University of Ghana  http://ugspace.ug.edu.gh
Experimental and Computational Study of
Transition Metal Doped Zinc Oxide.
by
Azimatu Seidu
(10285118)
This thesis is submitted to the University of Ghana, Legon, in partial fulfilment for the award
of the degree of
MASTER OF PHILOSOPHY (M.PHIL.)
in Physics
July 2016
University of Ghana  http://ugspace.ug.edu.gh
i
Conference Presentations
1. Azimatu Seidu, Martin Egblewogbe and G. H Gebreyesus, Theoretical and Experimental
study of the electronic and optical properties of transition metal doped Zinc Oxide, 8th
African International Conference of the African Materials Research Society.
6th – 11th December, 2015.
Materials Science and Engineering, College of Basic and Applied Sciences, School of
Engineering, University of Ghana, Legon.
2. Azimatu Seidu, Martin Egblewogbe and G. H Gebreyesus, Synthesis and characterization
of pure and transition metal-doped ZnO, College of Basic and Applied Sciences (CBAS)
Science and Development Platform.
18th February, 2016, Centre for African Wetlands, University of Ghana, Legon.
University of Ghana  http://ugspace.ug.edu.gh
Declaration
I, Azimatu Seidu, hereby declare that this project except for the references to the work of others
which have been duly acknowledged is the result of my own research as an M.Phil student of
the University of Ghana, under the supervision of Dr. M. N. Y. Egblewogbe and Dr. G. H.
Gebreyesus. It has not been presented for the award of any degree elsewhere.
Candidate
Azimatu Seidu
Signature .................................................. Date .........................
Supervisors
Dr. Martin N. Y. H Egblewogbe (Principal Supervisor)
Signature .................................................. Date .........................
Dr. Gebremedhn H. Gebreyesus
Signature .................................................. Date .........................
ii
University of Ghana  http://ugspace.ug.edu.gh
Dedication
This work is dedicated first and foremost to Allah for His mercies and guidance. I also dedicate
it to my mother for her tireless support throughout my life and education.
iii
University of Ghana  http://ugspace.ug.edu.gh
Acknowledgments
My sincere gratitude goes to my principal supervisor, Dr. M. N. Y. H. Egblewogbe for his
constant advice, guidance and support throughout my study in the university and within the
period of this research.
I am especially grateful to my second supervisor, Dr. G. H. Gebreyesus, for his tireless
effort to help me in understanding the computational aspect of my work which was entirely
new to me and to Dr. G. Nkrumah-Buandoh, head of department for his support throughout my
postgraduate study.
I would like to appreciate the support of the Carnegie Foundation for the award a research
scholarship which was instrumental in the completion of the research work.
I attended the 2016 African School for Electronic Structure Methods and Applications held
at the University of Ghana in June, and I would like to thank the organisers for the opportunity.
I acknowledge the support of the members of faculty of the department of Physics especially
Dr. A. A. Yankson for his friendship and encouragement, and the technicians who helped to
make this project a success. I especially mention Ms Beatrice Agyapomah, who did most of the
x-ray measurements. I salute the National Service personnel for their friendship and assistance.
I must say thank you as well to Dr. Regina Appiah-Opong of the Department of Clinical
Pathology at the Noguchi Memorial Institute for Medical Research, where I carried out the
photoluminescence measurements.
Most of the computational work was done on the nascent High Performance Cluster at the
University of Ghana Computing Systems, and I recognise their support.
Last but not the least, I would like to express my sincere gratitude to my family, for their
support and prayers.
iv
University of Ghana  http://ugspace.ug.edu.gh
Abstract
Pristine and transition metal (Fe, Mn, Co and Ni) doped ZnO powders were synthesised us-
ing a hydrothermal method, and characterised by x-ray powder diffraction, ultraviolet-visible
absorption spectroscopy, and photoluminescence (PL) emission spectroscopy. Synthesis was
carried out at pH 3 and pH 5, and the powders annealed at 280 ◦C and 600 ◦C. Doping was
done at concentrations of 1, 2, 4 and 8 mol %. Computational studies were also carried out on
undoped and transition metal (Fe, Mn, Co and Ni) doped ZnO using the Generalized gradient
Approximation (GGA) of the Density Functional Theory (DFT) implemented in the ab-initio
computational suite, Quantum ESPRESSO. Doping was done at concentrations of 5.56 % and
11.11 %. X-ray diffraction spectra showed the presence of secondary phases of Fe3O4, Co3O4,
Mn3O4, and NiO at higher doping concentrations. However, the results indicated that synthesis
at higher pH diminished the occurrence of these secondary phases. In the experimental work,
the band gap energies from the absorption edge of the ultraviolet-visible absorption spectra de-
creased with doping concentration for all dopants except for nickel. The band gap energy from
the near band edge of the PL spectra on the other increased.
The computational results and the experimental results both showed that the band energy in-
creased with doping concentration. The unit cell volume also increased for iron and manganese
doping in both the experimental and computational results. However, while the computational
results predicted a narrowing in the unit cell volume for cobalt and nicked doping, the experi-
mental results showed an increase for cobalt and a marginal increase for nickel.
v
University of Ghana  http://ugspace.ug.edu.gh
List of Abbreviations
AR Aspect Ratio
CPMD Car-Parrinello molecular dynamics
CVD Chemical Vapour Deposition
DFT Density Functional Theory
DMS Dilute Magnetic Semiconductor
FCC Face Centred Cubic
FPMD First principle molecular dynamics
FWHM Full width at half maximum
GGA Generalised Gradiennt Approximation
HF Hartree-Fock
HSE Heyd-Scuseria-Ernzerhof
LDA local density approximation
PBE Perdew-Burke-Ernzherof
PL Photoluminescence
PLD Pulsed Laser Deposition
PP Pseudopotential
vi
University of Ghana  http://ugspace.ug.edu.gh
vii
PW Plane Wave
PWSCF Plane wave self consistent field
QE Quantum ESPRESSO
TCO Transparent conducting oxide
TM Transition Metal
US-PP ultra-soft pseudopotential
UV-VIS Ultraviolet Visible
W-H Williamson-Hall
University of Ghana  http://ugspace.ug.edu.gh
List of Figures
1.1 Wurtzite structure of ZnO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.1 The band structure of ZnO . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2 A schematic diagram of a chemical vapour deposition reactor . . . . . . . . . . 16
2.3 A schematic of a ball milling reactor. . . . . . . . . . . . . . . . . . . . . . . . 18
2.4 A schematic of an electrochemical reactor . . . . . . . . . . . . . . . . . . . . 18
2.5 A schematic of PLD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.1 A flowchart of the procedure for the calculation of total energy from DFT . . . 33
4.1 A schematic of a UV-VIS absorption spectrometer. . . . . . . . . . . . . . . . 36
4.2 UV-spectrum of a ZnO sample. . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.3 A schematic diagram of a photoluminescence spectrometer . . . . . . . . . . . 38
4.4 A schematic of band transitions in a semiconductor . . . . . . . . . . . . . . . 39
4.5 A schematic of an x-ray powder diffractometer . . . . . . . . . . . . . . . . . 40
4.6 ZnO supercell containing 36 atoms . . . . . . . . . . . . . . . . . . . . . . . . 43
5.1 Williamson-Hall plot for 2 mol % Co-doped ZnO . . . . . . . . . . . . . . . . 47
5.2 XRD pattern for reference and as-prepared undoped ZnO . . . . . . . . . . . . 48
5.3 X-ray diffraction patterns for Fe- and Mn-doped ZnO. . . . . . . . . . . . . . . 49
5.4 X-ray diffraction patterns for Co- and Ni-doped ZnO. . . . . . . . . . . . . . . 50
5.5 Variation in the volume of the unit cell with doping concentration . . . . . . . 52
5.6 UV-VIS Absorption Spectra . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
5.7 PL Emission Spectra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
viii
University of Ghana  http://ugspace.ug.edu.gh
LIST OF FIGURES ix
5.8 XRD pattern for 4 and 8 mol % Fe - doped samples at pH 5 and 3 . . . . . . . 64
5.9 x-ray diffraction patterns of 4 and 8 mol % Mn-doping at pH 3 and 5 . . . . . . 65
5.10 X-ray diffractograms of 4 and 8 mol% Co-doping . . . . . . . . . . . . . . . . 66
5.11 XRD pattern for 280 ◦C annealing of 4 and 8 mol % Ni-doped samples at pH 5
and 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.12 UV-VIS absorption Spectra for 4 and 8 mol% doping at pH 5 and 3 . . . . . . . 70
5.13 PL emission spectra of 4 and 8 mol % doping at pH3 . . . . . . . . . . . . . . 71
5.14 XRD pattern for 4 and 8 mol % Fe-doped ZnO 600◦ annealing. . . . . . . . . . 75
5.15 XRD pattern for 4 and 8 mol % Mn-doped ZnO at 600◦ annealing. . . . . . . . 76
5.16 XRD pattern for 4 and 8 mol % Co-doped ZnO at 600◦ annealing. . . . . . . . 78
5.17 XRD pattern for 4 and 8 mol % Ni-doped ZnO at 600◦ annealing. . . . . . . . 79
5.18 PL emission spectra for 4 and 8 mol % doped ZnO at pH 3 . . . . . . . . . . . 84
5.19 PL emission spectra for 4 and 8 mol % doped ZnO at pH 5 . . . . . . . . . . . 85
5.20 Band structure and density of states for undoped ZnO . . . . . . . . . . . . . . 86
5.22 Band structure for 5.56 at. % doping . . . . . . . . . . . . . . . . . . . . . . . 89
5.23 Band structure for 11 at. % doping . . . . . . . . . . . . . . . . . . . . . . . . 90
5.24 Density of states for 5.56 at. % doped ZnO . . . . . . . . . . . . . . . . . . . 91
5.25 Density of states for 11 at. % doping . . . . . . . . . . . . . . . . . . . . . . . 92
A.2 PL Emission Spectra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
B.2 UV-VIS Absorption Spectra At 600 ◦C Annealing (pH 5) . . . . . . . . . . . . 111
B.3 UV-VIS Absorption Spectra At 600 ◦C Annealing (pH 3) . . . . . . . . . . . . 112
University of Ghana  http://ugspace.ug.edu.gh
List of Tables
5.1 Crystal parameters and reflections from (102) planes (Fe-doping) . . . . . . . . 51
5.2 Crystal parameters and reflections from the (102) plane (Mn-doping) . . . . . . 53
5.3 Crystal parameters reflections from (102) plane (Co-doping) . . . . . . . . . . 54
5.4 Crystal parameters and reflections from (102) plane (Ni-doping) . . . . . . . . 55
5.5 Aspect Ratios (AR) for doped ZnO . . . . . . . . . . . . . . . . . . . . . . . . 55
5.6 Changes in the unit cell volume with Doping . . . . . . . . . . . . . . . . . . 56
5.7 Aspect Ratios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
5.8 UV-VIS Band Gap Energy (Fe-doping) . . . . . . . . . . . . . . . . . . . . . 57
5.9 UV-VIS Band Gap Energy (Mn-doping) . . . . . . . . . . . . . . . . . . . . . 58
5.10 UV-VIS Band Gap Energy (Co - doping) . . . . . . . . . . . . . . . . . . . . . 59
5.11 UV-VIS Band Gap Energy (Ni - doping) . . . . . . . . . . . . . . . . . . . . . 59
5.12 Change in Band Gap Energy With Doping . . . . . . . . . . . . . . . . . . . . 62
5.13 Lattice parameters, crystallite size and strain for 4 and 8 mol % Fe-doping . . . 63
5.14 Lattice parameters, crystallite size and strain for 4 and 8 mol % Mn-doping . . 64
5.15 Lattice parameters, crystallite size and strain for 4 and 8 mol % Co-doping . . . 65
5.16 Lattice parameters, crystallite size and strain for 4 and 8 mol % doping. . . . . 66
5.17 Aspect ratio of 4 and 8 mol% Fe-doping . . . . . . . . . . . . . . . . . . . . . 67
5.18 UV-VIS Band Gap Energy for 4 and 8 mol % Fe-doping ZnO at pH 5 and pH 3 68
5.19 PL Band Gap Energy from PL emission spectra of 4 and 8 mol % Fe - doping . 69
5.20 UV-VIS Band Gap Energy for 4 and 8 mol % Mn - doping . . . . . . . . . . . 69
5.21 PL Band Gap Energy for 4 and 8 mol % Mn - doping ZnO at pH 5 and 3 . . . . 72
5.22 UV-VIS Band Gap Energy for 4 and 8 mol % Co-doping . . . . . . . . . . . . 72
x
University of Ghana  http://ugspace.ug.edu.gh
LIST OF TABLES xi
5.23 PL Band Gap Energy for 4 and 8 mol % Co-doping. . . . . . . . . . . . . . . . 72
5.24 UV-VIS Band Gap Energy for 4 and 8 mol % Ni - doping at pH 5 and 3 . . . . 73
5.25 PL Band Gap Energy for 4 and 8 mol % Ni - doping at pH 5 and 3 . . . . . . . 73
5.26 Lattice parameters, crystallite size and strain at pH 5 (Fe-doping) . . . . . . . . 75
5.27 Lattice parameters, crystallite size and strain at pH 3 (Fe-doping) . . . . . . . . 75
5.28 Lattice parameters, c/a ratio, crystallite size and strain for 4 and 8 mol % Mn-
doping at pH 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.29 Lattice parameters,crystallite size and strain for 4 and 8 mol % Mn-doping at
pH 5 with different annealing temperatures . . . . . . . . . . . . . . . . . . . 77
5.30 Lattice parameters, crystallite size and strain for 4 and 8 mol % Co-doping at
pH 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.31 Lattice parameters, crystallite size and strain for 4 and 8 mol % Co-doping at
pH 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.32 Lattice parameters, crystallite size and strain for 4 and 8 mol % Ni-doping at pH 5 79
5.33 Lattice parameters, crystallite size and strain for 4 and 8 mol % Ni-doping at pH 3 79
5.34 Aspect Ratios for 4 and 8 mol% Doping. A.T. = Annealing Temperature. . . . . 80
5.35 Band gap from UV absorption spectra for 4 and 8 mol % Fe-doping . . . . . . 81
5.36 Band gap from PL emission spectra for 4 and 8 mol % Fe-doping . . . . . . . . 81
5.37 Band Gap Energy from UV-VIS Absorption for 4 and 8 mol % Mn-doping . . . 82
5.38 Band gap Energy from PL Emission for 4 and 8 mol % Mn-doping . . . . . . . 82
5.39 Band Gap Energy from UV-VIS Absorption for 4 and 8 mol % Co-doping . . . 82
5.40 Band Gap Energy from PL Emission for 4 and 8 mol % Co-doping . . . . . . . 83
5.41 Band Gap Energy from UV-VIS Absorption for 4 and 8 mol % Ni-doping . . . 83
5.42 Band Gap Energy from PL Emission for 4 and 8 mol % Ni-doping . . . . . . . 83
5.43 Lattice parameters for Fe-doped ZnO . . . . . . . . . . . . . . . . . . . . . . . 86
5.44 Lattice parameters for Mn-doped ZnO . . . . . . . . . . . . . . . . . . . . . . 87
5.45 Lattice parameters for Co-doped ZnO . . . . . . . . . . . . . . . . . . . . . . 87
5.46 Lattice parameters for Ni-doped ZnO . . . . . . . . . . . . . . . . . . . . . . . 88
5.47 Band Gap Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
University of Ghana  http://ugspace.ug.edu.gh
Contents
List of Abbreviations vi
List of Figures viii
List of Tables x
1 Introduction 1
1.1 Zinc Oxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1.1 Nano-structured materials . . . . . . . . . . . . . . . . . . . . . . . . 4
1.1.2 Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.2 Motivation and Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2 Literature Review 8
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.2 Crystallographic and Electronic Properties . . . . . . . . . . . . . . . . . . . . 8
2.3 Optical Absorption and Luminescence . . . . . . . . . . . . . . . . . . . . . . 11
2.4 Doping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.4.1 Transition Metal Doping . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.5 Methods of Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.5.1 Chemical Vapour Deposition . . . . . . . . . . . . . . . . . . . . . . . 16
2.5.2 Ball Milling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.5.3 Electrochemical Synthesis . . . . . . . . . . . . . . . . . . . . . . . . 17
2.5.4 Epitaxial Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.5.5 Pulsed Laser Deposition (PLD) . . . . . . . . . . . . . . . . . . . . . 19
xii
University of Ghana  http://ugspace.ug.edu.gh
CONTENTS xiii
2.5.6 Wet-chemical methods . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3 Theory 22
3.1 Computational Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.1.1 Born-Oppenheimer Approximation . . . . . . . . . . . . . . . . . . . 24
3.1.2 Density Functional Theory . . . . . . . . . . . . . . . . . . . . . . . . 26
3.2 Quantum-ESPRESSO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4 Method 34
4.1 Synthesis Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
4.2 Characterisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.2.1 Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.3 Computational Routine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
5 Results and Discussion 45
5.0.1 Crystallographic Parameters . . . . . . . . . . . . . . . . . . . . . . . 46
5.0.2 Determination of the Band Gap Energy . . . . . . . . . . . . . . . . . 46
5.0.3 Size and Strain by the Williamson Hall Method . . . . . . . . . . . . . 46
5.0.4 Aspect Ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
5.1 Synthesis at pH 5 and annealing at 280 ◦C . . . . . . . . . . . . . . . . . . . . 47
5.1.1 Undoped ZnO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
5.1.2 Fe-doping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
5.1.3 Mn-doping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.1.4 Co-doping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
5.1.5 Ni-doping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
5.1.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
5.2 Band Gap Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
5.2.1 Fe-doping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
5.2.2 Mn-doping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
5.2.3 Co-doping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
5.2.4 Ni-doping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
University of Ghana  http://ugspace.ug.edu.gh
CONTENTS xiv
5.2.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
5.3 Synthesis at Elevated pH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.3.1 Lattice parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.3.2 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
5.3.3 Band Gap Energies . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
5.3.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
5.4 Annealing at 600 ◦C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
5.4.1 Lattice Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
5.4.2 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
5.4.3 Band Gap Energies . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
5.4.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
5.5 Computational Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.5.1 Undoped ZnO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.5.2 Lattice Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
5.5.3 Band Gap Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
5.5.4 Band Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
5.5.5 Density of states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
6 Conclusion 93
6.0.1 Prospects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
References 96
Appendices 109
A 110
A.1 PL emission spectra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
B 111
B.1 UV-VIS spectra at 600◦C annealing (pH 5) . . . . . . . . . . . . . . . . . . . . 111
University of Ghana  http://ugspace.ug.edu.gh
Chapter 1
Introduction
1.1 Zinc Oxide
Zinc oxide (ZnO) is found in nature as the mineral zincite, usually with the presence of certain
amounts of manganese and some other elements which gives it a yellow to red colour. It is a
chemically stable and biocompatible material (Devaramani et al. (2008)), and has been used for
various purposes for hundreds of years (Moezzi et al. (2012)). The reported uses dating from
2000 B.C. include the addition of ZnO to ointments used for treating skin conditions, and also
as a source of zinc metal. From the eighteenth century onwards, zinc oxide was produced as a
white pigment, and as an additive to rubber (Moezzi et al. (2012)). Zinc oxide has continued
to be used in various applications in electronics, photovoltaics, bio- and gas-sensing, as well as
in lighting. ZnO is thought of as one of the most important semiconductor oxides at present
(Ischenko et al. (2003)), and is a technological material of great promise.
ZnO is a II-VI n-type semiconductor material, typically crystallizing in the wurtzite poly-
morph, even though it has three crystal polymorphs, wurtzite, zinc blende and rock-salt. The
wurtzite ZnO unit cell belongs to the space group P63mc in the Hermann-Mauguin notation with
unit cell parameters a = 3.2488Å and c = 5.2066Å, with the c/a ratio to be between 1.593 and
1.603, and internal parameter u = 0.375 (Özgür et al. (2005)). Figure 1.1 shows the wurtzite
ZnO unit cell. The structure is made up of two inter penetrating hexagonal-close packed (hcp)
sub-lattices. Each of the sub-lattices consists of the atom of one element displaced with respect
to an atom of the second element along the threefold c-axis by an amount of u = 3 = 0.375.
8
1
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 1. INTRODUCTION 2
c-axis Oxygen  
Zinc 
a-axis
Figure 1.1: Wurtzite structure of ZnO
Created using the Xcrysden software.
Zinc oxide is readily synthesised by several methods, ranging from solution-based to gas-
phase methods. Some of these synthesis methods lead to the formation of ZnO crystals that
have a wide range of morphologies at the nanoscale such as wires, rings, cages, belts (Wang
(2004)), tetrapods (Fuxue et al. (2014)) and so on, showing that the range of morphologies that
can be synthesised is quite extensive.
ZnO exhibits high photo conductivity, and has pyroelectric properties (Wang et al. (2015)).
ZnO also exhibits piezoelectric properties (Wang (2007) and Klingshirn (2007)), which makes
possible the fabrication of ZnO based micro-electrochemical devices (Ko et al. (2003)).
In the last few decades research into wide-band materials has increased because of the im-
proved capabilities in tailoring material properties, even at the nanoscale (Wang (2004)). ZnO
is a material that has seen much of this research interest. Already well-known for applications
such as use in sun-screen (Barker and Branch (2008)), voltage surge arrestors (Thipprasert and
Sritakaew (2014)), the potential applications for ZnO now range from gas sensors (Krishnaku-
mar et al. (2009)), photovoltaics (Chu et al. (2011)) bio-sensors (Arya et al. (2012)), UV-sensors
(Xu et al. (2008)), opto-electronics (Gondoni et al. (2012)) to novel applications such as direct
drug delivery into diseased cells (Rasmussen et al. (2010) and Nie et al. (2006)).
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 1. INTRODUCTION 3
It is further expected that the properties of ZnO can be exploited in the fabrication of devices
for applications in lighting and optical switching devices (Huang et al. (2009)). The lumines-
cence properties of ZnO are well known, which have led to the use of ZnO powders in phosphor
applications for lighting (Shionoya and Yen (1999)). The well-known green photoluminescence
from ZnO has been of interest for the development of flat panel displays (Vanheusden et al.
(1996)). The luminescence property of ZnO under electrical or optical excitation, has been
found to span almost the entire optical range, and includes the near UV as well (Özgür et al.
(2005), Bhat and Deepak (2005), Shionoya and Yen (1999) and Mote et al. (2011)). With an
exciton binding energy of 60 meV at room temperature (Özgür et al. (2005), Shionoya and Yen
(1999)), it is expected that ZnO can be used in electroluminescent and photoluminescent appli-
cations, especially since single crystal ZnO has a high optical transparency (Chey (2015), Bhat
and Deepak (2005), Karmakar et al. (2007) and Mote et al. (2011)). This also makes ZnO a
good candidate for making transparent conducting oxide (TCO) materials, required for opto-
electronic applications (Raja et al. (2014) and Li et al. (2010)). The optical and luminescence
properties of ZnO can also be tailored by doping to enhance its optical and electronic properties.
With regards to applications exploiting the electrical properties of ZnO, it is already well
known that ZnO powders are already widely used in the manufacture of varistors (Kohan et al.
(2000)). It is expected that electronics based on ZnO (compared with the current silicon-based
electronic devices) will have wider applications and better performance on account of improved
electronic conductivity, lower power consumption, better stability, fast electron transfer kinetics
and improved storage capacity. In addition, with a band gap in the region of 3.3 eV (Meyer et al.
