Surface nanoengineering of titanium alloys for …...Surface Nanoengineering of Titanium Alloys for...
Transcript of Surface nanoengineering of titanium alloys for …...Surface Nanoengineering of Titanium Alloys for...
Surface Nanoengineering of
Titanium Alloys for Biomedical
Applications
A thesis submitted in total fulfilment of the requirements for the degree of
Doctor of Philosophy
By
Sepideh Minagar
Faculty of Science, Engineering and Technology
Swinburne University of Technology
Hawthorn, Melbourne
Australia
February 2015
ii
Abstract
Failure of a biomaterial implant occurs when it cannot be accepted by the body. In terms
of orthopaedic implants, this means that there is no bonding between the implant
material and bone cells. In this study, fabricated nanotubular layers on the surface of
Ti50Zr binary alloy was investigated for its bioactivity and osseointegration after
anodisation. This study investigates the effect of nanopatterning of the surface as metal
oxide nanotubes which convert the bio-inert metal surface to be bioactive. This
nanotubular layer was examined for its ability to induce bone-like calcium phosphate
(CaP), e.g. hydroxyapatite or other compositions such as octacalcium phosphate and
tricalcium phosphate through immersion in a modified simulated body fluid (m-SBF).
The biocompatibility of the nanotubular layer was also assessed by cell culture test
using human osteoblast-like cells (SaOS-2). As the bare TiZr metal exhibited potential
for metallic implant applications due to its excellent bioactivity and biocompatibility, it
is expected that the nanotubular layer plays a promising role in forming strong
adhesions with bone cells, allowing it to be used as an implant.
The first part of this study investigates the condition for fabrication of TiO2-ZrO2-
ZrTiO4 nanotubes. This includes the changing of the anodisation parameters such as the
applied potential, the concentration of F- ion in the electrolyte, the type of electrolyte
and the anodisation time. In this research project, a full characterisation of the TiO2-
ZrO2-ZrTiO4 nanotubes including morphological, thermal, topographical, chemical and
mechanical properties has been investigated using scanning electron microscopy
(SEM), thin film X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS)
analysis, nanoindentation and water contact angle measurements. Bioactivity and
biocompatibility assessment have been employed in order to demonstrate the
characteristics of the nanotubular layer to improve hydroxyapatite mineralisation and
bone cell integration.
It was found that in the aqueous electrolyte there was a distribution of inner diameter,
outer diameter and wall thickness of nanotubes in comparison to TiO2 nanotubes as well
as in non-aqueous electrolyte, due to the different oxidation rates of Ti and Zr. The
orientation of nanotubes followed the microstructure of the Ti50Zr alloy. This effect
was obvious in the aqueous electrolyte whilst in the non-aqueous electrolyte it was less
affected and the orientation can be observed when the water content of the electrolyte
increased from 5 to 10 wt %. The walls of nanotubes were not separated when they
iii
were fabricated in non-aqueous electrolyte. The roughness parameters of the
nanotubular layer of TiO2-ZrO2-ZrTiO4 were higher than the nanotubular layer of TiO2,
due to the different height of the grown nanotubes. It has been shown that there was a
direct link between roughness parameter and wettability of the nanotubular layer. The
wettability or hydrophilic properties of nanotubes increased with an increase in
roughness. The TiO2-ZrO2-ZrTiO4 nanotubes exhibited higher surface energy than TiO2
nanotubes. Hydroxyapatite mineralised on the TiO2-ZrO2-ZrTiO4 nanotubes with higher
Ca/P ratio and thickness than on the TiO2 nanotubes. When Sa of the as-formed TiO2-
ZrO2-ZrTiO4 nanotubes increased, the Ca/P ratio of HA increased in contrast to
annealed nanotube. Increasing the hydrophilic properties of either as-formed or
annealed nanotubes resulted in an increase of the Ca/P ratio of HA. After annealing, the
Ca/P ratio of HA increased. Results of the MTS assay indicated that the percentage of
cell adhesion on the nanotubes fabricated in aqueous and non-aqueous electrolyte was
critically affected by the nanoscale topographical parameters including the tube inner
diameter (Di), the tube wall thickness (Wt), the amplitude roughness (Sa) and the spacing
roughness (Sm) of the nanotubular surface. The highest percentage of cell attachment
was on the surface of TiO2-ZrO2-ZrTiO4 nanotubes with Di = 18 nm (optimum
nanospacing) and the lowest percentage of cell attachment was on the surface of TiO2-
ZrO2-ZrTiO4 nanotubes with Di = 59 nm (higher amplitude parameter of roughness, Sa)
when the nanotubes were fabricated in an aqueous electrolyte. It was found that a post-
treatment that reduces the contamination of nanotubular surface or mineralising
hydroxyapatite can increase cell attachment significantly from 41.0 % to 74.8 % and
59.7 %, respectively. When the nanotubes were fabricated in non-aqueous electrolyte,
bone cells behaved similarly. The highest percentage of cell attachment was on the
surface of TiO2-ZrO2-ZrTiO4 nanotubes with Di = 25 nm (optimum nanospacing) and
the lowest percentage of cell attachment were on the surface of TiO2-ZrO2-ZrTiO4
nanotubes with Di = 29 nm (higher spacing parameter of roughness, Sm).
iv
Acknowledgments
First and foremost I wish to express my deepest gratitude to my principal supervisor
Professor Cuie Wen for her continuous support, encouragement and unstinting
supervision in conducting research, and her optimistic manner and valuable guidance in
writing the research publications. I also appreciate her trust and friendship which means
a great deal to me. I would also like to express my sincere gratitude to my associate
supervisor Professor Christopher C. Berndt for his invaluable advice and guidance in
writing the research publications. I must thank my other associate supervisor Dr James
Wang for his continuous help and support in the metallographic and spectroscopy
laboratories. I am thankful to Professor Elena P. Ivanova for her supervision on the cell
biology aspect of this research. I also acknowledge the State Government of Victoria of
Australia and Swinburne University of Technology for supporting and funding this
research project through a Victorian International Research Scholarship (VIRS) and
Swinburne University Postgraduate Research Award (SUPRA).
I wish to thank Dr Yuncang Li for his guidance and advice in the cell culture study and
the nanoindentation research at the Institute for Frontier Materials, Deakin University,
Geelong. I also thank Dr Thomas Gengenbach for his help in XPS analysis at
CSIRO Materials Science and Engineering, Clayton.
The help of Dr Mehran Motamed Ektesabi, Dr Igor Sbarski, Dr Akbar Rhamdhani, Dr
Thomas Ameringer, Dr Mostafa Nikzad, Dr Yeannette Lizama and Dr Vi Khanh
Truong for permitting me to use their equipments and/or chemicals is highly
appreciated. My thanks are extended to Dr De Ming Zhu for training me to use the 3D-
Profilometer. The technical assistance of Messieurs Brian Dempster and Andrew Moore
and the technicians of the workshop of the Faculty of Science, Engineering and
Technology are gratefully acknowledged. I thank all my supervisor research group
members (PhD students and Research Fellows) for their scientific discussions and
advice. I wish to thank Ms Madeleine Bruwer, the liaison librarian of the Faculty of
Science, Engineering and Technology because she always answered my questions
patiently. My sincere thanks to Professor Syed H. Masood, Associate Professor Paul
Stoddart and Dr Ryan Cottam for their advice in my annual progress reviews.
v
I wish and must thank especially my close friend Ms Fatemeh Hamedi because of her
sister-like friendship and emotional support. She was, is and will always be willing to
help me no matter how hard the challenge. I would like to thank all my friends here and
in my home country for their friendship and support, which made the course of my PhD
studies an enjoyable experience. I would also like to sincerely thank Dr Gwyneth and
Prof Michael Asten, Ms Rosemarie and Mr John Worboys as they have not only helped
me improve my English but also made me feel right at home.
Last but not least, I should thank my mother Farazandeh and my father Mohammad
Hasan for their endless loves, prays, encouragements and support. I would also like to
express my love and gratitude to my sister Sima, my brother Sepehr and my niece Zahra
who encouraged me all the time.
vi
Declaration
This thesis contains no material which has been accepted for the award to the candidate
of any other degree or diploma, except where due reference is made in the text. I declare
that to the best of my knowledge this study contains no material previously published or
written by another person except where due reference is made in the text.
Sepideh Minagar
February 2015
vii
List of publications
Journal Papers:
Sepideh Minagar, Christopher C. Berndt, James Wang, Elena Ivanova, Cuie Wen. A
review of the application of anodization for the fabrication of nanotubes on metal
implant surfaces. Acta Biomaterialia, (2012), 8, 2875-2888.
Sepideh Minagar, James Wang, Christopher C. Berndt, Elena P. Ivanova, Cuie Wen.
Cell response of anodized nanotubes on titanium and titanium alloys - a review. Journal
of Biomedical Materials Research: Part A, (2013), 101 A (9), 2726-2739.
Sepideh Minagar, Christopher C. Berndt, Thomas Gengenbach, Cuie Wen. Fabrication
and characterization of TiO2-ZrO2-ZrTiO4 nanotubes on TiZr alloy manufactured via
anodization. Journal of Materials Chemistry B, (2014), 2 (1), 71-83.
Sepideh Minagar, Yuncang Li, Christopher C. Berndt, Cuie Wen. The influence of
titania-zirconia-zirconium titanate nanotube characteristics on osteoblast cell adhesion.
Acta Biomaterialia (2015), 12, 281–289.
Presentation at Research Conferences:
Poster presentation entitled “Simultaneous enhancing cell interaction and eliminating
bacterial infection in titania nanotubes”, Proceedings of the 4th International
Conference on Nanostructures (ICNS4), Kish Island, I.R. Iran, March 2012.
Oral presentation entitled “anodised nanotubes on titanium alloy”, 15th International
Conference on Advances in Materials & Processing Technologies Conference (AMPT),
Wollongong, Australia, September 2012.
viii
Table of contents
Abstract ............................................................................................................................. ii
Acknowledgments ............................................................................................................ iv
Declaration ....................................................................................................................... vi
List of publications .......................................................................................................... vii
Table of contents ............................................................................................................ viii
List of figures ................................................................................................................. xiii
List of tables .................................................................................................................. xxii
List of abbreviations ..................................................................................................... xxiv
Chapter 1 Introduction ...................................................................................................... 1
1.1 Overview ................................................................................................................. 1
1.2 Thesis objective .................................................................................................. 4
1.3 Thesis structure ................................................................................................... 5
Chapter 2 Literature review .............................................................................................. 7
2.1 Introduction ............................................................................................................. 7
2.1.1 Bone: structure, composition and properties .................................................... 8
2.1.2 Bone implant materials .................................................................................. 11
2.1.3 Surface treatment for implant materials ......................................................... 15
2.2 Anodic oxidation as a metallic implant surface treatment ............................... 16
2.2.1 The influence of the type and concentration of aqueous electrolyte on TiO2
nanotubes ................................................................................................................ 20
2.2.2 The influence of the non-aqueous electrolyte on TiO2 nanotubes ................. 23
2.2.3 The effect of pH value on the formation of TiO2 nanotubes.......................... 24
2.3 Nanotube oxide layer on titanium alloys and titanium alloying metals ........... 26
2.3.1 Anodisation of biocompatible Ti-Nb-Ta-Zr alloy.......................................... 26
2.3.2 Anodisation of binary titanium alloys for implant applications..................... 28
2.3.3 Anodisation of tantalum as a β stabiliser ....................................................... 33
ix
2.3.4 Anodisation of niobium as a β stabiliser ........................................................ 35
2.3.5 Anodisation of zirconium as a neutral titanium alloying element (or an
effective β stabilizer in multi-elementary Ti alloys) ............................................... 35
2.4 Factors that influence bone cell adhesion ......................................................... 37
2.4.1 The influence of surface physicochemical, mechanical and electrical
properties on bone cell behaviour ........................................................................... 39
2.4.2 Effect of topography on the bone cell behaviour ........................................... 41
2.5 Effect of the characteristics of anodised TiO2 nanotubes on bone cell behaviour
43
2.5.1 Effect of nano-spacing of the surface of TiO2 nanotubes on cell behaviour .. 43
2.5.2 Effect of crystalline phase of TiO2 nanotubes on bone cell behaviour .......... 49
2.5.3 Effect of hydroxyapatite (HA) coating on the surface of TiO2 nanotubes on
bone cell behaviour ................................................................................................. 52
2.6 In vivo effect of micro/nanostructure of the surface of TiO2 nanotubes .......... 54
2.7 Effect of TiO2 nanotubes on bacteria attachment ............................................. 56
2.8 Summary .......................................................................................................... 56
Chapter 3 Materials and methods .................................................................................... 59
3.1 Introduction ........................................................................................................... 59
3.2 Sample preparation ........................................................................................... 59
3.3 Fabrication of nanoporous and nanotubular layers .......................................... 59
3.4 Surface characterisation.................................................................................... 61
3.4.1 Surface morphology and chemical composition characterisation .................. 61
3.4.2 Surface topography and water contact angle and surface energy measurement
................................................................................................................................. 61
3.5 Bioactivity assessment by SBF soaking ........................................................... 62
3.6 Assessing cell responses on nanotubes with different nanoscale dimensions and
surface topographies .................................................................................................... 64
3.7 Nanohardness and elasticity measurements .......................................................... 65
x
Chapter 4 Nanotubes formed in aqueous electrolyte ...................................................... 67
4.1 Introduction ...................................................................................................... 67
4.2 Materials and methods ...................................................................................... 70
4.3 Results and discussion ...................................................................................... 71
4.3.1 Formation of TiO2-ZrO2-ZrTiO4 nanotubes .................................................. 71
4.3.2 Surface roughness of the TiO2-ZrO2-ZrTiO4 nanotubular surface ................ 95
4.3.3 Hydrophilic properties of the nanotubular surfaces ..................................... 100
4.3.4 Mechanical properties of the TiO2-ZrO2-ZrTiO4 nanotubes ........................ 102
4.3.5 Effect of annealing on the TiO2-ZrO2-ZrTiO4 nanotubes ............................ 104
4.4 Conclusions .................................................................................................... 107
Chapter 5 Nanotubes formed in non-aqueous electrolyte ............................................. 110
5.1 Introduction .................................................................................................... 110
5.2 Materials and methods .................................................................................... 112
5.3 Results and discussion .................................................................................... 113
5.3.1 Formation and characterisation of TiO2-ZrO2-ZrTiO4 nanotubes ............... 113
5.3.2 Surface roughness and hydrophilic property of the TiO2-ZrO2-ZrTiO4
nanotubular surface ............................................................................................... 125
5.3.3 Mechanical properties .................................................................................. 131
5.4 Conclusions .................................................................................................... 133
Chapter 6 Bioactivity of nanotubes fabricated in aqueous and non-aqueous electrolytes
....................................................................................................................................... 135
6.1 Introduction .................................................................................................... 135
6.2 Materials and methods .................................................................................... 137
6.3 Results and discussion .................................................................................... 139
6.3.1 Dimensional, physical and chemical properties of TiO2-ZrO2-ZrTiO4
nanotubes fabricated in aqueous electrolyte ......................................................... 139
6.3.2 Bioactivity of TiO2-ZrO2-ZrTiO4 and TiO2 nanotubes fabricated in aqueous
electrolyte .............................................................................................................. 146
xi
6.3.3 Dimensional, physical and chemical properties of TiO2-ZrO2-ZrTiO4
nanotubes fabricated in non-aqueous electrolyte .................................................. 155
6.3.4 Bioactivity of TiO2-ZrO2-ZrTiO4 and TiO2 nanotubes fabricated in non-
aqueous electrolyte ................................................................................................ 158
6.4 Conclusions .................................................................................................... 165
Chapter 7 Cell response of nanotubes formed in both aqueous electrolyte and non-
aqueous electrolyte ........................................................................................................ 167
7.1 Introduction .................................................................................................... 168
7.2 Materials and methods .................................................................................... 170
7.3 Results and discussion .................................................................................... 174
7.3.1 Dimensional, physical and chemical properties of TiO2-ZrO2-ZrTiO4
nanotubes fabricated in aqueous electrolyte ......................................................... 174
7.3.2 Cell adhesion and spreading on TiO2-ZrO2-ZrTiO4 nanotubes fabricated in
aqueous electrolyte ................................................................................................ 180
7.3.3 Dimensional, physical and chemical properties of TiO2-ZrO2-ZrTiO4
nanotubes fabricated in non-aqueous electrolyte .................................................. 201
7.3.4 Cell adhesion and spreading on TiO2-ZrO2-ZrTiO4 nanotubes fabricated in
non-aqueous electrolyte ........................................................................................ 205
7.4 Conclusions .................................................................................................... 210
Chapter 8 Nanoporous and nanotubular metal oxide layers on biocompatible metals of
Ta, Nb and Zr and their potential applications .............................................................. 212
8.1 Introduction .................................................................................................... 213
8.2 Materials and methods .................................................................................... 215
8.3 Results and discussion .................................................................................... 217
8.3.1 Process conditions and formation mechanism of nanoporous/nanotubular
metal oxides .......................................................................................................... 217
8.3.2 The dynamics of the anodisation process for tantala (Ta2O5), niobia (Nb2O5)
and zirconia (ZrO2) ............................................................................................... 218
xii
8.3.3 Physical characteristics of nanoporous Ta2O5 and Nb2O5 and nanotubular
ZrO2 layers ............................................................................................................ 222
8.3.4 Bioactivity of nanoporous and nanotubular metal oxide layers ................... 230
8.4 Conclusions .................................................................................................... 235
Chapter 9 Conclusions .................................................................................................. 237
9.1 Introduction ......................................................................................................... 237
9.2 Major findings ................................................................................................ 237
9.3 Recommendation for future work .................................................................. 244
References ..................................................................................................................... 246
xiii
List of figures
Fig. 2.1 Hierarchical structure of cortical bone (adapted from [32]) ................................ 9
Fig. 2.2 Growth of regular TiO2 nanotubes, a) cathodic reaction, b) anodic reaction, c)
transition state of TiO2 layer, d) starting of nanotube formation and e) titania nanotubes
......................................................................................................................................... 18
Fig. 2.3 Schematic current-time curve (adapted from [58])............................................ 19
Fig. 2.4 SEM and TEM images of TiO2 nanotubes (top view, cross-section and bottom):
a) formed in 1 M H3PO4 + 0.3 wt % HF at 15 V [61], b) in 1 wt % HF at 20 V for 15
and 30 min) [66], c) in 1 M (NH4)2SO4 + 0.5wt % NH4F at 20V [67], d) in ethylene
glycol + 0.3 wt % NH4F + 2 vol % H2O [68] ................................................................. 21
Fig. 2.5 SEM and TEM images (top view, cross-section and bottom) of anodic oxide
nanotube and nanoporous layer formed on a) β phase titanium alloy after anodisation at
20 V for 4000 s [99], d) Zr substrate using 1M (NH4)2SO4 + 0.5 wt % NH4F
electrolyte [102], c) niobium substrate as a function of anodisation temperature at 15 °C
[103], b) tantalum substrate using 1 M H2SO4 + 2 wt % HF electrolyte for 2 h with a
sweep rate 100 mVs-1 [104] ............................................................................................ 28
Fig. 2.6 Representation of the cell proteins involved in cell adhesion on biomaterial: (a)
immediately after implantation; (b) adsorbing proteins from body fluid; and (c) attached
bone cell on an implant material surface in higher magnification (adapted from [15]) . 38
Fig. 2.7 SEM image of filopodia of the SaOs-2 cells on 200 nm deep round concentric
grooves and ridges in quartz [157] .................................................................................. 42
Fig. 2.8 Schematic illustration of a bone cell (osteoblast) attached on titania nanotubes
with a diameter less than 100 nm (adapted from [109] and [35]). .................................. 46
Fig. 2.9 SEM images of extended MC3T3-E1 preosteoblast cell filopodia on nanotube
layers with different diameters: (a) 20 nm, (b) 50 nm, (c) 70 nm, (d) 100 nm, and (e)
120 nm (×70,000), (f) 120 nm (×30,000) [36] ................................................................ 51
Fig. 4.1 SEM images of TiO2-ZrO2-ZrTiO4 nanotubes: (a-1) microstructure of etched
Ti50Zr, (a-2) top view of patterned nanotubes of different phases exhibited in the
microstructure of Ti50Zr alloy, (b) nanoporous patches of different phases, (c) cross
section of the nanotubes showing the nanotube length, (d) view of the nanotubes from
xiv
the bottom, (e) top view and cross section of separated nanotubes of and (f) top view of
damaged nanotubes ......................................................................................................... 72
Fig. 4.2 SEM images of TiO2-ZrO2-ZrTiO4 nanotubes: (a) top view of as-formed
nanotubes anodised in 0.4 wt % NH4F at 20 V, (b) the nanotubes annealed at 500 ºC for
3 h, (c) nanotubes as-formed in 0.5 wt % NH4F anodised at 20 V, (d) the nanotubes
annealed at 500 ºC for 3 h, (e) nanotubes as-formed in 0.5 wt % NH4F anodised at 15 V,
and (f) the nanotubes annealed at 500 ºC for 3 h ............................................................ 74
Fig. 4.3 (a) Illustration of an electrochemical cell that indicates the electrolyte ions
species, (b) SEM image of top view of nanoporous TiO2-ZrO2-ZrTiO4 fabricated on a
Ti50Zr alloy after anodisation for 15 min in 0.1 wt % NH4F, 5V, and c) SEM image of
top view of TiO2-ZrO2-ZrTiO4 nanotubes fabricated on Ti50Zr alloy after anodisation
for 2.75 h in 0.3 wt % NH4F, 5V .................................................................................... 76
Fig. 4.4 Histograms of as-formed and annealed TiO2-ZrO2-ZrTiO4 nanotube parameters
anodised at 20 V for 2 h for different concentrations of fluorine anion (F-): (a,d) inner
diameter (Di), (b,e) outer diameter (Do), and (c,f) wall thickness (Wt), Note: The
nanotube size distribution graphs were generated from100 nanotubes on different
positions for each of three samples (300 measurements)................................................ 79
Fig. 4.5 Histograms of TiO2-ZrO2-ZrTiO4 nanotube parameters anodised in 0.5 wt %
NH4F electrolyte for 2 h at different applied potentials: (a, b) inner diameter (Di) as-
formed and after annealing, (c, d) outer diameter (Do) as-formed and after annealing, (e,
f) wall thickness (Wt) as-formed and after annealing, respectively ................................ 83
Fig. 4.6 Mean nanotube size of Ti50Zr and CP-Ti as a function of CF-: (a) Di for As-
formed, (b) Di for Annealed (c) Do for as-formed, (d) Do for Annealed, (e) Wt for as-
formed and (f) Wt for Annealed ...................................................................................... 84
Fig. 4.7 Mean nanotube size of Ti50Zr and CP-Ti as a function of applied potential: (a)
Di for as-formed, (b) Di for Annealed (c) Do for as-formed, (d) Do for Annealed, (e) Wt
for as-formed and (f) Wt for Annealed ............................................................................ 85
Fig. 4.8 The effect on nanotube length by: a) anodisation time and b) applied potential
......................................................................................................................................... 86
Fig. 4.9 SEM images of nanotubular layer fabricated at a) 20V - as-formed, b) 20V -
annealed, c) 25V - as-formed, d) 25 V - annealed, e) 30 V - as-formed, f) 30V -
annealed, g) 35 V - as-formed and h) 35 V - annealed ................................................... 87
xv
Fig. 4.10 Histogram and distribution of Di, Do and Wt of as-formed and annealed TiO2-
ZrO2-ZrTiO4 nanotubes fabricated at a,b) 20 V, c,d) 25 V, e,f) 30 V and g,h) 35 V ..... 90
Fig. 4.11 Histogram and fitted normal and Weibull distribution of Di of as-formed
nanotubular TiO2-ZrO2-ZrTiO4 fabricated at: a) 20 V, b) 25 V, c) 30 V, d) 35 V and e)
normal distribution of all four conditions ....................................................................... 93
Fig. 4.12 Histogram and fitted normal and Weibull distribution of Di, Do and Wt of
annealed nanotubular TiO2-ZrO2-ZrTiO4 fabricated at a) 10 V, b) 15 V, c) 20 V .......... 94
Fig. 4.13 EDS analysis for the (a) top and (b) bottom of TiO2-ZrO2-ZrTiO4 nanotubes
formed in 0.5 wt % NH4F at 20 V after 2 h .................................................................... 95
Fig. 4.14 The mean roughness (Sa) and the mean water contact angle (W.C.A.) of the
nanotubular surfaces of Ti50Zr alloy as a function of: (a), (b) F- concentration and (c),
(d) applied potential, respectively ................................................................................... 96
Fig. 4.15 The mean roughness (Sa) and the mean water contact angle (W.C.A.) of the
nanotubular surfaces of CP-Ti as a function of: (a), (b) F- concentration and (c), (d)
applied potential, respectively ......................................................................................... 96
Fig. 4.16 Loading-unloading forces versus the nanoindentation depths of a) as-formed
TiO2-ZrO2-ZrTiO4 and b) annealed TiO2-ZrO2-ZrTiO4 fabricated at the applied potential
20 to 35 V ...................................................................................................................... 102
Fig. 4.17 The nano mechanical properties of: a) hardness, b) reduced elastic Modulus
and c) elastic Modulus of TiO2-ZrO2-ZrTiO4 nanotubes fabricated at different applied
potential ......................................................................................................................... 103
Fig. 4.18 XRD patterns of the nanotube samples fabricated on Ti50Zr (α and β phases)
via anodisation. (a) as-formed amorphous TiO2 and ZrTiO4 and orthorhombic ZrO2; (b)
annealed at 500 °C for 3 h tetragonal anatase, srilankite (a mixture of orthorhombic
TiO2 and ZrO2) and orthorhombic ZrTiO4 .................................................................... 105
Fig. 4.19 XPS spectra for the as-formed and the annealed TiO2-ZrO2-ZrTiO4 nanotubes.
