Amun Amri, ST., MT. - Murdoch UniversityAmun Amri, ST., MT. This thesis is presented for the degree of Doctor of Philosophy of Murdoch University ... and EDX analyses corroborated

Structural, Optical and Mechanical Characterisations of Nanostructured Copper Cobalt Oxide Coatings

Synthesised via Sol-gel Method for Solar Selective Absorber

Amun Amri, ST., MT.

This thesis is presented for the degree of Doctor of Philosophy of

Murdoch University 2013

ii

Declaration

I declare that this thesis is my own account of my research and contains as its

main content work which has not previously been submitted for a degree at any

tertiary education institution.

(Amun Amri)

iii

To my father and my mother for your affection, encouragement and prayers...

To my wife and my children for your love, understanding and patience…

iv

ABSTRACT

The search for clean renewable energy sources to fulfil global energy needs, incorporating

environmentally-friendly technologies, is currently unabated. Solar thermal collectors are

technologies that harness unlimited solar radiation then convert it into usable heat for

numerous industries or domestic needs. The solar selective absorber thin film coating is the

key component in determining the efficiency of a solar thermal collector. Many challenges

still exist in terms of the fabrication of high quality selective absorber material, in order to

meet the criteria of better cost-effectiveness and environmentally-friendly characteristics

especially in the context of flat-plate absorbers.

In this study, novel copper cobalt oxide (CuxCoyOz) thin film coatings on highly reflective

aluminium substrate were synthesised via a facile, environmentally friendly and cost-

effective sol-gel dip-coating method. The structural, surface morphology and composition,

optical properties, mechanical properties and thermal durability were characterised using a

wide range of complementary techniques, namely, XRD, FESEM/SEM, EDX, AFM, XPS,

synchrotron radiation XPS and NEXAFS, UV-Vis-NIR, FTIR, nanoindentation and FEM

modelling, as well as an accelerated thermal durability test.

The copper cobalt oxide thin films showed a nano-sized grain-like morphology forming a

porous surface structure with distinctively high solar absorptance compared to the

manganese- and nickel- cobalt oxide coatings. XRD results demonstrated a relatively weak

crystallinity of copper cobalt coating through the annealing temperature of 500 °C, the XPS

and EDX analyses corroborated the existence of Cu-O and Co-O bonding structures within

different copper cobalt oxide composition ratios. The optimised solar absorptance value of

83.4% was achieved from the copper cobalt oxide thin films synthesised using 0.25 M of

copper acetate and 0.25 M cobalt chloride precursors ([Cu]/[Co]=1) with the withdrawal rate

of 120 mm/min by four dip-drying cycles at annealing temperature of 500°C. Higher

absorptance value could be accomplished by a thin film with [Cu]/[Co] of 0.5, however, its

reflectance spectra curve was less satisfactory in terms of a good selectivity curve profile.

The difference in [Cu]/[Co] ratios in the synthesis process has a direct influence on the

degree of porosity of surface morphology which slightly alters the surface compositions,

electronic structure and local coordination of the coatings. The mechanical properties analysis

such as the elastic modulus and hardness via a nanoindentation test revealed that the coatings

exhibit much higher wear resistance compared to the aluminium substrate especially for

[Cu]/[Co] = 1.0. Finite element modelling (FEM) indicated that, under spherical loading

conditions, the higher stress and the plastic deformation were predominantly concentrated

within the coating layer, with marginal effect on the substrate. The high absorptance value

(i.e. without an anti-reflective layer) accompanied by the high wear resistance of copper

cobalt oxide made it a very promising candidate for solar selective absorbers application.

Higher annealing temperatures treatment of up to 650 °C improved the crystalline structure of

copper cobalt oxide, but it relatively did not change the surface compositions and bonding

structures. The absorptance of coatings slightly increased with the annealing temperature up

to 550 °C and then decreased from 550 °C to 650 °C due to the increase of scattering from

larger crystallite. Even though the elastic modulus and the hardness improved, the wear

resistance was slightly decreased as temperature was increased.

v

To minimize the reflectance of absorber material and protect it from any degradation due to

external factors, a silica anti-reflection (AR) layer was fabricated on top of the copper cobalt

oxide coatings. The AR layer evidently changed the reflectance spectra which cause the

increase of the absorptance value in the UV-Visible-near infrared (UV-Vis-NIR) area and

unfortunately also increase the emittance value due to the strong phonon absorption by the

silica in the range from mid to far infrared. The optimum absorptance and emittance values

were 84.96 and 5.6 %, respectively. The accelerated thermal durability test revealed that the

degradation of the copper cobalt oxide with a silica AR layer was more governed by the

temperature regime fluctuations compared to the change in exposure time, indicating that the

coating is applicable for uses in the low temperature range solar collectors such as for

domestic solar water heater (≤ 150 oC).

The sol-gel dip-coating synthesised copper cobalt oxide thin film coatings present high

absorptance in UV-Vis-NIR range and low emittance (or high reflectance) in the mid – far

infrared range with good mechanical properties. All these attributes render the coatings

promising as a solar selective absorber for applications in the solar energy industry. However,

further research may require the development of an appropriate anti-reflective layer to

maximise absorptance and to minimise emission and then achieve a high selectivity of

coatings stack.

vi

TABLE OF CONTENTS

Title

Declaration ii

Abstract iv

Table of Contents vi

Acknowledgements ix

List of Publications x

List of Figures xi

List of Tables xvi

Symbols and Abbreviations xvii

Chapter 1. Introduction

1.1. Background 1

1.2. Objective and Scope of Study 4

Chapter 2. Theoretical Background

2.1. Solar Radiation, Thermal Radiation and Solar Selective Absorber 6

2.2. Optical Properties of Thin Film

2.2.1. Electromagnetic radiation absorption 9

2.2.2. Optical characterisation of selective solar absorber 11

2.3. Solar Selective Absorber Design 12

2.4. Antireflection Layer 17

2.5. Flat Plate Solar Collector 18

2.6. Mechanical Properties of Thin Film Coating and Modelling 21

2.7. Degradation of Selective Absorber and Accelerated Ageing Test 24

Chapter 3. Review of Sol-gel Selective Absorbers

3.1. Sol-gel Synthesis Process 27

3.2. State of the Art of the Sol-gel Selective Absorber Coatings

3.2.1. Metal oxide based selective absorber 28

3.2.2. Metal and carbon particles in dielectric matrix 37

3.2.3. Solar selective absorber surfaces using spinels 46

3.3. Effect of Silica Thickness 52

Chapter 4. Experimental Method

4.1. Film Coatings Preparation

4.1.1. Substrates preparation 53

4.1.2. Materials, sol-gel solution preparation and film coatings

deposition 54

4.2. Instrumentations and Characterisation Techniques

4.2.1. X-ray diffraction (XRD) 57

4.2.2. Scanning electron microscopy (SEM), energy dispersive

X-ray (EDX) and field emission scanning electron

microscopy (FESEM) 59

4.2.3. Atomic force microscopy (AFM) 61

4.2.4. X-ray photoelectron spectroscopy (XPS) 62

vii

4.2.5. Near edge X-ray absorption fine structure (NEXAFS)

spectroscopy 65

4.2.6. Optical characterisations via UV-Vis-NIR and FTIR

reflectance spectra 67

4.2.7. Mechanical characterisations: Nanoindentation test and

finite element modelling 69

4.2.8. The accelerated thermal durability test 71

Chapter 5. Characterisations of Cobalt-based Metal Oxide Thin Films Synthesised

Using Sol-Gel Dip-Coating Method: An Exploration Study

5.1. Introduction 74

5.2. Samples Preparation and Characterisation 76

5.3. Results and Discussion

5.3.1. XRD analysis 76

5.3.2. Surface topography and morphology 78

5.3.3. XPS analysis 81

5.3.4. Optical properties 86

5.3.5. Nanoindentation 89

5.4. Conclusions 91

Chapter 6. Solar Absorptance of Copper Cobalt Oxide Thin Film Coatings:

Optimization, Structural and Surface Compositions


6.2. Sample Preparation and Characterisation 94


6.3.1. EDX analysis 95

6.3.2. Solar absorptance properties 97

6.3.3. Synchrotron radiation XPS study of elevated

concentrations 101

6.4. Conclusions 109

Chapter 7. Surface and Mechanical Characterisations of Copper Cobalt Oxide Thin

Film Coatings Synthesised Using Different Compositions


7.2. Sample Preparation and Characterisation 113


7.3.1. Surface morphology 114

7.3.2. Synchrotron radiation XPS study 116

7.3.3. Synchrotron-based NEXAFS study 122

7.3.4. Mechanical nanoindentation test 126

7.3.5. Finite element modelling (FEM) 129


Chapter 8. Characteristics of Copper Cobalt Oxides Thin Film Coatings

Synthesised by Different Annealing Temperatures


8.2. Experimental 135


8.3.1. XRD analysis 136

8.3.2. XPS study 139

viii

8.3.3. Optical properties 145

8.3.4. Nanoindentation test 148

8.3.5. Finite element modelling (FEM) 150


Chapter 9. Optical Properties and Thermal Durability of Copper Cobalt Oxide

Thin Film Coatings with Integrated Silica Antireflection Layer


9.2. Samples Preparation and Characterisation 155


9.3.1. Reflectance spectra and solar absorptance 156

9.3.2. Emittance and selectivity 158

9.3.3. Accelerated thermal durability test 160


Chapter 10. Conclusions and Future Work 166

References 170

Appendix 1 182

Appendix 2 183

ix

ACKNOWLEDGEMENTS

I would to express my deepest gratitude to my supervisor, Dr. Zhong-Tao Jiang as well as my

co-supervisors, Dr. Chun-Yang Yin and Dr. Trevor Pryor for their encouragement,

supervision, inspiration and support. Without their encouragement and efforts, this thesis

would not have been completed.

I also would like to thank Dr. Alex Duan (The University of Melbourne), Dr. Xiaoli Zhao

(Edith Cowan University), Dr. Zonghan Xie (University of Adelaide), Dr. Bruce Cowie

(Australian Synchrotron, Melbourne), Dr. Sinisa Djordjevic (Murdoch University Energy

Research and Innovation Group, MUERI), Dr. G.E. Poinern (Murdoch Applied

Nanotechnology Research Group, MANRG), Prof. Jennifer Searcy, Prof. Philip Jennings,

Prof. Parisa A. Bahri and Mr. Ken Seymour for their assistance, valuable discussions and

comments during my research work and preparation of journal articles.

Appreciation is also extended to my colleagues: M. Mahbubur Rahman, Nick Mondinos, Hua

Guo, Shahidah Ali, Ravi Brundavanam, Brian Drake and Hantarto Widjaja for their support,

assistance, discussion and friendship. I would also like to express my gratitude and

appreciation to the lecturers and staff at the School of Engineering and Information

Technology - Murdoch University, Indonesian student community in Perth, and all those who

have helped me, either directly or indirectly, during my study in Australia. Last but not least,

I would like to thank the Indonesian Government for providing me with a Ph.D. scholarship.

Perth, June 2013

Amun Amri

x

LIST OF PUBLICATIONS

Journal Articles

1. A. Amri, Z.-T. Jiang, T. Pryor, C.-Y. Yin, Z. Xie, and N. Mondinos. Optical and

mechanical characterization of novel cobalt-based metal oxide thin films synthesized

using sol–gel dip-coating method. Surface and Coatings Technology, vol. 207, pp. 367-

374, 2012.

2. A. Amri, X. Duan, C.-Y. Yin, Z.-T. Jiang, M. M. Rahman, and T. Pryor. Solar

absorptance of copper–cobalt oxide thin film coatings with nano-size, grain-like

morphology: Optimization and synchrotron radiation XPS studies. Applied Surface

Science, vol. 275, pp. 127-135, 2013.

3. M.M. Rahman, X. Duan, Z.-T. Jiang, Z. Xie, A. Wu, A. Amri, B. Cowie, N. Mondinos,

and C.-Y. Yin. Near-edge X-ray absorption fine structure studies of Cr1-xMxN coatings.

Journal of Alloys and Compounds, vol. 578, pp. 362-368, 2013.

4. A. Amri, X. Duan, P.A. Bahri, Z.-T. Jiang, X. Zhao, Z. Xie, C.-Y. Yin, M.M. Rahman,

and T. Pryor. Surface electronic structure and mechanical characteristics of copper cobalt

oxide thin film coatings: Soft X-ray synchrotron radiation spectroscopic analyses and

modelling. The Journal of Physical Chemistry C, vol. 117, pp. 16457-16467, 2013.

5. A. Amri, Z.-T. Jiang, T. Pryor, C.-Y. Yin, and S. Djordjevic. Developments in flat plate

solar selective absorber materials synthesized by sol-gel methods: A review. Renewable

and Sustainable Energy Review. (Submitted in May 2013; Manuscript ID: RSER-D-13-

00712).

6. A. Amri, X. Zhao, Z.-T. Jiang, T. Pryor, C.-Y. Yin, M.M. Rahman, and N. Mondinos.

Tailoring the physicochemical and mechanical properties of optical copper cobalt oxide

thin films through annealing treatment. Journal of the American Ceramic Society.

(Submitted in June 2013; Manuscript ID: JACERS-33437).

7. A. Amri, Z.-T. Jiang, C.-Y. Yin, T. Pryor, and M.M. Rahman. Optical properties and

thermal durability of copper cobalt oxide thin film coatings with integrated silica

antireflection layer. Industrial & Engineering Chemistry Research. (Submitted in June

2013; Manuscript ID: ie-2013-02013c)

Conference Papers:

1. A. Amri, Z.-T. Jiang, T. Pryor, and C.-Y. Yin. Optical properties of copper cobalt metal

oxide thin films synthesized via sol-gel dip-coating method. Presented in: 7th

International Conference on Surfaces, Coatings and Nanostructured Materials, 18 - 21

September 2012, Prague, Czech Republic.

2. A. Amri, Z.-T. Jiang, T. Pryor, C.-Y. Yin, and N. Mondinos. Characterization of copper

cobalt oxide thin film coatings synthesized via sol-gel dip-coating method. Proceedings

of 3rd International Chemical and Environmental Engineering Conference, 21-23

December 2012, Kuala Lumpur, Malaysia.

xi

LIST OF FIGURES

Figure 2.1. Effects of Rayleigh scattering and atmospheric absorption

on the spectral distribution of solar irradiance 7

Figure 2.2. Solar hemispherical spectral irradiance for air mass 1.5 and

blackbody-like emission spectra at 100°C, 200°C and 300°C 9

Figure 2.3. Dark mirror absorber-reflector tandem design 13

Figure 2.4. Microstructure pictures of two examples of solar absorber composite

coatings 16

Figure 2.5. Cross-sectional view of a basic flat plate solar collector 19

Figure 2.6. Typical loading-unloading compliance curve from a nanoindentation

experiment with maximum load (Pmax) and depth beneath

the specimen free surface (hmax) 22

Figure 3.1. General strategy for synthesising metal oxide/spinels (route A) and

metal/carbon particles embedded in matrix (route B) solar selective

absorbers 51

Figure 4.1. Flow chart for the synthesis of copper cobalt oxide thin film coatings 55

Figure 4.2. Dip-coater (PTL-MM01, MTI Corporation) used in the present study 56

Figure 4.3. The incident and scattered X-rays make an angle of θ symmetric to the

normal of crystal plane in XRD analysis 58

Figure 4.4. Schematic diagram of SEM with a CRT display 60

Figure 4.5. Schematic diagram of hemispherical photoelectron energy analyser

in XPS instrument 63

Figure 4.6. Kratos Axis Ultra XPS spectrometer (Manchester, UK) with Mg Kα

radiation source 65

Figure 4.7. Samples transfer and analysis chamber in soft X-ray analysis end

station 66

Figure 4.8. Specular reflectance (Rs) and diffuse reflectance (Rd) in a reflectance

mode of integrated sphere 68

Figure 4.9. Flow chart of accelerated thermal durability test 72

Figure 5.1. XRD patterns of the prepared manganese–cobalt (i), copper–cobalt

(ii) and nickel–cobalt (iii) thin film coatings (6 dip-heating cycles) on

aluminum substrate and standalone heated and unheated aluminum

substrate (iv and v) respectively 77

Figure 5.2. Expanded XRD pattern region from 10o to 40° (intensity of observed

peaks are 0.3%–0.5% of maximum intensity peak of substrate from

panel in Figure 5.1) 77

xii

Figure 5.3. AFM images of the a) manganese–cobalt; b) copper–cobalt; and

c) nickel–cobalt thin film coatings (6 dip-heating cycles) 79

Figure 5.4. SEM micrographs of the a) manganese–cobalt; b) copper–cobalt;

and c) nickel–cobalt thin film coatings. FESEM micrograph for copper–

cobalt indicates the presence of nano-sized grain-like particles

(6 dip-heating cycles) 80

Figure 5.5. FESEM micrograph images (in magnifications of 200 nm and 100 nm)

for copper–cobalt indicate the presence of nano-sized grain-like

particles 81

Figure 5.6. Wide scan of XPS spectra of cobalt-based metal oxide film coatings 82

Figure 5.7. C1s and O1s XPS spectra of manganese–cobalt, copper–cobalt and

nickel–cobalt thin film coatings. Dashed lines correspond to fit

envelopes, while wavy lines correspond to data curves 83

Figure 5.8. Mn2p, Cu2p, Ni2p, and Co2p XPS spectra of manganese–cobalt,

copper–cobalt and nickel–cobalt thin film coatings 85

Figure 5.9. Absorbance spectra of thin film coatings on the glass substrates,

absorbance due to glass substrate was eliminated from the spectra 87

Figure 5.10. Reflectance spectra of thin film coatings on aluminium substrates

with corresponding solar absorptance (α) values 89

Figure 5.11. Elastic modulus and hardness of the thin films measured using the

nanoindentation 90

Figure 5.12. Typical load–displacement curves of the thin films measured using

the nanoindentation 91

Figure 6.1. EDX spectra of cobalt copper thin film coating on the top of aluminium

substrate synthesised using 0.15 M of copper-acetate and 0.15 M of

cobalt-chloride precursors (a), and aluminium substrate without

coating (b) 96

Figure 6.2. Reflectance spectra of copper–cobalt oxide thin film coatings on

aluminium substrates. Concentrations of reactants: (a) 0.15 M copper

and 0.15 M cobalt; (b) 0.2 M copper and 0.2 M cobalt; (c) 0.25 M

copper and 0.25 M cobalt; (d) 0.3 M copper and 0.3 M cobalt.

Four dip-heating cycles were carried out 98

Figure 6.3. Effect of Cu/Co concentration ratios on the reflectance of copper–

cobalt oxide thin film coatings. These include Cu/Co concentration

ratios of 0.5 (0.125 M copper and 0.25 M cobalt), 1 (0.25 M copper

and 0.25 M cobalt) and 2 (0.25 copper and 0.125 M cobalt),

respectively. The dip-speed is 120 mm/min with four dip-heating

cycles. 100

Figure 6.4. SEM micrograph picture of copper cobalt oxide thin film coating

synthesised using concentrations of 0.25 M copper acetate and

xiii

0.25 M cobalt chloride (Cu/Co ratio = 1) with dip-speed 120 mm/min

and four dip-heating cycles on the glass substrate 101

Figure 6.5. O 1s SR-XPS spectra of copper–cobalt thin film coatings synthesised

using concentrations of: (a) 0.15 M copper and 0.15 M cobalt, (b) 0.2 M

copper and 0.2 M cobalt, and (c) 0.25 M copper and 0.25 M cobalt 103

Figure 6.6. (a) Cu 2p SR-XPS spectra of copper–cobalt thin film coatings

synthesised using various concentrations, (b)–(d) decoupling of Cu 2p3/2

of copper–cobalt thin film coatings synthesised using various

concentrations 105

Figure 6.7. (a) Co 2p SR-XPS spectra of copper–cobalt thin film coatings

synthesised using various concentrations, (b)–(d) decoupling of Co 2p3/2

of copper–cobalt thin film coatings synthesised using various

concentrations 107

Figure 7.1. Surface morphologies of copper cobalt oxide coatings synthesised

using a) [Cu]/[Co]=0.5, b) [Cu]/[Co]=1 and c) [Cu]/[Co]=2 115

Figure 7.2. a) Cu 2p SR-XPS spectra of copper cobalt thin film coatings

synthesised using different [Cu]/[Co] concentration ratios,

b-d) decoupling of their corresponding Cu 2p3/2 peak 117

Figure 7.3. a) Co 2p SR-XPS spectra of copper cobalt thin film coatings

synthesised using different [Cu]/[Co] concentration ratios,

b-d) decoupling of their corresponding Co 2p3/2 peak 119

Figure 7.4. a) O 1s SR-XPS spectra of copper cobalt thin film coatings synthesised

using different [Cu]/[Co] ratios, b-d) decoupling of their corresponding

O 1s peaks and shoulders 121

Figure 7.5. Cu L2,3-edge NEXAFS in AEY mode spectra of copper cobalt oxide

thin film coatings 123

Figure 7.6. Co L2,3-edge NEXAFS in AEY mode spectra of copper cobalt oxide


Figure 7.7. O K-edge NEXAFS spectra in AEY mode for copper cobalt oxide


Figure 7.8. Load-displacement curves for the present coating samples 127

Figure 7.9. Mechanical properties of the as-deposited coatings derived

from the nanoindentation tests: (a) elastic modulus (b) hardness and

(c) H/E. The aluminium substrate is used for comparison 128

Figure 7.10. Stress distribution of the [Cu]/[Co]=1 sample obtained from

FEM simulations for different indentation depths: (a) 0.02 µm,

(b) 0.04 µm, (c) 0.06 µm, and (d) 0.08 µm 130

Figure 7.11. Change of the plastic zone size for the [Cu]/[Co] = 1.0 sample as

compared to the aluminium under increasing load as derived from

domain integration of the FEM results 131

xiv

Figure 8.1. (a) XRD patterns of the prepared copper–cobalt thin film

coatings on aluminum substrate at different annealing temperatures,

(b) Expanded XRD patterns within the region 30-42°. 137

Figure 8.2. Cu 2p XPS spectra of copper cobalt thin film coatings synthesised

at different annealing temperatures. 139

Figure 8.3. Decoupling of Cu 2p3/2 peaks of copper cobalt thin film coatings

synthesised at different annealing temperatures. 140

Figure 8.4. Co 2p XPS spectra of copper cobalt thin film coatings synthesised

at different annealing temperatures 142

Figure 8.5. Decoupling of Co 2p3/2 peaks of copper cobalt thin film

coatings synthesised at different annealing temperatures 143

Figure 8.6. O 1s XPS spectra and curve-fittings of copper cobalt thin

film coatings synthesised at different annealing temperatures 145

Figure 8.7. Reflectance spectra and solar absorptance of copper–cobalt oxide

thin film coatings on aluminium substrates synthesised at different

annealing temperatures 146

Figure 8.8. Typical load-displacement curves obtained from different coatings

synthesised at different annealing temperatures 148

Figure 8.9. Mechanical properties of the as-deposited coatings derived from

the nanoindentation tests, (a) elastic modulus, (b) hardness, and

(c) H/E. The wear resistance of aluminium are also displayed

for comparison purpose 149

Figure 8.10. Stress distribution of coating synthesised at annealing temperature

of 650oC, obtained from FEM simulations for different indentation

depths: (a) 0.03 μm, (b) 0.04 μm, (c) 0.05 μm, and (d) 0.06 μm.

The black lines close to the bottom of each model represent the

interface between the coating and the substrate 152

Figure 8.11. Variations of the plastic zone size in coatings synthesised at

annealing temperatures of 500-650oC compared to the aluminium

under increasing load, derived from domain integration of

the FEM results 152

Figure 9.1. Reflectance spectra of copper cobalt oxide thin film coatings with

and without silica AR layer within wavelength range of 0.3-2.7µm

with corresponding solar absorptance (α) values 157

Figure 9.2. Reflectance spectra of copper cobalt oxide thin film coatings

with and without silica AR layer within wavelength range of

3.0-15.4 µm with corresponding solar emittance (ɛ) values 159

xv

Figure 9.3. Reflectance spectrum of copper cobalt oxide thin film coatings

with AR layer (dip-speed of 10 mm/min) within wavelength range

of 3.0-15.4 µm 160


AR layer before and after accelerated thermal durability test at:

a) 265oC for 36 h and, b) 235oC for 179 h 163

Figure 9.5. Photograph pictures of physical condition of copper cobalt oxide

thin film coatings with silica AR layer before (a1) and after (a2)

thermal test at 265oC for 150 h, as well as before (b1) and

after (b2) thermal test at 235oC for 179 h 164

xvi

LIST OF TABLES

Table 3.1. Summary of absorptance (α) and emittance (ɛ) of various SSA materials

produced by sol-gel methods 51

Table 5.1. Metal composition analysis of film coatings using XPS 84

Table 6.1. The binding energies and the percentages of decoupling of Cu 2p3/2

and its satellites of copper–cobalt film coatings synthesised using

various concentrations 106

Table 6.2. The binding energies and the percentages of decoupling of Co 2p3/2

and its satellites of copper–cobalt film coatings synthesised using

various concentrations 108

Table 7.1. Correlation between the [Cu]/[Co] ratio and the porosity 115

Table 7.2. Binding energies and the percentage compositions derived from the

decoupling of Cu 2p3/2 peak and its satellites in the copper cobalt film

coatings 118

Table 7.3. Binding energies and the percentage compositions derived from the

decoupling of Co 2p3/2 peak and its satellites in the copper cobalt film

coatings 121

Table 7.4. Mechanical parameters used for FEM analysis 130

Table 8.1. Results of grain size from Debye-Scherrer formula for the (310) and

(301) lattice planes 136

Table 8.2. Residual stress within the coating layer, estimated by using the (301)

and (301) peak position data from the X-ray diffraction 138

Table 8.3. The curve-fittings results of Cu 2p3/2 and its satellite of copper cobalt


Table 8.4. The curve-fittings results of Co 2p3/2 and its satellite of copper cobalt


Table 8.5. Mechanical parameters derived from the nanoindentation and used for

FEM modelling 151

Table 9.1. Accelerated thermal durability parameter values obtained in the thermal

test 162

xvii

SYMBOLS AND ABREVIATIONS

α Absorptance

αc Absorption coefficient

β1, β2 First and second order flow heat loss coefficients

b0 Collector-specific incidence angle modifier constant

ε Emittance

Δε100 Emittance change measured at blackbody radiation standard of 100oC

ɛs Strain

σs Stress

σy Yield strength

σ Stefan-Boltzmann constant (σ= 5.6696⋅10-8

Wm-2

K-4

)

θ Angle between surface normal and incident irradiance

[ ] Concentration

λ Wavelength

η Conversion efficiency of a flat plate collector

ω Angular frequency

τα Optical transmittance-absorptance product

A Area of the contact made by the indenter

Ac Collector (absorber) area

c Speed of light in vacuum.

CuxCoyOz Copper cobalt oxides

d The thick of slab

D Elasticity matrix

E Elastic modulus/Young’s modulus

Eλb The spectral blackbody radiation

E0 Initial amplitude of the electromagnetic wave

Eb Hemispherical total emitted energy for an ideal blackbody

Eg Band gap energy

FR Collector heat removal factor

Gsc Extraterrestrial solar radiation

GT Total solar energy flux onto the collector surface

h Planck’s constant

H Hardness

HFB Sodium maleat-methyl methacrylates

I0 Intensity of radiation falling upon a material surface

I0 Photon flux incident

Ip Planck black-body distribution

Isol Incoming solar radiation

It Intensity of radiation transmitted through a material

Isat/Imain) Satellite peak intensity to main peak intensity ratio

MxCoyOz Metal-cobalt oxide

M-OH-M Hydroxo bonds with M=metal, H=Hydrogen, O=Oxygen

M(OR)z Metal alkoxides with where R is an alkyl group

P, Pmax Controlled and Maximum load

Qu Instantaneous thermal energy output delivered by collector

R Reflectance

Rs Specular reflectance

xviii

Rd Diffuse reflectance (Rd)

s Selectivity

Sa Arithmetic average height deviation

Sy Peak-to-peak parameter in AFM analysis

S Stiffness

Ti The mean fluid temperature in the collector

Ta The ambient air temperature

T1 Initial temperature in durability test

thkl Grain size

UL Overall heat-losses coefficient

v Poisson’s ratio of the indenter

3-APTES 3-aminopropyltriethoxy silane

AEY Auger Electron Yield

AFM Atomic Force Microscopy

AM Air Mass

AR Antireflection

Ac2O Acetic acid anhydride

BE Binding Energy

CB Carbon Blacks

CBD Chemical Bath Deposition

CD Cyclodextrins

CVD Chemical Vapour Deposition

DOFs Degrees of Freedom

DEA Diethanolamine

EDX Energy Dispersive X-ray

EM Electromagnetic

EOR Oxygen Evolutions Reaction

FEM Finite Element Modelling

FESEM Field Emission Scanning Electron Microscopy

FTIR Fourier Transform Infrared Spectroscopy

GPa Giga Pascal

h, hr hour

HCl Hydrogen Chloride

HPC Hydroxypropylcellulose

ICDD International Centre for Diffraction Data

IEA International Energy Agency

M Molar

MeOH Methanol

MJ Mega Joule

MTES Methyl trimethoxysilane

NEXAFS Near Edge X-ray Absorption Fine Structure

PC Performance Criterion

PEG Polyethylene glycol

PV Photovoltaic

PVD Physical Vapour Deposition

Sat Satellite

SEM Scanning Electron Microscopy

SHC Solar Heating and Cooling

SR-XPS Synchrotron Radiation X-ray Photoelectron Spectroscopy

xix

SSA/SSAs Solar Selective Absorber

T Transmittance

TEOS Tetraethoxysilane / Tetraethyl orthosilicate)

TMOS Tetramethylorthosilane

UV-Vis-NIR Ultra Violet – Visible – Near Infrared

XPS X-ray Photoelectron Spectroscopy

XRD X-ray Diffraction

1

Chapter One

INTRODUCTION

1.1. Background

The sun is an unlimited and environmentally-friendly source of energy. Solar

radiation can be converted into usable forms of energy such as electricity or heat.

Photovoltaic (PV) devices can be used for converting solar irradiation to electricity while

solar thermal collectors can convert solar irradiation to the heat which powers a steam

generator, i.e. solar thermal power. One type of solar thermal collector is the flat-plate solar

thermal collector usually used for water or air heating at low temperatures (< 150ºC) [1-3].

The key component of a flat-plate solar thermal collector is the solar-absorber surface, the

properties of which strongly affect the efficiency of the solar thermal conversion system.

Ideally, such surfaces absorb almost all of the incoming solar radiation (high absorptance)

without losing much of the thermal energy through re-radiation from the heated surface (low

emittance). However, no single material in nature can meet these criteria. As such, there is a

need to tailor the optical and structural properties of a surface through the use of a

combination of materials, the modification of the surface, or the synthesis of multilayer solar-

absorber materials to achieve the desired wavelength selectivity [4, 5]. Such surfaces are

called solar-selective absorber (SSA) surfaces.

Generally, SSA materials are categorized as materials with good optical performance

if they have absorptance values (α) greater than 90% in the solar wavelength range (0.3–2.5

µm) and thermal emittance values (ε) less than 10% in the mid/far-infrared wavelength

ranges (>2.5 µm) [6-8]. In the field of solar thermal energy, the efficiency of the

2

photothermal energy conversion could be enhanced by the development of new selective

absorber materials [9]. Other factors for consideration in the production of photothermal

absorbers are: Long thermal durability, simplicity, and cost-effectiveness in fabrication, as

well as minimal environmental impact in the production process.

Since the mid-fifties, when Tabor [10-12] proposed and demonstrated the usefulness

of selective surfaces for increasing the photothermal efficiency of solar collectors, many

types of solar absorbers have been reported and produced [2]. Electroplating/electrochemical

deposition (including chemical conversion/chemical bath deposition (CBD)) [13-16], vacuum

deposition (physical vapour deposition (PVD)/sputtering) [4, 17, 18], chemical vapour

deposition (CVD) [4, 19], mechanical grinding [7], and sol-gel methods [20-22], are some of

the techniques that have been used to synthesise absorber coatings, but only a handful of

these have been applied on an industrial scale [23].

The most widely-used industrial solar-selective absorbers today are metal particles in

ceramic (cermet) structures which are produced by electrochemical and vacuum deposition

methods. Some well-known examples include electroplated black chrome (Cr–Cr2O3) and

nickel-pigmented anodic Al2O3 (synthesised via the electroplating/electrochemical method)

as well as evaporated titanium nitride film (TiNOx) and nickel-nickel oxide (Ni-NiOx)

(synthesised via a vacuum deposition/sputtering method) [5, 18, 24-29].

Although a significant proportion of flat-plate solar hot water collectors have been

synthesised using these methods, they still have disadvantages. The electrochemical treatment

methods are relatively simple and have a low operating temperature, yet these methods utilize

large amounts of material and are not environmentally-friendly [7, 30, 31]. Vacuum and

sputtering deposition methods are low in material consumption, have good reproducibility

and low levels of environmental pollution; but they are, nonetheless, less cost-effective

because they require a large investment in quite complicated production equipment with high

3

operational costs and high energy intensity in production [4, 6, 7, 23, 30-34]. Other methods

of SSA production such as CVD and mechanical grinding also have their plus points and

drawbacks. In general, the CVD method has good potential in large-scale industrial

production, but there are difficulties associated with ensuring the stoichiometry of the metal

oxides produced [35] and with the vacuum condition. Mechanical grinding is a simple and

cost-effective method of SSA production but the selectivity of the absorber material is low [7,

36].

Recent developments in the synthesis of SSAs highlight the need for the development

of a material which has high selectivity and durability, and which also has a cost-effective

and environmentally-friendly synthesis process. In this context, sol-gel techniques meet these

criteria and they are potentially very promising techniques [24, 37-39]. However, the

application of these techniques to synthesis SSA materials is much less common than

electrochemical or vacuum-based techniques. The sol-gel methods are well-known, simple,

low cost, and environmentally friendly thin film fabrication techniques resulting in a uniform

chemical thin film composition [22, 40]. The sol-gel processes are a soft chemistry method

where the precursors are generally in the form of a colloidal solution that eventually

‘transforms’ into an extensive network of either discrete or continuously-linked molecules.

Sol-gel techniques facilitate control of the coating parameters such as absorber particle size,

particle size distribution, homogeneity, chemical composition and thickness of the film. The

techniques also show good potential for scaling up to an industrial scale [38, 41]. The

synthesis processes are low in material consumption and can be manufactured under ambient

pressure [21, 42]. Bostrom et al. [22] suggested that the sol-gel method in their research was

able to reduce production costs dramatically for absorber thin film fabrication compared to

the sputtering method, because the material cost for the coating itself could be neglected,

compared to the substrate cost. The most important advantage of sol-gel over conventional

4

coating methods is its ability to tailor the microstructure of the deposited film at low

temperatures [43].

1.2. Objective and Scope of Study

The objective of this thesis is to prepare a copper cobalt oxide thin-film coating on an

aluminium substrate as solar selective absorber synthesised via the sol-gel dip-coating

method with emphasis on the material characterisation aspects of structural, surface

morphology and compositions, as well as optical and mechanical properties. The basic

research question is: Can a cost-effective, durable, high performance SSA surface be

produced by using the sol-gel dip-coating method to coat an aluminium substrate with a

copper cobalt thin film?

This thesis consists of 10 chapters. Chapter One introduces the background and the

objective of the work. Chapter Two communicates the theoretical aspects related to the

interaction between the thin-film coating and the solar radiation and the designs for solar

selective absorbers. The mechanical properties, the degradation processes of selective

absorbers as well as the thermal stability aspects are also described. In Chapter Three the

state-of-the art with respect to sol-gel selective absorber coatings is presented. The theoretical

aspect of sol-gel synthesis is also briefly explained. Most of Chapter One and Chapter Three

have been submitted as a review paper for publication (Journal article 5). Chapter Four

explains the experimental methods, consisting of sample manufacturing, the equipment used

and the characterisation techniques. Chapter Five to Chapter Nine contain the main results

obtained from characterisation studies on the sample prepared and their further discussions.

The research results from Chapter Five, which contain the exploration results, have been

published in Surface and Coatings Technology journal (Journal article 1). Results from

Chapter Six, which contain the studies of surface morphology and surface composition of the

5

copper cobalt oxide coatings and the absorptance optimisation, have been published in

Applied Surface Science journal (Journal article 2). The research results from Chapter Seven,

containing the influence of composition in the synthesis process of copper cobalt oxide

coatings have been published in The Journal of Physical Chemistry C (Journal article 4). The

research results from Chapter Eight to Chapter Nine, containing the influence of annealing

temperature changes in the synthesis process of copper cobalt oxide coatings and the addition

of a silica antireflection layer, have been submitted to journals for publication (Journal article

6 and 7). Finally the last chapter is the conclusion. The need for future development, based on

results obtained, is also briefly discussed.

6

Chapter Two

THEORETICAL BACKGROUND

2.1. Solar Radiation, Thermal Radiation and Solar Selective Absorber

The Sun is one of the main energy sources for the Earth. The estimated total

electromagnetic radiation power of the Sun is about 3.8445⋅1020

MW. The extraterrestrial

solar radiation value (Gsc) prior to entering the Earth’s atmospheric absorption (just above the

Earth’s atmosphere) is almost constant with the numerical value of 4.921 MJ/m2hr [44]. This

extraterrestrial solar electromagnetic radiation consists of around 6.4% in the UV range, 48%

in the visible range and the rest in the near-infrared range [44, 45].

In the ionosphere, in the ozone layer or in the atmosphere, most of the radiation is

absorbed or scattered by nitrogen, oxygen, ozone, water vapour, and carbon-dioxide or by

other particles compounds in the atmosphere. Each of the atmospheric compounds absorbs

certain wavelengths causing absorption holes in the terrestrial solar spectrum, forming

Rayleigh attenuation as shown in Figure 2.1 [44]. Rayleigh scattering [45] refers to light

scattering by the molecules in the air. For smaller particles, as compared to the incoming

light wavelengths, regardless of their shape, a strong Rayleigh scattering that is symmetric in

forward and backward directions dominates. One example of Rayleigh scattering is the

scattering of solar radiation by atmospheric molecules which gives the blue colour to the sky.

Besides scattering and absorption, the terrestrial solar spectral distribution and

intensity are also influenced by the paths of the ray that traverse the atmosphere (i.e. the air

mass or AM) [46]. The air mass is defined as the ratio of optical mass at a slant path to the

vertical path. For example, the air mass has a value of 1 if the Sun is at the zenith (directly

7

overhead), i.e. its rays are normal/perpendicular to the horizontal surface of the Earth. Based

on the ISO 9845-1:1992 [47], when the Sun is about 41.8o above the horizon, the air mass is

1.5 (AM1.5). In this thesis, AM1.5 is used to characterise the absorptance value of solar-

selective absorbing surfaces on aluminium substrates as described by Duffie and Beckman

[44]. However, only small differences on the solar-weighted optical properties of selective

solar absorbers is found with variation in air mass application [48].

Figure 2.1. Effects of Rayleigh scattering and atmospheric absorption on the

spectral distribution of solar irradiance. Adapted from [46].

When a body surface becomes warmer than the surroundings due to receiving

amounts of solar energy or other energy sources, it actually has a net thermal electromagnetic

radiation transfer to the surroundings where the wavelengths and the intensity of the thermal

radiation depend on the temperature of the body and its optical characteristics. Thermal

radiation is usually referred to as “the blackbody radiation” which is an ideal surface that

absorbs all wavelengths of the incident radiation and emits the maximum amount of energy

8

according to Planck’s Law [49]. Theoretically, the spectral blackbody radiation Eλb is given

by formula:

]1[)/(5

1

2

TCbe

CE

(2.1)

where C1=3.7405⋅10-16

m2W and C2=0.0143879 mK are the first and the second Planck’s

radiation constants, respectively. T is the temperature of the blackbody in Kelvin. The Stefan-

Boltzmann law gives the hemispherical total emitted energy for an ideal blackbody as:

Eb = σT4 (2.2)

where σ is the Stefan-Bolzmann constant (σ= 5.6696⋅10-8

Wm-2

K-4

).

