3.4)
where ‫ܧ‬௕ is the known binding energies of the elements, ‫ܧ‬௞ is the kinetic energies of
the emitted electrons, ݄߱ is the energy of the X-ray photon, and ‫׎‬ௗ is the work function
of the detector.
The energy of the X-ray photons is finely tuned when performing XPS
measurements at synchrotron facilities. At the Australian Synchrotron facility, the
photon energy of the X-ray source covers wide range of energies from 90 eV to 1500
eV. By comparison, the XPS setups in normal laboratory utilize aluminium or
magnesium targets as X-ray source with photon energies of 1486.6 eV (Al KD) or 1253.6
eV (Mg KD). Therefore, performing XPS measurements at synchrotron facilities allows
for the XPS signal optimization at a specific photoemission by selecting the appropriate
photon energy.
XPS is a highly surface sensitive analysis technique. Although the
penetration power of X-ray photons in solids is in the range of 1 μm to 10 μm, the
electron interaction with matter is higher than that of photons. Thus, the electrons mean
free path is only tens of angstroms. The XPS signal is generated from only the
photoelectrons created directly below the surface.

3.7.2 X-Ray Absorption Near Edge Spectroscopy (XANES)

X-Ray Absorption Near Edge Spectroscopy (XANES) is accomplished by
measuring the X-rays absorption as a function of X-ray photon energy. Utilizing
Suranan Anantachaisilp Experimental Details/ 48

XANES, the oxidation state as well as element coordination number in a material, and
the elements site symmetry can be determined. In addition, the conduction band of the
materials information is also provided by XANES since it is probing empty states. The
X-ray photon energy is scanned over the core levels energy of primary elements in the
samples. The electrons are excited from their levels and the holes are then generated.
Subsequently, the electrons in a higher energy state recombine to the vacancy and
release the energy in the form of photons or excite the electrons from a higher state to
form Auger electrons.
XANES spectra cover over hundreds of electron volts above the core
absorption edge. The peaks in an oscillatory fine structure shape in XANES spectra
relate to the transitions of internal electrons between core levels. The oscillations in the
XANES spectrum reflect a standing wave formation causing scattering between the
photoelectron and neighbouring atoms.
XANES implemented in the Total Fluorescence Yield (TFY) and Total
Electron Yield (TEY) mode was performed around the N K-edge. The incident X-ray
beam was linearly polarized and perpendicular to the substrate surface. The photon
energy scale was calibrated against the Au 4f7/2 peak at 84 eV from a clean gold film in
electrical contact with samples.

3.8Electrode Deposition and Gas Sensor Device Assembly

In order to assemble a gas sensor device and study the electrical properties
of ZnO nanorods samples, the sensing ZnO nanorods arrays need two electrodes. A large
active area is needed to achieve the highest interaction with gases; therefore, several
pairs of electrodes are used in the sensor. The sensing layer is located between two sides
of electrodes. The most common electrode pattern providing a large active area is an
interdigitated pattern.
In this research, the electrodes are made from aluminium by a thermal
evaporation technique and shaped into twelve fingers interdigitated pattern, as shown in
Figure 3.6a. The dimension of each finger is 100 μm spacing and 900 μm length. The
white area represents the active area (ZnO nanorods arrays) and aluminium electrodes
are displayed in gray. To establish a device, the sensing layer with aluminium electrodes
University of Technology Sydney Experimental Details / 49

is attached to a polycarbonate board and wired at the connectors as shown in Figure

3.6b. Copper tape and silver paste were used at the connectors (black dot in Figure 3.6b).

3.9Current Voltage Characteristic Measurement

Current-Voltage analyses of ZnO nanorods samples with interdigitated
aluminium electrodes were investigated at room temperature in the measurement
chamber under an ambient air atmosphere utilizing source measurement unit; model
6157A electrometer/high resistance meter, Keithley. The measurement of current-
voltage is carried out by applying a DC voltage (V) between two electrodes and
monitoring the DC current (I).
An ideal current-voltage characteristic shows a linear curve obeying Ohm’s
law (Equation 3.5). General resistors reveal linear current voltage characteristic. The
resistance (R) is constant in an ideal resistor. In a sample with good Ohmic behaviour,
the junction resistance between metal (Al electrodes) and semiconductor (ZnO
nanorods) can be neglected.
V=I·R (3.5)
The charge carrier mobility of samples can be inferred by measuring
current-voltage characteristic since the electrical conductivity directly changes with the
charge carrier mobility as shown in Equation 3.6.
ı = qnμ (3.6)
where ı is the conductivity, q is carrier charge, n is carrier concentration and μ is drift
mobilityof the carriers.

3.10 Gas Sensing Measurement

The sensing performance of ZnO nanorods gas sensors were evaluated using
ammonia gas in air at room temperature. The gas sensor devices were investigated using
a constant voltage of 1V, and the current was subsequently recorded as a function of
time over the designed experiment cycles utilizing 6157A electrometer/high resistance
meter, Keithley.
Suranan Anantachaisilp Experimental Details/ 50

3.10.1 Gas Measurement System

The two main components in the gas measurement system are a gas flow
controller unit and gas response recorder. Figure 3.7 shows the diagram of the gas flow
controller unit consisting of a start switch (SW1), a stop switch (SW2), a programmable
logic controller (PLC), a solenoid valve (SV), and two mass flow controllers (MFC1,
and MFC2). The start and stop signals from SW1 and SW2 as well as the flow rates of
MFC1 and MFC2 were obtained by PLC. Additionally, SV and two output flow rates
were controlled by the PLC outputs. The second main component; the gas response
recorder consisted of a Keithley Co, Ltd. electrometer model 6157A wired into the
sensing chamber which was used to supply voltage and collect current.
The electrometer with 82350B PCI GPIB Interface Card from Agilent
Technologies Co, Ltd. is connected to a computer. LabVIEW software from National
Instrument Co, Ltd. was programmed in the computer to record the signal during the
measurement cycles.

Figure 3.6: (a) Twelve fingers of the interdigitated electrode (shown in gray). (b) Gas
sensor device comprised of ZnO sensing layer (light gray square), aluminum electrodes
(gray), polycarbonate board (outermost square) with conduction line (black line), and
connection wires (black dot).
University of Technology Sydney Experimental Details / 51

SW1 SW2

Input
PLC
Output

MFC1 MFC2
SV
Figure 3.7: Gas flow controller unit diagram; input channels are shown in dash lines
and output channels are shown in solid lines.

The gas measurement system consisted of 5 parts; gas sources, flow

regulators, sensor chamber, signal acquisition, and waste disposal. The homebuilt gas
measurement system used in this project is shown in Figure 3.8. The ammonia vapour
was generated by flowing carrier gas (air) into the liquid ammonia analyte reservoir.
The flow rates of ammonia gas and carrier gas were regulated by MFC1 and MFC2,
respectively. Both gases were introduced into the sensor chamber and subsequently
mixed and then exhausted into the waste reservoir. The ammonia gas concentration in
the sensor chamber was adjusted by controlling the flow rates of MFC1 and MFC2 in
the fixed total flow rate of 100 mL/min.

3.10.2 Designed Experimental Procedure

The gas sensor devices were put into the measurement chamber and the
electrical properties measured. A constant voltage of 1V is applied to the gas sensor
devices and then the current is recorded over the experiment cycles as a function of time
using a Keithley model 6157A electrometer/high resistance meter.
There are four steps to run experiments and control the external parameters:
(i) Leave the sensors under ambient air in the chamber for 1 hour.
(ii) Purge air for 30 minutes.
Suranan Anantachaisilp Experimental Details/ 52

(iii) Apply voltage and obtain baseline for 20 minutes.

(iv) Flow analyzed (NH3) and clean with air for 10 and 10 minute cycles.
The pre-measurement steps; step (i) and (ii), were done to control the
external parameters. To start the measurement cycles, the 20 ppm ammonia gas was
purged into the sensor chamber and switched with air during the cleaning period. Then
the ammonia concentrations systematically increase by 20 ppm in each cycle to highest
concentration 100 ppm. After the gas responses were received, the sensitivities were
then calculated by Equation 3.7.
ூ೒
S= (3.7)
ூೌ

where S is a gas sensitivity, ‫ܫ‬௚ is a current response at each ammonia concentration, and
‫ܫ‬௔ is an average current response at the first measurement period (base line period)

Vacuum pump
Pressure gate
Exhaust
MFC
1

Solenoid
Sensor
Analyte Valve, SV chamber
MFC
Carrier reservoir Waste
2
gas reservoir

Water

Figure 3.8: Homebuilt gas measurement system; dotted line represents the shortcut path
when the dry measurement starts.
University of Technology Sydney Gas Sensing Properties/ 53

CHAPTER IV
TAILORING DEEP LEVEL SURFACE DEFECTS IN ZnO
NANORODS FOR HIGH SENSITIVITY AMMONIA GAS SENSING

4.1 Summary
The influence of deep level surface defects on electrical and gas sensing
properties of ZnO nanorod NH3 (g) sensors was studied. ZnO nanorods 50í60 nm in
diameter were synthesized via low-temperature hydrothermal growth at 90 °C on
sapphire substrates. The as-grown nanorods exhibited a cathodoluminescence (CL) peak
centered at 1.90 eV (YL), attributed to LiZn deep acceptors or O interstitials. Subsequent
annealing in O2 at 1 atm and Znvap at 650 °C produced broad CL peaks centered at 1.70
eV (RL) and 2.44 eV (GL), respectively. The RL and GL have been ascribed to acceptor-
like VZn and donor-like VO related centers, respectively. Electrical and gas sensing
measurements established that the NH3 gas response sensitivity was 22.6 for O2 anneal
(RL), 1.4 for Znvap anneal (GL), and 4.1 for the as-grown (YL) samples. Additionally,
treatment in H-plasma quenched the RL and inverted the NH3 electrical response due to
the incorporation of H donors. Changes in the gas sensing response are explained by a
shift in the position of the ZnO Fermi level relative to the chemical potential of NH3 gas
due to the creation of near surface donor or acceptors. These data confirm that ZnO
nanorods arrays can be tailored to detect specific gas species.

4.2 Introduction
Zinc oxide (ZnO) has attracted great attention over the past few decades to
enable applications in optical devices, piezoelectric devices, solar cells, and gas
sensors.2, 4, 161 In general, the gas sensing can be measured by monitoring changes in the
surface electrical conductance of ZnO upon exposure, owing to the electron transfer
between the ZnO surface and target gas molecules.162 Because of their high surface area,
ZnO nanostructures such as nanoparticles, nanowires,16, 162-163
nanorods,164-165 and
Suranan Anantachaisilp Gas Sensing Properties/ 54

nanotubes166 are fabricated and used as gas sensors having a significantly higher
sensitivity in comparison to those using ZnO microcrystal thin films.165 A number of
researchers have confirmed that then sensitivity of these high surface area ZnO gas
sensors is strongly dependent on their operating temperature;115, 165, 167-168 i.e., the higher
the temperature, the higher the sensitivity. Accordingly, nanostructured ZnO materials
have been extensively reported as highly sensitive gas sensors when operating between
150 and 400 °C15, 163, 169-1705, or at above 400 °C.115, 171 Recently, room temperature gas
sensing devices composed of ZnO nanorods were developed to detect: H2 gas at a lowest
concentration of 150 ppm,165 alcohol gas at concentrations down to 0.5 ppm,172 and NH3
gas at concentrations of 300173 and 500 ppm.174 Here the gas sensing mechanism that
governs the ZnO surface conductivity has been attributed to low-temperature surface
reactions, involving adsorbed molecular or atomic oxygen as well as hydroxyl groups.
Several works have indicated that the electrical response can be significantly improved
in ZnO nanorods by controlling their nanostructural properties specifically their
morphology, particle size, and surface area17, 165 as well as the carrier concentration.15-
16, 18
Moreover, recent experimental and theoretical studies have shown that the surface
defect chemistry of ZnO nanorods has a significant effect on the sensitivity to different
gas species. Additionally, it has been suggested that the sensing capabilities of ZnO can
be improved by incorporating dopants such as Al, Sn, and Fe into the ZnO structure to
generate point defects that serve as catalysts enhancing the electron transfer between
ZnO and molecular gas169 or to stabilize oxygen vacancies.169-170 The exact role of these
surface defects on the surface conductivity remains controversial; however, a number
of models have been suggested, such as providing reactive sites for oxygen adsorption,
inducing surface band bending, and varying the carrier concentration by acting as donors
or acceptors.15-16, 163, 169 This research focuses on the systematic introduction of different
types of surface point defects in ZnO nanorods through controlled growth and thermal
processing in oxidizing (O2) and reducing (Zn vapor) atmosphere. The ZnO nanorods
arrays were formed on sapphire substrates without the use of a catalyst via a low-
temperature (90 °C) hydrothermal growth technique. Scanning cathodoluminescence
(CL) spectroscopy and imaging is a useful technique for studying surface defects in
semiconducting materials with high spatial resolution. Surface defects in the ZnO
nanorods were characterized by measuring low-temperature CL spectra at low
University of Technology Sydney Gas Sensing Properties/ 55

acceleration voltages, at which a surface layer of only a few tens of nanometers is

excited. ZnO nanorods having surface defects of various types were further employed
for fabrication of gas sensing devices and their electrical responses to gas exposure
probed and compared with the surface point defect chemistry. In this work, ammonia
gas was employed as a model reductive gas in order to illustrate the inter-relationships
between DL defect type, electrical properties, and sensing performance in
nanostructured ZnO. The results confirm that post-growth-controlled atmosphere heat
treatment can be used to mediate the Fermi level position of the ZnO nanorods through
the creation of acceptor or donor native near surface defects, which governs the sign and
magnitude of the electrical response to NH3 gas.

4.3 Experimental Section

4.3.1 Synthesis of ZnO Nanorods and Post-Growth Processing

Nucleation of ZnO was induced by applying a few drops of zinc acetate
solution (5 mM in ethanol) on a sapphire substrate, followed by sample baking at 250
°C for 20 min. The nucleated substrates underwent a hydrothermal treatment in seed
aqueous mixture containing 0.025 M zinc nitrate hexahydrate (Zn(NO3)2·6H2O, 98%
purity) and 0.025 M hexamethylenetetramine (C6H12N4, •99.0% purity) at 90 °C for 3
h to afford as-grown ZnO nanorods.175-176 The as-grown samples were then annealed at
650 °C for 30 min in either O2 or Zn vapor. Hydrogen plasma treatment was carried out
using a mild hydrogen radio-frequency plasma for 20 min (10 W, 370 K sample
temperature).

4.3.2 Analysis Techniques

The morphologies and crystal structures of as-grown and annealed ZnO
nanorods were characterized using field-emission scanning electron microscopy (FE-
SEM) and powder X-ray diffraction (XRD), respectively. CL spectra were collected on
a FEI Quanta 200 scanning electron microscope equipped with an Ocean Optics
QE6500 system. All CL spectra were corrected for the total response of the light
Suranan Anantachaisilp Gas Sensing Properties/ 56

collection system. Measurements of electrical and gas sensing were carried out with a
6157A electrometer (Keithley Co, Ltd.).

4.3.3 Fabrication of ZnO Nanorods Gas Sensor and Gas Sensing

Measurement
The 12 fingers interdigitated pattern of aluminum electrodes with 100 ȝm
spacing and 900 ȝm length was fabricated on the ZnO nanorods using a thermal
evaporation technique. To determine the gas response, the nanostructured ZnO-based
devices were subjected to ammonia gas exposure at various concentrations in an air
carrier gas, with an applied constant voltage (1.0 V) over specifically designed
measurement and cleaning cycles (10 min measuring and 10 min cleaning). The
electrical response upon exposure to ammonia gas was attained by the measuring current
as a function of exposure time.

4.4 Results and Discussion

Figure 4.1a shows SEM images of the as-grown ‫ۦ‬0001ۧ ZnO nanorod
sample, indicating that the hydrothermal growth treatment provided a high density of
uniform hexagonal ‫ۦ‬0001ۧ nanorods approximately 50í60 nm in diameter. The ‫ۦ‬0001ۧ
nanorods are aligned over a range of angles to the substrate normal so that their sidewalls
overlap each other, allowing multiple contacts between rods. Cross-sectional SEM
images (not shown) indicate that the typical length of the ZnO nanorods is around 2000
nm, which was considerably thicker than a nanoscale wetting layer. No change in the
size and shape of the ‫ۦ‬0001ۧ ZnO nanorods was observed following the controlled
atmosphere heating at 650 °C heating, confirming that no overgrowth layer was
deposited (SEM images are presented in Figure 4.2).
Powder XRD profiles of the as-grown and annealed samples are given in
Figure 4.1b, and diffraction peaks can be indexed to the wurtzite phase of ZnO,
hexagonal lattice (P63mc) with lattice constants of a = 0.3249 nm and c = 0.5207 nm as
calculated from ICSD using POWD-12++. The presence of the Bragg diffraction peaks
in addition to the (0002) appears because the ZnO nanorods are not strictly
perpendicular to the substrate plane. No significant change in the position of XRD peaks
University of Technology Sydney Gas Sensing Properties/ 57

was observed after annealing in O2 or Zn atmosphere, indicating that no microstrain or

extended defects were introduced during the thermal processing. The change in the
relative intensity of the XRD peaks after heating arises from slight reorientation of the
‫ۦ‬0001ۧ ZnO nanowires relative to the substrate plane.