(2004) and Srikant and Clarke (1998)), ZnO crystals are able to withstand much higher electric
fields than silicon crystals, such that electronic devices based on ZnO will be able to operate at
higher power and higher temperatures (Özgür et al. (2005)).
Globally, there is a current drive towards the generation of cheap and reliable energy from
photovoltaic sources. The realisation of this requires the development of new devices capable of
efficient energy harvesting, which in turn requires scientific study on especially the electronic,
optical and structural properties of the photo-active materials. ZnO is one of the materials that
has significant potential in this area, which makes an investigation into the electronic and optical
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 1. INTRODUCTION 4
properties a relevant endeavour. Closely associated with photovoltaic applications are applica-
tions concerned with lighting and luminescence. For instance, ZnO has been used as a substitute
for TiO2 in both dye sensitized and quantum-dot sensitized exciton solar cells (Malloci et al.
(2012)).
Research is on-going to harness the opto-electronic properties of ZnO in manufacturing
transparent conducting oxide (TCO) materials for opto-electronic devices. This is a promising
track since the electrical conductivity of ZnO is enhanced by doping elements such as alu-
minium and gallium (Raja et al. (2014), Özgür et al. (2005) and Klingshirn (2007)).
The theoretical prediction of above room temperature ferromagnetism in ZnO nano based
materials in the last decade by (Dietl et al. (2000)), potentially allowing for the fabrication of
spintronic devices such as light emitting diodes (Hao et al. (2012)) has led to further theoretical
and experimental work, involving the doping of ZnO with transition metals such as iron, cobalt
and manganese (Bhat and Deepak (2005) and Klingshirn (2007)).
The realisation of the full potential of ZnO as a technological material depends on successful
doping into p-type, and also achieving controlled n-type doping. Naturally ZnO tends to remain
n-type even when doped with electron deficient species in substitution for zinc. The reason for
this has been elusive, but current research seems to suggest that this behaviour is caused by
shallow donors such as hydrogen introduced inadvertently into the lattice during the synthesis
process (Janotti and van de Walle (2009)). In any case there remains much interest in delivering
simple and efficient doping methods for ZnO.
1.1.1 Nano-structured materials
Objects which have at least one spatial dimension in the range of 0.1 nm to 100 nm are classified
as nanomaterials. This includes films that are so thin as to be only a few nanometers thick, wires
that have cross-sections that are only a few nanometers across, or objects whose size is within
the nanometric range.
The interest in studying material properties at the nanoscale is mainly due to the enhance-
ment of the quantum effect which enables scientists to utilize the physical, mechanical and
optical properties of the material. Properties such as the melting point, electrical conductivity,
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 1. INTRODUCTION 5
mechanical strength, optical characteristics, and the chemical reactivity change as a function
of size (Özgür et al. (2005)). The possibility of controlling the physical properties (including
the morphology) of matter at sub-micron levels, and the opportunity for the scientific study of
novel materials has led to the increase in studies of nanostructured materials along with the ad-
vancement in synthesis and analytical methods over the first decade of the twenty-first century
(Zhang et al. (2013)).
For example, it is expected that devices can be fabricated that will enable direct drug delivery
to diseased cells, since there are nanomaterials with the ability to cross biological cell-barriers
without damaging them (Nguyen (2011), Zhang et al. (2013) and Bae et al. (2011)). Some
research has suggested that nano-sized drugs are able to effectively destroy cancer cells (Nikalje
(2015)). In addition, there is the expectation that biological, optical, and chemical sensors with
enhanced selectivity and sensitivity can be fabricated with nanostructured materials.
The changes in the electronic properties of nanostructured material is due to the spatial
confinement of their electronic and vibrational excitations (Devaramani et al. (2008)). This
typically appears as a change in the electronic band gap of the material, allowing for the precise
engineering of the band gap for specific purposes. Other basic physical effects such as increased
surface area can also cause the changes stated above for a nanoscaled material.
1.1.2 Synthesis
Even though zinc oxide occurs naturally, it is typical to prepare it in the laboratory for reasons
of purity and high crystallinity. The synthesis of pure ZnO nanomaterial can be achieved in the
solution, solid or gas phase. There are several methods of synthesis in the liquid, solid and gas
phases. These include hydrothermal synthesis, ball milling, and chemical vapour deposition.
These are described in more detail in Chapter 2.
In this work, ZnO powder was prepared with a hydrothermal method following the work of
(Jyoti et al. (2013)) using ZnCl2 as a precursor and KOH as precipitant. Doping was achieved
by adding the metal chlorides in the desired proportion during the synthesis. ZnO prepared
in this manner, was annealed at temperatures of 280◦ C and characterized with x-ray powder
diffraction, (UV-VIS ) absorption spectroscopy and photoluminescence from which the crystal-
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 1. INTRODUCTION 6
lographic parameters and band gap were calculated.
1.2 Motivation and Scope
In the preceding sections, the potential for applications based on ZnO materials have been dis-
cussed. It is clear that this potential is extensive given the remarkable properties of ZnO from
the macroscopic to the nanoscale. In order to harness these potentials, simple and efficient
methods of synthesis yielding high quality nanoparticles need to be studied and understood.
These methods should also allow doping across a wide range of concentrations and also allow
for the control of crystal size and modification of the band-gap energy. In addition to exper-
imentally establishing such methods, the theoretical basis of the behaviour of the synthesised
materials needs to be investigated.
The current work seeks to synthesize undoped and doped ZnO nanocrystals via a simple
hydrothermal method, and to study experimentally the effects of the doping on the crystal lattice
parameters, the band gap, and optical characteristics. In addition, computational work will be
carried out on the undoped and doped ZnO to determine the crystal lattice parameters and the
band gap.
Doping will be carried out with the transition metals iron (Fe), cobalt (Co), nickel (Ni) and
manganese (Mn) at doping concentrations varying from 0-8 mol %.
The as-prepared samples will be characterised by
1. X-ray powder diffraction, from which data the crystal lattice parameters will be deter-
mined. Also, the crystallite size and strain will be calculated using the Williamson-Hall
(W-H) method.
2. UV-VIS absorption spectroscopy, from which data the band gap energies will be calcu-
lated.
3. Photoluminescence (PL) spectroscopy, from which data the band gap energies will be
calculated. PL spectroscopy will also yield information about the presence of inter-band
transition states.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 1. INTRODUCTION 7
Computationally, th electronic properties of undoped ZnO and transition metal doped ZnO
will be determined using the Quantum ESPRESSO (QE) package.
Doping will be carried out with the transition metals iron (Fe), cobalt (Co), nickel (Ni) and
manganese (Mn) at doping concentrations 5.56 at.% and 11 at. % on a ZnO supercell containing
36 atoms. The crystal parameters to be determined computationally are a, c, the c/a ratio, and
the unit cell volume. Also, the band gap energies and the band structures will be calculated.
Common trends between the experimental and computational results will be determined.
University of Ghana  http://ugspace.ug.edu.gh
Chapter 2
Literature Review
2.1 Introduction
In this chapter, literature pertinent to this work is reviewed. Reference will be made to journal
publications, article reviews, books and Ph.D. thesis reports. The discussion will focus on
aspects of research into the properties of ZnO, ranging from experimental to the computational.
2.2 Crystallographic and Electronic Properties
There is a substantial body of literature available reporting on the experimental and computa-
tional studies into the properties of ZnO. The rising importance of this material has also led
to more work both experimentally and computationally (Özgür et al. (2005)), the latter being
aided by the more powerful computer architecture and software routines which are continuously
being developed.
Most group II-VI binary semiconductors like ZnO crystallize in either the cubic zinc-blende
or hexagonal wurtzite structures. Typically the bonds have covalent properties. However, the
bonding in ZnO is intermediate, having both covalent and ionic characteristics (Özgür et al.
(2005)). Many of the interesting properties as well as the challenges presented to the full utili-
sation of ZnO relate to this.
ZnO typically crystallizes in the wurtzite structure, and this has been experimentally verified
extensively (Özgür et al. (2005)). Some researchers have also experimentally determined the
8
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 2. LITERATURE REVIEW 9
high-pressure phases of ZnO, leading to interesting results showing that the NaCl (rocksalt)
phase is a high-pressure phase with transition starting from about 8 GPa accordingly as the
pressure increased (Karzel et al. (1996)). It has also been proposed that at even higher pressures,
ZnO might take on the cubic CsCl structure (Özgür et al. (2005)).
Under ambient conditions, the experimental values for the lattice parameters a and c fall
within the range of 3.2475 Å to 3.2501 Å and 5.2042 Å to 5.2075 Å (Özgür et al. (2005)).
The band structure of a crystalline semiconductor material gives details about the energy
states available for electronic population, and is therefore extremely useful in explaining many
of the observed electrical and optical properties of the material. The band structure is usually
calculated from approximations of the Schrödinger equation for a solid, as will be discussed
in more detail in Chapter 4. Figure 2.1 shows the band structure of ZnO calculated using the
Heyd-Scuseria-Ernzerhof (HSE) functional (Janotti and van de Walle (2009)). The abscissa
indicates the path in k-space in the Brilloun zone, and the ordinate the associated energy levels.
It can be seen that the lowest conduction band forms a ‘valley’ at the Γ point (where k = 0) and
the highest valence band forms a ‘hill’ at the same point. This makes ZnO a ‘direct’ bandgap
material, since electronic transitions are most probable at the Γ point and will occur without the
need for phonon assistance, as would be the case in an ‘indirect’ bandgap material.
Figure 2.1: The band structure of ZnO
From Janotti and van de Walle (2009)
There have been different approaches which have been used to calculate the band structure
of the different ZnO structures, and these are reported in the next section.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 2. LITERATURE REVIEW 10
Experimentally, the most directly accessible information about the band structure is the
band gap energy. There are various methods for measuring the band gap energy, such as photon
reflection, absorption and emission, in addition to electrical measurements involving the Hall
effect.
Mang et al. (1995) measured the bandgap of ZnO at a temperature of 6 K using the technique
of two photon absorption, which yielded a value of 3.4553 eV. Using reflectance methods,
(Reynolds et al. (1999)) determined the band gap at 2 K to be 3.4465 eV. Shionoya and Yen
(1999) reported the band gap as 3.436 eV at 4 K and 3.2 eV at room temperature.
The defect structure of ZnO has been the subject of research interest especially since many
of the interesting possibilities for luminescence applications are related to the defect-induced
interband states in ZnO (Shionoya and Yen (1999)). For example, Kohan et al. (2000) investi-
gated the intrinsic defects in ZnO using the (LDA) of the density functional theory (DFT). In
this study, the crystal and electronic structure and the formation energy of native point defects
were calculated. Janotti and van de Walle (2007) also studied the intrinsic defects in ZnO us-
ing the same method and concluded that the intrinsic defects in ZnO were not due to oxygen
and zinc vacancies but rather due to unintentional hydrogen doping. Similar conclusions were
drawn concerning the origin of the intrinsic defects in ZnO (van de Walle (2001)).
Calculations of the band gap energy have also been made using DFT-LDA and DFT-GGA
(Kohan et al. (2000), and Zhang et al. (2001). However, DFT is known for its under estimation
of the experimental band gap (Kohan et al. (2000)) as observed by Kohan et al. (2000) and
Janotti and van de Walle (2007) who reported for ZnO band gap energies of 0.91 eV and 0.8
eV respectively, which does not compare well with the experimental value of about 3.20 eV.
The LDA and generalised gradient approximation (GGA) use exchange correlation functionals
that suffer from an incomplete cancellation of the self interactions of the electrons and the
unavailability of integer discontinuity in the exchange and correlation energy upon addition of
an electron (Wróbel et al. (2009)). LDA and GGA also underestimates the binding energies
of the semicore 3d states of ZnO (Janotti and van de Walle (2007)). Further approximations
to DFT have been employed to correct the calculated value of the band gap energy. Some of
these approximations include LDA+U (Janotti and van de Walle (2007)) which gives a partial
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 2. LITERATURE REVIEW 11
correction to the band gap energy. For instance, using LDA+U, the band gap of ZnO has
been reported as 1.28 and 1.51 eV by Shih et al. (2010) and Janotti and van de Walle (2007)
respectively. Other studies have been made using DFT (Wang et al. (2013), Badaeva et al.
(2008) and Yao et al. (2012)). Badaeva et al. (2008) for example calculated the band gap
energy of Co-doped ZnO quantum-dots using different functionals and concluded that with the
hybrid PBE1 (PBE) level theory, the magnetism of DMS nanocrystals can be computationally
determined.
Using methods such as GW, better approximations to the band gap energy of semiconductors
can be calculated. Shih et al. (2010) reported a band gap energy of about 3.6 eV using the
plasmon pole model. Combinations such as LDA + GW and LDA + U + GW have also been
reported by Friedrich et al. (2011) to yield band gap energies of 3.4 and 3.6 eV respectively,
which are close to the experimental values.
The lattice parameters a and c have also been calculated by Janotti and van de Walle (2007)
who report the values to be 3.195 Å and 5.160 Å, which compares well with the experimental
values which are within the range of 3.2475 Å to 3.2501 Å and 5.2042 Å to 5.2075 Å (Özgür
et al. (2005)).
2.3 Optical Absorption and Luminescence
The optical characteristics of ZnO have been measured using the usual spectroscopic methods
of photon absorption and reflection spectroscopy, photoluminescence spectroscopy, spectro-
scopic ellipsometry, as well as extended spectroscopic methods such as cathodo-luminescence,
calorimetric luminescence among others (Özgür et al. (2005)).
ZnO crystals have been extensively characterised with photoluminescence (PL) spectroscopy,
especially on account of the green luminescence, the source of which has been in contention.
The room temperature PL spectrum of ZnO typically shows two distinct features, the near band
edge emission which occurs in the ultraviolet (UV) and typically one or more broad band emis-
sions due to defects or deep level inter-band states (Willander et al. (2010)). The deep level
emissions occur at wavelengths between 400 and 750 nm. These emissions are broad as a result
of superposition of many different deep levels. In ZnO, there are two main deep level emissions,
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 2. LITERATURE REVIEW 12
intrinsic and extrinsic deep levels. The intrinsic deep level in ZnO is believed to have originated
from oxygen vacancies (Vo), zinc vacancies (Vzn), oxygen interstitials (Oi), zinc interstitials
(Zni), oxygen anti-sites (Ozn) and zinc anti-sites (Zno) (Willander et al. (2010)). Combinations
of two point defects or an extrinsic and a point defect may also result in deep level emission.
One typical example of this is the VoZni cluster which occurs around 575 nm below the mini-
mum of the conduction band, corresponding to green light emission which is also suspected to
be due to Vzn (Willander et al. (2010)).
There are many such defects in synthesised ZnO, leading to the observed phenomenon of
ZnO crystals emitting almost across the visible spectrum in colours of red, yellow, blue, green,
orange, and even violet (Willander et al. (2009)).
With such rich features in optical characteristics, there is great interest in developing the
cheaper and non-toxic ZnO as an optical material in order to replace GaN compounds which
form the basis of blue and near UV optoelectronic devices (Moezzi et al. (2012)).
The utility of ZnO can be greatly extended if p-type doping can be reliably and reproducibly
achieved. This will enable the deployment of pn-junction devices, and open the door to a vast
range of applications. However, as has been mentioned, p-type doping has proven elusive.
2.4 Doping
Doping is the introduction of foreign atoms into a pure (or intrinsic) semiconductor to alter its
properties. Semiconductors are mainly doped to increase the charge carrier concentration to
enhance their conductivity. When a donor impurity is introduced into the crystal lattice of a
semiconductor, the concentration of electrons increases and the semiconductor is then n-type.
Introduction of an acceptor impurity in the crystal lattice of a semiconductor leads to an increase
in hole concentration which is known as p-type doping. When a semiconductor can be doped
into a p- and n-type, the creation of p-n junctions becomes possible.
Natively, ZnO is an n-type semiconductor even in the absence of intentional doping and as
such n-type doping of ZnO is easy to achieve (Janotti and van de Walle (2009)). The origin
of the native n-type of ZnO is still unclear. Özgür et al. (2005) in their review reported that
the origin of residual n-type in undoped ZnO can be partly attributed to the intrinsic defects
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 2. LITERATURE REVIEW 13
in ZnO from zinc and oxygen vacancies as well as interstitials in the crystal lattice. Others
such as Janotti and van de Walle (2009) have suggested that the origin of the native n-type
character of ZnO is due to unintentional incorporation of impurities such as hydrogen, pointing
out that the attribution of n-type ZnO to point defects is in contradiction to some experimental
results. Look et al. (2002) also suggested that the n-type conductivity is due to zinc intersti-
tials rather than oxygen vacancies, because the zinc interstitial is the dominant shallow donor
in ZnO. McCluskey and Jokela (2009), have also proposed that impurities such as hydrogen,
aluminium, boron, gallium, nitrogen, e.t.c., are mostly likely to be the source of the intrinsic
n-type conductivity in ZnO.
As a result of the native n-type character of ZnO, there have been difficulties in obtaining
a p-type doped ZnO despite much effort. For example, there have been some reports on suc-
cessful p-type doping (Nayak et al. (2009), Maksimov (2010)). Maksimov (2010) report recent
advances in the p-type doping of ZnO, noting that the difficulties with acheiving high level p-
type doping can be attributed to large acceptor activation energy, solubility of dopants (mainly
group I and V elements), and difficulty in overcoming the background n-type conductivity.
Doping ZnO with group I elements (Li, K, Na) has been predicted to result in p-type doped
ZnO because these elements have the ability to be substituted on the Zn-site (Özgür et al. (2005),
Janotti and van de Walle (2009)). This prediction has however not been experimentally con-
firmed (Özgür et al. (2005), ?). Another group of elements that were predicted to yield p-type
ZnO by substituting the oxygen are those in group V (nitrogen, phosphorus, arsenic and anti-
mony) (Özgür et al. (2005), Janotti and van de Walle (2009), Wang and Zunger (2003)). How-
ever, group V elements, with the exception of nitrogen, have larger bond lengths which leads
to the formation of anti-sites (Özgür et al. (2005)). The atomic size and electronic structure of
nitrogen makes it the best group V candidate for the substitution of oxygen for p-type doping
(Janotti and van de Walle (2009)). However, nitrogen appears to be insoluble in ZnO, which
difficulty researchers have tried to overcome using the method of ion implantation (Özgür et al.
(2005)).
In a first principle study using ab initio calculations, Huang et al. (2009) found that lithium,
sodium, and silver interstitials were stable in ZnO. The calculations were carried out with DFT
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 2. LITERATURE REVIEW 14
using the plane wave Vienna ab-initio simulating package (VASP).
Though there have been significant advances, full control over ZnO conductivity is yet to be
achieved and as such a comprehensive investigation of the fundamental properties of acceptors
in ZnO has been recommended (Gandomani et al. (2014)).
2.4.1 Transition Metal Doping
Apart from the creation of n- and p-type ZnO by doping, doping with transition metal TM)
ions has been predicted to yield dilute magnetic semiconductor DMS) materials. The predic-
tion of ferromagnetic properties by Dietl et al. (2000) in wide band gap semiconductors doped
with magnetic ions has resulted in increased research interest. Wide band gap semiconductors
such as ZnO are expected to be efficient DMS materials with curie temperatures above 300 K
(Pearton et al. (2007)) hence making their incorporation in electronic devices possible. Doping
ZnO with TMs ions such as cobalt, manganese, iron, nickel, titanium, vanadium and tin have
been realized by various synthesis methods (Ivill et al. (2008), Santos and Macedo (2012), Kar-
makar et al. (2012), Panigraphy et al. (2012), Saravanan et al. (2011), (Pearton et al. (2007)).
Room temperature ferromagnetism has been reported in cobalt- and manganese-doped zinc ox-
ide (Klingshirn (2007), Ivill et al. (2008), Panigraphy et al. (2012), Karmakar et al. (2012)).
However the results were not conclusive because the observed ferromagnetism could have been
as a result of the presence of cobalt clusters, zinc-cobalt complexes, grain boundaries, intrinsic
defects (McCluskey and Jokela (2009)) or the formation of Mn3O4 (Kim et al. (2004)).
Reports on the magnetic properties of transition metal doping often present conflicting
results, with antiferromagnetism, ferromagnetism, paramagnetism, superparamagnetism and
spin-glass all having been reported (Pearton et al. (2007), Ivill et al. (2008), Santos and Macedo
(2012) and Hao et al. (2012)).
There have been reports that Fe-doped ZnO has paramagnetic, antiferromagnetic and spin-
glass properties at 300 K (Santos and Macedo (2012). Wu et al. (2014) reported room temper-
ature ferromagnetism in Fe-doped ZnO synthesised using the hydrothermal method. Pearton
et al. (2007) also reported super-paramagnetic and ferromagnetic properties in Ni-doped ZnO.
Doping ZnO with transition metals such as Fe, Mn, Ni and Co has has been shown to be
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 2. LITERATURE REVIEW 15
viable (Klingshirn (2007) and Pearton et al. (2007)). However, the Mn2+ ion has been identified
as a candidate of choice for its ferromagnetic properties for DMS (Özgür et al. (2005)). This is
because the exactly half-filled 3d orbitals makes all of their five spins parallel by Hund’s rule.
As a result of the parallel alignment of their spins, a considerable amount of energy is required
to add an electron with opposite spin hence causing Mn2+ to have a complete 3d5 orbit leading
to a relatively large magnetic moment (Özgür et al. (2005)).
Transition metal doping can also cause changes in the bandgap energy of ZnO (Bhat and
Deepak (2005) and Karmakar et al. (2007). For instance Silambarasan et al. (2015) observed
an increase in the band gap energy for Fe-doped ZnO. Increases in the band gap energy have
also been reported for Ni-doped ZnO (Thakur et al. (2013)). There are however other reports by
researchers of the reduction in the band gap energy of transition metal doped ZnO. For example,
Bhat and Deepak (2005) and Ebrahimizadeh et al. (2010) observed a reduction in the band gap
energy for Mn-doped ZnO with increasing doping concentrations. He et al. (2013) and Iqbal
et al. (2014) also reported a reduction in the band gap energy for Co-doped ZnO.
In the literature (see, for example Kumar et al. (2011)), the reduction in band gap energy
in doped ZnO is explained by changes in the crystal field induced by exchange interactions
between the sp electrons of ZnO and the d electrons of the dopant ion (Özgür et al. (2005),
Mang et al. (1995)), leading to a narrowing of the band gap. This phenomenon is termed ‘sp-d
exchange interaction’.
On the other hand, an increase in band gap energies can be explained by the ‘Burstein-Moss
band filling effect’, which occurs at high electron concentrations as a result of doping, and
pushes the Fermi level to higher energies. ZnO exhibits this phenomenom, with the Fermi level
rising into the conduction band Kumar et al. (2011)) resulting in an apparent increase in the
band gap.
Therefore, it is possible to observe either narrowing or broadening of the band gap in doped
semiconductors. It is also possible for both narrowing and broadening of the band gap to occur
in the same semiconductor at different doping concentrations, as has been reported for example
by Bhat and Deepak (2005), who found band gap energy reduction for Mn-doped ZnO at 3 %
doping and an increase at 5 % and 7% doping, and also a reduction in the band gap energy of Ni-
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 2. LITERATURE REVIEW 16
doped ZnO nanoparticles at 1, 3 and 5 %. Likewise, Tan et al. (2014) reported an increase of the
band gap energy with increasing doping concentrations for Mn-doped ZnO nanoparticles, while
Kharroubi et al. (2012) mentioned an increase in the band gap energy with increasing doping
concentrations for Mn-doped ZnO thin films. He et al. (2012), Iqbal et al. (2014) and He et al.