The standard spectral solar flux incident at the surface of the Earth, after atmospheric

absorption, is limited to the range between 0.3 and 2.5 μm i.e. UV/Vis/NIR wavelength

ranges with the maximum solar intensity is around 0.55 µm, whereas the optical properties of

a real body in the infrared wavelength range can be characterised by its thermal emission

compared to the ideal blackbody. Figure 2.2 shows the solar hemispherical irradiance for

AM1.5 and blackbody-like emission spectra at 100°C, 200°C and 300°C. Basically, there is

no significant overlapping between the solar radiation confined in wavelengths range of 0.3-

2.5 μm and the emitted thermal radiation in wavelengths range of 2-30 μm especially for

temperatures below 200°C. From Figure 2.2 it also can be seen that if the temperature of the

blackbody increases, the amount of the emitted energy also increases, and the location of

peak power density shifts towards shorter wavelengths. These profiles suggest a possibility of

designing a thin film material which absorbs the maximum amount of incident solar

radiation, and re-emits a minimum amount of the absorbed energy, and it could be called a

solar selective absorber surface.

9

Figure 2.2. Solar hemispherical spectral irradiance for air mass 1.5 and blackbody-like

emission spectra at 100°C, 200°C and 300°C. Adapted from [6, 47].

2.2. Optical Properties of Thin Film

2.2.1. Electromagnetic radiation absorption

An electromagnetic wave propagation along the x-axis through an absorbing medium

at time t can be described by its electric field component, E(x,t), [50, 51],

))/(/exp(),( 0 tcxnicxkEtxE (2.3)

where E0 is the initial amplitude of the electromagnetic wave of angular frequency ω before

entering the medium, while c is the speed of light in vacuum.

The complex refractive index is given by equation:

iknN (2.4)

where the real part stated with n has relation to the phase of the wave. The imaginary part of

10

the equation (k), which is also known as “extinction coefficient”, describes damping

amplitude in the direction of propagation.

The intensity of electromagnetism is proportional to E2 which leads to Beer’s Law

which shows the change in intensity (I) of an electromagnetic wave when propagating

through a medium.

I=I0 exp(-αcx) (2.5)

where I0 is the initial intensity of the electromagnetic wave before entering the medium and

αc is the absorption coefficient.

The combination equation (2.3) and equation (2.5) results:

ckc 2 (2.6)

while the angular frequency ω can be expressed in:

c2 (2.7)

where λ is the wavelength of electromagnetic in vacuum. By assuming the medium as thin

slab d, then equation (2.5) in combination with equation (2.6) and (2.7) results:

)(4)(ln 0 kddxII (2.8)

which illustrating the relative intensity drop in a thin film medium.

The kd/λ factor in equation (2.8) is very important when designing a spectrally

selective absorber film. The film will be transparent when λ>>kd. The transition area from

low to high reflectance at wavelength area around of 2-3 µm, according to Figure 2.2, is

determined by the selection of k and d values. For an intrinsic type of selective absorber using

a single substance, which means that the condition is at certain wavelength, k is not a variable

and d is the only parameter which can be changed. The ceramic-metal (cermet) system

absorber offers more options. By altering the filling ratio or the size and the shape of the

metal particles, the k value of the cermet can be changed and subsequently the transition

11

profile can be approached by step function. Equation (2.8) does not consider thin film

interference which can be tuned to improve the selectivity marked by a sharper transition

curve profile [51].

2.2.2. Optical characterisation of selective solar absorber

The optical characterisation of selective absorber surface can be described in terms of

absorptance, reflectance, transmittance and emittance. For a certain angle of incoming

incident radiation, the absorptance (α), the reflectance (R) and the transmittance (T) are the

fractions of incident of radiation absorbed, reflected and transmitted by the absorber material,

respectively. The energy conservation gives

α(λ) + R(λ) + T(λ) = 1 (2.9)

For an opaque surface or thin film coating surface on reflective metal substrate, the angular

absorptance can be expressed in terms of the angular total reflectance

α(λ) = 1 - R(λ) (2.10)

For a range of solar wavelengths, the total solar absorptance is defined as a weighted

fraction between absorbed radiation and incoming solar radiation (Isol), while thermal

emittance (ε) is defined as a weighted fraction between emitted radiation and the Planck

black-body distribution (Ip), and both can be calculated in terms of the surface reflectance

(R(λ)) using equations [22]:

dIdRI solsol )())(1)((

5.2

3.0

5.2

3.0

(2.11)

20

5.2

20

5.2

)( )())(1)(( dIdRI ppT (2.12)

The solar spectrum (Isol) used here is defined according to the ISO standard 9845-1 (1992)

with an air mass of AM1.5. The absorptance and emittance values can also be measured

using tables of spectral distribution versus equal energy increments [44].

12

The parameter to evaluate the efficiency of the solar absorber surface in spectrally

selective solar absorber applications is the selectivity (s). The selectivity is described based

on the absorptance (α) of solar radiation and the emittance (ɛ) of thermal radiation [19, 40]:

s = α / ɛ (2.13)

The ideal solar selective absorber surface should have a maximum absorptance value

of near α=100% and a minimum an emittance value of near ɛ=0% (Figure 2.2).

Unfortunately, there is no ideal intrinsic absorber surface existing in nature; as such, there is a

need to tailor the optical and structural properties of a material approaching the ideal surface

through the various designs.

2.3. Solar Selective Absorber Design

There are several ways of designing the solar-selective absorbing surfaces on

substrate. The different designs result in the different optical absorption mechanisms such as

optical trapping absorbers, metal-dielectric multilayer absorber, absorber-reflector tandem,

quantum size effects, etc. The descriptions about various selective absorbers synthesised by

different methods can be found elsewhere [52-58]. Generally, the dark mirror absorber-

reflector tandem is the most common commercially available selective absorber design [58-

61].

An absorber-reflector tandem configuration is obtained by superimposing one or more

layers on the top of substrate in which the layer and the surface have different optical

properties. If the layers are highly-absorbing in the solar region and the substrate is non-

selective highly-reflecting material then this configuration is known as a dark mirror. Figure

2.3 describes the dark mirror absorber-reflector tandem design. If the layer is solar

transparent material and simultaneously also infrared reflector material whereas the surface is

highly solar absorber, then it is known as a heat mirror. In the dark mirror coatings, the

13

absorber layer on the top of substrate can be a semiconductor, absorber particles embedded in

dielectric matrix or selective paint. Several designs of materials for the construction of the

dark mirror absorber-reflector tandem are given below.

Figure 2.3. Dark mirror absorber-reflector tandem design

Semiconductor thin film coatings

A low band gap semiconductor coating has the characteristic of absorbing the solar

radiation and transmitting the mid-far infrared. It is because the semiconductor absorbs

photons having energies greater than the band gap and it will raise the material’s valence

electrons into the conduction band. Photons with energies less than the band gap energy are

transmitted through the coating [62]. If a semiconductor is deposited on the top of highly

infrared reflecting metal substrate, then a spectrally selective semiconductor coating will be

obtained.

According to the equation (2.14), to absorb all solar radiation below wavelengths of

λ=2.5 µm, the semiconductor should have a band gap (Eg) of 0.5 eV. However, the main

14

obstacle is the difficulty finding a suitable semiconductor. Most semiconductors have too

large band gaps, corresponding to too short wavelengths. Lead sulphide (PbS) is a suitable

semiconductor with a band gap of 0.4 eV [52]. Unfortunately, this material is very poisonous

for humans and the environment and is not commercially feasible.

gEhc (2.14)

where h is Planck’s constant and c is light speed in vacuum.

Another problem with semiconductors is that they have a high refracting index which

results in low absorptance in the air-coating surface interface. To obtain high solar

absorptance, the refractive index of the semiconductor should be as low as possible. The

absorptance can be increased by controlling the thickness of the coating to reduce the

interference effect or by applying an antireflection layer. An example of a semiconductor

metal tandem selective absorber is the chemical vapour deposition of silicon in a stack of

SiO2/Si33N4/Si/Cr2O3/Ag/Cr2O3 on stainless steel with an antireflection coating on top of the

multilayer stack [62, 63].

Composite thin film coatings

Certain metallic clusters embedded in a ceramic/dielectric matrix (cermet) composite

coating such as Cr-Cr2O3, Ni-Al2O3, Mo-Al2O3, or Ni-NiOx exhibit good solar spectral

selective absorption. The coatings strongly absorb solar radiation and are almost transparent

in the infrared region. The spectral selectivity of a cermet coating is enhanced by using a

highly infrared reflecting (poor thermal emitter) metal substrate [42, 64]. The concept of

using a cermet material to form a tandem structure with a poor thermal emitter metal

substrate has been investigated both theoretically and practically [18, 65].

Cermet selective absorbers usually consist of nanometer-sized metal particles (1–20

nm) [32] and the effective medium theories can be used to model the optical properties of the

15

film [66, 67]. Simulations have proved that a ceramic-metal solar absorber with an AR layer

could achieve absorptance values of 0.91–0.97 and emittance values of 0.02–0.07 [68]. The

metallic particles are usually transition metals which are uniformly distributed in the matrix

or gradient index with gradually increasing particles content from the upper-limit of the

matrix towards the substrate surface. Figure 2.4 shows the microstructures of two different

examples of composite coating solar selective absorbers. In the nickel pigmented anodic

aluminium oxide (Ni-Al2O3) microstructure, the particles are uniformly distributed in the

matrix, while in the sputtered nickel/nickel oxide (Ni-NiOx) microstructure the particles are

arrayed with graded index composition [18, 69, 70]. The metal particles in the cermet act as a

modifier for the optical response of the ceramic phase [71, 72]. The absorption in a cermet

coating is a result of light scattering by the boundaries between the metallic phase and the

oxide (dielectric) phase [73, 74].

The cermet system offers a high degree of flexibility with optical parameters which

can be tuned by controlling the metal content, the shape, orientation and size of the small

metallic clusters as well as the optical constants of the constituents. The thickness and

chemical nature of the dielectric phase can be adjusted to obtain the desired spectral

selectivity. The type of matrix also influences the quality of the film. In this regard, a porous

matrix is the optimum host for metal particle inclusion [2, 22, 42, 64, 75]. The surface

morphology of the cermet also plays a significant role in determining the surface absorptance

and can favour multiple reflections in the surface, thus enhancing the solar radiation

absorption [8]. By varying many of the parameters listed above, countless combinations can

be created, thus the required spectral selectivity can be easily achieved [22].

16

Figure 2.4. Microstructure pictures of two examples of solar absorber composite

coatings (adapted from [62] and [76]).

Selective paint, spinels and metal oxide absorber coatings

The selective paint absorber is a simple and less expensive selective absorber because

it can be produced by using the low-cost sol-gel process. This type of absorber is usually used

for a façade coating with a certain purpose. Factors determining the optical performance of

selective paint type absorber include intrinsic optical constants, particle size-dependent

scattering and paint binder [29, 62]. Generally, the selective paint absorber has a low

selectivity due to high thermal emittance.

The selective paint coating is composed of pigment absorber particles dispersed in a

resin/binder agent where they uniformly form the coating matrix. Some pigments, mostly

from transition-metal oxides, have high solar absoption which is due to the existence of

numerous spin-allowed electron transitions between partially filled d-orbital [29]. Polymer

binder, such as silicone, siloxane resin or phenoxy resin, is usually used in the selective paint

coating. Unfortunately, the binder agent absorbs strongly in the thermal IR range increasing

17

the thermal emittance significantly. Another disadvantage is that it is impossible to make

paint coatings thinner than 1–2 µm because the thickness of the paint layer is limited by the

size of the ground pigment particles [29]. Usually the pigment particles will agglomerate and

their size will be comparable to or larger than the incident wavelength of light, reducing the

paint performance. An example of commercial selective solar-absorbing paints is the

Solarect-ZTM

which is synthesised using siloxane resin and an inorganic pigment of

FeMnCuOx with a pigment volume fraction of about 0.2 [77].

Efforts to decrease the emittance value of paint coating have been done by other

researchers [20, 29]. They prepared the CuCoMnOx pigment coating without binder via the

sol-gel method, forming a spinel-type absorber. Other researchers prepared solar absorber

which consisted of less than three components of transition metal forming a spinels or metal

oxide absorber coating, and even the CuMn spinel oxide coating has reached a promising

performance for use on an industrial scale [24, 38, 78]. The review of metal oxides, spinels

and composite selective absorber coatings synthesised using the sol-gel method can be found

in Chapter 3.

2.4. Anti-reflection Layer

Light reflection is a phenomena occurring when light propagates across a boundary

between two media which have different refractive indices. In a spectrally selective absorber

application, the incident solar radiation should be absorbed to the maximum possible without

reflection from the surface as the reflection of light is undesired. One approach to achieve

low reflection is the use of antireflection (AR) deposited on top of the absorber surface [79].

The best refractive index value of the AR layer is when its refractive index is equal to the

square root of the refractive index of the material on which it is deposited, by assuming a

vacuum surrounding. However, this approach is only valid for non-absorbing dielectric

18

materials. The AR layer will increase solar absorptance without increasing the emittance

value for a sufficiently thin (<100 nm) AR layer [51]. Silica or Titania normally provides

suitable refractive indexes as AR thin films.

2.5. Flat Plate Solar Collector

Basically, there are at least four types of solar collector i.e., flat plate, stationary

parabolic, evacuated tube, and sun-tracking concentrating collector. Kalogirou, et al. [1] gave

a review of these collectors. In this thesis, we focus on the flat plate solar collector. A cross-

section view of a commonly used flat-plate solar collector design is shown in Figure 2.5.

Generally, the main components of the flat plate collector consist of a transparent

cover, plate absorber (film coating and metal substrate), fluid conduit and insulation. When

solar radiation passes through a transparent cover and hits the plate absorber, a large portion

of solar radiation is absorbed and converted into thermal energy; then the thermal energy is

transferred to the transport medium in the fluid conduit to be carried away to storage. The

fluid conduit can be welded to the absorbing plate, or it can be an integral part of the plate.

The insulation in the underside and the side of the casing is used to reduce the conduction

losses.

Besides conduction losses, heat losses can also be caused by convection and radiation

from the hot surface. To suppress the radiation heat loss, a transparent glazing cover is used.

A suitable transparent cover, like a pre-stressed low iron glass, is transparent for the solar

spectrum but opaque for thermal radiation/infrared wavelengths and thereby reduces thermal

radiation emitted from the absorber. The transparent cover also suppresses convection heat

loss, and to minimize losses, the spacing between the cover and absorber should be between

10-15 cm [44]. Convection losses can be reduced further by using an additional transparent

19

insulation material such as a thin transparent foil or honeycomb between the cover and the

absorber [62].

Figure 2.5 Cross-sectional view of a basic flat plate solar collector (adapted

from: [44] and [62]).

The conversion efficiency of a flat plate collector (η) limited by thermal losses is

given as [44]:

T

aiLR

R

Tc

u

G

TTUFF

GA

Q )()(

(2.15)

where Qu is the instantaneous thermal energy output delivered by collector, Ac is the collector

(absorber) area, GT is the total solar energy flux onto the collector surface, FR is the collector

heat removal factor namely a quantity that relates to the actual useful energy gain of a

collector to the useful gain if the whole collector surface were at the fluid inlet temperature

[44], τα is an optical transmittance-absorptance product measured from experiments that is

weighted according to the proportions of beam, diffuse, and ground-reflected radiation on the

collector, UL is overall heat-losses coefficient, Ti is the mean fluid temperature in the

collector and Ta is the ambient air temperature (°C).

If FR and UL are categorized as having a slight variation (not significant) in the

operation range of the collector and most of the radiation is beam radiation that is near

20

normal to the collector, then FR(τα) indicates the amount of energy absorbed and FRUL

indicates the amount of energy lost in Wm-2

K-1

. Equation (2.15) shows that the heat

collection efficiency is directly proportional to the solar absorptance, and it decreases with an

increase of the operating temperature. At high temperatures, radiation losses are dominant as

compared with conduction and convection losses. According to Haitjema [80], at high

collector temperatures the emittance determines the collector efficiency, while at a low

collector temperature the absorptance determines the efficiency. Hence it is important to note

that an increase in the solar absorptance is considered more important than an equal decrease

of the thermal emittance for better flat plate collector performance [81, 82].

Further, Perers [83, 84] developed a dynamic approach using a multiple regression

method to measure the collector efficiency made by taking into account the thermal

capacitance effects, incidence angle effects and the temperature, as well as the wind and sky

temperature dependency of the heat loss coefficient. The simplified version of the Perers’

formula to dynamically measure the energy output (Qu) is given as:

dt

dTmCTTTTGKQ

f

effaiaiTu )()()()( 2

210 (2.16)

where η0Kτα(θ)GT part is the optical efficiency of total irradiance, β1 and β2 are the first and

second order flow heat loss coefficients, respectively, and (mC)eff is the effective thermal

capacitance of the collector, while dTf /dt is the mean time derivative for the fluid temperature

(°C/s). The Kτα(θ) is an incidence angle modifier which is typically used for estimating the

dependency of a collector on the angle of incidence of impinging radiation [44]:

)1cos

1(1

)(

)()( 0

bK

n

(2.17)

where n is the surface normal (zero angle of incidence), θ is the angle between surface

21

normal and incident irradiance, and b0 is a collector-specific incidence angle modifier

constant obtained from the experiment.

2.6. Mechanical Properties of Thin Film Coating and Modeling

Mechanical properties such as elastic modulus/Young’s modulus (E) and hardness (H)

are needed to predict the wear resistance of optical solar selective absorber thin film coating

material. The wear resistance is important to maintain the performance and function of the

optical coatings during the time of service. The wear resistance of the coating can be

evaluated using the E and H values obtained from nanoindentation experiments [85].

Oliver and Pharr [86] introduced an improved method for determining hardness,

elastic modulus and resistance to deformation character from a set of load-displacement

nanoindentation curves (P-h curve) obtained during one cycle of loading-unloading without

the need to image the hardness impression. A controlled load (P) is applied through a

diamond indenter which is in contact with the film surface. During the loading-unloading

contact cycle, the load and penetration depth (h) are monitored. The penetration depth of the

indenter tip into the material or displacement is recorded as a function of the applied load

forming a P-h curve. From the P-h curve, maximum load (Pmax), maximum displacement

(hmax) and stiffness (S) can be determined (Figure 2.6).

The hardness (H) can be defined as how resistant solid matter is toward various

permanent shape changes when force is applied and it can be estimated from the equation:

H = Pmax / A (2.18)

where Pmax is the maximum load and A is the area of the contact made by the indenter, while

elastic modulus or Young’s modulus (E) is a material’s stiffness in the elastic region (where

Hooke’s Law applies). It is defined as the ratio of stress to strain. From the P-h curve (Figure

22

2.6), elastic modulus can be measured from the relationship between stiffness (S) and the

contact area (A):

AES eff

2

(2.19)

where β is a dimensionless constant taken as unity and Eeff can be defined as:

)11

(1

22

i

i

eff E

v

E

v

E

(2.20)

where E and v represents the elastic modulus and Poisson’s ratio of the indenter and Ei and vi

refers to the elastic modulus and Poisson’s ratio of the indented material, respectively [87].

Figure 2.6. Typical loading-unloading compliance curve from a nanoindentation

experiment with maximum load (Pmax) and depth beneath the specimen free

surface (hmax) [86].

To visualize the stress distribution within the coating and the substrate under the

indentation tests to assess the mechanical response of the coating system to external loading,

modelling is needed. Finite element modelling (FEM) is a versatile and reliable method to

implement this aim. The finite element modelling is a computational method that subdivides

23

an object into very small but finite-size elements. The physics of one element is

approximately described by a finite number of degrees of freedom (DOFs). Each element is

assigned a set of characteristic equations (describing physical properties, boundary

conditions, and imposed forces), which are then solved as a set of simultaneous equations to

predict the object’s behaviour.

In a material, the stress-strain relationship for linear conditions is given as:

σ = D ε (2.21)

where D is the 6×6 elasticity matrix, and the stress (σs) and strain (ɛs) components are

presented in vector form with the six stress and strain components in column vectors

described as:

xz

yz

xy

z

y

x

s

, and

xz

yz

xy

z

y

x

s

Of which the σs, τ, ɛs , and γ are the normal stress, shear stress, normal strain and shear strain

respectively, in x, y and z directions. And it is possible to completely describe the strain

conditions at a point with the deformation components -(u, v, w) in 3D and their derivatives

as well as the shear strain which can be expressed in a tensor form, εxy, εyz, εxz or in an

engineering form, γxy, γyz, γxz as:

z

w

y

v

x

uzyx

;;

;2

1

2

x

v

y

uxy

xy

;

2

1

2

y

w

z

vyz

yz

x

w

z

uxz

xz2

1

2

24

For isotropic materials, the inverse elasticity matrix (D-1

), which is also known as the

flexibility or compliance matrix, is described as:

)1(200000

0)1(20000

00)1(2000

0001

0001

0001

11

v

v

v

vv

vv

vv

ED (2.22)

where E is the modulus of elasticity and ν is Poisson’s ratio, which defines contraction in the

perpendicular direction. Inverting D-1

gives:

2

2100000

02

210000

002

21000

0001

0001

0001

)21)(1(

v

v

vvvv

vvv

vvv

vv

ED (2.23)

By inputting the parameters of mechanical properties of the coating system obtained from the

nanoindentation tests, stress distribution can be visualized using computational software.

2.7. Degradation of Selective Absorber and Accelerated Ageing Test

Besides efficiency, durability is also an important factor for selective absorber coating

in operation. The micro structure of a thin coating can change due to factors such as high

temperature, high air humidity, air pollutants such as Sulphur Dioxide, Sun (UV) radiation,

dirt, etc. resulting in the deterioration of coating optical selectivity quality [88, 89]. High and

long temperature exposure can quicken the oxidation processes; high levels of humidity and

air pollutant may trigger the corrosion processes and the Sun radiation may initiate

25

photochemical redox reactions. Therefore, the degradation under operation conditions (micro

climate) for a selective absorber coating should be tested.

There are at least two approaches to assess the durability of a solar absorber, namely

an in-situ test and an accelerated ageing test. The in-situ test is the more accurate method

used to evaluate the durability under normal working conditions. However, this test is very

hard to carry out because of the length of time required to get results [51]. The accelerated

ageing test was developed in place of exposing the absorber surface to its natural working

conditions for many years.

The International Energy Agency (IEA) developed an accelerated aging test to assess

the durability performance of a collector called performance criterion (PC) through the IEA

SHC Task X. The accelerated aging test is carried out to determine the estimated service

lifetime of a selective absorber surface for the standard of solar collector. This test procedure

assumes that the activation energy of a certain degradation process is high enough to ensure

absorber durability under natural working conditions of a flat thermal collector [38]. Only an

absorber with a minimum lifetime of 25 years can be categorized as qualified [90].

The PC value describes the influence of micro climate conditions to the change of

solar absorption (Δα) and emittance (Δε) [38, 88, 91]:

PC = −Δα + 0.25Δε100 ≤ 0.05 (2.24)

where the factor of 0.25 is a weighing factor that reduces the importance of a change in

thermal emittance compared to a change in solar absorptance. This formula implies that the

optical performance of the system would decrease less than 5% due to coating degradation

within an estimated service lifetime of 25 years [38, 88]. In the next revision of IEA SHC

Task 27, the weighing factor of 0.5 is found to be more appropriate [90];

PC = −Δα + 0.5Δε100 ≤ 0.05 (2.25)

26

It is possible to get a negative PC value, which indicates an actual improvement of the optical

selective properties of the surface.

In this study, only the accelerated thermal durability test using PC criteria of IEA

SHC Task 27 was carried out, since most of the real application of selective absorber was

strictly isolated under a transparent glass cover or in the vacuum tube condition, therefore, in

this situation, the thermal test became the most important factor determining the quality of

the absorber film.

27

Chapter Three

REVIEW OF SOL-GEL SOLAR SELECTIVE ABSORBERS

3.1. Sol-gel Synthesis Process

The regular sol-gel steps in synthesising SSA surfaces generally consist of substrate

surface cleaning, sol solution preparation, film deposition and heat treatment/calcination. In

the solution preparation step, Brinker et al. [43] briefly explains that the sol-gel process

involves the use of solid inorganic or metal organic compounds as raw ingredients

(precursors) in a solvent forming a colloidal dispersion. For more detail, in aqueous or

organic solvents, these precursors were hydrolyzed and condensed to form inorganic

polymers (network) composed of oxo (M-O-M) or hydroxo (M-OH-M) bonds. For inorganic

compounds, hydrolysis proceeded by elimination of a proton from an aquo ion [MONH2N]z+

to

form a hydroxo (M-OH) or oxo (M=O) ligand (M=metal). Condensation reactions involving

the hydroxo ligands resulted in inorganic polymers in which metal centers were bridged by

oxygens or hydroxyls [43]. Any precursors which form in an inorganic network subsequently

can be utilized in the sol-gel technique. The most frequently used metal organic compounds

were metal alkoxides M(OR)z, where R was an alkyl group CxH2X+1 [43]. Normally, the

alkoxide was dissolved in alcohol and hydrolyzed by the addition of water under acidic,

neutral, or basic conditions. Hydrolysis resulted in the substitution of an alkoxide with a

hydroxyl ligand [43]:

M(OR)z + H2O → M(OR)z-1OH + ROH (3.1)

In essence, the preparation of a sol-gel solution involves the use of inorganic or metal

organic compound aqueous organic/alcoholic solvent with the addition of an acid/base

28

conditioner as a catalyst [43]. In this review, any modifications in the solution preparation

method as elucidated earlier, such as the addition of a complexing agent, the addition of

metal oxide powder or the addition of other additives, as well as the solution preparation

without the addition of a catalyst (sol-gel-like) are also classified as sol-gel methods.

In the deposition step, there are several deposition options such as dip-, spin-, flow-,

spray- and roll-coating which can be used to coat a surface with a sol solution [92]. Using

these various deposition options, it is possible to prepare different materials in various forms:

monoliths, powders, fibers or thin films. Since the precursors are mixed at the beginning of

the synthesis, the processing temperatures are generally lower compared to equivalent solid-

state synthesis methods such as those mentioned earlier. In addition, by using the sol–gel

method it is possible to make multi-component films with a complex structure [29].

3.2. State of the Art of the Sol-gel Selective Absorber Coatings

This sub-chapter reviews the developments in the synthesis of flat-plate SSA

materials produced by sol-gel methods. There are three major categories of sol-gel synthesis

solar selective absorber materials.

3.2.1. Metal oxide based selective absorber

Metal oxide, either stand-alone or blended with other compounds can be simply

synthesised using sol-gel methods. Generally, the synthesis route is relatively short and

without the requirement for inert conditions in the calcination step (the heating step following

the coating step). This is the reason why research on this subject is relatively extensive. A

review of the synthesis and development of this type of SSA material is presented below.

3.2.1.1 Copper oxide-based absorber

Copper oxide (CuO), which is well-known for having good optical properties as SSA

material, is inexpensive and easy to process using sol-gel methods [93]. The other principal

29

oxide of copper, Cu2O, also exhibits good solar absorption, but its absorption is lower than

CuO [94]. Hottel and Unger [95] prepared bare CuO as an SSA coating on flat-plate

collectors. Coating deposition was carried out by spraying a dilute solution of cupric nitrate

onto an aluminium sheet, which converted the cupric nitrate to black cupric oxide by heating

it to above 170°C. This oxide film had α= 0.93 and ε=0.11 at 80°C. This selectivity value is

comparable to that obtained from various methods such as sputtering (α=0.75, ε=0.1) [94,

96], CVD (α=0.73-0.9; ε=0.04-0.52) [19, 96], electrochemical (α=0.94, ε=0.08) [16] and

combination methods (thermal, chemical and electrochemical) (α=0.97; ε=0.2) [97]. The bare

copper oxide experienced significant absorptance degradation after exposure to higher

temperatures (above 150oC) in air. This was associated with a chemical structure change [97]

and a decrease in the surface roughness of the coatings by heat [2, 98]. This has hindered a

more extensive application, so further modification is required to improve its durability.

Efforts to protect the bare CuO absorber have been made by many researchers using

various methods as summarised by Sathiaraj [99]. Barrera et al. [93] overcame the problems

associated with bare CuO by protecting it in a silica matrix forming a CuO-SiO2 composite

absorber using a sol-gel process. Silica was selected as the matrix due to its stable oxide state,

ease of manufacturability and cost-effectiveness. The sol was prepared by adding Cu-

propionate solution in a TEOS (tetraethoxysilane/tetraethyl orthosilicate) solution and

subsequently adding HCl as an acid catalyst. Film deposition was accomplished by dip-

coating on stainless steel substrates. In all cases, the final film was further annealed in air at

4500C for 4 hours [93]. Barrera-Calva et al. [93] suggest that during the annealing process,

copper-propionate complexes developed into particulate polycrystalline CuO dispersed in a

partially crystallized silica matrix. The thermal analysis of gel revealed that the synthesised

material might be stable up to 400oC. The solar parameters of such a system were strongly

influenced by the thickness and phase composition of the CuO-SiO2 film. Interestingly, the

30

best solar parameters (α = 0.92 and ε = 0.2) were associated with the thinnest films (one

dipping cycle) which comprised a CuO-Cu2O mixture embedded in a partially crystallized

silica matrix [93]. However, the relative high emittance values in this research was be due to

the strong silica phonon absorptions [71].

One way used to attain maximum solar absorption and reduce thermal emittance is to

synthesise an optimized porous antireflection (AR) layer or matrix with an optimized surface

roughness as suggested by Farooq and Lee [65]. The porosity reduces the refractive index of

the AR layer or matrix in which the refractive index can be optimized via tuning (i.e. square

root of the refractive index of the underlying material), whereas the increase in roughness of

the film up to 1 × 10-7

m rms (root mean square) increases the absorption linearly. Any

further increase of roughness raises the thermal emittance, because of the thermal radiation

absorption [65].

3.2.1.2. Cobalt oxide-based absorber

Aside from copper oxide, cobalt oxides (CoO or Co3O4) also have good optical

properties as SSAs and are comparatively easy to synthesis. The idea of using cobalt oxide as

a selective absorber material was first introduced by Gillette [100, 101]. In terms of sol-gel

techniques, many researchers are more interested in the synthesis of cobalt oxide than copper

oxide. This is attributed to the fact that cobalt oxide is more stable at high temperatures than

copper oxide. For example, two types of cobalt oxides, namely, Co3O4 and CoO are stable at

temperatures above 5000C [101-103]. However, cobalt oxide precursors are relatively more

expensive than the copper oxide precursors, though this cost is still negligible compared to

the substrate cost.

Choudhury et al. [104] synthesised a black cobalt selective surface by spray pyrolysis

on top of commercial aluminum and galvanized iron substrates. They found that the film had

31

a relatively good selectivity. Optimized films on aluminum substrate (about 0.21 μm thick)

had α=0.92 and ε100°C =0.13 while films on galvanized iron substrate (film thickness = 0.24

μm) had α=0.91 and ε100°C=0.12. Accelerated ageing studies indicated that these films had

good adhesion to the substrates. Nonetheless, the films were only stable up to 220°C and

there was degradation at higher temperatures [104, 105].

In a separate study, Chidambaram et al. [106] prepared cobalt oxide coatings by spray

pyrolysis on stainless steel substrates at 300°C. The coatings adhered strongly on the

substrate and were stable up to 600°C. Auger electron spectroscopy, X-ray photoelectron

spectroscopy and X-ray diffraction investigations revealed that the coatings consisted of an

upper layer of Co3O4 with a CoO layer nearest to the substrate. The integrated solar

absorptance value α was 0.93 and hemispherical emittance value ε (at 100°C) was 0.14.

However, heat treatment for a few hours at 600oC changed these absorptance and emittance

values to 0.89 and 0.19, respectively [101].

Chidambaram et al. [106] indicated that a lower substrate temperature of about 150°C

could be used for the preparation of coatings if an equimolar aqueous solution of cobaltous

acetate and thiourea was used. These coatings contained both cobalt oxide and cobalt

sulphide and exhibited comparable absorptance values, but they had higher emittance values.

The addition of cobalt sulphide rendered lower quality of the sulfured film and it became

worse after thermal annealing [107]. These coatings were stable only up to about 250-300°C

[106, 108]. Further work is necessary to improve the coating quality obtained using this

method. Barrera et al. [109] suggests the use of stainless steel containing nickel, or copper, as

a substrate.

Uma et al. [108] expected that if another stable oxide (like iron oxide) was added to

the cobalt oxide precursor solution system then higher stability and optical performance could

be achieved because the cobalt oxide-iron oxide coating was found to be stable up to 3000C.

32

Iron oxide has a lower refractive index than cobalt oxide and hence the combination

increased absorptance. They synthesised a cobalt oxide-iron oxide (CoFeO) solar selective

coating on stainless steel using a spray pyrolysis technique and found that the coating had an

absorptance of α = 0.94 and emittance of ε100 = 0.20. The coatings have been found to be

stable for temperatures up to 400°C.

Avila et al. [107] synthesised cobalt oxide thin films upon stainless steel and nickel-

stainless steel alloy using a spray pyrolisis technique at temperatures of 350-600oC during a

five-hour period. Cobalt nitrate dissolved in water-ethanol was used as the precursor. The

absorptance value of α=0.77 and emittance value of ε=0.20 were achieved when the stainless

steel substrate was used. Interaction between the stainless steel substrate and the coating

material was also detected, as evidenced by the presence of an iron austenite phase. Greater

thickness and roughness of the Co3O4 film also contributed to better absorptance. This

phenomenon was consistent with the research results obtained by Drasovean et al. [110] for

wavelengths between 300-800 nm. However, the greater thickness of Co3O4 also had a

negative effect, increasing thermal emittance. Other efforts to improve selectivity were

focused on changing the other experimental conditions. It was found that the higher the

annealing temperature, the higher the film roughness would be [107].

Efforts to improve the quality of the cobalt oxide selective absorber surface by using

simpler deposition techniques such as dip-coating have also been made. Cathro [101]

outlined that spray pyrolysis should be avoided because there were mechanical difficulties in

controlling the accuracy of the film thickness [111]. Cathro prepared SSA surfaces based on

cobalt oxide, either stand-alone or in an admixture with nickel or manganese oxide using a

sol-gel dip-coating process. The mild steel substrate was immersed in the ethanolic cobalt

nitrate solution and withdrawn at 10 mm/s before being pyrolised in a muffle furnace at

500ºC for 15 minutes. The addition of nickel nitrate to the cobalt oxide solution precursor

33

increased both absorptance and emittance of the final film, while the addition of manganese

decreased emittance. The use of a deposit containing 5% nickel afforded a solar absorptance

of 0.90 with a thermal emittance ranging from ε = 0.1 at 80°C to ε = 0.25 at 300°C. In other

conditions, the addition of colloidal silica to the solution improved the optical parameters of

the film. These surfaces were stable for at least 1000 h at 500°C [101]. Barrera et al. [102]

report that black cobalt (Co3O4) thin film made by sol-gel dip coating onto a stainless steel

substrate showed α = 0.88 and ε = 0.12. Cobalt acetate was used as the precursor and it would

become a gel in a few hours. Cobalt acetate was obtained from precipitation of CoCl2

aqueous solution by ammonia, and then it was dissolved in acetic acid to form a cobalt

acetate solution precursor. During the dipping process, the relative humidity was maintained

at 40% in the preparation chamber and the dipping speed was 1 mm/s. The coating colours

depended on the thickness of the films. Film thickness of around 0.08-0.25 μm could be

obtained depending on the viscosity of the precursor. Less viscous sols (<2 cp) produced a

film thickness of 0.08 μm/dipping. For more viscous sols, film thickness increased to 0.25

μm/dipping. The durability test showed the coating had good stability at high temperatures of

450ºC for 48 h. However, the multiple repetitions of the coating using a fix speed rate of dip-

coating had a weakness, namely, it created many defects which affected the optical and

mechanical properties of the films [33]. As a comparison, the electrochemical method used to

synthesis a cobalt oxide selective absorber on various substrates gives α=0.92-0.96 and

ε=0.04-0.18 [112-114]. Overall, it can be concluded that the cobalt oxide selective absorber

produced by the sol-gel dip-coating method are quite comparable with the electrochemical-

based cobalt oxide selective absorber.

To avoid the degradation performance of absorber material, the protecting layer is

required to cover the bare cobalt oxide absorber layer. Barrera et al. [21] synthesised the

cobalt oxide in a silicon matrix forming amorphous cobalt-silicon oxide thin film on the

34

stainless steel substrate using a sol-gel dip coating route. Cobalt (II) acetate tetrahydrate and

tetraethylorthosilicate (TEOS), respectively, were dissolved into the acidified ethanol.

Concentrated HCl was also added drop-wise, then the solution was stirred for 24 hours at

room temperature. The Co(II) ion in the sol was stable based on the FTIR experimental

results. There was Co(II) chelation between OAc- ion from HOAc with Co(II) stabilizing the

Co and avoiding precipitation. This solution was also used for the dipping procedure. After

the dipping process, all samples were heat-treated at 400ºC in order for the gels to adhere to

the substrates. The absorptance value of the thin film is not high (α = 0.82) but it shows high

thermal stability, because after heating up to 5000C, it maintained practically the same

absorptance values. The role of the silica matrix was to protect the cobalt oxide from

performance degradation. The sol–gel process was an adequate technique for preparation of a

homogeneous thin film; in this case, cobalt was incorporated homogeneously into the silica

matrix. FTIR detected the Co–O–Si bonds in the film, which indicated that homogeneity

extended to the molecular scale [21]. Unfortunately, the silica matrix absorbed too much EM

radiation in the IR wavelengths (around 8-10 μm) producing an increase in the emittance and

a decrease in selectivity [71, 115].

Barrera et al. [40] also used tin oxide (SnO) as a protecting layer for black cobalt. Tin

oxide was chosen because of its low emissivity [115-117] and high chemical stability [118].

Black cobalt and tin oxide were deposited by the sol-gel dip-coating method onto the various

substrates. Cobalt-propionate solution was used as the cobalt oxide precursor, while a

peptized tin carbonate aqueous solution was used as the tin oxide precursor. It was found that

the use of glass and stainless steel substrates improved selectivity slightly, while the use of a

nickeled stainless steel substrate, even though it only gave a moderate absorptance value,

decreased emittance values significantly where α = 0.72 and ε (at 1000C) = 0.037 [40].

Besides Co3O4, the Co2O3 compound also existed in the films. Large amounts of carbon both

35

as graphite particles and carbon bonded to metallic and oxygen atoms were also detected in

several configurations. In the tin oxide protecting layer, SnO2 phases and carbon particles

were also detected. Carbon presence was caused by the relative low annealing temperature

(400ºC) [40].

Barrera et al. [119] also tried a different approach to obtain a durable SSA by mixing

cobalt and copper oxide precursors without adding a protecting layer. They prepared

polycrystalline cobalt and copper oxide composites (cobalt oxide – copper oxide) thin films

on stainless steel (SS) substrates using the spray pyrolysis method. This preparation was

simple and required low consumption of reagents. A mixture of cobalt and copper nitrate

with the molar ratio of Co:Cu (5:1) in ethanol:water (3:1) solvent was used as the precursor

solution. After 3 minutes spraying deposition, the samples were heated to 300-6000C for 3

hours. The films were stable up to 4000C and showed good absorptance (α = 0.84) but the

emittance was also relatively high (ε = 0.28) reducing the performance of the selective

absorber. A complex chemical structure consisting of Co3O4, CuO and metallic copper

phases, as well as voids was detected by X-ray diffraction and ellipsometry studies.

3.2.1.3 Ruthenium oxide

Morales-Ortiz et al. [120] found that a ruthenium oxide (RuO2) thin film on the top of

an ASTM grade 2 titanium substrate produced the characteristics of a SSA. Ruthenium

chloride in alcoholic solution was used as a precursor solution. The deposition was carried

out using the dipping and spraying technique at room temperature before the sample was

heat-treated at a temperature of 450-5000C for 1 h. In the case of dipping a polished substrate,

the absorptance was 0.74 while the emittance was 0.12. For spray deposition onto a non-

polished substrate, the film exhibited a very high solar absorptance (α = 0.98), and also a very

high infrared emittance (ε = 0.8). Therefore to improve performance, a thin gold film was

36

added to the surface of the ruthenium oxide by evaporation, giving an absorptance value of

0.91 and an emittance value of 0.16. The conclusion was that close control of the deposition

parameters and the substrate surface roughness would allow for further improvement in

selectivity and reproducibility.