Figure 4.1: (a) SEM images of as-grown ZnO ‫ۦ‬0001ۧ nanorods on a sapphire substrate
and (b) PXRD patterns of as-grown, O2 and Zn atmosphere annealed ZnO nanorods
exhibiting the wurtzite crystal structure. The change in the relative intensity of the
diffraction peaks arises from a slight reorientation of the nanorods relative to the
substrate after thermal annealing.
Suranan Anantachaisilp Gas Sensing Properties/ 58

Figure 4.2: SEM images of ZnO nanorods after annealing in (a) O2 and (b) Zn
atmosphere.
Figure 4.3a highlights the CL spectra of the Zn vapor annealed, as-grown,
and O2 annealed ZnO nanorods at 80 K, which exhibit predominantly green
luminescence (denoted as GL), yellow luminescence (denoted as YL), and red
luminescence (denoted as RL) centered at 2.44, 1.90, and 1.70 eV, respectively. In order
to probe defects on the nanorod surface, CL spectra were collected at a low acceleration
voltage of 5 kV where the maximum photon generation depth is approximately 37 nm
using the CASINO simulation code.177 The CL spectra clearly show that each ZnO
sample exhibits a specific broad signature luminescence band, reflecting the different
types of DL defects at the ZnO surface following growth and postgrowth annealing. The
University of Technology Sydney Gas Sensing Properties/ 59

results in Figure 4.3a indicate that the hydrothermally synthesized ZnO nanorods
produce a visible emission corresponding to YL. On the other hand, annealing the as-
grown ZnO nanorods under an oxidative atmosphere resulted in ZnO producing RL,
whereas annealing under Zn vapor led to ZnO producing GL. These CL data confirm
that different DL types of point defect can be incorporated into the ZnO nanorods using
controlled gas atmosphere heat treatment.

Figure 4.3: (a) Normalized CL spectra collected from Zn annealed, as-grown, and O2
atmosphere annealed ZnO nanorods at 5 kV, 80 K, scan area = 50 ȝm × 43 ȝm, and
beam current of 0.2 ± 0.02 nA. The ratio of raw intensities of GL, YL, and RL centered
at 2.44, 1.90, and 1.70 eV, respectively, is 0.057:0.204:1. All spectra were normalized
Suranan Anantachaisilp Gas Sensing Properties/ 60

to the DL peaks. (b) Room temperature IíV characteristics of the Zn annealed, as-
grown, and O2 atmosphere annealed ZnO nanorods measured under an ambient
atmosphere.
The point defects formation reactions upon annealing the as-grown ZnO
nanorods under oxygen and Zn vapor atmosphere are given in equation 4.1 and 4.2. In
addition, the intensity ratio of near-band-edge (NBE) to DL emissions at approximately
3.3 eV was increased after annealing, especially in Zn atmosphere annealed ZnO
nanorods, suggesting that defect concentrations were lowered by thermal annealing.
ଶି
ܱଶሺ௚௔௦ሻ ൅ ܼܱ݊ ՜ ܼܱ݊ሺ௚௔௦ሻ ൅  ܸ௓௡ ൅  ʹ݄ା (4.1)
ܼ݊ሺ௚௔௦ሻ ൅ ܼܱ݊ ՜ ܼܱ݊ሺ௚௔௦ሻ ൅  ܸைଶା ൅  ʹ݁ ି (4.2)
The production of YL from the hydrothermally grown ZnO nanorods was
also observed and systematically studied by photoluminescence.79 It was also suggested
that the PL yellow emission was due to the presence of OH groups or Zn(OH)2 on the
surface of ZnO nanorods rather than oxygen interstitials. It has been shown that the YL
in hydrothermally grown ZnO could be related to deep LiZn acceptors.55, 83 Notably,
adsorption of OH groups on the ZnO surface is favourable under hydrothermal
treatments, and the chemical origin of YL produced by hydrothermally grown ZnO
could be different than those obtained by other preparation methods.
From the CL spectra (Figure 4.3a), the appearance of RL due to annealing
in an oxygen atmosphere indicates the generation of a different type of point defect at
the ZnO nanorod surface. It should be highlighted that the shift of YL to RL upon
oxidative annealing was in excellent agreement with Djurisic’s report, 79 suggesting that
RL originates from zinc vacancy defects. The red and orange-red emissions were
previously assigned as originating from oxygen interstitials (Oi),80, 82 zinc vacancies
(VZn),79 and substitutional nitrogen defects on oxygen sites.178 Our view is that the RL
is related to VZn or VZn related complex due to a surface chemical reaction where the
O2(g) reacts with the ZnO surface removing a Zn atom and creating a VZn and ZnO
vapor as shown in equation 4.1. This assignment is consistent with theoretical
calculations40 that VZn with the lowest formation energy can be easily formed in O-rich
ZnO. Electron paramagnetic resonance (EPR) and X-ray absorption spectroscopies were
employed to probe the presence of nitrogen defects. No EPR hyperfine signal
corresponding to nitrogen was observed. X-ray absorption near-edge structure
University of Technology Sydney Gas Sensing Properties/ 61

(XANES) spectra at the N K-edge (Figure 4.4) were collected from the as-grown and
annealed ZnO nanorods and compared with nitrogen-implanted ZnO nanoparticles and
bulk ZnO crystals. These data confirm the absence of N in the ZnO nanorods studied in
this work. To elucidate the chemical nature of the defect involved in the RL, the O2
atmosphere annealed ZnO was subjected to a hydrogen plasma treatment, causing the
RL to completely quench (Figure 4.5).

Figure 4.4: N K-edge XANES spectra of Nitrogen implanted ZnO crystal and ZnO
nanoparticles, and as grown, O2, and Zn atmosphere annealed ZnO nanorods.
The RL suppression can be explained by the passivation of VZn by hydrogen
donors through the formation of either (VZníH) or (HíVZníH) complexes.179 These
results suggest that the RL may originate from either VZn or a VZn-related center,
consistent with an assignment by Wang et al.180
The presence of the GL after annealing in a zinc-rich atmosphere is evidence
of a creation of another different type of defect type at the ZnO nanorod surface (see
equation 4.2). It is widely agreed that GL at around 2.5 eV is initiated from singly
ionized oxygen vacancies (ܸைା ), neutralܸை଴ , or Vo-related centers.181-183 Recently it has
suggested that this emission may be due to VZn centers;184 however, none of the
Suranan Anantachaisilp Gas Sensing Properties/ 62

experimental results in this work support this assignment in our low temperature grown
hydrothermal ZnO nanorods.
Figure 4.3b reveals the currentívoltage (IíV) characteristics, under an
ambient air atmosphere, of ZnO nanorod gas sensors fabricated from the as-grown, O2,
and Zn atmosphere annealed ZnO. The electrical conductivity of the samples
substantially increases after annealing, particularly in a Zn-rich atmosphere. In addition,
the maximum current (measured at 1.0 V) obtained from the as-grown ZnO based device
was 3.34 nA. In comparison, an approximately 100- and 1000-fold enhancement in
current resulted after annealing in O2 and Zn vapor atmosphere, respectively. It is
important to note that only a 20- fold enhancement in current was found following
annealing in an Ar atmosphere under identical conditions to the O2 and Zn vapor heat
treatments (Figure 4.6). This observation conclusively shows that the increase in
conductivity following O2 and Zn atmosphere annealing was the result of modification
to the surface defect chemistry with thermally induced improvement in the crystallinity
of the ZnO nanorods playing only a minor role.

Figure 4.5: CL spectra collected from O2 atmosphere annealed and H plasma of O2

annealed ZnO nanorods at 5kV, 80K. The spectra were normalized to the maximum
intensity peaks.
The ammonia gas sensing properties of ZnO nanorod sensors were
examined at room temperature in order to avoid any alteration of surface defects, which
may occur at elevated temperatures and to evaluate the performance of these ZnO
University of Technology Sydney Gas Sensing Properties/ 63

sensors under practical operating conditions. Note that all IíV curves in Figure 4.3 show
linear characteristics indicating good Ohmic behavior, which is ideal for sensor
applications. The NH3 gas response data (Figure 4.7) were collected in five
measurement cycles, with higher concentrations of gas being employed in subsequent
runs. The gas sensor was purged with air prior to each measurement cycle. The NH3
response sensitivity (S) can be determined using the expression ܵ ൌ  ‫ܫ‬௚ Ȁ‫ܫ‬௔ , where ‫ܫ‬௚ is
the current measured during NH3 gas flow and ‫ܫ‬௔ is the baseline current (air flow) before
exposure. From Figure 4.7, it was found as expected that the sensitivity was enhanced
when NH3 gas concentration increased from 20 to 100 ppm for all samples. Variation in
gas sensing properties of ZnO with different defect types (YL, RL, and GL) was
observed. At 100 ppm of NH3 gas, the highest sensitivity was observed from the sensor
fabricated from the O2 annealed ZnO (S = 22.6; RL) in comparison to the as-grown
nanorods (S = 4.1; YL) and the Zn vapor annealed ZnO (S = 1.4; GL). Although the
sensitivity of the as-grown nanorod-based sensor is relatively low, it is ca. 50 times
higher than the room temperature NH3 gas sensor based on hydrothermally grown ZnO
nanorods previously reported by Ang et al.,174 measured at 500 ppm (when S was
calculated by the same equation above). The sensitivity of O2 atmosphere annealed ZnO
based device was found to be ca. 280 times higher than that reported by Ang et al.174
More importantly, the NH3 gas sensitivity of hydrothermally grown oriented ZnO
nanorod arrays (S < 8.5)185 sensor at a temperature of 250 °C with a gas concentration
of 200 ppm was not significantly higher than our ZnO nanorod device operating at room
temperature. Possible surface reactions related to the NH3 gas response sensitivity of
sensors fabricated by the ZnO nanorods having particular DL defect emission are given
as follows.
ି
ܱଶሺ௚௔௦ሻ ൅  ݁ ି  ՞  ܱଶሺ௔ௗ௦ሻ ; T < 150 °C (4.3)
ି
Ͷܰ‫ܪ‬ଷሺ௚௔௦ሻ ൅  ͵ܱଶሺ௔ௗ௦ሻ  ՞ ͸‫ܪ‬ଶ ܱሺ௚௔௦ሻ ൅  ʹܰଶሺ௚௔௦ሻ ൅ ͵݁ ି (4.4)
The observed increase in electrical conductivity of the ZnO nanowires is
generally ascribed in the literature to the following surface chemistry model. Under air
flow, O2 molecules are absorbed on the ZnO surface and convert to oxygen ions through
depletion of electrons from the conduction band of ZnO, equation 4.3, lowering its
electrical conductivity.164-165, 186-187
Upon exposure to NH3 gas, the increase in the
electrical conductivity response of the ZnO gas sensor (‫ܫ‬௚ ) is attributed to surface
Suranan Anantachaisilp Gas Sensing Properties/ 64

reactions between NH3 with adsorbed oxygen ions. This process induces desorption of
the surface O2 as shown in equation 4.4, resulting in electrons being released back to the
ZnO conduction band subsequently reverting the electrical conductivity toward the
initial pre-NH3 exposure levels. This particular model, however, is inconsistent with the
CL spectroscopy analysis of the ZnO surface defects in this work. It is widely reported
in both theoretical188 and experimental189 studies that surface oxygen vacancies enhance
the adsorption of O2 via a dissociative adsorption process. Accordingly, the ZnO
nanorod sample annealed in Zn vapor with the high Vo concentration, as evidenced by
the strong GL, should exhibit the highest coverage of adsorbed O2 and consequently the
highest sensitivity to the NH3 gas. Conversely, the ZnO nanorod specimen annealed in
O2 with its VZn-rich surface, as indicated by the strong RL, should display a lower
sensitivity. However, the NH3 sensitivity results shown in Figure 4.7 reveal exactly
opposite behavior with the RL sample exhibiting the greatest NH3 sensitivity and the
GL sample the weakest. Significantly, RL samples exposed to H-plasma exhibited a
negative electrical response (Figure 4.7g,h). Given these data, an alternate model
involving annealing induced shifts of the Fermi level position of the ZnO nanorod is
presented to explain the NH3 conductivity and its relationship to the ZnO surface defect
chemistry.

Figure 4.6: Room temperature I-V characteristic of the Argon atmosphere annealed
ZnO nanorod gas sensor.
University of Technology Sydney Gas Sensing Properties/ 65

Figure 4.7: Sensitivity as a function of measurement time for (a) as-grown, (c) O2, (e)
Zn, and (g) H plasma after O2 annealed ZnO nanorod sensors and sensitivity as a
function of ammonia gas concentration for (b) as-grown, (d) O2, (f) Zn, and (h) H plasma
after O2 annealed ZnO nanorods sensors at room temperature. Labels in the bottom right
corner of (b), (d), (f), and (h) indicate the signature luminescence (YL, RL, GL, and RL
suppression) exhibited by ZnO nanorods employed in the fabrication of each gas sensor.
Suranan Anantachaisilp Gas Sensing Properties/ 66

Both bulk and nanostructured ZnO are n-type due to an autodoping process
inducing the inadvertent incorporation of intrinsic donors, such as H, Zni, and Vo as well
as their complexes that push the Fermi level (EF) toward the conduction band.40 Charge
will flow either to or from the ZnO surface and the adsorbed NH3 gas species, depending
on the position of the ZnO EF compared with chemical potential (ECP) of the adsorbed
NH3 gas.105, 190 If the EF is below the NH3 ECP, then charge with be transferred from
the gas species to the surface and vice versa for EF above the NH3 ECP. Accordingly,
annealing ZnO in O2 atmosphere produces VZn acceptors lowering the EF toward
valence below ECP, promoting electrons to move from the NH3 to the ZnO nanorod and
increasing the surface conductivity. Conversely, the generation of Vo donors by
annealing in Zn vapor raises EF toward the conduction band above ECP, causing
electrons to shift from the ZnO to the adsorbed NH3 and decreasing the surface
conductivity. Additionally, the negative electrical response (Figure 4.7g) following
hydrogen plasma treatment of the O2 anneal sample clearly confirms this alternate
model. The hydrogen in ZnO always acts as a donor.94, 191
Consequently, H
incorporation via the hydrogen plasma treatment pushes the EF toward the conduction
band above the NH3 chemical potential, leading to electron flow in the opposite
direction.

Scheme 4.1: The alteration of the ZnO Fermi level position owing to near surface
acceptor (RL: VZn) and donor (RL suppression: Hydrogen) after post growth processes.

4.5 Conclusions
In summary, the low temperature (90 °C) growth technique employed in the
preparation of ZnO nanorods in sensor devices shows great promise as an alternative
University of Technology Sydney Gas Sensing Properties/ 67

method to the existing high temperature routes due to its simplicity, application to
temperature sensitive substrates, energy efficiency, and reproducibility in obtaining
nanostructured ZnO for room temperature NH3 sensor applications. CL spectroscopy
confirmed that post-growth-controlled atmosphere heat treatment can be used to
mediate the Fermi level position of the ZnO nanorods through the creation of acceptor
or donor native near surface defects, which governs the sign and magnitude of the
electrical response to NH3 gas. Specifically a room temperature NH3 sensitivity of S =
22.6 was achieved by heating in O2 atmosphere at 650 °C that was 5 times greater than
S for as-grown ZnO nanorods. The reported findings on the inter-relationship between
surface point defects, electrical properties, and gas sensing properties of nanostructured
ZnO will facilitate the development of future, low cost, high-sensitivity, room
temperature gas sensor devices tailored for specific applications.
Suranan Anantachaisilp Surface Electronic Structure / 68

CHAPTER V
THE SURFACE ELECTRONIC STRUCTURE OF ZnO
NANORODS: THE ROLE OF SURFACE DEFECTS

5.1 Summary
The effect of native point defects on the surface electronic structure of low
temperature (90 qC) hydrothermally grown ZnO nanorods is investigated using X-ray
photoemission spectroscopy (XPS), X-ray Absorption Near-Edge Structure (XANES),
electron spin resonance (ESR), low voltage cathodoluminescence spectroscopy, sub
band gap photoluminescence spectroscopy as well as electrical I-V measurements. The
as-grown surface defect structure was altered by controlled atmosphere annealing at
650ºC in both O2 gas and Zn vapor environments. From the correlative characterization
measurements it is established that the degree of surface band bending and surface
conductivity is controlled by amount of adsorbed ambient gas species (O2, H2O and
OH), which is mediated by type and concentration of intrinsic surface point defects.
Specifically a direct relationship is established between peaks at 532.0 eV and 532.7 eV
in O1s XPS spectra attributed to presence of surface oxygen vacancies and oxygen in
adsorbed gas species. Singly ionized oxygen vacancies are confirmed to be stable on the
ZnO nanorod surface producing (i) a pronounced shoulder feature in the valence band
XPS spectra around 3 eV below the Fermi level, (ii) a strong broad green CL peak
positioned at 2.43 eV (300K) and (iii) an intense ESR line at g # 2.01 (10 K).

5.2 Introduction
A key attraction of ZnO relates to it useful and interesting surface electronic
properties, which are significantly enhanced in ZnO nanostructures with their extreme
surface-to-volume ratios.192-195 Gas – solid interfacial chemical reactions can produce
either a low conductivity surface electron depletion layer or a high conductivity surface
electron accumulation layer via physisorption and chemisorption of ambient or target
University of Technology Sydney Surface Electronic Structure/ 69

gas species on the ZnO surface. For instance, electron acceptor O2 gas molecules can
capture electrons from the bulk creating a negative surface charge above an electron
depletion layer with corresponding upward band bending. Conversely electron donor
gas species such as water vapor molecules and hydroxyl groups can cause the opposite
effect and accordingly form an electron accumulation layer with downward band
bending. In typical semiconductors the type of surface adsorbed gas depends on the bulk
electrical properties: n – type, acceptor surface states with upward band bending and
p – type, donor surface states with downward band bending.196-197 However, both
upward and downward band bending has been reported for bulk and nanostructured ZnO
despite being a naturally n-type semiconductor.196-200 This behaviour has been attributed
to variation in the type and density of native defects on the ZnO surface which affects
the adsorption efficiency of different ambient gas species.143, 197-208 To understand these
gas - surface mechanisms in detail, ZnO nanorod surfaces have been modified using
controlled atmosphere annealing techniques by design to produce surfaces with different
distributions of native point defect.202, 209-214 A range of complementary spectroscopy
techniques have been used to investigate interactions between native ZnO point defects
and ambient gas species to establish their effect on the surface optical and electrical
properties of ZnO nanorods.