(a) O1s, (b) Ti 2p, and (c) Zr 3d .................................................................................... 106
Fig. 5.1 SEM images of top view of nanotubes fabricated at: a) 5 wt % H2O at 20 V, b)
10 wt % H2O at 20V, c) 5 wt % H2O at 30 V and d) 10 wt % H2O at 30V in ethylene
glycol ............................................................................................................................. 114
Fig. 5.2 Histograms of TiO2-ZrO2-ZrTiO4 nanotube parameters anodised in 0.5 wt %
NH4F and ethylene glycol consisting of 5 and 10 wt % H2O for 90 min at 20 and 30V:
xvi
(a, d) inner diameter (Di) of as-formed and annealed nanotubes, (b, e) outer diameter
(Do) of as-formed and annealed nanotubes, (c, f) wall thickness (Wt) of as-formed and
annealed, respectively ................................................................................................... 116
Fig. 5.3 Histogram and fitted normal and Weibull distribution of Di, Do and Wt of as-
formed and annealed nanotubular TiO2-ZrO2-ZrTiO4 fabricated at a,e) 5 wt % H2O, 20
V, b, f) 10 wt % H2O, 20 V c, g) 5 wt % H2O, 30 V, d,h) 10 wt % H2O, 30 V ............ 120
Fig. 5.4 Mean nanotube sizes of Di, Do and Wt: a,c) as-formed and b,d) annealed TiO2-
ZrO2-ZrTiO4 and TiO2 nanotubes fabricated in two different water contents and two
different applied potentials in ethylene glycol .............................................................. 122
Fig. 5.5 EDS analysis for: (a) top and (b) bottom of TiO2-ZrO2-ZrTiO4 nanotubes
formed in 0.5 wt % NH4F, 5 wt % H2O in ethylene glycol at 20 V after 90 min ......... 123
Fig. 5. 6 The effect on TiO2-ZrO2-ZrTiO4 nanotubes length formed in organic
electrolyte by: a) anodisation time and b) applied potential ......................................... 124
Fig. 5.7 XRD patterns of the nanotube samples fabricated on Ti50Zr (α and β phases)
via anodisation in non-aqueous electrolyte. (a) as-formed amorphous TiO2 and ZrTiO4
and orthorhombic ZrO2; (b) annealed at 500 °C for 2 h tetragonal anatase, rutile,
srilankite (a mixture of orthorhombic TiO2 and ZrO2) and orthorhombic ZrTiO4 ....... 125
Fig. 5.8 a), c) The mean roughness (Sa), (Sq) and b), d) the mean water contact angle
(W.C.A.) of the nanotubular surfaces of TiO2-ZrO2-ZrTiO4 and TiO2 respectively, Note:
Each data point is an average of five measurements..................................................... 126
Fig. 5.9 Loading-unloading forces versus the nanoindentation depths of a) as-formed
TiO2-ZrO2-ZrTiO4 and b) annealed TiO2-ZrO2-ZrTiO4 fabricated at 5 and 10 wt % H2O
and the applied potential 20 and 30 V ........................................................................... 131
Fig. 5.10 The nano mechanical properties of of TiO2-ZrO2-ZrTiO4 nanotubes fabricated
at different water content and applied potential: a) hardness, b) reduced elastic Modulus
and c) elastic Modulus .................................................................................................. 132
Fig. 6.1 Mean nanotube size as a function of applied potential: (a) Ti50Zr - as-formed,
(b) Ti50Zr - annealed (c) CP-Ti - as-formed, (d) CP-Ti - annealed ............................. 140
Fig. 6.2 The mean roughness (Sa) of the nanotubular surfaces as a function of the
applied potential: (a) Ti50Zr alloy, (c) CP-Ti; and the mean water contact angle
(W.C.A.) of the nanotubular surfaces as a function of the applied potential: (b) Ti50Zr
alloy and (d) CP-Ti, Note: Each data point is an average of five measurements ......... 141
xvii
Fig. 6.3 Mineralisation of HA on the nanotubular surface of TiO2-ZrO2-ZrTiO4
fabricated at a), b) 20 V (Di = 37 nm)-as-formed, c), d) 20V (Di = 40 nm)- annealed, e),
f) 25 V (Di = 64 nm)-as-formed, g), h) 25V (Di = 59 nm)-annealed, i), j) 30 V (Di = 76
nm)-as-formed, k), l) 30V (Di = 64 nm)-annealed, m), n) 35 V (Di = 81 nm)-as-formed
and o), p) 35V (Di = 82 nm)-annealed .......................................................................... 147
Fig. 6.4 Mineralisation of HA on the nanotubular surface of TiO2 fabricated at a), b) 20
V (Di = 43 nm)-as-formed, c), d) 20V (Di = 51 nm)-annealed, e), f) 25 V (Di = 52 nm)-
as-formed, g), h) 25V (Di = 45 nm)-annealed, i), j) 30 V (Di = 65 nm)-as-formed, k), l)
30V (Di = 53 nm)-annealed, m), n) 35 V (Di = 68 nm)-as-formed and o), p) 35V (Di =
63 nm)-annealed ............................................................................................................ 148
Fig. 6.5 SEM images of cross section of mineralised HA layers on the surface of a)
TiO2-ZrO2-ZrTiO4 nanotubes and b) TiO2 nanotubes fabricated in aqueous electrolyte
....................................................................................................................................... 149
Fig. 6.6 XRD patterns of: a) TiO2-ZrO2-ZrTiO4 nanotubes anodised on Ti50Zr (α and β
phases) and annealed at 500°C for 3 h before and after immersion in m-SBF for 3
weeks, showing tetragonal anatase, srilankite (a mixture of orthorhombic TiO2 and
ZrO2) and orthorhombic ZrTiO4 with mineralised hydroxyapatite, b) TiO2 nanotubes
anodised on CP-Ti (α phase) and annealed at 500 °C for 3 h before and after immersion
in m-SBF for 3 weeks, showing tetragonal anatase with mineralised hydroxyapatite; and
c) hydroxyapatite ........................................................................................................... 150
Fig. 6.7 The percentage of atomic ratio of Ca/P of mineralised HA on four different
nanotubular surfaces of a) Ti50Zr and b) CP-Ti ........................................................... 151
Fig. 6.8 SEM images showing the transition of soluble salt crystal of present ions of m-
SBF into HA on the nanotubular surface of TiO2-ZrO2-ZrTiO4 with increasing
immersion time in m-SBF ............................................................................................. 153
Fig. 6.9 Schematic presentations of the process of bone-like apatite formation on
nanotubular surface in m-SBF ...................................................................................... 153
Fig. 6.10 The pH value of static m-SBF as a function of soaking time for nanotubular
TiO2-ZrO2 -ZrTiO4 fabricated at 20, 25, 30 and 35V .................................................... 154
Fig. 6.11 Loading-unloading forces versus the nanoindentation depths of HA on
nanotubular surface of TiO2-ZrO2-ZrTiO4 fabricated at the applied potential of 35 V 155
xviii
Fig. 6.12 Dimpled surface of TiO2-ZrO2-ZrTiO4 nanotubes after annealing and
removing the upper cracked layer a) 5 wt % H2O - 20 V, b) 10 wt % H2O - 20 V, c) 5 wt
% H2O - 30 V and d) 10 wt % H2O - 30 V ................................................................... 156
Fig. 6.13 a) The mean roughness (Sa), (Sq) and the mean water contact angle (W.C.A.)
of the nanotubular and dimpled surface of TiO2-ZrO2-ZrTiO4, respectively, Note: Each
data point is an average of five measurements.............................................................. 157
Fig. 6.14 SEM images of mineralisation of HA on the as-formed TiO2-ZrO2-ZrTiO4
nanotubes and dimpled surface fabricated in non-aqueous electrolyte: a, b) and c, d)
high and low magnification-5 wt % H2O-20V, e, f) and g, h) high and low
magnification-10 wt % H2O-20V, i, j) and k, l) high and low magnification-5 wt %
H2O-30V and m, n) and o, p) high and low magnification-10 wt % H2O-30V ............ 159
Fig. 6.15 SEM images of mineralisation of HA on the as-formed and annealed TiO2
nanotubes fabricated in non-aqueous electrolyte: a, b) and c, d) high and low
magnification-5 wt % H2O-20V, e, f) and g, h) high and low magnification-10 wt %
H2O-20V, i, j) and k, l) high and low magnification-5 wt % H2O-30V and m, n) and o,
p) high and low magnification-10 wt % H2O-30V ....................................................... 160
Fig. 6.16 SEM images of cross section of mineralised HA layers on the surface of a)
TiO2-ZrO2-ZrTiO4 nanotubes and b) TiO2 nanotubes fabricated in non-aqueous
electrolyte ...................................................................................................................... 161
Fig. 6.17 XRD patterns of the nanotube samples fabricated on Ti50Zr (α and β phases)
via anodisation in non-aqueous electrolyte, annealed dimpled surface (at 500 °C for 2 h)
of tetragonal anatase, rutile, srilankite .......................................................................... 162
Fig. 6.18 The percentage of atomic ratio of Ca/P of mineralised HA on a) as-formed and
annealed dimpled surface of TiO2-ZrO2-ZrTiO4 nanotubes and b) as-formed and
annealed TiO2 nanotubes fabricated in non-aqueous electrolyte .................................. 163
Fig. 6.19 The pH value of static m-SBF as a function of soaking time for nanotubular
TiO2-ZrO2-ZrTiO4 in a non-aqueous electrolyte with 5 and 10 wt % H2O at 20 and 30 V
....................................................................................................................................... 164
Fig. 6.20 Loading-unloading forces versus the nanoindentation depths of HA on
nanotubular surface of TiO2-ZrO2-ZrTiO4 fabricated at 10 wt % H2O and the applied
potential of 30 V............................................................................................................ 165
xix
Fig. 7.1 SEM images of annealed TiO2-ZrO2-ZrTiO4 nanotubes fabricated at a) 10 V, b)
15 V, c) 20 V, and d) 25 V, e) 30 V and f) 35 V........................................................... 174
Fig. 7.2 Mean nanotube size of Di, Do and Wt for annealed TiO2-ZrO2-ZrTiO4 nanotubes
anodised on Ti50Zr as a function of applied potential .................................................. 176
Fig. 7.3 Illustration of changing roughness parameters of Sa and Sq via changing Di of
TiO2-ZrO2-ZrTiO4 nanotubes anodised on Ti50Zr ....................................................... 177
Fig. 7.4 SEM images of SaOS2 cells cultured after 24 h on three different pore sizes of
TiO2-ZrO2-ZrTiO4 nanotubes fabricated at a) 10 V- low magnification, b) 10 V- higher
magnification, c) 15 V- low magnification, d) 15 V- higher magnification, e) 20 V- low
magnification and f) 20 V higher magnification ........................................................... 182
Fig. 7.5 SEM images of SaOS2 cells cultured after 24 h on three different pore sizes of
TiO2-ZrO2-ZrTiO4 nanotubes fabricated at a) 25 V- low magnification, b) 25 V- higher
magnification, c) 30 V- low magnification, d) 30 V- higher magnification, e) 35 V- low
magnification and f) 35 V higher magnification ........................................................... 183
Fig. 7.6 Schematic illustration of exploring tools of an osteoblast cell on nanotubular
surface ........................................................................................................................... 184
Fig. 7.7 Proposed illustration of the position of integrins on a) Di = 18 ± 6 nm best
nanospacing, b) rough surface of Di = 59 ± 17nm and c) Wt = 24 ± 7 nm most existing
of proteins ..................................................................................................................... 186
Fig. 7.8 Illustration of a) the length of filopodia, b) surface area and c) height of SaOS2
cells after 1 day and 7 days grown on different nanotube sizes (ρ < 0.05) ................... 189
Fig. 7.9 SEM images of SaOS2 cells cultured for 7 days on the four different pore sizes
of TiO2-ZrO2-ZrTiO4 nanotubes fabricated at a) 10 V- low magnification, b) 10 V-
higher magnification, c) 15 V- low magnification, d) 15 V- higher magnification, e) 20
V- low magnification and f) 20 V higher magnification ............................................... 191
Fig. 7.10 SEM images of SaOS2 cells cultured for 7 days on the four different pore
sizes of TiO2-ZrO2-ZrTiO4 nanotubes fabricated at a) 25 V- low magnification, b) 25 V-
higher magnification, c) 30 V- low magnification, d) 30 V- higher magnification, e) 35
V- low magnification and f) 35 V higher magnification ............................................... 192
Fig. 7.11 The percentage of cells attached to the surface of different pore sizes of TiO2-
ZrO2-ZrTiO4 nanotubes after 7 days of cell seeding using MTS assay (ρ < 0.05) ....... 194
xx
Fig. 7.12 Confocal images of stained SaOS2 cells presenting nuclei (blue) and actin
stress fibres (red) after 7 days on TiO2-ZrO2-ZrTiO4 nanotubes with a) Di = 18 nm, b)
Di =30 nm, c) Di =40 nm, d) Di = 59 nm, e) Di = 64 nm and f) Di = 82 nm ................. 195
Fig. 7.13 SEM images of SaOS2 cells cultured for 24 h on the TiO2-ZrO2-ZrTiO4
nanotubes (Di = 40 nm) a) with reduced contamination - low magnification, b) with
reduced contamination - higher magnification, c) with post-treated by HA - low
magnification and d) with post-treated by HA - higher magnification ......................... 196
Fig. 7.14 Illustration of a) the length of filopodia, b) surface area and c) height of
SaOS2 cells after 1 day and 7 days grown on different post-treated nanotubular surface
with the same Di ............................................................................................................ 198
Fig. 7.15 The percentage of cells attached to the different post-treated nanotubular
surface of TiO2-ZrO2-ZrTiO4 with the same Di after 7 days of cell seeding using MTS
assay .............................................................................................................................. 199
Fig. 7.16 SEM images of SaOS2 cells cultured for 7 days on the TiO2-ZrO2-ZrTiO4
nanotubes (Di = 40 nm) a) with reduced contamination - low magnification, b) with
reduced contamination - higher magnification, c) with post-treated by HA - low
magnification and d) with post-treated by HA - higher magnification ......................... 200
Fig. 7.17 Confocal images of stained SaOS2 cells presenting nuclei (blue) and actin
stress fibres (red) after 7 days on TiO2-ZrO2-ZrTiO4 nanotubes a) with Di = 40 nm
before post-treating b) reduced contamination, c) with post-treated by HA................. 201
Fig. 7.18 Mean nanotube size of Di, Do and Wt of dimpled surface after annealing and
removing the upper cracked layer fabricated in two different water contents and two
different applied potentials in ethylene glycol, Note: The nanotube size distribution
graphs were generated from 100 nanotubes on different positions for each of three
samples (300 measurements) ........................................................................................ 202
Fig. 7.19 SEM images of SaOS2 cells cultured for 7 days on the four different pore
sizes of TiO2-ZrO2-ZrTiO4 nanotubes fabricated at a) 5 wt % H2O - 20 V, b) 10 wt %
H2O - 20 V, c) 5 wt % H2O - 30 V, d) 10 wt % H2O - 30 V ......................................... 205
Fig. 7.20 Confocal images of stained SaOS2 cells presenting nuclei (blue) and actin
stress fibres (red) after 7 days on dimpled surface with a) Di = 25 nm, b) Di =29 nm, c)
Di =45 nm and Di =49 nm ............................................................................................. 207
xxi
Fig. 8.1 SEM images of nanoporous Ta2O5: (a) top view (b) the bottom view and the
cross section of multilayer, (c) shorted life nanotubes (arrow), (d) the top view of
multilayer formed on a Ta after anodisation for 120 min in 1M H2SO4 + 3.3 wt %
NH4F, 20V, SEM image of nanoporous Nb2O5 (e) top view and (f) cross section formed
on niobium after anodisation for 16 min in 1M H2SO4 + 3.3 wt % NH4F, 20V and SEM
image of ZrO2 nanotubes (g) top view, (h) cross section and (i) bottom view formed on
a zirconium after anodisation for 95 min in 1M (NH4)2SO4 + 0.3 wt % NH4F, pH=5 and
20V ................................................................................................................................ 219
Fig. 8.2 Demonstration of (a,c,e) changes of roughness (Sa) and (b,d,f) water contact
angle of bare tantalum, niobium, zirconium and titanium, after fabrication of
nanotubular and nanoporous layer, and subsequently annealing, respectively, Note:
Each data point is an average of five measurements..................................................... 223
Fig. 8.3 XRD patterns of (a) the nanoporous Ta2O5 and bare tantalum foil, (b) the
nanoporous Nb2O5 and bare niobium foil, (c) the ZrO2 nanotubes and bare zirconium
foil and (d) the TiO2 nanotubes and bare titanium foil ................................................. 229
Fig. 8.4 Illustration of pore size of the nanoporous and nanotubular layers ................. 230
Fig. 8.5 (a) low and (b) high magnifications of HA on as-formed nanoporous Ta2O5; (c)
low and (d) high magnifications of HA on annealed nanoporous Ta2O5 ...................... 231
Fig. 8.6 SEM images of HA on a) as-formed and b) annealed nanoporous Nb2O5 ...... 232
Fig. 8.7 (a) low and (b) high magnifications of HA on as-formed nanotubular ZrO2; (c,)
low and (d) high magnifications of HA on annealed nanotubular ZrO2 ....................... 232
Fig. 8.8 SEM images of cross section of mineralised HA layers on the surface of a)
nanoporous Ta2O5, b) nanoporous Nb2O5, c) nanotubular ZrO2 and d) nanotubular TiO2
....................................................................................................................................... 233
Fig. 8.9 Bioactivity of biocompatible nanoporous and nanotubular oxide metals after 3
weeks in m-SBF at 37 ˚C .............................................................................................. 234
Fig. 8.10 pH value of static m-SBF as a function of soaking time for: a) nanoporous
Ta2O5 and Nb2O5 and b) nanotubular TiO2 and ZrO2 ................................................... 235
xxii
List of tables
Table 2.1 Scale of sizes of the hierarchical bone components, in comparison to the scale
sizes of the implant structures ......................................................................................... 10
Table 2.2 Microstructures, mechanical properties and biological characteristics of
metallic implant materials ............................................................................................... 13
Table 2.3 Anodisation conditions, characteristics and properties of TiO2 nanotubes on
titanium ........................................................................................................................... 30
Table 2.4 Cell responses to different TiO2 nanotubes on titanium and the anodisation
conditions ........................................................................................................................ 47
Table 3.1 Compositions for the aqueous electrolyte used for fabricating nanoporous and
nanotubular layers ........................................................................................................... 60
Table 3.2 The reagents and their amounts for preparing 1000 mL m-SBF .................... 64
Table 4.1 The distribution of Di, Do and Wt according to normal and Weibull
distribution methods ........................................................................................................ 92
Table 4.2 Surface area index SI and roughness amplitude parameters of Sskw and Sku of
the nanotubular surfaces of Ti50Zr and CP-Ti anodised under various conditions, Note:
Each data point is an average of five measurements....................................................... 99
Table 4.3 Calculated surface energy of TiO2-ZrO2-ZrTiO4 nanotubular surfaces
fabricated on Ti50Zr and TiO2 nanotubular surfaces on CP-Ti, Note: Each data point is
an average of five measurements .................................................................................. 101
Table 5.1 The distribution of Di, Do and Wt according to normal and Weibull
distribution methods ...................................................................................................... 118
Table 5.2 Surface area index SI and roughness amplitude parameters of Sskw and Sku of
the nanotubular surfaces of Ti50Zr and CP-Ti anodised under various conditions, Note:
Each data point is an average of five measurements..................................................... 128
xxiii
Table 5.3 Calculated surface energy of as-formed and annealed dimpled surface TiO2-
ZrO2-ZrTiO4 nanotubular surfaces fabricated on Ti50Zr, Note: Each data point is an
average of five measurements ....................................................................................... 130
Table 6.1 Surface area index SI and roughness amplitude parameters of Sq, Sskw and Sku
of the nanotubular surfaces of Ti50Zr and CP-Ti anodised under various conditions,
Note: Each data point is an average of five measurements ........................................... 143
Table 6.2 Calculated surface energies of TiO2-ZrO2-ZrTiO4 nanotubular surfaces
fabricated on Ti50Zr and TiO2 nanotubular surfaces on CP-Ti, Note: Each data point is
an average of five measurements .................................................................................. 145
Table 7.1 Surface index and spacing parameters of TiO2-ZrO2-ZrTiO4 nanotubes
fabricated in different applied potential ........................................................................ 179
Table 7.2 Surface index, spacing parameters and calculated surface energy of TiO2-
ZrO2-ZrTiO4 nanotubes with dimpled surface fabricated in different water contents and
applied potentials, Note: Each data point is an average of five measurements ............ 204
Table 7.3 The percentage of cells attached to the surface of different pore sizes and
roughness parameters of TiO2-ZrO2-ZrTiO4 nanotubes fabricated in organic electrolyte
after 7 days of cell seeding using MTS assay (ρ < 0.05) .............................................. 208
Table 8.1 Surface area and volume index and roughness amplitude parameters of
nanoporous Ta2O5, Nb2O5 and nanotube ZrO2, Note: Each data point is an average of
five measurements ......................................................................................................... 224
Table 8.2 Water contact angle and surface energy of as-formed and annealed
nanoporous Ta2O5, Nb2O5 and nanotube ZrO2 in comparison to its bare metal, Note:
Each data point is an average of five measurements..................................................... 226
xxiv
List of abbreviations
ALP Alkaline phosphatase
ARB Accumulative roll bonding
at % Atomic percentage
BMD Bone mineral density
BSA Bovine serum albumin
Ca/P Atomic percentage of calcium to phosphate ratio
Ca5(PO4, CO3)3(OH) Carbonated apatite
CaP Calcium phosphate
CF- Concentration of fluorine anion
CP-Ti Commercial pure titanium
Di Inner diameter
Do Outer diameter
EB-PVD Electron-beam physical vapour deposition
ECM Extracellular matrix
EDM Electrical discharging machining
EDS Electron dispersive X-ray spectroscopy
FN Fibronectin
HA Hydroxyapatite
HEPES 2-(4-(2-hydroxyethyl)-1-piperazinyl) ethane sulfonic acid
HMDS Hexamethyldisilazane
ISQ Implant stability quotient values
xxv
L/W Length to width ratio
MC3T3-E1 Mouse osteoblast cells
MG63 Human osteosarcoma cells
mMSC Mouse bone marrow MSCs
m-SBF Modified simulated body fluid
MTS assay (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-
sulfophenyl)-2H-tetrazolium) assay
N2 Nitrogen gas
Nb Niobium
Nb2O5 Niobium pentoxide (niobia)
OCP Octacalcium phosphate
PBS Phosphate-buffered saline
PTES 1H, 1H, 2H, 2H-perfluorooctyl-triethoxysilane
RBA Rapid breakdown anodisation
RBF Rotating bending fatigue
real-time PCR Real-time polymerase chain reaction technique
Sa Mean roughness
SaOS-2 Human osteoblast-like cells
SAS.F Surface area of the actin stress fibre
SBF Simulated body fluid
Sc Core void volume (space roughness parameter)
SEM Scanning electron microscope
xxvi
SI Surface index
Sku Kurtosis (amplitude roughness parameter)
Sm Surface material volume(space roughness parameter)
Sp Largest peak height value from the reference surface within the
defined sampling area (amplitude roughness parameter)
Sq Root mean square (RMS) Roughness
Sskw Skewness (amplitude roughness parameter)
Sv Surface void volume (space roughness parameter)
Ta Tantalum
Ta2O5 Tantalum pentoxide (tantala)
TCP Tricalcium phosphate
TEM Transmission electron microscope
Ti Titanium
Ti50Zr Titanium zirconium alloy with 35 atomic percentage or 50 weight
percentage of zirconium
TiO2 Titanium dioxide (titania)
TiO2-ZrO2-ZrTiO4 Titanium-zirconium-zirconium titanate
TLM Ti-5Zr-3Sn-5Mo-15Nb
TNTZ Ti-Nb-Ta-Zr alloy
VI Volume index
Vol % Volume percentage
W.C.A or θW Water contact angle
Wt Wall thickness
xxvii
Wt % Weight percentage
XPS X-ray photoelectron spectroscope
XRD X-ray diffraction
Zr Zirconium
ZrO2 Zirconium dioxide (zirconia)
1
Chapter 1 Introduction
1.1 Overview
The fabrication, characterisation and application of biomaterials is an interesting field
which has been the focus of many studies in recent years. Although there are debates on
the exact kind of substance that can be classified as a biomaterial, in simple terms they
can be defined as implantable medical devices such as metals, ceramics and synthetic
polymers, biopolymers, self-assembled systems, nanoparticles, carbon nanotubes and
quantum dot, drug and gene delivery systems, tissue engineering and cell therapies,
organ printing and cell patterning, nanotechnology based imaging and diagnostic
systems and microelectronic devices, engineered tissues, cells, organs and even viruses
[1]. One of the applications of biomaterials is the repair of bone tissue, known as
orthopaedic biomaterials or implants [2]. These kind of biomaterials need to have
suitable mechanical and biological properties. Among different biomaterials used for
implants, such as ceramics, polymers, composites and natural products; metal implants
are preferred for load-bearing applications. Metal implants also exhibit excellent
mechanical properties such as Young’s modulus, tensile strength, ductility, fatigue life,
and wear resistance [3]. In addition, they have excellent chemical properties such as
corrosion resistance, biocompatibility and the ability to fuse and harmonise with other
implant materials [4]. These properties make them more suitable for long-term use in
hard tissue applications such as hip and knee joints [4, 5]. Stainless steels and cobalt-
chromium-molybdenum alloys are some examples alongside the more biocompatible
titanium and titanium alloys [2]. However, as bone is alive, an implant has to be
accepted by the bone cells in order to bond with it and to prevent loosening and
inflammation when it is inserted close to the bone. Surface modifications or treatments
can change the surface of the metal implant making it more stable and agreeable for
bone cells.
Surface modifications of metal implants play a major role in biomedical research. One
of the operations of surface modification techniques, either mechanical, physical or
chemical is the addition of substance to the surface such as plasma spray, physical
vapour deposition (PVD) and surface immobilisation of functional molecules. The
interface of the implant and body fluid can also be converted using ion implantation and
electrochemical oxidation. Another way to modify a surface is by removing undesirable
2
substances from the surface by surface machining, grit blasting and acid etching [6].
Based on the functionality and application of the metal implant, one or more surface
modification can be used. Considering the ability of titanium and titanium alloys to
form a protective oxide layer on their surfaces, electrochemical surface treatment is
utilised as a preferable technique.
Titanium, one of the best known transition metals, is of great interest in biomedical
research. It has an incompletely filled d shell in its electronic structure, which allows the
forming of solid solutions with other alloying elements. One of these elements is
zirconium which produces non-toxic β alloys with titanium. This β type alloy offers the
unique characteristic of mechanical biocompatibility such as low elastic modulus,
superior corrosion resistance, and good osseointegration [7, 8]. Titanium shape memory
alloys and porous titanium alloy are also of great interest because of their shape memory
function, or super-elastic function and bone tissue ingrowth into the porous network,
respectively[4, 9]. These metal implants are applied as plates, nails, screws and endo-
prostheses [8]. The protective oxide layer which forms naturally on the surface of
titanium and titanium alloy makes them bioinert. Therefore, electrochemical surface
modification such as anodisation is one of the suitable techniques available to make
them bioactive.
The significance of a surface treatment such as anodic oxidation, [10] which provides
the possibility of controlled nano-scale fabrication with suitable physico-chemical
properties on the metal surface, has been reported [11]. Highly ordered nanoporous
arrays of titanium dioxide that form on the titanium surface by anodic oxidation are
receiving increasing research interest due to their effectiveness in promoting
osseointegration. Uniform and controllable nano-patterned surfaces of titanium can be
fabricated using electrochemical anodic oxidation, which is an economical , simple and
versatile technique [12]. By applying a constant potential to titanium as a working
electrode in an electrochemical cell with a platinum counter electrode and a standard
reference electrode, the titania nanotubes can be obtained. Different TiO2 nanotube
diameters, wall thicknesses and lengths can be achieved by controlling the condition of
the anodising process, such as the applied potential, current density, anodisation time,
concentration of fluorine anion and type (aqueous and non-aqueous), pH value and the
3
viscosity of the electrolyte [13]. Such nano-scale surfaces exhibit similar properties to
those of physiological bone.
The important factors for cell response can be better controlled by understanding the
mechanism of cell attachment to the biomaterial surface. Various populations of cells,
cytokines, growth factors and the extracellular matrix affect the process of tissue
healing around implants. For example, chemotactic factors are essential for cell
recruitment and cell adhesion, and growth factors and cytokines are effective for the
proliferation and differentiation of cells. Protein adsorption at the implant surface
initiates interactions between cells and implants through a complex series of adsorption
and displacement steps. The interaction of cell and implant is followed by the
attachment phase, which is governed mainly by van der Waals forces. The next step,
termed the adhesion phase, is surface anchoring through fibronectin and vitronectin for
the formation of focal points at cell-membrane integrins. Filopodia and finger-like
protrusions are then produced to enable sensing of the optimum anchorage and
spreading toward the surface [14, 15].
In the specific instance of biomaterial applications, the highly ordered nanostructured
TiO2 promotes osseointegration because the size of integrins and focal contacts of cell-
implant interactions is also in the nano scale (~ 10 nm) [11, 16]. The effects of different
inner and outer diameters of nanotubes and the spaces between them on bone cells
during cell culture have been studied extensively [17-20]. The crystalline form of TiO2
nanotubes exhibited an encouraging effect on cell behaviour due to their enhanced
hydrophilic properties [21]. Furthermore, growing hydroxyapatite on the surface of
these so-formed nanotubes would improve cell behavior such as adhesion, proliferation,
morphology, migration, survival and differentiation due to beneficial changes in the
topographical and chemistry properties [22-25]. Cell biological responses of osteoblasts
on anodised titania zirconia nanotubes [26], the effect of surface energy for its
functionality [27], as well as the antibacterial properties of different sizes of titania
zirconia nanotubes [28] have also been investigated.
Several in vitro studies have examined the optimum nanotopography for cell response
and suggested that the optimum diameter of TiO2 nanotubes is less than 100 nm. The
15 nm diameter nanotubes revealed the best cell response; although there is controversy
[19, 20] regarding the optimum diameter required. The behaviour of the bone cells on
4
the nanotubular of oxide layer on the surface of Ti50Zr needs to be clarified.
Specifically, further studies are required to determine the optimum size, morphology,
physical and chemical properties of nanotubes on the biocompatible titanium-zirconium
(Ti50Zr) binary alloy for the adhesion of bone cell sensing elements. The current work
aims to improve the physicochemical and other properties of the nanotubular layer by
fabricating it on the surface of β alloy of Ti50Zr.
1.2 Thesis objective
This thesis was aimed at imparting new functionalities to the biocompatible Ti50Zr
alloy through surface modification. A nanotubular layer was anodised on the alloy with
enhanced bioactivity and biocompatibility because the nanoscale characteristics of the
nanotubes is able to interlock with bone cells, therefore improving cell adhesion,
proliferation and growth on the nanotubular surface. Extensive research has been
undertaken to fabricate and characterise the nanotubular layer on the surface of CP-Ti
and assess its bioactivity and biocompatibility. Although some research has been
undertaken to fabricate the nanotubular layer on the surface of Ti50Zr, few attempt have
been made to characterise the physical, chemical and mechanical properties of
nanotubular layer of TiO2-ZrO2-ZrTiO4 in respect of their influences on their bioactivity
and bone cell integration.
To accomplish the aforementioned objectives the following tasks were completed:
- Two different types of electrolytes were used to fabricate nanotubes with
different morphologies and physicochemical and mechanical properties.
- The effects of the applied potential and the concentration of fluoride ions in the
electrolyte on the morphology and other characteristics of the nanotubes were
systematically studied.
- The ability to induce hydroxyapatite formation was evaluated by immersing the
substrates with a nanotubular surface in the modified simulated body fluid.
- The cell adhesion was assessed by seeding osteoblast cells on the nanotubular
surfaces of Ti50Zr and CP-Ti.
Major contributions of the thesis include:
5
The TiO2-ZrO2-ZrTiO4 nanotubes fabricated on the surface of Ti50Zr alloy not only
revealed a distribution of nanotube sizes following the needle-like microstructure of the
alloy, but also exhibited a variety of nanotube height. These difference in the height of
nanotubes resulted in an interesting surface topography with a range of amplitude and
spacing roughness parameter. The link between roughness parameter and hydrophilic
properties of the nanotubular layer were evaluated. Hereby the surface of Ti50Zr alloy
was nanoengineered by fabricating the TiO2-ZrO2-ZrTiO4 nanotubes while the factors
which influence cell attachment were taken into consideration. This demonstrates that
the nanotubular layer not only patterned the surface in nano-scale but also created a
special topography in micro-scale. The surface chemistry and mechanical properties of
the substrate are also changed by the nanotubular layer. The TiO2-ZrO2-ZrTiO4
nanotubes were more bioactive than TiO2 nanotubes. The bone cell attachment
significantly increased on two different nanotubular layer of TiO2-ZrO2-ZrTiO4
fabricated in aqueous and non-aqueous electrolyte.
1.3 Thesis structure
Chapter 2 reviews the literature on anodisation of titanium, titanium alloy and transition
metal alloying elements to fabricate nanoporous and nanotubular layer after an
introduction of bone structure, metal implants and their surface modification. It presents
parameters which affect the morphology, topography, physical and chemical properties
of nanotubes. Then it describes in simple terms the mechanism of cell adhesion and the
factors that influence the adhesion of the bone cells on implant surface. Finally, it
highlights the characters of these nanotubular layers which influence bone cell
behaviour. Chapter 3 provides a methodological description of the materials and
methods for fabrication, characterisation and performance of the nanotubular layer. The
materials and methods of individual chapters are described as required in the relevant
chapter to avoid over-complication.
Chapter 4 presents the effect of anodisation parameters on the characteristics of the
TiO2-ZrO2-ZrTiO4 nanotubes fabricated in aqueous electrolyte such as inner, outer
diameter, wall thickness, roughness parameters, mechanical and hydrophilic properties
and surface energy. Chapter 5 presents the effect of anodisation parameters on the
characteristics of the TiO2-ZrO2-ZrTiO4 nanotubes fabricated in non-aqueous electrolyte
such as inner, outer diameter, wall thickness, roughness parameters, mechanical and
6
hydrophilic properties and surface energy. These characteristics are compared with TiO2
nanotubes fabricated on CP-Ti in the same condition to outline the significance of TiO2-
ZrO2-ZrTiO4 nanotubes.
Chapter 6 studies the effect of the characteristics of the nanotubular layer fabricated in
aqueous and non-aqueous electrolyte on their ability to induce hydroxyapatite. These
characteristics are compared with TiO2 nanotubes fabricated on CP-Ti in the same
condition. Chapter 7 investigates the cell adhesion on the surface of these nanotubular
layers and the influence of the morphology of the layer as well as the roughness
parameter, hydrophilic properties and the role of post-treatment on the percentage of
cell attachments is studied.
Chapter 8 presents the nanoporous and nanotubular layer fabricated on the surface of
some biocompatible transition metal such as tantalum, niobium and zirconium in
comparison to titanium. The characteristics of these nanoporous and nanotubular layer
such as roughness parameter, hydrophilic properties and surface energy is studied. The
ability of such a layer to induce hydroxyapatite is also assessed.
7
Chapter 2 Literature review
2.1 Introduction
Metal implants are the best choice for the long-term replacement of hard tissue, such as
hip and knee joints, because of their excellent mechanical properties. Titanium and its
alloys, due to their self-organised oxide layer that protects the surface from corrosion
and prevents ion release, are widely accepted as biocompatible metal implants. They
must also be accepted by bone cells, bonding with and growing on them to prevent
loosening. Surface modification is essential for the promotion of the osseointegration of
these biomaterials. Nanotubes fabricated on the surface of metal implants by
anodisation are currently receiving increasing attention for surface modification. The
response of bone cells to implant materials depends on the topography,
physicochemistry, mechanics and electronics of the implant surface and this influences
cell behavior, such as adhesion, proliferation, shape, migration, survival and
differentiation; for example the existing anions on the surface of a titanium implant
make it negative and this affects the interaction with negative fibronectin (FN).
Although optimal nanosize of reproducible titania nanotubes have not been found
significantly according to different protocols in the studies cell response has been
shown that were more sensitive on titania nanotubes with nanometer diameter and inter
space. By annealing amorphous TiO2 nanotubes change to a crystalline form and
become more hydrophilic, resulting in an encouraging effect on cell behaviour. The
crystalline size and thickness of the bone-like apatite that forms on the titania nanotubes
after implantation are also affected by the diameter and shape. This chapter provides an
overview of the employment of the anodisation for nanotubes fabricated on the surface
of titanium, titanium alloys and titanium alloying metals such as niobium, tantalum and
zirconium metal implants. It explains anodic oxidation and the manner by which
nanotubes form on the surface of the metals. It then assesses this topical research to
indicate how changes in anodising conditions influence nanotube characteristics such as
tube diameters and nanotube-layer thickness. Finally it describes how changes in
nanotube morphologies, such as the tube diameter, the thickness of the nanotube layer,
and the crystalline structure, influence the response of cells.
8
2.1.1 Bone: structure, composition and properties
Bone, as with other parts of our body, becomes damaged or weakened because of age,
accidents or disease. This damage, which includes bone fractures, low back pain,
osteoporosis, scoliosis and other musculoskeletal problems, usually occurs in elderly
people but not exclusively. Biomaterials such as implants are used for repairing injured
bones, cartilage or ligaments and tendons [2].
Weiner et al. [29] have described the basic bone composition that consists mostly of
fibrous protein collagen, carbonated apatite (Ca5(PO4, CO3)3(OH)) and water. The
crystal size and the proportion of these components change over time. As a result,
younger bones replace older bones. Uddin et al. [30] have emphasised that bones are
tissues that are alive and growing. Bone forming cells such as osteoblasts are
responsible for generating the synthesis and deposition of calcium phosphate crystals
that are required to confer hardness and strength in biomineralisation. A schematic
illustration of the hierarchical structure of cortical bone is shown in Fig. 2.1. It can be
seen that cortical bone contains many different structures that exist on several levels of
scale starting from sub-nanostructures. Mour et al. [31] described bone as a solid
material that is highly porous on a micrometre scale. Bone is a viscoelastic material due
to these pores, which are filled with fluid and cells; osteoblasts, osteoclasts, osteocytes
and bone lining cells that are regenerative. Bone, moreover, has values of compression
strength several times higher than, for example, concrete, but its low density is in the
range of aluminium. Table 2.1 lists the size scales of the hierarchical bone components,
in comparison to the size scales of the implant structures.
9
Fig. 2.1 Hierarchical structure of cortical bone (adapted from [32])
10
Table 2.1 Scale of sizes of the hierarchical bone components, in comparison to the scale
sizes of the implant structures
Structures Size
Ref. Micrometer Nanometer
Tissue
component
Osteon 10-500 - [32]
Osteoclast (lacunae) 100 or more - [33]
Osteoblast 20-30 - [33]
Lamella 3-7 - [32]
Collagen fibril 0.5 - [32]
Cell membrane - 10-100 [34]
Plate-like apatite crystals - 50×25×3 [32]
Integrins - 8-12 [35]
Proteins - 1-10 [34]
Collagen molecule - 1 [32]
Amino acids - 0.1-1 [34]
Water molecule - 0.1 [34]
Metal
Implant
Roughness of the surface >100 - [34]
titanium grain sizes 10-20 - [36]
Metal oxide - 5-15 [34]
Atom - 0.1 [34]
11
2.1.2 Bone implant materials
There are different types of implant materials, such as polymers, ceramics, metals,
composites and natural products. Metal implants can be load bearing up to several
decades and consequently are used as endoprostheses because of their mechanical
attributes. However, they need to be inserted into the human architecture and become
fixed and stable in the surrounding natural tissue. Furthermore, these implants need to
be bioinert within the highly corrosive and demanding environment of the human body.
Polymers, in comparison to the metals, give rise to inflammation because of monomers
intrinsic to their structure or that become available through infection; thereby creating
likelihood of degradation. However, polymers demonstrate excellent primary fixation
[2].
Ceramics and bioglass exhibit lower fracture toughness and higher elastic modulus than
bone, and also demonstrate property variations with respect to their formulation [37].
Their mechanical and biological properties depend on many factors during synthesising
such as maximum temperature, duration of the thermal steps, purity of the powder, size
and distribution of the grains and porosity; many of which cannot be accurately
controlled [2].