3.2.1.4. Nickel oxide – alumina

A film consisting of Nickel oxide (NiO) in the pores of alumina (NiO-Al2O3) on an

aluminium substrate also showed the characteristics of a solar selective absorber (SSA)

material. Ienei et al. [8] prepared NiO films obtained by sol-gel spray pyrolysis deposition

(SPD) using an aqueous solution of nickel acetate tetrahydrate embedded in a porous

structure of Al/Al2O3. The precursor solution concentrations and compositions, substrate

temperatures and annealing treatments were optimized to produce the best SSA. The

structural and morphological properties of the resultant films were investigated by X-ray

diffraction, atomic force microscopy (AFM) and contact angle measurements. The results

showed that the coatings had excellent spectral selective properties with a normal solar

absorptance of 0.92 and a normal thermal emittance of 0.03. A low thermal emittance value

was obtained after using hydrophobic polymer additives (sodium maleat-methyl metacrylate

(HFB)) and annealing treatment. The thermal emittance and solar absorptance of the

deposited films were correlated to the chemical composition, crystalline structure and

morphology. In terms of layer-by-layer deposition (e.g. substrate - selective surface -

antireflection (AR) layer), a high surface energy (low contact angles) of the intermediate

layer (absorber layer) was recommended to allow the deposition of the next layer from

aqueous/polar precursors. The surface of the last deposited layer should have large contact

angles, and thus low surface energy. This would ensure a non-wettable behavior and

37

therefore a cleaner surface which would prevent condensation of any water vapor that might

enter into the collector onto the surface of the thin films [8].

Another sol-gel route to synthesis a NiO-Al2O3 SSA film was reported by Qian et al.

[121]. An aluminium isopropoxide and nickel nitrate solution was used and deposited onto a

stainless steel substrate by dip-coating. The results showed that a compact and homogeneous

film was obtained when the withdrawal speed was 1 mm/s, the NiO content in the sol was

20% and the thermal treatment temperature was 700℃. The addition of a silica anti-reflection

layer on top of the absorbing layer could enhance the performance of the absorber. The

optimum performance with an anti-reflection-coated sample could reach a solar absorptance

of 0.84 [121].

3.2.2. Metal and carbon particles in dielectric matrix

3.2.2.1. Metal particles embedded in dielectric matrix

Many researchers in the field of SSA synthesis have investigated cermet selective

absorbers using various synthesising methods. This is because the cermet structure is unique

and is also one of the highest performance selective surfaces [75]. However, to the best of our

knowledge, the synthesis of this type of absorber using sol-gel methods is relatively scarce.

Eisenhammer et al. [122] patented the idea of metal/conductive particles in alumina, with

either Al65Cu20Ru15 in alumina or TiN in alumina, as a SSA. Each composite was obtained by

mixing the conductive particles with an alumina matrix sol precursor. The alumina sol

precursor was prepared by dissolving niobium chloride (NbCl5) in butanol and mixing with

sodium butoxide (Na(OBu)n) under reflux conditions. This produced Nb(OBun)5 which was

subsequently mixed with glacial acetic acid to form the alumina sol precursor. Eisenhammer

et al. [122] also investigated another route to prepare the alumina sol precursor by mixing

boehmite with HNO3 at 550C. For the synthesis of quasicrystal Al65Cu20Ru15 conductive

38

particles in alumina film, the particles were mixed with the alumina sol precursor solution

which was then sprayed onto a copper substrate and heat-treated at 6000C. For synthesis of

TiN conductive particles in alumina film, the particles were dispersed into the alumina sol

precursor solution, then coated on the copper substrate by centrifugation (spin) and finally

heat-treated at 6000C. The Al65Cu20Ru15 alumina layer had a thickness of 110 nm and a

volume fraction of 30%, whereas the TiN-alumina layer had a thickness of 130 nm and a

volume fraction of 20% [122]. However, they did not show any absorptance and emittance

values, but from the curves created in their patent, these two SSAs can be categorized as

having comparable selectivity values.

Bostrom and co-researchers [22, 37, 39, 51, 64, 92, 123, 124] have synthesised nickel

nanoparticles embedded in an alumina ceramic matrix (Ni-Al2O3) thin film on a smooth and

highly specular aluminium substrate using a sol-gel-like method. They reported that although

the sol-gel methods have been known to fabricate a wide variety of materials for many

decades, it was only in the last few years that the solution-chemistry or sol-gel science was

found to be a suitable method to produce nanoparticle composites appropriate for thermal

solar absorber applications [20, 22]. Precursor solutions of nickel and pure amorphous Al2O3

in different proportions were mixed to control the nickel to alumina ratio in the final

absorbing films [22]. Film deposition was conducted via spin-coating at 3700 rev/min for 20

seconds before the film was heat-treated to temperatures of 550-580oC in an oxygen-free

glass tube. During the heat treatment, solvents were evaporated and the only substances left

in the final film coating were alumina and metallic nickel [22, 51]. The thin films produced

were smooth and homogeneous with nickel content of up to 80% of the volume fraction. The

optimal single layer of coating had a nickel content of 65 volume %, a thickness of 100 nm

and particle size between 5-10 nm. This absorbing layer showed a solar absorptance of α =

0.83 and thermal emittance of ε = 0.03. The addition of a pure alumina anti-reflection (AR)

39

layer on top of the absorbing layer enhanced the performance of the absorber. The optimum

anti-reflection-coated sample reached a solar absorptance of = 0.93 and a thermal emittance

of = 0.04. These results showed that the Ni-Al2O3 cermet film had excellent spectrally

selective optical properties. They suggested that for constructing a Ni-Al2O3 absorber layer

more efficiently, the bottom part of the layer should have high nickel content while at the top

it should have minimum nickel content [22]. The use of a rough aluminium surface as a

substrate was also implemented in this research, but the results were less satisfactory than the

smooth substrate.

Further investigations by Bostrom and co-researchers [92] focused on improving the

selectivity and durability of the nickel-alumina cermet and enhancing the performance of the

AR coatings. They reported that the performance of the nickel-alumina selective absorber

thin film system was improved if a three-layer system was applied. This system was

composed of an 80% nickel and 20% alumina film with thickness of 103 nm at the base (first

layer), a 40% nickel – 60% alumina film with the thickness of 59 nm in the middle (second

layer) and a silica/hybrid-silica film with the thickness of 90 nm at the top (third layer/AR

layer). This optimal three-layer system showed a solar absorptance value of 0.97 and a

thermal emittance value of 0.05 [37, 39, 92, 124]. These results were comparable to

commercial products. These synthesis processes were simple and cost-effective but the

nickel-alumina solution was unstable and agglomerated to form precipitates within 24 hours,

thus reducing the reproducibility of this system, even though the stability can be enhanced for

up to one week in a methanol solution [51]. The calcination step also required strictly

oxygen-free conditions, which was troublesome. This absorber has been industrially

produced on a pilot scale since 2009 and the company is working on having a full-scale

process in the near future.

40

Another effort to improve the nickel-alumina SSA coatings synthesised using a sol-

gel-like method was carried out by Nejati [32]. Nejati used nickel nitrate and alumina powder

as precursors. Nickel nitrate was first dissolved in distilled water or ethanol, then, while

stirring, alumina powder was gradually added. The prepared mixture was then dispersed

mechanically using a dissolver and ultrasonication. To avoid agglomeration, the temperature

was strictly controlled and different additives such as a wetting agent; a coupling agent and a

dispersing agent were added to the suspension before dispersion. Cleaned aluminium

substrates were then dip-coated in the suspension at different speeds. The wet films were

dried for 30 minutes at 120oC and then quickly annealed for 1 hour at 450

oC in a hydrogen

atmosphere. Nejati found that the mechanical properties of a pure Ni-Al2O3 cermet composite

layer and the substrate were poor and the layers were easily removed during the tape test.

Nejati did not use only TEOS as a source of silica for the AR layer but also used it to improve

the bonding ability between the absorber thin film and the substrate (the silica was also used

as an underlayer). Adhesion and scratch resistance of the thin film was improved

significantly. The silica network formed after the addition of TEOS also enhanced the solar

absorption by lowering the effective refractive index of the film. However, although the

addition of the silica AR layer increased the solar absorptance value, it also increased the

emittance value slightly. The best result was shown by a sample with absorptance value of

= 0.94 and emittance value of = 0.11 [32]. Based on accelerated ageing and humidity

studies, Nejati estimated that the nickel-alumina absorber was suited for glazed collector

applications such as domestic solar water heaters operating at low temperatures. Due to the

promising optical performance and good thermal and humidity stability, the developed

absorber film could compete with sputtered absorber films [32].

41

3.2.2.2 Carbon particles in dielectric matrix

3.2.2.2.1. Carbon particles in silica

Katumba et al. [41] outlined the reasons for studying carbon-in-silica tandem selective

solar absorbers. Firstly, both carbon and silica are abundant, environmentally-friendly and

stable materials. Secondly the sufficiently small size of carbon particles, approximately 10

nm or less, have a high absorption cross-section for UV-VIS radiation [41, 125]. Finally,

carbon-silica composites could be synthesised easily via sol-gel techniques.

Mastai et al. [126] introduced a new concept for the design of carbon-silica based

SSA materials. This group showed that porous carbon-silica hybrid nanocomposites have

SSA characteristics. The synthesis of this composite involves a sol-gel-like method to

perform a direct carbonization in the nanoconfinement of porous silica leading to the

formation of nano-sized amorphous carbon particles. Materials used included sugar as a

precursor of carbon, and cyclodextrins (CD) and polystyrene-polyethylene oxide (SE) as

precursors of CD-based silica and SE-based silica, respectively. In such a structure, solar

radiation was absorbed and transferred into heat without infrared (IR) re-emission. The

carbon nanoparticles contributed to high absorptance and thermal stability, whereas silica

contributed a transparent matrix and binder material. Especially in the case of CD-based

silica, the overall processes were ideal because of cheap and “green chemistry” conditions.

Also, sugar was easily available and non-toxic. This composite was obtained under one-pot

synthesis conditions with the elimination of water. No removal or addition of any further

chemical was necessary to obtain the non-toxic carbon-containing silica. In addition, leaching

of the final material was practically impossible and if that did happen it would only release

materials that were already abundant in nature [126].

The absorptance and emittance values for the SE-based carbon-silica composite were

α = 0.93 and ε = 0.08 respectively, while that for the CD-based silica-carbon composite the

42

absorptance value was 0.92 and emittance value was 0.13. All samples showed good stability

under the influence of humidity and high temperatures. Based on the nature of the

components involved in this composite, it could be assumed that long-term stability of the

samples was likely to be high. The degradation of solar thermal absorber coatings which is

usually caused by thermal oxidation of metal particles did not happen in this composite [126].

In separate but related research, Katzen et al. [127] created a carbon-silica

nanocomposite film selective absorber on a glass substrate. The film was prepared using the

sol-gel spin-coating method. The silica sol preparation was followed by CD-based silica-

carbon composite preparation. β-methylated cyclodextrin (2 g) was dissolved in 3 g of

aqueous HCl (pH 2) and 4 g tetramethylorthosilane (TMOS) were added while stirring until a

homogeneous solution was produced (within a few minutes). The films were deposited by

spin-coating at 4000 rpm for 1–2 min. The films were dried at room temperature and

annealed in an oven under nitrogen (95%) flow at a temperature of 850 K (increasing at 20

K/min intervals). All films prepared by this method contained approximately 15% carbon. It

was found that the best thin film silica-carbon nanocomposite (thickness 1000 nm) showed α

= 0.94 and ε = 0.15. The films showed good stability under the influence of humidity, as they

were held above a water bath at 1000C for 5 hours and in a high temperature environment

(250–300oC) for 48 h [127].

Another sol-gel method used to synthesis carbon-silica thin film composites for solar

selective absorbers was reported by Katumba et al. [41]. The processes consisted of using a

silica-carbon precursor sol, which was spin-coated onto a metal (specular and rough

aluminum and stainless steel) substrate and carbonizing it in an inert atmosphere. Samples

were made from silica sols based on acid-catalysis of TEOS and water that were impregnated

with sucrose (SUC) as the carbon precursor. Four categories of samples were studied. These

were the tetraethyl-orthosilicate only (TEOS-only), methyl trimethoxysilane (MTES), acetic

43

acid anhydride (Ac2O) and soot (SOOT) samples. In this case, MTES and Ac2O functioned as

organic modifiers of inorganic silica.

The spin-coating technique produced films with very flat surfaces and uniform

thicknesses in the 1 μm range. The fine structure showed homogeneous mixing of the carbon

and silica in the TEOS-only samples, while the addition of both MTES and Ac2O resulted in

the segregation of silica and carbon at the nano-scale. However, the addition of 20 wt %

MTES or 15 wt % Ac2O to the TEOS-only sols helped to reduce the cracks in the TEOS-only

samples. The samples with 20 wt% MTES had a solar absorptance of α = 0.74 and thermal

emittance of ε = 0.30 while the corresponding values for samples with 15 wt % Ac2O had α =

0.81 and ε = 0.44 [41]. The addition of soot did not yield a net advantage.

3.2.2.2.2. Carbon particles in ZnO, NiO and TiO2 matrices

Carbon nanoparticles dispersed in ZnO and NiO dielectric matrices on aluminium

substrates, to be used as SSAs, have been prepared by Katumba et al. [6]. The sol-gel-like

method used to prepare these samples was closely related to the method of Liu et al. [128].

Appropriate amounts of zinc acetate dihydrate and nickel acetate tetrahydrate were separately

dissolved in 50 ml of anhydrous ethanol and stirred by a magnetic stirrer at room

temperature. Diethanolamine (DEA) was added as a chelating agent in a way that the molar

ratio of each type of acetate to DEA was maintained at 1:1. These solutions formed the ZnO

and NiO precursors. Sucrose was dissolved in distilled water in the mass ratio 1:1 prior to

mixing with the matrix precursor solutions. This constituted the carbon precursor solution.

The oxide and carbon precursor solutions were mixed and stirred again. After a period of

stirring, 1 g of polyethylene glycol (PEG) was added to the ZnO and NiO matrix precursor

sols. The resultant solution was stirred further until the formation of a sol which was

immediately spin-coated onto pre-cleaned aluminium substrates. The PEG was used as a

44

structure-directing template. The spin-coated samples were then calcined in a tube furnace

with nitrogen-flow at 550oC for 1 h to carbonize the carbon precursor and also to dry and

solidify the oxide matrix. This method ensured an even distribution of the carbon

nanoparticles in the oxide matrices [6]. The absorptance and emittance values achieved were

α = 0.84 and ε = 0.04 for C-NiO and α = 0.71 and ε = 0.06 for C-ZnO. SEM analysis revealed

a smooth surface for both C–ZnO and C–NiO samples, but other C–NiO samples showed

dendritic characteristics. The coatings contained amorphous carbon embedded in

nanocrystalline ZnO or NiO matrices. Explorations with a Selected Area Electron Diffraction

(SAED) instrument showed that a small amount of Ni grains of 30 nm diameter also existed

in the NiO matrix. Both C-ZnO and C-NiO also had grain sizes for the carbon clusters in the

range 55–62 nm and a crystallite size of 6 nm as indicated by Raman spectroscopy [129]. The

accelerated ageing tests in a weather chamber with a high relative humidity environment of

95% and a temperature of 450C for 600 h showed that the C–NiO sample maintained better

performance than the C–ZnO sample [6].

Titanium dioxide (TiO2) has also been used as the host for carbon particles for SSA

applications. Rincon et al. [42] synthesised carbon blacks (CB) and carbon nanotubes (CNT)

embedded in a TiO2 matrix deposited on polished stainless-steel substrates. The use of CNT

could bring interesting optical properties to the composite because it is highly anisotropic.

These researchers used a chemical method based on sol–gel techniques. They first prepared a

TiO2 matrix where the large refractive index of TiO2 was decreased by increasing the

porosity of the oxide with the addition and subsequent thermal elimination of polyethylene

glycol (PEG). The CB-TiO2 composite films were prepared by dissolving CB in an alcohol

solution prior to the addition of titanium tetraisopropoxide and then adding a minor amount

of dispersing agent. Then the solution was stirred for 24 h at room temperature. Films were

fabricated on stainless-steel substrates by dip-coating. Different dipping speeds and a number

45

of dipping and drying cycles were combined with a final annealing temperature to control the

thickness and microstructure of the deposited film.

The CNT-TiO2 film was prepared using a layer-by-layer strategy whereby the first

layer contained the TiO2-PEG coating and the second layer contained the CNT-PEG film.

The mixture of PEG, CNT and dispersing agent was sonicated for 90 min to obtain an ink, 30

ml of 2-propanol was added and the emulsion was centrifuged for 30 min. The supernatant

liquid was discarded and the nanotubes were re-dissolved in water and centrifuged again. In

the third centrifugation, the nanotubes could no longer be precipitated and the ink was stable

for weeks. This ink was kept under vigorous stirring for 2 days prior to being coated on the

TiO2 substrates (first layer). Five to fifteen immersions/drying cycles produced TiO2 film

thicknesses in the range of 150–650 nm after annealing. Once the first layer was obtained, it

was rinsed in distilled water and then in a mild solution of NH4OH before being rinsed again

in distilled water. This sequence eliminated the excess PEG and HCl from the sol–gel bath.

The TiO2-coated stainless-steel substrates were laid on a plane tilted at 8º to the horizontal

where a fixed amount of the CNT emulsion was cast and let dry for 24 h. Further drying at

1000C for 24 h, followed by annealing in N2 at 400

0C for 15 min, eliminated any remnants of

PEG and additive [42]. The research group found that the sol-gel-like method investigated in

their work proved successful in producing coatings with good spectrally selective properties,

but they did not give any specific absorptance and emittance values. The coating system was

thermally stable and free of corrosion problems due to the hydrophobic nature of carbon.

Despite many advantages in the use of carbon absorbers in the matrices described

above, the optical performance of these absorber materials has not yet reached a satisfactory

level. As such, there are still many opportunities for further research to improve their optical

properties and durability before commercialization. In addition to that, the processes involved

46

are relatively long and cumbersome since an inert environment is required in the

carbonization process.

3.2.3. Solar selective absorber surfaces using spinels

3.2.3.1 CuFeMnOx and CuCoMnOx spinels

During the last decade, spinels deposited on highly reflective metal substrates have

attracted considerable interest due to their promising properties as SSAs for solar thermal

collectors. The term “spinel” refers to a group of minerals which crystallize in a cubic

(isometric) crystal structure. Kaluza et al. [71] have succeeded in synthesising CuFeMnO4

black film spinel SSAs using sol-gel dip-coating and heat-treatment at 500 C. Mn-acetate,

Cu- and Fe- chloride precursors were used in a molar ratio of 3:3:1, respectively. To protect

the spinel from corrosion, a 3-aminopropyltriethoxy silane (3-APTES) silica precursor was

added to the Cu, Mn and Fe sol precursors with molar ratio of (Mn-Cu-Fe):silica = 1: 1.

Analytical results showed that the films consisted of two layers: the lower was amorphous

SiO2 and the upper was a spinel having the composition of Cu1.4Mn1.6O4. The films exhibited

absorptance values of around = 0.6 and emittance values of = 0.29–0.39. The low

performance was caused by the difference in the film thickness between the spinel and the

silica layer where the absorbing spinel layer film (200 nm) was much thinner than the

amorphous SiO2 layer (800 nm). The large thickness of the SiO2 layer increased the thermal

emittance of the composite films due to the strong phonon absorption of the Si-O stretching

modes at 1100 cm-1

. The absorptance value could reach 0.93 when a base catalyst (NH3)aq

was added to the precursor in the solution preparation, but the thermal emittance value

became very high (ε = 0.62) due to the presence of large SiO2 spherical particles (400 – 420

nm) [71].

47

Efforts have made to improve the optical performance of the CuFeMnO4 black film

spinels. Kaluza et al. [20] reported that the emittance value could be decreased by substituting

silica with zirconium oxide (ZrO2), but the presence of the ZrO2 generated a brown hue color

which caused the absorptance value to drop. This research group subsequently modified the

synthesis route using Fe-, Cu-, Mn-acetate precursors. After undertaking thermal hydrolysis

steps, they succeeded in making CuFeMn-oxide spinels which did not contain thermally

emitting components such as SiO2 or ZrO2, consequently enhancing the spectral selectivity of

the coatings. However, these CuFeMnOx films exhibited a reddish-brown hue, which

originated from the segregated Fe2O3 phase formed during heat treatment at 500 C. As a

consequence, the films showed a lower solar absorptance. To address this problem, Fe was

substituted with Co and it was expected that even if a segregated Co-oxide phase was formed,

the color of the oxides would be black due to the allowed interband transitions of Co-oxide

[20].

To prepare CuCoMnOx spinels, Kaluza et al. [20] used an ethanolic sol based on Mn-

acetate and Co- and Cu- chloride precursors. The solution was stirred at 60C until a viscous

sol (40 ml) was obtained. A part of the dark greenish viscous sol (6.5 g) was then diluted in a

MeOH/H2O mixture (40 g/4.6 g) to obtain optimum viscosity for dip-coating deposition. The

viscosity of the sol solution was further adjusted by the addition of thickening agent

hydroxypropylcellulose (HPC), which also contributed to the stability of the solution. The

films were dip-coated onto aluminum substrates with a dipping speed of 10 cm/min. To

obtain coatings with different spectral selectivities, the film thickness was varied by changing

the concentration of the thickening agent and the number of dipping/annealing cycles. Films

deposited on aluminum were then annealed at 5000C for either 15 min or 1 h [20]. The best

CuCoMnOx film demonstrated an absorptance of = 0.9 and an emittance of = 0.05 when

10 wt% of HPC was added to the sol precursor. This result proved that the CuCoMnOx spinel

48

was a promising candidate for a solar absorber coating material. Thermogravimetric analysis

showed that the xerogels became crystalline at 3160C while X-ray diffraction analysis

revealed that the coatings and powders consisted of predominantly CuCoMnOx spinels.

Rutherford back scattering (RBS) and transmission electron microscopic (TEM) studies,

combined with energy dispersive X-ray spectroscopy (EDXS) measurements, confirmed that

Cu, Mn and Co were present in the films in stoichiometric ratios close to that in the initial

sols. In addition, CuCoMnOx spinels exhibited relatively weak phonon absorptions at 600

cm-1

, i.e. below the peak of black-body thermal radiation [20].

To enhance the absorptance value of CuCoMnOx spinels, Vince et al. [29] attempted

to modify the spinel. They made two different types of films, namely, Ti-doped (up to 30%)

CuCoMnOx and undoped CuCoMnOx films. The precursor ratio of Co:Cu:Mn was 1:3:3. The

films were subsequently annealed at 450oC for 15 and 30 min in air. To improve the stability

(weather and abrasion resistance) of the films, two kinds of protective over-coatings were

tested: one over-coating was based on polysiloxane resin and the other based on the high-

density of silica. Results indicated that undoped CuCoMnOx films with SiOx protective over-

coatings exhibited absorptance values of = 0.85-0.91 and an emittance value of < 0.036

after just a single dipping/annealing cycle. All investigated films exhibited poor stability

during a boiling-water test (>2 hours) before protective over-coatings were applied. When an

over-coating based on high-density silica or polysiloxane resin was applied to either doped or

undoped CuCoMnOx films, both of them remained unaffected by the test.

Another shorter and easier method which was used to synthesis CoCuMn-spinel solar

selective absorbers was reported by Bayon et al. [130]. Copper, cobalt and manganese

nitrates were dissolved in absolute ethanol at various molar ratios. A complexing agent and a

wetting additive were also added to stabilize the solution and improve the film adherence.

Depositions were performed using dip-coating at different withdrawal rates on aluminium

49

foil, borosilicate glass and stainless steel substrates. The resulting layers were sintered in an

oven at 500ºC. A silica AR layer was also deposited by a sol-gel method on top of the spinel

absorber. The highest solar absorptance of the resulting film (α = 0.863) was reached with

only one layer of absorber material when the spinel was deposited at 15 cm/min and the

molar ratio in solution was 1Cu:0.5Co:1Mn. The solar absorptance was improved to 0.908

when a SiO2 antireflective layer was deposited onto the spinel. Long term stability studies

showed that the CuCoMn-spinel was a very stable material. This study showed that the

metallic ratios in the film were very close to the precursor ratios in the dipping solution. XPS

measurements have shown that different oxidation states can be found for the metals present

in the spinel: Cu+, Cu

2+, Co

2+, Co

3+, Mn

2+ and Mn

4+ [130]. Although the CoCuMnOx

synthesised via this method is often contaminated by some metal oxides, chlorides, and

oxychlorides, it is better than the co-precipitation method. This is because in the co-

precipitation method, it is difficult to control all of the metal cations that precipitate from the

solution and which, at the same time, result in composition segregation and low yield [131].

3.2.3.2 CuMnOx spinels/CuMn oxide

A simpler CuMnOx spinel which contains less than three metal components and is

derived from the CuCoMn-spinel also shows the characteristics of a SSA. Bayon et al. [24]

reported that CuMn-spinel thin films on aluminium foil synthesised by a sol-gel-like dip-

coating method and followed by air-sintering at 5000C could be used as a low temperature

application of SSA. Copper and manganese nitrate precursors were dissolved into the

absolute ethanol solutions with the addition of a complexing agent and a wetting additive to

stabilize the solution and improve the film adherence. Analysis of the composition showed

that the metallic ratio in the film was very close to the ratio in the dip solution and indicated

the formation of a spinel-like material with Cu1.5Mn1.5O4 stoichiometry. The annealing time

50

and temperature also influenced the film composition and optical properties. A solid-state

redox reaction occurred when temperatures higher than 450oC were applied [78]. The highest

solar absorptance of = 0.87 is reached by using a one layer film deposited from solutions

containing a molar ratio Cu/Mn = 1 and prepared with a withdrawal rate of 20 cm/min. The

optical property of the film was dramatically improved by subsequently depositing a SiO2

anti-reflective layer using a sol-gel technique onto the spinel. By optimizing the film

thickness of both CuMn-spinel and SiO2 layers, the best absorptance and emittance (at

100oC) values achieved were 94% and 6%, respectively. These results showed that it was

possible to obtain a very good selective absorber with only two layers (absorber layer and

anti-reflective coating) from cheap materials and by using a simple dip coating deposition

method [24]. Although the optical performance of this spinel oxide solar absorber was quite

promising, it was still not high enough to be competitive in the market. The absorptance of

this absorber surface could be improved to 0.95 by introducing an additional CuMn-oxide

absorber layer (a total of 3 layers) [38]. Thermal stability and humidity tests were conducted

based on the method developed by the International Energy Agency (IEA) within the Solar

Heating and Cooling (SHC) Program Task X for low-temperature SSAs [88, 91]. The results

of a preliminary up-scaling study revealed that it was possible to deposit CuMn-oxide

absorbers on large-area substrates and that they could be a good alternative to the materials

present today in the market, not only in terms of optical properties but also in terms of long

term durability [38].

Overall, the general strategy to implement sol-gel methods for the synthesis of

absorber-reflector tandem structures (non organic binder) suitable for SSA materials is shown

in Fig. 3.1. The absorptance, emittance and selectivity of various SSAs produced by sol-gel

methods to date are summarized in Table 3.1.

51

Figure 3.1. General strategy for synthesising metal oxide/spinels (route A) and

metal/carbon particles embedded in non-organic matrix/binder (route B) solar

selective absorbers.

Table 3.1. Summary of absorptance (α) and emittance (ɛ) of various SSA materials produced

by sol-gel methods

Sol-gels SSA Materials and Substrates α ɛ Reference

Metal Oxide Based Absorber

Bare CuO on aluminium 0.93 0.11 (80oC) [95]

CuO-SiO2 on stainless steel 0.92 0.2 [93]

Black cobalt on galvanized iron 0.91 0.12 (100oC) [104]

Cobalt oxide on stainless steel 0.93 0.14 (100oC) [106]

CoFeO on stainless steel 0.94 0.2 (100oC) [108]

Cobalt oxide on stainless steel 0.77 0.2 [107]

Cobalt oxide-nickel oxide on mild steel 0.9 0.1 (80oC) [101]

Black cobalt on stainless steel 0.88 0.12 [102]

Black cobalt-tin oxide on nickeled stainless steel 0.72 0.037 (100oC) [40]

Cobalt oxide-copper oxide on stainless steel 0.84 0.28 [119]

Ruthenium oxide on the ASTM grade 2 titanium 0.74 0.12 [120]

Nickel oxide - alumina on aluminium 0.92 0.03 [8]

Cermet based absorber

Nickel-alumina cermet on aluminium 0.97 0.05 [39]

Carbon-silica on glass 0.94 0.15 [127]

Carbon-NiO on aluminium 0.84 0.04 [6]

Carbon-ZnO on aluminium 0.71 0.06 [6]

Spinels based absorber

CuCoMnOx on aluminium 0.9 0.05 [20]

CuCoMnOx-SiOx on aluminium 0.91 0.036 [29]

CuMn oxide – SiO2 on aluminium 0.95 0.06 (100oC) [38]

52

3.3. Effect of Silica Thickness

Various SSAs, whether synthesised by sol-gel or other methods, often involve the

incorporation of silica to improve their selectivity or durability. The deposition of a silica

layer, especially silica as an AR layer, usually necessitates a sol-gel technique even though

the absorber film was deposited by other methods than sol-gel. In this review, the use of silica

(SiO2) either as anti-reflection (AR) layer, matrix or underlayer has been mentioned.

However, the use of silica as a protecting agent (matrix or underlayer) of the absorber film

has had an unfavourable influence on the optical performance. High emittance values are the

consequence of the incorporation of silica as a matrix and/or an underlayer because the silica

absorbs too much solar radiation in IR range [20, 32, 71, 93], while silica as an AR layer has

a more positive effect because it can improve absorptance with a non-significant influence on

the increase of the surface emittance value [24, 29, 92]. Silica as an AR layer is frequently

synthesised thinner than the silica as a matrix or underlayer, so, in the construction of a SSA

protective layer (matrix or underlayer) involving silica, the protective layer thickness should

be an important factor to be optimized. The AR layer or other protective upper coatings

should normally be within 50-70 nm or in the scale of tens of nanometers [132].

53

Chapter Four

EXPERIMENTAL METHOD

Experimental works undertaken in this study can be divided into two main parts,

namely, the film coating preparations and their characterisations. Film coating preparations

consist of substrates preparation, sol-gel solution preparation and film coatings deposition;

while the structural, surface morphology and surface composition/electronic structure as well

as the optical and mechanical characterisations were carried out using a wide range of

complementary techniques, including X-ray diffraction (XRD), scanning electron microscopy

(SEM) and field emission scanning electron microscopy (FESEM), energy dispersive X-ray

(EDX), atomic force microscopy (AFM), X-ray photoelectron spectroscopy (XPS),

synchrotron source of XPS and near edge X-ray absorption fine structure (NEXAFS)

spectroscopy, UV-Vis-NIR spectrophotometry, Fourier transform infrared spectroscopy

(FTIR), as well as nanoindentation and modelling, and the accelerated thermal durability test.

Details on substrates preparation and film coatings deposition are discussed in Section 4.1. In

Section 4.2, the instrumentations used to characterise the film coatings are reviewed and the

characterisation techniques are presented.

4.1. Film Coatings Preparation

4.1.1. Substrates preparation

Substrates used in this study were highly reflective/specular commercial flat plate

aluminium (size 2 × 4 cm2) supplied by Anofol. The aluminium substrates were cleaned

using an etching solution to minimise the alumina layer. The etching solution was produced

by adding 0.35 M chromium (VI) oxide with 0.327 M of phosphoric acid (85%) in distilled

54

water. The aluminium substrates were placed in the hot etching solution (85 °С) for 10 min

and subsequently rinsed in hot distilled water. This was followed by a flush of distilled water

and drying in a nitrogen stream. The glass microscope slide substrates were also used, but for

light/solar absorbance studies only. The glass substrates were rinsed using distilled water then

ultrasonically cleaned in acetone for 15 min and dried in a vacuum oven at 60 °С to remove

residual moisture.

4.1.2. Materials, sol-gel solution preparation and film coatings deposition

The main film coating used for selective absorber coating was copper-cobalt oxide,

but film coatings such as manganese-cobalt and nickel-cobalt oxides were also prepared for

comparison. A silica antireflection (AR) layer was also applied to the film coating to

facilitate optimum absorptance performance. To prepare these absorber coatings and

antireflection layer, the materials used consisted of cobalt (II) chloride (CoCl2.6H2O, APS

Chemical, >99%), copper (II) acetate monohydrate (Cu(OOCCH3)2.H2O, Alfa Aesar, 98%),

anhydrous manganese (II) acetate (C4H6MnO4, Alfa Aesar, >98%), nickel (II) acetate

tetrahydrate (Ni(OOCCH3)2.4H2O, Alfa Aesar, 98%), propionate acid (C2H5COOH, Chem

Supply, 99%), distilled water, tetraethoxysilane ((TEOS), Alfa Aesar, >99%), and absolute

ethanol (Merck). These materials were used as received.

The copper cobalt oxide film coatings were prepared by mixing certain concentrations

of copper acetate and cobalt chloride in a range of 0.15 - 0.3 M at different copper to cobalt

ratios ([Cu]/[Co] = 0.5, 1 and 2) using absolute ethanol as solvent. Propionate acid was then

added to the solution as a complexing agent and stirred for 2 hours in a sealed glass container

at room temperature. The resulting black solution was then used for thin film deposition on

aluminum substrates using a dip-coater at withdrawal rate of 60-180 mm/min at relative

humidity less than 55%, so that a relatively smooth and even film could be produced. The

55

adhesion of mixed metal oxide film on the substrate was immediate (ca 10 s) during the

dipping process. The wet coating was subsequently heated/dried at 150°C for 10 minutes in a

vacuum dryer for 10 seconds on the hot plate. Copper-cobalt thin films with different

thicknesses were prepared by repeating the dip-drying cycle and varying the withdrawal rates

before final atmospheric annealing in an oven furnace at 500 oC for 1 hour to remove

volatiles. The heating rate was 50oC/min. If the final temperature was set too low, residual

organic groups would not be completely removed, resulting in poor optical coating quality.

The scheme in Figure 4.1 illustrates the copper cobalt oxide thin film coatings synthesis

process while Figure 4.2 shows the dip-coater instrument.

Figure 4.1. Flow chart for the synthesis of copper cobalt oxide thin film coatings.

Ethanol

absolute Cobalt acetate Copper acetate

Propionate acid

Stirred for two

hours

Thin film deposition on

aluminum substrates by

dip-coating at different

withdrawal rates

Humidity

control

Drying at 150oC for 10

seconds on the hot plate

Dipping-drying

cycles to obtain films

with different

thicknesses

Final annealing in furnace

at 500oC for 1 hour

56

An analogous procedure was also applied to prepare the manganese–cobalt and

nickel–cobalt mixed metal oxides thin film coatings either on aluminium or glass substrates.

The annealing temperature synthesis in the range of 500-650oC were applied to assess their

characteristic in a higher temperatures. Annealing at temperatures higher than 650oC was not

possible since it was limited by the melting point of the aluminium substrates.

Figure 4.2. Dip-coater (PTL-MM01, MTI Corporation) used in the present study.

The silica antireflection layer was synthesised using a sol-gel route relatively similar

to the approach used by Bostrom, et al. [92]. Firstly, TEOS was mixed with ethanol and

dilute HCL (0.06 wt %) was gradually added to the TEOS-ethanol solution. The molar ratios

of ethanol and water to the TEOS were 5 and 4, respectively. To ensure full hydrolysis, the

resulting mixture was stirred for 24 hours at room temperature in a closed container. The

obtained solution was used for the antireflection layer deposition by the dip-coater at a

57

withdrawal rate of 10-40 mm/min. The wet AR layer was then stored in a desiccator before

being annealed to 400oC for 30 minutes in the oven furnace, and then allowed to cool to room

temperature.

4.2. Instrumentations and Characterisation Techniques

It was crucial that complementary techniques were utilized in the present study to

overcome the limitations set by a single characterisation technique. The crystalline structure

and surface of film coatings were confirmed by XRD and complemented with the EDX and

XPS, while the surface morphology and topography were imaged by SEM, FESEM and

AFM. Surface chemistry compositions and electronic structure are also important because

they can influence the light absorption properties in the transition-metal compound in relation

to the filling factor of d-orbital. The surface chemistry composition and electronic structure

were detected via surface oxidation states/surface electronic structure using conventional and

synchrotron source XPS. Further interfacial studies were also performed using synchrotron

radiation NEXAFS spectroscopy. The optical properties in the solar and infrared wavelengths

range were characterised by UV-Vis-NIR and FTIR, while the mechanical properties were

investigated by nanoindentation technique, finite element modelling and accelerated thermal

durability test. A brief introduction to the instruments and the characterisation techniques

carried out will be discussed in the following sections.

4.2.1. X-ray diffraction (XRD)

XRD is a non-destructive analysis tool for investigating the bulk crystallographic

structure, chemical composition and physical properties of natural or synthetic materials

[133]. It works based on the interaction between the monochromatic x-ray radiation and a

periodic crystal lattice. When X-ray radiation with a specific incident angle (θ) interacts with

58

a crystalline specimen, constructive interference of the scattered radiation can occur leading

to the formation of intense peaks (Bragg peaks) when the Bragg condition is satisfied:

nλ = 2d sinθ (4.1.)

where n is an integer determined by the reflection order, λ is the wavelength of the incident

X-ray, d (d-spacing) is the interplanar distance, and θ is the X-ray incidence angle between

the X-ray beam and the sample position (Bragg angle).

Figure 4.3. The incident and scattered X-rays make an angle of θ symmetric to

the normal of crystal plane in XRD analysis (adapted from [133, 134]).

By using a fixed wavelength, the interplanar distance can be determined based on the

scattering angle of the diffracted X-rays. For phase identification of an unknown material, the

interplanar distance values obtained can then be compared to those recorded in the powder

diffraction universal database (International Centre for Diffraction Data, ICDD).

In the present study, the mineralogical characteristics of the thin film coatings were

analysed using GBC EMMA X-ray Diffractometer and Bruker Advance D8 X-

Ray Diffractometer. The radiation source was CuKα (Cu Kα = 1.5406 Å) operating at 40 kV

59

and 30 mA. The diffraction patterns were collected over a 2θ range from 10° to 80° with an

incremental step size of 0.01° and speed of 1o/min. Another diffractometer, Bruker

Advance D8 X-Ray Diffractometer (XRD) equipped with a Lynx-Eye detector, Cu-tube and

operated at 40kV and 40mA was also used. Conditions of analysis for the latter XRD were as

follows: 15 rpm rotation, 10-60o of 2θ, 0.01 degree increment, 1.2 sec/step-time per step, 1

hour and 37 minutes scan time, 0.26 degree fixed divergence slit and 2.20 degree fixed anti-

scatter slit.

4.2.2. Scanning electron microscopy (SEM), energy dispersive X-ray (EDX), and field

emission scanning electron microscopy (FESEM)

The scanning electron microscope (SEM) is a type of microscope that images a

sample surface microstructure and morphology by scanning it with a well-focused beam of

high energy electrons. The interaction between the intense electron beam generated from an

electron gun (primary electron beam) and the sample will produce a number of signals

including secondary electrons, back-scattered electrons and X-rays [135]. In SEM, the

secondary electrons with low energy and back-scattered electrons are the greatest interest

signals; both provide information about sample morphology. These signals are detected and

captured by a photomultiplier detector and used to reproduce the surface morphology image

of the scanned area.

The X-rays generated by the interaction of the electron beam with atoms in the

specimen are the characteristic of the elements present in the specimen. It is because the X-

ray released energy is only dependent on the atomic structure. Hence, every atom exhibits a

characteristic X-ray emission spectrum. The X-rays generated may be separated in an energy

spectrum to identify the elemental composition of the specimen. This method is called energy

dispersive X-ray spectroscopy (EDX). The fact that a spectrum of interest from 0.1 to 20 keV

60

can be acquired in a relatively short time (10~100 s) allows EDX for fast analysis of

elemental constituents in the sample.

Figure 4.4. Schematic diagram of SEM with a CRT display [136].

Field emission SEM (FESEM) is similar to SEM, except that it is equipped with a

field-emission cathode in the electron gun. The electrons liberated from this cathode are then

accelerated in a high electrical field gradient, and within a high vacuum column they are

focused and deflected by electronic lenses to produce a narrow scan beam that bombards the

sample. Consequently, the secondary electrons are emitted from the sample and the angle and

velocity of these electrons corresponds to the surface structure of the sample. FESEM

provides narrower probing beams at low and high electron energy, resulting in both improved

spatial resolution and minimized sample charging and damage compared to SEM [137].

In this study, the surface morphologies of the film samples were examined using a

PHILLIPS XL 20 SEM linked with an EDX analysis column and a Zeiss Neon 40EsB

FESEM with a maximum EHT (extra high tension) voltage field emission gun of 30 kV. For

61

FESEM, the sample was mounted on the substrate holder using carbon tape, and sputter-

coated with platinum to reduce charging effects before analysis. InLens and SE2 detectors

were used to obtain FESEM images at various magnifications (2 to 10 kV).