5.3 Experimental section

ZnO nanorods were grown hydrothermally at 90°C for 3 hours in a seed

aqueous mixture containing 0.025 M zinc nitrate hexahydrate (Zn(NO3)2 6H2O, 98%
purity) and 0.025 M hexamethylenetetramine (C6H12N4, • 99.0% purity). Nucleation of
ZnO nanorods was achieved by applying a few drops of zinc acetate solution (5mM in
ethanol) on a sapphire substrate, which was then baked at 250°C for 20 minutes. The
hydrothermal synthesis parameters were chosen so that the nanorods elongated over a
range of angles with respect to the substrate normal to promote contact of the sidewall
surfaces, leading to good electrical conductivity across the entire nanorod ensemble.
The as-grown samples were annealed at 650°C for 30 minutes in O2 and in Zn vapor to
produce samples with three different surface electronic structures. The morphology of
Suranan Anantachaisilp Surface Electronic Structure / 70

ZnO nanorods was not changed by the post-growth heat treatment. The size and shape
as well as the crystal structure of as-grown and annealed ZnO nanorods were
characterized using field-emission scanning electron microscopy (FE-SEM) and powder
X-ray diffraction (XRD) respectively.
Electrical I-V curve measurements were conducted using a 6157A
electrometer (Keithley Co, Ltd). Aluminum electrodes consisting of pattern of 12
interdigitated fingers 900 μm length with 100 μm spacing, were deposited onto the
surface ZnO nanorod ensemble using a thermal evaporation technique.
Cathodoluminescence (CL) spectra were measured using a FEI Quanta 200
scanning electron microscope equipped with a retractable diamond machined parabolic
mirror light collector and a liquid nitrogen (80 K) cold stage. CL spectra were collected
using an Ocean Optics QE6500 optical spectrometer 300-1000 nm with a spectral
resolution of 0.9 nm. In all CL measurements, the accelerating voltage of the electron
beam was 5 kV, corresponding to a maximum CL generation depth of around 37 nm.
All CL spectra were corrected for the total response of the light collection system.
X-ray photoemission spectroscopy (XPS) and X-ray Absorption Near-Edge
Structure (XANES) were performed on the soft X-ray spectroscopy beamline at the
Australian Synchrotron. In photoemission experiments, the energy of incident X-ray
photon was adjusted to achieve a similar photoelectron kinetic energy with a ~ 3 nm
probing depth. The O1s spectra were acquired at hv = 650 eV, while the valence band
spectra were acquired at hv = 150 eV. XANES was implemented in the Total
Fluorescence Yield (TFY) mode around the O K-edge. The photon energy scale was
calibrated against the Au 4f7/2 peak at 84 eV from a clean gold film in electrical contact
with the samples. For each measurement, several scans were registered on the same spot
in order to monitor x-ray beam damage; no significant changes within the measurement
time (approximately 10 minutes) were detected.
Electron Spin Resonance (ESR) analysis was carried out at 10 K using a
Bruker Elexys E500 cw X-band ESR spectrometer equipped with an Oxford ITC605
temperature controller. All presented ESR spectra were measured using a power of
2.0 mW at a frequency of 9.4 GHz and modulation amplitude of 5.0 G.
The photoluminescence (PL) spectra were recorded using a home-built
confocal microscope with 500 nm resolution. Optical excitation was implemented using
University of Technology Sydney Surface Electronic Structure/ 71

continuous-wave blue (O = 405 nm) and green (O = 532 nm) lasers through a high
numerical aperture (NA = 0.9) objective. The PL signal was collected using the same
objective and sent to a Princeton Instruments, 300 lines/nm grating spectrometer. The
PL signal was measured using an Excelitas SPCM-AQRH-14Avalanche Photo Diode.
A dichoric mirror was used in all measurements to filter out the laser excitation
emission, and a band-pass filter was employed to select only the defect emission. The
measurements were done at room temperature and were corrected for system response.

5.4 Results and discussion

The 90qC hydrothermal growth technique yielded a continuous high-density
layer of uniform ZnO nanorods which exhibit a highly regular hexagonal cross-section
with a diameter and length of 55 ±5 nm and around 2000 nm respectively (Figure 5.1).
X-ray diffraction analysis (Figure 4.1(b)) confirmed that the nanorods are high quality
hexagonal wurtzite single crystals and that their growth axis is along the <0001>
direction.

Figure 5.1 shows SEM images of the as-grown hydrothermal ZnO nanorod ensemble
containing a high density of uniform hexagonal <0001> nanorods 50 ± 5 nm in diameter.
The oblique nanorod growth habit promotes multiple contact of the sidewalls providing
good electrical conductivity across the ensemble.
Suranan Anantachaisilp Surface Electronic Structure / 72

Typical low voltage CL spectra at 80 K are presented in Figure 5.2. The CL

peak at 3.37eV results from the ZnO near band edge (NBE) emission, containing highly
overlapped peaks from free exciton recombination and its associated phonon replicas.32
Deep level emission bands in the yellow (YL) at 1.90 eV, red (RL) at 1.70 eV and green
at 2.43 eV are observed in the as-grown, O2 and Zn vapor annealed nanorods
respectively. These broad CL peaks has been attributed in the literature to different types
of intrinsic defect centers: YL - oxygen interstitials (Oi) and LiZn acceptors,55, 79-80, 83, 215
RL - zinc vacancies (VZn)79, 82, 178, 180 and GL - oxygen and zinc vacancies VO40, 182-183,
216
. The low voltage CL measurements shown in Figure 5.2 confirm that each of the
three separate nanorod samples, as-grown, O2 and Zn vapor anneal, exhibit significantly
different surface defect structures. The appearance of the LO-phonon replica peaks on
the low energy side of the NBE indicates that the crystal quality of ZnO nanorods was
improved by the annealing process. This is also evidenced by the blue shift in the as-
grown NBE due to a decrease in carrier concentration due to the thermal removal of
inadvertently incorporated shallow donors, most likely hydrogen.217

Figure 5.2 Typical 5 kV CL spectra collected from as-grown, O2 atmosphere and Zn

vapor annealed ZnO nanorods at 80K, a scan area = 50 Pm x 43 Pm and beam current
of 0.2±0.02nA, showing yellow (YL) at 1.90 eV, green (YL) at 2.44 eV and red (RL) at
1.70 eV.
University of Technology Sydney Surface Electronic Structure/ 73

Differences in the surface defect structure are also evidenced by the I-V
results (Figure. 4.3(b)) which reveal large changes in the surface conductivity between
the three sample types. The low current observed in the as-grown samples arises from
the chemisorption of oxygen molecules which capture electrons forming charged
oxygen species (ܱଶି ሻand creating a resistive surface depletion layer.202, 204-206 After
annealing in Zn vapor a three orders of magnitude increase in the measured I-V current
was observed (Figure 5.3). This result is consistent with a high surface coverage of water
molecules and hydroxyl groups that release electrons to the ZnO nanorod surface
producing and electron accumulation layer, increasing the electrical conductivity.202, 204,
218-220
Here chemisorbed and physisorbed ambient water molecules efficiently bond and
dissociate on Zn sites at surface oxygen vacancies218 which renders the ZnO nanorod
surface hydrophilic. Lastly the O2 annealed nanorods produced a two orders of
magnitude current increase from the as-grown result due to the change in the type and
density of surface defect gas adsorption sites. This result emphasizes the effect of the
competitive nature of the oxygen and water vapor surface adsorption and its dependence
on the surface defect chemistry.

Figure 5.3 presents a log I vs log V showing the 1000x and 100x increase in the
measured current after Zn vapor and O2 annealing respectively.
Suranan Anantachaisilp Surface Electronic Structure / 74

(a) O2 g = 1.96 & 2.01

ESR Intensity (a.u.)

As-Grown
O2 annealed

As-Grown g = 1.96 & 2.01

1.92 1.96 2.00 2.04 2.08

g - value


6 (b) Zn vapor anneal

4
ESR Intensity (a.u.)

g = 2.00
2

-2

-4

-6

1.92 1.96 2.00 2.04 2.08

g - value

Figure 5.4 (a) ESR spectra of the as-grown and O2 annealed ZnO nanorods. Two main
paramagnetic ESR signals at g # 2.01 and 1.96 are assigned to singly ionized surface
oxygen vacancy defects and localized donors, respectively. (b) ESR spectrum following
Zn vapor annealing showing a significant intensity increase is g # 2.00 line after the
appearance of a broad green CL emission centered at 2.44 eV and a pronounced shoulder
in the valence band photoemission spectrum ~ 3 eV below the Fermi level.

ESR spectroscopy of the as-grown nanorods revealed weak signals at

g = 1.96 and g =2.01 (Figure. 5.4(a) and 5.4(b)). Both these ESR lines became slightly
University of Technology Sydney Surface Electronic Structure/ 75

sharper following annealing in O2 with no significant increase in intensity most likely

due to an increase in the mean spin-spin separation following the thermal enhancement
in the crystal quality.221 The g = 1.96 signal has been attributed to a number of different
point defects in ZnO, including zinc interstitials (ଶା
୧ ሻ,
222
oxygen interstitials (ି
୧ )
223

and zinc vacancies (ܸ୞୬ ି )223 as well as conduction band states related to shallow donor
impurities, such as H224 and Ga, In and Al.225 Centers involving ଶା
୧ and H can be ruled

out as these defects have high thermal mobilities in ZnO.226-227 Consequently in ZnO
nanostructures ଶା
୧ and H should easily diffuse to the surface and expelled during high

temperature heat treatment. Furthermore since there was no change in the intensity of
ି
g = 1.96 with O2 annealing ି
୧ and ܸ୞୬ can be excluded as well leaving Ga, In and Al

shallow donor impurities as the most likely chemical origin of the g = 1.96 ESR line.
The weak g # 2.01 signal has been attributed to singly ionized surface
oxygen vacancies ܸ௢ା in photo-ionization experiments,228 however, its assignment
remains controversial. In this work it was found that the intensity of the ESR line around
g # 2.02 increased significantly after the ZnO nanorods were annealed in Zn vapor
(Figure 5.4(b)). This result provides compelling support for the assignment of the
g = 2.02 line to surface ܸ௢ା since it is expected that Znvap annealing will cause strong
gas-solid surface reduction reactions that create oxygen vacancies as substantiated by
the presence of strong GL band in the CL analysis (Figure 5.2). Although the ܸ௢ା is
theoretically predicted to spontaneously transform to either a doubly ionized ܸ௢ଶା center
or a neutral ܸ௢ை by either emitting or capturing an electron respectively,40 the correlative
surface characterization results in this work strongly suggest that ܸ௢ା center is
energetically stable at the surface of ZnO nanorods.
The O1s XPS spectra for RL, GL and YL ZnO nanorod samples presented
in Figure 5.5. These XPS data were all curve fitted into three peaks using a pseudo-
Voigt function with a consistent set of peak position and FWHM curve fitting
parameters at 530.5 eV (Peak OI), 532.0 eV (Peak OII) and 532.7 eV (Peak OIII). Peak
OI is ascribed to O2- ions surrounded by fully co-ordinated by Zn and O atoms in the
fully stoichiometric wurtzite crystal structure, peak OII is related to either O2- ions in the
vicinity of VO defects or OH groups and peak OIII is attributed to oxygen in strongly
bound adsorbed gases, specifically H2O and O2 molecules.188, 229-232
The relative
Suranan Anantachaisilp Surface Electronic Structure / 76

percentage for each of these three peaks for the as-grown, Zn annealed and O2 annealed
specimens are as follows: as-grown OI – 31%, OII 41% and OIII 28%; O2 anneal OI –
38%, OII 38% and OIII 24%; and Zn vapour anneal - OI – 18%, OII 51% and OIII 31% (as
shown in table 5.1).

(a) As-Grown Measured

Fit
Background
Intensity (a.u.)

II
Peak I
Peak II
I Peak III
III

538 536 534 532 530 528 526

Binding Energy (eV)

(b) O2 anneal Measured

Fit
II I Background
Intensity (a.u.)

Peak I
Peak II
III Peak III

538 536 534 532 530 528 526

Binding Energy (eV)
University of Technology Sydney Surface Electronic Structure/ 77

(c) Zn vapor anneal Measured

Fit
Background

Intensity (a.u.)
II
Peak I
Peak II
Peak III
III

538 536 534 532 530 528 526

Binding Energy (eV)
Figure 5.5 The O1s XPS spectra for YL, RL and GL ZnO nanorod samples are shown
in Figures 5.5(a), 5.5(b) and 5.5(c) respectively. The XPS profiles are all curve fitted
into three peaks at 530.5 eV (Peak OI), 532.0 eV (Peak OII) and 532.7 eV (Peak OIII).
Peak OI results from O2- ions fully co-ordinated by Zn and O atoms in the fully
stoichiometric wurtzite crystal structure, peak OII is due to either O2- ions in the vicinity
of VO defects and peak OIII arises from oxygen in strongly bound adsorbed gases,
specifically H2O and O2 molecules.

Table 5.1 The relative percentage of the integrated intensity of the OI, OII and OIII peaks
from the O1s XPS emission for the as-grown, Zn annealed and O2 annealed specimens.

Peak OI Peak OII Peak OIII

Crystalline ZnO O Vacancies Absorbed Gas

As grown 31% 41% 28%

O2 anneal 38% 38% 24%

Zn anneal 18% 51% 31%

Suranan Anantachaisilp Surface Electronic Structure / 78

The XPS results show that annealing in O2 leads an increase in fully

coordinated oxygen and stoichiometry (peak OI) with a corresponding decrease in the
concentration of oxygen vacancies [VO] (peak OII). While annealing in Zn vapor
increases the [VO] from 30% to 50%. This direct correlation between the OI and OII
fractions is in agreement with their relationship to the degree of stoichiometry of the O2-
ions or more specifically the density of oxygen vacancies and is consistent with the ESR
and CL findings unequivocally confirming the presence of surface VO centers. These
data also strongly suggest that in our ZnO nanorods the OII peak is related to VO defects
rather than adsorbed surface OH groups. The ratio between the OIII / OII percentages is
around ~0.6 for all three samples indicating that the amount of the adsorbed gas species
(OIII) is directly proportional to the surface Vo concentration and that these centers play
an important role in surface gas adsorption as suggested by in other theoretical and
experimental work.189, 207, 218
Valence band X-ray Photoemission Spectroscopy (VB-XPS) spectra taken
from the ZnO nanorod samples are presented in Figure 5.6(a). The VB-XPS spectra for
all three samples consist of two dominant near edge peaks one centered at ~ 5 eV
ascribed to O 2p related states the other ~ 7 eV attributed to hybridized Zn 4s + O 2p
states.233-236 The Zn 3d peak is located ~13 eV at its signature position for wurtzite ZnO.
Using the VB-XPS spectra, the magnitude of surface band bending, VBB, can be
determined using the following relationship:
ୠୠ ൌ  ‫ܧ‬௚ െ ‫ܧ‬஻ா௖௨௧௢௙௙ െ ‫ܧ‬஼ி
where Eg is the ZnO band gap (= 3.37 eV) and ECF is the energy difference between the
୩ౘ ୘ ୒
Fermi level and the conduction band with ‫ܧ‬஼ி ൌ Ž ቂ ౙ ቃ, Nc is the conduction band
ୣ ୬౛

effective density of states [= 2.94 × 1018 cmí3 for ݉௘‫ כ‬ൌ ͲǤʹͶ݉௘ ], ne is the bulk carrier
density and e, T and kb all have their usual meaning. The EBEcutoff was found using the
intersection of extrapolated linear fits to the leading edge of valence band photoemission
and the photoemission background as shown in Figure 5.6(b). Using typically reported
literature values for the electron density in ZnO nanowires, ne = ~ 1 x 1017 cm-3,237
surface band bending values of - 0.42 ± 0.02V and - 0.24 ± 0.02V for the as-grown and
O2 annealed samples respectively. These negative values indicate downward band
bending due to surface electron accumulation on the ZnO nanorod, which is attributed
University of Technology Sydney Surface Electronic Structure/ 79

to adsorbed water molecules and hydroxyl groups (Figure. 5.7.). Less downward band
bending after O2 annealing indicates a smaller amount of negative charge transfer to the
surface by a reduced number of adsorbed water molecules and hydroxyl groups on the
ZnO nanorods due to the change in the surface defect chemistry.
Thermally induced changes in ne during O2 annealing can be dismissed
because heating at elevated temperatures should produce a reduction in ne either through
the loss of H or Zni donors because of their high mobility or via the formation of
compensating acceptor defects. Both these processes would cause a corresponding
increase in the degree of band bending rather than a decrease. Given this, the decrease
in negative band bending after O2 annealing is attributed to a reduction in the number
of adsorbed water and hydroxyl donors as established with the I-V results. However, it
is important to note that the I-V data was collected under ambient conditions with both
physisorbed and chemisorbed molecules on the ZnO nanorods whereas the XPS
measurements are taken under ultra high vacuum conditions were physisorbed species
are pumped away leaving strongly bound chemisorbed species.
Heating in zinc vapor produced a strong shoulder feature in the VB-XPS
spectra around ~ 3eV below the EF with a shift of the VB photoemission onset to ~ 1 eV
below the EF (Figure. 5.6(b)). These results together with the corresponding CL data
showing strong GL as well as the ESR results and the O 1s XPS data unequivocally
confirm the presence of electron populated surface oxygen vacancy states above the
valence band following Zn vapor annealing. The corresponding strong ESR signal
indicates that these centers must be paramagnetic singly ionized surface oxygen
vacancies. The presence of surface oxygen vacancies is also supported by the observed
broadening of the O 2p + 4 s and O 2p states due to lattice disordering (Figure. 5.6(b)).
Suranan Anantachaisilp Surface Electronic Structure / 80

(a) Zn3d
Zn anneal
as-grown
O2 anneal
Intensity (a.u.)
hQ = 160 eV

Zn4s + O2p EF
O2p

16 14 12 10 8 6 4 2 0
Binding Energy (eV)

Figure 5.6 Valence band X-ray Photoemission Spectroscopy (VB-XPS) spectra taken
from the as-grown, O2 anneal and Zn vapor anneal ZnO nanorod samples are presented
in Figure. 5.6(a). The VB-XPS spectra for all three samples consist of two dominant
near edge peaks one centered at ~ 5 eV ascribed to O 2p related states the other ~ 7 eV
attributed to hybridized Zn 4s + O 2p states. The leading edge of the valence band
photoemission showing the pronounced shoulder around 3 eV below the Fermi level
that forms after Zn vapour annealing is shown in Figure 5.6(b).
University of Technology Sydney Surface Electronic Structure/ 81

Figure 5.7 Downward band bending on the as-grown ZnO nanorod due to electron
accumulation and its decrease following O2 annealing using the valence band
photoemission spectroscopy results in Figure 5.6 (a) and 5.6 (b).
Suranan Anantachaisilp Surface Electronic Structure / 82

PL 405 nm laser as-grown

PL 405 nm laser O2 anneal
1.0 PL 532 nm laser as-grown
PL 532 nm laser O2 anneal
300 K

Normalised PL Intensity
0.8

0.6

0.4

0.2
1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3
Energy (eV)
Figure 5.8 Comparison of continuous-wave 532 and 405 nm sub band gap laser
excitation PL spectra at 300 K for ZnO nanorods as-grown and annealed in O2 gas. The
sub band gap excited spectra reveal a common broad OL emission at 1.95 eV except for
the O2 annealed sample illuminated with blue light which emits a broad RL around
1.83 eV indicated a different defect structure.