Biomaterials have progressed through three generations from the first, whose
particularities were matching its physical properties as a tissue replacement with the
least toxicity or biological inertness to the second generation that demonstrate bioactive
behaviour [38]. Surface treatments have been used to improve the bioactive nature of
these biomaterials, especially metals such as titanium and titanium alloys that change
the physico-chemical, mechanical and electrical properties of their surfaces [7]. The
third generation biomaterials are intended to promote specific cellular responses at the
molecular level [38]. Most of the metal implants used are stainless steels, cobalt
chromium alloy, and titanium and its alloys. SUS 316L stainless steel, with a Young’s
modulus of approximately 160 GPa, is the only stainless steel that is biocompatible;
however, its wear resistance is poor in comparison with other metal implants. When a
stainless steel alloy consists of nickel, it may possibly be allergenic [4] since the release
of metal ions can cause an adverse reaction due to corrosion and the soluble corrosion
products. Similar to stainless steel, cobalt alloys consist of nickel and chromium, and
consequently fabrication of nickel free stainless steels and cobalt chromium alloys is
12
under investigation since its corrosion resistance and mechanical properties are superior,
with the exception of the elastic modulus that is the same range as stainless steel [4].
When a metal implant material is able to form a stable and compact oxide layer that can
rebuild itself, then it can be described as bioinert since this surface can prevent
transmission of undesirable ions [7]. Other properties of a metal, which endow its
mechanical biocompatibility, are Young’s modulus, tensile strength, ductility, fatigue
life, fretting fatigue life and wear resistance [3]. As a result of the contact between the
inserted implant and bone, cyclic loading wear loss and fretting fatigue may occur and
the implant may become loose. Some oxide layers on metals increase the wear
resistance. A metal implant that has a lower Young’s modulus can homogeneously
transfer stress between itself and bone. However, as the modulus approaches that of the
bone, the strength of the implant will be lower than bone, and the possibility of failure
under large shear deformation will increase.
In vivo data show a smooth fracture callus for metals with a lower Young’s modulus
than that of bone. An in vivo fatigue test of implants is difficult due to the need to gather
data from tests that identically replicate in vivo conditions [3]. A standardised in vitro
fatigue test is therefore used. This includes tension/compression, bending, torsion, and
rotating bending fatigue (RBF) testing of the metal implant, which is then compared
with Ti6Al4V as a standard material [5]. Titanium and titanium alloys are more
biocompatible than other implant metals for the abovementioned reasons and these
metals have demonstrated ability to integrate with bone. The microstructures,
mechanical properties and biological characteristics of commonly used metallic implant
materials are summarised in Table 2.2.
13
Table 2.2 Microstructures, mechanical properties and biological characteristics of metallic implant materials
Alloy designation Microstructure
Nominal composition (wt%)
Ultimate
strength (MPa)
Elastic
modulus
E (GPa)
Yield
strength
(MPa)
Ref. Log polarisation resistance- Distinct form
Al/Cr V/Ni Mn/Ta Zr/Nb Others
Stainless steel 316L Austenite Cr=16-18
Ni=10-
14 Mn=2 - Mo=2-3
465-950 200 170-750 [5, 39]
7-8/ sequestration
Co-25Cr-6Mo Austenite(fcc)+hcp Cr=27-30/ Al=0.1 - - - Mo=5-7
600-1795 200-230 275-1585 [5, 39]
Titanium grade 2
(ASTM) Unalloyed
- - - - - 345 102.7 275 [39, 40]
- - - - -
Ti-6Al-4V α/β
Al=6 4 - - Fe=0.3
960-970 110 850-900 [5, 39] 4-5/ sequestration
5-6/
toxicity - -
4-5/
sequestration
Ti-6Al-7Nb α/β Al=6 - - Nb=7 -
900-1050 114 880-950 [39, 40] 4-5/ sequestration - - 6-7/
14
Inertness
Ti-Zr Cast ά/β
- - - Zr=50 -
900 - - [5, 39] - - -
7-8/
Inertness -
Ti-30Ta α″
- - Ta=29.6 - -
587 69 - [39, 41] - -
6-8
inertness - -
Ti-30Nb
- - - Nb=30 -
> 600 More than
120 530 [39, 42]
- - - 6-7/
Inertness -
Ti-29Nb-13Ta-4.6Zr Metastable β
- - Ta=13 Zr=4.6/N
b=29 -
911 80 864 [39, 40]
- - 6-8
inertness
7-8 and
6-8
Inertness
-
Bone Viscoelastic
composite 90-140 10-40 - [7, 39]
15
Researchers have paid more attention to titanium and its alloys for load bearing and
long term implants because of their good fatigue strength, relatively low elastic
modulus, low density, excellent corrosion resistance and biocompatibility; the last
property being reflected in low release of ions [39, 43]. Pure titanium and Ti-6Al-4V
ELI (Extra Low level of Interstitial content) are the prime metals used for biomedical
applications. However, aluminium is classified as belonging to a group of materials that
forms a capsule, or scar tissue, and vanadium is cytotoxic, and both can have
mutagenicity, cytotoxicity and allergic reactions. Research is active in developing
biocompatible Ti-Nb-Ta-Zr alloys. Recent β type titanium alloys have a lower elastic
modulus (Young’s modulus approximately 50 GPa) than commercial pure titanium and
Ti-6Al-4V (Young’s modulus approximately 105 and 110 GPa, respectively), which are
α and α+β types. The stability of β phase in titanium alloys depends on the alloying
element ratio and the β-transus temperature of the alloy. It has been established that
niobium and tantalum act as β-stabilisers and β phase exhibits a lower elastic modulus
in titanium alloys. In binary titanium alloys, it has been shown when β stabiliser
elements increase, the stability of β phase and also elastic modulus increase [4, 44-46].
2.1.3 Surface treatment for implant materials
The surface needs modification to optimise the properties of an implant and maximise
bioactivity when interfacing with natural tissue. The biological characteristics of
implants can be enhanced by adding material with desired properties, changing the
composition or removing unwanted material from the implant surface. These methods
are known as surface treatment or modification and can be classified into four
categories: mechanical, physical, chemical and biochemical surface modifications.
Machining, grinding, polishing and blasting are mechanical surface modifications.
Physical surface treatments include thermal spray, physical vapour deposition, ion
implantation and deposition, and glow discharge plasma. Chemical surface
modifications consist chemical treatment (acid, hydrogen peroxide, alkaline), anodic
oxidation, sol-gel and chemical vapour deposition. Some techniques have been
introduced for titanium and titanium alloys such as silanised titania, photochemistry,
self-assembled monolayers, protein resistance and protein immobilisation [6, 7].
A thin 2 - 5 nm surface layer of titania (TiO2) forms when titanium is exposed to air and
protects the surface from corrosion [47]. The layer that is in contact with the substrate
16
consists of TiO; the intermediate layer is Ti2O3 in composition and the outer layer is
TiO2. The titania outer layer is bioinert and cannot bond to bone cells easily since it is
encapsulated by fibrous tissue that isolates it from the bone cells. Such implants,
therefore, tend to become loose and are not recommended for lengthy periods [7]. On
the other hand, ceramics, glass and several other implant materials can bind and form
bone directly through the layer of apatite that forms on their surfaces [48]. Thus, surface
treatments are needed for titanium alloys to achieve better and more rapid bonding to
bone; i.e. osteoinduction. The optimum surface treatment would create a stable passive
layer on the surface of titanium implants with essential properties such as roughness,
wettability and electrostatic charges.
2.2 Anodic oxidation as a metallic implant surface treatment
Electrochemical anodic oxidation has been used to grow a thick and uniform oxide layer
on metals for a decade or so. It has been shown to play an excellent role in the
biocompatibility of metal implants [47]. A controllable and desired thickness is formed
by putting the metals that demonstrate a spontaneous protective oxide layer as the anode
of an electrochemical cell. An ordered oxide layer can be tailored when the type of
electrolyte, applied current density, electrolyte concentration, electrolyte temperature,
agitation speed, and cathode to anode surface area ratios are controlled [49]. It was
initially demonstrated that a nanoporous oxide layer could be formed on aluminium [50]
in such kind of conditions; thus seeding the fabrication of the same layer on titanium
[10], niobium [51], tantalum [52] and zirconium [53] has been undertaken. These are
termed as the “valve metals” that are able to form dense, stable, tightly adhering,
electrically insulating oxides on their surface.
A nanotube titanium oxide layer with a controlled and uniform diameter can be obtained
by anodising oxidation, which is a cost effective, versatile and simple technique. This
nanotube titania layer plays an important role in the enhancement of osseointegration
through improving the adhesion of the hydroxyapatite (HAP) coating that is deposited
onto TiO2. Oh et al. [54] indicated that due to the mechanical interlocking between the
HAP coating and the nanotube titanium oxide layer the cell adhesion could be improved
by up to 400 %.
17
Sibert [55] reported on the formation of a thick, adherent film of TiO2 on titanium
surface using anodic oxidation when the nature and concentration of electrolyte,
forming voltage, current density, and temperature were properly adjusted [55].
Mechanisms for apatite formation on TiO2 on titanium surfaces were indicated by
Kukubo [56]. Zwilling et al. [10] demonstrated the growth process of TiO2 nanotubes
on pure titanium and Ti6Al4V using anodisation in chromic acid (CA) electrolytes, with
or without the addition of fluorine species. The diameter and thickness of the nanotubes
were determined by the applied potential and time. Anodising can form TiO2 nanotubes
that are open at the top and closed at the bottom and of different sizes by changing the
concentration and type of electrolyte, the time and the voltage.
A standard two or three electrode setup can be employed for anodising the sample,
mostly commercial pure titanium. There are combinations of electrolytes: (i) an aqueous
acidic solution containing F-, (ii) an aqueous buffered solution, or (iii) a non-aqueous
electrolyte containing F- ion with or without a trace amount of H2O. Initially, compact
titanium dioxide will be created [57], as illustrated in Fig. 2.2.
18
Fig. 2.2 Growth of regular TiO2 nanotubes, a) cathodic reaction, b) anodic reaction, c) transition state of TiO2 layer, d) starting of nanotube
formation and e) titania nanotubes
19
Nanotubes then form as soluble fluoride complexes through the chemical dissolution of
the oxide in the presence of the fluoride ion and there will be an increase on current. A
stage is then reached where there is equilibrium between oxidation and dissolution; at
which point a nanostructured surface forms and the current flow is constant [57, 58]. A
schematic current-time curve is shown in Fig. 2.3.
Fig. 2.3 Schematic current-time curve (adapted from [58])
All chemical reactions respond to the in situ physical and chemical conditions. The
anodic oxidation mechanism is governed by the type and concentration of the
electrolyte, the applied potential or the applied current density, the temperature, and the
time of the electrochemical reaction. By changing one of these conditions when the
others remain constant, different morphologies of TiO2 nanotubes can be obtained on
commercial pure titanium. It has also been reported that the facing position of the
tubular shape working electrode (sample) towards the counter electrode influenced the
nanotube inner diameter and length. The nanotubes exhibited different diameters and
lengths at both inner and outer surface of the electrode, whether they were on the side
20
where faced the counter electrode or they were on the opposite side. The nanotubes
exhibited different diameters and lengths where they grew at the bottom or at the top of
the abovementioned tubular shaped working electrode. These observations were due to
the potential drop in the electrolyte, the field screening effect inside the tubular and
diffusion process [59]. Two different TiO2 morphologies; (i) the classic ordered
nanotubes and (ii) a nanoscale sponge layer were obtained when the rotated working
electrode was used [60].
2.2.1 The influence of the type and concentration of aqueous electrolyte on TiO2
nanotubes
When the anodisation of titanium takes place in an acidic solution, most of the
composition of hydrofluoric acid is in HF form. The diameter of the TiO2 nanotube that
grows in a HF-based electrolyte lies between 15 to 140 nm under conditions where
H3PO4 and H2SO4 were used and the applied potential changed from 1 to 25V [57, 61].
The optimum voltage was normally 20 V and the time of the anodisation, which had an
influence on the thickness of the oxide layer, was 12 h; a maximum length of ~ 500 nm
of TiO2 nanotube has been obtained [62]. The amorphous structure of nanotube arrays
was changed to anatase at 450 C inside the tube wall and it was entirely changed to
rutile at 600 C by annealing [63]. These phases exhibited different influences on the
apatite formation and/or cell fate [48, 64]. Titanium-based layers are more hydrophobic
than other layers formed in a buffered electrolyte. The hydrophilic value at the interface
of the implant surfaces with tissue is a crucial factor in cell adhesion and the best
hydrophilic value was reported to be 16 [65]. Fig. 2.4.a shows SEM images of TiO2
nanotubes (top view and cross-section) formed in 1 M H3PO4 + 0.3 wt % HF at 15 V
[61].
21
Fig. 2.4 SEM and TEM images of TiO2 nanotubes (top view, cross-section and bottom):
a) formed in 1 M H3PO4 + 0.3 wt % HF at 15 V [61], b) in 1 wt % HF at 20 V for 15
and 30 min) [66], c) in 1 M (NH4)2SO4 + 0.5wt % NH4F at 20V [67], d) in ethylene
glycol + 0.3 wt % NH4F + 2 vol % H2O [68]
Hydrofluoric acid selectively etches the surface of some metals. Therefore, the presence
of fluoride ions within the electrolyte during growth of titanium oxide allows the
formation of a nanotube oxide layer. A TiO2 nanotube of 70 - 100 nm in diameter forms
in less than 2 h when anodisation occurs in an electrolyte consisting of only HF at an
applied potential of 20 V [65, 66]. There is a certain concentration range of F- in which
regular uniform nanotubes can be formed; and beyond which nanotubes cannot be
uniformly fabricated. The nanotube diameter increases with an increase in voltage. It
has been reported that an oxide layer of TiO2 nanotube with a thickness of ~ 250 nm has
been formed in HF electrolyte [64, 69]. A combination of an amorphous and crystalline
22
structure for the TiO2 was observed; and the contact angle of the TiO2 nanotube that
grows in this type of electrolyte is still higher than the best proposed hydrophilic value
Fig. 2.4.b SEM and TEM images of TiO2 nanotubes (top view, cross-section and
bottom) formed in 1 wt % HF at 20 V for 15 and 30 min [66].
The bioactivity of the titanium oxide can be enhanced by immersion in a NaOH solution
so that sodium titanate nanostructures will be formed on the top edge of the nanotube
wall. Another process is to heat the TiO2 nanotube so that anatase will form; onto which
a nanostructured hydroxyapatite (HAP) phase can be created through soaking into a
simulated body fluid. The presence of the nanostructured TiO2 accelerates the formation
of HAP [64]. The adhesion of MC3T3-E1 mouse osteoblast cells was significantly
promoted with the filopodia of growing cells actually going into the nanotube pores,
producing an interlocked cell structure [54].
One proposal for the prevention of bacterial infection following orthopaedic
implantation is local antibiotic therapy, such as using bone cement or a polymer coating
that is loaded with a therapeutic agent [70]. It is possible to decrease the susceptibility
of the implant to bacterial infection due to limited chemical stability or controlled
kinetic release of the therapeutic agent. It is proposed that TiO2 nanotubes, following
phase transformation to anatase by annealing, can be loaded with antibiotics; for
instance 200, 400 and 600 µg of gentamicin, to minimise the initial bacterial adhesion
[70].
Schmuki et al. [67] have observed the electrolyte dissolubility and hydrolysis ability of
a titanium oxide layer and have indicated that smooth and longer nanotubes can be
formed in buffered electrolyte [((NH4)2SO4+NH4F) or (Na2SO4+NaF)]. The pH profile
within the tube was constructed to evaluate the effect of the electrochemical condition
on the geometry of TiO2 nanotubes using anodic oxidation in buffered electrolyte. The
dissolution current was adjusted and, as the electrolyte was neutral, acid was created at
the pore bottom. The pore thickness, regularity roughness, diameter and the crystalline
shape change with the variations in the concentration of F- in electrolyte, the sweep
potential rate and the applied potential. Under these conditions TiO2 nanotubes grew up
to ~ 2.5 µm, even 4 µm, in thickness and ~ 100 nm in diameter, based on a 20 V applied
potential [67, 71, 72]. The tube diameter is affected by the voltage but it is not affected
by time. However, the time of anodisation influences the thickness of the nanotube
23
layer. Fig. 2.4.c SEM images of TiO2 nanotubes (top view, cross-section and bottom)
formed in 1 M (NH4)2SO4 + 0.5wt % NH4F at 20V [67].
Apatite formation was evaluated on an annealed nanotube layer with a diameter of ~
100 nm and a thickness of 2 µm [48]. In comparison to a flat surface, such a nanotube
surface encourages the formation of apatite. The rate of apatite formation depends on
the thickness of the layer and the nanotube structure. Apatite forms more dominantly on
thick and crystalline nanotubes [48]. In cases where the oxide layers were not uniform,
precipitation was observed in buffered electrolytes consisting of K+ and Cs+ whereas no
precipitation was observed in Na+ and NH4+ electrolytes; thus, the electrolyte salts have
to be chosen carefully regardless of the importance of the final pH [73].
A TiO2 nanotube layer with a maximum layer thickness of 1 µm was formed in acidic
electrolyte consisting of fluoride salt at an applied potential of 20 V [74]. When the
length of the 35 - 70 nm tubes was increased, the Young’s modulus of TiO2 which was
approximately 36 - 43 GPa decreased because substrate as well as the functionality of
the coating influence measured Young’s modulus and hardness [75]. The modulus of
elasticity of nanotube oxide layer is more close to that of the substrate if it has lower
thickness or if it is closer to the substrate. An apatite coating can be obtained on the
TiO2 nanotube layer after immersing such an annealed layer in a solution with a high
calcium and phosphate concentration. A carbonated apatite crystal with a column-like
shape of nanometer size was formed, and the bond strength between the HA coating and
the nanotube layer was measured to be higher than 15.3 ± 2.5 MPa [74, 75].
2.2.2 The influence of the non-aqueous electrolyte on TiO2 nanotubes
An organic electrolyte has a low amount of oxygen in comparison to an aqueous
solution; therefore, the oxide chemical dissolution in an organic electrolyte depends on
the water concentration [68]. As a result, a thicker nanotube layer will form over a
longer time in an organic electrolyte and its viscosity, whether stirring [76] and the H2O
content influence the nanotube formation. Nanotubes with a diameter of 20 to 160 nm
were obtained at an applied potential ranging from 30 to 120 V using an organic
electrolyte [68, 76-79]. TiO2 nanotubes with a maximum height of 45 µm [68] or higher
(>150 µm) [77] in electrolytes such as glycerol [76, 80], ethylene glycol [68, 78] and
acetic acid [79] were formed after 1 to 15 h anodisation [68]. However, at short
24
anodisation time the tubes were connected by a ring around their outside walls, and with
a longer anodisation time they were smooth and not connected because of dissolution of
the rings [79]. A smooth layer of TiO2 nanotubes that were not connected by outside
walls was formed when NH4F in glycerol was used as the electrolyte [80]. Such an
electrolyte exhibits a lower diffusion constant, which has an influence on the pH value
of the pore tip and consequently on the current transient and dissolution rate. The
viscosity of the electrolyte is important too because in highly viscose electrolyte
diffusion constant decreases. The diameter of these nanotubes was 40 nm with a 7 µm
length and aspect ratio of 233 - 117 [80].
The shape of the nanotubes is influenced by the applied potential and the anodisation
time in a water-free electrolyte consisting of acetic acid and NH4F. Under low potential
conditions (i) the nanotube diameter is about 20 nm, (ii) the maximum length will be
100 nm and (iii) rings outside the walls connect the tubes [79]. Nanotubes with a coral
reef shape form and the rings dissolve when the applied potential and anodisation time
are increased. If the nonaqueous electrolyte consisting of glycerol and NH4F, and it is
stirred, the adhesion strength of the TiO2 nanotube (1.2 - 1.9 µm in thickness) to the
substrate decrease, and the contact angle with water ranges from 58 - 84 [76]. Fig.
2.4.d SEM images of TiO2 nanotubes (top view, cross-section and bottom) formed in
ethylene glycol + 0.3 wt % NH4F + 2 vol % H2O [68].
One interesting application of amphiphilic TiO2 nanotubes is the fabrication of a drug
release system. A hydrophobic cap using organic monolayer grafting was fabricated on
a TiO2 nanotube, which was treated using the anodising method in the presence of the
fluoride ion [81]. Another anodising process was then undertaken to fabricate a
hydrophilic TiO2 nanotube underneath the initial layer; and the hydrophilic segment was
loaded with hydrophilic drugs. Detachment of the cap would be performed using the
UV induced photocatalytic nature of TiO2; thereby permitting controlled drug release
[81].
2.2.3 The effect of pH value on the formation of TiO2 nanotubes
The pH value influences the dissolubility of the electrolyte and the hydrolysis ability of
titanium oxide layer. This behaviour has been surveyed for titanium foils that were
anodised in three electrolytes of different pH values and the mass loss of the anode
25
during the anodic oxidation was based on the current-time curve [82]. The time of
nanotube formation increases with increasing the pH; consequently the nanotubes have
a longer length compared to nanotubes formed in electrolytes with a lower pH [82]. A
length of 4.4 µm was achieved when the pH value was high but still acidic (pH=4.5)
[83]. As mentioned above, different tube sizes can be expected when the conditions of
the electrolyte and its pH, the applied potential and the anodisation time change.
Fabrication of TiO2 nanotubes can be classified as four generations based on the
anodising electrolyte: acidic, buffered, polar organic electrolytes and nonfluoride base
electrolyte [84]. An acidic electrolyte provides a suitable condition for an anodic
oxidation process. The dissolution rate based on nanometer per second can be
determined by measuring the XPS depth profiles when an anodised oxide layer etched
in different pH values. The highest dissolution rate matched the lower pH value. The
changes in pH values were observed during electrochemical anodisation [72].
To combine the in-growth of bone into micro/macro-sized (~ 1 to 20 µm) and
enhancing biological interaction by nanostructure (nanotubes with ~ 50 nm diameter)
Crawford et al. [85-87], recently fabricated a dual-porous hierarchical structure on
titanium using a simple anodic oxidation process. The anodisation time and pH
influence the nanotube layer length and random distribution of microscopic pores size
and density [86]. The fact that TiO2 nanotube growth can be affected by the orientation
of the titanium substrate have been reported by this group [87, 88]. They also have
investigated the mechanical properties and deformation behaviour of TiO2 nanotubes to
great detail and have measured the modulus of the nanotube coating [87, 89]. In another
study micro- and nano-patterned (MNP) surface was prepared via anodisation of
blasting titanium with large grit and acid etching in HF. The nanotubular layer with
mean inner diameter of 30 and 70 nm resulted in higher level of ALP activity,
mineralisation and osseointegration [90].