4.2.3. Atomic force microscopy (AFM)

Atomic force microscopy (AFM) is one of the most powerful tools for analysing and

imaging the surface topography and roughness of the material surface. It differs from optical

and electron microscopes which 'look' at the sample surface; the AFM works by scanning

probe microscope by 'feeling' the sample surface [138]. AFM operates based on the principle

of measuring the deflection of a sharp force-sensing tip which is attached to a flexible

cantilever with a specific spring constant as it probes the material’s surface. The sharp tip is

commonly made from silicon or silicon nitride [139]. The changes of cantilever deflection are

monitored by a four segment-photodiode detector. The computer processes the electrical

differential signal from the photodiodes and generates a feedback signal for a piezo-scanner

to maintain a constant force between the tip and the sample surface. The data obtained from

the cantilever moving vertically (z-direction) at each (x,y) point in the surface which is caused

by the changes in the surface contours are then processed by computer to form the

topographic image of surface [138, 139].

Surface topographic images of the thin films in this study were obtained using a

commercial atomic force microscope (AFM) (Ntegra Prima, NT-MDT Co., Moscow, Russia)

in semi-contact mode. The thin film samples were fixed on adhesive tape before AFM scans

were conducted. The probe used for the imaging contained a tetrahedral tip with height 14–16

μm and a typical curvature radius of 6 nm. The tip was mounted on a rectangular single

crystal silicon (N-type, antimony doped) cantilever with a thickness of 2 μm, a resonant

frequency of 140–390 kHz and a force constant of 3.1–37.6 N/m.

62

4.2.4. X-ray photoelectron spectroscopy (XPS)

XPS is a highly-sensitive surface technique based on the photoelectronic effect to

detect the chemistry composition and the electronic structure of the material surface. When

an atom in the surface is illuminated by a monoenergetic soft X-ray photon in an ultrahigh

vacuum chamber, an electron is ejected from an inner shell and this photoelectron has a

kinetic energy (Ek) equal to the following equation;

Ek = λυ – Eb – φ (4.2)

where λυ is the energy of the X-ray photon, Eb is the binding energy of the atomic orbital

from which the electron originates and φ is the work function, a value dependant on both

sample and spectrometer [140]. The ejected electrons are passed through the hemispherical

photoelectron energy analyser and selected at a given energy by electrostatic fields prior to

arriving at the detector.

An XPS spectrum is obtained by a plot of the electron counting rate versus their

binding energy. A peak at a particular energy would indicate the presence on a certain

element while the intensity of the peak corresponds to the concentration of the element in the

sample. Each element has a characteristic binding energy associated with each core atomic

orbital, yielding a set of discrete peaks in the photoelectron spectrum. The spectrum of a

mixture of elements may be considered as the sum of the peaks of the individual constituents.

Identification of chemical states can be obtained from an accurate estimation of the

separations and peak positions, as well as from certain spectral features. In the XPS, the

composition of the film can be determined by utilising the peak area and peak height

sensitivity factors. Fluctuations in the structural configuration or oxidation states of a

chemical element at the surface layers can be identified by examining the shifts in the binding

energies of a particular core level.

63

Figure 4.5. Schematic diagram of hemispherical

photoelectron energy analyser in XPS instrument [140].

XPS system performance is influenced by the power of source and its focusing ability.

Most common sources of photons are the MgKα and AlKα lines. The higher resolution of

XPS can be obtained from the synchrotron radiation source. The synchrotron radiation XPS

(SR-XPS) provides a continuous energy distribution over a large energy region with high

intensity and tuneability giving an optimal excitation energy instead of a fixed excitation

energy source (either AlKα or MgKα radiation from a sealed-off X-ray tube) used in

conventional XPS [141, 142]. The photon energies of the synchrotron source can be varied to

various escape depths of out-coming photon electrons and having a better photon ionization

cross-section [143]. Furthermore, the beam size of the synchrotron light source is much

smaller than the conventional photon sources, which assists to reduce the effects of the non-

uniformity of the sample surfaces [143].

64

The atomic percentages and surface bonding structures of samples in this study were

probed by Kratos Axis Ultra XPS spectrometer (Manchester, UK) with Mg Kα radiation (hν=

1253.6 eV). The samples were mounted, using double-sided Cu sticky tape, horizontally on

the holder and normal to the electrostatic lens. The vacuum pressure of the analyser chamber

was less than 10-9

Torr. The voltage and emission current of the X-ray source were held at 12

kV and 12 mA, respectively. Initial survey scans used pass energy of 80 eV. To ensure high

resolution and good sensitivity for the features of interests, pass energy of 10 eV was used.

The charging effects were corrected by using the C 1s of saturated carbon (C-C/C–H) peak as

reference for all samples at a binding energy (BE) of 284.8 eV. The electrostatic lens mode

and analyser entrance of the XPS instrument were selected using the Hybrid and Slot mode

(iris=0.6 and aperture=49), respectively. Charge neutralisation was employed during the XPS

measurements. The CASA XPS (V.2.3.15) software was utilised for quantification analysis

with Shirley background subtraction.

The analyses using the synchrotron source of XPS were conducted on the soft X-ray

beamline of the Australian Synchrotron under ring operation of 200 mA and 3 GeV. The

beamline was equipped with a collimated light plane grating monochromator SX700. The

1200 lines/mm grating and 15 μm entrance/exit slits were used. The samples were mounted

on a stainless steel sample holder and characterised under a background pressure 10-10

Torr in

the X-ray spectroscopy end-station. The Co 2p, Cu 2p and O 1s photoelectron lines were

measured in XPS mode using photon energy of 1253.6 eV. The XPS spectra energy scale was

calibrated using C 1s (284.8 eV; saturated carbon: C-C/C–H). The data were processed and

evaluated using SPECS (V2.75-R25274) and CasaXPS (V.2.3.15) softwares.

65

Figure 4.6. Kratos Axis Ultra XPS spectrometer (Manchester, UK) with

Mg Kα radiation source, Murdoch University.

4.2.5. Near edge X-ray absorption fine structure (NEXAFS) spectroscopy

NEXAFS is a powerful structural tool used to determine the electronic structure and

local co-ordination of the element of interest in the surface. When the X-ray beam from the

synchrotron radiation illuminates the sample, the electrons from the core will be ejected. The

resultant vacancies are then filled by the electrons from the higher energy level, resulting in

the ejection of Auger electrons (Auger decay). Auger electrons scatter secondary electrons

that escape from the sample. By measuring the intensity of Auger electrons as a function of

photon energy, sharp resonance features can occur which are associated with transitions from

core occupied states into the lowest unoccupied molecular states. These resonances appear at

energies near the absorption edge encompassing energy from the ionization edge with about

50 eV towards higher energies. When the sample is connected to earth, a drain current is

generated and established as total electron yield (TEY). By using an electron energy analyser,

66

an Auger electron yield (AEY) detection mode is obtained, where only elastically scattered

Auger electrons are recorded. The AEY mode is the best surface sensitivity among the

detection techniques in NEXAFS [144].

In this study, NEXAFS experiments were conducted at the soft X-ray beam-line of the

Australian Synchrotron facility in Melbourne under storage ring operation of 200 mA and 3

GeV. The beamline was equipped with a collimated light plane grating monochromator

SX700 with the 1200 lines/mm grating and 15 μm entrance/exit slits were used to

monochromatise the beam coming out of the storage ring. The samples were mounted on a

stainless steel sample holder using adhesive carbon tapes in order to avoid the surface

Figure 4.7. Samples transfer and analysis chamber in soft X-ray analysis end

station, Australian Synchrotron.

charging and characterised under a background pressure 1x10-10

Torr in the end-station X-ray

spectroscopy analysis chamber. The photon energy used was 1253.6 eV. The copper, cobalt

and oxygen X-ray NEXAFS absorption were measured in Auger Electron Yield (AEY) mode

67

by monitoring drain current and with a channeltron facing sample positioned 30° above the

incoming beam. To avoid erroneous interpretation of the results, the spectra obtained in AEY

mode were normalized by dividing the signal with the photon flux incident (I0). A gold mesh

was used to monitor photon flux incident (I0) on the sample. The measurements were

performed at room temperature and all samples were well-grounded and mounted on

adhesive carbon tapes to avoid the surface charging. The samples were characterised at a step

size of 0.1 eV over the energy region 920-980 eV, 770-820 eV, and 520-570 eV for Cu L-

edge, Co L-edge and O K-edge, respectively. The data were processed using SPECS (V2.75-

R25274) and CasaXPS (V.2.3.15) softwares.

4.2.6. Optical characterisations via UV-Vis-NIR and FTIR reflectance spectra

Optical performance of solar absorber coating on the aluminium substrates (opaque

surfaces) is calculated based on the absorptance (α) and emittance (ɛ) values. These values

can be obtained from the measurements of monochromatic reflectance in wavelengths area of

0.3µm – 2.5µm by using an UV-Vis-NIR and in wavelengths area of area more than 2.5µm

by using FTIR instruments [44].

The elemental parts of a UV-Vis-NIR spectrophotometer consist of light source,

sample holder, diffraction grating in monochromator or a prism to separate the different

wavelengths of light, and detector. The radiation source of a Tungsten filament (300-2500

nm) is often used. A photodiode detector and photomultiplier tubes are used with scanning

monochromators, which filter the light so that only light of a single wavelength reaches the

detector at one time. The scanning monochromator moves the diffraction grating to "step-

through" each wavelength so that its intensity may be measured as a function of wavelength.

The reflectance response by an opaque surface can be divided into specular and

diffuse reflectance. Specular reflectance (Rs) refers to a mirror-like reflection where the

68

incident polar angle is equal to the reflected polar angle, while a diffuse reflectance (Rd)

eliminates all directional characteristics of the incident radiation by distributing the radiation

uniformly in all directions. In practice, a highly polished surface approaches a specular

reflectance whereas a rough surface reflects diffusely. Most samples produce a combination

of specular and diffuse reflectance. It is possible to take separate measurements for specular

reflectance, diffuse reflectance or overall/total reflectance. An integrating sphere is needed to

measure overall reflectance which is the combination of Rs and Rd (Figure 4.8). The interior

wall in the integrating spheres is coated with diffusing and highly reflecting materials. In

solar wavelength range measurements the barium-sulphate (BaSO4) is usually used for

integrated sphere coated wall while in the IR wavelengths range the gold coated wall is

usually used [62, 145].

Figure 4.8. Specular reflectance (Rs) and diffuse reflectance (Rd) in a

reflectance mode of integrated sphere (adapted from [62, 76, 145]).

In this study, solar absorptance and emittance values of sample coating were

determined based on the reflectance data as described by Duffie and Beckman [44]. A

template of reflectance data against spectral distribution (air mass AM1.5) in equal energy

69

increment was created using an Excel worksheet to calculate the absorptance and emittance

values. The near normal hemispherical solar reflectance spectrum was recorded from 300 to

2700 nm using a UV–Vis-NIR Jasco V-670 double beam spectrophotometer with 60 mm

integrating sphere. Deuterium (300 to 350 nm) and Halogen (330 to 2700 nm) lamps were

used as the light source.

FTIR instrument basically consists of infrared source, interferometer, sample

compartments and detector. The IR energy is usually emitted from a glowing black-body

source. The beam passes through an aperture which controls the amount of energy presented.

The beam then enters the interferometer where the spectral encoding occurs. The resulting

interferogram signal beam then exits from interferometer to enter the sample compartment

where the beam is “transmitted through” or “reflected off” the sample surface (depending on

the FTIR type used). Finally, the beam passes to the detector for final measurement. The

measured signal is then digitised and sent to the computer where Fourier transformation

occurs.

In this study, the infrared reflectance spectra in wavelengths area of 2.5 to 15.4 µm

were obtained using a “reflected off” type of Perkin Elmer Spectrum 100 FTIR spectrometer

in a range of 4000 to 650 cm-1

. The coating surface was placed on the crystal surface area and

a pressure arm was positioned and locked at force of 80 N to maintain the coating surface

touching evenly onto the diamond surface. The reflectance spectrum was obtained after 4

times scans with resolution of 2 cm-1

. Background correction was performed before the

collection of each spectrum.

4.2.7. Mechanical characterisations: nanoindentation tests and finite element modelling

(FEM)

Nanoindentation test

70

One of the most popular applications of nanoindentation is to evaluate the mechanical

properties of thin film, either in lateral or depth dimension, without removing it from the

substrate. This instrument typically measures the depth of penetration using either an

inductance or capacitance displacement sensor. The crucial component in a nanoindentation

instrument is an indenter. The indenter should have high precision in geometry to facilitate

identification of contact area. Diamond is a commonly used indenter and it may be sharp,

spherical or flat-ended cylindrical. Sharp indenters including Berkovich, Vickers and cube

corner are preferred for nano- and microscale measurements especially in the characterisation

of the mechanical properties of thin films or small volume of sample such as hardness, elastic

modulus, resistance to deformation, etc.

In this study, a nanoindentation workstation (Ultra-Micro Indentation System 2000,

CSIRO, Sydney, Australia) equipped with a Berkovich indenter of 5 µm in radius was used to

determine the mechanical properties of the films according to the method proposed by Oliver

and Pharr [86, 87]. The procedure can be briefly described as follows: The area function of

the indenter tips was calibrated using a standard fused silica specimen. Nanoindentation was

conducted under load control with a maximum load of 0.5 mN. The indenter is pressed into

the sample surface under load-control, and the load and displacement are monitored during

the full loading-unloading contact cycle. For each test, 10 incremental and 10 decremental

steps were used. The maximum penetration depth during the tests was found to be <10% of

the film thickness, which ensured that only the film properties were measured. Twenty

indentations were performed for each sample.

Finite element modelling (FEM)

Finite element modelling (FEM) was used to visualize the stress distribution within

the coating and the substrate under spherical-tip indentation to assess the mechanical

71

response of the coating system to external loading. Mechanical properties of the coating

system, obtained from the nanoindentation tests, were used as input parameters. The

simulations were performed using COMSOL Multiphysics® Ver. 3.5a software. A two-

dimensional (2D) axisymmetric model was constructed with the loading direction along the

axial z axis. The details of the model set-up are briefly described here. The model consists of

a coating (1 µm thick) placed on top of aluminium substrate (49 µm thick), loaded under a

spherical tipped indenter with a radius of 5 µm. The simulation block is a rectangle

measuring 50 × 50 µm. Time-dependent deformation behaviours, such as creep, as well as

surface roughness and contamination were not considered in our simulations. The contact

between the indenter and the sample is assumed to be frictionless. The coating is assumed to

be bonded perfectly to the substrate. The boundary conditions are described below. The

bottom (z = 50 µm) is fixed in the z direction, while the right edge of the block (x = 50 µm) is

fixed in the x direction. The axisymmetric axis is placed on the left edge of the simulation

block (x = 0) to generate 3D effects. The tip of the indenter was located at z = 0 µm at the

beginning of the simulation. The indentation loading process is simulated as downward

movements in successive steps of 0.01 µm each, from 0 to 0.12 µm.

4.2.8. The accelerated thermal durability test

The accelerated thermal durability test in this study was carried out based on the PC

(performance criterion) value of IEA SHC Task 27 where the PC value is defined as

PC=−Δα+0.5Δε100 [90] as elucidated in Section 2.7, while the adhesion between the film

absorber and the substrate was evaluated by physical/cracking inspection before and after the

thermal test. The accelerated thermal durability test procedure is detailed below and

flowcharted in Figure 4.9. In the determination of PC value, the initial absorptance (α) and

emittance (ɛ) values of coatings were measured before the thermal test, and they became the

72

initial basis to determine the temperature (T1) applied in the thermal test (Appendix 1, [90]).

The PC value was evaluated after t =18, 36, 75, 150, 300 and 600 hours in an oven furnace.

Introduce t1 which is the last testing time where PC value is still less than 0.05 (Appendix 2,

[90]). Based on the PC value and the physical/cracking inspection, if:

Figure 4.9. Flow chart of accelerated thermal durability test in this study

i. PC value was similar to or less than 0.01 after t1=600 hour and without cracking, then

the absorber coating passed the accelerated thermal test, or

Measure α and ɛ, and determine T1 applied in

the test (Appendix 1)

Perform test at T1 and measure α and ɛ after the testing times 18, 36,

75, 150, 300, and 600 (Appendix 2). Introduce the time t1, which is

the last testing time where PC<0.05, then

If PC>0.05 at t1≤150h then

perform the additional test

using a fresh/new sample

in a lower temperature test

(T2) for t2 hours which

corresponds to the

previously determined t1

value to get the α and ɛ

values

If PC>0.05 at t1=300h

or PC>0.01 after

t1=600h then perform

the additional test using

a fresh/new sample in a

higher temperature test

(T3) for t3 hours which

corresponds to the

previously determined

t1 value to get the α and

ɛ values

If PC≤0.01 after

t1=600h and

without cracking,

then the absorber

passed the

thermal test

Determine the PC value,

if PC(T2,t2) ≤ PC(T1,t1) and the

coatings were without cracking,

then the absorber passed the

thermal test

Determine the PC value,

if PC(T3,t3) ≥ PC(T1,t1) and the

coatings were without cracking then

the absorber passed the thermal test

73

ii. if PC value was greater than 0.05 at t1≤150 hours, an additional test would be needed

using a lower temperature test (T2) for t2 hours (where t2 corresponded to the previously

determined t1 (see Appendix 2)). If PC (T2,t2) ≤ PC (T1, t1) and the coatings were

without cracking, then the absorber passed the accelerated thermal test.

iii. if PC value was more than 0.05 at t1=300 hour or PC value was higher than 0.01 after

t1=600 hour, an additional thermal test at a higher temperature (T3) for t3 hours would

be required (where t3 corresponded to the previously determined t1 (see Appendix 2)).

After this additional test, if PC(T3,t3)≥PC(T1,t1) and the coatings were without cracking,

then the absorber coating passed the accelerated thermal test.

74

Chapter Five

CHARACTERISATIONS OF COBALT-BASED METAL

OXIDE THIN FILMS SYNTHESISED USING SOL-GEL DIP-

COATING METHOD: AN EXPLORATION STUDY

5.1. Introduction

Cobalt-based metal oxides (MxCoyOz with M=Mn, Cu, Ni and derivatives) are highly

versatile functional materials, which have found widespread utilization in a variety of high-

tech applications. These include applications in catalytic processes [146-149],

electrochemistry [150-152], batteries and memory devices [153-155], solid oxide fuel cells

[156, 157] and electronics [158]. Manganese–cobalt oxides (MnxCoyOz) are a group of

widely-investigated metal oxides with particular emphasis on the influence of the synthesis

conditions on the oxidation states and cation distribution in the cubic and tetragonal phases as

typically described in the Mn–Co–O system [159]. The physicochemical properties of

manganese–cobalt oxide are greatly affected by the synthesis temperature, crystal structure,

anion oxidation states and composition [160]. Electrical properties of manganese–cobalt

oxide have also been investigated by Bordeneuve and co-researchers [161]. Likewise, there

have been numerous studies conducted to characterise the physicochemical, magnetic,

conductivity, electrochemical and thermal properties of copper–cobalt oxides [146, 162-165]

as well as nickel–cobalt oxides [152, 166, 167]. Nickel–cobalt oxides have been reported to

possess improved electronic conductivity properties, i.e., at least two orders of magnitude

higher than those of nickel or cobalt oxides [158].

An area of application in which these cobalt-based oxides are comparatively less-

studied is optical or solar-based coating, whereby optical performance of a surface can be

75

manipulated by depositing thin films with varying thicknesses and reflective indices. For

example, Kaluza and co-researchers [20] synthesised CuCoMnOx spinel coatings using a dip-

coating method which showed good potential as solar absorber coatings. In a more recent

study, Bayon et al. [24] reported deposition of CuMn-spinel layers on aluminum substrate as

solar selective absorbers using a similar sol–gel method. Incidentally, there are certainly

many knowledge gaps that need to be filled in terms of fundamental surface characteristics of

these thin films, especially in regard to their morphologies, binding states of metal oxides and

mechanical strengths. A technical understanding of these characteristics is an essential

component in the smart design and engineering context of thin film coatings for optical

applications.

In this chapter, manganese-, copper- and nickel–cobalt oxides thin film coatings on

commercial highly reflective aluminium substrates are synthesised using the sol–gel dip-

coating method. The sol–gel process is a soft chemistry method whereby the precursors are

often in a form of colloidal solution that ultimately ‘transforms’ into an extensive network of

either discrete or continuously-linked molecules. The sol-gel is selected due to its inherently

simple and safe characteristics in which solid-state synthesis could be accomplished at

relatively low temperatures [162, 168]. In this Chapter, the relationships between the surface,

optical and mechanical characteristics of the synthesised cobalt-based oxides film coatings

are analysed by employing X-ray diffraction (XRD), scanning electron microscopy (SEM),

atomic force microscopy (AFM), X-ray photoelectron spectroscopy (XPS),

spectrophotometry and nanoindentation techniques. Light absorption analysis for coatings on

glass substrate within a wavelength range of 300–1100 nm is also conducted. The

comprehensive and novel surface-based data with respect to thorough XPS and

nanoindentation analyses conducted in this study are significant as they afford a novel

understanding of the morphological and binding states/conditions of the aforesaid cobalt

76

based oxides which complement the existing knowledge on other mixed metal oxides. The

motivation behind this study is clear; detailed findings on surface analyses can be utilized to

aid the surface engineering designs of tuneable thin film metal oxides for a myriad of

industrial applications such as optical coatings and solar-selective absorbers.

5.2. Samples Preparation and Characterisation

The manganese-, copper-, and nickel-cobalt oxide samples here were prepared with a

similar procedure as mentioned in section 4.1. Typically, the concentration of each

manganese, copper, nickel and cobalt precursors was 0.15 M, and the withdrawal rate applied

was 60 mm/min. The characterisations carried out were structural analysis using XRD,

surface morphology analysis using AFM, SEM and FESEM, surface electronic structure and

composition analysis using XPS, optical properties analysis using UV-Vis and UV-Vis-NIR,

and mechanical properties analysis using nanoindentation. The elaborations about these

instruments and the characterisation techniques can be found in section 4.2.


5.3.1. XRD analysis

Figure 5.1 shows the XRD patterns of the prepared manganese–cobalt (i), copper–

cobalt (ii) and nickel–cobalt (iii) thin film coatings (6 dip-heating cycles) on aluminium

substrate and stand-alone heated and unheated aluminium substrate (iv and v), respectively.

The XRD patterns of the samples made with a 2 dip-heating cycle indicate marginal

differences from the aluminium substrate patterns (10° to 40° range) and as such only

samples from the 6 dip-heating cycle are considered for analysis.

The XRD pattern of the heated stand-alone substrate (Figure 5.1 (iv)) shows peaks at

approximately 45° (95%), 65° (100%) and 78° (19%)) where the percentages in brackets are

with respect to the observed main intensity peak. The XRD pattern of the unheated

77

Figure 5.1. XRD patterns of the prepared manganese–cobalt (i), copper–

cobalt (ii) and nickel–cobalt (iii) thin film coatings (6 dip-heating cycles) on

aluminum substrate and standalone heated and unheated aluminum substrate

(iv and v) respectively.

Figure 5.2. Expanded XRD pattern region from 10 to 40° (intensity of

observed peaks are 0.3%–0.5% of maximum intensity peak of substrate from

panel in Fig. 5.1).

78

stand-alone substrate (Figure 5.1 (v)) has two main peaks at approximately 65° and 78° and a

very small intensity peak (0.4%) at approximately 45°. Peaks from the MxCoyOz (M=Cu, Mn,

Ni) films are very hard to resolve in the 2θ-ranges observed for the substrates. Some peaks

with intensities of approximately 0.3%–0.5% of the main intensity peak of the substrates

shown in Fig. 5.2 (i)–(iii), when compared with the observed Al substrate plots of Fig. 5.2

(iv) and (v), are interpreted as due to cobalt-based metal oxides, MxCoyOz (M=Cu, Mn, Ni),

phases after comparing with ICDD database. The XPS and the EDX analyses in this chapter

or in the next chapters support this interpretation, indicating the presence of oxygen, cobalt

and metal (Cu, Ni or Mn) atoms on the surface forming the cobalt-based metal mixed oxides.

It is difficult to make conclusions on the stoichiometric formulation of the metal–oxide

phases present or on their crystallinity but it should be noted that poor crystallinity of

MxCoyOz (with M=Mn, Cu, Ni) synthesised by the sol–gel method has been reported by other

researchers [162, 169, 170].

5.3.2. Surface topography and morphology

AFM images indicate that the manganese–cobalt and nickel–cobalt coatings are

smoother than copper–cobalt coatings (Figure 5.3). The peaks and valleys (exhibited in the

form of contours) provide a quantitative indication of the surface roughness and absorptancy

of the coatings. In regard to surface roughness, the arithmetic average height deviation (Sa)

values for manganese-cobalt, nickel–cobalt and copper–cobalt are 17.61, 8.42 and 20.63 nm,

respectively. SEM and FESEM analyses corroborate this observation whereby the

morphology of a copper–cobalt coating shows the presence of nano-sized grain-like particles

(Figure 5.4). With close examination of the copper-cobalt coating micrograph, the

morphology of the grain-like particles is more obvious, with sizes ranging from 20 to 100

nm which are embedded within pores/trenches (Figure 5.5). These grain-like particles were

79

Figure 5.3. AFM images of the a) manganese–cobalt; b) copper–cobalt; and c)

nickel–cobalt thin film coatings (6 dip-heating cycles).

80

Figure 5.4. SEM micrographs of the a) manganese–cobalt; b) copper–cobalt;

and c) nickel–cobalt thin film coatings. FESEM micrograph for copper–cobalt

indicates the presence of nano-sized grain-like particles (6 dip-heating cycles).

also reported by Marsan et al. [171] and La Rosa-Toro et al. [172] for their porous copper–

cobalt oxide CuxCo3-xO4 layers. The former research group postulated that the porous/rough

morphology of the copper–cobalt oxide surface was attributed to the higher evolution of gas

volumes (NO2, O2) during the decomposition of concentrated nitrate coating. Concurrently,

the porous/rough surface of fabricated copper–cobalt oxide coating is attributed to the

evolution of O2 from high temperature decomposition of copper and/or cobalt oxides which

ultimately form a CuxCoyOz system [150, 173]. The thickness of film coating can be

approximated from the peak-to-peak (Sy) parameter from the AFM analysis result. Based on

peak-to-peak values, the manganese–cobalt, copper–cobalt and nickel–cobalt thin films

coatings are estimated to have the thickness per dipping of around 104, 221 and 172 nm,

81

respectively. The observation that copper–cobalt coating has the highest thickness is

consistent with the presence of sharp peaks as shown in the AFM image.

Figure 5.5. FESEM micrograph images (in magnifications of 200 nm and

100 nm) for copper–cobalt indicate the presence of nano-sized grain-like

particles.

5.3.3. XPS analysis

Figure 5.6 shows the wide-scan XPS spectra of cobalt-based metal oxides thin film

coatings. The wide XPS spectra confirm the existence of corresponding elements (Co, Cu,

Mn, and O) in relevant sample coatings as well as carbon.

Figure 5.7 shows the C1s and O1s XPS spectra of manganese–cobalt, copper–cobalt

and nickel–cobalt thin film coatings. The binding energies (BE) of each component are

evaluated by using a Gaussian–Lorentzian fit. The decoupling of the C1s spectra from all

82

three samples shows four carbon bonding states, i.e., (1) C-H and C-C (284.7–284.9 eV); (2)

C-O (286.4–286.6 eV); (3) O=C-O (288.2-288.5 eV); and (4) C–metal (i.e., metal carbide)

(281.7–283.9 eV). The C–metal bond of copper–cobalt coating consists only of cobalt

carbide (C-Co) bonding. This is understandable as carbon is neither miscible nor reactive to

the copper [174]. The manganese carbide binding energy is very rarely found in the

Figure 5.6. Wide scan of XPS spectra of cobalt-based metal oxide film coatings

literature. However, the binding energy of transition metals in carbides is normally in the

region of around 281–284 eV, and it can shift by 0.5–0.7 eV depending on the chemical

environment of the transition metal [175]. Ioffe et al. [175] proposed the binding energy of

manganese carbide (C–Mn) to be 282.5 eV. Our C1s XPS spectrum shows that a very strong

component located at around 281.7 (Figure 5.7. a1) is most likely assigned as manganese

carbide due to the high ratio of Mn/Co in the manganese–cobalt coating as seen in Table 5.1.

The decoupling of the O1s photoelectron spectrum shows three components, namely,

(1) lattice O2-

(O–metal bonds) (529.4–529.8 eV) [146, 162, 176]; (2) surface oxygen,

83

including adsorbed oxygen species, as hydroxyl (OH), carbonate group, etc. (531.1–531.2

eV) [146, 162]; and (3) subsurface O- species (531.9–532.2 eV) [177, 178]. For the

manganese–cobalt coating, the two components representing Co–O and Mn–O binding

energy are separated by 0.4 eV (Figure 5.7. b1) while for the other two coatings, the Co–O

and Cu–O / Ni–O binding energy peaks overlap. The manganese-cobalt coating does not have

subsurface (bulk structure near the surface) oxygen O-.

Figure 5.7. C1s and O1s XPS spectra of manganese–cobalt, copper–cobalt and nickel–cobalt

thin film coatings. Dashed lines correspond to fit envelopes, while wavy lines correspond to

data curves.

84

The manganese-cobalt coating does not have subsurface (bulk structure near the

surface) oxygen O-. The subsurface oxygen ions have lower electron density than the O2

-

ions. They can be associated with sites where the coordination number of oxygen ions is

smaller than in a regular site, with higher metal oxide bonds [177].

Table 5.1. Metal composition analysis of film coatings using XPS.

Coatings Atomic concentration (%) Atomic ratio

Manganese-cobalt Mn=14.93% Co=2.23% Mn:Co = 6.7:1

Copper-cobalt Cu =5.09% Co=10.62% Cu:Co = 0.5:1

Nickel-cobalt Ni =9.4% Co=7.84% Ni:Co = 1.2:1

Figure 5.8 shows the Mn2p, Cu2p, Ni2p, and Co2p XPS spectra of manganese–

cobalt, copper–cobalt and nickel–cobalt thin film coatings. Generally, each of the Mn 2p, Cu

2p, Ni 2p and Co 2p XPS spectra (Figure 5.8. a1–a3 and Figure 5.8. b1–b3) has two main

peaks representing 2p3/2 and 2p1/2. Based on their binding energy, these main peaks indicate

that the oxide states of Mn, Cu, Ni or Co are present in the surface coating [179]. Comparison

of the Mn 2p spectrum with the Co 2p spectrum (Figure 5.8. a1–b1) suggests that the

manganese–cobalt surface coating has much higher manganese concentration than cobalt,

while the copper–cobalt and nickel–cobalt surface coating have a comparable concentration

between copper or nickel and cobalt, respectively (Figure 5.8. a2–b2 and Figure 5.8. a3–b3).

Table 5.1 shows the metal compositions in the surface coatings. The observed higher

concentration of Mn in the manganese–cobalt system can be explained, based on the nature

of the surface cobalt itself where cobalt tends to leave more surface cation positions empty

than manganese, resulting in much lower cobalt concentration on the sample surface [180].

85

Figure 5.8. Mn2p, Cu2p, Ni2p, and Co2p XPS spectra of manganese–cobalt, copper–cobalt

and nickel–cobalt thin film coatings.

Gautier et al. [146] argued that the low intensity satellites of the Co2p spectra in the

copper–cobalt surface system indicated that the Co ions were present in a spinel-type lattice

arrangement, while the binding energy of Co 2p3/2 and 2p1/2 corresponded to the octahedral

diamagnetic Co(III) ions existing in a low-spin state. They further argued that the non-

satellite peaks in Cu2p spectra which had a binding energy difference of around 20 eV could

be attributed to Cu(II) ions whereas the strong satellites indicated Cu2+

as in CuO. If cobalt

were present as diamagnetic CoIII

ions and copper were present as Cu2+

ions, then the oxide

could be represented by the Cu2+

Co2III

O4 formula [146]. All of these characteristics

mentioned by Gautier et al. [146] match the results seen in Figure 5.8. a2–b2. A further

discussion of CuxCoyOz system will be given in Section 7.3.2.

86

The peak of Ni 2p3/2 in the nickel–cobalt coating (Figure 5.8. a3) could represent two

oxidation states: firstly, the Ni(II) ions present on the surface as NiO, and secondly the Ni(II)

ions or Ni(III) species present on the surface as Ni(OH)2, while the satellite at around 862 eV

indicates the presence of paramagnetic Ni ions [181]. Regarding the Co–2p curve in the

nickel–cobalt sample (Figure 5.8. b3), the low satellite band structure points to a surface

depletion of paramagnetic Co(II) ions [181].

5.3.4. Optical properties

The optical properties of the thin film coatings are evaluated on the basis of

absorptance (α). Absorptance is defined as a weighted fraction between absorbed radiation

and incoming radiation. The absorptance of a thin film on a substrate can be determined in

terms of reflectance (R) as described by Duffie and Beckman [44]. Low spectral reflectance

indicates high absorptance and vice versa. To the best of our knowledge, only a few previous

studies on the optical properties of cobalt-based metal oxides (MxCoyOz withM=Mn, Cu, Ni)

thin film coatings have been carried out [146, 167, 182, 183].

Before analysing the absorptance, the absorbance spectra for all thin film coatings are

needed to understand their intrinsic characteristic. Absorbance is a quantitative measure

expressed as a logarithmic ratio between the radiation falling upon a material (I0) and the

radiation transmitted through a material (It). It is influenced by the length of penetration and

the concentration of bulk film. The absorbance spectra for all thin film coatings on glass

substrates within a wavelength range of 300–1100 nm are shown in Figure 5.9. It is obvious

that all coatings show higher absorbance of ultraviolet (UV) light compared to visible light

with gradual increases in absorbance from the infrared (>740 nm) to the UV range (<400

nm). It is also clear that the nickel–cobalt thin film has the lowest absorptive capability

among all the films, albeit the increase of dip-heating cycles increases their absorptive

87

capability significantly. This implies that the thickness of the nickel–cobalt layer plays a

crucial role in UV and visible light absorption properties, i.e., this effect is even more

pronounced in nickel–cobalt compared to manganese–cobalt and copper–cobalt.

Figure 5.9. Absorbance spectra of thin film coatings on the glass substrates.

Absorbance due to glass substrate was eliminated from the spectra.

The absorbance spectra of manganese–cobalt film (Figure 5.9. a, b) show a broad

absorption band at around 570–750 nm with maximum at λ≈700 nm. We postulate that this

intrinsic band corresponds to a metal–metal charge transfer band between cobalt ions which

indicates the formation of segregated Co3O4 [146] in the manganese–cobalt spinel surface.

The copper–cobalt film spectra (Figure 5.8. c, d) seem to indicate that practically no band is

detected. This is attributed to the doping of Cu2+

ions into the lattice which causes

replacement at octahedral and tetrahedral Co sites forming the copper cobalt oxide spinel

structure, thus removing the orbital degeneracy and adding new orbital energy levels [146].

88

This distinguish the copper cobalt oxide from the Co3O4 [146]. This postulate can also be

extended for no-band detection in nickel–cobalt film (Figure 5.9. e, f) in terms of the

presence of Ni2+

ions.

Reflectance spectra of all thin film coating samples on aluminium substrates, together

with the corresponding solar absorptance values, are shown in Figure 5.10. Aluminium is

selected as the substrate because it is highly reflective and inexpensive. The prepared

coatings generally have low reflectance (<50%) of UV light, moderate reflectance (<80%) of

visible light and high reflectance (up to 100%) of infrared light. The copper–cobalt (with 6

dip-heating cycles) sample represents an anomaly amongst all the thin films because of its

rather low reflectance value which contributes to its comparatively high absorptance value of

71.3%. The inspections of figures 5.9 and 5.10 reveal that the choice of substrate has a

substantial influence on the absorptive property of the copper–cobalt film. It can be construed

that the reflectance property of a copper–cobalt layer is affected by both its thickness and the

reflectivity properties of the substrate surface. The copper–cobalt thin film on an aluminium

substrate experiences an increase of 29.5% in absorptance with a three-fold increase in dip-

heating cycles.

The discrepancies in the absorptance values of the samples can be explained by close

examination of the morphology and roughness of the deposited layers. Rincon and co-

researchers [42] argued that a rough surface reduced reflection of incident radiation at the

film surface, while surface pores contributed to the lower refractive index. As such, this

boosts the absorptance due to the interaction and relaxation mechanisms in the absorber as

well as multiple reflections and resonant scattering in the pores [42]. The observation that the

copper–cobalt film is rougher and more porous than the other samples in our SEM, FESEM

and AFM analyses corroborates our absorptance and reflectance results.

89

Figure 5.10. Reflectance spectra of thin film coatings on aluminum substrates with

corresponding solar absorptance (α) values.

5.3.5. Nanoindentation

The values of elastic modulus (E) and hardness (H) of thin films compared with those

of stand-alone commercial aluminium substrate are presented in Figure 5.11, while their load-

displacement curves are shown in Figure 5.12. There are marginal differences in terms of E

and H between the three thin films albeit all films exhibit significantly lower E (by 44–50%)

and hardness (by 68–83%) compared to the aluminium substrate. Among the three thin films

tested, nickel–cobalt film exhibits the highest average elastic modulus while the other two

have similar values of elastic modulus (i.e. similar stiffness). In addition, the nickel–cobalt

thin film sample is the hardest among the three films. There is an observed difference in

average hardness (ca 30 MPa) between nickel–cobalt and copper–cobalt films. The

representative load–displacement curves in Figure 5.12 show the responses of the three thin

90

films to increasing mechanical loadings, and indicate a trend that reflects the E and H results.

The level of resistance to deformation increases via the following sequence: manganese–

cobalt, copper–cobalt and nickel–cobalt.

Figure 5.11. Elastic modulus and hardness of the thin films measured using the

nanoindentation.

As far as we are aware, no mechanical properties have been measured on these types

of films, though such properties are important to their functions and durability. Nonetheless,

it is deemed appropriate that previous reported mechanical characteristics (values) for other

coatings be noted as references/benchmarks for our nanoindentation study. For example, the

SiO2/TiO2/ORMOSIL composite for optical coating synthesised by a similar sol–gel method

(annealing at 500 °C) has a hardness higher than the present film coatings by ca 2 GPa [184],

though the latter have higher E values (by approximately 40 GPa). Chan et al. [185]

investigated the thermal, optical and mechanical properties of sol–gel-derived silica-based

91

coatings on polyester substrates incorporated with organic and transition metal

oxides components.

Figure 5.12. Typical load–displacement curves of the thin films measured using the

nanoindentation.

Nanoindentation analysis revealed that their coatings have a surface hardness up to 2.5 GPa

and E values up to 13.6 GPa, which is approximately an order of magnitude higher than that

of the plastic surface. They also reported that the addition of transition metal oxides led to a

coating with reduced hardness due to the low density structure resulting from the rapid

condensation reactions of catalytic effect of transition metal oxides [185].

5.4. Conclusions

Cobalt-based metal oxide thin films have been successfully deposited on

commercially available highly reflective aluminium substrates as thin film coatings in cobalt-

based metal oxides mineralogical forms using the sol–gel dip-coating method. The distinctive

92

morphological (nano-sized, grain-like particles) and optical features of the copper–cobalt

coatings imply good prospects for future application as a solar absorber coating material,

though further engineering is needed to improve optimal performance. All three coatings

exhibit higher absorbance of UV light compared to visible light with gradual increases in

absorbance from the infrared (>740 nm) to the UV range (<400 nm) with an intrinsic band in

the absorbance spectrum of manganese–cobalt. In terms of reflectance, the films generally

have low reflectance (<50%) of UV light, moderate reflectance (<80%) of visible light and

high reflectance (up to 100%) of infrared light. Our findings can be used to aid the

engineering design of highly tuneable thin film metal oxides for numerous industrial

applications, such as optical coatings and solar-selective absorbers. Our obtained

nanoindentation data infer that the mechanical properties of the films are generally favourable

and can be applied in industrial conditions. Lastly, comprehensive surface data such as XPS

descriptions reported here for manganese- and nickel–cobalt coatings may be useful as the

basis for engineering design of thin film coatings for other important industrial applications.

93

Chapter Six

SOLAR ABSORPTANCE OF COPPER–COBALT OXIDE

THIN FILM COATINGS: OPTIMIZATION, STRUCTURAL

AND SURFACE COMPOSITIONS

6.1. Introduction

Cobalt–copper oxides (CuxCoyOz) have attracted attention from many researchers

worldwide for applications such as the catalysis in oxygen evolutions reaction (EOR),

Fischer–Tropsch process, synthesis of syngas-based alcohol, and thermoelectricity material

[146, 162, 171, 186-194]. Many studies have been conducted to characterise the various

properties of copper–cobalt oxides. De Koninck et al. [162] studied the physicochemical and

electrochemical properties of CuxCo3-xO4 powder as applied for EOR. They found that

CuCo2O4 particles had smaller crystalline structures (with crystallite size 10 times smaller)

than Co3O4. Furthermore, the CuCo2O4 composite electrode that contained the largest amount

of oxide particles had a high intrinsic electrocatalytic activity for the EOR. Volkova et al.