Finally sub-band gap excitation photoluminescence spectroscopy using both

a blue (3.06 eV) and green laser (2.39 eV) was employed to establish if there were any
additional defect structure differences between the as-grown and O2 annealed ZnO
nanorods. This technique enables the detection of defect centers that not usually
accessible with CL and above band gap PL excitation because of the presence of more
efficient recombination channels. Bulk defects can be probed as well. These sub-band
gap generated PL emission peaks are attributed to both excitation and radiative
relaxation of (i) electrons in ionized acceptor centers and (ii) valence band electrons to
shallow donor states178, 180 Normalized PL spectra at 300 K excited using blue and green
laser light from the as-grown and O2 annealed ZnO nanorods are shown in Figure. 5.8.
A broad orange emission band positioned at 1.95 eV (OL) is observed in as-grown ZnO
nanorods with both laser excitation sources. Green laser light illumination produced an
identical OL in the O2 annealed sample, however, the higher energy blue laser excited
a broad red emission centered around 1.83 eV comprised of a number of broad emission
University of Technology Sydney Surface Electronic Structure/ 83

peaks. Although it is highly likely that the RL and OL sub band gap emission bands
relate to native acceptor centers involving VZn or Oi defects respectively, the assignment
of these PL peaks to specific defects requires further work. Nevertheless, the sub-band
gap laser PL results confirm that O2 annealing changes the defect structure of the ZnO
nanorods through the formation surface VZn related centers. The presence of these
centers could account for the observed decrease in the magnitude of negative band
bending.

5.5 Conclusions
Optical and electrical characterization studies of ZnO nanorods before and
after annealing in O2 gas and Zn vapor have confirmed that the surface electronic
structure can be bespoke for specific sensor applications. These correlative spectroscopy
measurements reveal clear evidence of direct interrelationship between the electronic
properties of the ZnO nanorod surface and its surface defect structure. In particular, the
role of intrinsic point defects on the physisorption and chemisorption of water molecules
and hydroxyl groups which, through charge transfer mechanisms, govern both the
surface band bending effects and the surface conductivity. Additionally O1s XPS
spectra established that the number of adsorbed water and hydroxyl species on the ZnO
nanorod is directly proportional to the concentration of surface oxygen vacancies. An
important result of these studies is compelling experimental evidence for the existence
of stable singly ionized oxygen vacancies (ܸைା ) at the ZnO nanorod surface.
Furthermore, it was shown that (i) a broad green CL peak centered at 2.43 eV, (ii) an
intense ESR line at g # 2.01 and (iii) a pronounced shoulder in the valence band X-ray
photoemission spectra ~ 3.0 eV below the Fermi level, can all be attributed to centers
involving surface ܸைା centers.


Suranan Anantachaisilp Red Luminescence / 84

CHAPTER VI
NATURE OF RED LUMINESCENCE IN OXYGEN TREATED
HYDROTHERMALLY GROWN ZINC OXIDE NANORODS

6.1 Summary
A strong broad red luminescence (RL) peak centered at 1.69 eV
(FWHM = 0.57 eV) at 15K [1.78 eV (FWHM = 0.69 eV) at 300 K] is formed in ZnO
nanorods hydrothermally grown at low temperature following thermal annealing at
650°C for 30 minutes in an O2 gas environment. The optical properties of this peak were
comprehensively studied using a range of characterization techniques, including
photoluminescence and cathodoluminescence spectroscopy, X-ray absorption near edge
spectroscopy and electron paramagnetic resonance spectroscopy. With decreasing
temperature the RL peak position red shifted and its FWHM became narrower in
accordance with the configuration coordinate model. Using these results, the RL has
been assigned to highly lattice coupled VZn-related acceptor-like centers. No correlation
was found between the observed red luminescence and nitrogen impurities.

6.2 Introduction
Of particular importance is the luminescence of ZnO with optical emission
that spans the ultra-violet, visible and near infrared spectral range. In the ultra-violet the
emission is dominated by radiative near band edge (NBE) recombination mechanisms
including, free and bound excitons as well as free-to-bound and donor acceptor pair
transitions involving shallow defects. While deep level defects produce broad emission
peaks via a variety of recombination and internal mechanisms in the green (~2.5 eV and
2.3 eV) 181-183, 216, yellow (~2.1 eV) 55, 79, 83, orange (~1.9 eV) 80 and red (1.7 eV) 79, 82,
178
at 300 K.
Despite considerable research effort unequivocal assignment of these
luminescence peaks to specific dopants, impurities as well as defects and their
University of Technology Sydney Red Luminescence / 85

complexes has not been realized up to now. Nevertheless there is general agreement that
oxygen (VO) and zinc (VZn) vacancies and interstitials (Oi and Zni) play a major role in
these luminescence processes in undoped ZnO. To date most optical characterisation
studies have focused on the green luminescence and to a lesser extent the yellow and
orange emission in ZnO. However, there is currently growing interest in the red
luminescence (RL) for a number of reasons. First it has been recently suggested that the
RL could be a signature emission for N acceptors in ZnO [13] which is of importance
to the development of stable p-type material. However, RL has also been attributed to
zinc vacancies (VZn) 79, 180, oxygen interstitials (Oi) 80, 82, oxygen vacancies (VO) 238 as
well as the Zni-VO complex or VZnVO di-vacancies.88 Second gas sensor studies have
revealed a significant enhancement in the electrical sensitivity of the ZnO surface to
specific target gases, such as NH3, when RL dominates the luminescence spectrum.113
Finally the second-order correlation [g2(0)=0.1] of photons in the red-orange spectral
range confirming single photon emission in ZnO has been recently reported.215, 239
In this work the temperature and power density dependence of the RL in
ZnO have been systematically investigated using cathodoluminescence spectroscopy.
These data have been compared to ESR and photoluminescence NBE spectroscopy
results as well as ESR and XANES data. Collectively these correlative studies indicate
that the RL is associated with highly lattice coupled ܸ௓௡ -related acceptor –like radiative
recombination centers.

6.3 Samples and Experiments

6.3.1 Fabrication and Post Growth Processing of ZnO Nanorods

Zinc acetate solution (5mM in ethanol) was dropped on a silicon wafer to
induce ZnO nucleation and then baked at 250°C in air for 20 minutes. ZnO nanorods
were grown using a hydrothermal treatment by placing the nucleated substrates in seed
aqueous mixture containing 0.025 M zinc nitrate hexahydrate and 0.025 M
hexamethylenetetramine at 90°C for 3 hours.175-176 Analytical reagent grade zinc acetate
dihydrate (>98%, (C2H3O2)2Zn˜2H2O), zinc nitrate hexahydrate, (98%,
Zn(NO3)2 6H2O), and hexamethylenetetramine (HMT, •99.0%, C6H12N4) from Aldrich
Suranan Anantachaisilp Red Luminescence / 86

were employed as received with no further purification. The as-grown nanorod samples
were annealed at 650°C for 30 minutes in O2 gas environment. Hydrogen plasma
treatment was conducted at a sample temperature of 100 qC using a mild 10 W hydrogen
radio-frequency plasma for 20 minutes with 10 sccm H2 flow rate.

6.3.2 Characterization of hydrothermally grown ZnO Nanorods

6.3.2.1 Cathodoluminescence
The cathodoluminescence (CL) spectra were collected using a
FEI Quanta 200 scanning electron microscope equipped with the option of either an
Ocean Optics QE6500 system or a Hamamatsu S7011-1007 CCD sensor. The Ocean
Optics QE65000 system provides CL collection from 300-1000 nm with a spectral
resolution of 0.9 nm. CL was collected by a diamond machined parabolic mirror
positioned above the sample with an aperture to facilitate transmission of the electron
beam. In temperature resolved experiments, the sample was placed on a liquid helium
cold stage with the temperature varying from approximately12 K to 300 K. In all CL
experiments, the accelerating voltage of the electron beam was 5 kV which corresponds
to a maximum photon generation depth of around 37 nm.177 The electron beam current
was varied from around 0.013 nA to 2.54 nA in power-density measurements. All CL
spectra were corrected for the total response of the light collection system

6.3.2.2 Photoluminescence
High resolution PL was done utilizing the fourth harmonic
generation (266 nm) of a Nd:YAG laser. A monochromator with a 1200 lines/mm
grating (spectral resolution 0.2 meV) was used to disperse the emitted light. In
temperature resolved experiments, the PL measurements were performed in a liquid
helium bath cryostat allowing measurements from 7K to 300 K. A HeCd Laser with the
325 nm emission line and an output power of 36 mW was used to excite the ZnO
samples above the band gap. The emitted light was dispersed by a Spex-1404 double
monochromator (spectral resolution 50 μeV) and detected by a bi-alkali photodetector.
University of Technology Sydney Red Luminescence / 87

6.3.2.3 X-Ray Absorption Near Edge Spectroscopy (XANES)

Total Fluorescence Yield (TFY) and Total Electron Yield
(TEY) mode XANES analysis were performed around the N K-edge on the soft X-ray
Spectroscopy beamline at the Australian Synchrotron. The incident X-ray beam was
linearly polarized and perpendicular to the substrate surface. The photon energy scale
was calibrated against the Au 4f7/2 peak at 84 eV from a clean gold film in electrical
contact with samples.

6.3.2.4 Electron Spin Resonance

The ESR spectra were collected using a Bruker Elexys E500 cw
X-band ESR spectrometer equipped with an Oxford ITC605 temperature controller. All
presented ESR spectra were measured at 10 K using a power of 2.0 mW at a frequency
of 9.4 GHz and modulation amplitude of 5.0 G.

6.4 Results and Discussion

The low temperature 90 qC hydrothermal growth technique produced a
continuous high density film of ZnO nanorods approximately 55 nm in diameter and
2000 nm in length (Figure 6.1). XRD analysis (Figure 4.1(b)) confirmed that the ZnO
nanorods have a hexagonal wurtzite structure (P63mc) with lattice constants of
a = 0.3249 nm and c = 0.5207 nm and that their longitudinal axis was normal to the c-
plane. The <0001> ZnO nanorod ensemble are oriented at various angles to the normal
direction of the substrate plane. No change in the crystal structure, morphology or
orientation of the ZnO nanorods was observed after annealing at 650 ºC in either O2 or
Ar.
Suranan Anantachaisilp Red Luminescence / 88

Figure 6.1: SEM image of the as-grown <0001> ZnO nanorod ensemble, showing that
the low temperature hydrothermal growth treatment produced a high density of
uniform hexagonal <0001> nanorods approximately 55 nm in diameter orientated at
different angles to the normal direction of the substrate.

Figure 6.2a shows before and after annealing CL spectra at 300 K from the
“as-grown” and O2 heat treated nanorod films. The “as-grown” sample exhibits a ultra-
violet peak centered at 3.17 eV arising from a number of highly overlapped thermally
broadened LO-phonon replicas of the free exciton (FX) emission. A broad orange peak
positioned at 1.99 eV is also observed in the as-grown nanorods which has been
previously attributed to Li acceptors 240 and Oi related centers 215. After annealing in O2
at 900 qC the NBE emission increases in intensity by a factor of 5 due to a thermally
induced improvement in the quality of the ZnO nanorods following a reduction in the
number of point and extended defects that provide non-radiative recombination
pathways. Additionally and more significantly the O2 heat treatment created a strong,
highly symmetric red luminescence at 1.78 eV and FWHM of 0.69 eV. Figure 6.2b
shows that this emission was found to rapidly quench following exposure to a mild H2
plasma strongly suggesting that the chemical origin of the RL is related to acceptor-like
centers as hydrogen is a well-known donor in ZnO. The accompanying large increase
in the NBE could indicate that the RL centre acts as an efficient competitive
recombination channel to the NBE, however, further H passivation of non-radiative
pathways cannot be ruled out. At 15 K the RL peak position, see Figure 6.2c, slightly
red shifts to 1.69 eV and the FWHM narrows to 0.57 eV, which can be interpreted by a
University of Technology Sydney Red Luminescence / 89

241
configurational coordinate (CC) model . Furthermore the absence of any fine
structure in the peak at 15 K eliminates the assignment of the observed RL to transition
metal impurities, such as Fe, Co and Ni, which exhibit characteristic sharp features due
to internal 3d transitions 242-243. The presence of a broad emission at 2.5 eV has been
attributed to surface and bulk VO centers at 5 K 179, 215. In addition at 15 K the NBE blue
shifts revealing a structured UV emission at 3.31 eV. This peak and its phonon replicas
have been attributed to acceptors in ZnO involved in donor acceptor pairs and free-to-
bound acceptor transitions as well as to excitons bound to structural or surface or
defects. 244-245
CL spectra as a function of electron beam excitation current at 80 K and
300 K normalized at the maximum of the RL peak are shown in Figure 6.3a and 6.3b
respectively. The RL does not exhibit a blue shift with increasing power-density a
typical hallmark characteristic of the donor acceptor pair (DAP) recombination
mechanism. The CL intensity versus electron beam current displays a power-law
dependence on beam current with ICL v IBn. Log (ICL) versus log (Ib) plots at 300 K
using the data in Figure 6.4a and 6.4b provide power law exponent: nRL = 0.74 ± 0.03
and nNBE = 1.22 ± 0.04 at 80 K, and nRL = 0.81 ± 0.02 and nNBE = 1.08 ± 0.09 at 300 K.
The linear excitation power dependence of the NBE is expected as the
excitonic recombination rate is faster than the carrier injection rate by the energetic
electron beam. The nRL values are larger than expected for typical phonon-coupled
defect transitions which are typically around n ~ 0.5 because of the longer radiative
relaxation times. However, the slightly high n value of the RL could possibly be
explained by either a high RL defect concentration or possibly increased RL emission
due to excitation via an energy exchange mechanism involving the relaxation of free
excitons 240.
Suranan Anantachaisilp Red Luminescence / 90

Figure 6.2: Typical cathodoluminescence spectra collected with 5 kV, IB = 0.2 ±

0.02 nA and scan area = 50 Pm u43 Pm. Figure 6.2a ZnO nanorods before (“as-grown”)
and after annealing at 650 qC in O2 gas for 30 minutes, showing the creation of the red
luminescence (RL) peak centered at 1.78 eV with a FWHM = 0.69 eV. Figure 6.2b
illustrates rapid quenching of the RL following exposure to a mild hydrogen plasma
University of Technology Sydney Red Luminescence / 91

indicating that the RL relates to an acceptor-like center as hydrogen is a donor in ZnO.

Figure 6.2c. RL spectrum at 15 K showing the peak red shifted to 1.69 eV and narrowing
of the FWHM to 0.57 eV. The assignment of the RL to transition metal impurities is
ruled out because of the absence of fine structure arising from internal 3d transitions.


Figure 6.3: Normalized RL intensity versus excitation power (5 kV x IB) at 80 K (a) and
300 K (b). The absence of a blue peak shift with increasing beam current eliminates the
assignment of the RL to donor acceptor pair recombination transitions.

The temperature resolved CL spectroscopy of the O2 annealed ZnO

nanorods sample is presented in Figure 6.5 from 93 K increasing to 300 K. Figure 6.6
presents the temperature dependence of the peak position (Figure 6.6a), FWHM (Figure
6.6b), and emission intensity (Figure 6.6c), of the RL. The peak position ERL(T) exhibits
a roughly linear blue shift with increasing temperature with ERL(T) = E(0K) + ĮkbT,
Suranan Anantachaisilp Red Luminescence / 92

where E(0K) = 1.66 eV and Į = 0.296 ± 0.009 meV / K in agreement with the CC
model. The FWHM of the RL can be expressed below derived from the CC model.246

԰߱
‫ܯܪܹܨ‬ሺܶሻ ൌ ‫ܣ‬Ǥ ඨܿ‫ ݄ݐ݋‬൤ ൨
ʹ݇௕ ܶ

The vibrational energy of the excited state, ԰߱, for the RL was found to be
34.7 ± 0.6 meV, where FWHM(0) is the peak width approaching 0 K.

Figure 6.4: Plot of log (ICL) versus log (IB) at 80 K (a) and 300 K (b) with IB in nA. The
gradient, n, is the power law exponent in the relationship, ICL v IBn. The nNBE ~ 1 is
expected for excitonic recombination. The nRL ~ 0.8 is higher than expected for a lattice
coupled defect most likely due to a high concentration of RL centers.
The CL intensity (ICL) versus temperature dependence for RL is presented
in Figure 6.6c for temperatures decreasing from 300 to 93 K following the usual thermal
quenching relationship
University of Technology Sydney Red Luminescence / 93

‫ܫ‬ሺͲሻ
‫ܫ‬஼௅ ሺܶሻ ൌ 
െ‫ܧ‬
൤ͳ ൅ ‫ ݌ݔ݁ܣ‬ቀ ஺ ቁ൨
݇௕ ܶ
where EA is the activation energy for thermal quenching and A is a constant. An
activation energy of EA = 78.8 ± 5.1 meV was obtained from an Arrhenius plot (not
shown) using the data in Figure 6.6c which is most likely due to thermal quenching of
competitive non-radiative recombination centers. Remarkably a different ICL versus
temperature profile is observed when heating from 93 to 300 K compared to cooling
from 300 to 93 K. Specifically with increasing temperature a faster thermal quenching
of the luminescence is observed followed by an anomalous increase in ICL at around
180 K then further decay to 300 K. This thermal behavior can be explained by the
presence of competitive charge traps that are created during the O2 thermal annealing
process. These traps capture carriers at low temperature which are subsequently released
with further heating and recombine radiatively producing an increase in intensity of the
247
RL . The existence of excess carriers arising from thermal depopulation of charge
traps filled at low temperature is also evidenced by a corresponding “negative” thermal
quenching of the NBE as shown in Figure 6.4a and 6.4b.
The RL has been ascribed to both singly ionized N on an oxygen site, ܰைି
178 ଶି 180
and a doubly ionized zinc vacancy,ܸ௓௡ . Although the chemical origin of the
ionized acceptor is different in these two models, the underlying luminescence
mechanism is the same. Here, the electron at a singly (doubly) ionized acceptor, ‫ିܣ‬
(‫ܣ‬ଶି ) is photoexcited creating an electron in the conduction band, ݁஼஻ , and a neutral
(singly ionized) acceptor, ‫ܣ‬௢ (‫ ) ିܣ‬which induces a large lattice dilation, producing a
strong Jahn-Teller distortion. The excited ‫ܣ‬௢ (‫ ) ିܣ‬radiatively relaxes to ‫ܣ( ିܣ‬ଶି ), via
capture of a conduction band electron, generating RL due to the Stokes shift caused by
the Jahn-Teller distortion. The excitation and relaxation cycles for singly and doubly
ionized acceptors are given below.

ሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬԦ‫ ିܣ‬൅ ܴ‫ܮ‬

ሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬԦ‫ܣ‬଴ ൅  ݁஼஻ ‫ݐܽݔ݈ܽ݁ݎ‬ଓ‫݊݋‬
‫ ିܣ‬൅ ԰߱݁‫ܿݔ‬ଓ‫ݐܽݐ‬ଓ‫݊݋‬
ሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬԦ‫ܣ‬ଶି ൅ ܴ‫ܮ‬
ሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬԦ‫ ିܣ‬൅  ݁஼஻ ‫ݐܽݔ݈ܽ݁ݎ‬ଓ‫݊݋‬
‫ܣ‬ଶି ൅ ԰߱݁‫ܿݔ‬ଓ‫ݐܽݐ‬ଓ‫݊݋‬
Suranan Anantachaisilp Red Luminescence / 94

Figure 6.5: CL intensity versus temperature at 80 K for the RL (a) and NBE (b) with
300 K with 5 kV, IB = 0.2 ± 0.02 nA and scan area = 50 Pm u 43 Pm. The red shift and
narrowing of the RL with decreasing temperature is consistent with configuration
coordinate model for radiative recombination. The blue shift and appearance of phonon
replicas as the temperature approaches 93 K is expected for excitonic emission with the
bandgap shrinkage. The intensity of the RL and NBE exhibits “negative” thermal
quenching with increasing temperature.
It is important to note that in CL the recombination energy provides the
excitation energy for these processes. CL spectra of the NBE (Figure 6.7a) show three
peaks, the first at 3.36 eV ascribed to a donor bound exciton, DoX, and two others
positioned at 3.32 and 3.26 eV. The latter two peaks have been attributed to nitrogen
acceptors in DAPs 32 by other works, however, this assignment in our ZnO nanorod is
implausible as no nitrogen was detected using high sensitivity chemical analysis, as
University of Technology Sydney Red Luminescence / 95

discussed below. The corresponding CL spectra shown in (Figure 6.7b) reveal an anti-
correlation between the intensity of the RL and the NBE peaks at 3.32 and 3.26 eV. It is
also noteworthy that the peak position of these two NBE peaks is fixed as the
temperature increases from 7 to 100 K (Figure 6.8a and 6.8b) despite the red shift of the
DoX with the increase in the band gap due to electron-phonon scattering and lattice
expansion. The absence of a thermal shift can be accounted for by a free-to-bound
transition (FBT) not a DAP recombination mechanism. In FBT recombination at
temperatures from 7 to 100 K the thermal energy of the free carrier, kbT, roughly offsets
the red shift of the bandgap with increasing temperature with ‫ܧ‬ி஻் ൌ ‫ܧ‬௚ െ ‫ܧ‬஺ ൅ ݇௕ ܶ,
where ‫ܧ‬ி஻் is the position of the FBT peak and ‫ܧ‬஺ is the ionization energy of the
acceptor.
ZnO nanrods exposed to a mild nitrogen plasma show a well-defined set of
N kD peaks in the Total Fluorescence Yield (TFY) XANES spectroscopy results
(Figure 6.9) which are clearly absent in the as-grown and O2 anneal samples. TFY mode
was select to collect XANES data from the entire volume of the nanowire ensemble.
These results strongly suggest that the concentration of N in the ZnO nanorods studied
in this work is extremely low and consequently the RL at 1.76 eV and the NBE peaks
at 3.32 and 3.26 eV are all not related to recombination centers involving N. Electron
Spin Resonance (ESR) spectroscopy results are shown in Figure 5.4(a). No ESR
hyperfine signal corresponding to nitrogen was detected. Two very weak resonances at
g ~ 1.96 and g ~ 2.01 with similar intensities are observed before and after annealing in
O2. These ESR lines, attributed to shallow donor impurities and singly ionized surface
oxygen vacancies respectively215, clearly have no connection with the appearance of the
RL following O2 annealing and the RL has no correlated ESR resonance over the range
ଶି
of measured g-values supporting its assignment to the non-paramagnetic ܸ௓௡ center.
Suranan Anantachaisilp Red Luminescence / 96

Figure 6.6: The plot of the RL peak position and peak width as a function of temperature
are shown in Figure 6.6a and Figure 6.6b. Figure 5.6c displays the RL intensity
dependence on temperature. With decreasing temperature from 300 K to 93 K a typical
thermal response relationship with an activation energy of EA = 78.8 meV is found.
However, increasing the temperature from 93 K to 300 K “negative” thermal quenching
is observed arising from the presence of shallow levels that fill with carriers at low
temperature during electron beam excitation and depopulate with rising temperature
increasing the luminescence intensity.
University of Technology Sydney Red Luminescence / 97

Figure 6.7: CL spectra at 80 K of the NBE and RL before and after thermal annealing in
O2 are shown in Figure 6.7a with corresponding CL spectra in Figure 6.7b. Figure 6.7a
reveals emission due to donor bound exciton around 3.36 eV and peaks at 3.32 and
3.26 eV ascribed to the presence of acceptors in either DAP or free-to-bound transitions.
No interrelationship between the 3.32 and 3.26 eV NBE peaks and the RL in the CL
spectra is observed. Additionally there is no correlation between the free-to-bound NBE
and RL peaks and nitrogen as confirmed in the N kD XANES results in Figure 6.9.
Suranan Anantachaisilp Red Luminescence / 98

Figure 6.8: The temperature dependence of the DoX and peaks at 3.32 and 3.26 eV are
shown in Figure 6.8. The data show the usual red shift of the DoX with increasing
temperature as the band gap widens due to lattice expansion and the electron-phonon
effect. The absence of a corresponding shift of the 3.32 and 3.26 eV peaks between 7
and 100 K indicates that free carriers are involved in the recombination process
providing thermal energy (kbT) which offsets the band gap red shift. This result
suggests that these emission peaks arise from free-to-bound transition rather than a
DAP recombination process.

From the characterization studies above, the broad RL positioned at 1.78 eV

at 300 K in our O2 annealed ZnO nanorods has been cautiously ascribed to highly lattice
coupled VZn-related acceptor like centers. Its CL excitation and relaxation cycle is as
follows:
ଶି
ܸ௓௡ ሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬԦܸ௓௡
൅  ‫ܧ‬௥௘௖௢௠௕௜௡௔௧௜௢௡ ݁‫ܿݔ‬ଓ‫ݐܽݐ‬ଓ‫݊݋‬ ି ሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬሬԦܸ௓௡
൅  ݁஼஻ ‫ݐܽݔ݈ܽ݁ݎ‬ଓ‫݊݋‬ ଶି
൅ ܴ‫ܮ‬

This is assignment is consistent with: (i) the observed quenching of the RL

on exposure to a H plasma forming VZn-H complexes, (ii) the low formation energy of
VZn point defects under oxygen rich annealing conditions, (iii) the absence of nitrogen
in our ZnO nanorod samples, (iv) the large lattice dilation induced by creation of neutral
௢
ܸ௓௡ centers and (v) the temperature dependence of the RL peak position and FWHM
with the configurational co-ordinate model.
University of Technology Sydney Red Luminescence / 99

Figure 6.9: XANES spectra around the N kĮ energy from the “as-grown” and O2
annealed samples, as well as from ZnO nano-particles implanted with nitrogen, are
shown in Figure 6.9. The results show that no nitrogen is present in both the “as-grown”
and O2 annealed samples.

6.5 Conclusions
A strong broad red luminescence (RL) centered at 1.69 eV
(FWHM = 0.57 eV) at 15K and 1.78 eV (FWHM = 0.69 eV) at 300 K is formed in ZnO
nanorods following thermal annealing at 650°C for 30 minutes in an O2 gas
environment. The optical properties of this peak were comprehensively studied using a
range of characterization techniques, in particular photoluminescence and
cathodoluminescence spectroscopy. Using these results, the RL has been assigned to
highly lattice coupled VZn-related acceptor-like centers. No correlation was found
between the RL and nitrogen impurities.
Suranan Anantachaisilp Conclusions / 100

CHAPTER VII
CONCLUSIONS


The main aim of this thesis was to investigate the effect of deep level surface
point defects on optical and electrical properties, gas sensing properties and the surface
electronic structure of ZnO nanorods. Additionally, the emission properties of red
luminescence at 1.78 eV were systematically characterized as a function of temperature
and excitation density to gain a deeper insight into its chemical origin. These key finding
of this research are summarized below.
ZnO nanorods were successfully grown by low temperature (90 °C)
hydrothermal method. The various types of surface point defects were also
systematically introduce into ZnO nanorods by thermal treatments in oxidative and
reductive environments. The low temperature growth technique employed in the
preparation of ZnO nanorods in sensor devices shows great promise as an alternative
method to the existing high temperature routes due to its simplicity, application to
temperature sensitive substrates, energy efficiency, and reproducibility in obtaining
nanostructured ZnO for room temperature NH3 gas sensor applications. CL
spectroscopy confirmed that post-growth-controlled atmosphere heat treatment can be
used to mediate the Fermi level position of the ZnO nanorods through the creation of
acceptor or donor native near surface defects, which governs the sign and magnitude of
the electrical response to NH3 gas. Specifically a room temperature NH3 gas sensitivity
of S = 22.6 was achieved by heating in O2 atmosphere at 650 °C that was 5 times greater
than S for as-grown ZnO nanorods. The reported findings on the inter-relationship
between surface point defects, electrical properties, and gas sensing properties of
nanostructured ZnO will facilitate the development of future, low cost, high-sensitivity,
room temperature gas sensor devices tailored for specific applications.
Optical and electrical characterization studies of ZnO nanorods before and
after annealing in O2 gas and Zn vapor have confirmed that the surface electronic
structure can be bespoke for specific sensor applications. These correlative spectroscopy
measurements reveal clear evidence of direct interrelationship between the electronic
University of Technology Sydney Conclusions / 101


properties of the ZnO nanorod surface and its surface defect structure. In particular, the
role of intrinsic point defects on the physisorption and chemisorption of water molecules
and hydroxyl groups which, through charge transfer mechanisms, govern both the
surface band bending effects and the surface conductivity. Additionally O1s XPS
spectra established that the number of adsorbed water and hydroxyl species on the ZnO
nanorod is directly proportional to the concentration of surface oxygen vacancies. An
important result of these studies is compelling experimental evidence for the existence
of stable singly ionized oxygen vacancies (ܸைା ) at the ZnO nanorod surface.
Furthermore, it was shown that (i) a broad green CL peak centered at 2.43 eV, (ii) an
intense ESR line at g # 2.01 and (iii) a pronounced shoulder in the valence band X-ray
photoemission spectra ~ 3.0 eV below the Fermi level, can all be attributed to centers
involving surface ܸைା centers.
A strong broad red luminescence (RL) centered at 1.69 eV
(FWHM = 0.57 eV) at 15K and 1.78 eV (FWHM = 0.69 eV) at 300 K is formed in ZnO
nanorods following thermal annealing at 650°C for 30 minutes in an O2 gas
environment. The optical properties of this peak were comprehensively studied using a
range of characterization techniques, in particular photoluminescence and
cathodoluminescence spectroscopy. Using these results, the RL has been assigned to
highly lattice coupled VZn-related acceptor-like centers. No correlation was found
between the RL and nitrogen impurities.
Suranan Anantachaisilp References / 102

REFERENCES

1. Aslani, A.; Arefi, M. R.; Babapoor, A.; Amiri, A.; Beyki-Shuraki, K., Solvothermal
Synthesis, Characterization and Optical Properties of ZnO, ZnO–MgO and ZnO–
NiO, Mixed Oxide Nanoparticles. Applied Surface Science 2011, 257, 4885-4889.
2. Ozgur, U.; Alivov, Y. I.; Liu, C.; Teke, A.; Reshchikov, M. A.; Dogan, S.; Avrutin,
V.; Cho, S. J.; Morkoc, H., A Comprehensive Review of ZnO Materials and
Devices. Journal of Applied Physics 2005, 98, 041301-103.
3. Huang, M. H.; Mao, S.; Feick, H.; Yan, H.; Wu, Y.; Kind, H.; Weber, E.; Russo,
R.; Yang, P., Room-Temperature Ultraviolet Nanowire Nanolasers. Science 2001,
292, 1897-1899.
4. Li, H.; Huang, Y.; Zhang, Q.; Liu, J.; Zhang, Y., Influence of Electromechanical
Coupling and Electron Irradiation on the Conductivity of Individual ZnO Nanowire.
Solid State Sciences 2011, 13, 658-661.
5. Gao, P.; Wang, Z. Z.; Liu, K. H.; Xu, Z.; Wang, W. L.; Bai, X. D.; Wang, E. G.,
Photoconducting Response on Bending of Individual ZnO Nanowires. Journal of
Materials Chemistry 2009, 19.
6. Anderson, J.; Chris, G. V. d. W., Fundamentals of Zinc Oxide as a Semiconductor.
Reports on Progress in Physics 2009, 72, 126501.
7. Wei, A.; Pan, L.; Huang, W., Recent Progress in the ZnO Nanostructure-Based
Sensors. Materials Science and Engineering: B 2011, 176, 1409-1421.
8. Han, J.; Mantas, P. Q.; Senos, A. M. R., Defect Chemistry and Electrical
Characteristics of Undoped and Mn-Doped ZnO. Journal of the European Ceramic
Society 2002, 22, 49-59.
9. Charnhattakorn, B.; Charinpanitkul, T.; Sirisuk, A.; Pavarajarn, V., Controlled
Synthesis of Defects-Containing ZnO by the French Process Modified with Pulsed
Injection and Its Luminescence Properties. Ceramics International 2011, 37, 2021-
2024.
10. Lim, S. H.; Shindo, D.; Kang, H. B.; Nakamura, K., Study of Defects and Interfaces
ഥ 2 0) Al2O3 Grown by Electron Cyclotron
in Epitaxial ZnO Films on (1 ͳ
University of Technology Sydney References / 103

Resonance-Assisted Molecular Beam Epitaxy. Journal of Crystal Growth 2001,

225, 202-207.
11. Ament, W.; Huizenga, J.; Kort, E.; Van Der Mark, T.; Grevink, R.; Verkerke, G.,
Respiratory Ammonia Output and Blood Ammonia Concentration During
Incremental Exercise. International Journal of Sports Medicine 1999, 20, 71-77.
12. Narasimhan, L.; Goodman, W.; Patel, C. K. N., Correlation of Breath Ammonia
with Blood Urea Nitrogen and Creatinine During Hemodialysis. Proceedings of the
National Academy of Sciences 2001, 98, 4617-4621.
13. Kearney, D. J.; Hubbard, T.; Putnam, D., Breath Ammonia Measurement in
Helicobacter Pylori Infection. Digestive Diseases and Sciences 2002, 47, 2523-
2530.
14. Hassan, J.; Mahdi, M.; Chin, C.; Abu-Hassan, H.; Hassan, Z., Room Temperature
Hydrogen Gas Sensor Based on ZnO Nanorod Arrays Grown on a SiO2/Si Substrate
Via a Microwave-Assisted Chemical Solution Method. Journal of Alloys and
Compounds 2013, 546, 107-111.
15. Chang, C. M.; Hon, M. H.; Leu, I. C., Preparation of ZnO Nanorod Arrays with
Tailored Defect-Related Characterisitcs and Their Effect on the Ethanol Gas
Sensing Performance. Sensors and Actuators B: Chemical 2010, 151, 15-20.
16. Law, J.; Thong, J., Improving the NH3 Gas Sensitivity of ZnO Nanowire Sensors
by Reducing the Carrier Concentration. Nanotechnology 2008, 19, 205502.
17. Spencer, M. J., Gas Sensing Applications of 1D-Nanostructured Zinc Oxide:
Insights from Density Functional Theory Calculations. Progress in Materials
Science 2012, 57, 437-486.
18. Xue, X.; Nie, Y.; He, B.; Xing, L.; Zhang, Y.; Wang, Z. L., Surface Free-Carrier
Screening Effect on the Output of a ZnO Nanowire Nanogenerator and Its Potential
as a Self-Powered Active Gas Sensor. Nanotechnology 2013, 24, 225501.
19. Timmer, B.; Olthuis, W.; Van Den Berg, A., Ammonia Sensors and Their
Applications—a Review. Sensors and Actuators B: Chemical 2005, 107, 666-677.
20. Erisman, J. W.; Otjes, R.; Hensen, A.; Jongejan, P.; van den Bulk, P.; Khlystov, A.;
Möls, H.; Slanina, S., Instrument Development and Application in Studies and
Monitoring of Ambient Ammonia. Atmospheric Environment 2001, 35, 1913-1922.
Suranan Anantachaisilp References / 104