(2013) have also reported narrowing of band gap for Co-doped ZnO nano particles. Sahai et al.
(2014) reported narrowing and broadening of the band gap for Fe-doped ZnO nanostructures at
lower doping and higher doping concentrations respectively.
2.5 Methods of Synthesis
Several methods of synthesis have been employed in the synthesis of ZnO nanostructures. These
include solid, liquid and vapour synthesis methods. Some of the methods are described in the
following sections.
2.5.1 Chemical Vapour Deposition
The chemical vapour deposition (CVD) method employs the deposition of a solid material on a
substrate through a gaseous phase. The synthesis of ZnO nanocrystals via CVD is widely used
because the products are of high purity and quality, with few crystalline defects due to the high
synthesis temperature (usually around 900◦ C).
Precursor source
Reacting chamber
Gas outlet
Substrate
Inert gas
Figure 2.2: A schematic diagram of a chemical vapour deposition reactor
A typical CVD reactor consists of a tube furnace with an aluminium or quartz tube placed in
it, as well as an inlet and outlet for the inflow and exhaust of gas. Figure 2.2 shows a schematic
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 2. LITERATURE REVIEW 17
of a CVD reactor. In the CVD method the reactants are placed on substrates in ceramic boats
inside a tube furnace. The evaporation of the reactants, nucleation, and growth typically occur
within the different temperature zones within the tube. The reactions are often carried out in
different gas atmospheres introduced into the reaction chamber.
CVD methods yield ZnO crystals of varied morphology such as nanowires, nanotubes,
nanorods, tetrapods among others (Zhou et al. (2011), Chang et al. (2004), Liu and Zeng (2009),
Kononenko et al. (2012), Fuxue et al. (2014)). In the synthesis of ZnO, zinc ions are evaporated
from a precursor in the presence of oxygen gas and subsequently condensed into ZnO nano-
materials. CVD requires a close control on synthesis parameters such as temperature variation,
choice of substrate, carrier and reactant gas flow to ensure its success.
2.5.2 Ball Milling
The ball milling method is a high energy solid state synthesis method which involves the
mechanical milling of larger particles into smaller ones. It is a top-down method where the
nanocrystalline material is obtained from the bulk material through milling and can generate a
large yield of nanomaterial in short periods of time.
A typical ball mill consists of stainless steel or ceramic balls confined in a turn disc with
two or four bowls as shown in Figure 2.3. Rotation of the turn disc results in collisions which
crush the sample placed in the mill.
Some major disadvantages of this method are sample contamination and crystal damage
(Cao (2007)).
2.5.3 Electrochemical Synthesis
The electrochemical or electro-deposition method involves the deposition of ions on electrodes,
instantiated by the flow of current in an electrolytic cell. In this method, a voltage applied
across two electrodes in the cell drives the required chemical reaction. Electrochemical synthe-
sis are carried out at temperatures of 100 ◦C and below, which makes it relatively cheap and
straightforward method.
Figure 2.4 shows a set-up for the synthesis of ZnO. Zinc rods are used as the electrodes
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 2. LITERATURE REVIEW 18
Figure 2.3: A schematic of a ball milling reactor.
Credit: W. Cao, http://www.ssnano.com
Figure 2.4: A schematic of an electrochemical reactor
with the electrolyte being a salt solution through which oxygen is bubbled. ZnO nanowires
and nanorods grown on conducting substrates such as Zn foils have been reported (Ding et al.
(2013)). Jiangfeng et al. (2010) have also reported the synthesis of ZnO by the eletrochemical
method with gold coated silicon, indium-doped-titanium oxide and copper strip as substrates,
and have shown the ability to use variations in reactant concentrations to control the shape of
the products (Jiangfeng et al. (2010)).
2.5.4 Epitaxial Growth
Crystal synthesis by epitaxial growth is a bottom-up method which involves layer-by-layer de-
position of a crystalline material on a substrate. ZnO thin films or layers grown epitaxially have
smaller defect concentration compared to other methods (Janotti and van de Walle (2009)).
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 2. LITERATURE REVIEW 19
Epitaxial growth can be carried out in the solid, liquid or gaseous phase. Solid phase epi-
taxial growth is mostly used for healing damaged crystals. Gaseous phase methods of growth
include vapour phase epitaxy (VPE) and molecular vapour phase epitaxy (MOVEP), also known
as molecular beam epitaxy (MBE).
Several reports have been made regarding the use of epitaxial methods in the synthesis of
ZnO. For example, Zheng et al. (2006) successfully synthesised doped ZnO films on silicon
substrates and Jung et al. (2005) also grew epitaxial ZnO films on a c-Al2O3 substrate. Heinze
et al. (2007) and Ive et al. (2008) reported the successful growth of ZnO layers on GaN(0001)
templates.
2.5.5 Pulsed Laser Deposition (PLD)
The pulsed laser deposition (PLD) method is a gas-phase growth technique in which a high
energy laser pulse is used to evaporate a target material, forming an ‘ablation plume’ that sub-
sequently condenses into a thin crystalline film on a substrate. Figure 2.5 shows a diagram
of a typical PLD apparatus, where it can be seen that the substrate and the target material are
contained in a sealed chamber with the laser beam focused on the target through a window.
Figure 2.5: A schematic of PLD
Credit: chm.bris.ac.uk
For the synthesis of ZnO nanostructures and films, the target is bulk ZnO and the evapora-
tion is carried out in oxygen. PLD is able to deliver high quality crystals within a short time
(Yongning et al. (2007)), and allows control over growth rate, film thickness, and crystal orien-
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 2. LITERATURE REVIEW 20
tation (Sandana et al. (2009)) as well as on the morphology of crystals (Valerini et al. (2009)).
2.5.6 Wet-chemical methods
The wet-chemical method is a bottom-up liquid phase synthesis in which crystalline growth
occurs at temperatures at which the solvent remains a liquid. This is generally less than 100 ◦C.
When water is used, such a method is termed ‘hydrothermal’ and ‘solvothermal’ when other
solvents are used.
There are several reports on the synthesis of ZnO using wet-chemical methods. The general
procedure in these methods involves the dissolution of a zinc source (acetate, chloride, sulphate,
nitrate salts e.t.c.) in a solvent such as water, alcohol or a mixture of water and alcohol (Srinet
et al. (2014)). One of the early pioneers of the wet-chemical synthesis of zinc oxide nanocrystals
in alcohol were Spanhel and Anderson (1991), who used zinc acetate in ethanol to which a
solution of lithium hydroxide in ethanol was added at 0 ◦C to precipitate ZnO. Variations of this
method have been employed in the synthesis of ZnO.
Shafique et al. (2012) used ethanol as solvent in the synthesis of ZnO and cobalt doped
ZnO. In their approach, ZnO was obtained by dissolving ZnCl2 in ethanol and NaOH was used
as precipitant, leading to the formation of Zn(OH)2 which was transformed into ZnO at 200 ◦C.
Silambarasan et al. (2015) have reported the synthesis of ZnO and Fe-doped ZnO by using
a mixed solvent of ethanol and ethylene glycol in which zinc acetate dihydrate was dissolved.
The solution was transferred into a spirit lamp with an absorbent cotton wick, and subsequently
fired until the entire solution was converted into ash powder, which yielded wurtzite ZnO.
Sharma et al. (2006)) report a hydrothermal synthesis of ZnO starting with zinc acetate
dihydrate and oxalic acid dissolved in distilled water. The solution was sealed in a stainless
steel autoclave at 150 kPa for 3 hours. The resulting product was centrifuged, washed in double
distilled water and ethanol, dried at 60 ◦C and finally annealed at 400 ◦C for 2 hours to obtain
pure ZnO.
Chen et al. (2000) have also reported the synthesis of ZnO with ZnCl2 as precursor and
NaOH as precipitant. In their procedure, ZnCl2 and NaOH in the ratio of 1:2 was dissolved in
distilled water to form Zn(OH)2 as precipitate. The precipitate was then dispersed in distilled
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 2. LITERATURE REVIEW 21
water and the pH adjusted between 5–8 using HCl. The dispersion was stirred for 3 hours after
which it was poured into an autoclave and heated to temperatures ranging from 100-220 ◦C. The
obtained product was filtered and dried at room temperature. XRD and TEM analysis showed
wurtzite ZnO with different morphologies.
In this work, the hydrothermal synthesis procedure of ZnO followed after the work of Jyoti
et al. (2013), who in an experiment to study the role of temperature and NaOH concentration on
the synthesis of ZnO nanoparticles, used ZnCl2 as precursor and NaOH at 5 M as precipitant.
The synthesis was carried out in distilled water within a temperature range of 80–100 ◦C and
with varying NaOH concentrations. The obtained powder was washed 5 times in distilled water
and dried at 100 ◦C for 2 hours. The reaction temperature was maintained at 80 ◦C and the
concentration of NaOH varied from 2-10 M. X-ray diffraction analysis showed that the product
was pure ZnO. In addition, the crystallite size calculated using the Scherrer equation showed
an increase in size with increasing temperature and NaOH concentration. Electron microscopy
also showed that the morphology of the ZnO crystals changed with increasing temperature and
NaOH concentration.
The wet-chemical method was chosen for this work because of its numerous advantages,
some of which are as follows: highly crystallized particles can be synthesised with a narrow
grain size distribution; particles with high purity can be realised without high temperature treat-
ments; the morphology and size can be controlled by adjusting the reaction temperature, acidity,
concentration of reactants, and synthesis time (Chen et al. (2000)). Finally, the synthesis method
is relatively cheap and can be carried out rapidly.
University of Ghana  http://ugspace.ug.edu.gh
Chapter 3
Theory
3.1 Computational Studies
One of the challenges in theoretical physics and chemistry is the description of the features of
matter starting from the fundamental building blocks. Since matter is made up of atoms, and
by extension, electrons and nuclei, such a study reduces to a study of interactions at the atomic
level. The physical properties of molecules, proteins, crystalline solids, etc., can in principle be
elucidated by studying the problem at the atomic level.
Understanding the electronic structure of matter helps in the determination of many proper-
ties related to its stability. Properties such as the binding energy, magnetism, conductivity, crys-
tal structure molecular bond lengths and angles can be determined from the electronic structure
of a system. Dynamical properties such as optical excitations, can also be determined from the
electronic structure.
Since matter is made up of electrons and nuclei, a full appreciation of the properties of a
solid, for example, can be made by solving the many-body Schrödinger equation
ĤΨ = EΨ (3.1)
where Ψ is the total wave function of the many-body system, E is an energy eigenvalue and Ĥ
is the total Hamiltonian given by
22
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 3. THEORY 23
Ĥ = T̂e + T̂n + V̂ee + V̂nn + V̂ne (3.2)
where
∑ ℏ2
T̂e = − ∇2
∑ 2m ii iℏ2
T̂ 2n = − ∇
∑ 2M
I
I
I
1 e2
V̂ee =
2 ∑i̸=j ∑|r⃗i − r⃗j|ZIe2
V̂ne = −
∑I i |R⃗I − r⃗i|1 ZIZJe2
V̂nn = (3.3)
2
I≠ J |R⃗I − R⃗J |
T̂e is the sum of the electron kinetic energies, T̂n is the sum of the nuclei kinetic energies,
V̂ee is the sum of the potential energies due to electron-electron repulsion, V̂nn is the sum of the
potential energies from the nuclei interactions and V̂ne is the sum of the potential energies from
the nuclei-electron interactions. The indices i, j and I , J refer to the ith, jth electrons and I ,
J th nuclei respectively, Z is the nuclei charge, and e the electronic charge.
The total Hamiltonian is then
∑ 2 ∑ 2 ∑ 2 ∑∑ 2
Ĥ = − ℏ ∇2 − ℏ ∇2 1 e − ZIe
2m i
+ +
∑ i 2MI
I 2 |r⃗i − r⃗ |i jI i̸=j I i |R⃗I − r⃗i|
1 Z Z 2I Je
+ (3.4)
2
I≠ J |R⃗I − R⃗J |
where R⃗ represents the space coordinates of all the nuclei and r⃗ represents those of the
electrons.
The solution to the many-body Schrödinger equation is made complicated by the large num-
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 3. THEORY 24
ber of interacting particles, and therefore the practical approach to solving the equation is by
using approximations.
3.1.1 Born-Oppenheimer Approximation
The Born-Oppenheimer approximation, also known as the adiabatic approximation, is based on
the fact that the electrons can be considered to be mobile in a field of stationary nuclei. This is
because the mass of a nucleus is much larger (by a factor of about 2000) than the mass of an
electron.
This approximation makes it possible to split the many-body Schrödinger equation (Equa-
tion 3.4) into two parts which consists of the electron and nuclei as separate entities (White
(1934), Bassani and Parraviccini (1975)). The wave function can then be written as the product
of an ‘electronic’ wavefunction ψ(R⃗, r⃗) and a ‘nuclear’ wavefunction ϕ(R⃗),
Ψ(R⃗, r⃗) = ψ(R⃗, r⃗)ϕ(R⃗) (3.5)
allowing the total Hamiltonian to be split into two as follows:
Ĥe = T̂e + V̂ee + V̂ne (3.6)
Ĥn = T̂n + V̂nn (3.7)
The many-body Schrödinger equation of the electron is then,
Ĥeψ(R⃗, r⃗) = Eeψ(R⃗, r⃗) (3.8)
Even with the Born-Oppenheimer approximation, the solution to the many-body Schrödinger
equation remains difficult because of the electron-electron and electron-nuclear interactions
which have to be summed over N particles, leading to equations of 3N variables to be solved
simultaneously.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 3. THEORY 25
Hartree-Fock Approximation
The next approximation considered here is the Hartree-Fock (HF) approximation. This ap-
proximation was first proposed by Hartree in which the many-body wave-function is taken as
a product of one-electron wavefunctions, so that for a system of N particles, the ground state
wave function can be written as
Ψo(r⃗N , σN) = ψ1(r⃗1, σ1)ψ2(r⃗2, σ2)...ψN(r⃗N , σN) (3.9)
where r⃗i is the spatial co-ordinate of the electron and σi is the spin co-ordinate of the electron.
For simplicity, ψ(R⃗, r⃗) in Equation 3.8 is written as ψ(r⃗, σ).
The Hartree equation suffers the difficulty of being unable to account for the anti-symmetric
property of the electron. In the original Hartree equation, an exchange of any two spatial coordi-
nates of the electron, returns the same wave function which is a violation of the anti-symmetric
property of the wave function.
The HF approximation resolves this by the inclusion of an anti-symmetric term such that
the ground state wave function of a system of N -electrons is given by,
Ψo(r⃗N , σN) = A (ψ1(r⃗1, σ1)ψ2(r⃗2, σ2)...ψN(r⃗N , σN)) (3.10)
where A is the anti-symmetric operator
∑N !1
A = √ (−1)piPi (3.11)
N !
i
where (−1)pi is −1 or +1 for permutations of odd or even class respectively. The sum extends
to allN ! permutations Pi of the electronic coordinates (Grosso and Parravicini (2013)). The HF
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 3. THEORY 26
approximation replaces the ground state wave function with a Slater determinant
∣∣∣∣∣
∣∣
∣ ψ1(r⃗1σ1) ψ1(r⃗2σ2) · · · ψ1(r⃗NσN)∣ ∣
∣∣
1 ψ2(r⃗1σ1) ψ2(r⃗2σ2) · · · ψ2(r⃗ ∣NσN) ∣
Ψo(r̃1σ1, r̃2σ2, ...̃rNσN) = √ ∣ . . . . ∣N !
∣∣∣∣
∣ .. .. .. .. ∣∣∣∣
ψN(r⃗1σ1) ψN(r⃗2σ2) · · · ψN(r⃗NσN) ∣
giving the equation
ℏ2∇2 ∑ 2 ∑ ∫− − ZIe 2 ∗ |ψj(r⃗ σ 22 2)|ψj ψj + e ψ (r⃗σ) d(r⃗2σ2)−
2me
I |r⃗ − R⃗I |
i
∫ |r⃗ − r⃗∑ j 2
|
(occ)
ψ∗
(occ)
j (r⃗2σ2)ψi(r⃗2σ ) ∑
e2
2
ψj(r⃗σ) d(r⃗ σ ) = ε ψ| − 2 2 ij jr⃗ r⃗
j 2
|
j
The third term and the fourth terms on the LHS of the equation represents the Coulomb
potential and exchange correlation potential respectively.
The HF approximation has been used extensively in the computational study of solids such
as the investigation of surfaces (Jaffe and Hess (1993), George Kresse and Diebold (2003)),
elastic properties (Ahuja et al. (1998)), phonon dispersion (Serrano et al. (2004)), point defects
(Kohan et al. (2000)) as well as high-pressure phase transitions (Ahuja et al. (1998)).
However, the HF approximation does not work well in solving some problems. For a total
ground state energy of atoms such as the free electron gas, the HF approximation does not yield
reliable results (Bassani and Parraviccini (1975)). Corrections to the approximation are made
by using the screened Hartree-Fock approximation in which the Coulomb interaction remains
unchanged whiles the exchange interaction is screened with a desirable dielectric function (Car-
dona (1969)), leading to better results.
3.1.2 Density Functional Theory
The density functional theory is based on two theorems proved by Hohenberg and Kohn in 1964
(Walkosz (2011)):
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 3. THEORY 27
1. All the ground state properties of a many-body system can be uniquely determined by the
ground state charge density, n(r⃗).
2. If there exists a universal functional F [n(r⃗)] for the energy in terms of the charge density
such that E[n(r⃗)] is valid for any external potential Vext(r⃗), then for a given Vext(r⃗), the
charge density that minimizes the functional is the ground state density.
These theorems allowed the development of a one-to-one correspondence between n(r⃗) and
the external potential which made it possible to express n(r⃗) as a universal functional F [n(r⃗)].
With the universal functional, the existence of a functional energyE[n(r⃗)]was evident since the
Hamiltonian is fully determined (Walkosz (2011)). The functional energy assumes its minimum
value for the correct ground state density. The functional energy can be expressed in terms of
the universal functional and an external potential as,
∫
E[n(r⃗)] = Vext(r⃗, n(r⃗)dr + F [n(r⃗)] (3.12)
The ground state energy and charge density can be obtained by minimizing the functional en-
ergy, as was done by Kohn and Sham (1965) by mapping the interacting problem of Equation
3.12 with an auxiliary non-interacting equation such that the ground state charge density is
given by,
(∑occ)
n(r⃗) = |ψi(r⃗)|2 (3.13)
i
where ψi is a set of single-particle wavefunctions for the non-interacting electron gas with
charge density n(r⃗), summed over all the occupied single states. With this new ground state
charge density, the functional, F [n(r⃗)] is given by,
F [n(r⃗)] = Ekin[n(r⃗)] + EH [n(r⃗)] + Eexch[n(r⃗)] (3.14)
where the first term is the kinetic energy of non-interacting electron gas at n(r⃗), last term de-
notes the quantum mechanical exchange-correlation energy and the second term, EH [n(r⃗)] is a
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 3. THEORY 28
HF term which denotes the Coulomb interactions defined as,
2 ∫e n(r⃗)n(r⃗′)
EH [n(r⃗)] = drdr
′ (3.15)
2 |r⃗ − r⃗′|
The total energy of the Kohn-Sham equations which allows for the practical application of
DFT, can thus be written as,
∫
E[n(r⃗)]KS = Ekin[n(r⃗)] + EH [n(r⃗)] + n(r⃗)Vext(r⃗)dr + Eexch[n(r⃗)] (3.16)
The fourth term on the RHS of the equation is to account for the differences between the
actual interacting and non-interacting Kohn-Sham system (Kohn and Sham (1965)).
This approximation approach to DFT succeeded mainly because the contribution to the
many-body problem contained in Eexch[n(r⃗)] is a minimal fraction of the total energy (Walkosz
(2011)).
The Hamiltonian needed to solve the many-body problem of DFT is therefore given by
ℏ2
Ĥ 2DFT = − ∇r⃗ + veff (r⃗, n(r⃗)) (3.17)2me
where veff (r⃗, n(r⃗)) is the effective potential given by,
∫
2 n(r⃗
′) ′ δEexch[n(r⃗)]veff (r⃗, n(r⃗)) = Vext(r⃗) + e dr + (3.18)|r⃗ − r⃗′| δn(r⃗)
Hence the Schrödinger equation for the many-body problem in DFT is given by
ĤDFTψi(r⃗) = εiψi(r⃗) (3.19)
DFT calculations make use of an assumed density to obtain the effective potential which is
used to solve the Kohn-Sham equations. The effective potential is then used to calculate the
density of the system of interest using Equation 3.13. An iterative routine is then carried out to
obtain self consistency and to compute the total energy E[n(r⃗)]KS . Figure 3.1 is a flowchart of
the procedure for the calculation of total energy from DFT.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 3. THEORY 29
In essence, DFT reduces the N particle system of interacting particles with 3N degrees of
freedom, to a problem which employs the density functional of 3 variables (Walkosz (2011)).
Several improvements have been made to DFT since the work of Hohenberg-Kohn and
Kohn-Sham to make it an accurate and reliable tool for computational quantum chemistry and
solid state physics.
Exchange and correlation functionals
In principle, DFT would be exact if the exact form of the exchange correlation functional were
known. However, this is not the case in practice hence, there is the need to rely on approxi-
mations which relate the exchange correlation energy of the Kohn-Sham electrons to the real
electrons. The most widely used approximations in DFT are LDA and GGA.
LDA is an exchange correlation functional which relates locally, the exchange correlation
energy of a non-uniform system to that of a uniform electron gas with the same density n(r⃗) as,
∫
ELDA[n(r⃗)] = ϵLDAexch exch (n⃗(r))d
3r (3.20)
where ϵLDAexch (n⃗(r)) is the exchange-correlation energy density of the uniform electron gas.
LDA uses the ‘ultra-soft’ psuedopotential which allows the treatment of the electrons in the p,
d and s- orbitals as valence electrons.
Calculations of structural parameters (lattice parameters and atomic positions) with DFT-
LDA are quite accurate within some few percent of the experimental values (Payne et al. (1992),
Janotti and van de Walle (2007)). With the application of boundary conditions such as the
use of supercells with periodic boundary conditions, LDA has also been employed in DFT to
investigate defects in semiconductors (Kohan et al. (2000), Janotti and van de Walle (2007)).
The calculations of equilibrium concentrations have also been carried out with DFT-LDA (Van
de Walle and Neugebauer (2004)).
However, LDA tends to underestimate the bond and lattice distances as a result of overbind-
ing in molecules and solids (Laaksonen (2009)). In order to obtain a more precise exchange-
correlation functional, the addition of the gradient of the electron density is used, which then
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 3. THEORY 30
leads to the GGA method. The local exchange-correlation energy of the non-uniform system
which relates to that of the uniform gas in GGA is given by
∫
EGGA[n(r⃗)] = ϵGGA(n(r⃗),∇(nr⃗))d3exch exch r (3.21)
where EGGAexch [n(r⃗)] is the exchange-correlation energy density of the uniform electron gas and
∇(nr⃗) is the gradient of the density.
There are several forms of GGA, notable among which is the PBE. The PBE improves the
under estimation of the bond and lattice distances caused by LDA. However, PBE is gener-
ally not effective for solids and also has the problem of over correcting the ground state and
dissociation energies and the bond lengths (Perdew et al. (1992)).
