FIRST-PRINCIPLES CALCULATIONS ON SOL- GEL ZINC OXIDE NANOPARTICLES
OPTOELECTRONIC PROPERTIES
KAUSAR BINTI HARUN
UNIVERSITI SAINS MALAYSIA
2018
FIRST-PRINCIPLES CALCULATIONS ON SOL-GEL ZINC OXIDE NANOPARTICLES OPTOELECTRONIC PROPERTIES
by
KAUSAR BINTI HARUN
Thesis submitted in fulfillment of the requirements for the degree of
Master of Science
February 2018
ii
ACKNOWLEDGEMENT
Alhamdulillah, with all Allah’s will, I manage to complete my research and this thesis. A million thank dedicated to my supervisor, Associate Professor Dr. Ahmad Azmin Mohamad for his endless motivation, guidance, and patient in minoring my work. A sincere appreciation also goes to my co-supervisor, Professor Dr. Zainal Arifin Ahmad for being very helpful in giving suggestions and comments for my research.
I thank my dear colleagues from Battery Research Group for their supporting ideas and critics. I must thank the three great people from UiTM Shah Alam, Dr.
Mohamad Fariz, Dr. Muhamad Kamil and Mr. Abdul Wafi for a valuable knowledge sharing on first-principles calculation. This appreciation also deserved by all lecturers, technical, and administrative staffs of School of Materials and Mineral Resources Engineering. Their help in all areas has contributed to the success of my research.
I would also acknowledge the funding bodies, Universiti Sains Malaysia (USM) for awarding USM Fellowship Scheme. Finally, my hearties appreciation goes to my husband Dr. Mohd Amin, my kids Yusuf and Maryam and my entire family for their endless love and support. May Allah blessed and reunited us in His paradise, In Shaa Allah.
iii
TABLE OF CONTENTS
ACKNOWLEDGEMENT ii
TABLE OF CONTENTS iii
LIST OF TABLES vii
LIST OF FIGURES viii
LIST OF ABBREVIATIONS xiv
LIST OF SYMBOLS xvi
LIST OF CHEMICAL FORMULA xviii
ABSTRAK xix
ABSTRACT xx
CHAPTER ONE: INTRODUCTION 1.1 Study background
1.2 Problem statement 1.3 Objectives of study
1.4 Thesis outlines and significance of study
1 2 4 5
CHAPTER TWO: LITERATURE REVIEW 2.1 Introduction
2.2 Overview of ZnO as a semiconductor
2.3 Synthesis route of ZnO nanoparticles: Sol-gel method 2.4 Physical characterization of sol-gel synthesized ZnO
2.4.1 Sol-gel synthesized ZnO: Thermal properties
6 6 8 11 11
iv
2.4.2 Sol-gel synthesized ZnO: Phase and structural properties 2.4.3 Sol-gel synthesized ZnO: Microstructural and morphological
properties
2.4.4 Sol-gel synthesized ZnO: Optical properties
2.5 Background and development of first-principles calculation 2.6 Density functional theory
2.6.1 The Hohenberg-Kohn Theorems 2.6.2 Kohn-Sham method
2.7 Exchange-correlation functional 2.7.1 Local density approximation
2.7.2 Generalized gradient approximation 2.8 Hubbard-U scheme
2.9 First-principles calculation: Structure modeling of ZnO 2.10 First-principles calculation: Geometrical optimization of ZnO
structure
2.11 First-principles calculation: The electronic properties of ZnO 2.12 First-principles calculation: The optical properties of ZnO 2.13 Combined studies on first-principles and experimental
13 16
18 25 26 27 28 29 29 30 30 32 33
37 40 43
CHAPTER THREE: METHODOLOGY 3.1 Introduction
3.2 Experimental materials and apparatus
3.3 Synthesis of ZnO nanoparticles by sol-gel storage method 3.4 Characterization of experimentally-grown ZnO nanoparticles
44 44 47 48
v 3.4.1 Thermal analysis
3.4.2 Phase and structural analysis 3.4.3 Morphological analysis 3.4.4 Optical properties evaluation
3.5 First principles calculation of ZnO nanoparticles 3.5.1 ZnO structure modelling
3.5.2 Geometrical optimization
3.5.3 Energy calculation for electronic properties 3.5.4 First-principles calculation for defective ZnO 3.5.5 Optical properties calculation
48 49 50 51 52 52 54 57 58 58
CHAPTER FOUR: RESULTS AND DISCUSSIONS 4.1 Introduction