21. Mount, G. H.; Rumburg, B.; Havig, J.; Lamb, B.; Westberg, H.; Yonge, D.;
Johnson, K.; Kincaid, R., Measurement of Atmospheric Ammonia at a Dairy Using
Differential Optical Absorption Spectroscopy in the Mid-Ultraviolet. Atmospheric
Environment 2002, 36, 1799-1810.
22. Durbin, T. D.; Wilson, R. D.; Norbeck, J. M.; Miller, J. W.; Huai, T.; Rhee, S. H.,
Estimates of the Emission Rates of Ammonia from Light-Duty Vehicles Using
Standard Chassis Dynamometer Test Cycles. Atmospheric Environment 2002, 36,
1475-1482.
23. Budavari, S.; O’Neil, M.; Smith, A., The Merck Index: An Encyclopedia of
Chemicals, Drugs and Biologicals, Rahway, New Jersey, Merck and Co. Inc. 1989.
24. Moos, R.; Müller, R.; Plog, C.; Knezevic, A.; Leye, H.; Irion, E.; Braun, T.;
Marquardt, K.-J.; Binder, K., Selective Ammonia Exhaust Gas Sensor for
Automotive Applications. Sensors and Actuators B: Chemical 2002, 83, 181-189.
25. de la Hoz, R. E.; Schlueter, D. P.; Rom, W. N., Chronic Lung Disease Secondary
to Ammonia Inhalation Injury: A Report on Three Cases. American Journal of
Industrial Medicine 1996, 29, 209-214.
26. Lide, D. R., Crc Handbook of Chemistry and Physics; CRC press, 2004.
27. Schmidt-Mende, L.; MacManus-Driscoll, J. L., ZnO–Nanostructures, Defects, and
Devices. Materials today 2007, 10, 40-48.
28. Wöll, C., The Chemistry and Physics of Zinc Oxide Surfaces. Progress in Surface
Science 2007, 82, 55-120.
29. Özgür, Ü.; Alivov, Y. I.; Liu, C.; Teke, A.; Reshchikov, M.; Do÷an, S.; Avrutin,
V.; Cho, S.-J.; Morkoc, H., A Comprehensive Review of ZnO Materials and
Devices. Journal of applied physics 2005, 98, 041301.
30. Laskowski, R.; Christensen, N. E., Ab Initio Calculation of Excitons in ZnO.
Physical Review B 2006, 73, 045201.
31. Janotti, A.; Van de Walle, C. G., Fundamentals of Zinc Oxide as a Semiconductor.
Reports on Progress in Physics 2009, 72, 126501.
32. Meyer, B.; Alves, H.; Hofmann, D.; Kriegseis, W.; Forster, D.; Bertram, F.;
Christen, J.; Hoffmann, A.; Straßburg, M.; Dworzak, M., Bound Exciton and
Donor–Acceptor Pair Recombinations in ZnO. Physica Status Solidi (b) 2004, 241,
231-260.
University of Technology Sydney References / 105

33. Cobet, M.; Cobet, C.; Wagner, M. R.; Esser, N.; Thomsen, C.; Hoffmann, A.,
Polariton Effects in the Dielectric Function of ZnO Excitons Obtained by
Ellipsometry. Applied Physics Letters 2010, 96, 031904.
34. Varshni, Y., Temperature Dependence of the Energy Gap in Semiconductors.
Physica 1967, 34, 149-154.
35. Jagadish, C.; Pearton, S. J., Zinc Oxide Bulk, Thin Films and Nanostructures:
Processing, Properties, and Applications; Elsevier, 2011.
36. Shluger, A.; Foster, A.; Gavartin, J.; Sushko, P.; Greer, J.; Korkin, A., Models of
Defects in Wide-Gap Oxides: Perspective and Challenges. Nano and Giga
Challenges in Microelectronics 2003, 151-222.
37. Morris Jr, J., A Survey of Materials Science I. Structure University of California,
Berkeley 2007.
38. Jacobs, K., Fa Kröger. The Chemistry of Imperfect Crystals. Volume 1:
Preparation, Purification, Crystal Growth and Phase Theory. NorthǦHolland
Publishing CompanyǦAmsterdam/London 1973 American Elsevier Publishing
Company, Inc.ǦNew York 313 Seiten, Zahlreiche Abbildungen Und Tabellen,
Kunstleder Preis Dfl. 70.00. Kristall und Technik 1974, 9, K67-K68.
39. Kohan, A.; Ceder, G.; Morgan, D.; Van de Walle, C. G., First-Principles Study of
Native Point Defects in ZnO. Physical Review B 2000, 61, 15019.
40. Janotti, A.; Van de Walle, C. G., Native Point Defects in ZnO. Physical Review B
2007, 76, 165202.
41. Evans, S.; Giles, N.; Halliburton, L.; Kappers, L., Further Characterization of
Oxygen Vacancies and Zinc Vacancies in Electron-Irradiated ZnO. Journal of
Applied Physics 2008, 103, 043710-043710-7.
42. Reynolds, D.; Look, D.; Jogai, B.; Van Nostrand, J.; Jones, R.; Jenny, J., Source of
the Yellow Luminescence Band in GaN Grown by Gas-Source Molecular Beam
Epitaxy and the Green Luminescence Band in Single Crystal ZnO. Solid State
Communications 1998, 106, 701-704.
43. Reynolds, D. C.; Look, D. C.; Jogai, B.; Morkoc, H., Similarities in the Bandedge
and Deep-Centre Photoluminescence Mechanisms of ZnO and GaN. Solid State
Communications 1997, 101, 643-646.
Suranan Anantachaisilp References / 106

44. Janotti, A.; Van de Walle, C. G., Oxygen Vacancies in ZnO. Applied Physics Letters
2005, 87, 122102.
45. Erhart, P.; Klein, A.; Albe, K., First-Principles Study of the Structure and Stability
of Oxygen Defects in Zinc Oxide. Physical Review B 2005, 72, 085213.
46. Paudel, T. R.; Lambrecht, W. R., First-Principles Calculation of the O Vacancy in
ZnO: A Self-Consistent Gap-Corrected Approach. Physical Review B 2008, 77,
205202.
47. Erhart, P.; Albe, K.; Klein, A., First-Principles Study of Intrinsic Point Defects in
ZnO: Role of Band Structure, Volume Relaxation, and Finite-Size Effects. Physical
Review B 2006, 73, 205203.
48. Oba, F.; Togo, A.; Tanaka, I.; Paier, J.; Kresse, G., Defect Energetics in ZnO: A
Hybrid Hartree-Fock Density Functional Study. Physical Review B 2008, 77,
245202.
49. Janotti, A.; Van de Walle, C. G., Hydrogen Multicentre Bonds. Nature materials
2006, 6, 44-47.
50. Vanheusden, K.; Warren, W.; Seager, C.; Tallant, D.; Voigt, J.; Gnade, B.,
Mechanisms Behind Green Photoluminescence in ZnO Phosphor Powders. Journal
of Applied Physics 1996, 79, 7983-7990.
51. Vlasenko, L.; Watkins, G., Optical Detection of Electron Paramagnetic Resonance
in Room-Temperature Electron-Irradiated ZnO. Physical Review B 2005, 71,
125210.
52. Hofmann, D.; Pfisterer, D.; Sann, J.; Meyer, B.; Tena-Zaera, R.; Munoz-Sanjose,
V.; Frank, T.; Pensl, G., Properties of the Oxygen Vacancy in ZnO. Applied Physics
A 2007, 88, 147-151.
53. Sann, J.; Stehr, J.; Hofstaetter, A.; Hofmann, D.; Neumann, A.; Lerch, M.;
Haboeck, U.; Hoffmann, A.; Thomsen, C., Zn Interstitial Related Donors in
Ammonia-Treated ZnO Powders. Physical Review B 2007, 76, 195203.
54. Dingle, R., Luminescent Transitions Associated with Divalent Copper Impurities
and the Green Emission from Semiconducting Zinc Oxide. Physical Review Letters
1969, 23, 579.
55. Schirmer, O.; Zwingel, D., The Yellow Luminescence of Zinc Oxide. Solid State
Communications 1970, 8, 1559-1563.
University of Technology Sydney References / 107

56. Pantelides, S. T., Deep Centers in Semiconductors; CRC Press, 1992.

57. Glicksman, M., Diffusion in Solids: Field Theory, Solid-State Principles, and
Applications, 2000. John Wiley & Sons, Inc., New York.
58. Kador, L., Cathodoluminescence. Theory and Applications. ; Weinheim/VCH
Publishers, New York/Kodansha, Tokio, 1990.
59. Yacobi, B., Semiconductor Materials: An Introduction to Basic Principles;
Springer, 2003.
60. Reynolds, D.; Litton, C.; Collins, T., Zeeman Effects in the Edge Emission and
Absorption of ZnO. Physical Review 1965, 140, A1726-A1734.
61. Zwingel, D., Trapping and Recombination Processes in the Thermoluminescence
of Li-Doped ZnO Single Crystals. Journal of Luminescence 1972, 5, 385-405.
62. Boemare, C.; Monteiro, T.; Soares, M.; Guilherme, J.; Alves, E.,
Photoluminescence Studies in ZnO Samples. Physica B: Condensed Matter 2001,
308, 985-988.
63. Teke, A.; Özgür, Ü.; Do÷an, S.; Gu, X.; Morkoç, H.; Nemeth, B.; Nause, J.; Everitt,
H., Excitonic Fine Structure and Recombination Dynamics in Single-Crystalline
ZnO. Physical Review B 2004, 70, 195207.
64. Reynolds, D.; Look, D. C.; Jogai, B.; Litton, C.; Collins, T.; Harsch, W.; Cantwell,
G., Neutral-Donor–Bound-Exciton Complexes in ZnO Crystals. Physical Review B
1998, 57, 12151.
65. Alves, H.; Pfisterer, D.; Zeuner, A.; Riemann, T.; Christen, J.; Hofmann, D.; Meyer,
B., Optical Investigations on Excitons Bound to Impurities and Dislocations in
ZnO. Optical Materials 2003, 23, 33-37.
66. Thonke, K.; Gruber, T.; Teofilov, N.; Schönfelder, R.; Waag, A.; Sauer, R., Donor–
Acceptor Pair Transitions in ZnO Substrate Material. Physica B: Condensed Matter
2001, 308, 945-948.
67. Dietz, R.; Kamimura, H.; Sturge, M.; Yariv, A., Electronic Structure of Copper
Impurities in ZnO. Physical Review 1963, 132, 1559.
68. Yang, X.; Du, G.; Wang, X.; Wang, J.; Liu, B.; Zhang, Y.; Liu, D.; Liu, D.; Ong,
H.; Yang, S., Effect of Post-Thermal Annealing on Properties of ZnO Thin Film
Grown on C-Al2O3 by Metal-Organic Chemical Vapor Deposition. Journal of
Crystal Growth 2003, 252, 275-278.
Suranan Anantachaisilp References / 108

69. Korsunska, N.; Borkovska, L.; Bulakh, B.; Khomenkova, L. Y.; Kushnirenko, V.;
Markevich, I., The Influence of Defect Drift in External Electric Field on Green
Luminescence of ZnO Single Crystals. Journal of Luminescence 2003, 102, 733-
736.
70. Lin, B.; Fu, Z.; Jia, Y., Green Luminescent Center in Undoped Zinc Oxide Films
Deposited on Silicon Substrates. Applied Physics Letters 2001, 79, 943-945.
71. Vanheusden, K.; Seager, C.; Warren, W. t.; Tallant, D.; Voigt, J., Correlation
between Photoluminescence and Oxygen Vacancies in ZnO Phosphors. Applied
Physics Letters 1996, 68, 403-405.
72. Vanheusden, K.; Seager, C.; Warren, W.; Tallant, D.; Caruso, J.; Hampden-Smith,
M.; Kodas, T., Green Photoluminescence Efficiency and Free-Carrier Density in
ZnO Phosphor Powders Prepared by Spray Pyrolysis. Journal of luminescence
1997, 75, 11-16.
73. Leiter, F.; Alves, H.; Hofstaetter, A.; Hofmann, D.; Meyer, B., The Oxygen
Vacancy as the Origin of a Green Emission in Undoped ZnO. Physica Status Solidi
(b) 2001, 226, R4-R5.
74. Leiter, F.; Alves, H.; Pfisterer, D.; Romanov, N.; Hofmann, D.; Meyer, B., Oxygen
Vacancies in ZnO. Physica B: Condensed Matter 2003, 340, 201-204.
75. Studenikin, S.; Golego, N.; Cocivera, M., Fabrication of Green and Orange
Photoluminescent, Undoped ZnO Films Using Spray Pyrolysis. Journal of Applied
Physics 1998, 84, 2287-2294.
76. Kwok, W.; Djurišiü, A.; Leung, Y.; Chan, W.; Phillips, D., Time-Resolved
Photoluminescence Study of the Stimulated Emission in ZnO Nanoneedles. Applied
Physics Letters 2005, 87, 093108.
77. Liu, M.; Kitai, A.; Mascher, P., Point Defects and Luminescence Centres in Zinc
Oxide and Zinc Oxide Doped with Manganese. Journal of Luminescence 1992, 54,
35-42.
78. Greene, L. E.; Law, M.; Goldberger, J.; Kim, F.; Johnson, J. C.; Zhang, Y.;
Saykally, R. J.; Yang, P., Low-Temperature Wafer-Scale Production of ZnO
Nanowire Arrays. Angew Chem Int Ed Engl 2003, 42, 3031-3034.
University of Technology Sydney References / 109

79. Djurišiü, A.; Leung, Y.; Tam, K.; Hsu, Y.; Ding, L.; Ge, W.; Zhong, Y.; Wong, K.;
Chan, W.; Tam, H., Defect Emissions in ZnO Nanostructures. Nanotechnology
2007, 18, 095702.
80. Liu, X.; Wu, X.; Cao, H.; Chang, R., Growth Mechanism and Properties of ZnO
Nanorods Synthesized by Plasma-Enhanced Chemical Vapor Deposition. Journal
of Applied Physics 2004, 95, 3141-3147.
81. Cox, R.; Block, D.; Hervé, A.; Picard, R.; Santier, C.; Helbig, R., Exchange
Broadened, Optically Detected ESR Spectra for Luminescent Donor-Acceptor
Pairs in Li Doped ZnO. Solid State Communications 1978, 25, 77-80.
82. Alvi, N.; Ul Hasan, K.; Nur, O.; Willander, M., The Origin of the Red Emission in
N-ZnO Nanotubes/P-GaN White Light Emitting Diodes. Nanoscale research
letters 2011, 6, 1-7.
83. Li, D.; Leung, Y.; Djurišiü, A.; Liu, Z.; Xie, M.; Shi, S.; Xu, S.; Chan, W., Different
Origins of Visible Luminescence in ZnO Nanostructures Fabricated by the
Chemical and Evaporation Methods. Applied Physics Letters 2004, 85, 1601-1603.
84. Tarun, M.; Iqbal, M. Z.; McCluskey, M., Nitrogen Is a Deep Acceptor in ZnO. AIP
Advances, 2011, 1, 022105.
85. Ehret, W.; Greenstone, A., Red Zinc Oxide. Journal of the American Chemical
Society 1943, 65, 872-877.
86. Halliburton, L.; Giles, N.; Garces, N.; Luo, M.; Xu, C.; Bai, L.; Boatner, L. A.,
Production of Native Donors in ZnO by Annealing at High Temperature in Zn
Vapor. Applied Physics Letters 2005, 87, 172108.
87. Weber, M.; Parmar, N.; Jones, K.; Lynn, K., Oxygen Deficiency and Hydrogen
Turn ZnO Red. Journal of Electronic Materials 2010, 39, 573-576.
88. Chen, Y.; Xu, S.; Zheng, C.; Ning, J.; Ling, F.; Anwand, W.; Brauer, G.; Skorupa,
W., Nature of Red Luminescence Band in Research-Grade ZnO Single Crystals: A
“Self-Activated” Configurational Transition. Applied Physics Letters 2014, 105,
041912.
89. Hamby, D.; Lucca, D.; Klopfstein, M.; Cantwell, G., Temperature Dependent
Exciton Photoluminescence of Bulk ZnO. Journal of Applied Physics 2003, 93,
3214-3217.
Suranan Anantachaisilp References / 110

90. Hauschild, R.; Priller, H.; Decker, M.; Brückner, J.; Kalt, H.; Klingshirn, C.,
Temperature Dependent Band Gap and Homogeneous Line Broadening of the
Exciton Emission in ZnO. Physica Status Solidi (c) 2006, 3, 976-979.
91. Li, J.; Ong, H., Temperature Dependence of Surface Plasmon Mediated Emission
from Metal-Capped ZnO Films. Applied Physics Letters 2008, 92, 121107.
92. Morkoç, H.; Özgür, Ü., Zinc Oxide: Fundamentals, Materials and Device
Technology; John Wiley & Sons, 2008.
93. Adachi, S., Properties of Semiconductor Alloys: Group-Iv, Iii-V and Ii-Vi
Semiconductors; John Wiley & Sons, 2009; Vol. 28.
94. Van de Walle, C. G., Hydrogen as a Cause of Doping in Zinc Oxide. Physical
Review Letters 2000, 85, 1012.
95. Strzhemechny, Y. M.; Mosbacker, H. L.; Look, D. C.; Reynolds, D. C.; Litton, C.
W.; Garces, N. Y.; Giles, N. C.; Halliburton, L. E.; Niki, S.; Brillson, L. J., Remote
Hydrogen Plasma Doping of Single Crystal ZnO. Applied physics letters 2004, 84,
2545-2547.
96. Nickel, N.; Fleischer, K., Hydrogen Local Vibrational Modes in Zinc Oxide.
Physical review letters 2003, 90, 197402.
97. Ip, K.; Overberg, M.; Heo, Y.; Norton, D.; Pearton, S.; Stutz, C.; Luo, B.; Ren, F.;
Look, D. C.; Zavada, J., Hydrogen Incorporation and Diffusivity in Plasma-
Exposed Bulk ZnO. Applied physics letters 2003, 82, 385-387.
98. Oba, F.; Nishitani, S. R.; Isotani, S.; Adachi, H.; Tanaka, I., Energetics of Native
Defects in ZnO. Journal of Applied Physics 2001, 90, 824-828.
99. Zhang, S.; Wei, S.-H.; Zunger, A., Intrinsic N-Type Versus P-Type Doping
Asymmetry and the Defect Physics of ZnO. Physical Review B 2001, 63, 075205.
100.Look, D. C.; Farlow, G. C.; Reunchan, P.; Limpijumnong, S.; Zhang, S.; Nordlund,
K., Evidence for Native-Defect Donors in N-Type ZnO. Physical review letters
2005, 95, 225502.
101.Hoppe, U.; Stachel, D.; Beyer, D., The Oxygen Coordination of Metal Ions in
Phosphate and Silicate Glasses Studied by a Combination of X-Ray and Neutron
Diffraction. Physica Scripta 1995, 1995, 122.
102.Iliadis, A.; Vispute, R.; Venkatesan, T.; Jones, K., Ohmic Metallization Technology
for Wide Band-Gap Semiconductors. Thin Solid Films 2002, 420, 478-486.
University of Technology Sydney References / 111