As mentioned above, DFT is known to underestimate the band gap energy with LDA and
GGA. Both approximations are exchange-correlation functionals which have an incomplete
cancellation of the self interactions of the electrons. Additionally, the binding energy of the 3d
core electrons are underestimated by LDA and GGA (Janotti and van de Walle (2007)). Hence,
with regards to returning exact and experimentally confirmed values, LDA and GGA are not
good approximations despite their ability to accurately determine structural parameters.
3.2 Quantum-ESPRESSO
The computational work in this reasearch makes use of the Quantum ESPRESSO (QE) pack-
age. QE is a full ab-initio multi-purpose and multi-platform package which implements the
computational calculations of electronic structure and energy, linear responses in solids and
third order anharmonic perturbation theory. It has been employed for ab-initio calculations of
periodic and disordered systems (Scandolo et al. (2005)). The codes of the QE package makes
use of DFT, plane wave (PW) and pseudopontential (PP) description of the electronic ground
state. Ideally, it is suitable for structural optimizations at both zero and infinite temperatures.
The package can also calculate both the ground state energy and Kohn-Sham orbitals for insu-
lators, metals and many exchange correlation functions. The three main components of the QE
are plane wave self consistent field (PWSCF) (Baroni et al. (2001)), Car-Parrinello molecular
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 3. THEORY 31
dynamics (CPMD) (Pasquarello et al. (1992)), and first principle molecular dynamics (FPMD)
(Cavazzoni and Chiarotti (1999)). The CPMD and FPMD consist of molecular dynamic codes
whiles the PWSCF contains codes for the computations of the total energy and electronic struc-
ture of systems. The PWSCF uses both norm-conserving pseudopotentials (PP) and ultra-soft
pseudopotentials (US-PP) within DFT.
In this research, the QE package will be used to study the energy levels of undoped and
transition metal (Fe, Mn, Co and Ni) doped ZnO. The results from the energy levels will be
used in computations that will lead to plots of the band structures and density of states. The
lattice parameters of the undoped and doped samples will also be calculated from this study.
In addition to the elements of theory already discussed, the basic theory behind the use of
plane waves in QE comes from the Bloch function. In general, the electronic wave function in a
periodic crystal, Ψnk⃗(r⃗), can be written in terms of periodic Bloch functions unk⃗(r⃗) as follows:
Ψ ik⃗.r⃗nk⃗(r⃗) = e unk⃗(r⃗) (3.22)
with
∑
u (iG⃗.r⃗)nk⃗(r⃗) = cGe (3.23)
G
cG are constant coefficients of expansion and G⃗ is a vector that sums over all the reciprocal vec-
tors. This wavefunction is expanded using a plane wave basis which are summed over infinity.
Computationally, summing over infinity is not possible, so the plane wave basis is summed to
the limit of a finite number of G⃗ vectors. The G⃗ vectors are truncated with an appropriate finite
number known as the ‘plane wave cutoff’ parameter. Aside from the G⃗ vectors, the other im-
portant parameters in this approach are the ‘energy cutoff’ for the wavefunction and the number
of ‘k-points’. The ‘k-points’ are discretizations of k-space which allows for the definition of
an energy E(k⃗) at each k-point. Computationally, it is beneficial to select a minimum number
of k-points required to calculate the needed parameters. This will reduce the time and com-
puting resources required. There are methods of generating optimal k-point grids, such as the
Monkhorst-Pack mesh.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 3. THEORY 32
Near a nuclear core, a large number of plane waves is required to describe unk⃗(r⃗). To
minimize the computational resources required by having a large number of plane waves, QE
uses the method of truncating the electronic wave functions at a given energy value. This is the
‘energy cutoff’ parameter.
With these values properly identified, in principle the equations can be solved iteratively
until a self consistency is achieved through the diagonalization of an N × N matrix (N is the
number of basis functions in the Brillouin zone).
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 3. THEORY 33
Guess initial density 
n(r)
No
Yes
Figure 3.1: A flowchart of the procedure for the calculation of total energy from DFT
University of Ghana  http://ugspace.ug.edu.gh
Chapter 4
Method
4.1 Synthesis Procedure
Zinc chloride (ZnCl2), and the chlorides of cobalt (Co), manganese (Mn), nickel (Ni) and iron
(Fe) were used as precursors for the synthesis of undoped and doped ZnO nanoparticles with
potassium hydroxide (KOH) as precipitant. The reactions were carried out in distilled water.
ZnCl2 (97 % pure) was obtained from Central Drug House (CDH), MnCl2· 4H2O (98 % pure)
from Daejung, CoCl2· 6H2O (97 % pure) from Tianjin Fuchan Chemicals, FeCl2·4H2O (98
% pure) from Xilong Chemical , NiCl2· 6H2O (97% pure) from Research-Lab Fine Chem.
Industries . KOH (85 % pure) and ZnO (99 %) were obtained from CDH. The ZnO was used as
a standard. These chemicals were all used as supplied without further treatment.
For the synthesis of undoped ZnO, 1.5 g of ZnCl2 was crushed into fine powder in a porce-
lain mortar and transferred into a beaker. 40 ml distilled water was then added, following which
the solution was heated to 85◦C under constant stirring with a magnetic bead stirrer. The pH
of the solution was determined using a litmus paper which recorded the pH to be 5. Next, 2.5
g KOH was dissolved in 20 ml distilled water and added drop-wise to the ZnCl2 solution to
cause precipitation. This was done at 85 ◦C under constant magnetic stirring. The precipitate
was left to settle at room temperature. Following this, the supernatant was poured off and the
precipitate, washed five times in distilled water. The precipitate was then dried in air at 100 ◦C
for 12 hours. The resulting crystals were crushed into fine powder and annealed at 280 ◦C for 3
hours.
34
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 4. METHOD 35
The doped samples were obtained by adding the required molar percentages of Fe and Mn
(for 1, 2, 4 and 8 mol % doping) to ZnCl2 and crushed in a porcelain mortar. The rest of
the procedure followed as already described. Ni and Co doped ZnO were synthesised by first
dissolving 1.5 g ZnCl2 in 60 ml distilled water and heating to 100 ◦C under constant magnetic
stirring. The required molar percentages (1, 2, 4 and 8 mol %) of Ni and Co dissolved in 20 ml
distilled water, were added drop-wise to the ZnCl2 solution under constant magnetic stirring.
The rest of the procedure followed as described for the synthesis of the undoped ZnO. The 4
and 8 mol % transition metal doped ZnO at pH 5 after annealing at 280 ◦C for 3 hours, were
annealed again at 600 ◦C in air for 1 hour.
Synthesis of 4 and 8 mol % doping was also carried out at pH 3 by adding a few drops of
25% HCl. The synthesis procedure employed for the Ni and Co doped ZnO was used for the 4
and 8 mol % transition metal doped ZnO at pH 3. The 4 and 8 mol % transition metal doped
ZnO samples at pH 3 after annealing at 280 ◦C for 3 hours, were again annealed at 600 ◦C in
air for 1 hour. The electronic and optical properties of the as prepared samples were determined
using the x-ray powder diffractometer, UV-VIS absorption and photoluminescence emission
spectrometer.
4.2 Characterisation
The samples were characterised using x-ray diffraction, UV-VIS absorption spectroscopy and
photoluminescence emission spectroscopy to determine the crystal parameters and the band gap
energies of the as-prepared samples.
4.2.1 Equipment
UV-VIS Absorption Spectroscopy
UV-VIS absorption spectroscopy refers to absorption spectroscopy or reflectance spectroscopy
in the ultraviolet–visible spectral region. In UV-VIS absorption spectroscopy, the absorption co-
efficient α is obtained experimentally with an absorption spectrometer. Two lamps are used in a
typical spectrometer in order to extend the range of the spectrum. Usually, there is a deuterium
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 4. METHOD 36
lamp which provides ultra-violet (UV) radiation and a tungsten lamp which provides visible
and near infra-red (IR) radiation. At a set wavelength, the optics of the spectrometer are made
to switch from one lamp to the other.
Figure 4.1 is a schematic of the UV-VIS spectrometer. A single wavelength selected by the
monochromator is incident on the sample to be analysed. The wavelength is scanned from long
wavelengths to the short wavelengths through the sample. The detector collects the radiation
transmitted through the sample. The data from the detector can be visualized on a computer for
further analysis such as band gap calculations.
lamp enclosure
B
A
Monochromator
Mirror
Sample
Detector
Figure 4.1: A schematic of a UV-VIS absorption spectrometer.
A and B are deuterium and tungsten lamps respectively.
The UV-VIS absorption spectra for the undoped and doped ZnO were obtained from a Shi-
madzu 1240 UV-Mini UV-VIS absorption spectrometer equipped with deuterium and tungsten
lamps, with the lamp switching wavelength at 340 nm.
The scan was done from 230–800 nm. The band gap energy of the as-prepared samples
were determined from absorption edge of the spectra using the well known Tauc equation for
estimating direct band gap transition (Viezbicke et al. (2015)) given by,
1
(αhν)n = A(hν − Eg) (4.1)
where α is the absorption co-efficient, h is Planck’s constant, ν is the frequency, n = 1 (for
2
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 4. METHOD 37
direct allowed band gap), A is a constant and Eg is the optical band gap.
The method for calculating the band gap involves projecting the steepest end of the absorp-
tion curve to cut the axis at α = 0 as shown in Figure 4.2.
Figure 4.2: UV-spectrum of a ZnO sample.
The band gap is determined using the Tauc method by finding the value of hν at which α2 = 0.
Photoluminescence Emission Spectroscopy
Electrons that occupy energy levels in the valence band of a semiconducting material can be
excited to vacant energy levels in the conduction band if they receive energy from external ex-
citations in excess of the band gap energy. The excitation source could be from electromagnetic
radiation, an electric field, electron bombardment, and so on. Electronic transitions from the
valence band to the conduction band subsequently generates an electron-hole pair (EHPs) (i.e.,
a hole in the valence band and an electron in the conduction band). This transition causes the
material to be in an excited state and for equilibrium to be re-established, de-excitation occurs.
The de-excitations occur by the re-combination of EHPs which leads to the emission of photons.
However, if the de-excitation goes through a non-radiative path, there will be no photon emis-
sion. The de-excitation of an excited electron from the conduction to the valence band results
in the release of a photon of energy equivalent to the band gap energy (Kuzmany (2008)).
The phenomenon of emission of light by matter is termed luminescence. For solids in which
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 4. METHOD 38
charge carriers excited by an external light source leads to the subsequent emission of photons
by de-excitation, the phenomenon is called ‘photoluminescence’.
In photoluminescence spectroscopy for semiconductor materials, a beam of monochromatic
radiation from light source is used to excite the sample. Typically, the excitation energy should
be slightly greater than the band gap energy of the semiconductor so that some electrons in
the valence band, will receive enough energy to cross to the conduction band. The excited
electrons subsequently de-excite by emitting photons, which are then detected by an appropriate
device such as a photomultiplier tube. The output spectrum is then plotted and visualized on a
computer. Figure 4.3 is a schematic diagram of a photoluminescence spectrometer.
Figure 4.3: A schematic diagram of a photoluminescence spectrometer
Photoluminescence emission spectroscopy provides important information about the elec-
tronic structure of a semiconductor with regards to band gap energy and inter-band energy tran-
sition states. In Figure 4.4, a schematic of the band gap picture of a semiconductor is shown.
As is well known from basic solid state physics, electrons present in the valence band of a semi-
conductor can be excited across the band gap when they receive sufficient energy, which in the
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 4. METHOD 39
case of photoluminescence is from incident photons. Subsequent de-excitations of the electrons
returning to the valence band leads to emission of photons. De-excitations occurring directly
across the band gap lead to the emission of photons whose energy is close to the band gap en-
ergy. This emission is called the ‘near-band edge emission’ and may be assisted by phonons.
Impurities and defects can also lead to the formation of energy states within the band gap, called
inter-band states. In such cases, excitations and de-excitations can occur from the valence band
to impurity states, and likewise from the impurity states to the conduction band. These transi-
tions will lead to the emission of photons whose energy is much less than the band gap energy.
These are called deep-level emissions.
Figure 4.4: A schematic of band transitions in a semiconductor
. The pink and blue circles depict holes and electrons respectively
The electrons and holes in Figure 4.4 are also known as donor and acceptor states respec-
tively. The transitions as depicted are: a is known as band gap transition, b is a conduction band
to an acceptor state, c is a donor state to a valence band and d is known as a donor to an acceptor
state.
Photoluminescence emission measurements were carried out using the Tecan Infinite 200Pro.
The excitation wavelength was 320 nm and the emission was measured from 370 to 600 nm in
steps of 2 nm. The integration time was 2000 µs.
The samples were dispersed in methanol and stirred until completely dispersed using a spat-
ula and placed in a micro plate reader for measurement. The emission spectra were visualized
on a computer and the band gap as well as the inter-band transitions were calculated using the
equation Eg = hc where h = Planck’s constant, c = velocity of light, λ = peak wavelength.λ
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 4. METHOD 40
X-Ray Powder Diffraction
When an x-ray photon is incident on a crystal, reflections from the crystal planes occur because
the wavelength of the radiation is comparable to the spacing between the atoms. The x-rays
are in actual fact reflected as a result of interaction with the electron density distribution within
the crystal. The reflected radiation may interfere destructively or constructively. In a perfect
crystal, where x-rays scatter without loss of energy, incident waves are scattered in a specular
manner (such that the angle of incidence equals the angle of reflection) from parallel planes of
the atom, constructive interference may occur. For x-rays reflecting off equally spaced crystal
planes (distance d apart), constructive interference occurs when the path difference 2d sin θ is
an integral multiple of n of the wavelength λ, which is theBragg’s law, nλ = 2d sin θ. The
planes of atoms in the crystal that are responsible for Bragg reflections are called the Bragg
planes, and are indexed with the Miller indices (hkl). The series of planes corresponding to
n = 1, 2, 3, ... correspond to the first, second, third order Bragg reflections and so on.
In a typical x-ray powder diffractometer, a fine powder of the sample to be analysed is made
compact by compressing it in a powder sample holder. The sample holder with the powder
is then placed in a goniometer that rotates with a constant angular velocity to ensure that the
incident x-ray photons encounter all the possible planes in the powder. Figure 4.5 shows a
schematic of the x-ray powder diffractometer.
Figure 4.5: A schematic of an x-ray powder diffractometer
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 4. METHOD 41
In this work XRD data was taken with an Empyrean Panalytical x-ray powder diffractometer
equipped with a copper (Cu) tube. The x-ray diffraction was carried out with the Cu Kα line
at λ= 1.54 Å with a step size of 0.08◦ and the 2 θ scan range from 20–80◦. A nickel filter was
used to remove the Kβ line. Crystal phase identification was performed using the Panalytical
Highscore software version 3.05. The data obtained from the x-ray diffraction (XRD) was used
to determine the crystal phases and to calculate the lattice parameters of the unit cells. The
lattice parameters were calculated from the well known Bragg equation nλ = 2d sin θ and the
equation for indexing planes for the hexagonal crystal geometry (Mote et al. (2011)) as shown
below,
[ ]
1 4 h2 + hk + k2 l2
= + (4.2)
d2 3 a2 c2
Using the first order approximation n = 1 from the Bragg equation and substituting the result
into Equation 4.2 gives Equation 4.3 which can be used with the appropriate Miller indices
(hkl) to obtain a and c.
[ ]
λ2 4 (a)2
sin2 θ = (h2 + hk + k2) + l2 (4.3)
4a2 3 c
where n is the Bragg plane, λ is the wavelength, d is the inter-planar separation distance, θ is the
Bragg angle of diffraction, hkl are the Miller indices, a and c are the lattice parameters. Using
the lattice parameters a and c, the volume, V of the unit cell of the hexagonal crystal geometry,
for each sample was also calculated from Equation 4.4 (Mote et al. (2011)),
√
3
V = a2c (4.4)
2
The crystallite size and strain were determined using the Williamson Hall equation (Equa-
tion 4.5) calculated using a routine on the Panalytical Highscore Plus (V. 3.05). The instru-
mental broadening profile was determined using a pressed silicon disk standard, and the profile
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 4. METHOD 42
subtracted from the experimentally measured peak profile. The shape factorK was set to 1.0.
Kλ
βT cos θ = + (4ϵ) sin θ (4.5)
D
where D is the diameter of a (quasi) spherical crystallite, λ is the x-ray wavelength (here, Cu-
Kα at 1.54 Å), βT is the corrected full-width at half-maximum (FWHM) of a selected peak, and
θ the Bragg angle.
The FWHM was corrected with a quadratic broadening correction,
β2 2 2T = [βmeasured − βinstrumental] (4.6)
where βmeasured is the measured FWHM and βinstrumental is the FWHM due to the
instrumental profile.
4.3 Computational Routine
Quantum ESPRESSO
The expected electronic parameters from the Quantum ESPRESSO package are the band gap
energy, the lattice parameters, the band structure and the density of state plots of undoped and
transition metal (Fe, Mn, Co and Ni) doped ZnO. A unit cell containing 4 atoms was used for
the calculations of the parameters for the undoped ZnO. A ZnO supercell containing 36 atoms
(18 atoms each of zinc and oxygen) was used for the doped samples. The PBE of the GGA
exchange correlation functional was used.
Doping was done at 5.56 at. % and 11.11 at. % by the substitution of Zn atoms. In the
5.56 at. % doping, one Zn atom was substituted with the dopants at the (0.00, 0.667, 0.9829)
position. The 11.11 at. % doping was achieved by replacing an additional Zn atom at (0.444,
0.556, 0.483). A gaussian smearing with 0.01 width was used for both the undoped unit cell
and doped supercells. The supercell is shown in Figure 4.6.
The computation of lattice parameters, band structure and density of states started with
a series of optimization procedures for the undoped cell and doped supercells. The process
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 4. METHOD 43
OXYGEN
ZINC
Figure 4.6: ZnO supercell containing 36 atoms
started with the convergence test for the cutoff energy (ecut) and the k-point.
Optimization of Input Parameters
A ‘self consistent field’ (scf) calculation was first carried out, by testing a range of values for
‘ecut’ to select an optimal value. This was done using the pw.x code with ‘ecut’ values from 10
Ry to 100 Ry in steps of 10 Ry. The optimal value of ‘ecut’ selected was 50 Ry.
The ‘scf’ calculation was also carried out with the pw.x code to obtain the minimum num-
ber of k-points. The range of values for the convergence test of k-points were in the format,
kx, kx, kx − 2 from kx = 4 to kx = 16 Convergence was achieved at the k-points (8, 8, 6). The
optimized k-points and the ‘ecut’ values were then used as part of the input for the ‘relaxation’
routine.
Relaxation
The ‘ecut’ and the k-points obtained from the convergence test were used as part of the input file
with assumed values of the lattice pararmeters. By setting ‘calculations’ = ‘vc-relax’, the pw.x
code was used in the relaxation of the system. From the output of the relaxation, the atomic
positions were extracted and used in the input file.
The lattice parameters a and c were also calculated from the output of the relaxation. All
the calculated parameters were then used to create an input file for the subsequent calculations.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 4. METHOD 44
Band Structure
The optimized parameters earlier computed (‘ecut’, k-points and lattice parameters a and c) as
well as the atomic positions from the output of the relaxation, were used in an ‘scf’ calculation.
The number of electrons of the core the valence shell were obtained from the output of the scf
calculation based on the pseudo potential used for the calculation.
Calculations for the band structure preceded by modifying the input file from the ‘scf’ cal-
culation. In this work, extra bands were calculated so the number of bands used was 240. A 501
k-point grid was obtained using the high symmetry path (A, L, M, Γ, A, H, K, Γ) of the Bril-
louin zone for the hexagonal unit cell. The new input file was generated and the band structure
calculations were run using the pw.x code. A post processing code was used for the plotting of
the band structure.
Density of States
The density of states (DOS) calculation was performed by running a non-self consistent field
(‘nscf’) calculation, following which the ‘dos.x’ code was run reading input generated from the
earlier calculation. The data for plotting the density of state is stored in the dos.dat file.
University of Ghana  http://ugspace.ug.edu.gh
Chapter 5
Results and Discussion
In this chapter, results from the experimental and computational work are analysed and dis-
cussed.
Section 5.0 introduces the relevant parameters to be discussed in this chapter. Section 5.1
and 5.2 deals with the crystallographic parameters and band gap energies for undoped and doped
ZnO powders. These samples were prepared at pH 5 and were all annealed at a temperature of
280 ◦C. Fe-, Mn-, Co-, and Mn- doping were carried out at 1, 2, 4, and 8 molar percentages.
Band gap energies calculated from the UV-VIS absorption and PL spectra are presented and
discussed.
In order to confirm that synthesis at elevated pH delivers better crystalline products, the
powders were synthesised at pH 3 (instead of pH 5 as done earlier) at doping concentrations of
4 and 8 mol %. The powders were further annealed at 280 ◦C. Crystallographic parameters and
band gap energies were determined as before. The results are presented in section 5.3.
In section 5.4, results are presented for 4 and 8 mol % doped samples annealed at a tempera-
ture of 600 ◦C. The data from these results are compared to those at 280 ◦C annealing. This was
done to determine the effect of high temperature annealing, which is well known to improve the
crystalline character of the samples.
Section 5.5 discusses the lattice parameters and band gap energies calculated using DFT
implemented in the Quantum ESPRESSO package. The results from the computational and
experimental work are compared.
45
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 46
5.0.1 Crystallographic Parameters
The crystallographic parameters were calculated from the X-ray diffraction data as discussed in
Chapter 4. The parameters of interest extracted for discussion in this section are a, c and the
unit cell volume.
5.0.2 Determination of the Band Gap Energy
The methods for determination of the band gap energies from the UV-VIS absorption edge and
the PL near band edge emission spectra have been described in previous sections. Both methods
were used in this work, but as expected, did not yield exactly the same values for all the samples.
With regards to the band gap energy for the reference ZnO powder, the value from the UV-VIS
calculation was 3.19 ± 0.01 eV and that from the PL calculation 3.23 ± 0.03 eV. These values
are the same, within the range of error. However, for the as-prepared undoped sample, the value
from UV-VIS was 3.10± 0.01 eV while that from the PL calculation was 3.20± 0.03 eV. This
difference was attributed to the nature of the as-prepared samples, which would naturally be of
poorer quality than the reference and behave differently under the different methods. In any
case, comparisons are made with values calculated from the same method, i.e., UV-VIS band
gap values are compared with other UV-VIS band gap values, and PL values with PL values.
5.0.3 Size and Strain by the Williamson Hall Method
As discussed in Chapter 4, the crystallite size and strain were determined using the quadratic
Williamson-Hall (W-H) method. For each sample, a plot was made of β cos θ on the y axis and
sin θ on the x-axis, which yielded the crystallite size from the y-intercept and the strain from
the gradient of the fitted straight line. A representative W-H plot (for 8 mol % Co-doping) is
shown in Figure 5.1. The errors in the values of the size and strain are determined from the
Panalytical Highscore software which was used in the W-H analysis. The average error in the
strain was ±1 × 10−3. The average error in the size for annealing at 280 ◦C was 40 % for all
the doped samples and 36 % for annealing at 600 ◦C, with the exception of the nickel-doped
samples which had an average error of 24 %.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 47
Figure 5.1: Williamson-Hall plot for 2 mol % Co-doped ZnO
5.0.4 Aspect Ratio
The aspect ratio (AR) as defined in this work is that which is commonly used, being the ratio of
the length of the hexagonal crystallite to its width. This ratio can be determined from the x-ray
peaks by noting that the intensity of reflections from any set of crystal planes is representative
of the ‘number’ of planes involved in the diffraction. Therefore, a more intense peak in general
suggests that there are more planes available in the given direction. For ZnO, x-ray diffraction
from the (002) planes are representative of the c direction, while the diffraction from the (100)
planes are representative of the a direction. Therefore the intensity of the peaks can be used as
an indicator of the length of the crystallite along the c axis as compared with the width along
the a axis. The ‘aspect ratio’ then gives an idea of the length versus the width of the crystallite.