[193] have used CuCoO2 as a precursor for Cu–Co alloy selective catalyst in the higher

alcohol synthesis from syngas. They analysed the peculiarities of formation and destruction

of Cu–Co alloy and Co2C to understand their roles in higher alcohol synthesis. The role of

Cu–Co alloy consisted of formation of cobalt carbide which was able to activate CO

undissociatively that led to oxygenate synthesis. Beekman et al. [195] had prepared the

delafossite type of CuCoO2 by ion exchange (metathesis) solid-state reaction between CuCl

and LiCoO2. The analyisis of electrical transport and magnetic susceptibility data for CuCoO2

showed that the transport and magnetic susceptibility data for polycrystalline CuCoO2 were

consistent with formal charge assignments of Cu+ and Co

3+ for the transition metal

94

constituents, and corroborated recent density functional theory calculations. Singh [194]

studied the electronic and thermoelectric properties of CuCoO2 by density functional

calculations. Application of the Boltzmann transport theory to the calculated band structure

shows high thermopowers comparable to NaxCoO2, an established material for thermoelectric

power generation application for both p- and n-type doping [194].

Therefore, in order to harness the beneficial properties of such metal oxides, we

conducted deposition of new cobalt-based metal oxide thin films (MxCoyOz with M = Mn,

Cu, Ni) on commercial aluminum substrates using the sol–gel dip-coating method as seen in

Chapter 5 of this thesis. The CuxCoyOz thin film coatings exhibited favorable optical

properties, albeit it was clear that further optimization study would be required to facilitate

commercialization of these coatings. To the best of our knowledge, solar-based optical

properties and optimization aspects of copper–cobalt oxides thin film coating are

comparatively less studied [146] and as such, these features form the basis for the present

study.

The objective of this chapter is to optimize the solar absorptance of cobalt copper

oxides thin films via the dip-coating method. The parameters studied are concentrations of

copper/cobalt and dip-speed whereby these are directly correlated to the thickness of the thin

films which ultimately influences their solar absorptance. The bulk and surface film coatings

compositions/electronic structures are also characterised using energy dispersive X-ray

spectroscopy (EDX) and synchrotron radiation X-ray photoelectron spectroscopy (SR-XPS),

respectively.


The copper-cobalt oxide coating samples on aluminium substrates were prepared with

a similar procedure as that mentioned in section 4.1 where the concentrations of each of the

95

copper-acetate and cobalt-chloride precursors were 0.15 M, 0.2 M, 0.25 M and 0.3 M

([Cu]/[Co] ratios were 1). Other copper to cobalt concentration ratios of 0.5 and 2 were also

made for an additional absorptance optimization study.

The characterisations carried out were bulk composition analysis using EDX, surface

chemistry composition analysis using synchrotron radiation SR-XPS and optical properties

characterisation using UV-Vis-NIR. To obtain the thickness of film coating, a SEM

micrograph picture was taken using a PHILLIPS XL 20 scanning electron microscopy

(SEM). In SEM analysis, before measurement, a glass sample was mounted on the substrate

holder using carbon tape then sputter-coated with gold using Balzers Union sputter coater to

reduce charging effects before analysis. SE detectors were used to obtain SEM images at

magnifications 10240x. The elaborations about these instruments and the characterisation

techniques can be found in section 4.2.


6.3.1. EDX analysis

The bulk compositions of copper cobalt film coating and the aluminium substrate

identified using EDX can be seen in Figure 6.1. Aluminium from the substrate also appeared

in EDX spectra of coating as a major component, and is much higher than the copper and

cobalt components as seen in Figure 6.1a. This shows that the copper cobalt thin film

thickness on the top of aluminium substrate is much less than 1 μm where the thicknesses of

around 1 μm are the around maximum thicknesses which the bulk compositions can be

probed by the EDX instrument [172].

By eliminating the aluminium component in the calculations, the normalization result

shows that the coating consists of ~3.35% of Cu, ~4.05% of Co, and ~92.6% of oxygen. It

96

Figure 6.1. EDX spectra of cobalt copper thin film coating on the top of aluminium

substrate synthesised using 0.15 M of copper-acetate and 0.15 M of cobalt-chloride

precursors (a), and aluminium substrate without coating (b).

has been pointed out by many researchers that the oxygen excess is inherent in the family of

cobalt mixed oxide [196, 197] and could also be related to the structural defects involved in

the conduction mechanism [172]. The presence of a subsurface of oxygen as identified in the

XPS result in Chapter Five could be one example showing the excessive amount of oxygen in

the bulk near the surface. The copper to cobalt ratio in the EDX result is relatively similar to

the copper to cobalt precursors ratio in the synthesis process indicating that the copper

precursor and the cobalt precursor in the synthesis process were mixed well. Based on this

97

fact, the resulted compound would have an approximately similar in stoichiometric value of

copper and cobalt.

6.3.2. Solar absorptance properties

The optical properties of copper-cobalt thin film coatings are evaluated on the basis of

absorptance. The solar absorptance (α) values of the synthesised copper–cobalt thin film

coatings were determined using the reflectance (R) data in the wavelength range of 300-2700

nm as described by Duffie and Beckman [44]. Low spectral reflectance essentially implies

high absorptance and vice versa. An important point is that the thickness of the absorber

coating layer influences the final absorptance of the system [22, 78, 93]. In the case of dip-

coating, film thickness can be easily controlled and optimized by altering the dip-

drying/heating cycles or the withdrawal rate [43]. This being the case, the dip-heating process

was conducted at several pre-fixed cycles. Figure 6.2 shows the reflectance spectra of

copper–cobalt oxide thin film coatings on aluminium substrates with equimolar copper and

cobalt concentrations of 0.15 M, 0.2 M, 0.25 M and 0.3 M, respectively, with dip-speed of

60-180 mm/min. Four dip-heating cycles were selected since they basically afford an

optimized reflective system compared to other numbers of cycles (for reason of brevity,

results for other cycles is not shown).

Interestingly, a wavy curve with peak (interference peak) and valley (absorption edge)

was detected at the shorter wavelengths range of spectra. A similar phenomenon was also

reported by other researchers [22, 24]. Generally, the interference peak and the absorption

edge shift towards longer wavelengths when increasing the dip-speed and concentrations. On

the other hand, the amplitude of the wavy curves oscillation tends to decrease with an

increase in the concentrations of metal ions. The absorptance values are higher for faster dip-

98

Figure 6.2. Reflectance spectra of copper–cobalt oxide thin film coatings on aluminium

substrates. Concentrations of reactants: (a) 0.15 M copper and 0.15 M cobalt; (b) 0.2 M

copper and 0.2 M cobalt; (c) 0.25 M copper and 0.25 M cobalt; (d) 0.3 M copper and

0.3 M cobalt. Four dip-heating cycles were carried out.

speeds or higher concentrations. Nonetheless, in order to obtain optimized spectra with

selective solar absorption character, the position of reflectance value in the cut-off area

(wavelength at around 2500 nm) has to be relatively high, i.e. at least 50% of the reflectance

99

value. The film coating synthesised using concentrations of 0.25 M copper acetate and 0.25

M cobalt chloride (Cu/Co ratio = 1) with dip-speed 120 mm/min (Figure 6.2c, α = 83.4%)

can be construed as the optimum coating design in this study.

In addition, we have examined the relative effect of concentration ratios on the

reflectance of copper–cobalt oxide thin film coatings. The Cu/Co concentration ratios

investigated are 0.5 (0.125 M copper and 0.25 M cobalt), 1 (0.25 M copper and 0.25 M

cobalt) and 2 (0.25 copper and 0.125 M cobalt), respectively. The dip-speed is 120 mm/min

with four dip-heating cycles. This method shows that the positions of the interference peaks

and the absorption edges shift towards the longer wavelengths as the Cu/Co ratio is decreased

(Figure 6.3). The absorption edge of the film coated with Cu/Co ratio = 0.5 reaches a

reasonably long wavelength at ca 1500 nm with lower interference peak and absorption edge

than the spectrum for Cu/Co ratio = 1. These render the absorptance value of the film coating

with Cu/Co = 0.5 the highest among the three spectra, at about 86.77%. However, even

though its absorptance is the highest among the coatings, its reflectance position in the cut-off

area is relatively lower than the coating with Cu/Co ratio = 1, so it is not attractive enough to

be a solar selective absorber. The curve profile of film coating synthesised with Cu/Co ratio =

0.5 can be considered as optimum since it exhibits the higher reflectance position in the cut-

off area.

The absorptance value of around 83.4% - 86.77% appears to be comparatively

promising for the application of selective absorber systems (i.e. before the addition of an anti-

reflective layer). By comparison, other reported coatings synthesised using the complicated

sol–gel method exhibited absorptance values of α = 83% [22], 80–85% [20] and 80% [40].

We surmise that the nano-sized grain-like particles surrounded by pores-trenches in our

samples are capable of providing a conducive surface morphology for absorption of

incidental solar radiation due to the multiple reflections that can occur inside the

100

pore/aggregate [8]. As such, this boosts the absorptance due to the interaction and relaxation

mechanism in the absorber as well as multiple reflections and resonant scattering in the pore

[8, 42]. At the same time, the highly reflective aluminium substrate functions to reflect back

the lower energy radiation (infrared) that penetrates the film coating.

Figure 6.3. Effect of Cu/Co concentration ratios on the reflectance of

copper–cobalt oxide thin film coatings. These include Cu/Co

concentration ratios of 0.5 (0.125 M copper and 0.25 M cobalt), 1

(0.25 M copper and 0.25 M cobalt) and 2 (0.25 copper and 0.125 M

cobalt), respectively. The dip-speed is 120 mm/min with four dip-

heating cycles.

To measure the thickness of film coating that is showing an optimum absorptance

value (α = 83.4%), the coating is fabricated on the glass substrate using a similar parameter

when it is fabricated on aluminium substrate. Figure 6.4 shows the copper cobalt oxide film

coating synthesised using a similar parameter with the coating that is showing the optimum

101

absorptance value. It shows the average of film thickness is 320 nm. This result is consistent

with our prediction earlier (Section 6.3.1) which indicates that the film thicknesses are less

than 1 µm. Some similar research could indicate that the film thicknesses were around 100-

200 nm [20, 24, 71].

Figure 6.4. SEM micrograph picture of copper cobalt oxide thin film coating

synthesised using concentrations of 0.25 M copper acetate and 0.25 M cobalt

chloride (Cu/Co ratio = 1) with dip-speed 120 mm/min and four dip-heating cycles

on the glass substrate.

6.3.3. Synchrotron radiation XPS study of elevated concentrations

A complementary high resolution synchrotron radiation X-ray photoelectron

spectroscopy (SR-XPS) study affords detailed information on the compositions/electronic

structure in the surface of the thin films. The synchrotron radiation source of XPS provides a

continuous energy distribution over a large energy region with high intensity and tune-ability

102

giving an optimal excitation energy instead of a fixed excitation energy source (either Al Kα

or Mg Kα) radiation from a sealed-off X-ray tube) used in conventional XPS [141, 142]. The

photon energies of the synchrotron source can be varied to various escape depths of out-

coming photon electrons having a better photon ionization cross-section [143]. Furthermore,

the beam size of the synchrotron light source is much smaller than the conventional photon

sources, which helps to reduce the effects of the non-uniformity of the sample surfaces [143].

Figure 6.5 shows the O 1s SR-XPS spectra of copper–cobalt oxide film coating synthesised

using various concentrations. The O 1s spectra exhibit a strong peak with a shoulder at a

higher binding energy. The decoupling of the three O 1s spectra generally gives four curve-

fitting components in the spectrum from each sample. The peaks (labelled “i”) at binding

energy (BE) around 529.4–529.6 eV could be attributed to lattice O2-

(Cu–O, Co–O) [146,

162, 176], the peaks (denoted “ii”) at BE around 530.5–530.8 eV may be treated as the

surface oxygen from a wide variation of species such as adsorbed oxygen O- and/or OH-like

species, as hydroxyl, and carbonate groups [146, 162, 198-201], while the adjacent peaks

(labelled “iii” and “iv”) at BE around = 531.4–531.5 eV and at BE around 531.8–532.0 eV,

respectively, could be assigned as subsurface O- species [177, 178]. A relatively flat peak

shoulder of O 1s peak (Figure 6.5c) is due to the high percentage of surface oxygen

and the low percentage of subsurface oxygen in the coating surface

synthesised using a higher concentration of copper and cobalt precursors.

Figure 6.6 shows the Cu 2p spectra and the peak-fitting of their Cu 2p3/2 of copper–

cobalt oxide film coatings synthesised using various concentrations. The two main peaks as

Cu 2p3/2 and Cu 2p1/2 peaks and the satellites on the high energy side of each Cu 2p3/2

and Cu 2p1/2, respectively, are found in every spectrum (Figure 6.6a). The binding energy

difference between Cu 2p1/2 and Cu 2p3/2 peaks in every sample, which is around 19.8 eV,

103

indicates the presence of a low oxidation state of copper, while the satellite peak between Cu

2p3/2 and Cu 2p1/2 confirms the presence of Cu2+

ions.

Figure 6.5. O 1s SR-XPS spectra of copper–cobalt thin film

coatings synthesised using concentrations of: (a) 0.15 M copper

and 0.15 M cobalt, (b) 0.2 M copper and 0.2 M cobalt, and (c)

0.25 M copper and 0.25 M cobalt.

104

The decoupling of Cu 2p3/2 and its satellite in every coating gives five curve-fitting

components (Figure 6.6b–d) and the quantification analysis is presented in Table 6.1. It is

commonly recognized that the Cu 2p3/2 photoelectron peak around 933.3–934.0 eV with

shake-up satellite is due to the CuO (or Cu2+

), while many researchers identified that the Cu

2p3/2 photoelectron peak at around of 932.5–932.8 eV is from the tetrahedral Cu+

with its

counterpart peak from octahedral Cu+ located below the tetrahedral one [172, 202-210]. From

Table 6.1, the octahedral and tetrahedral Cu+ as well as the octahedral and paramagnetic Cu

2+

oxidation states are detected with the tetrahedral Cu+ being the most prominent. These results

are relatively different from the copper acetate precursor used, which has the Cu2+

oxidation

state only. It is widely known that the increase in temperature changes the characteristic of

oxidation states of a surface. In relatively low temperatures (under 150oC), the copper oxide

obtained from the alcohothermal process of copper acetate precursor undergoes a binding

energy shifting of Cu2+

oxidation state from A-sites (tetrahedral coordination) towards B-sites

(octahedral coordination) as temperatures are increased [207, 211]. However, in

this temperatures range, no reduction of Cu2+

occurred [211]. In a copper–cobalt oxides

environment, the reduction of Cu2+

to Cu+ is detected at a temperature of 350

oC [172, 181].

The tetrahedral Cu+

species could be from the direct reduction of Cu2+

at tetrahedral sites

[172]. In regard to the copper–cobalt oxide surface samples, the presence of oxidation states

of copper with different co-ordinations is originally estimated from the evolution of Cu2+

A-

sites due to decomposition/deligandation during the high-temperature (500oC) calcination. In

addition, the low intensities of paramagnetic satellites of cupric oxide (Figure 6.6a) indicate

that a part of octahedral Cu2+

undergoes a further reduction, forming the octahedral Cu+ [207,

212].

A broad and asymmetric of line shapes core-level main peaks profile (Figure 6.7a)

and the presence of cuprous (Cu+) in the copper–cobalt oxide environment have been widely

105

Figure 6.6. (a) Cu 2p SR-XPS spectra of copper–cobalt thin film coatings synthesised

using various concentrations, (b)–(d) decoupling of Cu 2p3/2 of copper–cobalt thin film

coatings synthesised using various concentrations.

recognized as typical of monophasic Cu–Co mixed oxides [172]. In Cu–Co mixed oxides, the

Cu2+

ions incorporate into the surface octahedral vacancy, then could share oxygen with

106

adjacent Co2+

ions which the Cu2+

ions are filling interstitial sites within the structure of the

cobalt oxide forming surface Cu–O–Co species [172, 213].

Figure 6.7 shows the Co 2p spectra and the decoupling of their Co 2p3/2 of copper–

cobalt oxide film coatings synthesised using various concentrations. The two main peaks

assigned as Co 2p3/2 and Co 2p1/2 peaks and a low intensity satellite between these two main

peaks were found in every spectrum (Figure 6.7a). Qualitatively, the Co 2p3/2 peak and Co

2p1/2 peak separated by a spin–orbit splitting of around 15.1 eV indicates the presence

of the mixed Co(II) and Co(III), while the low intensity satellite of the Co 2p spectra in an

area of around 789 (Figure 6.7a) on the copper–cobalt oxide surface system indicates that the

cobalt ions are present in a partial spinel-type lattice arrangement. The observed asymmetry

in the Co 2p1/2 peak confirms the existence of Co(II) and Co(III) ions.

Table 6.1. The binding energies and the percentages of decoupling of Cu 2p3/2 and its

satellites of copper–cobalt film coatings synthesised using various concentrations.

Coatings

synthesised

using

concentrations

Binding energies and the percentages of

the components of Cu 2p3/2 photoelectron

line

Binding energies and the

percentages of satellites

Label: i ii iii Sat. I Sat. II

[Cu]=[Co]=

0.15 M

931.2 eV

(3.8 at%)

932.8 eV

(53.7 at%)

933.8 eV

(31.9 at%)

940.7 eV

(5.6 at%)

943.2 eV

(5 at%)

[Cu]=[Co]=

0.2 M

931.2 eV

(2.9 at%)

932.7 eV

(48.7 at%)

933.7 eV

(35.4 at%)

941.3 eV

(10.6 at%)

943.8 eV

(2.4 at%)

[Cu]=[Co]=

0.25 M

931.4 eV

(4.9 at%)

932.8 eV

(56 at%)

933.6 eV

(32.1 at%)

940.5 eV

(3.1 at%)

943.2 eV

(3.8 at%)

Attributions Octahedral

Cu+

[207-210]

Tetrahedral

Cu+

[172, 202-

206]

Octahedral

Cu2+

[162, 172,

212-216]

Paramagnetic

Cu2+

[162, 172,

181, 202,

215, 216]

Paramagnetic

Cu2+

[162, 172,

181, 202,

215, 216]

The curve-fitting of Co 2p3/2 and its satellites from every coating gives five peak

components (Figure 6.7b–d). The peaks in the region of 779.1–780.0 eV are mostly due to

107

Co2O3 (or Co(III)) and Co3O4 (or mixed Co(II,III)) bonding states. The peak binding energy

of Co 2p3/2 above 780.0 eV with a shake-up satellite is the characteristic of CoO (or Co(II)).

Figure 6.7. (a) Co 2p SR-XPS spectra of copper–cobalt thin film coatings

synthesised using various concentrations, (b)–(d) decoupling of Co 2p3/2 of copper–

cobalt thin film coatings synthesised using various concentrations.

In a copper–cobalt oxide environment, the types of coordination (octahedral/tetrahedral) of

Co 2p3/2 have been specifically identified by some researchers [146, 162, 213]. The

quantitative analysis of Co 2p3/2 is presented in Table 6.2. From Table 6.2, the tetrahedral and

108

paramagnetic Co(II), octahedral Co(III) as well as mixed Co(II,III) oxidation states are

detected with the tetrahedral Co(II) ions predominant.

The presence of oxidation states here is quite different from the oxidation state of the

cobalt chloride precursor used. Cobalt chloride has been widely known to contain the

octahedral cobalt (II) oxidation state surrounded by six chloride ions. To understand our

results, it is necessary to understand the changes in the oxidation state due to the thermal

influence on the cobalt oxide surface synthesised from the cobalt chloride precursor. In a

relatively low temperature synthesis process (under 100oC), the cobalt oxide surface

synthesised using the cobalt chloride precursor has already shown a differentiation in the

composition of the oxidation states by the presence of both Co(II) and Co(III) species [217].

In a higher temperature treatment of around 330–350oC, the mixed oxidation states of cobalt

(Co3O4), which has a normal spinel crystal structure based on a close-packed face centred

cubic configuration of O2-

ions where Co(II) ions occupy the one-eighth of the tetrahedral A-

sites and Co(III) ions occupy one-half of the octahedral B sites, are detected [218]. Further,

Table 6.2. The binding energies and the percentages of decoupling of Co 2p3/2 and its

satellites of copper–cobalt film coatings synthesised using various concentrations.

Coatings

synthesised

using

concentrations

Binding energies and the percentages of the

components of Co 2p3/2 photoelectron line

Binding energies and the

percentages of satellites

Label i ii iii Sat. I Sat. II

[Cu]=[Co]=

0.15 M

779.2 eV

(11.5 at%)

779.8 eV

(25 at%)

780.8 eV

(48.4 at%)

785.6 eV

(9.2 at%)

789.1 eV

(5.9 at%)

[Cu]=[Co]=

0.2 M

779.1 eV

(17.15 at%)

779.9 eV

(23.7 at%)

780.9 eV

(45.5 at%)

785.6 eV

(8.3 at%)

788.9 eV

(5.3 at%)

[Cu]=[Co]=

0.25 M

779.4 eV

(11.8 at%)

780 eV

(30.4 at%)

781 eV

(46.4 at%)

786 eV

(5.9 at%)

789 eV

(5.4 at%)

Attributations Octahedral

Co(III)

[146, 162,

198, 219]

Co(II,III)

[213, 220-

222]

Tetrahedral

Co(II)

[162, 212,

213, 219]

Paramagneti

c

Co(II)

[162, 219]

Paramagn

etic

Co(II)

[162, 219]

109

the octahedral Co(III) ions here have a more significant role determining the surface activity

rather than the tetrahedral Co(II) [218, 219]. The relative amount of Co(II) ions in

tetrahedral sites is found to increase with the increase of calcination temperature from 450oC

to 650oC [220] due to the reduction of Co(III). Finally, it is widely known that at around

950oC, the mixed cobalt(II,III) oxidation states convert fully to cobalt(II) oxide with the

following reaction: 2Co3O4→ 6CoO + O2. In line with this evolution, it can be understood

that the presence of various cobalt oxidation states corresponds with different types of

coordination in our samples, and the tetrahedral Co(II) ions are predominant since the

temperature synthesis is above 450oC.

In a copper–cobalt mixed oxides system, the Co(II) ions are partially substituted by

Cu2+

ions [162, 221]. If Cu2+

ions and octahedral Co(III) ions are present in the copper–cobalt

oxide system, then the oxide could be represented by the Cu2+

Co2III

O4, a form of copper–

cobalt spinel structure.

6.4. Conclusions

In this study, copper–cobalt oxides thin film coatings have been successfully

deposited on highly reflective aluminium substrates using a sol–gel dip-coating method. EDX

analysis reveals that the coating contained copper, cobalt and oxygen compounds with an

excessive amount of copper and oxygen which were characteristics for the copper cobalt

oxides spinels family. SR-XPS, showed that (i) the oxygen consists of lattice, surface and

subsurface oxygen, (ii) the copper consists of octahedral and tetrahedral Cu+, octahedral and

paramagnetic Cu2+

oxidation states, and (iii) the cobalt consists of tetrahedral and

paramagnetic Co(II), octahedral Co(III) as well as mixed Co(II,III) oxidation states. The

EDX and XPS analysis results indicate the presence of the copper cobalt oxide family which

corroborates XRD analysis results in Chapter Five. Absorptance optimization study reveals

110

that the coating synthesised using 0.25 M copper and 0.25 M cobalt (Cu/Co ratio = 1) with

dip-speed 120 mm/min (four cycles) represented the optimal coating design with absorptance

value of α = 83.4%, where the average of film thickness was around ~320 nm. The simplicity

of the dip-coating system which facilitated the sol–gel process implies that, additionally, such

a system could be extended for the coating of other mixed metal oxides. Our data can be used

to aid the engineering design of highly tuneable thin film metal oxides for numerous

industrial applications.

111

Chapter Seven

SURFACE AND MECHANICAL CHARACTERISATIONS OF

COPPER COBALT OXIDE THIN FILM COATINGS

SYNTHESISED USING DIFFERENT COMPOSITIONS

7.1. Introduction

Copper cobalt mixed oxides (CuxCoyOz) such as spinel-type and delafossite-type of

copper cobalt oxide have attracted much attention, and have been studied by many

researchers for a wide range of applications, such as oxygen evolution reactions (OER), the

Fischer-Tropsch process, the synthesis of syngas-based alcohol and as thermoelectric

materials [146, 162, 171, 172, 186-192, 194, 195, 223, 224]. Such intensive focus is

attributed to their high stabilities and high surface catalytic activities, good corrosion

resistance, cost-effectiveness and availabilities [172, 224]. Many studies have revealed their

physicochemical, electrochemical, magnetic, conducting as well as thermal properties [146,

162-165], all of which are essential to afford functionalities and enhance application

performances for these materials.

Physicochemical properties of copper cobalt oxides such as their surface

morphologies and surface electronic structures play an important role in governing

electrocatalytic reactions or thermoelectric applications. For spinel electrocatalysis, it was

reported that the mixed oxidation states of the cations placed in the octahedral sites (the

external sites) are the main contributing factor in an increase of the electrical conductivity

that facilitates the adsorption of the oxygen (O2 gas or OH- ions) by providing donor-acceptor

levels (d-orbitals) for chemisorption, which in turn enhances electrocatalyst activity [146].

On the other hand, it has been well established that the distribution of the mixed cationic

112

valences, which is the key determinant of catalytic activity and other physicochemical

properties, is erratic because the distribution is strongly influenced by the preparation

method, i.e. the chemical nature and the compositions of the precursors, as well as the

annealing atmosphere [146, 162, 172]. For delafossite, the mixed cationic valences are also

critical in determining the high electrical conductivity. Beekman et al. [195] characterised the

delafossite-type of CuCoO2 prepared by ion exchange (metathesis) solid-state reaction

between CuCl and LiCoO2. They found that the electrical transport and magnetic

susceptibility data for polycrystalline CuCoO2 were consistent with formal charge

assignments of Cu+ and Co

3+ for the transition metal constituents and corroborated recent

density functional theory calculations for this material.

Mechanical properties such as hardness (H), elastic modulus (E) and elastic strain to

failure (related to the ratio of hardness and elastic modulus, H/E) are the important

parameters required to estimate the wear resistance of nanostructured surfaces or coatings.

The hardness itself is a good measure of resistance against abrasive wear; however when

taking into account the presence of plastic deformation mechanisms, the H/E ratio is a more

suitable parameter [85]. The mechanical properties of copper cobalt oxides are less studied.

However, metal oxides generally show stability at high temperatures in air, are inert and do

not inter-diffuse at working temperatures. In fact, some metal oxides such as Alumina,

Chromia and Titania exhibit high hardness and elastic modulus [225-227]. A material with a

high hardness and a lower elastic modulus is deemed to have a better toughness when plastic

deformation is dominant and is therefore better suited for optimising the wear resistance of

‘real’ industrial surface materials [85]. Hence by lowering the elastic modulus while

maintaining hardness, an increase of the resistance against cracking can be achieved.

This chapter analyses and discusses the surface morphology, surface

composition/electronic structure and its local coordination as well as the mechanical

113

properties of the copper cobalt oxide thin film coatings synthesised using different copper to

cobalt concentration ratios ([Cu]/[Co]=0.5, 1 and 2). It is needed to fully comprehend the

nature of the optimised coating obtained in Chapter Six. To this end, field emission scanning

electron microscopy (FESEM), high resolution synchrotron radiations X-ray photoelectron

spectroscopy (SR-XPS) in combination with the synchrotron-based near edge X-ray

absorption fine structure (NEXAFS) spectroscopy and mechanical nanoindentation analysis

have been employed for characterisations. As the deposited coatings on aluminium substrates

exhibit nano-sized grain-like morphology with superior wear-resistant characteristics

compared to the aluminium substrate. Finite element modelling (FEM) has been used to

complement the existing experimental data and to establish their load-bearing ability.


The copper-cobalt oxide coating samples were prepared using a similar procedure

described in Section 4.1.2. For more specific, copper (II) acetate monohydrate and cobalt (II)

chloride (0.125 to 0.3 M) were mixed in absolute ethanol using propionic acid as complexing

agent to produce solutions with [Cu]/[Co] concentration ratios of 0.5, 1 and 2. Thin film

coating depositions on aluminium substrates were carried out with withdrawal rate 120

mm/min and four dip-heating cycles. For the nanoindentation test, the thicker coatings were

fabricated by more dip-heating cycles (30 times) to minimise substrate effects.

Characterisations were surface morphology analysis using FESEM, surface chemistry

composition analysis using synchrotron radiation SR-XPS in combination with the

synchrotron-based near edge X-ray absorption fine structure (NEXAFS) spectroscopy for

local coordination study as well as the mechanical properties analysis using nanoindentation

test and finite element modelling (FEM). Further elaborations on these instruments and the

characterisation techniques can be found in Section 4.2.

114


7.3.1. Surface morphology

Figure 7.1. (a-1, b-1, c-1) are micrographs illustrating the surface morphologies of

copper cobalt oxide thin films synthesised using different copper to cobalt concentrations

ratios. Generally, all surfaces exhibit coarse surface morphologies except for the [Cu]/[Co]=2

sample which has a smoother surface (Figure 7.1.c-1). With closer examination (Figure 7.1.

a-2, b-2, c-2), the surfaces consist of grain-like nanoparticles with sizes ca 10 - 60 nm. For

[Cu]/[Co] = 1, the particles appear to agglomerate to form a coral-like morphology embedded

within pores/trenches (Figure 7.1.b-2). Interestingly, relatively homogeneous particle sizes

and arrangement are observed for [Cu]/[Co] = 0.5 as compared to [Cu]/[Co] = 1, implying a

more pronounced particle agglomeration for the latter. Such morphologies were previously

reported by others researchers [171, 172, 200] for their porous copper–cobalt oxide layers

synthesised via thermal decomposition of copper and cobalt nitrate precursors for electro-

catalytic application. Marsan and co-researchers [171] suggested that the porous/rough

morphology of the copper-cobalt oxide surface was attributed to the higher evolution of gas

volumes (NO2, O2) during the decomposition of the concentrated nitrate coating. In line with

their analysis, the observed morphologies of our copper-cobalt oxide coatings could be

attributed to the evolution of O2 from the high temperature decomposition of copper and/or

cobalt oxides which ultimately form the copper cobalt oxide [150, 173].

These porous and rough surface structures may explain the different absorptance

performances exhibited by the coatings as described in Chapter Six. Low porosity of coating

exhibited by [Cu]/[Co]=2 sample has a direct correlation to the decrease of ~10% in

absorptance value as seen in Table 7.1. The higher porosity and rougher surface are more

conducive structures for higher absorption of incidental solar radiation due to the higher

multiple reflections, resonant scattering and relaxation mechanisms that occurred on the

115

surface and inside the pore/aggregate [8, 42]. However, the influence of surface morphology

to the optical behaviour is not significant. The optical property of material mainly depends on

its electronic structure or band structure as that will be discussed in Section 8.3.3.

Table 7.1. Correlation between the [Cu]/[Co] ratio and the porosity

[Cu]/[Co] ratio Porosity (qualitatively) Absorptance (see Section 6.3.2)

0.5 good-fair 86.77%

1.0 fair 83.4%

2.0 low 74.13%

Figure 7.1. Surface morphologies of copper cobalt oxide coatings

synthesised using a) [Cu]/[Co]=0.5, b) [Cu]/[Co]=1 and c) [Cu]/[Co]=2.

116

7.3.2. Synchrotron radiation XPS study

High resolution SR-XPS was used to afford detailed information regarding the

electronic structure of the thin film coatings. Figure 7.2 shows the Cu 2p SR-XPS spectra and

the decoupling of Cu 2p3/2 peaks of copper cobalt oxide film synthesised using different

concentrations ratios. In each spectrum, the two main peaks of Cu 2p3/2 and Cu 2p1/2 and the

satellites on the high energy side of each of the main peaks can be observed (Figure 7.2a).

The presence of these satellites represents evidence of an open 3d9 shell of Cu

2+ [228]. In

each spectrum, the binding energy difference between Cu 2p1/2 and Cu 2p3/2 which is around

19.9 eV, and the shake satellite on the high energy side of Cu 2p3/2 peak, confirm the

presence of Cu2+

ions. The shorter separation between the Cu 2p3/2 line and its satellite peak,

and the higher value of satellite intensity to Cu 2p3/2 main peak intensity (Isat/Imain) ratio,

signify a decrease in the covalent character of the Cu-O bond in copper cobaltite as compared

to CuO [172]. Relatively higher intensity of Cu 2p satellites in the sample with a [Cu]/[Co]

ratio of 2 also indicates the higher number of Cu2+

ions which are not incorporated into the

copper cobalt spinel structure compared with the other two concentration ratios.

The decoupling of Cu 2p3/2 peak and its satellite in each coating is shown in Figure

7.2.b-d. The decoupling provide five curve-fitting components except for sample with

[Cu]/[Co] ratio of 2 which does not contain any component below the binding energy of

932 eV (Figure 7.2. d). The quantitative analysis results are presented in Table 7.2. It is

commonly recognised that the Cu 2p3/2 photoelectron peaks at ca 933.3–934.0 eV are due to

the Cu2+

. Many researchers [229] identify that the Cu 2p3/2 photoelectron peak at ca 932.5–

932.8 eV is attributed to the tetrahedral Cu+ with its counterpart peak from the octahedral Cu

+

located below the tetrahedral one. From Table 7.2, it can be observed that the tetrahedral Cu+

is the more prominent oxidation state in each sample except for the [Cu]/[Co] = 2 sample

whereby the numbers of tetrahedral Cu+ and octahedral Cu

2+ are relatively balanced. The

117

increase of copper content tends to promote the formation of octahedral Cu2+

and to reduce

the formation of octahedral Cu+. It is shown that for the [Cu]/[Co] = 2 sample, there is no

reduction of Cu2+

in the octahedral environment compared to the other two samples. In

Figure 7.2. a) Cu 2p SR-XPS spectra of copper cobalt thin film coatings synthesised using

different [Cu]/[Co] concentration ratios, b-d) decoupling of their corresponding Cu 2p3/2

peak.

addition, the absence of octahedral Cu+ implies that there are less amounts of typical

monophasic Cu-Co mixed oxides in the coating for the [Cu]/[Co] = 2 sample compared to the

other two samples [172]. The effect of the absence of the octahedral Cu+ on the electronic

structure of CuxCoyOz films can be explained by studying the local environment of

118

components via NEXAFS study as seen in Section 7.3.3. The presence of Cu2+

ions in an

octahedral environment is in contrast with the copper spinel structure, where copper occupies

predominantly the tetrahedral sites [228]. This suggests that in the copper cobalt system, the

Cu2+

ions themselves are ‘guests’ which partially substitute the tetrahedral Co2+

in the cobalt

structure host [229]. Based on these observations, it can be suggested that in the copper-

cobalt system with a [Cu]/[Co] ratio of 2, the elevated concentrations of copper have

increased the competitiveness of octahedral Cu2+

ions which facilitates occupation of the

Co2+

sites in the cobalt structure host while minimizing the driving-force of Cu2+

to undergo

reductions.

Table 7.2. Binding energies and the percentage compositions derived from the decoupling of

Cu 2p3/2 peak and its satellites in the copper cobalt film coatings.

Film coatings Binding energy and percentage of the

components of Cu 2p3/2 photoelectron line

Binding energy and the

percentage of satellites

p q r Satellite I Satellite II

[Cu]/[Co]=0.5 931.5 eV

(3.7 at%)

932.8 eV

(45.5 at%)

933.8 eV

(35.7 at%)

940.9 eV

(11.9 at%)

943.6 eV

(3.2 at%)

[Cu]/[Co]=1 931.5 eV

(2.8 at%)

932.9 eV

(51.4 at%)

933.9 eV

(35.8 at%)

941.5 eV

(7.6 at%)

944 eV

(2.5 at%)

[Cu]/[Co]=2 - 932.7eV

(40.6 at%)

933.8 eV

(42.0 at%)

941.0 eV

(13 at%)

943.6 eV

(4.4 at%)

Attributions: Octahedral

Cu+

Tetrahedral

Cu+

Octahedral

Cu2+

Paramagneti

c Cu2+

Paramagneti

c Cu2+

The profile of the Co 2p spectra is shown in Figure 7.3a. In each spectrum, the two

main peaks are attributed to Co 2p3/2 and Co 2p1/2, and the weak satellites located on the high

energy side of each these main peaks are also found. Qualitatively, the presence of satellite

on the high energy side of the Co 2p3/2 peak indicates the presence of Co2+

ions. The Co 2p3/2

peak and Co 2p1/2 peaks separated by a spin-orbit splitting of ~15 eV correspond to the mixed

Co2+

and Co3+

ions, while the weak intensity satellite located in between the Co 2p3/2 and Co

2p1/2 indicates that the Co ions are present in a partial spinel-type lattice arrangement. The

weak satellite structures are characteristic of spinel structures in which 3+ cations occupy

119

octahedral lattice sites with diamagnetic, filled t2g and empty eg levels, and 2+ cations are in

tetrahedral sites [218]. The observed asymmetry in the Co 2p1/2 peak confirms the existence

of both Co2+

and Co3+

ions.

Figure 7.3. a) Co 2p SR-XPS spectra of copper cobalt thin film coatings synthesised

using different [Cu]/[Co] concentration ratios, b-d) decoupling of their corresponding Co

2p3/2 peak.

In every spectrum, the decoupling of the Co 2p3/2 peak and the satellite on the high

energy side of this peak provides five curve-fitting components (Figure 7.3. b-d). The binding

energy and the percentage of each component are tabulated in Table 7.3. The peaks in the

120

region of 779.1–780.0 eV are mostly due to Co3+

in an octahedral environment and mixed

Co2+,3+

bonding states. Due to the covalence and final state effects, the binding energy of

Co2+

is higher than the Co3+

[230] and it is found mostly above 780.0 eV with a characteristic

shake-up satellite [229]. In Table 7.3, it can be seen that in all samples, the tetrahedral Co2+

ions dominate. However, in a copper-cobalt mixed oxide system, the Co2+

ions are partially

substituted by Cu2+

ions, forming a lower crystallization of copper-cobalt spinel particles

[162]. If Cu2+

ions and octahedral Co3+

ions are present in the copper-cobalt oxide system,

then the oxide could be represented by Cu2+

Co23+

O4, a form of copper–cobalt spinel structure.

Table 7.3. Binding energies and the percentage compositions derived from the decoupling of

Co 2p3/2 peak and its satellites in the copper cobalt film coatings.

Film coatings Binding energy and the percentage of the

components of Co 2p3/2 photoelectron

line

Binding energy and the

percentage of satellites

p q r Satellite I Satellite II

[Cu]/[Co]= 0.5 779.2 eV

(10.4 at%)

780 eV

(22.9 at%)

780.8 eV

(51.75 at%)

785.9 eV

(8.4 at%)

789.1eV

(6.6 at%)

[Cu]/[Co]= 1 779.4 eV

(9.7 at%)

780.1 eV

(26.6 at%)

781.1eV

(45.8 at%)

786 eV

(11.5 at%)

789.3 eV

(6.5 at%)

[Cu]/[Co]= 2 779.2 eV

(8.4 at%)

779.9 eV

(24.8 at%)

780.7 eV

(52.7 at%)

786.1 eV

(8.4 at%)

789.4 eV

(5.7 at%)


Co3+

Mixed

Co2+,3+

Tetrahedral

Co2+

Paramagneti

c Co2+

Paramagneti

c Co2+

Figure 7.4 shows the O 1s SR-XPS spectra of copper cobalt oxide film coatings and

the corresponding curve-fitting resulting from the decoupling of the 1s peak. The O 1s

spectrum exhibits a strong peak with a shoulder at the high binding energy side of O 1s peak,

except for the [Cu]/[Co] = 2 sample, where a relatively lower intensity peak is identified

(Figure 7.4.a). The apparent shoulder at the higher energy side of the O 1s main peak is a

typical feature of copper-cobalt mixed oxides [162]. The decoupling of O 1s photoelectron

spectrum in each sample results in four curve-fittings grouped into three components (Figure

7.4.b-d). The curve-fitting peaks at binding energy (BE) at around 529.4-529.5 eV (labelled

121

“p”) are mostly attributed to the lattice O2-

in a Co3O4 spinel structure. The presence of this

spinel as the main structure reaffirms that in the mixed copper-cobalt system here the cobalt

oxide itself becomes the spinel host with copper ions as “guests”, and then a copper cobalt

oxide spinel structure is formed as elucidated earlier.