In the process of the fabrication of a TiO2 nanotube, it has been shown that when the pH
value is slightly acidic then smooth and long tubes can be formed [83]. An irregular
porous layer will be obtained if an inappropriate applied potential is used. The thickness
of the nanotube oxide layer can be increased by increasing the anodisation time;
however there is a limit beyond which no further changes will be observed or the oxide
26
layer might possibly dissolve. The time limit for a nonaqueous electrolyte is longer than
for an aqueous electrolyte.
There are few study on the effect of the nanotube length and wall thickness on adhesion,
wettability, or surface energy and consequently on cell adhesion, proliferation and
differentiation. Although a longer nanotube has been obtained by improving the method
and the condition, the question of the optimum length required to promote cell response
has not been reported to date except the work of Bauer et al. [91] that have indicated the
tube length does not affect the cell behaviour.
2.3 Nanotube oxide layer on titanium alloys and titanium alloying metals
2.3.1 Anodisation of biocompatible Ti-Nb-Ta-Zr alloy
Some titanium alloying elements are beta (β) stabilisers; e.g., molybdenum, vanadium,
niobium, tantalum (isomorphous; completely miscible elements in the phase), iron,
tungsten, chromium, silicon, cobalt, manganese (eutectoid; elements with eutectoid
temperature as 335 ºC below the unalloyed titanium transformation temperature); and
zirconium, a neutral (a continuous solid solution) alloying element. Beta titanium alloys
exhibit lower elastic modulus compared to α, and α+β titanium alloys [7, 92]: with Ti-
Nb-Ta-Zr (TNTZ) being such an alloy that is the choice for orthopaedic endoprostheses
[93]. Since TNTZ alloy cannot form apatite on its surface by the usual chemical and
heat-treatment processes, then another surface modification method is required [4, 93].
Nakada et al. [94] suggested using Ti-Nb-Ta-Zr alloy for orthopaedic implants because
of their excellent mechanical properties, anticorrosion ability, cytocompatibility and
biocompatibility. An alloy of Ti15Nb4Ta4Zr was compared to conventional Ti6Al4V in
terms of new bone formation and bone mineral density (BMD) [95]. It was
demonstrated that bone mineral density around Ti15Nb4Ta4Zr was equivalent to, or
even higher [94], than that of Ti6Al4V.
Hard tissue replacements, especially knee and hip joints, are eroded over time due to
micro-movement. As a result, microscopic wear particles gather at the implant-tissue
interface and give rise to inflammation, osteolysis, infection and, potentially, implant
loosening. The wear resistance of Ti29Nb13Ta4.6Zr was measured after different heat
treatments and oxidation treatments [96]. Wear resistance was improved by increasing
27
the niobium content, especially by means of an oxidation treatment, since the Nb2O5
layer is hard and exhibits lubricity; but no improvement was observed by heat treatment
[96].
Anodic oxidation of TNTZ alloy exhibits different reaction rates because of the
different electrochemical oxidation rates of elements in this alloy [97]. Consequently,
dissolution was more selective and a homogeneous, self-organised pattern of anodic
tube layers grew to different sizes. Each tube with a large diameter was surrounded by
nearly eight tubes of smaller diameter. The process was initiated with two layers; an
outer nanoporous layer and a nanotube layer when the sweep potential of the
potentiometer was employed. The top nanoporous layer dissolved when the sample was
held in the electrolyte and additional time was also necessary to grow nanotubes when
the applied potential was low [97, 98].
Homogeneous oxide layers of 100 nm thickness were formed on the α phase of
Ti6Al4V due to electrochemical anodisation of the titanium alloy [99]. On the other
hand, due to selective dissolution of oxide layers during anodisation, the β phase of
Ti6Al4V exhibited oxide layers that were inhomogeneous in terms of morphology and
diameter. An inhomogeneous oxide layer that was nanoporous and a mixture of
nanopores and nanoscratches were observed for a complex α + β phases. The element
niobium was observed in nanotubes formed on the surface of alloys consisting of
niobium [99, 100]. Two-size-scale structures of anodised nanotube oxide layer were
observed on Ti-28Zr-8Nb alloy using buffered electrolyte [101]. In this chapter, two
distinctly different tube diameters of 175 nm large nanotubes surrounded by several 98
nm small nanotubes at the bottom were formed with a length of 8.5 µm. The thickness
of this oxide layer depended on both the anodisation time and the applied potential and
it reached 22 µm by extending the time to 8 h; and increased from 3.2 µm to 11.3 µm
by increasing the applied potential from 10 to 40 V for 1 h anodisation [101]. This study
concluded that the nanotube length increased almost linearly with the anodisation time
in the first 60 min, and the rate slowed down with a longer time; whilst the nanotube
length increased linearly with the applied potential up to 40 V. Further studies are
needed concerning the anodic oxidation conditions to tailor the optimum nanotubes on
the titanium alloys, and to evaluate the cell behaviour with regard to their orthopedic
applications on this new pathway of multi-scale nanotubular geometry. Fig. 2.5.a SEM
28
and TEM images (top view, cross-section and bottom) of anodic oxide nanotube layer
formed on a) β phase titanium alloy after anodisation at 20 V for 4000 s [99].
Fig. 2.5 SEM and TEM images (top view, cross-section and bottom) of anodic oxide
nanotube and nanoporous layer formed on a) β phase titanium alloy after anodisation at
20 V for 4000 s [99], d) Zr substrate using 1M (NH4)2SO4 + 0.5 wt % NH4F electrolyte
[102], c) niobium substrate as a function of anodisation temperature at 15 °C [103], b)
tantalum substrate using 1 M H2SO4 + 2 wt % HF electrolyte for 2 h with a sweep rate
100 mVs-1 [104]
2.3.2 Anodisation of binary titanium alloys for implant applications
Studies [105, 106] on binary titanium alloys such as Ti-xTa, Ti-yZr and Ti-Al have
investigated the effect of alloy composition on the morphology of anodic oxide layers.
Uniform nanotube arrays of oxide layers were formed on Ti-Zr alloy and with
increasing zirconium content, the diameters of the tubes decreased and the length
increased. Tube separation occurred when the titanium content of the Ti-Al alloy
increased [105]. In the case of Ti-50Ta, the morphology and length of the nanotubes
29
depended on the phases of the alloy. It was revealed that the nanoporous oxide layers
that were formed first were dissolved after the growth of nanotubes beneath the
substrate surface. This dissolution arose because a local electrochemical was established
due to the different compositions of the complex phases [106].
Titanium-zirconium alloys (Ti-20Zr, Ti-50Zr, Ti-80Zr), which had been fabricated by
accumulative roll bonding (ARB), were treated electrochemically [107]. The ARB
process did not influence the diameter, length or arrangement degree of the tubes during
anodisation. Tubular oxide layers were formed at all ranges of zirconium content
without any transition from porous to tubular forms. As well, the tube diameter
decreased, although the tube length increased, when the zirconium content was
increased. Two dissimilar diameter tubes were observed on Ti-50Zr and Ti-80Zr. These
have been observed previously as forming on Ti-29Nb-13Ta-4.6Zr and Ti-50Ta [107].
Table 2.3 summarised the anodisation conditions, characteristics and properties of TiO2
nanotubes on titanium investigated by different research groups, to date.
30
Table 2.3 Anodisation conditions, characteristics and properties of TiO2 nanotubes on titanium
Anodisation condition Characteristics of TiO2 nanotubes Properties Ref.
Electrolyte
Applied
potential
(V)
Time Others
Nanotube
diameter
(nm)
Wall
thickness
(nm)
Nanotube
layer
thickness
Phase
structure Other treatment
Hardness / bonding
strength
Wetting
angle
(water)
Roughness
H2SO4/HF 20 2 h 50-100 amorphous HTa to Anatase &
Rutile [63]
H3PO4/NaF 20 28 min 30C
Stirred 100 amorphous
HTa to Anatase-
Deposition; Ca to
P=1.67, pH=4.5
bonding strength
;7.41 MPa [108]
H2SO4 /HF 20 3 h ~100 500 nm amorphous HTa to Anatase &
Rutile
enhanced apatite
formation Rsm=25nm [48]
H3PO4/HF 20 2 h 400 77.45 Ra=14 nm [65]
H2SO4/NaF/ Citric
acid 20
15min
-4 h pH=4.5 35-70 10-18
230-670
nm
Young’s modulus=
36-43 GPa [75]
HF /NH4H2PO4 20 1 h 20C
Stirred 100 19 1 µm
HTa to Anatase- HA
deposition by
immersing
bonding strength;
15.3 ± 2.5 MPa [74]
HF 20 2 h 100 83.39 Ra=19nm [65]
HF/NH4F 20 2 h 60 55.45 Ra=16nm [65]
31
HF 20 20min 80 10 250 nm amorphous HTa to Anatase &
Rutile [69]
Na2SO4 /NaF 20 3 h ~100 2 µm amorphous HTa to Anatase &
Rutile
significantly
enhanced apatite
formation
Rsm=77nm [48]
(NH4)2SO4/NaF 20 2 h 120 16° Ra=23nm [65]
Glycerol/NH4F 30 30 min 30C
Stirred
50-65
ripple 1.20 µm amorphous 3.5 ± 0.2 MPa 58 ± 2
Ra= 0.32 ±
0.2 µm [76]
Glycerol/NH4F 30 60 min 30C
Stirred 60-75 ripple 1.45 µm amorphous 4.1 ± 0.2 MPa 73 ±2
Ra= 0.49 ±
0.3 µm [76]
Glycerol/NH4F 30 120
min
30C
Stirred 75-85 ripple 1.7 µm amorphous
3.7±0.3
MPa 78 ± 2
Ra= 0.17 ±
0.2µm [76]
Glycerol/NH4F 30 240
min
30C
Stirred 110-130 ripple 1.9 µm amorphous 3.6 ± 0.2 MPa 84 ± 1
Ra= 0.13 ±
0.2 µm [76]
NH4F and H2O in
Ethylene glycol 60 18 h 100 25 45µm amorphous
HTa to Anatase &
Rutile [68]
Ethylene glycol/HF 120 15 h 70 45 261 µm [78]
CH3COOH-0.2%
H2O/NH4F 30 1 h 10 200 nm [79]
32
KF/NaHSO4/citric
Acid 25 20 h pH=4.5 110-120 4.4 µm amorphous
HTa to Anatase &
Rutile [83]
H2SO4/NaF/ citric
acid 20 4 h pH=4.5 51.1 50.58 600nm amorphous
HTa to Anatase &
Rutile
1.5-1.7 GPa (36-43
GPa)
Water=7.5±
3
Cell media=
2±1º
(surface
energy of
332.6
mJ/m2)
Rms=
51.83±7.77 [109]
acetic
acid/HF 5-20 30-100 amorphous HTa to Anatase 11,9,7,4º
Ra=13.0,12.7
,13.5,13.2
nm
[20]
HTa = Heat Treatment
33
An equation was proposed to describe the linear relationship between the outer diameter
of zirconium-titanate nanotube and the anodisation potential from 1 to 100 V in a
buffered electrolyte [110]. This study also indicated that the concentration of the
fluoride ions influenced the wall thickness and the composition of the nanotube layer
because the zirconium titanate nanotubes grew as a result of a competition between an
electrochemical oxide formation and chemical dissolution of oxide by fluoride ions. The
wall thickness was smaller for high fluoride ion concentrations due to the accelerated
dissolution rate of the oxide wall. The nanotube length depended on the anodisation
charge until etching of nanotube top was significantly observed [110]. Multiple
nanotube layers that can be formed using two step anodising have potential application
such as ultrafiltration or drug release systems. The lower layer of nanotubes started
forming in the gaps between the upper layer of nanotubes formed in the previous
anodisation process [111]. Zirconium titanate nanotubes that were formed on Ti-50Zr
(wt %) alloy surface were amorphous and they were converted to crystalline form after
annealing although TiO2 nanotubes on pure titanium were amorphous and ZrO2
nanotubes on pure zirconium were crystalline [112]. The nanotube oxide layer that grew
on Ti-45Nb (wt %) alloy were thicker and more regular in comparison to the nanotube
oxide layer formed on pure titanium and niobium [113]. Under the same anodisation
conditions, the oxide layer of TiO2 on pure titanium exhibited a thickness of 1.1 µm and
tube diameter 100 nm; and the Nb2O5 nanotubes on pure niobium exhibited a thickness
of 0.6 µm with 15 nm irregular pore diameter; whilst 2.9 µm in length uniform
nanotube oxide layer with 60 nm tube diameter grew on Ti-45Nb alloy after 2 h
anodisation and increased to 6.3 µm at 20 h [113].
2.3.3 Anodisation of tantalum as a β stabiliser
Tantalum is more expensive and has much higher density (16.6 g/cm3) [114] than
titanium (4.5 g/cm3) [115], and therefore it is not appropriate for implants. However, its
chemical and physical properties, such as biocompatibility [116], radiopacity, dielectric
properties [117] and high corrosion resistance to body fluids, are attractive [118].
Tantalum has applications for optical devices, storage capacitors for very large-scale
integrated circuits, and for electronic and sensor devices. Tantalum oxide has received
much attention due to its corrosion resistance as a protective coating material. In
34
addition to aluminium, tantalum is one of several metals onto which a porous oxide
layer can be formed by anodising [52, 119].
A porous oxide layer (Ta2O5) can be formed on tantalum metal when it is
electrochemically treated in acid that contains HF as the electrolyte. Pore diameters are
from 2 to 35 nm and the layer thickness is 350 - 400 nm. Similar to titanium, the
structure and distribution of the pores and the thickness of the porous layer can be
altered by changing the concentration of HF, the time allowed for anodic oxidation and
the anodisation voltage. The equilibrium between porous oxide formation and its
dissolution, as well as the time scale, controls the film thickness for the development of
porosity and ordering effects [52, 104, 117]. Before making dimpled tantalum metal
surfaces in H2SO4 and HF electrolyte without organic additive, tantala nanotubes can be
fabricated if the time of anodisation adjusted between 5 to 120 s. The reason of
nanotube loss from the surface was shown to be the formation of a thin, fluoride rich
(likely TaF5) layer builds up at the Ta/ Ta2O5 interface [120].
Tantalum oxide nanoporous layer can be fabricated in an organic electrolyte consisting
of ethylene glycol and glycerol with a background salt such as (NH4)2SO4 plus suitable
amount of NH4F to a maximum thickness of 11 µm and the pore diameters increased
from 9 to 16 nm with increasing the voltage from 10 to 40 V [121]. A highly ordered
Ta2O5 nanoporous layer grew on tantalum foils by anodisation using glycerol (1,2,3-
propanetriol) and NH4F as electrolyte . It was found that the average pore diameter and
the thickness of the porous layers were affected by the NH4F concentration. The
optimum conditions for forming a homogenous nanoporous oxide layer with high aspect
ratio and a pore diameter of 17 nm and a thickness of 16 µm were a 0.2 M NH4F
electrolyte and an applied potential of 20 V [122]. Fig. 2.5.b SEM images (top view
and cross-section) of anodic oxide nanoporous layer formed on tantalum substrate using
1 M H2SO4 + 2 wt % HF electrolyte for 2 h with a sweep rate 100 mVs-1 [104].
Tantalum has been deposited on the titanium surface by electrodeposition [123, 124],
and a sol-gel process [125] so that the good bulk properties of titanium can be enhanced
by the good surface properties of tantalum oxide. The aim was to increase the corrosion
resistance as well as the bond strength between the bone tissue and titanium.
35
2.3.4 Anodisation of niobium as a β stabiliser
The functionality of niobium for its application in gas sensors, catalysts, optical,
electrochromic devices, and biocompatible prostheses, has been investigated with
regard to increasing the local surface area. Nanostructured niobium oxide has been
fabricated by reactive sputtering [126], sol-gel processes [127], templating techniques
[128] and anodic oxidation [129]. Sieber et al. [51] were the first group of researchers to
fabricate self-organised porous niobium oxide by anodising oxidation in an electrolyte
containing F-. Uniform layers of porous niobium oxide with 20 - 30 nm pore diameters
were obtained. The HF concentration in the electrolyte and the anodisation time
influenced the thickness and microstructure of the porous layer. A steady state porous
film formation and film dissolution limited the film thickness to about 500 nm [51].
The effects of mixed electrolytes, applied potential and anodisation time were
investigated to obtain a high aspect ratio of ordered porous niobium oxide layers. It was
revealed that less than 10 nm in diameter of ordered porous niobium oxides were
formed by anodisation of niobium foils at 2.5 V in a mixture of 1 wt % HF and 1 M
H3PO4 for 1 h. Nanoporous niobium oxides could not formed on niobium if a single HF
electrolyte was used since the pore wall would dissolve as they form [130].
Choi et al. [103] targeted a thickness of porous niobium oxide more than 500 nm by
employing double anodising and one step annealing. The annealing step was designed
to form a protective oxide layer that would delay the chemical dissolution as part of the
anodic oxidation in fluorinated electrolytes. The thickness of the outer layer formed in
the first step was around 90 - 130 nm; and the inner layer of around 300 - 400 nm was
grown underneath the first layer after annealing [103]. Fig. 2.5.c SEM images (top view
and cross-section) of anodic oxide nanoporous layer formed on niobium substrate as a
function of anodisation temperature at 15 °C [103].
2.3.5 Anodisation of zirconium as a neutral titanium alloying element (or an
effective β stabilizer in multi-elementary Ti alloys)
Zirconium is one of the valve metals reported to form porous oxide layers on their
surface. Zirconium oxides have noteworthy properties, such as chemical and thermal
stability, mechanical strength, wear resistance, good ion-exchange and excellent
biocompatibility. Zirconium is used as a catalyst where an increase in its surface area
36
improves its functionality. Other applications include magnetic recording media, optical
devices, functional electrodes, and waveguides [131].
Tsuchiya et al. [53] were among the first researchers to report on the fabrication of
sponge-like porous zirconium oxide layers manufactured by electrochemical
anodisation in acidic electrolyte containing F-. An applied potential between 20 and 40
V was needed and the pore diameter was several 10 nm with a thickness of several 10
µm.
The formation mechanism of a self-organised nanotube oxide layer has been described
in detail [132] with regard to (i) the effect of changing the concentration of fluoride ions
and pH, (ii) the composition of the electrolyte, and (iii) the potential sweep rate and
holding potential, on the morphology and geometry of the porous oxide structures. By
improving the method, smooth high aspect ratio zirconia nanotubes with a diameter of ~
50 nm and a length of ~ 17 µm were obtained using a buffering electrolyte, instead of
the wavy and irregular wall morphology of pore arrays [102, 132]. Organic electrolytes
were also used to grow significantly thicker, and continuous and smooth ZrO2 nano-
tubular layers up to 40 µm on zirconium substrates [133-135]. It has been shown that
water content of the organic electrolyte influenced the chemical etching behaviour of
the electrolyte and the porous to tubular morphology, which is important in different
applications [134].
Pretreatments such as dip etching [133], electropolishing [135] and two step anodising
[133] were used in the investigations of the effect of pre-treatment of zirconium surface
on the anodised nanotubes. Regular nanotubes that formed on dip etched surface
indicated that the mechanical deformation layer and/or zirconium surface impurities
were the reason of the irregular, patchy (coherent) oxides. The needle-like (nano-pillars)
structures on the top of the tube walls were also observed due to thinning and partially
etching-out [133]. Highly regular hexagonal ordered dimples of the right spacing were
formed on zirconium surface after removing the layer resulted from the first anodisation
step. The nanotubular oxide layer exhibited a thickness of 3 µm and a diameter of 60
nm with highly ordered hexagonally porous which can be removed [133]. The
roughness of zirconium surface decreased from ~ 18.6 nm of bare surface to ~ 1.1 nm of
electropolished surface and this reduction had influence on a homogeneous electric field
distribution over the entire metal surface during the anodisation step, thus nanotubes
37
grew regularly on the entire surface [135]. Fluorine can be removed from the ZrO2
nanotubes by heat treatment at 300 °C for 6 h and it has been shown that although 800
°C heating will lead to crystallisation; collapse of the nanotubes will be observed [136].
Fig. 2.5.d SEM and TEM images (top view, cross-section and bottom) of anodic oxide
nanotube layer formed on Zr substrate using 1M (NH4)2SO4 + 0.5 wt % NH4F electrolyte
[102].
2.4 Factors that influence bone cell adhesion
The extracellular matrix (ECM) secreted by cells plays a role as the interface of a
biomaterial surface. ECM is synthesised and degraded by cells as a dynamic
surrounding substance that consists of 90 % collagenic proteins (type I collagen and
type V collagen) and 10 % non-collagenic proteins (NCP, for example osteocalcin,
osteonectin, bone sialoproteins, proteoglycans, osteopontin, fibronectin, growth factors
and bone morphogenetic proteins etc.) [15]. Anselme et al. [137] indicated that the cells
are always in contact with a biomaterial surface that has previously adsorbed water and
proteins from biological fluids, rather than with a bare surface. Specific receptor
proteins such as integrins, as well as the junctions of adherents containing cadherins
mediating cell-cell adhesions, are important for cell-substrate adhesion. Focal contacts
or adhesion plaques are junction locations of about 10 - 15 nm where adherent cells and
material surfaces are joined. Integrins are on the external side of focal contacts and they
can translate the attachment of external ligands to internal information that induces
adhesion, spreading, or cell migration, growth and differentiation [15]. It has been
shown that the vinculin and puxillin can play an essential role in continuing adhesion
and upcoming differentiated cell functions [138]. The cell proteins involved in cell
adhesion on biomaterial are illustrated in Fig. 2.6.