A larger aspect ratio suggests a more elongated structure along the c-axis.
5.1 Synthesis at pH 5 and annealing at 280 ◦C
5.1.1 Undoped ZnO
Undoped ZnO powder was prepared using the method outlined in Chapter 4 with the pH of
the solution at pH 5. The resulting powder was annealed at 280 ◦C and characterised by x-ray
powder diffraction.
Figure 5.2 shows the x-ray diffractograms for the undoped ZnO and the ZnO reference.
X-ray diffraction peaks were observed at 2θ angles of 31.8◦, 34.4◦, 36.3◦, 47.6◦, 56.7◦, 62.9◦,
66.4◦, 68.0◦, 69.2◦, 72.6◦ and 77.0◦, corresponding to the (100), (002), (101), (102), (110),
(103), (200), (112), (201), (004) and (202) planes of hexagonal wurtzite ZnO, indexed to ZnO
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 48
in the ICSD database ZnO with reference code 01-073-8765.
Figure 5.2: XRD pattern for reference and as-prepared undoped ZnO
5.1.2 Fe-doping
X-ray diffractograms for undoped and doped ZnO samples at 1, 2, 4 and 8 mol % are shown
in Figure 5.3a. ZnO was the primary crystalline phase present for all the samples. However,
for doping at 4 and 8 mol %, secondary peaks were observed at 2θ angles of 29.75◦, 35.04◦,
42.6◦ and 61.9◦ (marked with dots in the figure). These peaks were indexed to the face centred
cubic (FCC) phase of magnetite (Fe3O4) with ICSD number 01-078-3148. Also, the peaks in
the x-ray diffraction patterns of Fe-doped ZnO were shifted toward lower 2θ angles. The peak
corresponding to reflections from the (102) plane has been used as a marker to highlight this
trend. This peak shift was present for all doped samples and is indicative of an enlarged unit
cell, or increased strain, or both.
Table 5.1 lists the lattice parameters of the unit cell, the crystallite size, strain and the shift in
the diffraction peaks towards lower 2θ angles. The value of the unit cell volume increased with
the doping percentage from 0 - 2 mol %. A slight decrease in the volume of the unit cell was
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 49
(a) XRD for reference and Fe-doped ZnO.
Secondary phase peaks marked with dots are due to Fe3O4.
(b) XRD for Mn-Doped ZnO.
Secondary phase peaks marked with dots are due to Mn3O4.
Figure 5.3: X-ray diffraction patterns for Fe- and Mn-doped ZnO.
The vertical line at 2θ ≈ 47.6◦ in each chart is a marker and guide to the eye.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 50
(a) XRD for undoped and Co-doped ZnO at pH 5.
No secondary crystalline phases are present.
(b) XRD for reference and Ni-doped ZnO at pH 5.
No secondary crystalline phases are present.
Figure 5.4: X-ray diffraction patterns for Co- and Ni-doped ZnO.
The vertical line at 2θ ≈ 47.6◦ in each chart is a marker and guide to the eye.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 51
recorded at 4 mol % with an increase at 8 mol %. The value of the unit cell at 4 mol % despite
the decrease, was however larger than that of the 0 mol %. The c/a ratio remained constant at
1.62.
Table 5.1: Crystal parameters and reflections from (102) planes (Fe-doping)
Doping (Mol %) a (Å) c (Å) V (Å)3 Size (nm) Strain ×10−3 c/a 2θ (102)
Reference 3.216 5.196 46.541 92 0 1.62 47.63
0 3.218 5.197 46.607 30 3 1.62 47.58
1 3.221 5.204 46.757 34 4 1.62 47.52
2 3.221 5.206 46.775 26 3 1.62 47.53
4 3.220 5.205 46.737 29 3 1.62 47.54
8 3.222 5.205 46.795 33 1 1.62 47.52
The lattice strain was about 3.0 × 10−3 for all the doped samples except at 8 mol % where
it was 1.0 × 10−3 with an error of ±1 × 10−3. The crystallite sizes of the doped and undoped
samples were all within the range of 26 - 34 nm, which, with an average error of 40%, means
that the sizes were within the same size range.
The changes in unit cell volume with the doping concentration for all the samples are shown
in Figure 5.5. An initial sharp increase in volume is followed by a more gradual increase as the
doping percentage increased.
5.1.3 Mn-doping
The x-ray diffractograms of undoped and Mn-doped ZnO at 1, 2, 4 and 8 mol % are shown in
Figure 5.3b. ZnO was the primary crystalline phase observed, but there was also a secondary
phase with peaks at 2θ ≈ 29.6◦ (4 and 8 mol % doping) and 61.7◦ (8 mol% doping). This
secondary phase was indexed to the spinel phase of manganese oxide (Mn3O4) with ICSD
number 00-013-0162.
As was observed with Fe-doping, the diffraction peaks were observed to shift towards lower
2θ angles (Figure 5.3b), suggesting an increase in the volume of the unit cell. Peak shifts to
lower 2θ were observed for all Mn-doped samples.
The crystallographic parameters are presented in Table 5.2, along with the crystallite size
and strain. The volume of the unit cell increased with doping concentration from 0-2 mol %
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 52
(a) Fe-doped (b) Mn-doped
(c) Co-doped (d) Ni-doped
Figure 5.5: Variation in the volume of the unit cell with doping concentration
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 53
with a decrease at 4 mol % with a slight increase at 8 mol %. Just as was observed for the Fe-
doping, the value at these doping concentrations, were larger than that of the 0 mol %. The c/a
ratio remained constant at 1.62 for all samples. The change in the lattice parameters is reflected
in the increase in the volume of the unit cell (Figure 5.5b). Just as for the case of Fe-doping,
there was a sharp increase in the unit cell volume at the onset of doping, but the rate of increase
fell off, dropping at 4 mol % doping and reaching a value of 46.746 Å3 at 8 mol % doping. Such
increases in the lattice parameters have also been reported by Deka and Joy (2007) and Mote
et al. (2011) for Mn-doped ZnO nanowires and nanoparticles.
Table 5.2: Crystal parameters and reflections from the (102) plane (Mn-doping)
Doping (Mol %) a (Å) c (Å) V (Å)3 Size (nm) Strain ×10−3 c/a 2θ (102)
Reference 3.216 5.196 46.541 92 0 1.62 47.63
0 3.218 5.197 46.607 30 3 1.62 47.58
1 3.220 5.210 46.782 25 3 1.62 47.53
2 3.221 5.206 46.775 33 3 1.62 47.53
4 3.220 5.204 46.728 41 2 1.62 47.54
8 3.220 5.206 46.746 27 2 1.62 47.53
The crystallite sizes for the doped samples did not appear to depend on the doping concen-
tration, and were between 25 - 41 nm, with the error of ±40%. The lattice strain was constant
at about 3× 10−3.
5.1.4 Co-doping
Figure 5.4a shows the x-ray diffraction patterns for undoped and Co-doped samples. The only
phase observed was wurtzite ZnO.
The crystallographic parameters are detailed in Table 5.3. The diffraction peaks were shifted
to lower 2θ angles with respect to the undoped ZnO at 47.63◦ but remained constant at 47.53◦
for all doped samples. While the parameter a was constant for all the doped samples, c tended
to decrease, leading to a decrease in the unit cell volume of 46.728 Å3 at 8 mol %. The sharp
increase in the unit cell volume observed in the Fe- and Mn-doped samples was also present in
the Co-doped ZnO samples (Figure 5.5c). The c/a ratio was constant at 1.62 for both undoped
and doped samples. The absence of secondary phases and the relatively constant volume of the
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 54
unit cell with increasing doping concentration is not surprising, since cobalt is reported to be
one of the most soluble transition metals in ZnO (Jin et al. (2000)).
The size of the crystallites for the doped samples ranged from 31 to 37 nm (±40%) while
the lattice strain was maximum at 4×10−3 (4 mol %) and minimum at 2×10−3 (8 mol %) with
an error of 1× 10−3.
Table 5.3: Crystal parameters reflections from (102) plane (Co-doping)
Doping (Mol %) a (Å) c(Å) V (Å)3 Size (nm) Strain ×10−3 c/a 2θ (102)
Reference 3.216 5.196 46.541 92 0 1.62 47.63
0 3.218 5.197 46.607 30 3 1.62 47.58
1 3.220 5.206 46.746 31 3 1.62 47.54
2 3.221 5.206 46.775 31 3 1.62 47.53
4 3.220 5.204 46.728 37 4 1.62 47.54
8 3.220 5.204 46.728 31 2 1.62 47.54
5.1.5 Ni-doping
The x-ray diffractograms of undoped and Ni-doped ZnO are shown in Figure 5.4b. All the
diffraction peaks which appear are indexed to the wurtzite phase of ZnO and no secondary
phases were observed. However, the diffraction peaks shifted towards lower 2θ angles for the
doped samples.
Table 5.4 lists the crystallographic parameters of Ni-doped ZnO as well as the size and strain
of the crystallites. The volume of the unit cell increased from 46.607 Å3 for the undoped sample
to 46.77 Å3 at 8 mol % doping. However, as seen in Figure 5.5d, unlike for the other dopants,
the volume of the unit cell did not increase until doping was at 4 mol %. The crystallite size of
the Ni-doped ZnO was between 30 and 38 nm (±40%). The strain in the doped samples was
constant at about 3× 10−3 except at 8 mol %, when it was 1× 10−3. The c/a ratio was constant
at 1.62 for both undoped and doped samples.
Aspect Ratios
Table 5.5 shows the aspect ratios (AR) for the doped samples. The aspect ratios for the reference
ZnO powder and the undoped ZnO are also listed for comparison. It can be seen that AR for the
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 55
Table 5.4: Crystal parameters and reflections from (102) plane (Ni-doping)
Doping (mol %) a (Å) c (Å) V (Å)3 Size (nm) Strain ×10−3 c/a 2θ (102)
Reference 3.216 5.196 46.541 92 0 1.62 47.63
0 3.218 5.197 46.607 30 3.0 1.62 47.58
1 3.216 5.198 46.559 38 2.0 1.62 47.61
2 3.216 5.200 46.576 30 2.0 1.62 47.60
4 3.221 5.207 46.784 32 2.0 1.62 47.53
8 3.221 5.207 46.784 32 1.0 1.62 47.53
doped and undoped samples are in general higher than that of the reference ZnO, but the values
do not exhibit a dependence on the doping concentration.
Table 5.5: Aspect Ratios (AR) for doped ZnO
Doping (mol %) AR (Fe) AR (Mn) AR (Co) AR (Ni)
Reference 0.73 0.73 0.73 0.73
0 1.02 1.02 1.02 1.02
1 1.21 1.08 1.16 1.10
2 1.30 0.99 1.02 1.04
4 1.20 0.87 1.07 1.04
8 1.05 0.88 0.97 1.03
5.1.6 Summary
In the preceding sections, the results of x-ray analysis on samples prepared at pH 5 and annealed
at 280 ◦C have been presented. The appearance of secondary phases in some of the x-ray
diffractograms of the doped samples is due to the inadvertent reaction of the dopant ions to
form their oxides during the synthesis procedure. In the literature, reports of secondary phases
in wet-chemical synthesis are not uncommon, as for example Silambarasan et al. (2015) have
reported in Fe-doped ZnO nanoparticles and (Sharma et al. (2006)) in Mn-doped ZnO.
In order to minimize the occurrence of these secondary phases, synthesis was also carried
out at elevated pH of 3 instead of 5, and the results are presented in section 5.3.
The x-ray diffraction patterns for all the doped samples showed that the unit cell invariably
enlarged with the introduction of the dopant. This increase was not monotonic, however, but
in sum, the apparent effect of doping on the lattice parameters was an increase in the volume
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 56
of the unit cell for all dopant ions. Even though the increase in the dimensions of the unit cell
could be due to the substitution of a zinc ion by a dopant cation with a larger radius, the case of
cobalt and nickel doped samples are different on account of the fact that their ionic sizes are less
than that of Zn2+. From Table 5.6, the unit cell volumes for the doped samples are compared
with the unit cell volume for the undoped sample. It can be seen that the increase in the unit
cell appears to depend on the radius of the of the dopant cation. However, the data presented
so far indicated that even the cations that are smaller than the host Zn cation seemed to cause
the unit cell to expand. Again, there is insufficient experimental data to determine the correct
oxidation state of the dopant cations, and as is well known, the size of the ion within the crystal
depends on its oxidation state. This expansion could be due to strain in the crystallite as well.
It would have been expected that the crystallite strain would increase with increasing doping
concentration. However, this was not generally the case. The values of the strain (which were
quite small for all samples) did not correspond to doping concentration. In some cases the strain
at higher doping concentrations were lower than for lower doping. In order to gain more insight
into this, the samples were annealed at 600 ◦C, the results of which are presented in Section 5.4.
This expansion of the unit cell will again be discussed after the results from the computa-
tional work are presented in the latter part of Chapter 5.
Table 5.6: Changes in the unit cell volume with Doping
Dopant Maximum Volume (Å3) Average Volume, Vav (Å3) Vav − V 30 (Å ) Cation Radius (Å)
Fe2+ 46.795 46.766 0.159 0.77
Mn2+ 46.782 46.758 0.151 0.80
Co2+ 46.775 46.744 0.137 0.72
Ni2+ 46.784 46.676 0.069 0.69
Vo is the volume of the unit cell of undoped ZnO.
Table 5.7 lists the maximum and average aspect ratios for the doped samples. Compared
with the standard ZnO, the aspect ratios are seen to be higher. It would appear that the method
of synthesis delivers crystallites with an aspect ratio of about 1, regardless of dopant.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 57
Table 5.7: Aspect Ratios
Sample Maximum Value Average Value
ZnO Reference 0.73 0.73
Undoped ZnO 1.02 1.02
Fe-doped 1.30 1.19
Mn-doped 1.08 0.96
Co-doped 1.16 1.06
Ni-doped 1.10 1.05
5.2 Band Gap Energy
5.2.1 Fe-doping
The band gap energies calculated from the UV-VIS absorption spectra and the PL near band
edge emission for Fe-doped samples are shown in Table 5.8. The UV-VIS absorption spectra
for undoped and Fe-doped ZnO are shown in Figure 5.6a. From the UV-VIS data, a general
narrowing of the band gap with respect to the reference ZnO was observed. At 4 and 8 mol
%, the reduction in the band gap energy was 0.19 eV compared to the undoped sample. On
the other hand, the band gap energies calculated from PL shows a different picture. Contrary
to data from UV-VIS spectroscopy, the band gap energy increased with doping concentration,
reaching 3.30 eV at 4 and 8 mol % doping.
The PL emission spectra of 8 mol % Fe-doped ZnO is shown in Figure 5.7a. Two peaks
corresponding to interband transitions can be seen at 494 and 579 nm.
PL emission spectra for the other doping percentages of iron are shown in Appendix B.
Table 5.8: UV-VIS Band Gap Energy (Fe-doping)
Doping UV-VIS PL
Mol % Band Gap Energy (±0.01) eV Band Gap Energy (±0.03) eV
Reference 3.19 3.23
0 3.10 3.20
1 3.06 3.19
2 3.08 3.27
4 2.91 3.30
8 2.95 3.30
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 58
5.2.2 Mn-doping
The band gap energies calculated from the UV-VIS absorption edge and the PL near band edge
emission are shown in Table 5.9. The band gap energy from UV-VIS reduced with doping
concentration, dropping by a value of 0.45 eV at 4 mol % and by 0.26 eV at 8 mol %. However,
the band gap energies calculated from PL increased with doping concentration, reaching 3.31
eV at 8 mol %, an increase of 0.11 eV over the band gap energy for the undoped sample.
The UV-VIS absorption spectra of undoped and Mn-doped ZnO are shown in Figure 5.6b.
Figure 5.7b shows the PL emission spectrum of 4 mol % Mn-doped ZnO. Interband transi-
tions showed up as peaks at 489 nm and 538 nm.
PL emission spectra for the other Mn-doping concentrations are shown in Appendix B.
Table 5.9: UV-VIS Band Gap Energy (Mn-doping)
Doping UV-VIS PL
Mol % Band Gap Energy (±0.01) eV Band Gap Energy (±0.03) eV
Reference 3.19 3.23
0 3.10 3.20
1 3.08 3.29
2 2.99 3.30
4 2.65 3.26
8 2.84 3.31
5.2.3 Co-doping
For cobalt doping, the values obtained from the UV-VIS absorption edge and the PL near band
emission are presented in Table 5.10. The band gap energies from UV-VIS showed a steady
reduction, with the exception of 4 mol % doping, which at 3.14 eV was higher than the values
for the 2 mol % doping and 8 mol %, but lower than the value for the undoped sample. The
band gap energies from PL increased with doping concentration, reaching a value of 3.30 eV at
8 mol %.
Figure 5.6c shows the UV-VIS absorption spectra for undoped and Co-doped ZnO, while
Figure 5.7c shows the PL emission spectra of 8 mol % Co-doped ZnO. The peak at 491 nm is
due to an interband transition.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 59
Table 5.10: UV-VIS Band Gap Energy (Co - doping)
Doping UV-VIS PL
Mol % Band Gap Energy (±0.01 ) eV Band Gap Energy (±0.03) eV
Reference 3.19 3.23
0 3.10 3.20
1 2.96 3.17
2 2.98 3.31
4 3.14 3.32
8 2.86 3.30
PL emission spectra for the other doping concentrations are presented in Appendix B.
5.2.4 Ni-doping
The band gap energies calculated from the UV-VIS absorption and PL emission spectra for
undoped and Ni-doped ZnO are shown in Table 5.11. Contrary to data from the Fe-, Mn-, and
Co-doped samples, the band gap energies of the Ni-doped samples from the absorption edge of
the UV-VIS spectra did not appear to change much. The band gap energy at 0 mol % doping
was 3.10 eV and the largest deviation from this value was a reduction by 0.04 eV, for 4 mol %
doping. Within the range of error, the band gap energies obtained from PL for Ni-doped ZnO
did not show much changes either.
Table 5.11: UV-VIS Band Gap Energy (Ni - doping)
Doping UV-VIS PL
Mol % Band Gap Energy (±0.01) eV Band Gap Energy (±0.03) eV
Reference 3.19 3.23
0 3.10 3.20
1 3.07 3.16
2 3.11 3.20
4 3.06 3.25
8 3.09 3.26
Figure 5.6d depicts the UV-VIS absorption spectra. The PL spectrum of 8 mol % Ni-doping
(Figure 5.7d) shows the presence of an interband state with a peak at 600 nm.
PL spectra for the other Ni-doped percentages are shown in Appendix B.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 60
(a) Fe-doped (b) Mn-doped
(c) Co-doped (d) Ni-doped
Figure 5.6: UV-VIS Absorption Spectra
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 61
(a) Fe-doped (b) Mn-doped
(c) Co-doped (d) Ni-doped
Figure 5.7: PL Emission Spectra
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 62
5.2.5 Summary
The data presented in the foregoing section shows that the band gap energies calculated from
the UV-VIS absorption spectra decreased with doping concentration, with the exception of the
nickel doped samples. On the other hand, the band gap energies calculated from the near band
emission of the PL spectra showed an increase in the band gap energies, again with the excep-
tion of the nickel doped samples, which were generally constant for both the UV-VIS and PL
experiments. In this work, it was noticed that for the doped samples which showed secondary
phases, the band gap energy calculated from the UV-VIS absorption invariably reduced with
doping concentration, leading to the conclusion that the secondary phases were having an effect
on the band gap energies determined by UV-VIS absorption. This observation was consistent,
and will be discussed more thoroughly in the subsequent sections and in the conclusion. On the
other hand, the band gap energies calculated from the PL emission spectra did not change with
the secondary phases, leading to the conclusion that the band gap energies from PL are more
reliable in this sense.
In order to clarify the main results of this section, a quantity∆Eg is defined as the maximum
difference between the band gap energy for the doped and undoped samples. The band gap
energies calculated from the PL spectra presented earlier are used, and the data is presented in
Table 5.12. It is seen that the maximum change in the band gap energy are within the same range
for all the dopants with the exception of nickel. This trend will be discussed in the conclusion
after the data from the computational work has been presented and discussed.
Table 5.12: Change in Band Gap Energy With Doping
Dopant ∆Eg (eV)
Fe 0.10
Mn 0.11
Co 0.12
Ni 0.06
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 63
5.3 Synthesis at Elevated pH
The oxidation state of a metal ion in solution is strongly dependent on the pH of the solution.
For instance Fe2+, (Morgan and Lahav (2007)) and Mn2+ (Wekesa et al. (2011)) are known to
be the most dominant states in aqueous solutions with high pH (in the pH range of 4 and below).
In order to use this property in improving the quality of doping, the synthesis was repeated for
4 and 8 mol % doping at pH 3. The samples were further annealed at 280 ◦C and results for the
crystallographic and optical characterisation are presented in this section.
5.3.1 Lattice parameters
Fe-doping
Figure 5.8 compares the x-ray diffractograms of 4 and 8 mol % Fe-doping at both pH 3 and
5. At 4 mol %, the secondary phase of Fe3O4 appears at pH 5 but not at pH 3. At 8 mol %,
the peaks for Fe3O4 are present in both diffractograms but with diminished intensity at pH 3,
suggesting that the formation of Fe3O4 is less favoured at pH 3 as compared to pH 5.
The crystallographic parameters are listed in Table 5.13. The volume of the unit cell in-
creased at 4 mol % doping but decreased for 8 mol % doping from pH 5 to 3. It is observed that
the crystallite size reduced by 9 nm at pH 3 for both doping concentrations, while the strain did
not change with pH.
Table 5.13: Lattice parameters, crystallite size and strain for 4 and 8 mol % Fe-doping
Doping (Mol %) a (Å) c (Å) V (Å)3 c/a Size (nm) Strain ×10−3
4(pH 5) 3.220 5.204 46.728 1.62 29 3
4 (pH 3) 3.221 5.203 46.748 1.62 20 3
8 (pH 5) 3.222 5.204 46.786 1.62 30 2
8 (pH 3) 3.221 5.205 46.766 1.62 21 2
Mn-doped ZnO
Figure 5.9 shows the x-ray diffraction patterns for 4 and 8 mol% Mn-doping. The 4 mol %
Mn-doped ZnO at pH 3 showed only the primary phase of the wurtzite ZnO, however, at pH
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 64
Figure 5.8: XRD pattern for 4 and 8 mol % Fe - doped samples at pH 5 and 3
5 a secondary phase appeared. The 8 mol % Mn-doped at pH 3 and pH 5 also showed this
secondary phase, which has earlier been indentified as Mn3O4.
The crystallographic parameters are presented in Table 5.14, along with the crystallite size
and the strain. The c/a ratio of the unit cell was also constant at 1.62. The crystallite sizes at 8
mol % were the same at both pH 5 and pH 3, but at 4 mol %, the size decreased by 19 nm at pH
3.