Figure 7.4. a) O 1s SR-XPS spectra of copper cobalt thin film coatings synthesised using

different [Cu]/[Co] ratios, b-d) decoupling of their corresponding O 1s peaks and shoulders.

The curve-fitting peaks at BE in range of 530.4-531.4 eV (labelled “q” and “r”) may

be attributed to surface oxygen from a wide variety of species such as chemisorbed oxygen

122

O-, oxygen containing surface contamination, and/or OH-like species, as hydroxyl, carbonate

groups, etc [146, 162, 172, 198-201], while the curve-fitting peaks at BE above 531.4 eV

(labelled “s”) could be assigned to subsurface O- species [177, 178, 231]. Dupin et al. [177]

suggested that the subsurface (bulk structure near the surface) oxygen ions had lower electron

density than the lattice O2-

ions. They could be associated with sites where the coordination

number of oxygen ions was smaller than in a regular site, with higher a covalence of the M-O

bonds [177].

7.3.3. Synchrotron-based NEXAFS study

Further interfacial studies to detect the influence of copper to cobalt concentration

ratios to the local coordination of the electronic structure were performed using synchrotron

radiation NEXAFS spectroscopy. Figure 7.5 shows the Cu L-edge NEXAFS absorption

spectra in Auger Electron Yield (AEY) mode with monitoring photon flux incident (I0). The

AEY mode provides the best surface sensitivity compared to other modes [144]. Two main

peaks, i.e. Cu-L2 and Cu-L3 are observed at all spectra. All peaks and shoulders are found to

exist around the same photon energy. Generally, it can be seen that there is no significant

change in the spectral line-shapes with the change in copper and cobalt concentrations except

for the [Cu]/[Co] = 2 sample where significantly higher peak intensities of Cu-L2 and Cu-L3

are observed. This indicates that the local environment of Cu remains relatively invariant in

samples except for the [Cu]/[Co] = 2 sample.

The Cu-L3 and Cu-L2 absorption peaks are observed at photon energies of ~930.4 and

950.2 eV, respectively. These main peaks arise from the dipole transitions of the Cu 2p1/2 for

L2 and Cu 2p3/2 for L3 into the empty d-states [232]. The Cu-L3 peak is more sensitive to the

local environment than the Cu-L2 peak, which can be attributed to the Cu2+

ions [233-237].

This observation reinforces the conclusion espoused in the previous XPS analysis that there

123

would be changes in the local coordination if the copper/cobalt concentration ratio were

higher than 1, due to the loss of octahedral Cu+.

Figure 7.5. Cu L2,3-edge NEXAFS in AEY mode spectra of copper cobalt oxide

thin film coatings.

The absorption peaks of the Co L2,3-edge NEXAFS spectra for copper cobalt oxide

film coatings in AEY mode are shown in Figure 7.6. Generally, all spectra exhibit relatively

similar line-shapes and intensities which indicate that the local co-ordination of Co remains

relatively unchanged in all samples. It has been known that the Co L-edge feature is sensitive

to the change in electronic configuration; particularly it will change drastically with the

change in Co oxidation states and the spin-state transition [238, 239]. Based on this evidence,

the local environments of Co here are independent of the change of the copper to cobalt

concentrations ratios.

In Figure 7.6, each spectrum has two main prominent peaks with shoulders. The Co-

L3 peak which has absorption at ~779.6 eV has a shoulder on the low energy side with a

124

shoulder corner at ~ 777.8 eV and a thin shoulder/asymmetric behaviour on the high energy

side at area of around 781.6 eV. Another absorption peak as Co-L2 is found at 793.9 eV with

a relatively more pronounced shoulder on the high energy side. In addition, their features

show that the Co-L3 peak form is relatively narrow with regard to the low spin state [238].

The distance separation between L3 and L2 peaks is around ~14.3 eV, while no satellite peak

between the L3 and L2 peak can be found. These features are indicative of the presence of

Co3+

. The thin shoulder on the high photon energy side of Co-L3 peak confirms the absence

of CoIV

in the structure [240]. The low spin state is also confirmed by the branching ratio of

the L2 and L3 peak intensities, namely I(L3)/[I(L3) + I(L2)], which are around 0.5-0.55 (below

the statistical value) in each spectrum showing a low spin state Co3+

[241, 242]. Gautam et al.

[243] revealed that the similarity in form between the high energy side shoulder of Co-L3 and

Co-L2 peaks was due to the Coulomb and exchange interaction of 2p core holes with the 3d

electrons.

Figure 7.6. Co L2,3-edge NEXAFS in AEY mode spectra of copper cobalt oxide

thin film coatings.

125

Figure 7.7 shows the O K-edge NEXAFS spectra for copper cobalt film coatings in

AEY mode. All spectra have a relatively similar trend and it is clearly evident that the local

environments of O are relatively similar and they are relatively independent to the changes in

copper to cobalt concentrations. The O K-edge main absorption peaks are found at photon

energy values of around ~530.5 and ~542.4 eV.

The O K-edge NEXAFS feature can be used to investigate the hybridisation of the

metal 3d orbital with the host O 2p orbital. The O K-edge spectra in the binary metal mixed

oxide involving Co system emerge mainly due to the transition of the O 1s electron to the

conduction band near the Fermi surface, which is dominated by the O 2p and transition metal

3d hybridized orbital [244]. In each spectrum (Figure 7.7), the peak at ~530.5 eV is attributed

to the hybridization of Co 3d states with O 2p states which is close to the conduction-band

minimum and influenced by the density of unoccupied Co 3d states. Relatively sharp and

narrow peaks at ~530.5 eV are due to the low spin configuration [238]. The peak at around

542.4 eV is attributed to the transitions to the non-dispersive O 2pz and 2px+y. The shoulder

on the low energy side of the 542.4 eV peak could be due to the presence of the O vacancies

and Co, while the shoulder on the high energy side of this peak and the area above this

shoulder are attributed to O 2p hybridized with Co 4 sp states and O 2p states that extend to a

Co higher orbital, respectively [243-245]. The area between the two main peaks is attributed

to O 2p hybridisation with Co 3d states that form the bottom of the conduction band [244]. In

the case of copper-cobalt oxide system with cobalt as the host structure, these cobalt ionic

structure features suggest that the copper ions, as guests, have an interaction with the cobalt

host and that they are tetrahedrally coordinated with the ligand O atoms [243].

126

Figure 7.7. O K-edge NEXAFS spectra in AEY mode for copper cobalt oxide

thin film coatings.

7.3.4. Mechanical nanoindentation test

The representative load-displacement curves obtained from the nanoindentation

experiments on the different coating samples are shown in Figure 7.8. The elastic modulus

(E), hardness (H) and hardness to modulus ratio (H/E) values of the thin film coatings were

derived from the nanoindentation results and are presented in Figure 7.9 (a), (b) and (c),

respectively. The sample with [Cu]/[Co]=2.0 exhibits the highest average elastic modulus

while the samples with [Cu]/[Co]= 1 and 0.5 show a lower average elastic modulus values

with slightly differentiation (Figure 7.9a). In addition, it is evident that the elastic

moduli of the three samples here are significantly lower than that of the aluminium

substrate, which is consistent with our previous findings on coatings with similar

compositions (Section 5.3.5) . A different trend is shown by the hardness properties where

the sample with the intermediate [Cu]/[Co] ratio of 1.0 has the highest average hardness (~3.6

127

GPa) among all the three samples (Figure 7.9.b). Interestingly, the hardness is about twice

that of the aluminium substrate.

Figure 7.8. Load-displacement curves for the present coating samples.

The main factors governing the mechanical properties of nanostructure materials

include structural composition and the chemical nature [246]. From the elastic modulus and

the hardness results, the morphology and porosity factors in the surface of coatings as

described in Section 7.3.1 seem to have an influence on the mechanical properties, which is

reflected from the fact that there is a spread of the data points that results to an error about

~10%. This could be because the porosities or densities are not uniform either in the surface

or in the bulk. The increase of copper component in the copper cobalt oxide coating tends to

increase the elastic modulus. In fact, it is widely known that the elastic modulus value of

cobalt metal is higher than elastic modulus value of copper metal. If it is assumed that the

composition in the bulk of the coating near the surface (at the depth of penetration achieved

128

by the indenter) can be represented by the surface composition, then the elastic modulus of

the metal mixed oxide should decrease with the increase of the copper component in the

Figure 7.9. Mechanical properties of the as-deposited coatings derived

from the nanoindentation tests: (a) elastic modulus (b) hardness and

(c) H/E. The aluminium substrate is used for comparison.

coating. The fact that the opposite trend is observed indicates that in the mixed copper cobalt

oxide, the copper and cobalt components do not reflect their individual metal/oxide elasticity

129

characteristics. A new mixed metal oxide has been formed with different elastic modulus

characteristic compared to the individual components. Further, using the similar route of

synthesis, it could be estimated roughly that in the mixed copper cobalt oxide thin film

coatings, the coatings with minimum content of cobalt component would have a higher

elastic modulus compared to the coatings containing minimum amount of copper component.

The wear resistance of the coating can be evaluated using the E and H values obtained

from the nanoindentation experiments. The H/E ratio is considered to be an important

parameter for predicting wear resistance [85]. Compared to the aluminium substrate, the

present coatings are envisaged to have better wear resistance (Figure 7.9.c), particularly for

the [Cu]/[Co] = 1.0 sample which shows the highest H/E value (> 0.05). However, the

possible underling factors of chemical composition on mechanical properties are not clear

and further investigation is needed. The superior wear resistance exhibited by our coatings

has direct implications in terms of sustaining performance and function of the optical devices

during routine maintenance and service as mechanical contact can always be expected. For

comparison, the wear resistance or the toughness of our coating here is better than reported

values for other mixed metal oxides, for example (Ca3Co4O9) ceramic developed for

thermoelectric application which had a reported value of H/E of 0.03-0.04 [247, 248].

7.3.5 Finite element modelling (FEM)

To investigate the load-bearing ability of the as-deposited coatings, FEM was

conducted using the parameters listed in Table 7.4 as obtained from our nanoindentation

experiments. Simulation results for a coating synthesised using [Cu]/[Co]=1.0 are shown in

Figure 7.10, where the stress distribution under progressive loading is presented. The

concentrations of higher stress as well as the plastic zone were primarily restricted within the

coating layer by up to an indentation depth of 0.08 µm. This is expected because all the three

130

coating layers have lower elastic moduli but higher hardness than that of the aluminium

substrate in a way that there is little plastic deformation within the aluminium substrate below

Figure 7.10. Stress distribution of the [Cu]/[Co]=1 sample obtained from

FEM simulations for different indentation depths: (a) 0.02 µm, (b) 0.04 µm,

(c) 0.06 µm, and (d) 0.08 µm.

the interface. As a result, the coating delamination is suppressed, which typically occurs at the

interface between the coating and plastically deformed substrate during unloading [249].

Stress distributions for the other two coatings, i.e. [Cu]/[Co] = 0.5 and 2 samples, are similar

Table 7.4. Mechanical parameters used for FEM analysis.

Parameters (GPa) [Cu]/[Co]=0.5 [Cu]/[Co]=1.0 [Cu]/[Co]=2.0 Aluminium

E 66 67 77 132

H 3.3 3.7 3.2 1.455

Yield strength 1.1 1.2 1.05 0.5

131

to that to the [Cu]/[Co] = 1 sample, and hence they are not shown here for brevity. However,

when the same loading is applied directly onto the Al substrate, a marked difference is

observed, as shown in Figure 7.11, where the plastic zone of the loaded samples (coated and

uncoated) are evaluated from FEM results using domain integration and plotted against the

indentation depth. The size of the plastic zone resulting from the same loading has increased

5-7 times, indicating a significant increase in the plastic deformation, which is detrimental to

the integrity of the coating/substrate system. Hence in terms of load-bearing performance,

improvement can be expected when the coating layer with higher H/E is applied.

Figure 7.11. Change of the plastic zone size for the [Cu]/[Co] = 1.0 sample as

compared to the aluminium under increasing load as derived from domain

integration of the FEM results.

7.4. Conclusions

Copper-cobalt oxide thin films with different compositions have been successfully

deposited on aluminium substrates using a sol-gel dip-coating method and characterised via

FESEM, SR-XPS, NEXAFS and nanoindentation analyses. The surfaces of the [Cu]/[Co] =

132

0.5 and 1 samples typically composed of granular nanoparticles, while the [Cu]/[Co] = 2

sample had a smoother surface. The SR-XPS analyses showed that the copper electronic

structure consisted of octahedral Cu+ (except for [Cu]/[Co] = 2), tetrahedral Cu

+ and

octahedral and paramagnetic Cu2+

oxidation states. The cobalt electronic structure comprised

tetrahedral and paramagnetic Co2+

, mixed Co2+,3+

, and octahedral Co3+

oxidation states, in

which the tetrahedral Co2+

was predominant. The oxygen electronic structure consisted of

lattice, surface and subsurface oxygen. The increase of copper concentration in the synthesis

process tended to promote the formation of octahedral Cu2+

which minimised the formation

of octahedral Cu+ as well as increased the competitiveness of octahedral Cu

2+ ions to

substitute the Co2+

site in cobalt structure host. NEXAFS spectra revealed that the local

environments of Co, Cu and O were not significantly influenced by the change in the copper

to cobalt concentration ratios except for the [Cu]/[Co] = 2 sample where the local

coordination appeared to slightly change due to the loss of octahedral Cu+. Compared to the

aluminium substrate, the present coatings have significantly improved wear resistance,

particularly for the [Cu]/[Co] = 1.0 sample which shows the highest H/E value (> 0.05). FEM

modelling indicated that, under spherical loading conditions, the higher stress and the plastic

deformation were primarily concentrated within the coating layer, without spreading further

into the substrate. This would reduce the probability of delamination of the coating layer

during unloading phase. Our findings can be used to aid in the engineering design of metal

oxides coatings with superior wear-resistance for numerous industrial applications, such as

optical coatings and solar-selective absorbers.

133

Chapter Eight

CHARACTERISTICS OF COPPER COBALT OXIDES THIN

FILM COATINGS SYNTHESISED BY DIFFERENT

ANNEALING TEMPERATURES

8.1. Introduction

Copper cobalt oxides are a family of metal oxides which have found important

applications in electro-catalytic reactions and as thermoelectric material [146, 162, 171, 186-

192, 194, 195, 223]. To enable improved designs for optimal performance in these

applications, their physicochemical, electrochemical, magnetic, conductivity as well as

thermal properties have been intensely studied, in conjunction with their structural

characteristics [146, 162-165, 195]. From these previous studies, it can be construed that

temperature change in the synthesis process or application has substantial influence on their

physicochemical properties.

The temperature effect on the structural, magnetic and electronic structure properties

in the delafossite-type of copper cobalt oxides were established by Beekman et al. [195]. The

thermal analysis showed that the compound was stable until 680oC, whereupon a phase

transition event commenced. A weak temperature dependent magnetic susceptibility exists,

which remains negative in the temperature range from ~20 K to 300 K. There is not any

ferromagnetic or paramagnetic impurity contribution from samples at temperatures as low as

2 K [195]. The temperature independent diamagnetism reported for this type of copper cobalt

oxide is in agreement with formal charge assignments of Cu+

(d10

) and Co3+

(d6, low spin) as

suggested by Shannon et al. [250], as well as the analysis of the electronic band structure

determined by density functional theory (DFT) calculations [194, 195]. The spinel-type of

copper cobalt oxides tends to form a low crystallized single phase of copper cobalt oxide with

134

a partially inverted spinel structure and minor segregations of new cobalt and/or copper oxide

phases, which depend on the Cu/Co ratio in the precursor salt as well as the calcination

temperature [162, 172]. The increase of calcination temperature is typically accompanied by

an increase in the degree of crystallinity of phases in copper cobalt oxides [165]. Nonetheless,

the opposite result was observed by Shaheen [165], where the degree of crystallinity of

detected phase in copper cobalt oxide synthesised by lower content of copper compared to

cobalt decreased. Indeed, this discrepancy can be addressed by considering the dissolution of

more cobalt species in the lattice of the copper cobaltite phase, thus producing a more

homogeneous solid solution [165].

Compared to the above mentioned properties, the mechanical properties of optical

copper cobalt mixed oxides are seldom studied and, to the best of our knowledge, there is no

integrated experimental and modelling study on the mechanical properties of copper cobalt

oxides coatings. This is quite surprising, in view of the fact that mechanical strength and

durability are important in extending their service life. In Chapter Five and Six it shown that

the copper cobalt oxide coatings exhibited distinctive optical properties with a spectrally

selective profile in UV-Vis-NIR wavelengths region. There are, however, still many

unresolved engineering issues, especially those regarding to understanding the influence of

annealing temperatures on the physicochemical and mechanical properties of the coatings.

Therefore, the aim of this work is to investigate the structural, surface compositions, optical

and mechanical properties of copper cobalt oxides thin film coatings synthesised by different

annealing temperatures using XRD, XPS, UV-Vis-NIR and nanoindentation. Moreover, these

experimental results are used to evaluate the mechanical behaviour of the coatings by Finite

Element Modelling (FEM). The high absorptance value accompanied by the high mechanical

robustness of the copper cobalt oxide coating renders these coatings a promising material for

various applications, especially for solar selective absorption.

135

8.2. Experimental

Copper-cobalt oxides thin film coatings were deposited using a sol-gel dip-coating

technique described in Section 4.1 with some variations as elucidated in the following. The

copper and cobalt precursors (at 0.25 M for each) were mixed using absolute ethanol and

propionate acid. The resulting solution was then used for deposition on aluminium substrates

using a dip-coater at a withdrawal rate of 120 mm/min with relative humidity being

controlled below 55%, and subsequently heated on hot plate at 150°C for 10 seconds. A four

dip-heating cycles was conducted before final annealing in a close (atmospheric) oven

furnace at temperatures within the range of 500-650oC for 1 hour since it basically afford an

optimized reflective system compared to other number of cycles as seen in Chapter Six. If the

annealing temperature was set lower, residual organic groups would not be completely

removed, while temperatures higher than 650oC could also not be applied since it was limited

by the melting point of aluminium substrate. The increase of temperature to the final

annealing temperatures was conducted at ramp-rates of 50oC/min while cooling to room

temperature was allowed to occur naturally inside the closed furnace overnight.

Mineralogical characteristics of the thin films were analysed in a Bruker Advance D8

X-Ray Diffractometer (XRD) equipped with a Lynx-Eye detector. The surface bonding

structures were probed by XPS (Kratos Axis Ultra XPS spectrometer, Manchester, UK) with

Al Kα radiation (hν=1486.6 eV). The solar absorptance was recorded from 300 to 2700 nm

using a UV–Vis-NIR Jasco V-670. A nanoindentation workstation (Ultra-Micro Indentation

System 2000, CSIRO, Sydney, Australia) equipped with a Berkovich indenter, was used to

determine the mechanical properties of the films. Finite element modelling (FEM) was used

to simulate the physical response of the coating system under external loading. The details of

the model set-up have been given in Section 4.2.

136


8.3.1. XRD analysis

Figure 8.1a shows the XRD patterns of coating samples synthesised on aluminium

substrates and treated at different annealing temperatures. The peaks at ca 45° are attributed

to the aluminium substrate as also seen in Chapter Five. The peaks from the coatings are

found within the 2θ range of 30-42o, as seen in Figure 8.1b. They exhibit low intensities and

indicated poor crystallinity compared to peaks from the aluminium substrate. The low

crystallinity of the copper cobalt oxides synthesised by sol–gel technique was also reported

by other researchers [162]. Analyses of the peaks intensities and the d-spacing show that the

peaks from the coatings at around 35.3o (0 1 1), 36.9

o (3 1 0) and 40.2

o (3 0 1) (Figure 8.1b)

are assigned to CoCu2O3 (ICDD 76-0442) with the lattice parameters are in good agreement

with the orthorhombic crystal system (Space Group (#59) = Pmmn). The peaks at

approximately 31.3o and 38.5

o could be attributed to CuCoO2 (ICDD 21-0256) and CoCuO2

(ICDD 74-1855) phases. It clearly shows that the crystallinity along the direction of (301) of

CoCu2O3 increases extensively.

Analysis of domain size from the (310) and (301) peaks using the Debye-Scherrer

B

hklB

Kt

cos (8.1)

where K is the crystallite-shape factor (K=0.94 [251-253]); B=FWHM.

Table 8.1. Results of grain size derived using Debye-Scherrer formula from the

(3 1 0) and (3 0 1) lattice planes.

Annealing

Temperatures (oC)

Domain size (nm)

(310) plane (301) plane

500 26 221

550 53 252

600 61 196

650 101 196

137

Figure 8.1. (a) XRD patterns of the prepared copper–cobalt thin

film coatings on aluminum substrate at different annealing

temperatures, (b) Expanded XRD patterns within the region 30-42°.

The black dot of at around 2θ=38.5o belong to CoCuO2, while the

black dot of around 2θ=38.6o belong to CuCoO2

138

formula (equation 8.1) is tabulated in Table 8.1. The results indicate that as annealing

temperature increases, the domain measured perpendicular to the (3 1 0) lattice plane

increases, while the domain measured perpendicular to the (3 0 1) plane basically remains

unchanged.

It has been established that, the strain within a material may be evaluated by

measuring the d-spacing of the crystal planes using X-ray diffraction [254]:

z = (dn-d0)/d0 (8.2)

wherez is the stress component normal to the surface, d0 and dn are the strain free and

measured d-spacing, respectively. Within a coating layer of ~1 m thickness, the residual

stress z is normally zero [255]. As such, we have [256]:

z = - (x+y) = - (/E)(x+y) (8.3)

where is the Poisson’s ratio, E is the Young’s modulus, x and y are the in-plane principal

stresses along the x and y directions, respectively. Combining Equation (2) and (3), and

assuming that the coating layer is isotropic, i.e., x=y, we obtain:

2x= - (E/) (dn-d0)/d0 (8.4)

from which the in-plane residual stress can be estimated. In this work, E value is obtained

from nanoindentation as seen in Section 8.3.5, and a Poisson’s ratio of 0.3. The residual stress

within the range of the annealing temperature is in the order of ~0.5 GPa. Increasing

annealing temperature seems to reduce the tensile residual stress slightly (Table 8.2).

Table 8.2. Residual stress within the coating layer, estimated by using the (301) and

(310) peak position data from the X-ray diffraction

Annealing

temperature (oC)

2θ for (310) peak 2θ for (301) peak Tensile residual

stress, σx (GPa)

500 36.920 40.354 0.63

550 36.990 40.324 0.65

600 36.920 40.252 0.52

650 36.860 40.191 0.40

Reference 36.445 40.243 0.00

139

8.3.2. XPS study

Figure 8.2 and Figure 8.3 show the Cu 2p XPS spectra and the decoupling of Cu 2p3/2

peaks of copper cobalt oxide film coatings synthesised at different annealing temperatures,

respectively. In every spectrum, the two main peaks of Cu 2p3/2 and Cu 2p1/2 and the satellites

on the high energy side of these two main peaks can be found (Figure 8.2). Qualitatively, in

every spectrum, the binding energy difference between Cu 2p1/2 and Cu 2p3/2, which is around

19.8 eV, indicates the presence of a low oxidation state of copper, while the satellite peak

between Cu 2p3/2 and Cu 2p1/2 confirms the presence of Cu2+

. It is widely established that this

satellite arises due to the shake-up transition by a ligand metal 3d charge transfer that does

not occur with Cu+

species which have completely filled 3d shells. From Figure 8.2, it can be

seen that the Cu 2p3/2 satellite intensity to Cu 2p3/2 main peak intensity ratio (Isat/Imain) varies

slightly from 0.1 to 0.15 as the annealing temperature is increased from 500oC to 650

oC,

indicating that there is a decrease in the covalent character of the Cu-O bond in copper cobalt

mixed oxide [172].

Figure 8.2. Cu 2p XPS spectra of copper cobalt thin film coatings

synthesised at different annealing temperatures.

140

The decoupling of Cu 2p3/2 peak and its satellite in every coating is shown in Figure

8.3.a-d. Overall, the curve-fittings result in four components in every spectrum and they are

quantified in Table 8.3. It is commonly recognized that the photoelectron peak at around

932.3-932.4 eV of Cu 2p3/2 is usually from the tetrahedral Cu+. The components at around

933-934 eV with their satellites characteristic are due to the octahedral Cu2+

. From Table 8.3,

it can be seen that the tetrahedral Cu+ ions remain more prominent compared to the

octahedral Cu2+

ions, even though the annealing temperature is increased. The

Figure 8.3. Decoupling of Cu 2p3/2 peaks of copper cobalt thin film coatings synthesised

at different annealing temperatures.

141

increase of annealing temperature generally does not change the copper bonding structure in

the surface. The absence of a component at the low energy side of the Cu 2p3/2 peak indicates

that natural cooling overnight to room temperature inside the closed oven furnace might

prevents the reduction of octahedral Cu2+

compared to the relatively faster cooling outside the

furnace as reported in Chapter Six.

Table 8.3. The curve-fittings results of Cu 2p3/2 and its satellite of copper cobalt film coatings


Annealing

temperature

Binding energy and percentage

Cu 2p3/2

photoelectron line

Satellite I Satellite II

500 oC 932.3 eV

(43.7 %)

933.5 eV

(38.0 %)

940.5 eV

(9.4 %)

943.0 eV

(8.9 %)

550 oC 932.3 eV

(42.9 %)

933.5 eV

(38.9 %)

940.6 eV

(10.3 %)

943.1 eV

(7.9 %)

600 oC 932.3 eV

(47.1 %)

933.5 eV

(36.2 %)

940.6 eV

(8.1 %)

943.1 eV

(8.6 %)

650 oC 932.4 eV

(47.5%)

933.5 eV

(37.4 %)

940.6 eV

(7.4 %)

943.2 eV

(7.7 %)

Attributions: Tetrahedral Cu+ Octahedral Cu

2+ Cu

2+ characteristic satellites

Figure 8.4 shows the profile of Co 2p spectra for samples synthesised at different

annealing temperatures. Similarly, in every spectrum, the two main peaks can be attributed to

Co 2p3/2 and Co 2p1/2 and the satellites located in the high energy sides of these two main

peaks are also found. Qualitatively, the Co 2p3/2 and Co 2p1/2 peaks separated by a spin-orbit

splitting of ~15.9 eV and the Co 2p1/2 to Co 2p3/2 intensities ratio of 0.5 correspond to the

Co2+

ions [230]. The presence of a characteristic satellite on the high energy side of Co 2p3/2

confirms this bonding structure. Relatively low intensities satellites located in between Co

2p3/2 and Co 2p1/2 indicate that Co ions are present in a partial spinel-type lattice arrangement

142

and these low intensities satellite could also indicate the presence of Co3+

ions in mixing with

Co2+

ions [146]. The asymmetry in the Co 2p1/2 peak confirms the existence of Co2+

and Co3+

ions.

Figure 8.4. Co 2p XPS spectra of copper cobalt thin film coatings


The decoupling of the Co 2p3/2 peak and the satellite at the high energy side of this

peak in every spectrum provides five curve-fitting components (Figure 8.5. a-d). The peaks in

the region lower than 779.8 eV are mostly due to Co3+

in octahedral coordination while the

peaks around 780 eV are predominantly attributed to the mixed Co(II,III) bonding states. The

peak with binding energy of Co 2p3/2 above 780.0 eV with a shake-up satellite is

characteristic of Co2+

in tetrahedral coordination. The binding energy and the percentage of

each component are tabulated in Table 8.4. It can be seen that, in all samples, the tetrahedral

Co2+

ions dominate. Nonetheless, even though they are prominent in a copper-cobalt mixed-

oxides system, these Co2+

ions are partially substituted by Cu2+

ions [162, 221] forming

143

copper–cobalt oxide structures [146]. The increases of annealing temperatures between

500oC to 650

oC generally do not influence the cobalt bonding structure in the surface.

Figure 8.5. Decoupling of Co 2p3/2 peaks of copper cobalt thin film coatings


Figure 8.6. shows the O 1s XPS spectra and curve-fittings of copper cobalt oxide film

coatings synthesised at different annealing temperatures. In every spectrum, the O 1s exhibits

a strong peak with a shoulder at its higher binding energy side. The decoupling of the O 1s

144

photoelectron spectrum of samples results in four curve-fittings grouped into three

components. The component at binding energy around 529.3-529.4 eV (denoted “i”) is

Table 8.4. The curve-fittings results of Co 2p3/2 and its satellite of copper cobalt

film coatings synthesised at different annealing temperatures.

Annealing

temperature

Binding energy and percentage

Co 2p3/2 photoelectron line satellites

i ii iii iv v

500oC 779.0 eV

(13.3 %)

780.0 eV

(27.5 %)

781.9 eV

(31.7 %)

785.7 eV

(15.1 %)

787.7 eV

(12.4 %)

550oC 778.9 eV

(10.6 %)

779.9 eV

(24.6 %)

781.7 eV

(35.9 %)

785.5 eV

(13.3 %)

787.6 eV

(15.6 %)

600oC 778.9 eV

(9.24 %)

779.8 eV

(27.7 %)

781.7 eV

(34.5 %)

785.6 eV

(14.8 %)

787.9 eV

(13.7 %)

650oC 778.9 eV

(11.1 %)

779.9 eV

(26.1 %)

781.6 eV

(34.8 %)

786.0 eV

(22.7 %)

789.3 eV

(5.3 %)


Co(III)

Co(II,III) Tetrahedral

Co(II)

Co(II) characteristic

satellites

attributed to lattice O2-

in the structure, while the components at BE around 530.4-531.5 eV

(denoted “ii” and “iii”) may be due to the surface oxygen from a wide variety of species such

as chemisorbed oxygen O-, oxygen containing surface contamination, and/or OH-like species,

as hydroxyl, carbonate groups, etc [146, 162, 172, 198-201]. The component at BE around

531.8-532.5 eV (denoted “iv”) could be assigned to subsurface (bulk structure near surface)

O- species [177, 178]. The apparent shoulders at the high energy side of the O 1s main peaks

are the characteristic feature of the copper-cobalt mixed oxides family which distinguishes

them from O 1s on Co3O4 [162]. Overall, there is no change in the oxygen surface

compositions when the surfaces are treated at different annealing temperatures from 500oC to

650oC.

145

Figure 8.6. O 1s XPS spectra and curve-fittings of copper cobalt thin film coatings


8.3.3. Optical properties

The optical properties of the copper cobalt thin film coatings are evaluated on the

basis of absorptance (α) within the wavelength range of 0.3-2.7 µm. Absorptance is defined

as a weighted fraction between absorbed radiation and incoming radiation. The absorptance

of a thin film on a substrate can be determined in terms of reflectance as described by Duffie

and Beckman [44]. Low spectral reflectance indicates high absorptance and vice versa. The

reflectance spectra of all the thin film coatings on highly reflective aluminium substrates

146

synthesised at different annealing temperatures, together with their corresponding solar

absorptance values, are shown in Figure 8.7a. The prepared coatings exhibit low to moderate

reflectance with wavy curves consist of interference peaks at around 1.0-1.2 µm and

absorption edges at around 1.5-1.7 µm. The spectra essentially form solar selective absorber

curve profiles within UV-Vis-NIR wavelengths area. Similar phenomena of the presence of

Figure 8.7. Reflectance spectra and solar absorptance of copper–cobalt oxide thin film

coatings on aluminium substrates synthesised at different annealing temperatures.

interference peaks and absorption edges have also been reported by others researchers [22,

24]. The increases in temperature generally tends to raise the interference peaks and the

absorption edges positions that lower the absorptance values, except for the spectrum of

sample annealed at 550oC where the interference peak and absorption edge approach each

other leading this spectrum to have the smallest wavy curve amplitude and the highest

147

absorptance value among the coatings (α = 84.4%). Significant increases of the interference

peak and the absorption edge positions are indicated by the coatings synthesised at annealing

temperatures of 600oC and 650

oC that decrease the absorptance values up to about 8%

compared to the maximum absorptance value (Figure 8.7b).

From Figure 8.7a, it can be seen that the more significant changes of reflectance

spectra occur near the infrared (NIR) wavelength region (> 0.8 µm). It can be expressed that

the reflectance property of copper–cobalt oxide layer in the NIR wavelengths area is affected

by at least three factors; (1) the thickness of film coating, (2) the intrinsic properties of

coating material, and (3) the reflectivity property of the substrate. For coatings with similar

thicknesses, the reflectance curve features showed in Figure 8.7a are due to the integrated

factors of solar wavelengths absorptions/scattering by the coating material (intrinsic

properties) and the back-reflections of the NIR radiations transmitted through the coating

material by the highly reflective aluminium substrate. The increase of annealing temperature

from 500 to 650oC enhances the crystallinity of the coating material that subsequently could

increases the scattering by the larger crystallite leading to the decrease of absorption by the

coating.

The choice of substrate also has a substantial influence on the reflectance property of

the coatings. It is widely accepted that the longer the NIR wavelength, the more radiation will

be transmitted through the semiconductor coating material due to the smaller energy owned

by the radiations/photons, which makes them easier to pass the coating material without

being absorbed. This transmitted-through radiation will be then reflected back by the

reflective substrate (dark mirror absorber-reflector tandem concept). In view of this, it seems

that our coatings behave akin to a semiconductor material. The increase of annealing

temperature, i.e. more than 550oC in the coating synthesis process might increase the intrinsic

“band gap energy” of the coating. As such, smaller number of the incident NIR photons are

148

absorbed through the transition across the band gap, while more photons are transmitted

through the coating. This transmitted radiation will then be reflected back by the reflective

substrate which eventually increases the reflectance and decreases the absorptance.

8.3.4. Nanoindentation tests

Figure 8.8 shows representative load-displacement curves obtained from

nanoindentation experiments on the thin film coatings treated at different annealing

temperatures. From these curves the values of elastic modulus (E), hardness (H) of the thin

films and their wear resistance index (H/E) were derived and presented in Figure 8.9. From

Figure 8.8. Typical load-displacement curves obtained from coatings

treated at different annealing temperatures.

Figure 8.8, the level of resistance to deformation of copper cobalt oxide thin film coatings

increases with the increase of the annealing temperature; the coating annealed at temperature

650oC exhibits the highest resistance to deformation. The elastic moduli of all the coatings

149

Figure 8.9. Mechanical properties of the as-deposited coatings

derived from the nanoindentation tests, (a) elastic modulus, (b)

hardness, and (c) H/E. The wear resistance index of aluminium is

also displayed for comparison purpose.

150

are lower than that of the aluminium substrate, consistent with our previous findings as seen

in Chapter Five. In addition, the obtained hardness values of the present study are generally

consistent with those reported by other researchers [184, 185]. Following heat treatment,

there is an increasing trend in both elastic modulus and hardness for the coatings, albeit this is

not so pronounced for hardness. Hence, it can be construed that the heat treatment exerts a

positive impact on the mechanical properties of the coating layer. The spread of the

measurement results, and the associated errors in both the modulus and hardness, may be due

to the surface roughness and the porosity of the coatings as seen in Chapter Five and Seven.

Wear resistance is vital to the performance and reliability of the optical coatings

during service, where mechanical contacts are always expected. Previous studies indicated

that the hardness to modulus ratio, H/E, is an important parameter for predicting the wear

resistance [85]. Even though there is a decreased tendency in H/E ratio of the coatings with

the increase in annealing temperature, all coatings prepared in this work are envisaged to

have superior wear resistance when compared with the aluminium substrate (Figure 8.9(c)).

8.3.5. Finite Element Modelling (FEM)

FEM simulation was conducted using parameters listed in Table 8.5. The results for

coating annealed at a temperature of 650oC are shown in Figure 8.10, where the stress

distribution under progressive loading is presented. Notably, the higher stress, as well as the

associated plastic zone, was primarily concentrated within the coating layer, up to an

indentation depth of 0.06 µm. For the loading conditions modelled, only about half of the

maximum stress level could expand into the substrate. This is because all the coating layers

have lower elastic modulus but higher hardness than that of the aluminium substrate (EAl =

131.41 GPa, HAl = 1.455 GPa) as seen in Chapter Five, and consequently, little plastic

deformation would result within the Al substrate. Considering the fact that for all the

151

samples, variations in both Young’s modulus and Poisson’s ratio only occur within a narrow

range, contact-induced stress distributions in the other coatings are similar. Therefore, two

implications can be derived from the above analysis: a) coating delamination would be

suppressed, which typically occurs at the interface between the coating and plastically

deformed substrate during unloading [249, 257], and b) mechanical damage, once induced,

would be confined within the coating under moderate loading conditions. In contrast,

when the same loading is applied directly onto the Al substrate, a marked difference can be

observed in Figure 8.11, where the plastic zones of the loaded samples (coated and uncoated)

are determined from FEM results using domain integration and plotted against the indentation

depth. The size of the plastic zone resulting from the same loading has increased by 5-7

times, indicating a significant increase in the plastic deformation. It is worth noting that the

plastic deformation is detrimental to the integrity of the coating/substrate system. From these

results, improvement in the load-bearing performance is expected when applying to the Al

substrate with a coating layer having higher H/E, such as the coatings being developed and

studied here.

Table 8.5. Mechanical parameters derived from the nanoindentation and used for FEM

modelling

Parameters Temperatures (oC)

500 550 600 650

E (GPa) 91 101 102 105

H(GPa) 3.2 3.2 3.2 3.2

Yield strengthy) (GPa) 1.1 1.1 1.1 1.1

H/E 0.035 0.032 0.032 0.030

152

Figure 8.10. Stress distribution of coating treated at annealing temperature of 650

oC,

obtained from FEM simulations for different indentation depths: (a) 0.03 μm, (b)

0.04 μm, (c) 0.05 μm, and (d) 0.06 μm. The dark lines close to the bottom of each

model represent the interface between the coating and the substrate.

Figure 8.11. Variations of the plastic zone size in coatings synthesised at

annealing temperatures of 500-650oC compared to the aluminium under

increasing load, derived from domain integration of the FEM results.

153

8.4. Conclusions

The copper-cobalt oxides thin film coatings were deposited on the aluminium

substrates and then treated at different annealing temperatures within the range 500-650oC.

The resultant coatings were characterised via XRD, XPS, UV-Vis-NIR and nanoidentation

methods. An increase in the means size of the crystalline domains of the coatings was found

with the increase of annealing temperature. The chemical binding structures in the surface

characterised by XPS remained relatively unaltered with the change in the annealing

temperature. The copper electronic structure consisted primarily of tetrahedral Cu+ in

addition to octahedral Cu2+

. The cobalt electronic structure comprised tetrahedral Co2+

ions,

octahedral Co3+

and mixed Co(II,III) in oxidation states. The oxygen in oxidation states

consisted mainly of lattice O2-

with minor surface and subsurface oxygen. Optical properties

characterised by UV-Vis-NIR revealed that the increase of the annealing temperature to

550oC increased the absorptance which reaching the maximum value of α = 84.4%, while

further increases of temperature decreased the absorptance. This transition was caused by the

integrated effects of the intrinsic properties of coating material and the substrate surface

optical properties. Mechanical properties measured by nanoindentation tests revealed that

both the elastic modulus and the hardness had an increasing trend but there was a slight

decrease in H/E ratio as the annealing temperature was increased. However, by using H/E as

an indicator, the wear resistance of all these coating materials was expected to be superior to

that of the aluminium substrate. FEM modelling showed that, under mechanical loading

conditions, stress and plastic deformation were primarily concentrated within the coating

layers. This would reduce the likelihood of delamination of the coating layer upon unloading.

154

Chapter Nine

OPTICAL PROPERTIES AND THERMAL DURABILITY OF

COPPER COBALT OXIDE THIN FILM COATINGS WITH

INTEGRATED SILICA ANTIREFLECTION LAYER

9.1. Introduction

Solar thermal collectors such as the ubiquitous solar hot water panels are designed to

collect solar radiation and convert it into useful heat energy for various industrial and

domestic applications. A significant component that affects the efficiency of a solar thermal

collector system is the solar selective absorber (SSA) coating [1] which, ideally, should

absorb the incoming solar radiation (high solar absorptance) as much as possible with

concurrent low thermal emittance. The most frequently used industrial SSAs in recent years

are the metal particles in ceramic (cermet) structures which can be synthesised via

electroplating/electrochemical or sputtering/vacuum deposition techniques [6, 24]. Though

these techniques are effective, they are, nonetheless, not environmentally-friendly [7, 30, 31]

and sputtering/vacuum deposition processes are technically complicated and not cost-

effective [4, 6, 23, 32-34]. Concerted efforts by materials scientists are currently underway in

seeking alternative SSA materials which can improve on these characteristics.