103.Sheng, H.; Emanetoglu, N.; Muthukumar, S.; Feng, S.; Lu, Y., Nonalloyed Al
Ohmic Contacts to MgxZn1í Xo. Journal of Electronic Materials 2002, 31, 811-
814.
104.Brillson, L. J.; Lu, Y., ZnO Schottky Barriers and Ohmic Contacts. Journal of
Applied Physics 2011, 109, 121301.
105.Zhang, D.; Li, C.; Liu, X.; Han, S.; Tang, T.; Zhou, C., Doping Dependent NH3
Sensing of Indium Oxide Nanowires. Applied Physics Letters 2003, 83, 1845-1847.
106.Wan, Q.; Wang, T., Single-Crystalline Sb-Doped SnO2 Nanowires: Synthesis and
Gas Sensor Application. Chemical Communications 2005, 3841-3843.
107.Li, Q.; Liang, Y.; Wan, Q.; Wang, T., Oxygen Sensing Characteristics of Individual
ZnO Nanowire Transistors. Applied Physics Letters 2004, 85, 6389-6391.
108.Basu, S.; Dutta, A., Modified Heterojunction Based on Zinc Oxide Thin Film for
Hydrogen Gas-Sensor Application. Sensors and Actuators B: Chemical 1994, 22,
83-87.
109.Rout, C. S.; Raju, A.; Govindaraj, A.; Rao, C., Hydrogen Sensors Based on ZnO
Nanoparticles. Solid state communications 2006, 138, 136-138.
110.Wang, H. T.; Kang, B. S.; Ren, F.; Tien, L. C.; Sadik, P. W.; Norton, D. P.; Pearton,
S. J.; Lin, J., Hydrogen-Selective Sensing at Room Temperature with ZnO
Nanorods. Applied Physics Letters 2005, 86, -.
111.Kang, B.; Heo, Y.; Tien, L.; Norton, D.; Ren, F.; Gila, B.; Pearton, S., Hydrogen
and Ozone Gas Sensing Using Multiple ZnO Nanorods. Applied Physics A 2005,
80, 1029-1032.
112.Sarala Devi, G.; Bala Subrahmanyam, V.; Gadkari, S. C.; Gupta, S. K., Nh3 Gas
Sensing Properties of Nanocrystalline ZnO Based Thick Films. Analytica Chimica
Acta 2006, 568, 41-46.
113.Anantachaisilp, S.; Smith, S. M.; Ton-That, C.; Osotchan, T.; Moon, A. R.; Phillips,
M. R., Tailoring Deep Level Surface Defects in ZnO Nanorods for High Sensitivity
Ammonia Gas Sensing. The Journal of Physical Chemistry C 2014, 118, 27150-
27156.
114.Nanto, H.; Minami, T.; Takata, S., ZincǦOxide ThinǦFilm Ammonia Gas Sensors
with High Sensitivity and Excellent Selectivity. Journal of Applied Physics 1986,
60, 482-484.
Suranan Anantachaisilp References / 112

115.Kim, J.; Yong, K., Mechanism Study of ZnO Nanorod-Bundle Sensors for H2S Gas
Sensing. The Journal of Physical Chemistry C 2011, 115, 7218-7224.
116.Wang, C.; Chu, X.; Wu, M., Detection of H2s Down to ppb Levels at Room
Temperature Using Sensors Based on ZnO Nanorods. Sensors and Actuators B:
Chemical 2006, 113, 320-323.
117.Chang, J. F.; Kuo, H. H.; Leu, I. C.; Hon, M. H., The Effects of Thickness and
Operation Temperature on ZnO:Al Thin Film CO Gas Sensor. Sensors and
Actuators B: Chemical 2002, 84, 258-264.
118.Gong, H.; Hu, J. Q.; Wang, J. H.; Ong, C. H.; Zhu, F. R., Nano-Crystalline Cu-
Doped ZnO Thin Film Gas Sensor for CO. Sensors and Actuators B: Chemical
2006, 115, 247-251.
119.Gao, T.; Wang, T. H., Synthesis and Properties of Multipod-Shaped ZnO Nanorods
for Gas-Sensor Applications. Applied Physics A 2005, 80, 1451-1454.
120.Wang, X.-H.; Ding, Y.-F.; Zhang, J.; Zhu, Z.-Q.; You, S.-Z.; Chen, S.-Q.; Zhu, J.,
Humidity Sensitive Properties of ZnO Nanotetrapods Investigated by a Quartz
Crystal Microbalance. Sensors and Actuators B: Chemical 2006, 115, 421-427.
121.Zhou, X.; Zhang, J.; Jiang, T.; Wang, X.; Zhu, Z., Humidity Detection by
Nanostructured ZnO: A Wireless Quartz Crystal Microbalance Investigation.
Sensors and Actuators A: Physical 2007, 135, 209-214.
122.Sberveglieri, G.; Nelli, P.; Groppelli, S.; Quaranta, F.; Valentini, A.; Vasanelli, L.,
Oxygen Gas Sensing Characteristics at Ambient Pressure of Undoped and Lithium-
Doped ZnO-Sputtered Thin Films. Materials Science and Engineering: B 1990, 7,
63-68.
123.Martins, R.; Fortunato, E.; Nunes, P.; Ferreira, I.; Marques, A.; Bender, M.;
Katsarakis, N.; Cimalla, V.; Kiriakidis, G., Zinc Oxide as an Ozone Sensor. Journal
of Applied Physics 2004, 96, 1398-1408.
124.Baratto, C.; Sberveglieri, G.; Onischuk, A.; Caruso, B.; di Stasio, S., Low
Temperature Selective NO2 Sensors by Nanostructured Fibres of ZnO. Sensors and
Actuators B: Chemical 2004, 100, 261-265.
125.Cho, P.-S.; Kim, K.-W.; Lee, J.-H., NO2 Sensing Characteristics of ZnO Nanorods
Prepared by Hydrothermal Method. Journal of Electroceramics 2006, 17, 975-978.
University of Technology Sydney References / 113

126.Hung, N. L.; Ahn, E. S.; Jung, H. C.; Kim, H. J.; Kim, D. J., Synthesis and Gas
Sensing Properties of ZnO Nanostructures. Journal of the Korean Physical Society
2010, 57, 1784-1788.
127.Huang, X.-J.; Choi, Y.-K., Chemical Sensors Based on Nanostructured Materials.
Sensors and Actuators B: Chemical 2007, 122, 659-671.
128.Fang, X.; Hu, L.; Ye, C.; Zhang, L., One-Dimensional Inorganic Semiconductor
Nanostructures: A New Carrier for Nanosensors. Pure and Applied Chemistry
2010, 82, 2185-2198.
129.Kolmakov, A.; Zhang, Y.; Cheng, G.; Moskovits, M., Detection of CO and O2
Using Tin Oxide Nanowire Sensors. Advanced Materials 2003, 15, 997-1000.
130.Williams, D. E., Semiconducting Oxides as Gas-Sensitive Resistors. Sensors and
Actuators B: Chemical 1999, 57, 1-16.
131.Kolmakov, A.; Moskovits, M., Chemical Sensing and Catalysis by One-
Dimensional Metal-Oxide Nanostructures. Annu. Rev. Mater. Res. 2004, 34, 151-
180.
132.Huang, M. H.; Wu, Y.; Feick, H.; Tran, N.; Weber, E.; Yang, P., Catalytic Growth
of Zinc Oxide Nanowires by Vapor Transport. Advanced Materials 2001, 13, 113-
116.
133.Lu, C.; Qi, L.; Yang, J.; Tang, L.; Zhang, D.; Ma, J., Hydrothermal Growth of
Large-Scale Micropatterned Arrays of Ultralong ZnO Nanowires and Nanobelts on
Zinc Substrate. Chemical Communications 2006.
134.Wu, J. J.; Liu, S. C., Low-Temperature Growth of Well-Aligned ZnO Nanorods by
Chemical Vapor Deposition. Advanced Materials 2002, 14, 215-218.
135.Park, W. I.; Yi, G. C.; Kim, M.; Pennycook, S. J., ZnO Nanoneedles Grown
Vertically on Si Substrates by Non-Catalytic Vapor-Phase Epitaxy. Advanced
Materials 2002, 14, 1841-1843.
136.Tang, C. C.; Fan, S. S.; Chapelle, M. L. d. l.; Li, P., Silica-Assisted Catalytic
Growth of Oxide and Nitride Nanowires. Chemical Physics Letters 2001, 333, 12-
15.
137.Pan, Z. W.; Dai, Z. R.; Wang, Z. L., Nanobelts of Semiconducting Oxides. Science
2001, 291, 1947-1949.
Suranan Anantachaisilp References / 114

138.Wang, Z.; Li, H. L., Highly Ordered Zinc Oxide Nanotubules Synthesized within
the Anodic Aluminum Oxide Template. Applied Physics A 2002, 74, 201-203.
139.Miao, L.; Ieda, Y.; Tanemura, S.; Cao, Y.; Tanemura, M.; Hayashi, Y.; Toh, S.;
Kaneko, K., Synthesis, Microstructure and Photoluminescence of Well-Aligned
ZnO Nanorods on Si Substrate. Science and Technology of Advanced Materials
2007, 8, 443-447.
140.Dalal, S. H.; Baptista, D. L.; Teo, K. B. K.; Lacerda, R. G.; Jefferson, D. A.; Milne,
W. I., Controllable Growth of Vertically Aligned Zinc Oxide Nanowires Using
Vapour Deposition. Nanotechnology 2006, 17, 4811.
141.Satoh, Y.; Ohshio, S.; Saitoh, H., Photoluminescence Spectroscopy of Highly
Oriented Y2O3: Tb Crystalline Whiskers. Science and Technology of Advanced
Materials 2005, 6, 215-218.
142.Yasuda, T.; Segawa, Y., Zinc Oxide Thin Films Synthesized by Metal Organic
Chemical Reactions. Physica Status Solidi (b) 2004, 241, 676-679.
143.Soci, C.; Zhang, A.; Xiang, B.; Dayeh, S. A.; Aplin, D.; Park, J.; Bao, X.; Lo, Y.-
H.; Wang, D., ZnO Nanowire UV Photodetectors with High Internal Gain. Nano
Letters 2007, 7, 1003-1009.
144.Lagashetty, A.; Havanoor, V.; Basavaraja, S.; Balaji, S.; Venkataraman, A.,
Microwave-Assisted Route for Synthesis of Nanosized Metal Oxides. Science and
Technology of Advanced Materials 2007, 8, 484-493.
145.Spanhel, L., Colloidal ZnO Nanostructures and Functional Coatings: A Survey. J
Sol-Gel Sci Technol 2006, 39, 7-24.
146.Ma, X.; Zhang, H.; Ji, Y.; Xu, J.; Yang, D., Sequential Occurrence of ZnO
Nanopaticles, Nanorods, and Nanotips During Hydrothermal Process in a Dilute
Aqueous Solution. Materials Letters 2005, 59, 3393-3397.
147.Kohls, M.; Bonanni, M.; Spanhel, L.; Su, D.; Giersig, M., Green Eriii
Luminescence in Fractal ZnO Nanolattices. Applied Physics Letters 2002, 81, 3858-
3860.
148.Xu, H.; Wang, H.; Zhang, Y.; Wang, S.; Zhu, M.; Yan, H., Asymmetric Twinning
Crystals of Zinc Oxide Formed in a Hydrothermal Process. Crystal Research and
Technology 2003, 38, 429-432.
University of Technology Sydney References / 115

149.Shingubara, S., Fabrication of Nanomaterials Using Porous Alumina Templates.

Journal of Nanoparticle Research 2003, 5, 17-30.
150.Krunks, M.; Mellikov, E., Zinc Oxide Thin Films by the Spray Pyrolysis Method.
Thin Solid Films 1995, 270, 33-36.
151.Ayouchi, R.; Martin, F.; Leinen, D.; Ramos-Barrado, J. R., Growth of Pure ZnO
Thin Films Prepared by Chemical Spray Pyrolysis on Silicon. Journal of Crystal
Growth 2003, 247, 497-504.
152.Lee, J.-H.; Leu, C.; Chung, Y.-W.; Hon, M.-H., Fabrication of Ordered ZnO
Hierarchical Structures Controlled Via Surface Charge in the Electrophoretic
Deposition Process. Nanotechnology 2006, 17, 4445.
153.Xu, S.; Wang, Z. L., One-Dimensional ZnO Nanostructures: Solution Growth and
Functional Properties. Nano Research 2011, 4, 1013-1098.
154.Tam, K. H., et al., Defects in ZnO Nanorods Prepared by a Hydrothermal Method.
The Journal of Physical Chemistry B 2006, 110, 20865-20871.
155.Amin, G.; Asif, M. H.; Zainelabdin, A.; Zaman, S.; Nur, O.; Willander, M.,
Influence of PH, Precursor Concentration, Growth Time, and Temperature on the
Morphology of ZnO Nanostructures Grown by the Hydrothermal Method. Journal
of Nanomaterials 2011, 2011.
156.Tao, Y.; Fu, M.; Zhao, A.; He, D.; Wang, Y., The Effect of Seed Layer on
Morphology of ZnO Nanorod Arrays Grown by Hydrothermal Method. Journal of
Alloys and Compounds 2010, 489, 99-102.
157.Usman Ali, S. M.; Nur, O.; Willander, M.; Danielsson, B., A Fast and Sensitive
Potentiometric Glucose Microsensor Based on Glucose Oxidase Coated ZnO
Nanowires Grown on a Thin Silver Wire. Sensors and Actuators B: Chemical 2010,
145, 869-874.
158.Israr, M. Q.; Sadaf, J. R.; Asif, M.; Nur, O.; Willander, M.; Danielsson, B.,
Potentiometric Cholesterol Biosensor Based on ZnO Nanorods Chemically Grown
on Ag Wire. Thin solid films 2010, 519, 1106-1109.
159.Hafner, B., Energy Dispersive Spectroscopy on the SEM: A Primer.
Characterization Facility, University of Minnesota 2006, 1-26.
160.Australian Synchrotron, Synchrotron Science. http://www.synchrotron.org.au/
synchrotron-science (accessed October 23, 2014).
Suranan Anantachaisilp References / 116

161.Ozgur, U.; Hofstetter, D.; Morkoc, H., ZnO Devices and Applications: A Review
of Current Status and Future Prospects. Proceedings of the IEEE 2010, 98, 1255-
1268.
162.Zhang, D.; Chava, S.; Berven, C.; Lee, S. K.; Devitt, R.; Katkanant, V.,
Experimental Study of Electrical Properties of ZnO Nanowire Random Networks
for Gas Sensing and Electronic Devices. Applied Physics A 2010, 100, 145-150.
163.Ahn, M. W.; Park, K. S.; Heo, J. H.; Park, J. G.; Kim, D. W.; Choi, K. J.; Lee, J.
H.; Hong, S. H., Gas Sensing Properties of Defect-Controlled ZnO-Nanowire Gas
Sensor. Applied Physics Letters 2008, 93, 263103-263103-3.
164.Hassan, J. J.; Mahdi, M. A.; Chin, C. W.; Abu-Hassan, H.; Hassan, Z., Room-
Temperature Hydrogen Gas Sensor with ZnO Nanorod Arrays Grown on a Quartz
Substrate. Physica E: Low-dimensional Systems and Nanostructures 2012, 46, 254-
258.
165.Hassan, J. J.; Mahdi, M. A.; Chin, C. W.; Abu-Hassan, H.; Hassan, Z., Room
Temperature Hydrogen Gas Sensor Based on ZnO Nanorod Arrays Grown on a
SiO2/Si Substrate Via a Microwave-Assisted Chemical Solution Method. Journal
of Alloys and Compounds 2013, 546, 107-111.
166.Wang, J.; Sun, X.; Yang, Y.; Wu, C., N–P Transition Sensing Behaviors of ZnO
Nanotubes Exposed to NO2 Gas. Nanotechnology 2009, 20, 465501.
167.Wang, D.; Chu, X.; Gong, M., Hydrothermal Growth of ZnO Nanoscrewdrivers
and Their Gas Sensing Properties. Nanotechnology 2007, 18, 185601.
168.Xiangfeng, C.; Dongli, J.; Djurišic, A. B.; Leung, Y. H., Gas-Sensing Properties of
Thick Film Based on ZnO Nano-Tetrapods. Chemical Physics Letters 2005, 401,
426-429.
169.Han, N.; Chai, L.; Wang, Q.; Tian, Y.; Deng, P.; Chen, Y., Evaluating the Doping
Effect of Fe, Ti and Sn on Gas Sensing Property of ZnO. Sensors and Actuators B:
Chemical 2010, 147, 525-530.
170.Li, L. M.; Du, Z. F.; Wang, T. H., Enhanced Sensing Properties of Defect-
Controlled ZnO Nanotetrapods Arising from Aluminum Doping. Sensors and
Actuators B: Chemical 2010, 147, 165-169.
University of Technology Sydney References / 117