Table 5.14: Lattice parameters, crystallite size and strain for 4 and 8 mol % Mn-doping
Doping (Mol %) a (Å) c (Å) V (Å)3 c/a Size (nm) Strain ×10−3
4(pH 5) 3.220 5.204 46.728 1.62 41 2
4 (pH 3) 3.220 5.206 46.746 1.62 22 3
8 (pH 5) 3.220 5.206 46.746 1.62 27 2
8 (pH 3) 3.221 5.207 46.784 1.62 26 2
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 65
Figure 5.9: x-ray diffraction patterns of 4 and 8 mol % Mn-doping at pH 3 and 5
Co-doped ZnO
Figure 5.10 shows the x-ray diffractograms for 4 and 8 mol % Co-doping at pH 5 and 3. No
secondary phases were observed.
The crystallographic parameters are listed in Table 5.15. Interestingly, even though at 4 mol
% doping the crystal parameters a and c reduced at pH 3, strain was constant, along with the
c/a ratio at 1.62 at both pH 3 and pH 5. The c/a ratio and the strain were constant, as was the
crystallite size within the range of the average error of 40 %.
Table 5.15: Lattice parameters, crystallite size and strain for 4 and 8 mol % Co-doping
Doping (Mol %) a (Å) c (Å) V (Å)3 c/a Size (nm) Strain ×10−3
4(pH 5) 3.220 5.204 46.728 1.62 22 2
4 (pH 3) 3.216 5.198 46.559 1.62 26 2
8 (pH 5) 3.220 5.203 46.719 1.62 31 2
8 (pH 3) 3.220 5.205 46.737 1.62 27 3
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 66
Figure 5.10: X-ray diffractograms of 4 and 8 mol% Co-doping
Ni-doped ZnO
The x-ray diffractograms for 4 and 8 mol % Ni-doping at pH 3 and 5 are presented in Figure
5.11. Here too, as for the case of the Co-doping, no secondary phases were observed.
Table 5.16 lists the crystallographic parameters for 4 and 8 mol % doping. The volume of
the unit cell decreased for 4 mol % from pH 5 to pH 3 and increased for 8 mol % from pH 5 to
pH 3. The crystallite size was constant within the range of error (at 40 % average error). c/a
was constant at 1.62 and the strain did not vary from 2× 10−3.
Table 5.16: Lattice parameters, crystallite size and strain for 4 and 8 mol % doping.
Doping (Mol %) a (Å) c (Å) V (Å)3 c/a Size (nm) Strain ×10−3
4(pH 5) 3.221 5.207 46.784 1.62 32 2
4 (pH 3) 3.221 5.206 46.775 1.62 37 2
8 (pH 5) 3.221 5.206 46.775 1.62 32 1
8 (pH 3) 3.221 5.208 46.793 1.62 34 2
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 67
Figure 5.11: XRD pattern for 280 ◦C annealing of 4 and 8 mol % Ni-doped samples at pH 5
and 3
Table 5.17: Aspect ratio of 4 and 8 mol% Fe-doping
Dopant Doping (mol %) AR (pH 5) AR (pH 3)
4 1.20 1.20
Fe
8 1.05 1.07
4 0.89 0.99
Mn
8 0.88 0.88
4 1.07 1.06
Co
8 0.97 1.16
4 1.04 1.04
Ni
8 1.03 1.10
Aspect Ratios (AR)
Table 5.17 shows the AR for 4 and 8 mol % doping. For Fe, it can be seen that at 4 mol %
doping, the aspect ratio was 1.20, compared with 0.73 for the ZnO standard and 1.02 for the
as-prepared annealed ZnO. For 4 mol % Mn-doping the crystallites were longer along the c
axis, but at 8 mol % the aspect ratio remained constant at 0.88. For Co, the AR at 4 mol% was
constant whereas at 8 mol % the crystallites are elongated along the c axis for pH 3. For Ni, the
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 68
AR at 4 mol % doping is seen to be constant, with an increase from 1.03 to 1.10 with pH for 8
mol % doping.
5.3.2 Summary
In the light of the foregoing, it can be concluded that the effects on the crystal parameters of
synthesis at elevated pH are as follows:
1. At higher pH, the formation of secondary phases was less likely
2. The crystallite sizes saw slight reductions in the Fe-doped samples, though the associated
error in the values (40 %) was large enough to conclude that the sizes were essentially the
same. For Mn-doping, the crystallite size reduced by about 50% for 4 mol % doping at
pH 3 as compared to pH 5, but remained within the same range for the 8 mol % doping.
For Co- and Ni-doping, crystallite sizes for the doped samples were within the same range
at pH 3 and pH 5.
3. Even though the aspect ratios of the samples indicate that the crystallites are elongated
along the c-axis as compared to the reference ZnO, the change in the values with respect
to pH are not considered to be significant.
5.3.3 Band Gap Energies
Fe-doping
Figure 5.12a shows the UV-VIS absorption spectra for 4 and 8 mol % Fe-doped ZnO at pH 3
and pH 5. As can be seen in Table 5.18, the band gap energies at pH 3 were larger than the
corresponding values at pH 5.
Table 5.18: UV-VIS Band Gap Energy for 4 and 8 mol % Fe-doping ZnO at pH 5 and pH 3
Doping pH 5 pH 3
Mol % Band Gap Energy (±0.01) eV Band Gap Energy (±0.01) eV
4 2.91 3.15
8 2.95 3.20
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 69
Table 5.19: PL Band Gap Energy from PL emission spectra of 4 and 8 mol % Fe - doping
Doping pH 5 pH 3
Mol % Band Gap Energy (±0.03) eV Band Gap Energy (±0.03) eV
4 3.30 3.28
8 3.30 3.38
The PL emission spectra of the 4 and 8 mol % Fe-doped ZnO at pH 3 are presented in Figure
5.13a, and the values of the band gap energy presented in Table 5.19. The band gap energy at
pH 3 increased from 3.28 eV to 3.38 eV for 4 and 8 mol % doping respectively. For the 8 mol
% doping, the band gap increased from 3.30 eV to 3.38 eV at pH 3. An interband state can be
observed at 498 nm in Figure 5.13a.
Mn-doped ZnO
The UV-VIS absorption spectra for 4 and 8 mol % Mn-doping at pH 3 and pH 5 are shown in
Figure 5.12b, and the values of the band gap energies listed in Table 5.21. The band gap energies
for both 4 and 8 mol % doping are seen to increase from pH 5 to pH 3. The increase in band
gap energy at 4 mol % doping with pH supports the assertion that the presence of secondary
phases affected the position of the absorption edge in the UV-VIS spectrum. It must be recalled
that there were no secondary phases observed at pH 3 for 4 mol % doping. The smaller increase
in band gap energy at 8 mol % doping can also be attributed to this same fact, since secondary
phases were observed at pH 3 for 8 mol % doping.
The band gap energies are listed in Table 5.21. At 8 mol % doping, there was no difference
in band gap at pH 3 and pH 5, whereas the band gap energy reduced for 4 mol % doping. The
PL spectrum (Figure 5.13) shows an inter-band peak at 488 nm for 4 mol % doping, and two
interband states at 488 and 556 nm for the 8 mol % doping.
Table 5.20: UV-VIS Band Gap Energy for 4 and 8 mol % Mn - doping
Doping pH 5 pH 3
Mol % Band Gap Energy (±0.01 eV) Band Gap Energy (±0.01 eV)
4 2.65 2.98
8 2.84 2.86
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 70
(a) Fe-doped (b) Mn-doped
(c) Co-doped (d) Ni-doped
Figure 5.12: UV-VIS absorption Spectra for 4 and 8 mol% doping at pH 5 and 3
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 71
(a) Fe-doped (b) Mn-doped
(c) Co-doped (d) Ni-doped
Figure 5.13: PL emission spectra of 4 and 8 mol % doping at pH3
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 72
Table 5.21: PL Band Gap Energy for 4 and 8 mol % Mn - doping ZnO at pH 5 and 3
Doping pH 5 pH 3
Mol % Band Gap Energy (±0.01 eV) Band Gap Energy (±0.01 eV)
4 3.29 3.24
8 3.31 3.31
Co-doped ZnO
The UV-VIS absorption spectra for 4 and 8 mol % Co-doping at pH 3 and pH 5 are presented in
Figure 5.12c. Table 5.22 lists the band gap energies calculated from the spectra. The band gap
energy for 4 mol % doping reduced from 3.14 eV (pH 5) to 2.88 eV (pH 3) but increased for 8
mol % doping from 2.86 eV to 3.06 eV. The reduction in band gap observed for 4 mol % at pH 3
was unexpected in the light of the behaviour of the Fe-doped and Mn-doped samples, especially
since there were no secondary phases observed in the x-ray diffractogram for the sample.
Table 5.22: UV-VIS Band Gap Energy for 4 and 8 mol % Co-doping
Doping pH 5 pH 3
Mol % Band Gap Energy (±0.01) eV Band Gap Energy (±0.01) eV
4 3.14 2.88
8 2.86 3.06
Table 5.23: PL Band Gap Energy for 4 and 8 mol % Co-doping.
Doping pH 5 pH 3
Mol % Band Gap Energy (±0.03 eV) Band Gap Energy (±0.03 eV)
4 3.32 3.31
8 3.30 3.35
The PL emission spectra are presented in Figure 5.13. Interband states appear at 498 nm
and 490 nm for the 8 mol % and 4 mol % doping respectively. The corresponding band gap
energies (Table 5.23) are constant at 4 mol % but showed a 5 nm increase at pH 3 for 8 mol %
doping.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 73
Ni-doping
Figure 5.12d shows the UV-VIS absorption spectra for 4 and 8 mol % Ni - doping at pH 5 and
3, and Table 5.24 lists the band gap energies. A decrease in the band gap energy was observed
for both samples, which was an unexpected result due to the absence of secondary phases in the
x-ray spectra. However, this result is similar to the case for 4 mol % Co-doping.
The PL emission spectra (Figure 5.13d) and band gap energies (Table 5.25) of 4 and 8
mol% however show a different picture, with the band gap energy constant within the error
range. From Figure 5.13d, an interband transition state was observed at a wavelength of 493
nm for 4 mol %.
Table 5.24: UV-VIS Band Gap Energy for 4 and 8 mol % Ni - doping at pH 5 and 3
Doping pH 5 pH 3
Mol % Band Gap Energy (±0.01) eV Band Gap Energy (±0.01) eV
4 3.06 2.90
8 3.09 3.01
Table 5.25: PL Band Gap Energy for 4 and 8 mol % Ni - doping at pH 5 and 3
Doping pH 5 pH 3
Mol % Band Gap Energy (±0.03) eV Band Gap Energy (±0.03) eV
4 3.25 3.24
8 3.26 3.24
5.3.4 Summary
The observed trend in band gap energy for the 4 and 8 mol% Fe-doped ZnO at pH 3, supports
the earlier argument that the observed decrease in the band gap energy for these samples at pH 5,
as measured by UV-VIS, was due to the presence of secondary phases. Though at 8 mol % and
pH 3 Fe-doping showed a secondary phase, the reduction in intensities in the x-ray spectrum
showing a diminished presence of the secondary phases, resulted in the increased band gap
energy.
It is clear from the trend of band gap observed for the 4 and 8 mol % Fe and Mn-doped ZnO
that UV-VIS absorption spectra is sensitive to the formation of secondary phases. Though there
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 74
were no observed secondary phases in Ni and Co-doped ZnO at pH 5 and 3, a decrease was
observed in the band gap energies of these samples at pH 3. These decreases can however not
be explained directly by the argument about the presence of secondary crystalline phases, since
none were observed in the x-ray diffractograms. (It must be indicated that these measurements
were repeated several times with the same results). However, the fact that as the secondary
phases reduced, the band gap as measured by UV-VIS increased, supports this proposition in
all the other cases.
5.4 Annealing at 600 ◦C
As is well known, the hydrothermal method, because of the relatively low temperatures used in
synthesis, typically produces crystals with a large number of defects and surface contaminants.
Some of the surface contaminants are removed by washing, as was done in this work. However,
it is usual to carry out annealing at elevated temperatures as well. The data that has been reported
so far has been for samples annealed at 280 ◦C. Further annealing at 600 ◦C was carried out on
the 4 mol % and 8 mol % doped samples, and the results of the characterisation are presented
in this section.
5.4.1 Lattice Parameters
Fe-doping
Figure 5.14 shows the x-ray diffractograms of undoped, 4 and 8 mol % Fe-doped ZnO prepared
at pH 5 and 3. The observed secondary phase in the samples at pH 5 and 3 was indexed to the
face centred cubic (FCC) of magnetite (Fe3O4).
In Table 5.26, the crystallographic parameters of the undoped, 4, and 8 mol% Fe-doped
ZnO prepared at pH 5 are shown, while Table 5.27 lists the values for pH 3. At both pH 3 and
pH 5, the volume of the unit cell decreased at the annealing temperature of 600 ◦C. At pH 5,
the change in the crystallite size due to the different annealing temperatures was 1 nm, much
smaller than the average error in the size, which was 37 %. However, at pH 3, annealing at 600
◦C caused an increase in the crystallite size, almost tripling in size (4 mol %) and doubling in
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 75
size (8 mol %). Also, at pH 5 the strain was constant for both annealing temperatures, but at pH
3 the strain was significantly diminished (Table 5.27). The c/a ratio of the unit cell was constant,
and in general, the crystallographic parameters of the samples at 600 ◦C annealing were close
to the parameters of undoped ZnO, indicating that annealing the doped crystals improved the
crystalline quality.
(a) Fe-doping at pH 5 (b) Fe-doping at pH 3
Figure 5.14: XRD pattern for 4 and 8 mol % Fe-doped ZnO 600◦ annealing.
Table 5.26: Lattice parameters, crystallite size and strain at pH 5 (Fe-doping)
Doping (Mol %) a (Å) c (Å) V (Å)3 c/a Size (nm) (Strain ±1)× 10−3
0 (600◦C) 3.216 5.198 46.559 1.62 62 1.5
4 (280◦C) 3.220 5.204 46.728 1.62 29 3.0
4 (600◦C) 3.218 5.198 46.603 1.62 31 3.0
8 (280◦C) 3.222 5.205 46.795 1.62 30 2.0
8 (600◦C) 3.217 5.196 46.571 1.61 29 2.0
Table 5.27: Lattice parameters, crystallite size and strain at pH 3 (Fe-doping)
Doping (Mol %) a (Å) c (Å) V (Å)3 c/a Size (nm) (Strain ±1)× 10−3
0 (600◦C) 3.216 5.198 46.559 1.62 62 1.5
4 (280◦C) 3.221 5.203 46.748 1.62 20 3.0
4 (600◦C) 3.218 5.198 46.614 1.62 63 0.7
8 (280◦C) 3.221 5.205 46.766 1.62 21 2.0
8 (600◦C) 3.219 5.199 46.649 1.62 50 0.4
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 76
Mn-doping
Figure 5.15a shows the x-ray diffractograms for undoped, 4 and 8 mol % Mn-doping at pH 5.
The peak at 2θ angle of 29.8◦ was identified as belonging to the spinel phase of Mn3O4.
Figure 5.15b shows x-ray diffractograms of undoped, 4, and 8 mol% Mn-doping at pH 3.
Here, too, the peaks observed at 2θ angles of 30.08, 43.02 and 53.65◦ for 4 mol % doping were
indexed to the spinel phase of Mn3O4 whereas the peaks at 8 mol % doping at 2θ angles of 37.08
and 43.07◦ were indexed to the hexagonal phase of MnO2 (ICSD reference code 01-089-5171).
Table 5.28 lists the crystallographic parameters at pH 3 and Table 5.29 lists the values at pH
5 for both the 4 and 8 mol %Mn-doped ZnO samples. At pH 5 and pH 3, the volume of the unit
cell reduced for both 4 and 8 mol % doping. The lattice strain and the c/a ratio were constant
at pH 5, but the crystallite size changed by up to 2 nm, which is insignificant considering the
average error in the size of 36 %. At pH 3 however, the crystallite size increased at 600 ◦C by
5 nm (4 mol % doping) with the strain 33 % decrease in the strain, and by 13 nm (8 mol %
doping) with no change in the strain. The c/a ratio of the unit cell remained constant.
Again, the values of the lattice parameters are in agreement with that of the undoped ZnO
at 600 ◦C annealing.
(a) Mn-doping at pH 5 (b) Mn-doping at pH 3
Figure 5.15: XRD pattern for 4 and 8 mol % Mn-doped ZnO at 600◦ annealing.
The aspect ratio decreased with increasing annealing temperature (Table 5.34), which was
also observed for the Fe-doped samples, and are close to the values for undoped ZnO at 600 ◦C.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 77
Table 5.28: Lattice parameters, c/a ratio, crystallite size and strain for 4 and 8 mol % Mn-
doping at pH 3
Doping (Mol %) a (Å) c (Å) V (Å)3 c/a Size (nm) (Strain ±1)× 10−3
0 (600◦C) 3.216 5.198 46.559 1.62 62 1.5
4 (280◦C) 3.220 5.206 46.746 1.62 22 3.0
4 (600◦C) 3.218 5.199 46.632 1.62 37 1.0
8 (280◦C 3.221 5.207 46.784 1.62 26 2.0
8 (600◦C) 3.220 5.202 46.707 1.62 49 2.0
Table 5.29: Lattice parameters,crystallite size and strain for 4 and 8 mol % Mn-doping at pH 5
with different annealing temperatures
Doping (Mol %) a (Å) c (Å) V (Å)3 c/a Size (nm) (Strain ±1)× 10−3
0 (600◦C) 3.216 5.198 46.559 1.62 62 1.5
4(280◦C) 3.220 5.204 46.728 1.62 41 2.0
4 (600◦C) 3.217 5.199 46.596 1.62 39 2.0
8(280◦C) 3.220 5.206 46.746 1.62 27 2.0
8 (600◦C) 3.215 5.197 46.527 1.62 26 2.0
Co-doping
Figure 5.16a shows the x-ray diffractograms for undoped, 4 and 8 mol % Co-doped ZnO pre-
pared at pH 5. Interestingly, at 600 ◦C annealing, a secondary phase was observed, which was
not present in the samples annealed at 280 ◦C. This secondary phase with peaks at 2θ angles of
59.30 and 65.21◦ (4 mol% Co-doping) and an additional peak at 36.83◦C (8 mol % doping) was
matched with the cubic phase of cobalt oxide (Co3O4) with ICSD reference code 01-074-2120.
This same secondary phase was observed in the Co-doped samples prepared at pH 3 (Figure
5.16b).
From Table 5.30 it can be seen that the unit cell volume decreased at 600 ◦C annealing while
the crystallite size increased by 9 nm (4 mol %) and by 31 nm (8 mol %). The lattice strain
reduced by half at 4 mol% doping but remained constant at 8 mol%.
Table 5.31 lists the lattice parameters and the crystallite size and strain for undoped, 4 mol
%, and 8 mol% Co-doped ZnO synthesised at pH 3. The unit cell volume reduced while the
crystallite size increased by 16 nm for 4 mol % doping and by 31 nm for 8 mol % doping. The
strain was constant for 4 mol % doping but decreased by a third for 8 mol % doping at 600 ◦C.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 78
The c/a ratio was constant.
(a) Co-doping at pH 5 (b) Co-doping at pH 3
Figure 5.16: XRD pattern for 4 and 8 mol % Co-doped ZnO at 600◦ annealing.
Table 5.30: Lattice parameters, crystallite size and strain for 4 and 8 mol % Co-doping at pH 5
Doping (Mol %) a (Å) c (Å) V (Å)3 c/a Size (nm) (Strain ±1)× 10−3
0 (600◦C) 3.216 5.198 46.559 1.62 62 1.5
4 (280◦C) 3.220 5.204 46.728 1.62 22 2.0
4 (600◦C) 3.218 5.198 46.616 1.62 31 1.0
8(280◦C) 3.220 5.203 46.719 1.62 31 2.0
8 (600◦C) 3.219 5.198 46.645 1.62 61 2.0
Table 5.31: Lattice parameters, crystallite size and strain for 4 and 8 mol % Co-doping at pH 3
Doping (Mol %) a (Å) c (Å) V (Å)3 c/a Size (nm) (Strain ±1)× 10−3
0 (600 ◦C) 3.216 5.198 46.559 1.62 62 1.5
4 (280 ◦ C 3.216 5.198 46.559 1.62 26 2.0
4 (600 ◦C) 3.217 5.196 46.561 1.62 42 2.0
8 (280 ◦ C) 3.220 5.205 46.737 1.62 27 3.0
8 (600 ◦C) 3.220 5.205 46.737 1.62 58 1.0
Ni-doping
The x-ray diffraction patterns for undoped, 4 and 8 mol % Ni-doped ZnO at pH 5 with 600 ◦C
annealing are shown in Figure 5.17a. A secondary phase was observed, with peaks at 2θ angles
43.16◦ for 4 and 8 mol % with an extra peak at 37.14◦ for 8 mol% which was indexed to the
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 79
monoclinic phase of nickel oxide (NiO) with ICSD reference code 98-007-6670. Figure 5.17b
shows the x-ray diffractograms of undoped, 4 and 8 mol % Ni-doped ZnO at pH 3. Here, too,
the extra peaks at 2θ angles of 43.16 ◦ (4 mol %) and 28.38, 37.18, 40.53 and 43.17◦ (8 mol %)
showed the existence of the same monoclinic phase of NiO.
(a) Ni-doping at pH 5 (b) Ni-doping at pH 3
Figure 5.17: XRD pattern for 4 and 8 mol % Ni-doped ZnO at 600◦ annealing.
Table 5.32: Lattice parameters, crystallite size and strain for 4 and 8 mol % Ni-doping at pH 5
Doping (Mol %) a (Å) c (Å) V (Å)3 c/a Size (nm) (Strain ±1)× 10−3
0 (600 ◦C) 3.216 5.198 46.559 1.62 62 1.5
4(280 ◦C) 3.221 5.207 46.784 1.62 32 2.0
4 (600 ◦C) 3.217 5.200 46.606 1.62 35 1.0
8(280 ◦C) 3.221 5.207 46.784 1.62 32 1.0
8 (600 ◦C) 3.215 5.193 46.468 1.62 65 0.0
Table 5.33: Lattice parameters, crystallite size and strain for 4 and 8 mol % Ni-doping at pH 3
Doping (Mol %) a (Å) c (Å) V (Å)3 c/a Size (nm) Strain ×10−3
0 (600 ◦C) 3.216 5.198 46.559 1.62 62 1.5
4(280 ◦C) 3.221 5.206 46.775 1.62 37 2.0
4 (600 ◦C) 3.217 5.199 46.596 1.62 42 1.4
8(280 ◦C) 3.221 5.208 46.793 1.62 34 2.0
8 (600 ◦C) 3.217 5.198 46.587 1.62 43 1.8
Tables 5.32 and 5.33 show the lattice parameters, volume of the unit cell, c/a ratio, crystal-
lite size and strain at pH 3 and pH 5. The crystallite size increased with increasing annealing
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 80
temperature. At 600 ◦C, there was a reduction in strain for both 4 and 8 mol%, and the unit cell
volume reduced as well. The c/a ratio was however constant.
The values obtained for the lattice parameters at 600 ◦C, compare well with that of the
undoped ZnO annealed at 600 ◦C. This is also the case for the aspect ratio (Table 5.34), where
the values compare well with that of the undoped ZnO.
Aspect Ratio
Table 5.34 lists the values of the aspect ratio (AR) for all the samples. AR was found to decrease
for the annealed samples, approaching the value of 0.80 of the undoped ZnO.
Table 5.34: Aspect Ratios for 4 and 8 mol% Doping. A.T. = Annealing Temperature.