Cobalt copper oxides are versatile metal oxides which have applications in a variety

of important catalytic reactions such as conversion of syngas to higher alcohols, oxidation of

carbon monoxide (CO) by O2, oxygen evolutions reaction (EOR), Fischer-Tropsch synthesis

and for thermoelectric power generation material [146, 162, 186, 190, 191, 194]. Numerous

studies have been conducted to establish the physicochemical, magnetic, conductivity,

electrochemical and thermal properties of copper-cobalt oxides [146, 162-165]. On the other

155

hand, solar-based optical properties of the copper-cobalt oxides thin film coating are

comparatively less well-studied [146].

In previous chapters, we have prepared the copper cobalt oxides thin film coatings

deposited on highly reflecting aluminium substrate via sol-gel dip-coating route. To increase

the absorptance and to protect the coatings from any degradation due to external factors, an

antireflection (AR) layer on the top of the absorber layer is needed. The aims of this chapter

are to prepare the copper cobalt oxides thin film coatings with a silica antireflection layer

(CuxCoyOz - SiO2) and to investigate their optical properties, selectivity and accelerated

thermal durability. The durability data obtained from this study are useful in establishing the

physical (wear and thermal) resistance of the coatings against extreme weather and external

conditions.

9.2. Sample Preparation and Characterisations

The copper-cobalt oxide coating samples and the silica antireflection layer were

prepared using a similar procedure described in Section 4.1.2. Specifically, the cobalt copper

oxides thin film coatings were prepared from their respective chemical; precursors (0.25 M

copper acetate and 0.25 M cobalt chloride) using a sol-gel dip-coating method. The dip-

coating withdrawal rate was fixed at 120 mm/min. The final annealing was conducted at

500oC for 1 hour. The heating rate of the annealing process was 50

oC/min, while cooling was

performed inside the furnace for 10 minutes before allowed to cool to room temperature.

The silica antireflection layer was prepared by mixing the TEOS with ethanol, while

0.06 wt% HCl solution was gradually added to the TEOS-ethanol solution. The molar ratios

of ethanol and water to the TEOS were 5 and 4, respectively. To ensure complete hydrolysis

process, the resulting mixture was stirred for 24 hours in a closed container. The obtained

solution with pH of 2.1 was used for the AR layer deposition by dip-coater with withdrawal

156

rates ranging from 10 to 40 mm/min. The wet AR layer was subsequently stored in a

desiccator before final annealing to 400oC for 30 minutes in an oven furnace and finally

allowed to cool to room temperature overnight inside the furnace.

The optical performance of the coatings with the silica AR layer on reflective

aluminium substrates (opaque surfaces) was calculated based on the absorptance (α) and

emittance (ε) values. These values were obtained from the measurements of monochromatic

reflectance in the wavelength area from 0.3 to 2.7 µm by using an Ultra violet– visible-near

infrared (UV-Vis-NIR) spectrometer and wavelength area more than 2.7 µm by using a FTIR

spectrometer. The accelerated thermal durability test was conducted using an oven furnace

based on the PC (performance criterion) value of IEA SHC Task 27 [90], while the adhesion

effectiveness between the film absorber and substrate was evaluated by the physical/cracking

inspection before and after thermal test.

Further elaborations on these instruments and the characterisation techniques can be

found in Section 4.2.


9.3.1. Reflectance spectra and solar absorptance

Figure 9.1 shows the reflectance spectra (wavelength range of 0.3-2.7µm) of the

copper cobalt oxide thin film coatings with the silica AR layer synthesised at different

withdrawal rates. The spectrum of copper cobalt oxide coating without a silica AR layer is

also presented for comparison purposes. Interestingly, it can be clearly seen that the addition

of a silica AR layer at increasing dip-coating speeds has a substantial effect on the reflectance

curves profile. The relatively significant changes occur when the AR layer was deposited at

withdrawal rates of 20 and 40 mm/min where their interference peaks (labelled “*”) and the

absorption edges (labelled “#”) positions shift to the lower wavelengths area compared to the

157

coating with silica synthesised using withdrawal rate 10 mm/min or coating without silica

(see the lines labelled “2” and “3” versus lines labelled “0” and “1” in Figure 9.1). Likewise,

the distance between the interference peak and absorption edge (amplitude) of these two

former silica coatings decreases ~60% compared to the amplitude of coating with silica

synthesised using the withdrawal rate of 10 mm/min or the coating without silica.

Theoretically, the increase of the withdrawal rate would increase the thickness of the silica

AR layer [43]. These observations evidently imply that the reflectance spectra profile

significantly alters with the increase in the thickness of the silica AR layer.


and without the silica AR layer within a wavelength range of 0.3-2.7µm with

corresponding solar absorptance (α) values.

Low spectral reflectance in the solar wavelength area indicates high absorptance and

vice versa. The increases of absorptance values for the three coatings with silica AR layers

158

are ca 1.0-1.5% while the AR layer synthesised with dip-speed of 10 mm/min exhibits the

highest absorptance among all the samples. The increase of absorptance after the addition of

silica is attributed to the enhanced solar absorption by the silica network which lowers the

effective refractive index of the film [32]. Further increases of withdrawal rates, however,

seem to have marginal effect on the absorptance values which suggests that the AR layer

thickness does not significantly influence absorptance (as opposed to reflectance spectra

profile, especially in the NIR area (>0.8 µm)). This may be elucidated based on the

characteristic of the Duffie and Beckman method [44] in counting the absorptance value

where the denser count occurred in the UV-Vis wavelengths area compared to the NIR area

due to the denser spectral distribution of solar irradiance in the UV-Vis area compared to the

NIR area [44, 46, 47], therefore the reflectance spectra curve profile below 0.8 µm becomes

crucial to note.

9.3.2. Emittance and selectivity

The reflectance spectra of copper cobalt oxide thin film coatings with and without a

silica AR layer within the mid-far infrared wavelength range are presented in Figure 9.2.

Similarly, the addition of silica AR layer at increasing dip-coating speeds has a substantial

effect on the reflectance curves profile. Within the wavelength range of around 3 - 8 µm, the

addition of an AR-layer increases the reflectance. However, within the wavelength range of

around 8-10 µm, the AR layer absorbs too much infrared light, thus increasing the thermal

emittance (ɛ). Within this range, the thicker the AR layer, the higher the infrared absorption.

This phenomenon is due to the strong phonon absorption of the Si-O stretching modes as also

reported by other researchers [20, 71]. The relatively weaker phonon absorption in the

wavelength range of around 15 µm is also observed and can be attributed to the phonon

159

absorption typically exhibited by the copper cobalt oxide coating which was as also reported

by Kaluza et al. for their CuCoMnOx coating [20].

The emittance (ε) value is defined as a weighted fraction between emitted radiation

and the Planck black body distribution. It may be determined based on the reflectance

spectrum data [44]. This parameter is generally used to explain the performance of a solar

selective absorber in the mid-far infrared wavelength range. High spectral reflectance within

this range indicates low thermal emittance and vice versa. From Figure 9.2, the overall

thermal emittance of copper cobalt oxide thin film coatings with a silica AR layer increases

with the increase of dip-speed. The significant increase of emittance at ca 50% occurs when

the dip-speed is increased from 20 to 40 mm/min which implies increased heat loss from the

coating surface and decreased efficiency. The marginal increase of emittance value is shown

by the coating with the AR layer synthesised using dip-speed 10 mm/min.

Figure 9.2. Reflectance spectra of copper cobalt oxide thin film

coatings with and without silica AR layer within wavelength range of

3.0-15.4 µm with corresponding solar emittance (ɛ) values.

160

Figure 9.3 shows the optimal reflectance curve of copper cobalt oxide thin film

coating with silica AR layer (dip-speed of 10 mm/min) within the wavelength range of 0.3-

15.4 µm corresponding to the optimum absorptance value of α = 84.96% and emittance value

of ε = 5.63% (selectivity, s = 15.1). An emittance value below 10% can be categorized as a

good emittance performance for a selective absorber material [6-8]. The dashed line in Figure

9.3 is the extrapolation line created in place of ‘noisy‘ spectrum at the end of the spectrum

measurement range generated by the equipments used (UV-Vis-NIR and FTIR).

Figure 9.3. Reflectance spectrum of copper cobalt oxide thin film coatings with

AR layer (dip-speed of 10 mm/min) within wavelength range of 3.0-15.4 µm.

9.3.3. Accelerated thermal durability test

The accelerated thermal durability test was conducted to determine the estimated

service lifetime of a selective absorber surface based on its thermal behaviour at a high

161

temperature range. This is because the real application of a selective absorber is strictly

insulated under a transparent glass cover; therefore the thermal behaviour became the most

essential factor determining the quality of the absorber film. The International Energy

Agency (IEA) developed an accelerated thermal durability test to assess the thermal collector

performance called performance criterion (PC) through the IEA SHC Task 27 [90]. This test

procedure assumes that the activation energy of a certain degradation process is sufficient to

ensure absorber durability under natural working conditions of a flat thermal collector [38].

Table 9.1 shows the PC values of the coatings with a silica AR layer based on the

IAEA SHC Task 27. Based on this procedure, we selected the temperature 265oC as applied

in the thermal test (T1) for several testing times (t1) as our coating showed the optimum

absorptance value of around 84-85% and emittance value of around 5-6% (see Section 4.2.8,

Appendix 1 and Appendix 2). From Table 9.1, it can be seen that the variations in emittance

before and after the thermal test (∆ε) are much higher than the variations of absorptance (∆α).

The change of emittance (∆ε) values increases with the increase of testing time, while the

change of absorptance (∆α) values is relatively small and negligible. As such, the PC values

absorber coatings are more influenced by the changes in emittance. The PC values increase

with the increase in the testing times while for the testing times of 75 and 150 hours, the

recommended PC value (PC = 0.05) for a qualified absorber was exceeded. As such, longer

testing times were not required. Instead, an additional test was carried out using a lower

temperature (T2 = 235oC) for t2=179 hours. It is observed that the PC (235

oC; 179h) is less

than PC (265oC; 36h). Based on the PC value criteria, the coatings with the silica AR layer

pass the accelerated thermal durability test. Figure 9.4 shows the reflectance spectra of

copper cobalt oxide thin film coatings with an AR layer within the wavelength range of 0.3-

15.4 µm before and after the thermal test at 265oC for 36 h and 235

oC for 179 h.

162

Table 9. 1. Accelerated thermal durability parameter values obtained in the thermal test.

Parameters Testing times (t1) at 265oC

18 h 36 h 75 h 150 h

∆α -0.005 0.002519 0.000628 0.001132

∆ε 0.0283 0.08412 0.10895 0.14197

PC 0.01915 0.039541 0.053847 0.069853

Thermal test at 235oC for 179 h

∆α -0.0019

∆ε 0.060247

PC 0.032024

None of the samples showed observable visual changes and cracks before or after the

thermal test. This indicates the high thermal endurance of coatings and the strong adhesion

between the coating and the substrate. Figure 9.5 shows the physical conditions of the tested

coatings before and after the thermal test at 265oC for 150 h and at 235

oC for 179 h. Based on

these physical and cracking inspections, the coatings with AR layers fulfil the performance

criteria as good material for solar selective absorber. Nonetheless, the results in Table 9.1

reveal that the degradation of the coatings with the silica AR layer is more governed by the

temperature regime than the exposure time. This may be detected by comparing the PC

values between temperatures 235°C and 265°C where the PC value can be retained remain

low (<0.05) by decreasing the temperature test even though the test was carried out at a

longer of exposure time. As such, the coatings are qualified for uses within an extended

timeframe but in low temperature range applications (≤150°C) such as for domestic solar

water heating systems (Flat Plate Solar Selective Absorber) [3]. The flat-plate could reduce

the surface temperature and maintain its surface temperature at below 150 oC, so that this

coating material is suitable for such device.

163

Figure 9.4. Reflectance spectra of copper cobalt oxide thin film

coatings with AR layer before and after accelerated thermal

durability test at: a) 265oC for 36 h and, b) 235

oC for 179 h.

164

Figure 9.5. Photograph pictures of physical condition of copper cobalt oxide thin

film coatings with silica AR layer before (a1) and after (a2) thermal test at 265oC for

150 h, as well as before (b1) and after (b2) the thermal test at 235oC for 179 h.

165

9.4. Conclusions

The copper cobalt oxide thin film coatings with the silica antireflection (AR) layer

have been successfully deposited on reflective aluminium substrates using the sol-gel dip-

coating method. The addition of silica changed the reflectance spectra of coatings within the

wavelength range of 0.3-15.4 µm. The absorptance values increase slightly compared to the

coating without silica, but the increase of withdrawal rate of silica in the synthesis process

unfortunately also increased the emittance values due to the strong phonon absorption by the

silica, leading the optimum absorptance of α = 84.96% and emittance of ε=5.63% or

selectivity s=15.1 showed by coating with silica AR layer synthesised at withdrawal rate of

10 mm/min. The PC values results and physical inspections showed that the coatings with

silica AR layer passed the accelerated thermal durability test without any cracking detected.

The degradation of the copper cobalt oxide thin film coating with silica AR layer is governed

more by the temperature changes regime than the exposure time indicating that the coating is

qualified for long periods of uses in low temperature applications such as for domestic solar

water heater systems (≤150o).

166

Chapter Ten

CONCLUSIONS AND FUTURE WORK

In the exploration stage, manganese-, copper- and nickel-cobalt coatings on highly

reflective aluminium substrates were synthesised and characterised. Even though all coatings

demonstrated relatively weak crystallinity at the annealing temperature of 500°C, the XPS

and EDX analyses corroborated the existence of metal-oxide bonding structures within the

coatings. The copper cobalt oxide coatings exhibited the more distinctive morphological

(nano-sized, grain-like particles) features and better optical properties compared to the

manganese- and nickel- cobalt oxides coatings. The copper–cobalt coatings seemed to

provide good prospects for future application as a solar absorber coating material, and so

were chosen for more detailed analysis.

The optical properties of the copper cobalt oxide coatings showed that the optimised

solar absorptance value of 83.4% could be achieved with an average film thickness of around

~320 nm, synthesised using 0.25 M of copper acetate and 0.25 M cobalt chloride precursors

([Cu]/[Co]=1), with a withdrawal rate of 120 mm/min by using four dip-drying cycles, and an

annealing temperature of 500°C. Higher absorptance values could be accomplished by a thin

film with [Cu]/[Co] of 0.5; however, its reflectance spectra curve was less satisfactory in

terms of a good selectivity curve profile. Surface composition analysis showed that oxygen

exists as lattice, surface and subsurface oxygen; the copper consists of octahedral and

tetrahedral Cu+, together with octahedral and paramagnetic Cu

2+ oxidation states, and; the

cobalt consists of tetrahedral and paramagnetic Co2+

, octahedral Co3+

as well as mixed Co2+,3+

oxidation states.

167

Changes in [Cu]/[Co] ratios in the synthesis process have a direct influence on the

surface morphology and composition as well as the mechanical properties. The surfaces

produced with the [Cu]/[Co] ratio of 0.5 and 1 were typically composed of granular

nanoparticles, while the surface produced with the [Cu]/[Co] ratio of 2 had a smoother

surface. XPS analyses showed that the electronic structure of the coatings did not change

much except for the coating with [Cu]/[Co] = 2, which did not indicate the presence of

octahedral Cu+. The increase of copper concentration in the synthesis process tended to

promote the formation of octahedral Cu2+

which minimised the formation of octahedral Cu+

as well as increased the competitiveness of octahedral Cu2+

ions to substitute the Co2+

site in

cobalt structure host. NEXAFS spectra revealed that the local environments of Co, Cu and O

were not significantly influenced by the change in the copper to cobalt concentration ratios

except for the [Cu]/[Co] = 2 sample, where the local coordination appeared to slightly change

due to the loss of octahedral Cu+. Compared to the aluminium substrate, the present coatings

have significantly improved wear resistance, particularly for the [Cu]/[Co] = 1.0 sample

which showed the highest H/E value (0.055). FEM modeling indicated that, under spherical

loading conditions, the higher stress and the plastic deformation were primarily concentrated

within the coating layer, without spreading further into the substrate. This would reduce the

probability of delamination of the coating layer during the unloading phase.

The use of higher annealing temperatures of up to 650 °C improved the crystalline

structure of the copper cobalt oxide coatings. The XRD pattern of coatings indicated the

mineralogical forms of CoCu2O3 and could be CuCoO2 and CoCuO2. The chemistry binding

structures in the surface characterised by XPS remained relatively unchanged with the change

in annealing temperature. However, the cooling method employed could have had an

influence on the surface composition. Optical properties characterised by UV-Vis-NIR

revealed that the increase of annealing temperature until 550oC increased the absorptance

168

with a maximum absorptance value of α = 84.4%, while further increases of temperature

decreased the absorptance values. This fluctuation was caused by a combination of factors

relating to the intrinsic properties of coating material and the substrate surface optical

properties. Mechanical properties probed by nanoindentation tests revealed that both the

elastic modulus and the hardness showed an increasing trend, while the H/E ratio showed a

slight decrease as the annealing temperature was increased. However, using H/E as an

indicator, the wear resistance of all these coating materials was expected to be superior to that

of the aluminium substrate material. Similar to the temperature of 500oC, the FEM modelling

results of samples synthesised at higher annealing temperatures also indicated that, under

mechanical loading conditions, the higher stress and the plastic deformation were primarily

concentrated within the coating layers. This would reduce the likelihood of delamination of

the coating layer upon unloading.

To increase the absorptance of the absorber material and protect it from any

degradation due to external factors, a silica anti-reflection (AR) layer was fabricated on top of

the copper cobalt oxide coatings. The addition of silica changes the reflectance spectra of the

coatings in the wavelengths range of 0.3-15.4 µm. The absorptance values increase slightly

compared to the coating without the silica layer, but the increase of the withdrawal rate of

silica in the synthesis process unfortunately also increased the emittance values due to the

strong phonon absorption by the silica, leading the optimum absorptance of α = 84.96% and

emittance of ε=5.63% (or selectivity s=15.1) showed by the coating with silica AR layer

synthesised at a withdrawal rate of 10 mm/min. Both the PC values results and the physical

inspection of the surfaces showed that the copper cobalt oxide thin film coated with silica AR

layer passed the accelerated thermal durability test without any cracking detected. The

degradation of the copper cobalt oxide thin film coating with the silica AR layer was more

governed by the temperature changes regime than the exposure time regime, indicating that

169

the coating is well suited for long periods of operation in the relatively low temperature range

experienced in applications such as for domestic solar water heating (≤150oC).

Overall, the sol-gel dip-coating synthesised copper cobalt oxide thin film coatings,

which can be produced using relatively simple, low-cost and environmentally friendly

processes, exhibited high absorptance in the UV-Vis-NIR range and low emittance (or high

reflectance) in the mid – far infrared range, together with good mechanical properties. All

these attributes render the coatings promising as a solar selective absorber for applications in

the solar energy industry. However, the fundamental challenge faced by the CuxCoyOz is still

low in the absorptance value. There are some solutions to overcome this challenge, namely

modifying the absorber layer by addition another absorber layer, the addition a proper

antireflection layer on the top of the SSA coating, or changing the synthesis route of

CuxCoyOz. Future research may require the development of a more appropriate additional

absorber layer, appropriate anti-reflective layer and changing the synthesis route of CuxCoyOz

to maximise absorptance, minimise emissivity and hence increase the selectivity of coatings

stack.

170

REFERENCES

[1] S. A. Kalogirou, "Solar thermal collectors and applications," Progress in Energy and

Combustion Science, vol. 30, pp. 231-295, 2004. [2] C. E. Kennedy, "Review of mid- to high-temperature solar selective absorber materials,"

National Renewable Energy Laboratory, Golden, Colorado, USA2002. [3] E. Barrera, T. Viveros, A. Montoya, and M. Ruiz, "Titanium–tin oxide protective films on a

black cobalt photothermal absorber," Solar Energy Materials and Solar Cells, vol. 57, pp. 127-140, 1999.

[4] H. C. Barshilia, N. Selvakumar, K. S. Rajam, D. V. Sridhara Rao, and K. Muraleedharan, "Deposition and characterization of TiAlN/TiAlON/Si3N4 tandem absorbers prepared using reactive direct current magnetron sputtering," Thin Solid Films, vol. 516, pp. 6071-6078, 2008.

[5] P. Oelhafen and A. Schüler, "Nanostructured materials for solar energy conversion," Solar Energy, vol. 79, pp. 110-121, 2005.

[6] G. Katumba, L. Olumekor, A. Forbes, G. Makiwa, B. Mwakikunga, J. Lu, and E. Wäckelgård, "Optical, thermal and structural characteristics of carbon nanoparticles embedded in ZnO and NiO as selective solar absorbers," Solar Energy Materials and Solar Cells, vol. 92, pp. 1285-1292, 2008.

[7] P. Konttinen, P. D. Lund, and R. J. Kilpi, "Mechanically manufactured selective solar absorber surfaces," Solar Energy Materials and Solar Cells, vol. 79, pp. 273-283, 2003.

[8] E. Ienei, L. Isac, C. Cazan, and A. Duta, "Characterization of Al/Al2O3/NiOx solar absorber obtained by spray pyrolysis," Solid State Sciences, vol. 12, pp. 1894-1897, 2010.

[9] Z. Crnjak Orel, M. Klanjšek Gunde, A. Lenček, and N. Benz, "The preparation and testing of spectrally selective paints on different substrates for solar absorbers," Solar Energy, vol. 69, Supplement 6, pp. 131-135, 2000.

[10] H. Tabor, "Selective radiation. I. Wavelength discrimination: A new approach to the harnessing of solar energy," Bull. Res. Council Isr., vol. 5A, pp. 119-128, 1956.

[11] H. Tabor, "Selective radiation. II. Wavefront discrimination: A new approach to the harnessing of solar energy," Bull. Res. Council Isr., vol. 5A, pp. 129-134, 1956.

[12] H. Tabor, Transactions of the Conference on the Use of Solar Energy, vol. Vol 11, pp. Section A1-23, University of Arizona Press, Tucson., 1956.

[13] M. Zemanová, M. Chovancová, Z. Gáliková, and P. Krivošík, "Nickel electrolytic colouring of anodic alumina for selective solar absorbing films," Renewable Energy, vol. 33, pp. 2303-2310, 2008.

[14] A. Wazwaz, J. Salmi, and R. Bes, "The effects of nickel-pigmented aluminium oxide selective coating over aluminium alloy on the optical properties and thermal efficiency of the selective absorber prepared by alternate and reverse periodic plating technique," Energy Conversion and Management, vol. 51, pp. 1679-1683, 2010.

[15] M. G. Hutchins, P. J. Wright, and P. D. Grebenik, "Comparison of different forms of black cobalt selective solar absorber surfaces," Solar Energy Materials, vol. 16, pp. 113-131, 1987.

[16] X. Xiao, L. Miao, G. Xu, L. Lu, Z. Su, N. Wang, and S. Tanemura, "A facile process to prepare copper oxide thin films as solar selective absorbers," Applied Surface Science, vol. 257, pp. 10729-10736, 2011.

[17] N. Selvakumar, H. C. Barshilia, K. S. Rajam, and A. Biswas, "Structure, optical properties and thermal stability of pulsed sputter deposited high temperature HfOx/Mo/HfO2 solar selective absorbers," Solar Energy Materials and Solar Cells, vol. 94, pp. 1412-1420, 2010.

[18] E. Wäckelgård and G. Hultmark, "Industrially sputtered solar absorber surface," Solar Energy Materials and Solar Cells, vol. 54, pp. 165-170, 1998.

171

[19] M. Toshiro, "Copper oxide thin films prepared by chemical vapor deposition from copper dipivaloylmethanate," Solar Energy Materials and Solar Cells, vol. 56, pp. 85-92, 1998.

[20] L. Kaluza, B. Orel, G. Drazic, and M. Kohl, "Sol-gel derived CuCoMnOx spinel coatings for solar absorbers: Structural and optical properties," Solar Energy Materials and Solar Cells, vol. 70, pp. 187-201, 2001.

[21] E. Barrera, A. Avila, J. Mena, V. H. Lara, M. Ruiz, and J. Méndez-Vivar, "Synthesis of cobalt–silicon oxide thin films," Solar Energy Materials and Solar Cells, vol. 76, pp. 387-398, 2003.

[22] T. Bostrom, E. Wackelgard, and G. Westin, "Solution-chemical derived nickel-alumina coatings for thermal solar absorbers," Solar Energy, vol. 74, pp. 497-503, 2003.

[23] B. Japelj, A. Š. Vuk, B. Orel, L. S. Perše, I. Jerman, and J. Kovač, "Preparation of a TiMEMO nanocomposite by the sol–gel method and its application in coloured thickness insensitive spectrally selective (TISS) coatings," Solar Energy Materials and Solar Cells, vol. 92, pp. 1149-1161, 2008.

[24] R. Bayón, G. San Vicente, C. Maffiotte, and Á. Morales, "Preparation of selective absorbers based on CuMn spinels by dip-coating method," Renewable Energy, vol. 33, pp. 348-353, 2008.

[25] R.-C. Juang, Y.-C. Yeh, B.-H. Chang, W.-C. Chen, and T.-W. Chung, "Preparation of solar selective absorbing coatings by magnetron sputtering from a single stainless steel target," Thin Solid Films, vol. 518, pp. 5501-5504, 2010.

[26] M. Adsten, R. Joerger, K. Järrendahl, and E. Wäckelgård, "Optical characterization of industrially sputtered nickel-nickel oxide solar selective surface," Solar Energy, vol. 68, pp. 325-328, 2000.

[27] S. Zhao, C.-G. Ribbing, and E. Wäckelgård, "Optical constants of sputtered Ni/NiO solar absorber film--depth-profiled characterization," Solar Energy Materials and Solar Cells, vol. 84, pp. 193-203, 2004.

[28] S. Zhao and E. Wäckelgård, "Optimization of solar absorbing three-layer coatings," Solar Energy Materials and Solar Cells, vol. 90, pp. 243-261, 2006.

[29] J. Vince, A. Šurca Vuk, U. O. Krašovec, B. Orel, M. Köhl, and M. Heck, "Solar absorber coatings based on CoCuMnOx spinels prepared via the sol–gel process: structural and optical properties," Solar Energy Materials and Solar Cells, vol. 79, pp. 313-330, 2003.

[30] S. Zhao and E. Wäckelgård, "The optical properties of sputtered composite of Al-AlN," Solar Energy Materials and Solar Cells, vol. 90, pp. 1861-1874, 2006.

[31] Y. Yin, Y. Pan, L. X. Hang, D. R. McKenzie, and M. M. M. Bilek, "Direct current reactive sputtering Cr-Cr2O3 cermet solar selective surfaces for solar hot water applications," Thin Solid Films, vol. 517, pp. 1601-1606, 2009.

[32] M. Nejati, "Cermet Based Solar Selective Absorbers: Further Selectivity Improvement and Developing New Fabrication Technique," Saarländische Universitäts und Landesbibliothek, 2008.

[33] O. A. Harizanov, K. A. Gesheva, and P. L. Stefchev, "Sol-gel and CVD-metal oxide coatings for solar energy utilization," Ceramics International, vol. 22, pp. 91-94, 1996.

[34] M. Voinea, C. Bogatu, G. C. Chitanu, and A. Duta, "Copper cermets used as selective coatings for flat plate solar collectors," Rev. Chim. (Bucuresti), vol. 59, pp. 659-663, 2008.

[35] S. S. Kanu and R. Binions, "Thin films for solar control applications," Proceedings of the Royal Society A: Mathematical, Physical and Engineering Science, vol. 466, pp. 19-44, January 8, 2010 2010.

[36] P. Konttinen and P. D. Lund, "Microstructural optimization and extended durability studies of low-cost rough graphite-aluminium solar absorber surfaces," Renewable Energy, vol. 29, pp. 823-839, 2004.

172

[37] T. Boström, S. Valizadeh, J. Lu, J. Jensen, G. Westin, and E. Wäckelgård, "Structure and morphology of nickel-alumina/silica solar thermal selective absorbers," Journal of Non-Crystalline Solids, vol. 357, pp. 1370-1375, 2011.

[38] R. Bayon, G. San Vicente, and A. Morales, "Durability tests and up-scaling of selective absorbers based on copper-manganese oxide deposited by dip-coating," Solar Energy Materials and Solar Cells, vol. 94, pp. 998-1004, 2010.

[39] T. Boström, G. Westin, and E. Wäckelgård, "Optimization of a solution-chemically derived solar absorbing spectrally selective surface," Solar Energy Materials and Solar Cells, vol. 91, pp. 38-43, 2007.

[40] E. Barrera, L. Huerta, S. Muhl, and A. Avila, "Synthesis of black cobalt and tin oxide films by the sol-gel process: surface and optical properties," Solar Energy Materials and Solar Cells, vol. 88, pp. 179-186, 2005.

[41] G. Katumba, J. Lu, L. Olumekor, G. Westin, and E. Wäckelgård, "Low cost selective solar absorber coatings: characteristics of carbon-in-silica synthesized with sol-gel technique," Journal of Sol-Gel Science and Technology, vol. 36, pp. 33-43, 2005.

[42] M. E. Rincón, J. D. Molina, M. Sánchez, C. Arancibia, and E. García, "Optical characterization of tandem absorber/reflector systems based on titanium oxide-carbon coatings," Solar Energy Materials and Solar Cells, vol. 91, pp. 1421-1425, 2007.

[43] C. J. Brinker, G. C. Frye, A. J. Hurd, and C. S. Ashley, "Fundamentals of sol-gel dip coating," Thin Solid Films, vol. 201, pp. 97-108, 1991.

[44] J. A. Duffie and W. A. Beckman, Solar Engineering of Thermal Processes, third ed. New Jersey: John Wiley & Sons Inc., 2006.

[45] L. Rayleigh, "On the light from the sky, its polarization and colour," Philos. Mag., vol. 41, pp. 107-120, 274-279, 1971.

[46] M. Iqbal, An introduction to solar radiation: Academic Press, Ontario, Canada, 1983. [47] ISO-9845-1, "Solar Energy. Reference solar spectral irradiance at the ground at different

receiving conditions. Part 1: Direct normal and hemispherical solar irradiance for air mass 1.5," 1992.

[48] M. A. Lind, R. B. Pettit, and K. D. Masterson, "The Sensitivity of Solar Transmittance Reflectance and Absorptance to Selected Averaging Procedures and Solar Irradiance Distributions," 1980.

[49] F. K. Richtmyer and E. H. Kennard, Introduction to Modern Physics vol. 4th ed. New York: McGraw-Hill 1947.

[50] O. S. Heavens, Optical properties of thin solid films. New York: Dover Publications, 1991. [51] T. Boström, "Solution-Chemically Derived Spectrally Selective Solar Absorbers: With System

Perspectives on Solar Heating," Acta Universitatis Upsaliensis, Uppsala, 2006. [52] O. P. Agnihotri and B. K. Gupta, Solar Selective Surfaces. New York: Wiley-Interscience, 1981. [53] C. M. Lampert, "Coatings for enhanced photothermal energy collection I. Selective

absorbers," Solar Energy Materials, vol. 1, pp. 319-341, 1979. [54] M. M. Koltun, Selective Optical Surfaces for Solar Energy Converters New York: Allerton

Press, Inc., 1979. [55] C. G. Granqvist, Materials Science for Solar Energy Conversion Systems Oxford: Pergamon

Press, 1991. [56] W. Bogaerts and C. Lampert, "Materials for photothermal solar energy conversion," Journal

of Materials Science, vol. 18, pp. 2847-2875, 1983/10/01 1983. [57] L. E. Murr, Solar Materials Science. New York: Academic Press, Inc., 1980. [58] J. Gordon, "Solar Energy The state of the art - ISES position papers," James & James,

London2001. [59] B. O. Seraphin, Ed., Topics in Applied Physics. Berlin, : Springer-Verlag, 1979, p.^pp. Pages.

173

[60] C. M. Lampert, "Heat mirror coatings for energy conserving windows," Solar Energy Materials, vol. 6, pp. 1-41, 1981.

[61] R. A. Buhrman, Physics of solar selective surfaces vol. 3. New York: ASES Plenum Press, 1986. [62] T. Tesfamichael, "Characterization of selective solar absorbers: Experimental and theoretical

modeling.," Doctoral Thesis, Faculty of Science and Technology, Uppsala University, Sweden, 2000.

[63] B. O. Seraphin, "Chemical vapor deposition of thin semiconductor films for solar energy conversion," Thin Solid Films, vol. 39, pp. 87-94, 1976.

[64] T. K. Bostrom, E. Wackelgard, and G. Westin, "Durability tests of solution-chemically derived spectrally selective absorbers," Solar Energy Materials and Solar Cells, vol. 89, pp. 197-207, 2005.

[65] M. Farooq and Z. H. Lee, "Computations of the optical properties of metal/insulator-composites for solar selective absorbers," Renewable Energy, vol. 28, pp. 1421-1431, 2003.

[66] G. A. Niklasson, C. G. Granqvist, and O. Hunderi, "Effective medium models for the optical properties of inhomogeneous materials," Appl. Opt., vol. 20, pp. 26-30, 1981.

[67] G. Niklasson, "Optical properties and solar selectivity of coevaporated Co-Al2O3 composite films," J. Appl. Phys., vol. 55, p. 3382, 1984.

[68] M. R. Nejati, V. Fathollahi, and M. Khalaji Asadi, "Computer simulation of the optical properties of high-temperature cermet solar selective coatings," Solar Energy, vol. 78, pp. 235-241, 2005.

[69] A. Andersson, O. Hunderi, and C. G. Granqvist, "Nickel pigmented anodic aluminum oxide for selective absorption of solar energy," Journal of Applied Physics, vol. 51, pp. 754-764, 1980.

[70] E. Wäckelgård, T. Chibuye, and B. Karlsson, Improved Solar Optical Properties of a Nickel Pigmented Anodized Aluminum Selective Surface: Pergamon Press, Oxford, UK, 1990.

[71] L. Kaluza, A. Šurca-Vuk, B. Orel, G. Dražič, and P. Pelicon, "Structural and IR spectroscopic analysis of sol-gel processed CuFeMnO4 spinel and CuFeMnO4/silica films for solar absorbers," Journal of Sol-Gel Science and Technology, vol. 20, pp. 61-83, 2001.

[72] G. A. Niklasson and C. G. Granqvist, "Surfaces for selective absorption of solar energy: an annotated bibliography 1955–1981," Journal of Materials Science, vol. 18, pp. 3475-3534, 1983.

[73] Y. M. Lu, W. S. Hwang, J. S. Yang, and H. C. Chuang, "Properties of nickel oxide thin films deposited by RF reactive magnetron sputtering," Thin Solid Films, vol. 420-421, pp. 54-61, 2002.

[74] Z. He, Z. Ji, S. Zhao, C. Wang, K. Liu, and Z. Ye, "Characterization and electrochromic properties of CuxNi1-xO films prepared by sol-gel dip-coating," Solar Energy, vol. 80, pp. 226-230, 2006.

[75] M. Voinea, C. Vladuta, C. Bogatu, and A. Duta, "Surface properties of copper based cermet materials," Materials Science and Engineering: B, vol. 152, pp. 76-80, 2008.

[76] P. Konttinen, "Characterization and Aging Studies of Selective Solar C/Al2O3/Al Absorber Surfaces," Doctoral thesis, Department of Engineering Physics and Mathematics, Helsinki University of Technology Espoo, Finland, 2004.

[77] Z. Crnjak Orel, "Characterisation of high-temperature-resistant spectrally selective paints for solar absorbers," Solar Energy Materials and Solar Cells, vol. 57, pp. 291-301, 1999.

[78] R. Bayón, G. San Vicente, C. Maffiotte, and Á. Morales, "Characterization of copper-manganese-oxide thin films deposited by dip-coating," Solar Energy Materials and Solar Cells, vol. 92, pp. 1211-1216, 2008.

[79] D. Chen, "Anti-reflection (AR) coatings made by sol–gel processes: A review," Solar Energy Materials and Solar Cells, vol. 68, pp. 313-336, 2001.

[80] H. Haitjema, "Spectrally selective tinoxide and indiumoxide coatings," Doctoral, Technische Universiteit Delft, The Netherlands, 1989.

174

[81] R. A. Buhrman and H. G. Craighead, Solar Materials Science. New York: Academic Press, Inc., 1980.

[82] G. A. Niklasson, Chalmers Univ. Tech., Gothenburg, 1982. [83] B. Perers, "Dynamic method for solar collector array testing and evaluation with standard

database and simulation programs," Solar Energy, vol. 50, pp. 517-526, 1993. [84] B. Perers, "Optical modelling of solar collectors and booster reflectors under non stationary

conditions," Doctoral, Uppsala University, Sweden, 1995. [85] A. Leyland and A. Matthews, "On the significance of the H/E ratio in wear control: a

nanocomposite coating approach to optimised tribological behaviour," Wear, vol. 246, pp. 1-11, 2000.

[86] W. C. Oliver and G. M. Pharr, "Improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments," Journal of Materials Research, vol. 7, pp. 1564-1583, 1992.

[87] W. C. Oliver and G. M. Pharr, "Measurement of hardness and elastic modulus by instrumented indentation: Advances in understanding and refinements to methodology," Journal of Materials Research, vol. 19, pp. 3-20, 2004.

[88] B. Carlsson, U. Frei, M. Kohl, and K. Moller, "Accelerated life testing of solar energy materials - Case study of some selective solar absorber coating materials for DHW Systems," Borås, Sweden1994.

[89] C. M. Lampert, "Failure and degradation modes in selected solar materials: A review," Lawrence Berkeley Lab., CA (USA)1989.

[90] B. Carlsson, "Recommended qualification test procedure for solar absorber surface durability," Borås, Swedish2004.

[91] B. Carlsson, U. Frei, M. Kohl, K. Moller, and S. Brunold, "Qualification test procedure for solar absorber surface durability," Solar Energy Materials and Solar Cells, vol. 61, pp. 255-275, 2000.

[92] T. K. Bostrom, E. Wackelgard, and G. Westin, "Anti-reflection coatings for solution-chemically derived nickel-alumina solar absorbers," Solar Energy Materials and Solar Cells, vol. 84, pp. 183-191, 2004.

[93] E. Barrera-Calva, J. Mendez-Vivar, M. Ortega-Lopez, L. Huerta-Arcos, J. Morales-Corona, and R. Olayo-Gonzalez, "Silica-copper oxide composite thin films as solar selective coatings prepared by dipping sol-gel," Research Letters in Materials Science, vol. 2008, pp. 1-5, 2008.

[94] T. Karlsson and A. Roos, "Optical properties and spectral selectivity of copper oxide on stainless steel," Solar Energy Materials, vol. 10, pp. 105-119, 1984.

[95] H. C. Hottel and T. A. Unger, "The properties of a copper oxide - aluminum selective black surface absorber of solar energy," Solar Energy, vol. 3, pp. 10-15, 1959.

[96] A. Márquez, G. Blanco, M. E. Fernandez de Rapp, D. G. Lamas, and R. Tarulla, "Properties of cupric oxide coatings prepared by cathodic arc deposition," Surface and Coatings Technology, vol. 187, pp. 154-160, 2004.

[97] A. Scherer, O. T. Inal, and A. J. Singh, "Investigation of copper oxide coatings for solar selective applications," Solar Energy Materials, vol. 9, pp. 139-158, 1983.

[98] A. Scherer, O. T. Inal, and R. B. Pettit, "Modelling of degradation in black copper photothermal collector coatings," Journal of Materials Science, vol. 23, pp. 1923-1933, 1988.

[99] T. S. Sathiaraj, "Solar selective properties of copper-aluminium composite films," Indian Journal of Pure & Applied Physics, vol. 45, pp. 613-617, 2007.

[100] R. B. Gillette, "Selectively emissive materials for solar heat absorbers," Solar Energy, vol. 4, pp. 24-32, 1960.

[101] K. J. Cathro, "Preparation of cobalt-oxide-based selective surfaces by a dip-coating process," Solar Energy Materials, vol. 9, pp. 433-447, 1984.

175

[102] E. C. Barrera, T. G. Viveros, and U. Morales, "Preparation of selective surfaces of black cobalt by the sol-gel process," Renewable Energy, vol. 9, pp. 733-736, 1996.

[103] T. Maruyama and T. Nakai, "Cobalt oxide thin films prepared by chemical vapor deposition from cobalt (II) acetate," Solar Energy Materials, vol. 23, pp. 25-29, 1991.

[104] C. Choudhury and H. K. Sehgal, "Black cobalt selective coatings by spray pyrolysis for photothermal conversion of solar energy," Solar Energy, vol. 28, pp. 25-31, 1982.

[105] C. Choudhury and H. K. Sehgal, "High temperature degradation in cobalt oxide selective absorber," Solar Energy, vol. 30, pp. 291-292, 1983.