4.2 Synthesis of ZnO from sol-gel method 4.3 Characterization of sol-gel synthesized ZnO
4.3.1 Thermal analysis
4.3.2 Qualitative phase and structural analysis 4.3.3 Quantitative phase and structural analysis 4.3.4 Proposed growth mechanism
4.3.5 Morphological analysis
4.3.6 Absorption and energy band gap evaluation 4.3.7 Luminescence properties of ZnO
4.4 First-principles calculation of ZnO
4.5 Structure modeling of perfect ZnO unit cell
59 59 61 61 64 66 70 74 79 83 85 85
vi
4.5.1 Optimization and energy calculation of perfect ZnO unit cell 4.5.2 Electronic density of state of perfect ZnO unit cell
4.6 Structure modeling of defective ZnO
4.6.1 Energy and electronic calculation of defective ZnO 4.7 Optical properties of ZnO
4.8 Summary
87 95 98 99 102 104
CHAPTER 5: CONCLUSIONS AND RECOMMENDATIONS 5.1 Conclusions
5.2 Recommendations for future work
106 107
REFERENCES 109
LIST OF PUBLICATIONS APPENDICES
vii
LIST OF TABLES
Page Table 2.1 List of precursors, solvents and additives used during sol-gel
synthesis of ZnO nanoparticles for various applications from literature
10
Table 2.2 Summary on the characterization techniques of sol-gel synthesized ZnO
23
Table 2.3 The optimized lattice constant and energy band gap obtained from various approximations
34
Table 2.4 The convergence, input criteria and calculated energy band gap obtained from first-principles calculation of ZnO system
36
Table 4.1 Refined structural parameters of ZnO nanoparticles aged at various time and the corresponding agreement indices. The standard ICSD (98-010-6787) is presented for comparison and relative deviation is indicated in bracket
69
Table 4.2 Variation of particle size measured from FESEM and HRTEM images
76
Table 4.3 The calculated lattice parameter using conventional functional with relative deviation from experimental approach. The band gap obtained from the UV-VIS plot is compared with calculated energy band gap
88
Table 4.4 The calculated lattice parameter using corrected Hubbard-U method with relative deviation from experimental approach.
The band gap obtained from the UV-VIS plot is compared with calculated energy band gap
92
viii
LIST OF FIGURES
Page Figure 2.1 The unit cell of ZnO (a) wurtzite, (b) zinc-blende and
(c) rock salt structure. The grey sphere represent Zn atom and red sphere represent O atom [32]
7
Figure 2.2 (a) Basic TGA curve of dry ZnO gel derived from zinc acetate dihydrate and methanol (adapted from Ref. [64]) and (b) TGA measurement of ZnO-DMLT nanoparticles with the TGA curve of pure DMLT serving as reference to ensure the full decomposition of DMLT (adapted from Ref. [65])
12
Figure 2.3 (a) The comparison of XRD patterns between bulk and nanosize ZnO, clear broadening observed in nanosized ZnO (adapted from Ref. [67]), (b) basic XRD patterns of as-prepared ZnO nanoparticles synthesized from zinc acetate, ethanol and lithium hydroxide (adapted from Ref. [56]) and (c) the diffraction patterns of ZnO nanoparticles with the effect of precipitation and washing on nanoparticle purity (adapted from Ref. [69])
14
Figure 2.4 Rietveld refinement on ZnO which the observed (red dots) profiles represent the experimental data and the calculated data (black line) represent the modelled fit curve (adapted from Ref. [73])
16
Figure 2.5 FESEM images of ZnO powder synthesized using different procedures (a) sol–gel storage and (b) sol–gel centrifugation (adapted from Ref. [48]), (c) microstructural images using TEM for EDTA-capped ZnO, and (d) HRTEM image of EDTA-capped ZnO
17
ix (adapted from Ref. [83])
Figure 2.6 (a) The UV-Vis absorption spectra of the obtained ZnO quantum dot sols prepared by different solvent: A- methanol, B-ethanol and C-hexanol (adapted from Ref.