Dopant Doping pH 5 pH 5 pH3 pH 3
Mol % A.T. = 280 ◦C A.T. = 600 ◦C A.T. = 280 ◦C A.T. = 600 ◦C
None 0 1.08 0.80 1.08 0.80
4 1.20 0.94 1.20 0.70
Fe
8 1.05 1.09 1.07 0.70
4 0.89 0.83 0.99 0.70
Mn
8 0.88 0.87 0.88 0.73
4 1.20 0.94 1.20 0.70
Co
8 1.05 1.09 1.07 0.70
4 1.04 0.87 1.04 0.76
Ni
8 1.03 0.72 1.10 0.78
5.4.2 Summary
The x-ray diffraction peaks were sharper at 600 ◦C annealing, and along with the fact that the
crystallographic parameters did not deviate much from those of the undoped ZnO at 600 ◦C,
indicates that annealing the doped samples resulted in crystals of better quality, as expected.
5.4.3 Band Gap Energies
Fe-doping at pH 5 and 3
Table 5.35 shows the band gap energies calculated from the absorption edge of the UV-VIS
absorption spectra. At pH 5 and 600 ◦C annealing, the band gap energy reduced for both 4 and
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 81
8 mol % doping. At pH 3, the band gap energies also reduced for 4 mol % but increased for 8
mol % doping.
In comparison, the band gap energies calculated from the near band emission of the PL
spectra at pH 3 (Table 5.36) showed an increase in the band gap energy at 4 mol % and a
decrease at 8 mol %, while the band gap energies at pH 5 were fairly constant.
Table 5.35: Band gap from UV absorption spectra for 4 and 8 mol % Fe-doping
Doping pH 5 pH 3
Mol % Band Gap Energy (±0.01 eV) Band Gap Energy (±0.01 eV)
4 (280◦C) 2.91 3.15
4 (600◦C) 2.76 3.00
8 (280◦C) 2.95 3.20
8 (600◦C) 2.87 3.00
Table 5.36: Band gap from PL emission spectra for 4 and 8 mol % Fe-doping
Doping pH 5 pH 3
Mol % Band Gap Energy (±0.03 eV) Band Gap Energy (±0.03 eV)
4 (280◦C) 3.30 3.27
4 (600◦C) 3.29 3.32
8 (280◦C) 3.30 3.38
8 (600◦C) 3.29 3.30
Mn-doping at pH 5 and 3
Table 5.37 shows the band gap energies for Mn-doping at pH 5 and pH 3. At pH 5 for 4 mol
% doping (600 ◦C annealing) there was drop in the band gap energy from 2.65 eV to 1.93 eV,
while at 8 mol % doping, the reduction in band gap energy was from 2.86 eV to 2.55 eV. At pH
3, there was again a reduction in the band gap energy for 4 mol % doping, but at 8 mol % an
increase was recorded from 2.86 eV to 2.91 eV.
However, the PL emission spectra (Table 5.38) showed a rather different picture, with the
band gap energy changing by only 0.01 eV and 0.02 eV nm at pH 5 and pH 3 respectively.
Here again, there is the observation that the band gap energy from the UV-VIS absorption
spectrum was sensitive to the presence of secondary crystalline phases. It will be recalled from
Figure (5.15) that annealing the samples at 600 ◦C showed the presence of Mn3O4 and MnO2.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 82
Table 5.37: Band Gap Energy from UV-VIS Absorption for 4 and 8 mol % Mn-doping
Doping pH 5 pH 3
Mol % Band Gap Energy (±0.01) eV Band gap Energy (±0.01) eV
4 (280◦C) 2.65 2.86
4 (600◦C) 1.93 2.20
8 (280◦C) 2.86 2.86
8 (600◦C) 2.55 2.91
Table 5.38: Band gap Energy from PL Emission for 4 and 8 mol % Mn-doping
Doping pH 5 pH 3
Mol % Band Gap Energy (±0.03) eV Band Gap Energy (±0.03) eV
4 (280◦C) 3.29 3.31
4 (600◦C) 3.27 3.32
8 (280◦C) 3.31 3.31
8 (600◦C) 3.29 3.32
Co-doping at pH 5 and 3
Table 5.39 shows the band gap energies for 4 and 8 mol % Co-doping at pH 5 and pH 3.
The band gap energy from the UV-VIS absorption spectra for 4 and 8 mol % at both pH 5
and 3 reduced with increasing annealing temperature. Table 5.40 shows the band gap energies
calculated from the near band emission of the PL spectrum. A decrease in band gap energy was
observed at 4 mol % for samples at pH 5 with an increase for the samples at pH 3. However,
the samples at 8 mol % for both the pH 5 and 3 for the 8 mol % remained constant.
Table 5.39: Band Gap Energy from UV-VIS Absorption for 4 and 8 mol % Co-doping
Doping pH 5 pH 3
Mol % Band Gap Energy (±0.01) eV Band gap Energy (±0.01) eV
4 (280◦C) 3.14 2.88
4 (600◦C) 2.69 2.12
8(280◦C) 2.86 3.06
8 (600◦C) 2.14 2.62
Ni-doping at pH 5 and 3
UV absorption spectra for 4 and 8 mol % Ni-doping at pH 5 showed narrowing of band gap with
increasing annealing temperature (Table 5.41). Table 5.42 shows band gap energies calculated
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 83
Table 5.40: Band Gap Energy from PL Emission for 4 and 8 mol % Co-doping
Doping pH 5 pH 3
Mol % Band gap Energy (±0.03) eV Band gap Energy (±0.03) eV
4 (280◦C) 3.32 3.31
4 (600◦C) 3.26 3.36
8(280◦C) 3.30 3.35
8 (600◦C) 3.31 3.34
from the near band emission of the PL spectrum 4 and 8 mol % Ni-doping. It can be seen that
the values were constant within the range of error at both pH 3 and pH 5.
Table 5.41: Band Gap Energy from UV-VIS Absorption for 4 and 8 mol % Ni-doping
Doping pH 5 pH 3
Mol % Band Gap Energy (±0.01 eV) Band gap Energy (±0.01 eV)
4 (280◦C) 3.06 2.90
4 (600◦C) 3.04 2.65
8 (280◦C) 3.09 3.01
8 (600◦C) 3.08 2.68
Table 5.42: Band Gap Energy from PL Emission for 4 and 8 mol % Ni-doping
Doping pH 5 pH 3
Mol % Band gap Energy (±0.03 eV) Band gap Energy (±0.03 eV)
4 (280◦C) 3.25 3.25
4 (600◦C) 3.28 3.24
8 (280◦C) 3.26 3.24
8 (600 ◦C) 3.26 3.22
5.4.4 Summary
The band gap energies calculated from the PL emission spectra for all the doped samples were
constant within the range of error. This was regardless of whether there were interband states or
not. However, all the doped samples had band gap energies that were larger than (or equal to,
in the case of nickel doping) to the band gap energy of the reference and undoped ZnO at 3.20
eV and 3.23 eV respectively.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 84
(a) Fe-doped (b) Mn-doped
(c) Co-doped (d) Ni-doped
Figure 5.18: PL emission spectra for 4 and 8 mol % doped ZnO at pH 3
The band gap energies calculated from the absorption edge of the UV-VIS spectrum signif-
icantly varied. It can be seen from the data that there were reductions in the band gap energy
for doped samples for which secondary phases were observed in the X-ray data. This leads
(for the second time) to the conclusion that the secondary crystalline phases variously of the
iron, manganese, cobalt, and nickel oxides served to reduce the absorption edge as measured by
UV-VIS absorption.
Finally, annealing did not have any effect of diminishing the secondary crystalline phases.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 85
(a) Fe-doped (b) Mn-doped
(c) Co-doped (d) Ni-doped
Figure 5.19: PL emission spectra for 4 and 8 mol % doped ZnO at pH 5
5.5 Computational Results
The computational results from the DFT-GGA calculations, being the lattice parameters and the
band gap energies are presented and discussed in this section.
Doping was carried out on a 36-atom supercell. By replacing one zinc atom with the re-
quired transition metal atom, doping at 5.56 at. % was achieved. Replacing two zinc atoms led
to doping at 11.11 at. %.
5.5.1 Undoped ZnO
Figure 5.20 shows the band structure and density of states for undoped ZnO. The Fermi energy
was set to zero in the figure. The calculation was performed on a 4-atom unit cell of ZnO.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 86
The highest valence band and the lowest conduction band both occurred at the Γ k-point,
which shows that ZnO is a direct band gap semiconductor. The calculated band gap energy was
0.7537 eV. Similar values have been reported in the literature (Janotti and van de Walle (2007)).
As has been discussed in Chapter 3, the method of DFT-LDA and DFT-GGA underestimates
the band gap energy, and this result is not surprising.
The lattice parameters and the band gap energy obtained from this calculation will be pre-
sented alongside the data from the doped samples.
(a) Band structure (b) Density of states
Figure 5.20: Band structure and density of states for undoped ZnO
Fermi energy of band structure is set to zero
5.5.2 Lattice Parameters
Fe-doping
The lattice parameters for undoped and Fe-doped ZnO from the DFT-GGA calculations are
presented in Table 5.43. The unit cell volume increased with increasing doping concentration,
while the c/a ratio remained approximately constant.
Table 5.43: Lattice parameters for Fe-doped ZnO
Doping (at. %) a (Å) c (Å) V (Å)3 c/a
0 3.280 5.291 49.288 1.61
5.56 3.279 5.292 49.278 1.61
11.11 3.301 5.282 49.849 1.60
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 87
Mn-doping
From Table 5.44, which details the values of the lattice parameters for Mn-doping, it can be
observed that the unit cell volume increased with doping concentration. The c/a ratio was
however constant.
Table 5.44: Lattice parameters for Mn-doped ZnO
Doping (at. %) a (Å) c (Å) V (Å)3 c/a
0 3.280 5.291 49.288 1.61
5.56 3.280 5.306 49.426 1.62
11.11 3.284 5.318 49.683 1.62
Co-doping
Table 5.45 lists the lattice parameters for the Co-doped ZnO. It can be seen that the volume
decreased from 49.288 Å3 at 0 % doping to 49.157 Å3 at 5.56 % doping, then increased again
at 11.11 % doping to 49.162 Å3. The c/a ratio was however constant.
Table 5.45: Lattice parameters for Co-doped ZnO
Doping (at. %) a (Å) c (Å) V (Å)3 c/a
0 3.280 5.291 49.288 1.61
5.56 3.281 5.274 49.157 1.61
11.11 3.278 5.283 49.162 1.61
Ni-doping
In Table 5.46, the lattice parameters for the Ni-doped ZnO are listed. The volume decreased
monotonically from 0 % doping to 11.11 % doping. The c/a ratio changed from 1.61 at 5.56 %
doping to 1.62 at 11.11 % doping. Approximately, the c/a ratio was constant.
5.5.3 Band Gap Energy
The band gap energies from the computational calculations are presented in Table 5.47. It can
be seen that for each dopant, there was an increase in the band gap energy with increasing
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 88
Table 5.46: Lattice parameters for Ni-doped ZnO
Doping (at.%) a (Å) c (Å) V (Å)3 c/a
0 3.280 5.291 49.288 1.61
5.56 3.282 5.277 49.218 1.61
11.11 3.280 5.291 49.183 1.62
concentration. Again, it must be noted that the method of calculation underestimates the value
of the band gap energy. In this analysis, the important item is the variation of the band gap
energy with doping concentration. Hence, the data is examined for the trend of band gap energy
change with doping concentration and not the absolute values of the band gap energies. The
band gap energies were estimated by subtracting the value of at the maximum valence band
band from that of the minimum conduction band.
It must be mentioned that this behaviour of increasing band gap energies with the doping
concentration was observed in the experimental results, where the band gap energy for the
doped samples was found to increase.
Table 5.47: Band Gap Energy
Doping Fe-doped Mn-doped Co-doped Ni-doped
at. % Energy (eV) Energy (eV) Energy (eV) Energy (eV)
0 0.7537 0.7537 0.7537 0.7537
5.56 0.9027 0.8753 0.9454 0.9408
11.11 1.0348 1.0137 1.0046 1.0838
5.5.4 Band Structure
The band structure for the doped samples are presented in Figures ??, 5.22 and 5.23. It can be
seen from the band structures of all dopants that, the doped ZnO samples have direct band gap:
the maximumvalence and minimum conduction bands occured at the γ point. The Fermi energy
in each figure has been set to zero.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 89
(a) Fe-doped (b) Mn-doped
(a) Co-doped (b) Ni-doped
Figure 5.22: Band structure for 5.56 at. % doping
Fermi level is set to zero
5.5.5 Density of states
Figures 5.24 and 5.25 show the calculated density of states for the doped samples. The density
of state plots show the occupied and unoccuppied states in the materials. The peaks in the
density of states plots represent the occuppied states.
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 90
(a) Fe-doped (b) Mn-doped
(c) Co-doped (d) Ni-doped
Figure 5.23: Band structure for 11 at. % doping
Fermi level is set to zero
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 91
(a) Fe-doped (b) Mn-doped
(c) Co-doped (d) Ni-doped
Figure 5.24: Density of states for 5.56 at. % doped ZnO
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5. RESULTS AND DISCUSSION 92
(a) Fe-doped (b) Mn-doped
(c) Co-doped (d) Ni-doped
Figure 5.25: Density of states for 11 at. % doping
University of Ghana  http://ugspace.ug.edu.gh
Chapter 6
Conclusion
Undoped and transition metal (Fe, Mn, Co and Ni) doped ZnO were synthesised and anal-
ysed using x-ray powder diffraction, UV-VIS absorption spectroscopy, and PL emission spec-
troscopy. The crystal lattice parameters and the band gap energies of the doped and undoped
ZnO were also calculated using DFT-GGA on QE.
In the previous chapter, the crystallographic parameters and the band gap energies for the
undoped and doped ZnO samples were presented. The first set of data, detailing synthesis at
pH 5 and annealing at 280 ◦C, showed that doping with TMs invariably caused an increase in
the unit cell volume. The fact that there were increases in the volume of the unit cell regardless
of the size of the dopant cation suggests that the mechanism for doping might not have been
a simple drop-in replacement for Zn2+. Also, the dopant cations have different solubilities
in ZnO which would lead to differences in doping efficiency. From the W-H analysis, the
crystallite sizes remained in the same range within the limit of error, while the strain did not
vary significantly nor appear to follow any trend.
Band gap energies were determined using UV-VIS absorption and PL emission spectroscopy.
With regards to the values from UV-VIS measurements, the band gap energies decreased with
increasing doping concentration for all the dopants except nickel. This was different from
the values obtained from the PL measurements, where the band gap energies increased for all
dopants again with the exception of nickel. The difference was attributed to the presence of sec-
ondary crystalline phases within the samples as confirmed by the x-ray diffractograms. These
secondary phases would have had the effect of creating an apparent reduction in the band gap
93
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 6. CONCLUSION 94
as a result of the method used. The band gap energies from the PL spectra, however, which
were determined from the near band edge emission increased with doping, again with the ex-
ception of nickel doping. Data from the PL emission is deemed to be more reliable and this
work concludes that the band gap energies were increasing.
In order to reduce the effect of the secondary phases, synthesis was carried out at pH 3.
It was noticed that this reduced the occurrence of secondary phases. It also provided partial
confirmation of the assertion that the presence of secondary phases made the UV-VIS band
gap energies of the doped ZnO appear smaller. This was clear from the fact that at pH 3,
along with the decrease in the secondary phases, the band gap energy from UV-VIS increased,
approaching the value calculated from the photoluminescence spectra. This was the case for all
dopants except nickel and cobalt at 4 mol %.
The aspect ratios were greater at pH 3 than at pH 5, indicating that growth along the c-axis
was favoured at higher pH.
In order to improve their crystallinity, 4 and 8 mol % doped samples were annealed at 600
◦C. The XRD results showed better crystals, with sharper peaks, larger crystallite sizes with
smaller error ranges, and smaller strain. The aspect ratios at pH 3 and 600 ◦C annealing were
close to 0.7, which is the value for the ZnO reference. With regards to the band gap energies, the
values calculated from UV-VIS reduced as has been earlier explained, whereas the values from
PL emission were constant, but larger than the band gap energy for the synthesised undoped
ZnO.
It is concluded that the best results were for synthesis at pH 3 with annealing at 600 ◦C.
It is interesting to note that the results from the computational work match those from the ex-
perimental work. The band gap energies calculated for the doped samples invariably increased
at all doping concentrations. This was confirmed by experiment. Furthermore, the increase
in the unit cell volume observed experimentally for iron and manganese doping were also ob-
served in the computational results. With regards to the computational results for cobalt and
nickel doping, the unit cell volume however decreased. In the case of the experimental work,
the unit cell volume showed a marginal increase for nickel, whereas an increase was recorded
for cobalt. This observation cannot be fully explained from the data in this work, but might be
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 6. CONCLUSION 95
related to the higher solubilities of cobalt and nickel in ZnO, leading to a higher uptake of the
ion experimentally and probably greater lattice strain.
6.0.1 Prospects
The work presented here lays the groundwork for more detailed research into the field, both
computationally and experimentally. The initial synthesis method had some shortfalls identi-
fied and improvements proposed included the use of higher pH solutions and post-synthesis
annealing at 600 ◦C. Photoluminescence emission spectroscopy was also shown to be the pre-
ferred method for measuring the band gap energy. Further work can be carried out to improve
the doping of ZnO using this hydrothermal method.
In principle, the trend of the band gap energies obtained from the absorption edge of the UV-
Vis and the emission edge of the photoluminescence spectrometers should not be different even
though the values may differ. Hence there will be the need to reinvestigate the measurements or
try other equipment to validate the results.
With regards to the computational work, it is proposed that investigations be carried out at
more dilute doping concentrations and also to study the effect of symmetry in the placement of
the dopant cation within the supercell.
University of Ghana  http://ugspace.ug.edu.gh
References
Ahuja, R., Fast, L., Eriksson, O., Wills, J. M., and Johansson, B. (1998). Elastic and high
pressure properties of ZnO. Journal of Applied Physics, 83:8065.
Arya, S. K., Saha, S., Vick, J. E. R., Gupta, V., Bhansali, S., and Singh, S. P. (2012). Recent
advances in ZnO nanostructures and thin films for biosensor applications: Review. Analytica
Chimica Acta, 737:1 – 21.
Badaeva, E., Feng, Y., Gamelin, D. R., and Li, X. (2008). Investigation of pure and Co2+ -
doped ZnO quatum dot electronic structures using the density functional theory:chosing the
right functional. New Journal of Physics, 10:055013(12).
Bae, K. H., Chung, H. J., and Park, T. G. (2011). Nanomaterials for cancer therapy and imaging.
Molecules and Cells, 31:295 – 302.
Barker, P. J. and Branch, A. (2008). The interaction of modern sunsrceen formulations with
surface coatings. Progress in Organic Coatings, 62:313 – 320.
Baroni, S., de Gironcoll, S., and Corso, A. D. (2001). Phonons and related crystal properties
from density-functional perturbation theory. Reviews of Modern Physics, 73:515(48).
Bassani, F. and Parraviccini, G. P. (1975). Electronic states and Optical Transitions in solids.
Pergamon Press, 1 edition.
Bhat, S. V. and Deepak, F. L. (2005). Tuning the bandgap of ZnO by substitution with Mn2+,
Co2+ and Ni2+. Solid State Communications, 135:345 – 347.
Cao, W. (2007). Synthesis of Nanomaterials by High Energy Ball Milling.
96
University of Ghana  http://ugspace.ug.edu.gh
REFERENCES 97
Cardona, M. (1969). Modulation spectroscopy, in Solid State Physics. Academic Press.
Cavazzoni, C. and Chiarotti, G. L. (1999). A parallel and modular deformable cell Car-
Parrinello code. Computer Physics Communications, 123:56 –76.
Chang, P. C., Fan, Z., Wang, D., Tseng, W. Y., Chiou, W. A., Hong, J., and Lu, J. G. (2004).
ZnO nanowires synthesized by vapor trapping CVD method. Chemical Materials, 16:5133 –
5137.
Chen, D., Jiao, X., and Cheng, G. (2000). Hydrothermal synthesis of zinc oxide powdrs with
different morphologies. Solid State Communications, 113:363–366.
Chey, C. O. (2015). Synthesis of ZnO and transition metals doped ZnO nanostructures, their
characterization and sensing applications. PhD thesis, Linköping University – Physical Elec-
tronics and Nanotechnology Department.
Chu, S., Li, D., Chang, P.-C., and Lu, J. G. (2011). Flexible Dye-Sensitized Solar Cell Based
on Vertical Nanowire Arrays. Nanoscale Reaserch Letters, 6(38).
Deka, S. and Joy, P. A. (2007). Synthesis and magnetic properties of Mn doped ZnO nanowires.
Solid State Communications, 142:190 – 194.
Devaramani, B. S., S, R. Y., Manjasetty, B. A., and Nair, T. R. G. (2008). The Novelty of Syn-
theses and Varied Applications of ZnO nano systems. International Conference on Frontiers
in Chemical Research, pages 206 – 212.
Dietl, T., Ohno, H., Matsukukra, F., Cibert, J., and Ferrand, D. (2000). ZenerModel Description
of Ferromagnetism in Zinc – Blende Magnetic Semiconductors. Science, 287:1019 – 1022.
Ding, L., Zhang, R., and Fan, L. (2013). Electrochemical route to the synthesis of ZnO mi-
crostructures: its nestlike structure and holding of ag particles. Nano Express, 8(78):1 –
7.
Ebrahimizadeh, M., Hosseini, S. M., Kakhki, E. A., and Kompany, A. (2010). Structural
and optical properties of zinc oxide nanopowders doped with Mn. Physica Status Solidi
C, 7(6):1595–1598.
University of Ghana  http://ugspace.ug.edu.gh
REFERENCES 98
Friedrich, C., Müller, M. C., and Blügel, S. (2011). Band convergence and linearization error
correction of all-electron GW calculations: The extreme case of zinc oxide. Physical Review
B, 83:081101(4).
Fuxue, W., Xiaolong, C., Dawei, Y., Zhaomin, Z., and Xiaofeng, G. (2014). Luminescence
properties of tetrapod ZnO nanostructures. Journal of Semiconductors, 35(6):063004.
Gandomani, S. K., Yousefi, R., Sheini, F. J., and Huang, N. M. (2014). Optical and electrical
properties of p - type Ag - doped ZnO nanostructures. Ceramics International, 40:7957 –
7963.
George Kresse, O. D. and Diebold, U. (2003). Competing stabilization mechanism for polar
ZnO (0001) - Zn surface. Physical Review B, 68:245409.
Gondoni, P., Ghidelli, M., Fonzo, F. D., Carminati, M., Russo, V., Bassi, A. L., and Casari, C. S.
(2012). Structure – dependent optical and electrical transport properties of nanostructured Al
– doped ZnO. Nanotechnology, 23:365706.
Grosso, G. and Parravicini, G. P. (2013). Solid State Physics. Academic Press, 2 edition.
Hao, Y.-M., Zhou, S.-Y., Yuan, R.-J., Zhu, G.-Y., and Li, N. (2012). Structural, optical and mag-
netic studies of manganese–doped zinc oxide hierarchical microspheres by self–assembly of
nanoparticles. Nanoscale Research letters, 100(7):1 – 9.
He, R., Hocking, R. K., and Tsuzuki, T. (2012). Co - dpoed ZnO nanopowders: Location
of cobalt and reduction in photocatalytic activity. Materials and Chemistry and Physics,
132:1035 – 1040.