38
Fig. 2.6 Representation of the cell proteins involved in cell adhesion on biomaterial: (a)
immediately after implantation; (b) adsorbing proteins from body fluid; and (c) attached
bone cell on an implant material surface in higher magnification (adapted from [15])
39
The adhesion force at the cell/implant interface determines phenomena that involve
short term and long term adhesion due to (i) physicochemical linkages such as ionic
forces and van der Walls forces and (ii) different biological molecules such as
extracellular matrix proteins, cell membrane proteins, and cytoskeleton proteins. The
adhesion can be measured using a variety of methods such as micropipettes aspiration,
centrifugation (the force necessary to separate cells from a substrate is provided by
centrifugal force) [139] , and fluid flow; enzymatic procedures; optical or magnetic
tweezers (measure local viscoelastic properties of the surface of adhering cell by a
magnetic bead microrheometer [140]); and microcantilevers (detaching adhered cell by
applying a lateral load using a microcantilever and measuring the detachment force
[141]) for the detachment of cell populations by micromanipulation [137]. By
micropipettes aspiration the strength of cell adhesion can be measured by vertically
oscillating micropipette while it makes contact with cell and the pressure within the
micropipette is reduced to a level that cell can attach to the pipette by suction [142].
Glass, silane-adherent cells [143] which deposited in glass and silane capillary tubes can
be detached by calibrated laminar shear flows with a highly viscous dextran solution.
This method is known as fluid flow [144].
There are different opinions concerning the preferred bone growth direction around
implants, for example, towards the implant, on the implant surface or both together
[137, 145]. In addition to the cell type, the surface chemistry, mechanical cues and
topography influence cell behaviours such as adhesion, proliferation, shape, migration,
survival and differentiation [137, 145, 146]. It has also been reported that there are
different mechanisms of adhesion for blood cells, fibroblasts or osteoblasts (connective
tissues), and endothelial vascular cells or keratinocytes (endothelia or epithelia), such as
adherence and proliferation rate [137].
2.4.1 The influence of surface physicochemical, mechanical and electrical
properties on bone cell behaviour
Surface chemistry plays an important role in the cell response at interfaces. The effect
of surface energy, water contact angle or wettability (hydrophilicity and
hydrophobicity), and the zeta potential of different surfaces on cell behaviour have been
reviewed [137]. Protein adsorption, cell adhesion, proliferation and osteoblastic
differentiation will be increased by a decrease in the water contact angle or an increase
40
in surface energy. Electrical charges on titanium surfaces that are measured as zeta
potential are important in the interaction with negatively charged proteins such as
fibronectin (FN) [14, 147]. A titanium surface, normally, is negatively charged due to
the adsorption of anions (OH-, F- etc. from the electrolyte). On the other hand, cell
membranes are also negatively charged, so positively charged proteins at the cell
surface interface can play an important role, which is proposed as a dynamic model for
osteoblast attachment [147]. The adhesion constant and binding efficiency of adsorbed
proteins on different functional groups (e.g. OH, COOH, NH2 and CH3) demonstrate
different adhesion strengths [14]. The cell adhesion, proliferation and gene expression
were improved after UV irradiation due to the introduction of Ti-OH groups on the
nanostructured surface and the further enhanced hydrophilicity of nanostructured
surface [148].
Changing the surface topography to nano scale does not change the surface chemically
with respect to its bulk, but it might change the surface chemistry and energy at the
interface. Rising hydrophobicity is one example of the influence of surface density on
surface chemistry. Decuzzi et al. [149] proposed a mathematical model and introduced
three regimes for surface energy including small, intermediate and large. They then
investigated the effects of surface roughness on cell adhesion in these three regimes, and
found that cell adhesion will deteriorate with an increase in roughness for small energy
surfaces, exhibit maximum adhesion for an optimal roughness yet will be scarcely
affected by roughness for intermediate surface energies [149]. In the line of the
abovementioned model, stable cell adhesion and proliferation for four different cell
lines have been reported to be increased on moderately rough Brownian substrates with
nearly similar surface energy but the generality of the results need additional studies to
verify [150].
Osteoprecursor cell line attachment and growth behaviour on TiO2 nanotubes with ~ 50
nm diameter were also studied [109]. The effects of the physicochemical properties of
the surface, for instance, roughness, contact angle and surface energy, on the cell
behaviour were investigated and it was found that materials with a high nano scale
roughness, low contact angle and high surface energy showed the same positive effect
on cell behaviour, such as early differentiation and significant bone cell proliferation,
which means an improvement in cell adhesion and spreading [109]. Kim et al. [151]
41
indicated that cell behaviour is affected by the incorporated anions in the oxide layer
rather than the morphology of surface. It was shown that the composition of the
electrolyte, e.g., HF, HF+H3PO4, and H3PO4 not only determines the oxide
morphologies (dot-like structures, granules, nanopowders and nanotubes), but it also
affects the incorporation levels of the ions of F- and PO43-, which cause different cell
behaviours: the PO43- enriched oxides enhance cell proliferation in 7 days and the F-
enriched oxides stimulate initial cell attachment [151].
Mechanical cues, for instance, rigidity, stiffness and resilience, have been shown to
influence cell behaviour [137]. Influence of microenvironment stiffness on stem cell
specifications has been observed when the cells commit to the lineage specified by
matrix elasticity after several weeks in culture in inverse to the initial week in culture
that could be reprogrammed these lineages by adding soluble induction factors [152].
Mechanical feedback regarding substrate rigidity is essential for the cell to shape, grow
and survive, but recognition by the cells of their microenvironment and elasticity and its
influence on the structure and function of the cells is still under investigation [137, 153-
155].
2.4.2 Effect of topography on the bone cell behaviour
Chen et al. [156] reported that a decrease in the surface roughness of Ti resulted in an
increase in its corrosion resistance and a decrease in ion release. The roughness of a
metallic implant surface and its uniformity in the horizontal or vertical direction
influence its favorable mechanical locking to tissues. Several hypotheses regarding the
mechanisms of cell responses to surface topography were proposed in a review by
Anselme et al. [157]. Extracellular matrix protein adsorption acts as an energy barrier
for cells to modify their orientation, adhesion and spread. Changes in gene expression
and in cellular, cytoskeletal, and focal adhesions were observed in a study on cell
behaviour on micro- and nano-anisotropic topographies, regardless of the long-term
turnover [157]. Although there is an absence of comparable studies, cell operation is
motivated when the size is ~ 10 nm in height or depth [158], and cell operation is not
noticeable when the scale increases to ~ 100 nm.
The principal aim of all research in this field is to improve the wound healing that
accompanied by the promotion of osseointegration. Large macrostructure human bones
42
consist of nanosized organic and mineral phases, such as ions, DNA, proteins and the
viruses in the body. There has been insufficient study concerning the retention of
proteins on nanoscale surfaces, because fabrication of nano features or nano surfaces
cannot be easily reproduced. Nevertheless, investigation has shown that the structure
and function of absorbed proteins on smaller nanoparticles can be readily retained in
comparison to larger nanoparticles, but the details of the adsorption mechanisms are not
clear [157, 159]. As cell-substrate adhesion is based on integrins that have nanoscale
features, it can be predicted that cells will respond better to nanoscale surfaces. Further
research needs to be undertaken to clarify this hypothesis [137, 160]. In general it has
been shown that cells respond to nanosurfaces because the pores, ridges, and fibres of
the basement membranes have nano-scale characteristics [161]. Fig. 2.7 shows filopodia
of SaOs-2 cells on 200 nm deep round concentric grooves and ridges in quartz [157].
Fig. 2.7 SEM image of filopodia of the SaOs-2 cells on 200 nm deep round concentric
grooves and ridges in quartz [157]
43
Nanotubular layer is well-suited to an investigation of the mechanism of a bone cell
response to the various morphologies of a titanium and titanium alloy surfaces [11, 162-
164]. In revealing the effect of nanotubular pattern of Ti-30Ta on cell response in
comparison to a surface of Ti-30Ta alloy, cellular adhesion, proliferation, viability,
cytoskeletal organisation and morphology of human dermal fibroblasts (HDF, neonatal)
were investigated. Nanotube architecture anodised at 35 V for 40 min showed
enhancement of improved cellular functionality [165]. Good bioactivity for
mesenchymal stem cell adhesion and spreading and fast formation of extracellular
matrix (ECM) materials on the Ti-Nb-O nanotubes formed on Ti-35Nb has been
observed [166]. There were also some studies on ternary and higher titanium alloys
[167-170]. Nearly the same cell adhesion rate, the spreading out of attached stem cells,
the flat shapes and numerous filopodia after longtime culture occurred for mouse bone
marrow MSCs (mMSC) on Ti-35Nb-5Zr in comparison to Ti-35Nb [168]. Human
osteoblast cells growth on anodised nanotubes on Ti-6Al-7Nb in two inorganic and
hybrid electrolyte was found to be higher than that of cells cultured on untreated Ti-6Al-
7Nb alloy [169]. The protein adsorption, initial cell adhesion, cell differentiation and
osteogenesis related gene expression have been enhanced on sparsely distributed
nanotubes (SNT) on the surface of a near β titanium alloy Ti-5Zr-3Sn-5Mo-15Nb
(TLM) compared to the polished same alloy [170].
2.5 Effect of the characteristics of anodised TiO2 nanotubes on bone cell
behaviour
The size and distance between nanofeatures on an implant, in addition to adsorbed
proteins, as well as the shape or organisation of these nanofeatures, influence the cell
response [157, 159]. Although there are methods such as lithography and
nanoimprinting to fabricate uniform and reproducible nanofeatures on polymeric or
other material surfaces, these are not convenient for metals [14]. Anodic oxidation is a
method to tailor nanoscale patterned surfaces for metals and alloys [10].
2.5.1 Effect of nano-spacing of the surface of TiO2 nanotubes on cell behaviour
Park et al. [17] evaluated the adhesion, spreading, growth, and differentiation of rat
mesenchymal stem cells on the surface of TiO2 nanotubes of 15, 20, 30, 50, 70, and 100
nm diameters. Cells were found to adhere and spread widely on 15 nm tubes but the
44
adhesion decreased with an increase in the nanotube diameter. Focal contact formation
on nanotubes with a diameter ≤ 30 nm was higher than on those with larger diameters.
The cell proliferation rate decreased as the diameter of the nanotubes increased. The
highest cell differentiation was also observed on 15 nm tubes. This result was compared
with polished TiO2 surfaces, which showed that the 15 nm tubes offered optimal
spacing for cell proliferation, migration, and differentiation resulting from integrin
clustering and focal contact formation. Peri-implant bone formation of commercial Ti
covered by 30 nm TiO2 nanotubes was investigated in vivo in pigs and it was found that
the nanostructures not only promoted osteoblast formation in the initial step but also
resisted implant insertion shear force [18]. Osteoblasts and osteoclasts, which are
responsible for the formation and resorption of bone cells, demonstrated the same
reaction to the one-dimension surface nanotopography of 15 nm TiO2 nanotubes [19],
even though it is well known that their activities are different due to their different
working mechanisms. It has been claimed that the essential nanosize of TiO2 nanotubes
is less than 100 nm and the best diameter is 15 nm for adhesion and differentiation of
various cells [19].
Although MC3T3-E1 mouse osteoblast cells promoted higher adhesion on ~ 30 nm
TiO2 nanotubes, increased elongation of the cells and enhanced alkaline phosphatase
activity were observed on nanotubes of greater diameter between 70 - 100 nm that are
known to have greater bone forming ability [20]. This conclusion is different from the
results reported by Park et al. [19], in which the optimal diameter of nanotubes is 15
nm. However, the increase in the bone forming ability of a nanotube of ~ 100 nm
diameter was caused by crystallisation of the nanotube during heat treatment [20, 171].
The adhesion was accelerated by ~ 300 - 400 % increase in the number of the adhered
cells due to its significantly increased surface area and the presence of fluid between
annealed nanotubes or those soaked in sodium hydroxide [54, 172]. The optimum size
of a TiO2 nanotube for osteoconductivity and osseointegration ability has been reported
as 70 nm after assessments using an in vivo real-time polymerase chain reaction (real-
time PCR) technique, fluorochrome labels and histological analysis [173], but this is not
completely consistent with the studies mentioned above. Also in a study of the
relationship between Zr content and nanotube diameter in the Ti-xZr (x=10,20,30,40)
and Ti-30Ta-xZr (x=3,7,15) alloy, it has been reported that when Zr content increased
nanotube diameter decreased from 200 nm to 150 nm for Ti-xZr alloy and it decreased
45
from 200 nm to 50 nm for the ternary alloys. Good MC3T3-E1 mouse osteoblast cell
proliferation, migration and differentiation have been observed on the 50 nm diameter
of nanotubes formed on the surface of Ti-30Ta-15Zr, better than on Ti alloy with low Zr
content [167]. Although titania nanotubes formed on α phase and β phase of the
titanium alloys due to the content of alloying elements [167] were self-organised,
irregular [169, 174] and showed different diameters [175] which affected the HA
formation [176] there is still few study on the cell behaviour to different phases.
In summary, cells respond to the topography, physicochemistry, and the electrical and
mechanical properties of the implant surface. They can recognise the topography from a
few nanometers to several hundred micrometers following different chemical
treatments, especially anodic oxidation where the nanotube layer is reproducible and
can be fabricated in a cost effective way. Table 2.4 lists the different cell responses to
various TiO2 nanotubes on titanium and the anodisation condition from different
studies, to date. The interaction between the cells and the TiO2 nanotubes provides the
possibility of controlling the cell culture by ordering the physicochemical properties of
the surface. Fig. 2.8 schematically illustrates a bone cell attached to titania nanotubes
showing a hypothetical model of adhesion.
46
Fig. 2.8 Schematic illustration of a bone cell (osteoblast) attached on titania nanotubes
with a diameter less than 100 nm (adapted from [109] and [35]).
47
Table 2.4 Cell responses to different TiO2 nanotubes on titanium and the anodisation conditions
Nanotube characteristic Anodisation conditions
Hydroxyapatite deposition / cell response Ref. Nanotube
diameter
(nm)`
Wall
thickness
(nm)
Nanotube
layer
thickness
(nm)
Phase
structure Other treatment Electrolyte
Applied
potential (V)
Time
(h) Others
51.1 50.58 600 Amorphous HT* to anatase
and rutile
H2SO4/NaF/
citric acid 20 4 pH=4.5
Osteoblastic precursor cell line
(OPC1) /Dense focal contacts
Proliferation 0.342
High activity
Early differentiation in day 5
[109]
>100 - Short
length - - H3PO4 /HF 20 1
Stirred
(≈180 rpm)
MC3T3-E1 cells/good initial
attachment of cells
not changed through culturing for 7 days
[151]
15-20-30-
50 - - - - H3PO4/HF 1-20 - -
Rat mesenchymal stem cells /adhesion and
spreading, proliferation and differentiation were
highest for 15 nm diameter [17]
70-100 - - - - H3PO4/HF 1-20 - - Rat mesenchymal stem cells /adhesion and spreading
was impaired without stable extension of filopodia
lowest impaired [17]
30-100 - - Amorphous HT* to anatase Acetic 5-20 - - MC3T3-E1/increasing nanotube diameters [20]
48
acid/HF led to increased elongation/stretching of cell bodies,
increased levels of alkaline phosphatase and greater
bone forming ability
30-100 - Amorphous HT* into
crystalline phase
Acetic
acid/HF 5-20 0.5 -
Human mesenchymal stem cells (hMSCs)/
promoting adhesion without noticeable
differentiation at ~30 nm, eliciting a dramatic
stem cell elongation at ~70-100 nm, induced
cytoskeletal stress and selective differentiation into
osteoblast-like cells
[21]
100 - - - - H3PO4/HF 1-20 1 - Human primary osteoblasts and osteoclast/15nm was
recognised at least by both [19]
70 15 250 Amorphous HT* to anatase HF 20 0.5 - MC3T3-E1/Adhesion increase ~300-400% [54]
80 - 400 - HT* to anatase-
load with
gentamicin HF 20 0.75 -
MC3T3-E1/40% increase in the number of cells after
1 day
65% more cells present after 7 days
50% increase in ALP levels after 3 weeks
[70]
~70 ~15 ~250 -
Immersed in
NaOH then HT*
to anatase
HF 20 0.5 -
Soaked in SBF/ MC3T3-E1/
the kinetics of hydroxyapatite growth=7
The number of
the adhered cells increases= 400%
[172]
100 19 1000 - HT* to anatase
and rutile
NH4H2PO4/
NH4F 20 1 Stirred
HA with thickness of 6µm/bonding
strength;15.3±2.5 MPa [74]
*HT: Heat Treatment
49
2.5.2 Effect of crystalline phase of TiO2 nanotubes on bone cell behaviour
TiO2 nanotubes that were tailored on the surface of titanium using anodic oxidation
could be easily removed by a moderate touch. Heat treatment has been used to increase
the adhesion of nanotubes to the titanium substrates [22]. Crystalline structures of
titania, such as anatase, rutile and a mixture of both, have been observed after this
process, which are dependent on the temperature used [23]. At higher temperatures (≥
600 oC), rutile is the remaining structure along with changes in the shapes of the
nanotubes. However at a lower temperature, for instance 450 ºC, the growth of anatase
crystallites occurs along the length and curvatures of the nanotubes, thus the
morphology remains stable [24]. Loss of fluorine can occur during heat treatment,
which in turn affects cell adhesion and proliferation [151, 177]. The stability and
hydrophilic properties of TiO2 nanotubes can also be changed through annealing. The
diameter of the titania nanotubes annealed at 500 ºC for 2 h decreased by 34.43 %, the
wall thickness of the nanotubes increased by 35.65 % for a selected nanotube diameter
of 25 - 40 nm, and the measured contact angle value changed from 73.15 ° to 17.61 °
after annealing [178]. Although heat treatment at different temperatures did not
influence the surface roughness of the titania nanotubes, the average roughness of the
annealed nanotubes was higher than the as-formed nanotubes [177]. Another interesting
study [179] indicated that the anatase phase of TiO2 nanotubes was fabricated by hybrid
thermal heating using a microwave oven after only 2 or 3 min.
In addition to annealing that changes the hydrophilicity of TiO2 nanotubes (contact
angle value decreased from 24.62 ± 5.23 ° to 10.76 ± 2.35 ° [180]), aging also
influences the hydrophobicity properties of TiO2 nanotubes. Due to the formation of
Ti(OH)4, instead of TiO2, after anodising titanium in an electrolyte based on ethylene
glycol, the hydrophobicity properties of both as-formed and annealed titania nanotubes
increased after 92 days of aging; the contact angle value increased from 24 ° to 42 ° and
from 10 ° to 27 ° for the as-formed and annealed nanotubes, respectively [180]. It can
be deduced that after transferring to a more stable state of TiO2 after aging, Ti(OH)4
becomes more hydrophobic [180]. Anti-aging TiO2 nanotubes were obtained through
optimising the anodising and annealing conditions. It was reported that TiO2 nanotubes
fabricated at the applied potential of 60 V followed by 600 °C annealing maintained
their hydrophilicity longer than those annealed at 300 °C [181]. The viability,
50
proliferation of MC3T3-E1 preosteoblasts, and mineralisation on a mixture of anatase
and rutile crystallite were higher and more regular than on anatase only and the structure
was amorphous [177].
As anatase TiO2 is more biocompatible than amorphous TiO2, the adhesion, ALP
activity and mineralisation of MC3T3-E1 preosteoblast were investigated on titania
nanotubes with a diameter ranging from 20 to 120 nm and a length of 100 to 400 nm.
SEM images of extended MC3T3-E1 preosteoblast cell filopodia on nanotube layers
with different diameters are shown in Fig. 2.9: (a) 20 nm, (b) 50 nm, (c) 70 nm, (d) 100
nm, and (e) 120 nm (×70,000), (f) 120 nm (×30,000) [36]. Despite the increase in cell
proliferation with increases in diameter, there was no other cell behaviour observed for
nanotubes with a diameter of 70 nm and greater [36]. It has been proposed that this
phenomenon is related to the effect of length and roughness of the nanotubes and to the
sensitivity of the filopodia of the cells to the anatase structure [36]. The crystalline form
of TiO2 nanotubes showed a lower tendency to corrosion in Hank’s solution in
comparison to the amorphous phase, due to the thicker layer of oxide at the interface of
the metal surface and the bottoms of the nanotubes [24] and it has been demonstrated
that annealing helps to stabilise the surface of nanotubes, leading to a higher corrosion
resistance [178].
51
Fig. 2.9 SEM images of extended MC3T3-E1 preosteoblast cell filopodia on nanotube
layers with different diameters: (a) 20 nm, (b) 50 nm, (c) 70 nm, (d) 100 nm, and (e)
120 nm (×70,000), (f) 120 nm (×30,000) [36]
52
2.5.3 Effect of hydroxyapatite (HA) coating on the surface of TiO2 nanotubes on
bone cell behaviour
A bone-like apatite layer forms on the modified surface of biomaterials after
implantation in the body. This has been identified as an indispensable stage for bone-
bonding. The nanotubular structure of TiO2 promotes the tailoring of a nano-sized
apatite structure, which has been shown to improve the bioactivity and osteoblast
functions. An increased surface area, and pathways for fluid in between the nanotubes,
is the most obvious reasons for such functionality of titania nanotubes [25]. There are
several methods for calcium phosphate or hydroxyapatite deposition using physical and
chemical techniques, such as plasma spray [182], sputtering [183, 184], electron-beam
physical vapour deposition (EB-PVD) [185], electrodeposition [186-188] and a simpler
method: immersion in a simulated body fluid (SBF) such as a Hank’s solution [22, 189].
A large diameter of 200 nm and a small diameter of 100 nm nanotubes composed of
TiO2, Nb2O5 and ZrO2 formed on the surface of Ti-29Nb-5Zr have been coated by TiN,
HA and HA/TiN using a RF magnetron sputtering system. TiN coating had a columnar
shape and was well spread onto the nanotubes, exhibiting a high film nucleation and
growth rate compared to the HA-coated surface with their coating particles slightly
covered on the tops of the nanotubes and well spread on the surface. Multi-layer coating
of HA/TiN for 1 h deposition of HA and 5 min deposition of TiN showed an entirely
spherical shape and the coatings did not block the nanotube tips which could be an
optimum condition for enhance high Ca/P absorption, in comparison to the coatings of
HA/TiN for 2 h deposition of HA and 10 min deposition of TiN [190]. The
hydroxyapatite (HA) coating made of tooth ash has been coated on 150 and 100 nm
nanotubes on the surface of Ti-xHf by EB-PVD to examine their corrosion behaviours.