[106] K. Chidambaram, L. K. Malhotra, and K. L. Chopra, "Spray-pyrolysed cobalt black as a high temperature selective absorber," Thin Solid Films, vol. 87, pp. 365-371, 1982.

[107] A. Avila G, E. Barrera C, L. Huerta A, and S. Muhl, "Cobalt oxide films for solar selective surfaces, obtained by spray pyrolisis," Solar Energy Materials and Solar Cells, vol. 82, pp. 269-278, 2004.

[108] C. Uma, L. Malhotra, and K. Chopra, "Cobalt oxide-iron oxide selective coatings for high temperature applications," Bulletin of Materials Science, vol. 8, pp. 385-389, 1986.

[109] E. Barrera, T. Viveros, A. Avila, P. Quintana, M. Morales, and N. Batina, "Cobalt oxide films grown by a dipping sol-gel process," Thin Solid Films, vol. 346, pp. 138-144, 1999.

[110] R. Drasovean, R. Monteiro, E. Fortunato, and V. Musat, "Optical properties of cobalt oxide films by a dipping sol–gel process," Journal of Non-Crystalline Solids, vol. 352, pp. 1479-1485, 2006.

[111] D. J. Goyal, C. M. Agashe, B. R. Marathe, M. G. Takwale, and V. G. Bhide, "Effect of precursor solution concentration on the structural properties of sprayed ZnO films," Journal of Materials Science Letters, vol. 11, pp. 708-710, 1992.

[112] S. John, N. Nagarani, and S. Rajendran, "Black cobalt solar absorber coatings," Solar Energy Materials, vol. 22, pp. 293-302, 1991.

[113] E. Barrera, I. González, and T. Viveros, "A new cobalt oxide electrodeposit bath for solar absorbers," Solar Energy Materials and Solar Cells, vol. 51, pp. 69-82, 1998.

[114] F. Kadirgan and M. Sohmen, "Development of black cobalt selective absorber on copper for solar collectors," Turk J Chem, vol. 23, pp. 345-351, 1999.

[115] H. J. Brown-Shaklee, W. Carty, and D. D. Edwards, "Spectral selectivity of composite enamel coatings on 321 stainless steel," Solar Energy Materials and Solar Cells, vol. 93, pp. 1404-1410, 2009.

[116] Z. Crnjak Orel, B. Orel, and M. Klanjšek Gunde, "Spectrally selective SnO2: film on glass and black enamelled steel substrates: spray pyrolytical deposition and optical properties," Solar Energy Materials and Solar Cells, vol. 26, pp. 105-116, 1992.

[117] F. Simonis, A. J. Faber, and C. J. Hoogendoorn, "Porcelain enamelled absorbers, coated by spectral selective tin oxide," Journal of Solar Energy Engineering, vol. 109, pp. 22-25, 1987.

[118] A. P. Rizzato, L. Broussous, C. V. Santilli, S. H. Pulcinelli, and A. F. Craievich, "Structure of SnO2 alcosols and films prepared by sol–gel dip coating," Journal of Non-Crystalline Solids, vol. 284, pp. 61-67, 2001.

[119] C. E. Barrera, G. A. Avila, S. Rodil, and L. Huerta, "Cobalt-copper oxide compound for selective solar absorber," presented at the World Renewable Energy Congress (WREC) VIII, 2004.

[120] U. Morales-Ortiz, A. Avila-García, and V. Hugo Lara C, "Ruthenium oxide films for selective coatings," Solar Energy Materials and Solar Cells, vol. 90, pp. 832-840, 2006.

[121] W. Qian, C. Xu-dong, W. Hui, and M. Tao, "Study on preparation and properties of NiO-Al2O3 solar selective absorption films," Surface Technology, vol. 40, p. 1, 2011.

[122] T. Eisenhammer, H. Schellinger, and M. Lazarov, "Process for producing selective absorbers," United States Patent, 1999.

176

[123] T. K. Boström and E. Wäckelgård, "Optical properties of solution-chemically derived thin film Ni–Al2O3 composites and Si, Al and Si–Ti oxides," Journal of Physics: Condensed Matter, vol. 18, p. 7737, 2006.

[124] T. Boström, J. Jensen, S. Valizadeh, G. Westin, and E. Wäckelgård, "ERDA of Ni-Al2O3/SiO2 solar thermal selective absorbers," Solar Energy Materials and Solar Cells, vol. 92, pp. 1177-1182, 2008.

[125] N. Etherden, T. Tesfamichael, G. A. Niklasson, and E. Wackelgard, "A theoretical feasibility study of pigments for thickness-sensitive spectrally selective paints," Journal of Physics D: Applied Physics, vol. 37, pp. 1115-1122, 2004.

[126] Y. Mastai, S. Polarz, and M. Antonietti, "Silica–carbon nanocomposites—A new concept for the design of solar absorbers," Advanced Functional Materials, vol. 12, pp. 197-202, 2002.

[127] D. Katzen, E. Levy, and Y. Mastai, "Thin films of silica-carbon nanocomposites for selective solar absorbers," Applied Surface Science, vol. 248, pp. 514-517, 2005.

[128] Z. Liu, Z. Jin, W. Li, and J. Qiu, "Preparation of ZnO porous thin films by sol–gel method using PEG template," Materials Letters, vol. 59, pp. 3620-3625, 2005.

[129] G. Katumba, B. Mwakikunga, and T. Mothibinyane, "FTIR and Raman spectroscopy of carbon nanoparticles in SiO2 , ZnO and NiO matrices," Nanoscale Research Letters, vol. 3, pp. 421 - 426, 2008.

[130] R. Bayón, G. S. Vicente, C. Maffiotte, and Á. Morales, "Development of solar absorbers based on spinel-type materials deposited by dip-coating," presented at the Solar PACES, 2006.

[131] Q.-F. Geng, X. Zhao, X.-H. Gao, and G. Liu, "Sol–gel combustion-derived CoCuMnOx spinels as pigment for spectrally selective paints," Journal of the American Ceramic Society, vol. 94, pp. 827-832, 2011.

[132] M. Kozelj, A. S. Vuk, I. Jerman, and B. Orel, "Corrosion protection of Sunselect, a spectrally selective solar absorber coating, by (3-mercaptopropyl)trimethoxysilane," Solar Energy Materials and Solar Cells, vol. 93, pp. 1733-1742, 2009.

[133] B. D. Cullity and S. R. Stock, Elements of X-ray Diffraction vol. 3: Prentice hall Upper Saddle River, NJ, 2001.

[134] B. B. He, Two-dimensional X-ray Diffraction: Wiley, 2011. [135] R. F. Egerton, Physical principles of electron microscopy: an introduction to TEM, SEM, and

AEM: Springer, 2005. [136] R. F. Egerton, Physical principles of electron microscopy: Springer, 2005. [137] H. Schatten and J. B. Pawley, Biological low voltage field emission scanning electron

microscopy: Springer, 2008. [138] G. Binnig, C. F. Quate, and C. Gerber, "Atomic force microscope," Physical Review Letters,

vol. 56, pp. 930-933, 1986. [139] Y. Jiabao, "Structure and magnetic properties of Ni/NiO, Co/CoO composite films," Doctoral

thesis, Department of Materials Science, National University of Singapore, Singapore, 2007. [140] J. F. Watts, J. Wolstenholme, and J. Wiley, An introduction to surface analysis by XPS and

AES: Wiley Online Library, 2003. [141] K. Laajalehto, I. Kartio, and E. Suoninen, "XPS and SR-XPS techniques applied to sulphide

mineral surfaces," International Journal of Mineral Processing, vol. 51, pp. 163-170, 1997. [142] T. M. Rosseel, T. A. Carlson, R. E. Negri, C. E. Beall, and J. W. Taylor, Synchrotron radiation as

a source for quantitative XPS: advantages and consequences, 1986. [143] H. Jensen, A. Soloviev, Z. Li, and E. G. Søgaard, "XPS and FTIR investigation of the surface

properties of different prepared titania nano-powders," Applied Surface Science, vol. 246, pp. 239-249, 2005.

[144] G. Hahner, "Near edge X-ray absorption fine structure spectroscopy as a tool to probe electronic and structural properties of thin organic films and liquids," Chemical Society Reviews, vol. 35, pp. 1244-1255, 2006.

177

[145] A. Roos, "Use of an integrating sphere in solar energy research," Solar Energy Materials and Solar Cells, vol. 30, pp. 77-94, 1993.

[146] J. L. Gautier, E. Trollund, E. Ríos, P. Nkeng, and G. Poillerat, "Characterization of thin CuCo2O4 films prepared by chemical spray pyrolysis. Study of their electrochemical stability by ex situ spectroscopic analysis," Journal of Electroanalytical Chemistry, vol. 428, pp. 47-56, 1997.

[147] B. Cui, H. Lin, J.-B. Li, X. Li, J. Yang, and J. Tao, "Core–Ring Structured NiCo2O4 Nanoplatelets: Synthesis, Characterization, and Electrocatalytic Applications," Advanced Functional Materials, vol. 18, pp. 1440-1447, 2008.

[148] M. A. Carreon, V. V. Guliants, L. Yuan, A. R. Hughett, A. Dozier, G. A. Seisenbaeva, and V. G. Kessler, "Mesoporous Nanocrystalline Mixed Metal Oxides from Heterometallic Alkoxide Precursors: Cobalt–Nickel Oxide Spinels for Propane Oxidation," European Journal of Inorganic Chemistry, vol. 2006, pp. 4983-4988, 2006.

[149] T. Nissinen, M. Leskela, M. Gasik, and J. Lamminen, "Decomposition of mixed Mn and Co nitrates supported on carbon," Thermochimica Acta, vol. 427, pp. 155-161, 2005.

[150] J. L. Gautier, E. Rios, M. Gracia, J. F. Marco, and J. R. Gancedo, "Characterisation by X-ray photoelectron spectroscopy of thin MnxCo3-xO4 (1 > x > 0) spinel films prepared by low-temperature spray pyrolysis," Thin Solid Films, vol. 311, pp. 51-57, 1997.

[151] D. Klissurski and E. Uzunova, "Synthesis of nickel cobaltite spinel from coprecipitated nickel-cobalt hydroxide carbonate," Chemistry of Materials, vol. 3, pp. 1060-1063, 2011/12/20 1991.

[152] M. Guene, A. A. Diagne, M. Fall, M. M. Dieng, and G. Poillerat, "Preparation of nickel-cobalt spinel oxides NixCO3-xO4. Comparison of two physical properties stemming from four different preparation methods and using carbon paste electrode," Bulletin of the Chemical Society of Ethiopia, vol. 21, pp. 255-262, 2007.

[153] S. Guillemet-Fritsch, C. Tenailleau, H. Bordeneuve, and A. Rousset, "Magnetic properties of cobalt and manganese oxide spinel ceramics," Advances in Science and Technology, vol. 67, pp. 143-148, 2010.

[154] A. V. Chadwick, S. L. P. Savin, S. Fiddy, R. Alcántara, D. Fernández Lisbona, P. Lavela, G. F. Ortiz, and J. L. Tirado, "Formation and Oxidation of Nanosized Metal Particles by Electrochemical Reaction of Li and Na with NiCo2O4: X-ray Absorption Spectroscopic Study," The Journal of Physical Chemistry C, vol. 111, pp. 4636-4642, 2007/03/01 2007.

[155] R. Alcántara, M. Jaraba, P. Lavela, and J. L. Tirado, "NiCo2O4 Spinel: First Report on a Transition Metal Oxide for the Negative Electrode of Sodium-Ion Batteries," Chemistry of Materials, vol. 14, pp. 2847-2848, 2002/07/01 2002.

[156] Z. Yang, G. Xia, S. P. Simner, and J. W. Stevenson, "Thermal Growth and Performance of Manganese Cobaltite Spinel Protection Layers on Ferritic Stainless Steel SOFC Interconnects," Journal of The Electrochemical Society, vol. 152, pp. A1896-A1901, 2005.

[157] A. Balland, P. Gannon, M. Deibert, S. Chevalier, G. Caboche, and S. Fontana, "Investigation of La2O3 and/or (Co,Mn)3O4 deposits on Crofer22APU for the SOFC interconnect application," Surface and Coatings Technology, vol. 203, pp. 3291-3296, 2009.

[158] T.-Y. Wei, C.-H. Chen, H.-C. Chien, S.-Y. Lu, and C.-C. Hu, "A Cost-Effective Supercapacitor Material of Ultrahigh Specific Capacitances: Spinel Nickel Cobaltite Aerogels from an Epoxide-Driven Sol–Gel Process," Advanced Materials, vol. 22, pp. 347-351, 2010.

[159] E. Vila, R. M. Rojas, J. L. Martin de Vidales, and O. Garcia-Martinez, "Structural and Thermal Properties of the Tetragonal Cobalt Manganese Spinels MnxCo3-xO4 (1.4 < x < 2.0)," Chemistry of Materials, vol. 8, pp. 1078-1083, 2011/12/20 1996.

[160] H. T. Zhang and X. H. Chen, "Size-dependent x-ray photoelectron spectroscopy and complex magnetic properties of CoMn2O4 spinel nanocrystals," Nanotechnology, vol. 17, p. 1384, 2006.

178

[161] H. Bordeneuve, C. Tenailleau, S. Guillemet-Fritsch, R. Smith, E. Suard, and A. Rousset, "Structural variations and cation distributions in Mn3-xCoxO4 (0 ≤ x ≤ 3) dense ceramics using neutron diffraction data," Solid State Sciences, vol. 12, pp. 379-386, 2010.

[162] M. De Koninck, S.-C. Poirier, and B. Marsan, "CuxCo3-xO4 Used as Bifunctional Electrocatalyst," Journal of The Electrochemical Society, vol. 153, pp. A2103-A2110, 2006.

[163] R. N. Singh, J. P. Pandey, N. K. Singh, B. Lal, P. Chartier, and J. F. Koenig, "Sol-gel derived spinel MxCo3-xO4 (M=Ni, Cu; 0≤x≤1) films and oxygen evolution," Electrochimica Acta, vol. 45, pp. 1911-1919, 2000.

[164] M. Hamid, A. A. Tahir, M. Mazhar, K. C. Molloy, and G. Kociok-Köhn, "Copper–cobalt heterobimetallic ceramic oxide thin film deposition: Synthesis, characterization and application of precursor," Inorganic Chemistry Communications, vol. 11, pp. 1159-1161, 2008.

[165] W. M. Shaheen and A. A. Ali, "Characterization of solid–solid interactions and physico-chemical properties of copper–cobalt mixed oxides and CuxCo3-xO4 spinels," Materials Research Bulletin, vol. 36, pp. 1703-1716, 2001.

[166] J. G. Kim, D. L. Pugmire, D. Battaglia, and M. A. Langell, "Analysis of the NiCo2O4 spinel surface with Auger and X-ray photoelectron spectroscopy," Applied Surface Science, vol. 165, pp. 70-84, 2000.

[167] C. F. Windisch Jr, G. J. Exarhos, K. F. Ferris, M. H. Engelhard, and D. C. Stewart, "Infrared transparent spinel films with p-type conductivity," Thin Solid Films, vol. 398-399, pp. 45-52, 2001.

[168] R. Mechiakh, N. B. Sedrine, J. B. Naceur, and R. Chtourou, "Elaboration and characterization of nanocrystalline TiO2 thin films prepared by sol–gel dip-coating," Surface and Coatings Technology, vol. 206, pp. 243-249, 2011.

[169] A. Restovic, E. Rios, S. Barbato, J. Ortiz, and J. L. Gautier, "Oxygen reduction in alkaline medium at thin MnxCo3-xO4 (0 < x < 1) spinel films prepared by spray pyrolysis. Effect of oxide cation composition on the reaction kinetics," Journal of Electroanalytical Chemistry, vol. 522, pp. 141-151, 2002.

[170] M. El Baydi, S. K. Tiwari, R. N. Singh, J.-L. Rehspringer, P. Chartier, J. F. Koenig, and G. Poillerat, "High Specific Surface Area Nickel Mixed Oxide Powders LaNiO3 (Perovskite) and NiCo2O4 (Spinel) via Sol-Gel Type Routes for Oxygen Electrocatalysis in Alkaline Media," Journal of Solid State Chemistry, vol. 116, pp. 157-169, 1995.

[171] B. Marsan, N. Fradette, and G. Beaudoin, "Physicochemical and Electrochemical Properties of CuCo2O4 Electrodes Prepared by Thermal Decomposition for Oxygen Evolution," Journal of The Electrochemical Society, vol. 139, pp. 1889-1896, 1992.

[172] A. La Rosa-Toro, R. Berenguer, C. Quijada, F. Montilla, E. Morallon, and J. L. Vazquez, "Preparation and Characterization of Copper-Doped Cobalt Oxide Electrodes," The Journal of Physical Chemistry B, vol. 110, pp. 24021-24029, 2011/12/19 2006.

[173] K. Petrov and L. Markov, "Preparation of copper-cobalt oxide spinels by thermal decomposition of copper-cobalt basic nitrate mixed crystals," Journal of Materials Science, vol. 20, pp. 1211-1214, 1985.

[174] G. Speranza, L. Minati, and M. Anderle, "The C1s core line in irradiated graphite," Journal of Applied Physics, vol. 102, pp. 043504-7, 2007.

[175] L. M. Ioffe, P. Bosch, T. Viveros, H. Sanchez, and Y. G. Borodko, "Natural manganese oxides as catalysts for oxidative coupling of methane: a structural and degradation study," Materials Chemistry and Physics, vol. 51, pp. 269-275, 1997.

[176] F. Hao, J. Zhong, P.-L. Liu, K.-Y. You, C. Wei, H.-J. Liu, and H.-A. Luo, "Preparation of mesoporous SiO2-Al2O3 supported Co or Mn catalysts and their catalytic properties in cyclohexane nitrosation to É-caprolactam," Journal of Molecular Catalysis A: Chemical, vol. 351, pp. 210-216, 2011.

179

[177] J.-C. Dupin, D. Gonbeau, P. Vinatier, and A. Levasseur, "Systematic XPS studies of metal oxides, hydroxides and peroxides," Physical Chemistry Chemical Physics, vol. 2, pp. 1319-1324, 2000.

[178] S. Royer, A. Van Neste, R. Davidson, S. Mcintyre, and S. Kaliaguine, "Methane oxidation over nanocrystalline LaCo1-xFexO3: Resistance to SO2 poisoning," Industrial & engineering chemistry research, vol. 43, pp. 5670-5680, 2004.

[179] J. F. Moulder, W. F. Stickle, P. E. Sobol, and K. D. Bomben, Handbook of X-ray Photoelectron Spectroscopy: A Reference Book of Standard Spectra for Indentation and Interpretation of XPS Data. Minnesota-USA: Perkin Elmer Corporation, 1992.

[180] D. G. Klissurski and E. L. Uzunova, "Cation-deficient nano-dimensional particle size cobalt-manganese spinel mixed oxides," Applied Surface Science, vol. 214, pp. 370-374, 2003.

[181] A. C. Tavares, M. I. da Silva Pereira, M. H. Mendonça, M. R. Nunes, F. M. Costa, and C. M. Sá, "XPS and voltammetric studies on Ni1-xCuxCo2O4 spinel oxide electrodes," Journal of Electroanalytical Chemistry, vol. 449, pp. 91-100, 1998.

[182] J. Charles F. Windisch, K. F. Ferris, and G. J. Exarhos, "Synthesis and characterization of transparent conducting oxide cobalt–nickel spinel films," Journal of Vacuum Science and Technology A, vol. 19, pp. 1647-1651, 2001.

[183] P. Nkeng, G. Poillerat, J. F. Koenig, P. Chartier, B. Lefez, J. Lopitaux, and M. Lenglet, "Characterization of Spinel-Type Cobalt and Nickel Oxide Thin Films by X-Ray Near Grazing Diffraction, Transmission and Reflectance Spectroscopies, and Cyclic Voltammetry," Journal of The Electrochemical Society, vol. 142, pp. 1777-1783, 1995.

[184] W. Que, Z. Sun, Y. Zhou, Y. L. Lam, S. D. Cheng, Y. C. Chan, and C. H. Kam, "Preparation of hard optical coatings based on an organic/inorganic composite by sol–gel method," Materials Letters, vol. 42, pp. 326-330, 2000.

[185] C. M. Chan, G. Z. Cao, H. Fong, M. Sarikaya, T. Robinson, and L. Nelson, "Nanoindentation and adhesion of sol-gel-derived hard coatings on polyester," Journal of Materials Research, vol. 15, pp. 148-154, 2000.

[186] K. Fujimoto and T. Oba, "Synthesis of C1-C7 alcohols from synthesis gas with supported cobalt catalysts," Applied Catalysis, vol. 13, pp. 289-293, 1985.

[187] J. E. Baker, R. Burch, and S. E. Golunski, "Synthesis of higher alcohols over copper/cobalt catalysts: Influence of preparative procedures on the activity and selectivity of Cu/Co/Zn/Al mixed oxide catalysts," Applied Catalysis, vol. 53, pp. 279-297, 1989.

[188] X. Xiaoding, E. B. M. Doesburg, and J. J. F. Scholten, "Synthesis of higher alcohols from syngas - recently patented catalysts and tentative ideas on the mechanism," Catalysis Today, vol. 2, pp. 125-170, 1987.

[189] S. Angelov, D. Mehandjiev, B. Piperov, V. Zarkov, A. Terlecki-Baricˇevic´, D. Jovanovic´, and Z. Jovanovic´, "Carbon monoxide oxidation on mixed spinels CuXCo3-XO4 (0< × < 1) in the presence of sulphur compounds," Applied Catalysis, vol. 16, pp. 431-437, 1985.

[190] R. Bonchev, T. Zheleva, and S. C. Sevov, "Morphological and compositional characterization of copper-cobalt spinel made by mechanochemical reactions," Chemistry of Materials, vol. 2, pp. 93-95, 1990/03/01 1990.

[191] G. Fornasari, S. Gusi, F. Trifiro, and A. Vaccari, "Cobalt mixed spinels as catalysts for the synthesis of hydrocarbons," Industrial & Engineering Chemistry Research, vol. 26, pp. 1500-1505, 1987/08/01 1987.

[192] G. G. Volkova, T. A. Krieger, L. M. Plyasova, V. A. Zaikovskii, and T. M. Yurieva, "Copper-cobalt catalysts for higher alcohols synthesis from syngas," in Studies in Surface Science and Catalysis. vol. Volume 107, R. L. E. C. P. N. J. H. S. M. de Pontes and M. S. Scurrell, Eds., ed: Elsevier, 1997, pp. 67-72.

180

[193] G. G. Volkova, T. M. Yurieva, L. M. Plyasova, M. I. Naumova, and V. I. Zaikovskii, "Role of the Cuâ€“Co alloy and cobalt carbide in higher alcohol synthesis," Journal of Molecular Catalysis A: Chemical, vol. 158, pp. 389-393, 2000.

[194] D. J. Singh, "Electronic and thermoelectric properties of CuCoO2: Density functional calculations," Physical Review B, vol. 76, p. 085110, 2007.

[195] M. Beekman, J. Salvador, X. Shi, G. S. Nolas, and J. Yang, "Characterization of delafossite-type CuCoO2 prepared by ion exchange," Journal of Alloys and Compounds, vol. 489, pp. 336-338, 2010.

[196] G.-H. Li, L.-Z. Dai, D.-S. Lu, and S.-Y. Peng, "Characterization of copper cobalt mixed oxide," Journal of Solid State Chemistry, vol. 89, pp. 167-173, 1990.

[197] S. Angelov, G. Tyuliev, and T. Marinova, "XPS study of surface composition of polycrystalline CuxCo3−xO4 (0<x<1) obtained by thermal decomposition of nitrate mixtures," Applied Surface Science, vol. 27, pp. 381-392, 1987.

[198] T. J. Chuang, C. R. Brundle, and D. W. Rice, "Interpretation of the x-ray photoemission spectra of cobalt oxides and cobalt oxide surfaces," Surface Science, vol. 59, pp. 413-429, 1976.

[199] G. Tyuliev, D. Panayotov, I. Avramova, D. Stoichev, and T. Marinova, "Thin-film coating of Cu-Co oxide catalyst on lanthana/zirconia films electrodeposited on stainless steel," Materials Science and Engineering: C, vol. 23, pp. 117-121, 2003.

[200] P. Stefanov, I. Avramova, D. Stoichev, N. Radic, B. Grbic, and T. Marinova, "Characterization and catalytic activity of Cu-Co spinel thin films catalysts," Applied Surface Science, vol. 245, pp. 65-72, 2005.

[201] G. Tyuliev and S. Angelov, "The nature of excess oxygen in Co3O4+ϵ," Applied Surface Science, vol. 32, pp. 381-391, 1988.

[202] A. C. Tavares, M. A. M. Cartaxo, M. I. da Silva Pereira, and F. M. Costa, "Effect of the partial replacement of Ni or Co by Cu on the electrocatalytic activity of the NiCo2O4 spinel oxide," Journal of Electroanalytical Chemistry, vol. 464, pp. 187-197, 1999.

[203] J. J. Shim and J. G. Kim, "Copper corrosion in potable water distribution systems: influence of copper products on the corrosion behavior," Materials Letters, vol. 58, pp. 2002-2006, 2004.

[204] Y. Peng, B. Wang, and A. Gerson, "The effect of electrochemical potential on the activation of pyrite by copper and lead ions during grinding," International Journal of Mineral Processing, vol. 102–103, pp. 141-149, 2012.

[205] S. Zhao and Y. Peng, "The oxidation of copper sulfide minerals during grinding and their interactions with clay particles," Powder Technology, vol. 230, pp. 112-117, 2012.

[206] A. Salvi, F. Langerame, A. Macchia, M. Sammartino, and M. Tabasso, "XPS characterization of (copper-based) coloured stains formed on limestone surfaces of outdoor Roman monuments," Chemistry Central Journal, vol. 6, pp. 1-13, 2012/05/02 2012.

[207] Y. Li, J. Zhao, J. Han, and X. He, "Combustion synthesis and characterization of NiCuZn ferrite powders," Materials Research Bulletin, vol. 40, pp. 981-989, 2005.

[208] M. Zimowska, A. Michalik-Zym, R. Janik, T. Machej, J. Gurgul, R. P. Socha, J. Podobiński, and E. M. Serwicka, "Catalytic combustion of toluene over mixed Cu–Mn oxides," Catalysis Today, vol. 119, pp. 321-326, 2007.

[209] G.-X. Qi, X.-M. Zheng, J.-H. Fei, and Z.-Y. Hou, "A novel catalyst for DME synthesis from CO hydrogenation: 1. Activity, structure and surface properties," Journal of Molecular Catalysis A: Chemical, vol. 176, pp. 195-203, 2001.

[210] Y. D. Premchand and S. A. Suthanthiraraj, "Structural investigation of (CuI)0.45–(Ag2WO4)0.55 solid electrolyte using X-ray photoelectron and laser Raman spectroscopies," Electrochemistry Communications, vol. 6, pp. 1266-1269, 2004.

[211] Z.-s. Hong, Y. Cao, and J.-f. Deng, "A convenient alcohothermal approach for low temperature synthesis of CuO nanoparticles," Materials Letters, vol. 52, pp. 34-38, 2002.

181

[212] Y. Lv, L. Liu, H. Zhang, X. Yao, F. Gao, K. Yao, L. Dong, and Y. Chen, "Investigation of surface synergetic oxygen vacancy in CuO–CoO binary metal oxides supported on c-Al2O3 for NO removal by CO," Journal of Colloid and Interface Science, vol. 390, pp. 158-169, 2013.

[213] Y. Lv, H. Zhang, Y. Cao, L. Dong, L. Zhang, K. Yao, F. Gao, L. Dong, and Y. Chen, "Investigation of the physicochemical properties of CuO–CoO binary metal oxides supported on γ-Al2O3 and their activity for NO removal by CO," Journal of Colloid and Interface Science, vol. 372, pp. 63-72, 2012.

[214] M. F. Al-Kuhaili, "Characterization of copper oxide thin films deposited by the thermal evaporation of cuprous oxide (Cu2O)," Vacuum, vol. 82, pp. 623-629, 2008.

[215] S. Yokoyama, H. Takahashi, T. Itoh, K. Motomiya, and K. Tohji, "Elucidation of the reaction mechanism during the removal of copper oxide by halogen surfactant at the surface of copper plate," Applied Surface Science, vol. 264, pp. 664-669, 2013.

[216] J. Zhou, C. Soto, M.-S. Chen, M. Bruckman, M. Moore, E. Barry, B. Ratna, P. Pehrsson, B. Spies, and T. Confer, "Biotemplating rod-like viruses for the synthesis of copper nanorods and nanowires," Journal of Nanobiotechnology, vol. 10, pp. 1-12, 2012/05/01 2012.

[217] A. A. Athawale, M. Majumdar, H. Singh, and K. Navinkiran, Synthesis of Cobalt Oxide Nanoparticles/Fibres in Alcoholic Medium using y-ray Technique vol. 60, 2010.

[218] J. Xu, P. Gao, and T. S. Zhao, "Non-precious Co3O4 nano-rod electrocatalyst for oxygen reduction reaction in anion-exchange membrane fuel cells," Energy & Environmental Science, vol. 5, pp. 5333-5339, 2012.

[219] P. Shelke, Y. Khollam, K. Patil, S. Gunjal, S. Jadkar, M. Takwale, and K. Mohite, "Studies on Electrochemical Deposition and Characterization of Co 3 O 4 Films," Journal of Nano-and Electronic Physics, vol. 3, pp. 486-498, 2011.

[220] N. Srisawad, W. Chaitree, O. Mekasuwandumrong, P. Praserthdam, and J. Panpranot, "Formation of CoAl2O4 Nanoparticles via Low-Temperature Solid-State Reaction of Fine Gibbsite and Cobalt Precursor," Journal of Nanomaterials, vol. 2012, p. 8, 2012.

[221] G. Fierro, M. Lo Jacono, M. Inversi, R. Dragone, and P. Porta, "TPR and XPS study of cobalt–copper mixed oxide catalysts: evidence of a strong Co–Cu interaction," Topics in Catalysis, vol. 10, pp. 39-48, 2000.

[222] Z. Lendzion-Bieluń, M. M. Bettahar, S. Monteverdi, D. Moszyński, and U. Narkiewicz, "Effect of Cobalt on the Activity of CuO/CeO2 Catalyst for the Selective Oxidation of CO," Catalysis Letters, vol. 134, pp. 196-203, 2010/02/01 2010.

[223] G. G. Volkova, T. M. Yurieva, L. M. Plyasova, M. I. Naumova, and V. I. Zaikovskii, "Role of the Cu–Co alloy and cobalt carbide in higher alcohol synthesis," Journal of Molecular Catalysis A: Chemical, vol. 158, pp. 389-393, 2000.

[224] J. Jia, X. Li, and G. Chen, "Stable spinel type cobalt and copper oxide electrodes for O2 and H2 evolutions in alkaline solution," Electrochimica Acta, vol. 55, pp. 8197-8206, 2010.

[225] F. C. Zhang, H. H. Luo, T. S. Wang, S. G. Roberts, and R. I. Todd, "Influence factors on wear resistance of two alumina matrix composites," Wear, vol. 265, pp. 27-33, 2008.

[226] H. Liu, J. Tao, J. Xu, Z. Chen, and Q. Gao, "Corrosion and tribological behaviors of chromium oxide coatings prepared by the glow-discharge plasma technique," Surface and Coatings Technology, vol. 204, pp. 28-36, 2009.

[227] C. A. Freyman and Y.-W. Chung, "Synthesis and characterization of hardness-enhanced multilayer oxide films for high-temperature applications," Surface and Coatings Technology, vol. 202, pp. 4702-4708, 2008.

[228] M. Brandhorst, J. Zajac, D. J. Jones, J. Rozière, M. Womes, A. Jimenez-López, and E. Rodríguez-Castellón, "Cobalt-, copper- and iron-containing monolithic aluminosilicate-supported preparations for selective catalytic reduction of NO with NH3 at low temperatures," Applied Catalysis B: Environmental, vol. 55, pp. 267-276, 2005.

182

[229] A. Amri, X. Duan, C.-Y. Yin, Z.-T. Jiang, M. M. Rahman, and T. Pryor, "Solar absorptance of copper–cobalt oxide thin film coatings with nano-size, grain-like morphology: Optimization and synchrotron radiation XPS studies," Applied Surface Science, vol. 275, pp. 127-135, 2013.

[230] R. Amadelli, L. Samiolo, A. Maldotti, A. Molinari, M. Valigi, and D. Gazzoli, "Preparation, Characterisation, and Photocatalytic Behaviour of Co-TiO2 with Visible Light Response," International Journal of Photoenergy, vol. 2008, 2008.

[231] Y.-Y. Peng, T.-E. Hsieh, and C.-H. Hsu, "White-light emitting ZnO–SiO2 nanocomposite thin films prepared by the target-attached sputtering method," Nanotechnology, vol. 17, pp. 174-180, 2006.

[232] B. R. Lewandowski, K. L. Lusker, Z. M. LeJeune, D. A. Lytle, P. Zhou, P. T. Sprunger, and J. C. Garno, "Impact of pH, Dissolved Inorganic Carbon, and Polyphosphates for the Initial Stages of Water Corrosion of Copper Surfaces Investigated by AFM and NEXAFS," CheM, vol. 1, pp. 16-26, 2011.

[233] A. N. Buckley, W. M. Skinner, S. L. Harmer, A. Pring, and L.-J. Fan, "Electronic environments in carrollite, CuCo2S4, determined by soft X-ray photoelectron and absorption spectroscopy," Geochimica et Cosmochimica Acta, vol. 73, pp. 4452-4467, 2009.

[234] G. van der Laan, R. A. D. Pattrick, C. M. B. Henderson, and D. J. Vaughan, "Oxidation state variations in copper minerals studied with Cu 2p X-ray absorption spectroscopy," Journal of Physics and Chemistry of Solids, vol. 53, pp. 1185-1190, 1992.

[235] G. van der Laan, R. A. D. Pattrick, J. M. Charnock, and B. A. Grguric, "Cu L2,3 X-ray absorption and the electronic structure of nonstoichiometric Cu5FeS4," Physical Review B, vol. 66, p. 045104, 2002.

[236] R. A. D. Pattrick, G. Laan, D. J. Vaughan, and C. M. B. Henderson, "Oxidation state and electronic configuration determination of copper in tetrahedrite group minerals by L-edge X-ray absorption spectroscopy," Physics and Chemistry of Minerals, vol. 20, pp. 395-401, 1993.

[237] R. A. D. Pattrick, J. F. W. Mosselmans, J. M. Charnock, K. E. R. England, G. R. Helz, C. D. Garner, and D. J. Vaughan, "The structure of amorphous copper sulfide precipitates: An X-ray absorption study," Geochimica et Cosmochimica Acta, vol. 61, pp. 2023-2036, 1997.

[238] J. G. Chen, "NEXAFS investigations of transition metal oxides, nitrides, carbides, sulfides and other interstitial compounds," Surface Science Reports, vol. 30, pp. 1-152, 1997.

[239] M. Abbate, J. C. Fuggle, A. Fujimori, L. H. Tjeng, C. T. Chen, R. Potze, G. A. Sawatzky, H. Eisaki, and S. Uchida, "Electronic structure and spin-state transition of LaCoO3," Physical Review B, vol. 47, pp. 16124-16130, 1993.

[240] T. Kroll, A. A. Aligia, and G. A. Sawatzky, "Polarization dependence of x-ray absorption spectra of NaXCoO2 : Electronic structure from cluster calculations," Physical Review B, vol. 74, p. 115124, 2006.

[241] G. van der Laan, B. T. Thole, G. A. Sawatzky, and M. Verdaguer, "Multiplet structure in the L2,3 x-ray-absorption spectra: A fingerprint for high- and low-spin Ni2+ compounds," Physical review. B, Condensed matter, vol. 37, pp. 6587-6589, 1988.

[242] T. Mizokawa, L. H. Tjeng, P. G. Steeneken, N. B. Brookes, I. Tsukada, T. Yamamoto, and K. Uchinokura, "Photoemission and x-ray-absorption study of misfit-layered (Bi,Pb)-Sr-Co-O compounds: Electronic structure of a hole-doped Co-O triangular lattice," Physical Review B, vol. 64, p. 115104, 2001.

[243] S. Gautam, P. Thakur, K. H. Chae, G. S. Chang, M. Subramanain, and R. Jayavel, "Electronic Structure of Co-doped ZnO Thin Films by X-ray Absorption and Emission Spectroscopy," Journal of the Korean Physical Society, vol. 55, pp. 167-172, 2009.

[244] S. Abhinav Pratap, K. Ravi, P. Thakur, N. B. Brookes, K. H. Chae, and W. K. Choi, "NEXAFS and XMCD studies of single-phase Co doped ZnO thin films," Journal of Physics: Condensed Matter, vol. 21, p. 185005, 2009.

183

[245] J. H. Guo, L. Vayssieres, C. Persson, R. Ahuja, B. Johansson, and J. Nordgren, "Polarization-dependent soft-x-ray absorption of highly oriented ZnO microrod arrays," Journal of Physics: Condensed Matter, vol. 14, p. 6969, 2002.

[246] F. Mammeri, E. L. Bourhis, L. Rozes, and C. Sanchez, "Mechanical properties of hybrid organic-inorganic materials," Journal of Materials Chemistry, vol. 15, pp. 3787-3811, 2005.

[247] D. Kenfaui, G. Bonnefont, D. Chateigner, G. Fantozzi, M. Gomina, and J. G. Noudem, "Ca3Co4O9 ceramics consolidated by SPS process: Optimisation of mechanical and thermoelectric properties," Materials Research Bulletin, vol. 45, pp. 1240-1249, 2010.

[248] D. Kenfaui, D. Chateigner, M. Gomina, and J. G. Noudem, "Texture, mechanical and thermoelectric properties of Ca3Co4O9 ceramics," Journal of Alloys and Compounds, vol. 490, pp. 472-479, 2010.

[249] M. T. Tilbrook, D. J. Paton, Z. Xie, and M. Hoffman, "Microstructural effects on indentation failure mechanisms in TiN coatings: Finite element simulations," Acta Materialia, vol. 55, pp. 2489-2501, 2007.

[250] R. D. Shannon, D. B. Rogers, C. T. Prewitt, and J. L. Gillson, "Chemistry of noble metal oxides. III. Electrical transport properties and crystal chemistry of ABO2 compounds with the delafossite structure," Inorganic Chemistry, vol. 10, pp. 723-727, 1971/04/01 1971.

[251] H. P. Klug and L. E. Alexander, "X-ray diffraction procedures: for polycrystalline and amorphous materials," X-Ray Diffraction Procedures: For Polycrystalline and Amorphous Materials, 2nd Edition, by Harold P. Klug, Leroy E. Alexander, pp. 992. ISBN 0-471-49369-4. Wiley-VCH, May 1974., vol. 1, 1974.

[252] B. Heying, X. Wu, S. Keller, Y. Li, D. Kapolnek, B. Keller, S. P. DenBaars, and J. Speck, "Role of threading dislocation structure on the x‐ray diffraction peak widths in epitaxial GaN films," Applied Physics Letters, vol. 68, pp. 643-645, 1996.

[253] H. Uwe and G. Neil, "The Scherrer equation versus the 'Debye-Scherrer equation'," Nature Nanotechnology, vol. 6, pp. 534-534, 2011.

[254] I. C. Noyan and J. B. Cohen, Residual stresses: measurement by diffraction and interpretation ; with 31 tables: SPRINGER VERLAG GMBH, 1987.

[255] L. B. Freund and S. Suresh, Thin Film Materials: Stress, Defect Formation and Surface Evolution: Cambridge University Press, 2003.

[256] A. P. Boresi and R. J. Schmidt, Advanced mechanics of materials: John Wiley & Sons, 2003. [257] R. K. Singh, M. T. Tilbrook, Z. H. Xie, A. Bendavid, P. J. Martin, P. Munroe, and M. Hoffman,

"Contact damage evolution in diamondlike carbon coatings on ductile substrates," Journal of Materials Research, vol. 23, pp. 27-36, 2008.

184

Appendix 1 Table of initial absorptance (α) and emittance (ɛ) values and the corresponding determined

maximum temperature (T1, oC) applied in the thermal test. The line entitled “α(AR)>" is used

for solar absorber surfaces with antireflective layer.

185

Appendix 2

Table of test conditions for the different accelerated temperature tests used in the

qualification of solar absorber surface

Amun Amri, ST., MT. - Murdoch UniversityAmun Amri, ST., MT. This thesis is presented for the degree of Doctor of Philosophy of Murdoch University ... and EDX analyses corroborated

Documents