171.Ueda, T.; Plashnitsa, V. V.; Nakatou, M.; Miura, N., Zirconia-Based Amperometric
Sensor Using ZnO Sensing-Electrode for Selective Detection of Propene.
Electrochemistry Communications 2007, 9, 197-200.
172.Cheng, X.; Zhao, H.; Huo, L.; Gao, S.; Zhao, J., ZnO Nanoparticulate Thin Film:
Preparation, Characterization and Gas-Sensing Property. Sensors and Actuators B:
Chemical 2004, 102, 248-252.
173.Dhingra, M.; Kumar, L.; Shrivastava, S.; Kumar, P.; Annapoorni, S., Impact of
Interfacial Interactions on Optical and Ammonia Sensing in Zinc Oxide/Polyaniline
Structures. Bull Mater Sci 2013, 36, 647-652.
174.Wei, A.; Wang, Z.; Pan, L.-H.; Li, W.-W.; Xiong, L.; Dong, X.-C.; Huang, W.,
Room-Temperature NH3 Gas Sensor Based on Hydrothermally Grown ZnO
Nanorods. Chinese Physics Letters 2011, 28.
175.Greene, L. E.; Law, M.; Tan, D. H.; Montano, M.; Goldberger, J.; Somorjai, G.;
Yang, P., General Route to Vertical ZnO Nanowire Arrays Using Textured ZnO
Seeds. Nano Letters 2005, 5, 1231-1236.
176.Xu, S.; Lao, C.; Weintraub, B.; Wang, Z. L., Density-Controlled Growth of Aligned
ZnO Nanowire Arrays by Seedless Chemical Approach on Smooth Surfaces.
Journal of Materials Research 2008, 23, 2072-2077.
177.Drouin, D.; Couture, A. R.; Joly, D.; Tastet, X.; Aimez, V.; Gauvin, R., Casino V2.
42—a Fast and EasyǦtoǦUse Modeling Tool for Scanning Electron Microscopy and
Microanalysis Users. Scanning 2007, 29, 92-101.
178.Tarun, M.; Iqbal, M. Z.; McCluskey, M., Nitrogen Is a Deep Acceptor in ZnO. AIP
Advances 2011, 1, 022105.
179.Ton-That, C.; Weston, L.; Phillips, M., Characteristics of Point Defects in the Green
Luminescence from Zn-and O-Rich ZnO. Physical Review B 2012, 86, 115205.
180.Wang, X.; Vlasenko, L.; Pearton, S.; Chen, W.; Buyanova, I. A., Oxygen and Zinc
Vacancies in as-Grown ZnO Single Crystals. Journal of Physics D: Applied Physics
2009, 42, 175411.
181.Morkoç, H.; Özgür, Ü., Zinc Oxide: Fundamentals, Materials and Device
Technology; Wiley, 2009.
182.Foley, M.; Ton-That, C.; Phillips, M. R., Cathodoluminescence Inhomogeneity in
ZnO Nanorods. Applied Physics Letters 2008, 93, 243104.
Suranan Anantachaisilp References / 118

183.Phillips, M. R.; Gelhausen, O.; Goldys, E., Cathodoluminescence Properties of Zinc

Oxide Nanoparticles. Physica Status Solidi (a) 2004, 201, 229-234.
184.Fabbri, F., et al., Zn Vacancy Induced Green Luminescence on Non-Polar Surfaces
in ZnO Nanostructures. Sci. Rep. 2014, 4.
185.Wang, J.; Sun, X.; Yang, Y.; Huang, H.; Lee, Y.; Tan, O.; Vayssieres, L.,
Hydrothermally Grown Oriented ZnO Nanorod Arrays for Gas Sensing
Applications. Nanotechnology 2006, 17, 4995.
186.Lin, Y.; Wang, D.; Zhao, Q.; Li, Z.; Ma, Y.; Yang, M., Influence of Adsorbed
Oxygen on the Surface Photovoltage and Photoluminescence of ZnO Nanorods.
Nanotechnology 2006, 17, 2110.
187.Bao, J.; Shalish, I.; Su, Z.; Gurwitz, R.; Capasso, F.; Wang, X.; Ren, Z.,
Photoinduced Oxygen Release and Persistent Photoconductivity in ZnO
Nanowires. Nanoscale Research Letters 2011, 6, 1-7.
188.Yan, Y.; Al-Jassim, M.; Wei, S.-H., Oxygen-Vacancy Mediated Adsorption and
Reactions of Molecular Oxygen on the ZnO (10ͳത0) Surface. Physical Review B
2005, 72, 161307.
189.Reyes, P. I.; Ku, C.-J.; Duan, Z.; Xu, Y.; Garfunkel, E.; Lu, Y., Reduction of
Persistent Photoconductivity in ZnO Thin Film Transistor-Based UV
Photodetector. Applied Physics Letters 2012, 101, 031118.
190.Li, C.; Zhang, D.; Lei, B.; Han, S.; Liu, X.; Zhou, C., Surface Treatment and Doping
Dependence of In2O3 Nanowires as Ammonia Sensors. The Journal of Physical
Chemistry B 2003, 107, 12451-12455.
191.Lavrov, E.; Herklotz, F.; Weber, J., Identification of Two Hydrogen Donors in ZnO.
Physical Review B 2009, 79, 165210.
192.Fan, Z.; Dutta, D.; Chien, C.-J.; Chen, H.-Y.; Brown, E. C.; Chang, P.-C.; Lu, J. G.,
Electrical and Photoconductive Properties of Vertical ZnO Nanowires in High
Density Arrays. Applied Physics Letters 2006, 89, 213110.
193.Fonoberov, V. A.; Balandin, A. A., ZnO Quantum Dots: Physical Properties and
Optoelectronic Applications. Journal of Nanoelectronics and Optoelectronics
2006, 1, 19-38.
194.Pan, H.; Feng, Y. P., Semiconductor Nanowires and Nanotubes: Effects of Size and
Surface-to-Volume Ratio. ACS Nano 2008, 2, 2410-2414.
University of Technology Sydney References / 119

195.Zhang, X.; Qin, J.; Xue, Y.; Yu, P.; Zhang, B.; Wang, L.; Liu, R., Effect of Aspect
Ratio and Surface Defects on the Photocatalytic Activity of ZnO Nanorods.
Scientific Reports 2014, 4.
196.Mönch, W., Semiconductor Surfaces and Interfaces, Springer Series in Surface
Sciences; Springer-Verlag Berlin Heidelberg, 2001; Vol. 26.
197.Zhang, Z.; Yates Jr, J. T., Band Bending in Semiconductors: Chemical and Physical
Consequences at Surfaces and Interfaces. Chemical Reviews 2012, 112, 5520-5551.
198.Allen, M.; Swartz, C.; Myers, T.; Veal, T. D.; McConville, C. F.; Durbin, S., Bulk
Transport Measurements in ZnO: The Effect of Surface Electron Layers. Physical
Review B 2010, 81, 075211.
199.Heinhold, R.; Williams, G.; Cooil, S.; Evans, D.; Allen, M., Influence of Polarity
and Hydroxyl Termination on the Band Bending at ZnO Surfaces. Physical Review
B 2013, 88, 235315.
200.Heinhold, R.; Cooil, S. P.; Evans, D. A.; Allen, M. W., Stability of the Surface
Electron Accumulation Layers on the Nonpolar (101ࡄ0) and (112ࡄ0) Faces of ZnO.
The Journal of Physical Chemistry C 2014, 118, 24575-24582.
201.Chen, M.-W.; Retamal, J. R. D.; Chen, C.-Y.; He, H., Photocarrier Relaxation
Behavior of a Single ZnO Nanowire UV Photodetector: Effect of Surface Band
Bending. Electron Device Letters, IEEE 2012, 33, 411-413.
202.Liao, Z.-M.; Liu, K.-J.; Zhang, J.-M.; Xu, J.; Yu, D.-P., Effect of Surface States on
Electron Transport in Individual ZnO Nanowires. Physics Letters A 2007, 367, 207-
210.
203.Li, Q.; Gao, T.; Wang, Y.; Wang, T., Adsorption and Desorption of Oxygen Probed
from ZnO Nanowire Films by Photocurrent Measurements. Applied Physics Letters
2005, 86, 123117.
204.Fan, Z.; Wang, D.; Chang, P.-C.; Tseng, W.-Y.; Lu, J. G., ZnO Nanowire Field-
Effect Transistor and Oxygen Sensing Property. Applied Physics Letters 2004, 85,
5923-5925.
205.Li, Y.; Della Valle, F.; Simonnet, M.; Yamada, I.; Delaunay, J.-J., Competitive
Surface Effects of Oxygen and Water on UV Photoresponse of ZnO Nanowires.
Applied Physics Letters 2009, 94, 023110.
Suranan Anantachaisilp References / 120

206.Chen, C.-Y.; Retamal, J. R. D.; Wu, I.-W.; Lien, D.-H.; Chen, M.-W.; Ding, Y.;
Chueh, Y.-L.; Wu, C.-I.; He, J.-H., Probing Surface Band Bending of Surface-
Engineered Metal Oxide Nanowires. ACS Nano 2012, 6, 9366-9372.
207.Bao, J.; Shalish, I.; Su, Z.; Gurwitz, R.; Capasso, F.; Wang, X.; Ren, Z.,
Photoinduced Oxygen Release and Persistent Photoconductivity in ZnO
Nanowires. Nanoscale Res. Lett 2011, 61, 404.
208.Xu, H.; Fan, W.; Rosa, A.; Zhang, R.; Frauenheim, T., Hydrogen and Oxygen
Adsorption on ZnO Nanowires: A First-Principles Study. Physical Review B 2009,
79, 073402.
209.Gurwitz, R.; Cohen, R.; Shalish, I., Interaction of Light with the ZnO Surface:
Photon Induced Oxygen “Breathing,” Oxygen Vacancies, Persistent
Photoconductivity, and Persistent Photovoltage. Journal of Applied Physics 2014,
115, 033701.
210.Moormann, H.; Kohl, D.; Heiland, G., Work Function and Band Bending on Clean
Cleaved Zinc Oxide Surfaces. Surface Science 1979, 80, 261-264.
211.An, W.; Wu, X.; Zeng, X., Adsorption of O2, H2, CO, NH3, and NO2 on ZnO
Nanotube: A Density Functional Theory Study. The Journal of Physical Chemistry
C 2008, 112, 5747-5755.
212.Spencer, M. J.; Wong, K. W.; Yarovsky, I., Surface Defects on ZnO Nanowires:
Implications for Design of Sensors. Journal of Physics: Condensed Matter 2012,
24, 305001.
213.Collins, R.; Thomas, D., Photoconduction and Surface Effects with Zinc Oxide
Crystals. Physical Review 1958, 112, 388.
214.Chen, D.; Wang, Z.; Ren, T.; Ding, H.; Yao, W.; Zong, R.; Zhu, Y., Influence of
Defects on the Photocatalytic Activity of ZnO. The Journal of Physical Chemistry
C 2014, 118, 15300-15307.
215.Willander, M.; Nur, O.; Sadaf, J. R.; Qadir, M. I.; Zaman, S.; Zainelabdin, A.; Bano,
N.; Hussain, I., Luminescence from Zinc Oxide Nanostructures and Polymers and
Their Hybrid Devices. Materials 2010, 3, 2643-2667.
216.Fabbri, F.; Villani, M.; Catellani, A.; Calzolari, A.; Cicero, G.; Calestani, D.;
Calestani, G.; Zappettini, A.; Dierre, B.; Sekiguchi, T., Zn Vacancy Induced Green
University of Technology Sydney References / 121

Luminescence on Non-Polar Surfaces in ZnO Nanostructures. Scientific Reports

2014, 4.
217.Giles, N.; Xu, C.; Callahan, M.; Wang, B.; Neal, J. S.; Boatner, L. A., Effects of
Phonon Coupling and Free Carriers on Band-Edge Emission at Room Temperature
in N-Type ZnO Crystals. Applied Physics Letters 2006, 89, 251906-251906-3.
218.Hu, H.; Ji, H.-F.; Sun, Y., The Effect of Oxygen Vacancies on Water Wettability of
a ZnO Surface. Phys. Chem. Chem. Phys. 2013, 15, 16557-16565.
219.Kiasari, N. M.; Soltanian, S.; Gholamkhass, B.; Servati, P., Room Temperature
Ultra-Sensitive Resistive Humidity Sensor Based on Single Zinc Oxide Nanowire.
Sensors and Actuators A: Physical 2012, 182, 101-105.
220.Chen, Q.; Lu, W.; Wu, Y.-k.; Ding, H.-y.; Wang, B.; Chen, L., Probe the Effects of
Surface Adsorbates on ZnO Nanowire Conductivity Using Dielectric Force
Microscopy. Chinese Journal of Chemical Physics 2014, 27, 582-586.
221.Murphy, D. M., Metal Oxide Catalysis (Edited by S. D. Jackson and J. S. J.
Hargreaves); Wiley VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, 2008.
222.Djurišiü, A. B.; Leung, Y. H.; Choy, W. C.; Cheah, K. W.; Chan, W. K., Visible
Photoluminescence in ZnO Tetrapod and Multipod Structures. Applied Physics
Letters 2004, 84, 2635-2637.
223.Pöppl, A.; Völkel, G., ESR and PhotoǦESR Investigations of Zinc Vacancies and
Interstitial Oxygen Ions in Undoped ZnO Ceramics. Physica Status Solidi (a) 1991,
125, 571-581.
224.Larina, L.; Tsvetkov, N.; Yang, J.; Lim, K.-S.; Shevaleevskiy, O., Electron
Paramagnetic Resonance Studies of Shallow Donors Behavior in Hydrogenated
ZnO Films. ECS Transactions 2010, 28, 161-167.
225.Geisler, C.; Simmons, G., High Temperature Induced EPR Signals in Zinc Oxide.
Physics Letters 1964, 11, 111-112.
226.Neumann, G., Current Topics in Materials Science, Edited by E. Kaldis. North-
Holland, Amsterdam 1981, 269.
227.Erhart, P.; Albe, K., Diffusion of Zinc Vacancies and Interstitials in Zinc Oxide.
Applied Physics Letters 2006, 88, 201918.
228.Smith, J.; Vehse, W., ESR of Electron Irradiated ZnO Confirmation of the F+
Center. Physics Letters A 1970, 31, 147-148.
Suranan Anantachaisilp References / 122

229.Kotsis, K.; Staemmler, V., Ab Initio Calculations of the O1s XPS Spectra of ZnO
and Zn Oxo Compounds. Physical Chemistry Chemical Physics 2006, 8, 1490-
1498.
230.Kumar, V.; Swart, H.; Ntwaeaborwa, O.; Kroon, R.; Terblans, J.; Shaat, S.; Yousif,
A.; Duvenhage, M., Origin of the Red Emission in Zinc Oxide Nanophosphors.
Materials Letters 2013, 101, 57-60.
231.Kim, Y.-S.; Tai, W.-P.; Shu, S.-J., Effect of Preheating Temperature on Structural
and Optical Properties of ZnO Thin Films by Sol–Gel Process. Thin Solid Films
2005, 491, 153-160.
232.Wei, X.; Man, B.; Liu, M.; Xue, C.; Zhuang, H.; Yang, C., Blue Luminescent
Centers and Microstructural Evaluation by XPS and Raman in ZnO Thin Films
Annealed in Vacuum, N2 and O2. Physica B: Condensed Matter 2007, 388, 145-
152.
233.Li, G.; Yi, Z.; Wang, H.; Jia, C.; Zhang, W., Factors Impacted on Anisotropic
Photocatalytic Oxidization Activity of ZnO: Surface Band Bending, Surface Free
Energy and Surface Conductance. Applied Catalysis B: Environmental 2014, 158,
280-285.
234.Ton-That, C.; Phillips, M. R.; Foley, M.; Moody, S. J.; Stampfl, A. P., Surface
Electronic Properties of ZnO Nanoparticles. Applied Physics Letters 2008, 92,
261916.
235.Girard, R.; Tjernberg, O.; Chiaia, G.; Söderholm, S.; Karlsson, U. O.; Wigren, C.;
Nylen, H.; Lindau, I., Electronic Structure of ZnO (0001) Studied by Angle-
Resolved Photoelectron Spectroscopy. Surface Science 1997, 373, 409-417.
236.Chiou, J.; Tsai, H.; Pao, C.; Kumar, K. K.; Ray, S.; Chien, F.; Pong, W.-F.; Tsai,
M.-H.; Chen, C.-H.; Lin, H.-J., Comparison of the Electronic Structures of Zn1í
XCoxO and Zn1í XMgxO Nanorods Using X-Ray Absorption and Scanning
Photoelectron Microscopies. Applied Physics Letters 2006, 89, 043121.
237.Song, S.; Hong, W.-K.; Kwon, S.-S.; Lee, T., Passivation Effects on ZnO Nanowire
Field Effect Transistors under Oxygen, Ambient, and Vacuum Environments.
Applied Physics Letters 2008, 92, 263109.
University of Technology Sydney References / 123

238.Stavale, F.; Pascua, L.; Nilius, N.; Freund, H.-J., Luminescence Properties of
Nitrogen-Doped ZnO. The Journal of Physical Chemistry C 2014, 118, 13693-
13696.
239.Choi, S.; Johnson, B. C.; Castelletto, S.; Ton-That, C.; Phillips, M. R.;
Aharonovich, I., Single Photon Emission from ZnO Nanoparticles. Applied Physics
Letters 2014, 104, 261101.
240.Kik, P.; Polman, A., Exciton–Erbium Energy Transfer in Si Nanocrystal-Doped
SiO2. Materials Science and Engineering: B 2001, 81, 3-8.
241.Shionoya, S.; Goldberg, P., Luminescence of Inorganic Solids. Academic, New
York 1966, 205.
242.Alemán, B.; Ortega, Y.; García, J. Á.; Fernández, P.; Piqueras, J., Fe Solubility,
Growth Mechanism, and Luminescence of Fe Doped ZnO Nanowires and Nanorods
Grown by Evaporation-Deposition. Journal of Applied Physics 2011, 110, 014317.
243.Müller, S.; Zhou, M.; Li, Q.; Ronning, C., Intra-Shell Luminescence of Transition-
Metal-Implanted Zinc Oxide Nanowires. Nanotechnology 2009, 20, 135704.
244.Kurbanov, S. S.; Kang, T. W., Spectral Behavior of the Emission around 3.31 eV
(a-Line) from ZnO Nanocrystals. Journal of Luminescence 2010, 130, 767-770.
245.Tainoff, D.; Masenelli, B.; Mélinon, P.; Belsky, A.; Ledoux, G.; Amans, D.;
Dujardin, C.; Fedorov, N.; Martin, P., Competition between Exciton-Phonon
Interaction and Defects States in the 3.31 eV Band in ZnO. Physical Review B 2010,
81, 115304.
246.Bebb, H. B.; Williams, E., Semiconductors and Semimetals. Academic Press, New
York 1972, 8, 181.
247.Shibata, H., Negative Thermal Quenching Curves in Photoluminescence of Solids.
Japanese Journal of Applied Physics 1998, 37, 550.