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
6.1. Introduction 93
6.2. Sample Preparation and Characterisation 94
6.3. Results and Discussion
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
7.1. Introduction 111
7.2. Sample Preparation and Characterisation 113
7.3. Results and Discussion
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
7.4. Conclusions 131
Chapter 8. Characteristics of Copper Cobalt Oxides Thin Film Coatings
Synthesised by Different Annealing Temperatures
8.1. Introduction 133
8.2. Experimental 135
8.3. Results and Discussion
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
8.4. Conclusions 153
Chapter 9. Optical Properties and Thermal Durability of Copper Cobalt Oxide
Thin Film Coatings with Integrated Silica Antireflection Layer
9.1. Introduction 154
9.2. Samples Preparation and Characterisation 155
9.3. Results and Discussion
9.3.1. Reflectance spectra and solar absorptance 156
9.3.2. Emittance and selectivity 158
9.3.3. Accelerated thermal durability test 160
9.4. Conclusions 165
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
thin film coatings 124
Figure 7.7. O K-edge NEXAFS spectra in AEY mode for copper cobalt oxide
thin film coatings 126
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
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) 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
film coatings synthesised at different annealing temperatures 141
Table 8.4. The curve-fittings results of Co 2p3/2 and its satellite of copper cobalt
film coatings synthesised at different annealing temperatures 144
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. Results and Discussion
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.
6.2. Samples Preparation and Characterisation
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. Results and Discussion
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.
7.2. Samples Preparation and Characterisation
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. Results and Discussion
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%)
Attributions Octahedral
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. Results and Discussion
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
synthesised at different annealing temperatures.
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
synthesised at different annealing temperatures.
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
synthesised at different annealing temperatures.
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 %)
Attributions Octahedral
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
synthesised at different annealing temperatures.
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. Results and Discussion
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.
Figure 9.1. Reflectance spectra of copper cobalt oxide thin film coatings with
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 Cu–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