[85]) and (b) the UV–Vis-IR absorption spectrum of ZnO nanoparticles for different annealing temperatures.
The inset is Tauc’s plot showing the Eg spectrum for different temperatures (adapted from Ref. [12])
20
Figure 2.7 Comparison of luminescence emission between bulk, nanocrystal and quantum dots ZnO [90]
21
Figure 2.8 (a) The 1×1×1 unit cell of ZnO, (b) 2×3×4 ZnO supercell with Cd as foreign atom, (c) 2×2×4 ZnO supercell with Cd as foreign atom and (d) 1×3×3 ZnO supercell with Cd as foreign atom [112]
32
Figure 2.9 The band gap variation along with lattice parameter for (a) at fixed Up,O = 0 eV with varied Ud,Zn and (b) at fixed Ud,Zn = 10 eV with varied Up,O [32]
35
Figure 2.10 The calculated band structure using different XC functional (a) LDA with red circle show underestimated band gap (adapted from Ref. [125], (b) GGA and GGA+U and (c) the enlarged energy gap (adapted from Ref. [32])
37
Figure 2.11 The calculated DOS of ZnO (a) combined DOS using LDA (adapted from Ref. [101]), (b) schematic representation of individual and total DOS by GGA (adapted from Ref. [128]), (c) spin-up and spin-down DOS of ZnO with oxygen vacancy by GGA and (d)
40
x
spin-up and spin-down DOS of ZnO with oxygen vacancy by GGA+U (adapted from Ref. [129])
Figure 2.12 (a) The imaginary part of dielectric function of ZnO calculated by LDA+U (adapted from Ref.[97]) and (b) the calculated absorption of ZnO by LDA, GGA and GGA-EV (adapted from Ref. [128])
42
Figure 3.1 Flowchart of the project starting with ZnO synthesis followed by first-principles calculation
46
Figure 3.2 Synthesis of ZnO nanoparticles (a) initial clear solution of Zn(CH3COO)2.2H2O dissolved in CH3OH, (b) milky- white slurry after titration with 1 M NaOH and (c) sedimentation of ZnO after 48 h aging
47
Figure 3.3 Build crystal window for space group insertion 52 Figure 3.4 Selection for lattice parameter insertion (a and c for
ZnO)
53
Figure 3.5 (a) Adding atomic coordination of Zn atom and (b) adding atomic coordination of O atom
53
Figure 3.6 The CASTEP calculation window for geometrical optimization task with specified XC functional
54
Figure 3.7 (a) The convergence tolerance in optimization process and (b) the specified energy cutoff and k-point basis set
55
Figure 3.8 The CASTEP calculation window for optimization with Hubbard-U correction scheme. The red arrow indicated the activation of Hubbard-U in the calculation
56
xi
Figure 3.9 The selection of Coulomb energy U for (a) Zn-3d and (b) O-2p state
57
Figure 4.1 Synthesis of ZnO nanoparticles by the sol-gel method;
(a) fresh pasty-like product, (b) gel obtained after 48 h aging, (c) xerogel produced after drying, (d) grounded powder before calcination and (e) grounded powder after calcination
60
Figure 4.2 The mass-loss traces from TGA curves of as- synthesized ZnO at different aging time
62
Figure 4.3 The TGA curve for precursor zinc acetate dihydrate, Zn(CH3COO)2.2H2O heated in normal air from – at min-1
63
Figure 4.4 The diffraction patterns of a as-synthesized nO and nO after calcination at
65
Figure 4.5 Rietveld-refinement profiles of ZnO phase synthesized by sol-gel method aged at various time. Iobs refers to the intensity of observed data (experimental data) and Icalc refers to intensity of calculated model. The difference plot (blue line) was obtained from Iobs – Icalc which represent the residue between observed and calculated model
67
Figure 4.6 The schematic of ZnO growth from sol-gel method at different aging time
73
Figure 4.7 The FESEM images of ZnO synthesized at different aging time (a) 0, (b) 6, (c) 12, (d) 24, (e) 36 and (f) 48 h
75
xii at magnification 30 000×
Figure 4.8 The HRTEM images of ZnO synthesized at various aging time at magnification 97 000 × and the scale bar for each images are 50 nm
77
Figure 4.9 Plot of particle size of ZnO measured by FESEM and HRTEM