He, R., Tang, B., and That, C. T. (2013). Physical structure and optical properties of Co - doped
ZnO nanoparticles prepared by co - precipitation. J Nanopart. Res, 15(2030):1 –8.
Heinze, S., Krtschil, A., Bläsing, J., Hempel, T., Veit, P., Dadgar, A., Christen, J., and Krost,
A. (2007). Homoepitaxial growth of ZnO by metalorganic vapor phase epitaxy in two -
dimensional growth mode. Journal of Crystal Growth, 308:170 – 175.
University of Ghana  http://ugspace.ug.edu.gh
REFERENCES 99
Huang, H., Fang, G., Mo, X., Long, H., Yuan, L., Dong, B., Meng, X., and Zhao, X. (2009).
Zno based ultraviolet light emitting diodes with a low operation voltage. IEEE electron device
Letters, 30(10):1063 – 1065.
Iqbal, J., Janjua, R. A., and Jan, T. (2014). Structural, optical and maggnetic properties of
Co-doped ZnO nanoparticles prepared via a wet chemical route. International Journal of
Modern Physics B, 28(25):1450158.
Ischenko, V., Polarz, S., Grote, D., Stavarache, V., Fink, K., and Driess, M. (2003). Zinc Oxide
Nanoparticles with Defects. Advanced Functional Materials, 15:1945 –1954.
Ive, T., Yaacov, T. B., Mural, A., Asamizu, H., van de Walle, C. G., Mishra, U., DenbBaara,
S. P., and Speck, J. S. (2008). Metalorganic chemical vapor deposition of ZnO(0001) thin
films on GaN(0001) templates and ZnO(0001) substrate. Physica Status Solidi, 5(9):3091 –
3094.
Ivill, M., j Pearton, S., Rawal, S., Leu, L., Sadik, P., Das, R., Hebard, A. F., Chisholm, M.,
Budai, J. D., and Norton, D. P. (2008). Structure and magnetism of cobalt - doped ZnO thin
films. New Journal of Physics, 10:065002.
Jaffe, J. E. and Hess, A. C. (1993). Hartree-Fock study of phase changes in ZnO at high pressure.
Physical Review B, 48(11):7903 – 7909.
Janotti, A. and van de Walle, C. G. (2007). Native point defects in ZnO. Physical Review B,
76:165202 – 165222.
Janotti, A. and van de Walle, C. G. (2009). Fundamentals of zinc oxide as a semiconductor.
Report on Progress in Physics, 72(126501):29.
Jiangfeng, G., Zhaoming, D., Qingping, D., Yuan, X., and Weihua, Z. (2010). Controlled
synthesis of ZnO nanostructures by electrodeposition method. Journal of Nanomaterials,
pages 1 – 7.
Jin, Z., Murakami, M., Fukumura, T., Matsumoto, Y., Ohtomo, A., Kawasaki, M., and Koinuma,
University of Ghana  http://ugspace.ug.edu.gh
REFERENCES 100
H. (2000). Combinatorial laser MBE synthesis of 3d ion doped epitaxial ZnO thin films.
Journal of Crystal Growth, 214 - 2015.
Jung, Y. S., Kononenko, O., Kim, J. S., and Choi, W. K. (2005). Two - dimensional growth
of ZnO epitaxial films on c - Al2O3 (0001) substrates with optimized growth temperature
and low - temperature buffer layer by plasma - assisted molecular beam epitaxy. Journal of
Crystal Growth, 274:418 – 424.
Jyoti, M., Vijay, D., and Radha, S. (2013). To study the role of temperature and sodium hy-
droxide concentration in the synthesis of zinc oxide nanoparticles. International Journal of
Scientific and Research Publications, 3(11):2250 – 3153.
Karmakar, D., Mandal, S. K., Kadam, R. M., Paulose, P. L., Rajarajan, A. K., Nath, T. K., Das,
A. K., Dasgupta, I., and Das, G. P. (2007). Ferromagnetism in Fe – doped ZnO nanocrystals:
Experiment and theory. Physical Review B, 75:144404.
Karmakar, R., Neogi, S. K., Banerjee, A., and Bandopadhyay, S. (2012). Structural; morpjol-
ogy; optical and magnetic properties of Mn doped ferromagnetic ZnO thin film. Applied
Surface Science, 263:671 – 677.
Karzel, H., Potzel, W., Köfferlein, M., Schiessl, W., Steiner, M., Hiller, U., Kalvius, G. M.,
Mitchell, D. W., Das, T. P., Blaha, P., Schwarz, K., and Pasternak, M. P. (1996). Lattice
dynamics and hyperfine interactions in ZnO and ZnSe at high external pressures. Physical
Review B, 53(17):11425–11438.
Kharroubi, B., Baghdad, R., Abdiche, A., aand M Bousquet, M. B., Zeinert, A., Marssi, M. E.,
Zellama, K., and Hamzaoui, S. (2012). Mn doping effect on the structural properties of
ZnO - nanastructured films deposited by the ultrasonic spray pyrolysis method. Phys. Scr.,
86(015805):1 – 7.
Kim, Y. M., Yoon, M., Park, I. W., Park, Y. J., and Lyou, J. H. (2004). Synthesis and magnetic
properties of Zn1−xMnxO films prepared by the sol - gel method. Solid State Communica-
tions, 129(3):175 – 178.
University of Ghana  http://ugspace.ug.edu.gh
REFERENCES 101
Klingshirn, C. F. (2007). Zno: Material, Physics and Applications. Chemphyschem, 8:782 –
803.
Ko, S. C., Kim, Y. C., Lee, S. S., Choi, S. H., and Kim, S. R. (2003). Micromachined piezo-
electric membrane acoustic device. Sensors and Actuators A, 103:130 – 134.
Kohan, A. F., Ceder, G., Morgan, D., and de Walle, C. G. V. (2000). First – principles study of
native point defects in ZnO. Physical Review B, 61(22):15019.
Kohn, W. and Sham, L. J. (1965). Self-Conistent Equations Including Exchange and Correlation
Effects*. Physical Review, 140(4A).
Kononenko, O. V., Redkin, A. N., Baranov, A. N., Panin, G. N., Kovalenko, A. A., and Firsov,
A. A. (2012). ZnO nanorods: Synthesis by catalyst - free CVD and thermal growth from salt
compites and application to nanodevices, chapter 3. Nanotechnology and Nanomaterials.
Krishnakumar, T., Jayaprakash, R., Pinna, N., Donato, N., Bonavita, A., Micali, G., and Neri, G.
(2009). CO gas sensing of nanostructures synthesised by an assisted microwave wet chemical
route. Sensors and Actuators B, 143:198 – 204.
Kumar, S., Mukherjee, S., Singh, R. K., Chatterjee, S., and Ghosh, A. K. (2011). Structural and
optical properties of sol - gel derived nanocrystalline Fe - doped ZnO. Journal of Applied
Physics, 110:103508.
Kuzmany, H. (2008). Solid-State Spectroscopy: An Introduction. Springer, 2nd edition.
Laaksonen, K. (2009). Computational studies of III-V compound semiconductors. PhD thesis,
Helsinki University of Technology - Department of Applied Physics.
Li, H., Avrutin, V., Izyumskaya, N., Ö zgur, Ü., and Morkoç, H. (2010). Transparent conducting
oxides for electrode applications in light emitting and absorbing devices. Superlattices and
Microstructures, 48(5):458 – 484.
Liu, B. and Zeng, H. C. (2009). Direct growth of enclosed ZnO nanotubes. Nano Research,
2:201 – 209.
University of Ghana  http://ugspace.ug.edu.gh
REFERENCES 102
Look, D. C., Reynold, D. C., Litton, C. W., Jones, R. L., Eason, D. B., and Cantewell, G. (2002).
Characterization of homoepitaxial p – type ZnO grown by molecular beam epitaxy. Applied
Physics Letters, 81(10):1830 –1832.
Maksimov, O. (2010). Recent advances and novel approaches of p – type doping of zinc oxide.
Rev. Adv. Material Science, 24:26 – 34.
Malloci, G., Chiodo, L., Rubio, A., and Mattoni, A. (2012). Structural and Optoelectronic Prop-
erties of Unsaturated ZnO and ZnS Nanoclusters. Journal of Physical Chemistry, 116:8741
– 8746.
Mang, A., Reiman, K., and Rübenacke, S. (1995). Band gaps, crystal - field splitting, spin -
orbit coupling and exciton binding energies in ZnO under hydrostatic pressure. Solid State
Communications, 94(4):251 – 254.
McCluskey, M. D. and Jokela, S. J. (2009). Defects in ZnO. Journal of Applied Physics,
106:071101.
Meyer, B. K., Alves, H., Hoffmann, D. M., Kriegseis, W., Forster, D., Bertram, F., Christen, J.,
Hoffmann, A., Stra"sburg, M., Dworzak, M., Haboeck, U., and Rodina, A. V. (2004). Bound
exciton and donor–acceptor pair recombinations in ZnO. phys. stat. sol. (b), 241(2):231–260.
Moezzi, A., McDonagh, A. M., and Cortie, M. B. (2012). Zinc oxide particles: Synthesis,
properties and applications. Chemical Engineering Journal, 185–186:1–22.
Morgan, B. and Lahav, O. (2007). The effect of pH on the kinetics of spontaneous Fe(II) oxi-
dation by O2 in aqueous solution –basic principles and simple heuristic description. Chemo-
sphere, 68:2080–2084.
Mote, V. D., Purushotham, Y., and Dole, B. N. (2011). Structural and morphological studies on
Mn substituted ZnO nanometer—sized crystals. Crystal Research Technology, 46(7).
Nayak, J., Kimura, S., and Nozaki, S. (2009). Enhancement of the visible luminescence from
the ZnO nanocrystals by Li and Al co – doping. Journal of Luminiscence, 129:12 – 16.
University of Ghana  http://ugspace.ug.edu.gh
REFERENCES 103
Nguyen, K. T. (2011). Targeted Nanoparticles for Cancer Therapy: Promises and Challenges.
Journal of Nanomedicine and Nanotechnology, 2(5):1 – 2.
Nie, L., Gao, L., Feng, P., Zhang, J., Fu, X., Liu, Y., Yan, X., and Wang, T. (2006). Three
Dimensional Functionalised Tetrapod-like ZnO Nanostructures for Plasmid DNA delivery.
Small, 2(5):621–625.
Nikalje, A. P. (2015). Nanotechnology and its Applications to Medicine. Medicinal Chemistry,
5(2):081 – 089.
Özgür, Ü., Alivov, Y. I., Liu, C., Ateke, A., Resshchikov, M. A., Doǧan, S., Avrutin, V., Cho,
S. J., and Morkoç, H. (2005). A comprehensive review of ZnOmaterials and devices. Journal
of Applied Physics, 98:041301.
Panigraphy, B., Aslam, M., and Bahadur, D. (2012). Effect of Fe doping concentration on
optical and magnetic properties of ZnO nanorods. Nanotechnology, 23:115601.
Pasquarello, A., Laasonen, K., Car, R., Lee, C., and Vanderbilt, D. (1992). Ab initio Molec-
ular Dynamics for d-Electron Systems: Liquid Copper at 1500 K. Physical Review Letters,
69(13):1982 – 1985.
Payne, M. C., Teter, M. P., and Allan, D. C. (1992). Iterative minimization technique for ab initio
total-energy calculations: molecular dynamics and conjugate gradients. Review of Modern
Physics, 64(4):1045–1097.
Pearton, S. J., Norton, D. P., Ivill, M. P., Hebard, A. F., Zavada, J. M., Chen, W. M., and
Buyanova, I. A. (2007). ZnO doped transition metal ions. IEEE Transactions on Electron
Devices, 54(5):1040 – 1048.
Perdew, J. P., Chevary, J. A., Vosko, S. H., Jackson, K. A., Pederson, M. R., Singh, D. J., and
Fiolhais, C. (1992). Atoms, molecules, solids and surfaces: Applications of the generalized
gradient approximation for exchange and correlation. Physical Review B, 46(11):6671(17).
Raja, K., Ramesh, P. S., and Geetha, D. (2014). Structural FTIR and photoluminiscence studies
University of Ghana  http://ugspace.ug.edu.gh
REFERENCES 104
of Fe doped ZnO nanopowder by co-precipitation method. Spectrochimica Acta Part A:
Molecular and Biomolecular Spectroscopy, 131:183 – 188.
Rasmussen, J. W., Martinez, E., Louka, P., and Winget, D. G. (2010). Zinc oxide nanoparticles
for selective destruction of tumor cells and potential for drug delivery applications. Expert
Opin Drug Deliv., 7(9):1063 – 1077.
Reynolds, D. C., Look, D. C., Jogai, B., Litton, C. W., Cantwell, G., and Harsch, W. C. (1999).
Valence-band ordering in ZnO. Physical Review B, 60(4):2340–2344.
Sahai, A., Kumar, Y., Agarwal, V., Méndez, S. F. O., and Goswami, N. (2014). Dopoing
concentration driven morphological evolution of Fe doped ZnO nanostructures. Journal of
Applied Physics, 116:164315.
Sandana, V. E., Rogers, D. J., Teherani, F. H., McClintock, R., Bayram, C., Razeghi, M.,
Drouhin, H. J., Clochard, M. C., Sallet, V., Garry, G., and Falyouni, F. (2009). Comparison of
ZnO nanostructures grown using pulsed laser deposition, metal organic chemical vapor de-
positio, and physical vapor transport. JVST B - Microelectronics and Nanometer Structures,
27(3):1678 – 1683.
Santos, D. A. A. and Macedo, M. A. (2012). Study of the magnetic and structural properties of
Mn - , Fe - , and Co - doped ZnO powder. Physica B, 407:3229 – 3232.
Saravanan, R., Santhi, K., Sivakumar, N., Narayanan, V., and Stephen, A. (2011). Synthesis
and characterization of ZnO and Ni doped ZnO nanorods by thermal decompostion method
for spintronics application. Materials characterization, 67:10 – 16.
Scandolo, S., Giannozzi, P., Cavazzoni, C., de Gironcoli, S., Pasquarello, A., and Baroni, S.
(2005). First-principles codes for computational crystallography in the quantum-espresso
package. Z Kristallogr., 220:574 – 579.
Serrano, J., Romero, A. H., Manjón, F. J., Lauck, R., Cardona, M., and Rubio, A. (2004).
Pressure dependence of the lattice dynamics of ZnO: An ab initio approach. Physical Review
B, 69:094306.
University of Ghana  http://ugspace.ug.edu.gh
REFERENCES 105
Shafique, M. A., Shah, S. A., Nafees, M., Rasheed, K., and Ahmad, R. (2012). Effects of
doping concentration on absorbance, structural, and magnetic properties of cobalt – doped
ZnO nanocrstallites. International Nano Letters, 2(31):1 – 7.
Sharma, V. K., Xalxo, R., and Varma, G. D. (2006). Structural and magnetic studies of Mn -
doped ZnO. Cryst. Res. Technol., 42(1):34 – 38.
Shih, B.-C., Xue, Y., and Zhang, P. (2010). Quasiparticle Band Gap of ZnO: High accu-
racy conventionalaccuracy from the Conventional G0W0 approach. Physical Review Letters,
105:146401.
Shionoya, S. and Yen, W. M., editors (1999). Phosphor Handbook. CRC Press.
Silambarasan, M., Saravanan, S., and Soga, T. (2015). Effect of Fe–doping on the structural,
morphology and optical properties of ZnO nanoparticles synthesized by solution combustion.
Physica E, 71.
Spanhel, L. and Anderson, M. A. (1991). Semiconductor clusters in sol - gel process: Quan-
tized, aggregation, gelation and crystal growth in concentrated ZnO colloids. Journal of
American Chemical Society, 113:2826 – 2833.
Srikant, V. and Clarke, D. R. (1998). On the optical band gap of zinc oxide. Journal of Applied
Physics, 83(10):5447–5451.
Srinet, G., Kumar, R., and Sajal, V. (2014). Effects of aluminium doing on structural and
photoluminiscence properties of ZnO nanoparticles. Ceramics International, 40:4025 – 4031.
Tan, T. L., Lai, C. W., and Hamid, S. B. A. (2014). Tunable band gap energy of Mn-doped ZnO
nanoparticles using the coprecipitation technique. Journal of Nanomaterials, 2014:1–6.
Thakur, S., Kumar, J., Sharma, J., Sharmar, N., and Kumar, P. (2013). Structural and optical
study of nickel doped ZnO nanoparticles and thin films for dye sensitized solar applications.
Journal of Optoelectronics and Advanced materials, 15(7 – 8):989 – 994.
University of Ghana  http://ugspace.ug.edu.gh
REFERENCES 106
Thipprasert, W. and Sritakaew, P. (2014). Leakage currents of zinc oxide surge arresters in 22
kV distribution system using thermal image camera. Journal of Power and Energy Engineer-
ing, 2:712 – 717.
Valerini, D., Caricato, A. P., Creti, A., Lomasolo, M., Romano, F., Taurino, A., Tunno, T., and
Martino, M. (2009). Morphology and photplumiscence properties of zinc oxide films grown
by pulsed laser deposition. Applied Surface Science, 255:9680 – 9683.
van de Walle, C. G. (2001). Defect analysis and engineering in ZnO. Physica B, 308–310:899–
903.
Van de Walle, C. G. and Neugebauer, J. (2004). First principle calculations for defects and
impurities: applications to III-nitrides. Journal of Applied Physics, 95(8):3851 – 3879.
Vanheusden, K., Warren, W. L., Seager, C. H., Tallant, D. R., and Voigt, J. A. (1996). Mecha-
nism behind green photoluminiscence in ZnO phosphor powders. Journal of Applied Physics,
79(7983):7983 – 7990.
Viezbicke, B. D., Davis, S. P. B. E., and Birnir, D. P. (2015). Evaluation of the tauc method for
optical absorption edge determination: ZnO thin films as a model system. Phys. Status Solidi
B, 252(8):1700 – 1710.
Walkosz, W. (2011). Atomic Scale Characterization and Principles Studies of Si3N4 Interfaces,
volume 10. Springer Science+Business Media.
Wang, L. G. and Zunger, A. (2003). Cluster - Doping approach for wide-gap semiconductors:
The case of p-type ZnO. Physical Review Letters, 90:256401(4).
Wang, Q.-B., Zhou, C., Wu, J., and Lü, T. (2013). A GGA + U study of the optical properties
of vanadium doped ZnO with and without single intrinsic vacancy. Optics Communications,
297:79–84.
Wang, Z., Yu, R., Pan, C., Li, Z., Yang, J., Yi, F., and Wang, Z. L. (2015). Light – induced
pyroelectric effect as an effective approach for ultrafast ultraviolet nanosensing. Nature Com-
munications, 6(8401):1–7.
University of Ghana  http://ugspace.ug.edu.gh
REFERENCES 107
Wang, Z. L. (2004). Zinc oxide nanostructures: growth, properties and applications. J. Phys:
Condensed Matter, 16:829 – 858.
Wang, Z. L. (2007). The New Field of Nanopiezoelectronics. Materials Today, 10(5).
Wekesa, M., Jamaluddin, M., and Sobhi, H. F. (2011). An insight into Mn(II) Chemistry:
A study of reaction kinetics under alkaline conditions. Internationa Journal of Chemistry
Research, 2(4):32–37.
White, H. E. (1934). Introduction to Atomic Spectra. McGraw-Hill.
Willander, M., Nur, O., Sadaf, J. R., Qadir, M. I., Zaman, S., Zainelabdin, A., Bano, N., and
Hussain, I. (2010). Luminescence from zinc oxide nanostructures and polymers and their
hybrid devices. Materials, 3:2643 – 2667.
Willander, M., Nur, O., Zhao, Q. X., Yang, L. L., Lorenz, M., Cao, B. Q., Pérez, J. Z., Czekalla,
C., Zimmermann, G., Grundmann, M., Bakin, A., Behrends, A., Al-Suleiman, M., El-Shaer,
A., Mofor, A. C., Postels, B., Waag, A., Boukos, N., Travlos, A., Kwack, H. S., Guinard, J.,
and Dang, D. L. S. (2009). Zinc oxide nanorod based photonic devices: recent progress in
growth, light emitting diodes and lasers. Nanotechnology, 20:332001 (40pp).
Wróbel, J., Krzydlwoski, Hummer, K., Kresse, G., and Piechota, J. (2009). Calculations of ZnO
properties using the Hyed-Scuseria-Ernzerhorf screened hybrid density functional. Physical
Review B, 80:155124(8).
Wu, X., Wei, Z., Zhang, L., Wang, X., Yang, H., and Jiang, J. (2014). Optical and Magnetic
Properties of Fe Doped ZnO Nanoparticles obtained by Hydrothermal Synthesis. Journal of
Nanomaterials, 2014:1 – 6.
Xu, J., Han, J., Zhang, Y., ’an Sun, Y., and Xie, B. (2008). Studies on alcohol sensing mecha-
nism of ZnO based gas sensors. Sensors and Actuators B, 132(1):334 – 339.
Yao, G. Y., Fan, G. H., Zhao, F., Chen, M. J., w Zeng, S., He, L. F., and Zhang, T. (2012). In
assisted realization of p-type C-doped ZnO: A first principle study. Physica B, 407:3539–
3542.
University of Ghana  http://ugspace.ug.edu.gh
REFERENCES 108
Yongning, H., Jingwen, Z., Xiaodong, Y., Qingan, X., ChangChun, Z., and Xun, H. (2007).
Study on pulsed laser ablation and deposition of ZnO thin films by L - MBE. Science in
China Series E: Technological Sciences, 5(3):290 – 301.
Zhang, S. B., Wei, S. H., and Zunger, A. (2001). Intrinsic n − type versus p − type doping
asymmetry and the defect physics of ZnO. Pysical Review B, 63:075205 (7).
Zhang, Y., Nayak, T. R., Hong, H., and Cai, W. (2013). Biomedical Application of Zinc oxide
Nanomaterials. National Institutes of Health Public Access, 13(10):1633 –1645.
Zheng, W., Liao, Y., Li, L., Yu, Q., Wang, G., Li, Y., and Fu, Z. (2006). Structure and properties
of ZnO films grown on Si substrates with low temperature buffer layers. Applied Surface
Science, 253:2765 – 2769.
Zhou, Z., Zhan, C., Wang, Y., yanjie Su, Yang, Z., and Zhang, Y. (2011). Rapid mass production
of ZnO nanowires by a modified carbothermal reduction method. Materials Letters, 65:832
– 835.
University of Ghana  http://ugspace.ug.edu.gh
Appendices
109
University of Ghana  http://ugspace.ug.edu.gh
Appendix A
A.1 PL emission spectra
(a) Fe-doped (b) Mn-doped
(a) Co-doped (b) Ni-doped
Figure A.2: PL Emission Spectra
110
University of Ghana  http://ugspace.ug.edu.gh
Appendix B
B.1 UV-VIS spectra at 600◦C annealing (pH 5)
(a) Fe-doped (b) Mn-doped
(a) Co-doped (b) Ni-doped
Figure B.2: UV-VIS Absorption Spectra At 600 ◦C Annealing (pH 5)
111
University of Ghana  http://ugspace.ug.edu.gh
APPENDIX B. 112
UV-VIS absorption spectra at 600 ◦C annealing (pH 3)
(a) Fe-doped (b) Mn-doped
(c) Co-doped (d) Ni-doped
Figure B.3: UV-VIS Absorption Spectra At 600 ◦C Annealing (pH 3)