It has been shown that the diameter of nanotubes decrease with increasing Hf content
and HA coating covered the tip of nanotubes in lower content of Hf (Ti-20Hf), in
reverse to Ti-40Hf. Good corrosion resistance has been obtained by HA coating on top
of nanotubes in comparison to nanotubes without coating [176]. High corrosion
resistant HA coating using EB-PVD on top of nanotubes has been reported on the
surface of Ti-30Ta-xZr and Ti-30Nb-xZr alloys; with nearly complete covering on Ti-
30Ta-15Zr/Ti-30Nb-15Zr alloys, in comparison to a non-uniformly covering on Ti-
30Ta-3Zr/Ti-30Nb-3Zr alloys [175]. The Ca/P ratio of EB-PVD coated HA on Ti-
53
35Nb-10Zr alloy after heat treatment at 500 °C was around 1.67 and the HA exhibited a
thickness of 150 nm [191].
Uniform flaky and hexagonal columnar calcium phosphate ceramics layer, depending
on the pH of the electrolyte (pH=4, 6 respectively) has been grown on anodised titanium
surface with a pre-alkaline treatment nanotubular oxide layer by electrodeposition.
Bond strength of the calcium phosphate crystals with 100 - 200 nm in diameter and
about 2 µm long which have been started to deposit at the bottom of the nanotubes of
the oxide was from 16 to 19 MPa, whereas calcium phosphate coating onto a polished
non-anodised titanium surface have been easily removed by washing due to its weak
bonding and nonuniformity [192]. In another work [193] the bond strength between HA
and alkaline treated titania nanotubes that has been deposited by pulse electrodeposition
method was from 16 to 44 MPa. The bond strength of the as-deposited HA on
nanotubes was low; it increased after alkali treatment and annealing at 450 and 600 °C.
The ring like structure of sodium titanate that has been formed in alkaline treatment at
the neck of nanotubes has played an important role to improve the bond strength of the
coating with the substrate alongside with the annealing of the HA. Calcium hydrogen
phosphate (CaHPO4·2H2O, DCPD) crystals in nanometer scale precipitated on anodised
titanium with and without annealing when electrochemically polarised in SBF solution
transformed to HA coating in alkaline solution. The bond strength between the HA
coating and the substrate was 7.41 MPa because of both the anchoring in and between
the tubes, in comparison to Ti without anodic oxidation which exhibited a bond strength
of 3.29 MPa [108].
Calcium phosphate coatings grew uniformly on TiO2 nanotubes that were 40 - 110 nm
in diameter and about 0.7 µm in length that were annealed at a temperature up to 600 ºC
and immersed in Hank’s solution, due to the negative charge nature of titania nanotubes
[22]. The Ca/P ratio of the coating layer after immersion in Hank’s solution was 1.16
for 2 days and 1.37 for 7 days, which indicated that the coating layer tended to
incorporate more calcium with immersion time. The adsorption of bovine serum
albumin (BSA) on a high surface area, such as with a calcium phosphate coating,
increased in comparison to the titanium surface covered with native oxide film, and this
enhanced cell adhesion [22]. An alternative immersion method, preloading the TiO2
nanotubes with synthetic hydroxyapatite, was used to efficiently deposit synthetic
54
hydroxyapatite, which is affected by pore size and the ion content in the oxide material
[189]. The Ca/P ratio obtained through this method, which consists of 20 cycles of
alternating immersion in a saturated calcium solution followed by 1 min holding time
and 1 min washing with ultrapure water, was about 1.58, which is close to 1.67, the
Ca/P ratio of synthetic apatite. The deposition of calcium phosphate on annealed TiO2
nanotubes increased as the diameter increased from 15 to 100 nm and the length
increased from 880 nm to 1.2 µm [189].
Apatite formation differed in terms of the nucleation and growth on the TiO2 nanotubes,
which were formed using two different conditions of stirring electrolyte: i) a bath stirred
using a magnetic pellet, and ii) an ultrasonated bath [25]. The diameter of titania
nanotubes was 75 - 110 nm with a thickness of 700 nm when formed in an ultrasonated
bath and 50 - 90 nm with a thickness of 900 nm when formed in a magnetically stirred
bath. The apatite coatings formed on the larger diameter TiO2 nanotubes (75 - 110 nm)
after soaking in SBF for 504 h exhibited a smaller crystallite size (15 nm) and a
complete covering of the surface; whilst the apatite that formed on smaller diameter
nanotubes (50 - 90 nm) exhibited a larger crystallite size of 25 nm and an islands-
covering of the surface. The protein activity of MG63 human osteosarcoma cells was
higher on the finer apatite coating [25]. The thickness of the nanotubes showed only an
insignificant influence on the cell activity. Meanwhile, hydroxyapatite growth as well as
adhesion/proliferation of osteoblasts occurred more effectively on a mixture of anatase
and rutile structures, due to annealing, than on anatase only [24], although it has been
reported that a 10 µm thick hydroxyapatite coating was formed after 4 days immersion
in SBF on the as-formed TiO2 nanotubes that were treated using several dip-and-dry
steps, by which the TiO2 nanotubes were filled and covered with calcium phosphate
nucleation sites [194].
2.6 In vivo effect of micro/nanostructure of the surface of TiO2 nanotubes
According to the highly organised hierarchical structures of bone natural tissues that
consist of nano-, micro-, and macroscale building blocks [32] a few attempts have been
undertaken to study these phenomena through implantation in the body of selected
animals. To translate the role of the micro and nanotopographies on cell functions when
primary osteoblasts were seeded on the HF acid treated/anodised nanotubes substrate,
the micro/nano-textured surface topographies showed more biologically friendly
55
rendering, with more balanced promotion in multiple cell functions than the
microtopography [195]. The HF acid treated/anodised nanotubes substrates with screw-
shape were placed in mandibles of ovariectomised sheep for 12 weeks [196]. The
implant stability quotient (ISQ) values, the maximum pull-out forces, and the bone-
implant contact (BIC) have been investigated. Significantly increased bone volume ratio
and the trabecular number with decreasing trabecular separation evidenced the fact that
the osseointegration of titanium implant could be improved by a hierarchical
micro/nanotextured surface [196]. In another work [197] screw-shaped and anodised
titanium implants have been implanted in the femur condyle close to the knee joint of
New Zealand white male rabbits. The mean pore size of titania nanotubes was ~ 108
nm, the pore size distribution (PSD) was ~ 58 - 150 nm, the porosity was 60 % , the
roughness (Sa) value was 0.65 (± 0.02) µm and the equivalent values for developed
surface area (Sdr) was 14.3 % (± 0.9). The results of the animal studies based on the
osseointegration strengths, new bone formation and bone/cell contact at the bone-
implant interface in comparison to the blasted, moderately rough implants demonstrated
potential applications of titania nanotubes for bone tissue engineering [197]. To
examine the effect of the nanotube diameter on cellular activity, five different nanotube
diameters (15, 30, 50, 70 and 100 nm) have been implanted in the frontal skull of an
adult domestic pig [198]. The BIC for the 50, 70 and 100 nm groups were greater than
control group (untreated) and the BMP-2 expression within the 50, 70 and 100 nm
groups was different. This study showed the important effect of nanotube diameters for
the controlled formation of peri-implant bone around medical implant devices.
Annealed titania nanotubes with pore size of approximately 80 nm and length of 400
nm have been used to investigate the short- and long-term performance of MSCs in
vitro and implanted in the scruff region of the neck male Lewis rats in vivo [199].
Higher cell adhesion, proliferation and viability up to 7 days of culture, higher ALP
activity and 50 % higher calcium and phosphorous concentrations have been observed
in comparison to pure titanium and tissue culture polystyrene as a control system with
no adverse immune response occurred under in vivo conditions [199]. The in vitro
outcome of accelerated osteoblast adhesion by ~ 300 - 400 % [54] of annealed TiO2
nanotubes has been followed by an investigation of in vivo bone bonding in earloop of
rabbits for 4 weeks. Bone bonding strength, bone-implant contact area, new bone
formation, and calcium and phosphorus levels on the nanotube surfaces were increased
56
in comparison to TiO2 grit-blasted surfaces [200]. Further studies have to be undertaken
to investigate the influence of nanoarchitecture and properties of titanium dioxide
nanotubes on bone cell behaviour after implantation in the body of animals to confirm
the in vitro results.
2.7 Effect of TiO2 nanotubes on bacteria attachment
Another important factor in implantation is bacterial infection. It has been reported that
bacterial adherence is lower on titania nanotubes with a smaller diameter (20 nm) rather
than a larger diameter (100 nm), although both Staphylococcus aureus and
Staphylococcus epidermidis exhibited different adhesions on both anodised and un-
anodised Ti6Al4V surfaces [201]. Yu et al. reported that bacteria adhesion on titania
nanotubes increase with an increase in nanotube diameter and reaches the highest value
on 120 nm nanotubes, which is 300 - 400 % increase compared to smooth Ti; although
the proliferation of canine bone marrow stromal cells also increased by as much as some
300 % in contrast to the smooth Ti [202]. The nanotubular layer fabricated on the
surface of Ti-Nb-Zr-Mo (β titanium alloy) with inner diameter of 30 - 50 nm revealed
good antibacterial activity to resist actinomyces viscosus [203]. The smallest nanotubes
fabricated by two-step anodisation on the surface of TiZr alloy with inner diameter of
20 nm exhibited the most efficient antibacterial behaviour against Escherichia coli
bacterium than those with inner diameter of 100 nm [204]. To limit bacteria attachment
whereas cell fate still promote, loading nanotubes with gentamicine [70], silver
nanoparticle [205] and hydrophobicity treatments (using 1H, 1H, 2H, 2H-
perfluorooctyl-triethoxysilane (PTES) [206] or heat treatment [207]) was explored.
2.8 Summary
If bonding between an implant material and bone cannot be formed initially, then
acceptance of the implant by the body, or more precisely the bone cell, fails. It is well
known that titanium and its alloys, in comparison with other metal implants, are more
biocompatible and of these, the Ti-Nb-Ta-Zr alloys have lower Young’s modulus and
higher biocompatibility. If there is fluoride ion in electrolyte, a controllable nanotube
TiO2 can be fabricated by anodising oxidation on titanium, titanium alloying metals
(niobium, tantalum and zirconium) and biocompatible titanium alloys, that is, Ti-Nb-
Ta-Zr alloys and binary Ti-Nb, Ti-Ta, and Ti-Zr alloys. The nanotubes on titanium
57
promote bonding to bone due to their high surface area and the ability of cell-
interlocking.
The tube diameter and the length of TiO2 nanotubes would depend on the conditions of
electrochemical oxidation such as the type, concentration and the pH of the electrolyte,
the applied potential and the anodisation time. The tube diameter can be increased with
increasing the applied potential if the pH value remains constant; and the tube diameter
is pH-independent if the applied potential remains constant. In comparison to using an
aqueous electrolyte, a higher applied potential is required for obtaining the same
diameter nanotubes if an organic electrolyte is used. The tube length or the thickness of
the nanotube layer increases with the time of anodisation but there is always a point at
which the thickness remains constant with further increasing time. It is worth noting
that in an acidic electrolyte with pH value less than 1, although nanotubes form, their
length will not increase by increasing the time of anodisation. In an organic electrolyte,
although the process needs longer time, considerable and longer nanotubes can be
tailored beyond a certain amount of water content.
Nanotubes fabricated on implant material surfaces provide great potential in promoting
cell adhesion, proliferation and differentiation. A high surface area, especially the
spacing between nanofeatures that is provided by the nanotubes, offers convenient
conditions for interlocking with bone cells and the penetration of body fluid. Cells
respond to nano scale structures and nanopatterns are also sensible to the chemical and
mechanical properties of the surface, such as surface energy, water contact angle and
zeta potential. A low contact angle influences cell differentiation and proliferation.
Titania nanotubes after annealing existed in a mixture of anatase and rutile, exhibited a
more stable morphology, and demonstrated a higher proliferation of MC3T3-E1. A
hydroxyapatite layer with a thickness of up to 10 µm on the surface of TiO2 nanotubes,
with different nanotube sizes and forms, can be obtained. In order to establish the
optimum nanotopography for cell response, some investigations have been undertaken.
The results of in vitro studies suggested that the optimum diameter of TiO2 nanotubes is
less than 100 nm and that 15 nm is the best although there are some controversy
discussions. If the mechanism of the cell response is clarified, the cell fate at the surface
boundary will be more controllable. Nanotubes also offer possibility in bacterial
infection control through loading the tubes with antibacterial agents. More studies are
58
needed in order to find the optimum size of nanotubes in length and diameter for the
sensing element of a bone cell to be recognised and adhered to and more in vivo
investigations to confirm the in vitro results, especially for the new biocompatible
titanium binary, ternary and quaternary alloys.
59
Chapter 3 Materials and methods
3.1 Introduction
This chapter describes the sample preparation, surface modification and characterisation
procedures for the investigation of nanotubular layer on the surface of biocompatible
metals. The materials and methods requirements for of the assessment of bioactivity i.e.
mineralisation of hydroxyapatite and biocompatibility i.e. cell adhesion of osteoblast
cell (SaOS2) are also elaborated in this chapter.
3.2 Sample preparation
The main starting material used in this study was Ti50Zr alloy (wt % hereafter). Alloy
disc samples with a diameter 8 mm and 2 mm thick were cut from ingot using electrical
discharging machining (EDM). The other samples were tantalum (Ta, 10×10×0.1 mm),
niobium (Nb, 10×10×0.05 mm), zirconium (Zr, 10×10×0.05 mm) and commercially
available pure titanium (CP-Ti, 10×10×0.05 mm) foils (Baoji Boxin Metal Materials
Co. Ltd., Shaanxi, China).
All the surfaces of the disc samples were ground using silicon-carbide papers to 2400
grit and one of the surfaces was polished down to a finish of 1 μm diamond paste. All
the samples were degreased by sonification in methanol, isopropanol, acetone and
ethanol for 15 min each progressively between grinding and polishing processes.
Finally, the samples were washed with deionised water and dried using a nitrogen (N2)
gas stream.
3.3 Fabrication of nanoporous and nanotubular layers
In order to set the electrochemical connections for anodisation, the Ti50Zr discs were
spot-welded to titanium foil and connected to the anode of an electrochemical cell; the
Ta, Nb, Zr and CP-Ti foils were connected to the anode directly as working electrode.
Two electrode configurations were set up using a DC power supply with a 1 cm2
platinum plate as a counter electrode, and placed 4 cm from the working electrode.
Electrochemical experiments were performed at room temperature with stirring. The
compositions for the aqueous electrolyte were listed in Table 3.1. All the chemicals
were of Sigma-Aldrich reagent grades.
60
Table 3.1 Compositions for the aqueous electrolyte used for fabricating nanoporous and
nanotubular layers
Type of sample Concentration of NH4F Concentration of
(NH4)2SO4
Others
Ti50Zr 0.1 - 0.5 wt % 1 M -
CP-Ti 0.1 - 0.5 wt % 1 M -
Ta 3.3 wt % - 1 M H2SO4
Nb 3.3 wt % - 1 M H2SO4
Zr 0.3 wt % 1 M addition of H2SO4
to attain pH 5
In terms of non-aqueous electrolyte, 0.5 wt % of NH4F was used and 5 and 10 wt % of
pure water were added to ethylene glycol. Different applied potentials, i.e., 5, 10, 15,
20, 25, 30 and 35 V, were used to create a variety of nanoscale sizes and
nanotopographies of nanoporous and nanotubular layers. After the electrochemical
treatment the samples were rinsed with deionised water for 5 min and dried with N2
stream. The samples which were treated in non-aqueous electrolyte were rinsed with
acetone. For cell culture studies samples were firstly pretreated by immersion in 0.5 M
NaOH at 50 ˚C for 2 min, then rinsed with deionised water to increase
biomineralisation. Some samples were soaked at 70 °C for 14 days in pure water with
periodic replacements (three times per day) to reduce any contamination of remaining
chemicals from the electrolyte. The nanotubular layer formed in non-aqueous
electrolyte were post-treated by immersion in 0.5 M HCl at room temperature for 1 min,
then rinsed with deionised water for 1 min to dissolve the remained upper nanoporous
layer.
Annealing of samples was undertaken by heating in air at 500 ˚C for 3 h at a heating
rate of 2 ˚C/min in a conventional muffle furnace (Nabertherm LT15/13/P330;
Nabertherm GmbH, Lilienthal, Germany) after any necessary treatments after
anodisation. The nanotubular layers formed in non-aqueous electrolyte were annealed
61
by heating in air at 500 ˚C for 2 h at the heating rate of 2 ˚C/min. The annealing of
nanoporous and nanotubular layer on the surface of Ta, Nb and Zr was performed in air
at 290 °C for 10 min.
3.4 Surface characterisation
3.4.1 Surface morphology and chemical composition characterisation
Metallographic characterisation of the samples after each surface modification was
carried out using a field-emission scanning electron microscope (FESEM, ZEISS
SUPRA 40 VP) equipped with electron dispersive X-ray spectroscopy (EDS). The SEM
observation was conducted directly on the surface of nanoporouse and nanotubular
layers and mineralised hydroxyapatite (HA) at 3 kV. EDS analysis was performed with
an accelerating voltage of 20 kV. Determination of the surface crystallographic structure
and phases were performed by means of X-ray Diffraction (XRD, Bruker D8; Bruker
Pte Ltd, Singapore) using Cu Kα incident radiation at 40 kV and 40 mA with a 2θ
scanning step size of 0.02 ° from 10 to 90 °. The surface chemistry was analysed using
X-ray photoelectron spectroscopy (XPS, AXIS Ultra DLD, Kratos Analytical Inc.,
Manchester, UK) with a monochromated Al Ka source at a power of 180 W (15 kV and
12 mA) and a hemispherical analyser operating in the fixed transmission mode with a
standard aperture (analysis area: nominally 0.3 mm × 0.7 mm). The nanotube size
distribution measurements were generated from 100 nanotubes on different positions for
each of three samples (300 measurements). ImageJ software (Wayne Rasband, National
Institutes of Health, America) was used for measurements. By plotting the histogram of
nanotube sizes their distributions were presented by normal and Weibull distribution
using origin lab software. The highest percentage of normal distribution of nanotube
sizes represented by mean ± variance while it presented by scale parameter (β) ± shape
parameter (α) for Weibull distribution [208, 209].
3.4.2 Surface topography and water contact angle and surface energy
measurement
Roughness parameters were measured using a 3D-Profilometer (Bruker, Contour GT-
K1; Bruker Pte Ltd, Singapore) and analysed using the SurfVision software (Veeco
Instruments Inc.; Plainview, NY, USA). Surface topography was characterised by
several roughness parameters. Mean Roughness (Sa) and Root Mean Square (RMS)
62
Roughness (Sq) are some of the amplitude parameters studied in this thesis. These
parameters measure the vertical characteristics of surface deviation over the 3D surface.
A comprehensive description of the surface textures such as peaks, valleys and the
spacing can be represented by these parameters [210]. The third central moment of
profile amplitude, skewness of a 3D surface texture (Sskw), represents the dominant
nature of topography. Thus, a value of Sskw > 0 indicates high peaks about the mean
plane and Sskw < 0 indicates deep valleys such as would be formed from scratches [210].
The sharpness of the height distribution is indicated by a fourth central moment of
profile amplitude; i.e., the kurtosis (Sku) of the 3D surface texture where Sku = 3.0 for a
normal distribution of heights. However, when a few high peaks are spread out over a
3D surface, then the distribution is defined as ‘platykurtoic’ and Sku < 3.0; whereas Sku
> 3.0 in the instance of the surface exhibiting a high proportion of high peaks and low
valleys and this distribution is described as being ‘leptokurtoic’ [210].
The water contact angle was measured by sessile drop technique at room temperature
using a goniometer (NRL C.A. Goniometer, Ramé-hart, Inc.; Succasunna, NJ-USA).
The drop size was 5 µL. Five droplets of each liquid (pure water and glycerol) were
used for each sample. Surface energy is calculated based on Owens-Wendt (OW)
method [211], as in the following equation:
√
√
(3-1)
Where L is the liquid surface tension and Ld, L
p are its dispersive and polar
components, respectively. Sd and S
p refer to the dispersive and polar components of the
solid surface tension, S, which is the sum of Sd and S
p. The surface tension, dispersive
component, and polar component for water are =72.8 mJm-2, d=21.8 mJm-2, and
p=51.0 mJm-2 and for glycerol, =63.4 mJm-2, d=37.0 mJm-2, and p=26.4 mJm-2,
respectively [212]. An average of five readings per sample was acquired for the
roughness parameters, water contact and surface energy measurements.
3.5 Bioactivity assessment by SBF soaking
The in vitro bioactivity assessment of metal samples were carried out by soaking the as-
formed, annealed metal oxides samples and their respective bare metals in simulated
body fluid (SBF) and incubating at 37 ˚C for 1 day, 1 week, 2 weeks and 3 weeks. An
63
SBF solution is known as a solution with ion concentrations approximately equal to
those of human blood plasma. A modified (m-SBF) [213] was prepared by dissolving
chemicals in Table 3.2 in pure water as previously reported [213]. The m-SBF was
buffered at pH 7.4 at 37˚C using the prepared solution according to the Table 3.2 of
HEPES and 1 M NaOH. There are some important notes which have to be considered
while preparing m-SBF which has been picked up from reference [214]. An appropriate
preparation method has to be undertaken to avoid precipitation of apatite during the
preparation because the solution is supersaturated with respect to apatite. A plastic
beaker without any scratch should be used in this respect. When the temperature of 700
ml pure water in a water bath reaches 36.5 ± 0.5 ºC under stirring the reagents of the 1st
to the 6th order were dissolved into the solution one by one in the order given in Table
3.2. It means that a reagent should be dissolved only after the preceding one (if any) is
completely dissolved. The hygroscopic reagents such as KCl, K2HPO4 - 3H2O, MgCl2 -
6H2O, CaCl2, Na2SO4 should be measured in as short a period as possible. Then the
electrode of the pH meter should be inserted into the solution before adding HEPES
solution little by little taking careful note of the pH change. The pH has to be kept
between 7.42 and 7.45 at 36.5 ± 0.5 ºC by adding NaOH 1M until all the 100 ml of
HEPES solution to be added because increase or decrease in local pH (out of 7.42 -
7.45) of the solution can lead to the precipitation of calcium phosphate. Then the
temperature of the solution should be adjusted to 36.5 ± 0.2 ºC and the pH of the
solution should be adjusted by dropping 1M NaOH little by little at a pH of 7.42 ± 0.01.