78
Figure 4.10 The absorbance of ZnO nanoparticles obtained from UV-VIS spectroscopy: (a) aged at 0 and 6 h and (b) aged at 12, 24, 36, and 48 h
80
Figure 4.11 Tauc plots of ZnO at synthesized at different aging time:
(a) aged at 0 and 6 h and (b) aged at 12, 24, 36 and 48 h
82
Figure 4.12 Photoluminescence spectra of ZnO at different aging time
84
Figure 4.13 The ZnO structure drawn on Material Studio Visualizer (a) the unit cell of ZnO and (b) tetrahedral coordination of Zn and O atom
86
Figure 4.14 Variation of energy band gap and lattice parameter calculated using different functionals (a) LDA+U, (b) GGA-PBE+U and (c) GGA-PBESol+U. In all calculation, the Ud,Zn value was fixed to 10 eV and Up,O was varied accordingly
91
Figure 4.15 Calculated energy band structure of synthesized ZnO using different functionals (a) LDA, GGA-PBE and GGA-PBESol and (b) LDA+U, GGA-PBE+U and GGA-PBESol+U. The right figure demonstrate enlarged
94
xiii
energy band structure focussing at highest valence band and lowest conduction band along G-G path
Figure 4.16 Comparison of density of state calculated from (a) GGA-PBE functional and (b) GGA-PBE+U functional with Ud,Zn = 10 eV and Up,O = 6.1 eV. The grey dotted line was the Fermi level located at 0 eV
96
Figure 4.17 The ZnO structure drawn on Material Studio Visualizer with (a) unit cell, (b) 3 × 3 × 2 supercell structures and (c) the top view from (002) plane consisting oxygen vacancy at the dotted circle
99
Figure 4.18 The calculated band gap of 3 × 3 × 2 ZnO structure containing oxygen vacancy using GGA-PBE+U
100
Figure 4.19 (a) The calculated density of state of ZnO structure containing oxygen vacancy and (b) schematic view of the band structure
101
Figure 4.20 The calculated dielectric function of ZnO based on GGA-PBE+U approximation
102
Figure 4.21 The calculated absorption of ZnO based on GGA- PBE+U approximation
104
xiv
LIST OF ABBREVIATIONS
Abbreviation Description
CASTEP Cambridge Serial Total Energy Package
CB Conduction band
CBM Conduction band minimum
DFT Density functional density
DOS Density of state
FESEM Field emission scanning electron microscope
GGA Generalized gradient approximation
GGA-PBE Generalized gradient approximation with Perdew-Burke- Ernzerhof scheme
GGA-PBE+U Generalized gradient approximation with Perdew-Burke- Ernzerhof and Hubbard-U correction scheme
GGA-PBESol Generalized gradient approximation with Perdew-Burke- Ernzerhof scheme for solid
GGA-PBESol+U Generalized gradient approximation with Perdew-Burke- Ernzerhof scheme for solid and Hubbard-U correction scheme
GOF Goodness of fit
ICSD Inorganic Crystal Structure Database
LDA Local density approximation
xv
LDA+U Local density approximation with Hubbard-U correction scheme
PDOS Partial density of state
PL Photoluminescence
SGC Sol-gel centrifuge
SGS Sol-gel storage
SIC Self-interaction correction
TGA Thermal gravimetric analysis
TEM Transmission electron microscope
UV Ultra violet
UV-VIS Ultra violet-visible
VB Valence band
VBM Valence band maximum
XC Exchange-correlation
XRD X-ray diffraction
3D Three dimension
xvi
LIST OF SYMBOLS
Symbols
% Percentage
° Degree
°C Degree Celcius
°C min-1 Degree Celcius per minute
Ψ Wavefunction
ε1 Real part of dielectric function
ε2 Imaginary part of dielectric function
ω Photon frequency
∇ Gradient of electron density
λ Wavelength
Å Angstrom
α Absorption coefficient
a Lattice parameter in x-axis
c Lattice parameter in z-axis
e Charge of electron
ehom Homogenous electron gas
m Mass of ion
n(r) Particle density at point r
h Planks’s constant
h Hour
j Jam
xvii
r Coordinates of electron
v Frequency of light
Eg Energy band gap
EHF Hartree-Fock energy
Ĥ Hamiltanion operator
M Mass of electron
P1 Momenta of ion
P2 Momenta of electron
R Coordinate of ion
R Residual factor
Rexp Expected profile residual
Rwp Weighted profile residual
U Coulomb repulsion energy
Z Charge of ion
GPa Giga Pascal
nm Nano meter
eV Electron Volt
meV Mili electron Volt
xviii
LIST OF CHEMICAL FORMULA
CH3OH Methanol
H2O Water
NaOH Sodium hydroxide
Zn(CH3COO)2.2H2O Zinc acetate dihydrate
Zn(OH)2 Zinc hydroxide
Zn(OH)42- Zincate ion
OH- Hydroxide ion
Zn2+ Zinc ion
O2- Oxygen ion
ZnO Zinc oxide
TiO2 Titanium oxide
CdS Cadmium sulfide
CdSe Cadmium selenide
SnO2 Tin oxide
xix
PENGIRAAN PRINSIP-PERTAMA TERHADAP CIRI-CIRI OPTOELEKTRONIK PARTIKEL NANO SOL-GEL ZINK OKSIDA
ABSTRAK
Diagnostik berkesan antara eksperimentasi dan pengiraan teori adalah perlu untuk memastikan sinergi antara kedua pendekatan. Kajian ini menggunakan input struktur daripada experimentasi ke dalam rangka kerja teori. Permulaannya, partikel nano ZnO telah disintesis melalui kaedah sol-gel pada waktu penuaan berbeza. Analisa fasa dan struktur mengesahkan penghasilan struktur ZnO wurtzit heksagon dengsn sampel dituakan selama 36 j menunjukkan penghabluran tertinggi dan memberikan visual tepat terbaik dalam analisa Rietveld. Pemerhatian morfologi menunjukkan penghasilan partikel nano sfera yang seragam pada masa penuaan melebihi 6 j manakala variasi yang kecil direkodkan pada jurang jalur tenaga antara 3.08 – 3.12 eV. Jalur kependarkilauan menunjukkan pelepasan hijau kerana kekosongan oksigen.
Di dalam pengiraan prinsip pertama, sel unit ZnO dibina berdasarkan parameter struktur daripada analisa Rietveld bagi menghubungkan kajian eksperimental.
Beberapa fungsi penukaran-korelasi termasuk LDA, GGA-PBE, GGA-PBESol, LDA+U, GGA-PBE+U dan GGA-PBESol+U. Fungsi GGA-PBE+U (Ud,Zn = 10 eV dan Up,O = 6.1 eV) menunjukkan sisihan kekisi terendah dan berjaya mengulang jurang jalur tenaga eksperimentasi. Struktur ZnO super sel bersama kekosongan oksigen menunjukkan kedudukan kecacatan lebih nyah-setempat dan berada pada 1.90 eV dari atas jalur konduksi. Posisi ini menepati tenaga pembebasan foton (2.06 eV) seperti terlihat di spektrum kependarkilauan. Dapatan ini bermanfaat dalam rekabentuk anod sel solar bagi meningkatkan penyerapan cahaya nampak.
xx
FIRST-PRINCIPLES CALCULATIONS ON SOL-GEL ZINC OXIDE NANOPARTICLES OPTOELECTRONIC PROPERTIES
ABSTRACT
An efficient diagnostic between experimental and theoretical calculation is essential to ensure the synergy between these two approaches. This study made attempt to use structural input from experimental in the theoretical framework.
Initially, ZnO nanoparticles were synthesized by sol-gel method at different aging time. The phase and structural analyses confirmed the formation of hexagonal wurtzite ZnO structure at which sample aged at 36 h showed highest crystallinity and gave the best visual fit in Rietveld analysis. Morphological observation revealed spherical nanoparticles were formed at aging time higher than 6 h while only small variation in energy band gap recorded between 3.08 – 3.12 eV. The photoluminescence spectra revealed a green emission due to oxygen vacany. In first- principles calculation, the ZnO unit cell was built based on structural parameter from Rietveld analysis in order to provide a bridge with experimental study. Several exchange-correlation functional including LDA, GGA-PBE, GGA-PBESol, LDA+U, GGA-PBE+U and GGA-PBESol+U were tested. The GGA-PBE+U (Ud,Zn = 10 eV and Up,O = 6.1 eV) showed lowest lattice deviation and successfully reproduced the experimental band gap. ZnO supercell structure with oxygen vacancy showed that defect state were more delocalized and appeared at 1.90 eV from top of conduction band. This position was close to the photon energy released due to recombination of electron (2.06 eV) as observed in luminescence spectra. The results are beneficial in designing photoanode material in solar cell that will enhance visible light absorption.
1
CHAPTER ONE INTRODUCTION
1.1 Study background
The 21st century has marked a tremendous research work focusing on potential clean and renewable energy technology. The new generation of solar cell known as dye-sensitize solar cell (DSSC) is an example of energy device that actively studied.
In DSSC, the photoanode consist of a metal oxide semiconductor plays important role that contributes to overall efficiency. It serves as a scaffold that supports the dye molecules and transferring electrons [1]. Zinc oxide (ZnO) has become a potential photoanode material pertaining to its unique and comparable properties from its former counterpart.
ZnO is a II-VI semiconductor with a wide energy band gap (3.3 eV) and high electron mobility with magnitude larger than anatase TiO2 μTiO2 = 0.1-4 cm2 Vs-1, μZnO = 200-300 cm2 Vs-1) [2]. To date, issue on the incapability of ZnO to fully utilize visible light due to its wide band gap has limited its potential use especially in solar cell.
Several attempts have been conducted such as the introduction of a doping element and monitoring the native defects [3, 4]. These work in return involved number of experiments before the ideal properties can be achieved.
Pure ZnO nanoparticles can be obtained through several synthesis routes such as solid state reaction [5], hydrothermal [6] and sol-gel methods [7-9]. Notably, the sol–
gel method has been favoured for the synthesis of nO ecause it can take place at a
2
lower temperature , involves simple starting materials, and produces nO with excellent chemical homogeneity. The synthesis condition including solution pH [8, 10], type of starting materials [11], and pre- and post-heat treatment [12] are found to give impact on properties of sol-gel derived ZnO.
Meanwhile, the current practice used first-principles calculations based on the density functional theory (DFT) to study the properties of ZnO. DFT has become the preferred computational method due to the simplicity of the software and its ability to calculate the ground state properties with predictive accuracy. The principles of DFT are based on two theorems pioneered by Hohenberg-Kohn [13] and Kohn-Sham [14] that simplify the complexity of the many-body Schrodinger equation. By considering the electron density instead of many-body wave function, DFT has made the computational work much easier to be solved [15].
A number of theoretical studies have been conducted to simulate the optoelectronic properties ZnO [16-19]. Based on this method, fast and accurate results have been achieved, along with reduced trial and error, as often happens in experimental work. However, calculations based on the DFT are sensitive as the varying of unnecessary parameters may lead to unphysical and misinterpreted results.
1.2 Problem Statement
Previous studies have shown that intensive investigations on the properties of ZnO have been carried out by means of experimental and theoretical methods.
Hence, it is necessary to verify these two approaches to ensure the synergy of each
3
work and in return leading to a significant improvement. One intriguing approach is to integrate the theoretical calculation with the input from the experimental result.
This approach used lattice parameters and atomic coordination obtained from experimentation to build the ZnO crystal structure in theoretical framework.
However, the reported studies had used lattice inputs from random literature during the structure modelling stage [20, 21]. This strategy successfully created a ZnO model, but it did not offer a close representation of experimentally-grown ZnO. As a result, no bridging is attained and the calculated optoelectronic properties are merely belong another system.
To obtain an exact crystal structure is a challenging task. The refined diffracted profile from X-ray diffraction analysis offered structural information that is close representation to the synthesized version. Therefore, a well synthesized ZnO must be produced with a controlled parameter and carefully characterized.
In the sol-gel method, several processes involved such as hydrolysis, condensation, nucleation and aging. The growth of ZnO mainly occurred during aging [22] and if the gel is freely aged over time, the formation of ZnO nanoparticles could be investigated. Previous literature has noted that stabilized ZnO can be obtained after short-time aging lasting 0–36 h [23], 48 h [7, 8], and even after a month [24]. The range of aging time is rather very wide and may lead to difficulties when the optimum aging is to be chosen for practical consideration. Hence, aging time must be carefully examined to allow complete formation of ZnO.