The final pH adjustment after reaching 1000 ml should be 7.40 exactly at 36.5 ºC. Then
the solution should be poured from the beaker into 1000 ml volumetric flask and to be
kept in water to cool down to 20 ºC when it can be filled with pure water to the marked
line. The m-SBF solution should be kept in plastic container at 5 - 10 ºC in a refrigerator
and shall be used within 30 days after preparation.
64
Table 3.2 The reagents and their amounts for preparing 1000 mL m-SBF
Reagents Amount (g) Supplier
NaCl 5.403 Merck
NaHCO3 0.504 Sigma-Aldrich
Na2CO3 0.426 Merck
KCl 0.225 Chem-Supply
K2HPO4- 3H2O 0.230 Merck
MgCl2 - 6H2O 0.311 Chem-Supply
2-(4-(2-hydroxyethyl)-1-
piperazinyl) ethane sulfonic acid
(HEPES)
17.892 g HEPES has dissolved in
100 ml 0.2 M NaOH
aMResCO
CaCl2 0.293 Sigma-Aldrich
Na2SO4 0.072 Sigma-Aldrich
NaOH 1 M Merck
The ability of the pre-treated samples to form apatite was evaluated in a static SBF
environment. The samples were removed after incubation for 1 day, 1 week, 2 weeks
and 3 weeks in m-SBF; then rinsed with deionised water and dried at room temperature
for 24 h. The change of pH value during incubation was measured using pH/mV &
Ion/pH meter Series (Oakton, Eutech instruments Pte Ltd, USA). The atomic percentage
of calcium to phosphate ratio of HA is calculated using EDS results.
3.6 Assessing cell responses on nanotubes with different nanoscale
dimensions and surface topographies
All the samples for the cell culture studies were sterilised in a muffle furnace at 180 ˚C
for 3 h. Subsequently they were placed in a well in the cell culture plate. Osteoblast-like
cell (SaOS2) (Barwon Biomedical Research, Geelong Hospital, Victoria, Australia)
were seeded on the surfaces of the nanotubular TiO2-ZrO2-ZrTiO4 samples with
65
different inner and outer diameter, wall thickness and different surface roughness
parameters. The density of the seeded cell was 5×103 cells per well (100 mm2 area). To
measure the in vitro proliferation of the SaOS2 cells a MTS assay was used. The cells
were dehydrated by immersion in a buffer solution in which the ethanol concentrations
were increased progressively every 10 min to 60, 70, 80, 90 and 100 %. This procedure
was followed by chemical drying using hexamethyldisilazane (HMDS, Sigma - Aldrich,
Australia) for 10 min. The morphology of dehydrated and dried cells was observed
using SEM after cell culturing and coating with a gold layer. The cell morphology was
also observed using a confocal microscopy (Leica SP5, Leica Microsystems, Germany).
The cell-seeded samples after cell culture were fixed in paraformaldehyde, and then
permeabilised with triton-X 100 in phosphate-buffered saline (PBS) (Sigma - Aldrich,
Australia) for 10 min each at room temperature for the confocal microscopy
observation. Then the samples were stained with 1 % phalloidin and 40-6-diamidino-2-
phenylindole for 40 min at room temperature. Between each of the steps three washes
by PBS were included. If storage of stained samples was required they were stored in
PBS at 4 °C and the observations were conducted within a week of staining. Significant
differences in the cell number were analysed using one-way ANOVA (ρ < 0.05).
3.7 Nanohardness and elasticity measurements
By recording the penetration depth of an indenter into a sample material while the
applied load is measured, the area contact and consequently the hardness of the material
can be calculated as well as many other mechanical properties. These mechanical
properties can be the strain hardening index, fracture toughness, yield strength and
residual stress [215].The nano-mechanical characteristics of the nanotubular layers such
as hardness and elastic modulus were measured by using a nanoindenter (IBIS, Fischer-
Cripps Laboratories Pty Ltd, Sydney, Australia) with a three-side pyramid Berkovich
diamond tip which was calibrated against fused silica standard. The face half-angle of
the most popular geometry for nanoindentation testing; the three-sided Berkovich
indenter is 65.27 ° - 65.3 ° [215]. The samples were put on the platform and then used
load-unload nanoindentation testing using a needle that was pressed into the sample
surface. The maximum effective load was 10 mN. Once the maximum prescribed force
was reached, loading stopped and the load was maintained constant for 10 second
before unloading. The data reflects the mean values from multiple indentations (a
66
matrix of 2×2) separated by 10 μm which were made in each sample. Oliver-Pharr
method [216] used to draw the curve of loading and unloading results to calculate the
hardness, reduced elastic modulus and elastic modulus. The following equations were
used for calculation:
hc = hmax – ε Pmax / S (3-2)
where ε is a constant of 0.75 that depends on the geometry of the Berkovich indenter. S
is the slope of the unloading curve at maximum point before start of unloading. The
contact area A is then
A = 24.5 hc2
(3-3)
the hardness is estimated as below when the contact area is determined as above
H = Pmax / A (3-4)
According to the relationship of the elastic modulus to contact area and the measured
unloading stiffness (S) it is calculated as follows:
√ √ ⁄ (3-5)
Young’s modulus E of nanotubular layer with Poisson’s ratio υ of 0.27 [217] is
calculated using Eq. (3-6):
1/Er = 1 - υ2 / E – 1 - υi2 / Ei (3-6)
Where Ei and υi are elastic constants and Poisson’s ratio of indenter; 1141 GPa and 0.07
respectively.
67
Chapter 4 Nanotubes formed in aqueous electrolyte
Abstract
Titanium and its alloys are able to grow a stable oxide layer on their surfaces and have
been used frequently as substrates for anodisation in an electrochemical surface
treatment. A nanotubular oxide layer is formed in the presence of fluorine anion (F-) via
anodisation due to the competition between oxide formation and solvatisation. In this
chapter, a highly ordered titania-zirconia-zirconium titanate (TiO2-ZrO2-ZrTiO4)
nanotubular layer was formed on the surface of Ti50Zr alloy via anodic oxidation in an
F- containing electrolyte. The sizes of the nanotubes (i.e., the inner and outer diameters,
and wall thicknesses), morphology, crystal structure, hydrophilic properties and
components of the TiO2-ZrO2-ZrTiO4 nanotubular layer before and after annealing were
examined by scanning electron microscopy (SEM), thin film X-ray diffraction, X-ray
photoelectron spectroscopy (XPS) analysis and water contact angle measurements.
The results indicated that the mean inner diameter, outer diameter and wall thickness of
the as-formed TiO2-ZrO2-ZrTiO4 nanotubes were distributed in the ranges of 13 -
81 nm, 32 - 131 nm and 10 - 22 nm, respectively, and depended on the F- concentration
of the electrolyte and the applied potential during anodisation. The number of smaller
nanotubes increased with increasing F- concentration and the mean nanotube inner and
outer diameters increased with increasing applied potential. The as-formed TiO2 and
ZrTiO4 nanotubes exhibited an amorphous structure and the as-formed ZrO2 nanotubes
displayed an orthorhombic structure. These phases transformed into anatase TiO2 and
orthorhombic ZrO2 and ZrTiO4 after annealing. The hydrophilic properties of the TiO2-
ZrO2-ZrTiO4 nanotubular layer were affected by the size distribution of the nanotubes.
The surface roughnesses and the nanotubular character transformed the nanotubes to
exhibit superhydrophilic properties after annealing. The TiO2-ZrO2-ZrTiO4 nanotubular
surface on Ti50Zr alloy exhibited higher surface energy than that of the TiO2
nanotubular surface on commercially pure (CP) titanium.
4.1 Introduction
Titanium (Ti) is the fourth most plentiful metal on earth, lighter (4.5 g/cm3), stronger
(480 MPa yield strength) and more expensive than steels (~ 7.8 g/cm3 and 175 -
750 MPa yield strength, respectively) [218]. Titanium alloys possess exceptional high
68
specific strength and excellent corrosion resistance because of their ability to form a
protective oxide layer on their surfaces [219]. Applications in aerospace, biomedical
and cryogenic industries are examples where titanium and its alloys are used other than
as an alloying element for other metals [218, 219]. Titania (TiO2) that forms on titanium
at room temperature is amorphous [11] but in crystalline form it exists in 3 phases as
brookite, anatase and rutile. Titania has a wide range of applications, such as paint
pigment, whitener for paper and rubber and as a finishing agent in porcelain enamels
[218]. The electrical and optical properties of TiO2 depend on their band gaps, that is
3.0 eV for rutile, 3.2 eV for anatase and 3.2 - 3.5 eV for the amorphous phase; all of
which are influenced by the relative Ti3+ content [11]. The TiO2 surface has been
employed in gas phase adsorption and catalysis functions; for example CO oxidation,
selective reduction of NOx and O2 and water decomposition. The reactive sites are the
surface defects such as OH groups, oxygen vacancies and the presence of Ti3+ [11].
Improved functionalities of TiO2 in the above mentioned applications can be obtained
by increasing its specific surface area.
An important example where surface modification can be implemented to beneficially
alter material characteristics concerns the fabrication of TiO2 nanotubes by the anodic
oxidation method. The nanotube arrays are able to impart the substrates with novel
functionalities in various applications such as photoelectrochemistry, photocatalysis,
dye sensitised solar cells, and electrochromic devices. Anodisation, an electrochemical
method that has been used for fabrication of nanotubes for over a decade [17, 19, 21,
172, 220, 221]; is simple, reproducible and versatile. Titanium oxide nanotubes with
desirable morphological characteristics can be obtained by controlling the
electrochemical conditions such as the applied potential, time of anodisation and type
and concentration of the electrolyte. The shape and the capillary force of the nanotubes,
along with a highly ordered morphology and large surface area, make them an excellent
choice for the before-mentioned applications. Post processing by annealing the TiO2
nanotubes on the titanium alloy surface transforms the phase structure from amorphous
to anatase or a mixture of anatase and rutile. Other applications of TiO2 nanotubes
include gas sensing (CO, H2, NOx), solution sensing (O2 gas), a substitute for carbon
based supports in methanol fuel cells, as nano test tubes and biosensors [11].
69
Zirconium is one of the sister metallic elements of titanium and hafnium, also forms a
very stable, cohesive and protective oxide on its surface [222]. Zirconium has excellent
mechanical, chemical and thermal properties that enable it to be used as catalyst and
corrosion resistant structure materials such as pressure vessels, heat exchangers, pumps
and valves as well as alloying element for β titanium type alloys [223]. Acid and NOx
reduction catalyst, optoelectronic and biomedical application have been reported for
zirconium oxide known as zirconia [224-226]. Zirconia nanotubes can grow on the
surface of zirconium by anodisation in aqueous and organic electrolyte containing F- ion
under different formation conditions [53, 132, 133].
There has been great interest in fabricating nanotubes onto the surface of binary
titanium alloys; for example, Ti-Zr [105, 110], Ti-Ta [105, 165], Ti-Nb [105, 113], Ti-
Al [105], Ti-Mo [227] and Ti-W [228]. Metal oxide nanotubes have been grown on the
surfaces of these alloys and the diameter, shape and distribution of nanotubes were
related to the alloy element concentrations [105-107, 110-113]. Only a few studies have
examined the fabrication of ZrTiO4 nanotubes for applications as microwave resonant
components, optical devices and refractory ceramics [110-112, 229]. The functionalities
of nanotubes fabricated on titanium alloys for biomedical applications have also been
encouraging; hence the focus of this current investigation.
It has been reported that Ti50Zr (wt %) alloy possesses superior mechanical properties
[230] and excellent biocompatibility [231, 232]. Nanotubular oxide layer has been
fabricated on its surface and the effect of Zr content on the morphology of anodic oxide
layers [105, 233] and their corrosion behaviours [162, 232], crystal structure and
optical properties [229] have been discussed previously. The formation of single layer
[112] or multilayers [111] of self-organised zirconium titanate nanotubes, and the
control of their morphologies and compositions [110] have also been reported.
This chapter investigated the fabrication of TiO2-ZrO2-ZrTiO4 nanotubes onto the
surface of the Ti50Zr alloy via anodisation and contrasted to TiO2 nanotubes fabricated
on the surface of a commercial pure titanium (CP-Ti). A new approach for the
characterisation of nanotubes, for the first time, in terms of two different methods of the
size distribution of the nanotubes was described and this was affected by the fluoride
ion (F-) concentration of the electrolyte and the applied potential. The hydrophilic
properties and the roughness changes of the nanotubular surfaces before and after
70
annealing were characterised using a new set of surface roughness parameters such as
Sskw, Sku and SI to describe the topography of nanotubular surface.
4.2 Materials and methods
The disc samples of Ti50Zr (8 2 mm) prepared as described in Chapter 3 were
polished down to a finish of 1 μm diamond paste and then degreased by sonification in
methanol, isopropanol, acetone and ethanol. Finally, the samples were washed with
deionised water and dried using a nitrogen (N2) gas stream. CP-Ti foils 10×10 mm were
also degreased and washed by means of the identical procedure.
The Ti50Zr discs were spot-welded to titanium foil and connected to the anode of an
electrochemical cell; the CP-Ti foils were connected to the anode directly. Two
electrode configurations were set up using a DC power supply with a 1 cm2 platinum
plate as a counter electrode, and placed 4 cm from the working electrode.
Electrochemical experiments were performed at room temperature with the electrolyte
composed of 1M (NH4)2SO4 with the addition of small amounts of NH4F (Sigma-
Aldrich reagent grades). Electrolytes with different concentrations of fluorine anions (F-
) ranging from 0.1 to 0.5 wt % were used to investigate its effect on the nanotube
morphologies. After the electrochemical treatment the samples were rinsed with
deionised water for 5 min and dried with N2 stream.
The metallographic characterisation of the samples was carried out using a field-
emission scanning electron microscope (FESEM, ZEISS SUPRA 40 VP). Annealing of
the samples was carried out in air at 500 °C for 3 h in a muffle furnace (Nabertherm
LT15/13/P330). The nanotube size distribution measurements were generated from 100
nanotubes on different positions for each of three samples (300 measurements). Phase
characterisation was performed by means of X-ray Diffraction (XRD, Bruker D8) using
Cu Kα incident radiation at 40 kV and 40 mA with a scanning step size of 0.02 ° from
10 to 90 ° (2θ). The surface chemistry was analysed using X-ray photoelectron
spectroscopy (XPS, AXIS Ultra DLD, Kratos Analytical Inc., Manchester, UK) with a
monochromated Al Ka source at a power of 180 W (15 kV and 12 mA) and a
hemispherical analyser operating in the fixed transmission mode with a standard
aperture (analysis area: nominally 0.3 mm × 0.7 mm). Roughness parameters were
measured using a 3D-Profilometer (Bruker, Contour GT-K1) and analysed using the
71
SurfVision software. Water contact angle was measured using a goniometer (NRL C.A.
Goniometer, Ramé-hart, Inc.). The surface energy was calculated using Eq. 3-1. An
average of five readings per sample was acquired for the roughness parameters, water
contact and surface energy measurements. A laboratory nanoindenter (IBIS, Fischer-
Cripps Laboratories Pty Ltd, Sydney, Australia) was applied to measure hardness and
elastic modulus of nanotubular layer. A maximum force of 10 mN was chosen for the
nanoindentation. The loading and unloading curves were drawn and the hardness,
reduced elastic modulus and elastic modulus were calculated using Oliver-Pharr method
[216] (Chapter 3). A matrix of 2 × 2 μm separated by 10 μm measurement was applied
for each sample.
4.3 Results and discussion
4.3.1 Formation of TiO2-ZrO2-ZrTiO4 nanotubes
The nanotube inner and outer diameters, layer thickness and the shape (wavy or smooth
walls) were dependent on the applied potential, time of anodisation and the type and
concentration of the electrolyte. Four categories of electrolytes for anodising titanium
dioxide nanotubes have been reported; i.e., (i) acidic, (ii) buffered, (iii) polar organic
electrolytes, and (iv) nonfluoride based electrolyte [84]. For all of these electrolytes, the
diameter of the anodised nanotubes was controlled by the applied potential and the
nanotube diameter increased with increasing voltage. The nanotube length (in other
words, the nanotube layer thickness) is controlled by the anodisation time and it
increased with increasing time until a steady state was reached at which stage there was
no growth in thickness [110].
Titania-zirconia-zirconium titanate (TiO2-ZrO2-ZrTiO4) nanotubes were fabricated on
the surface of Ti50Zr alloy via anodisation at an applied potential of 5 - 20 V and time
of 2 h in the electrolyte of 0.1 - 0.5 wt % of NH4F. Fig. 4.1 shows the SEM images of
the nanotubes on the Ti50Zr alloy. Fig. 4.1(a-1) shows the microstructure of Ti50Zr
alloy after etching which exhibited a needle-like structure. Fig. 4.1(a-2) shows two
images; the left shows the surface after peeling of the nanotubes, demonstrating that the
original grain structure of the alloy was reflected in the subsequent growth of the
nanotubes. A needlelike morphology composed of different orientations of the
nanotubes can be observed from the top view of the nanotubes on the right of Fig. 4.1(a-
72
2); and a high magnification top view image, as shown in Fig. 4.1(b), shows the
equivalent patch-like regions when the time or the applied potential was not sufficient
enough. The cross-section view of the nanotubes, Fig. 4.1(c), indicates an average
thickness (i.e., nanotube length) of 34.1 µm. The bottom view of the nanotubes reveals
different sizes; in particular, features located along the grain boundaries appeared to be
larger than those inside phases, Fig. 4.1(d).
Fig. 4.1 SEM images of TiO2-ZrO2-ZrTiO4 nanotubes: (a-1) microstructure of etched
Ti50Zr, (a-2) top view of patterned nanotubes of different phases exhibited in the
microstructure of Ti50Zr alloy, (b) nanoporous patches of different phases, (c) cross
section of the nanotubes showing the nanotube length, (d) view of the nanotubes from
the bottom, (e) top view and cross section of separated nanotubes of and (f) top view of
damaged nanotubes
73
The effect of the F- concentration in the electrolyte on the nanotube parameters such as
inner diameter (Di), outer diameter (Do), wall thickness (Wt), roughness (Sa, Sq, Sskw,
Sku), wettability and surface energy at constant applied potential and anodisation time,
as well as the effects of the applied potential on the nanotube parameters at constant F-
concentration in the electrolyte and the anodisation time were investigated. The SEM
images of the TiO2-ZrO2-ZrTiO4 nanotubes fabricated on Ti50Zr alloy under different
anodisation conditions and subsequently annealed at 500 oC for 3 h are shown in Fig.
4.2. The top view of the TiO2-ZrO2-ZrTiO4 nanotubes fabricated in 0.4 wt % NH4F at
20 V for 2 h is shown in Fig. 4.2(a). The nanotubes after annealing at 500 oC for 3 h
exhibited a slightly larger mean inner and outer diameter, Fig. 4.2(b). By increasing the
F- concentration to 0.5 wt % NH4F, the nanotubes fabricated at the same applied
potential of 20 V for 2 h revealed slightly smaller mean inner diameters and larger mean
outer diameters, Fig. 4.2(c). The cross-section view of the nanotubes fabricated at 0.5 wt
% NH4F at a lower applied potential of 15 V for 2 h is shown in Fig. 4.2(e) and the top
view of the nanotubes annealed at 500 oC for 3 h is exhibited as Fig. 4.2(f).
74
Fig. 4.2 SEM images of TiO2-ZrO2-ZrTiO4 nanotubes: (a) top view of as-formed
nanotubes anodised in 0.4 wt % NH4F at 20 V, (b) the nanotubes annealed at 500 ºC for
3 h, (c) nanotubes as-formed in 0.5 wt % NH4F anodised at 20 V, (d) the nanotubes
annealed at 500 ºC for 3 h, (e) nanotubes as-formed in 0.5 wt % NH4F anodised at 15 V,
and (f) the nanotubes annealed at 500 ºC for 3 h
The application of a suitable voltage to the surface of a metal (M) allows the formation
of Mn+ ions that can (i) dissolve, or (ii) react with existing O2- and form an insoluble
metal oxide layer on the surface of metal, or (iii) partially dissolve due to the
composition and conditions of the electrolyte [234].
75
Titania nanotubes can be formed on the surface of titanium when it is used as working
electrode in an electrochemical cell with platinum as counter electrode and applying a
constant voltage between 1 - 30 V in an aqueous electrolyte, or between 5 - 150 V in an
organic electrolyte [234, 235]. The oxidation and reduction reactions at anode and
cathode of the cell forms titanium dioxide in the early minutes, according to the
reactions described by Eqs. (4-1) - (4-4). The reaction in aqueous electrolyte for water
can be given:
2H2O 2O2- + 4H+ (4-1)
The oxidation reaction at the interface of the titanium surface and electrolyte at a
constant applied potential is:
Ti Ti4+ + 4 e- (4-2)
Then,
Ti4+ + 2O2- TiO2 (4-3)
The reduction reaction at the surface of counter electrode, normally platinum, under
these conditions is:
4H+ + 4e- 2H2 (4-4)
In terms of formation of TiO2-ZrO2-ZrTiO4 or TiZrO4 the following reactions are
possible:
Zr Zr4+ + 4 e- (4-5)
Zr4+ + 2O2- ZrO2 (4-6)
2Ti4+ + 2Zr4+ + 8O2- TiO2 + ZrO2 + ZrTiO4 or TiZrO4 (4-7)
The presence of F- allows a complex of titanium and fluoride that is soluble in the
electrolyte to be created according to the mechanism proposed in Eqs. (4-8) and (4-9)
and a porous layer forms. Nanotubes grow when there is equilibrium between the
oxidation of titanium and dissolution of the hexafluorotitanate anion [11].
Ti4+ + 6F- [TiF6]2- (4-8)
Zr4+ + 6F- [ZrF6]2- (4-9)
76
If fluorine anions are not present in the electrolyte, then the thickness of the oxide layer
is limited and the current exponentially decays due to decreasing availability of Ti4+,
Zr4+ and O2- at the interface between the titanium alloy surface and the titanium-
zirconium oxide layer. The Ti4+ and Zr4+ at the interface of the oxide layer and
electrolyte will be precipitated as a hydroxide layer if a soluble condition is not
provided. In the presence of fluorine anions, a soluble complex of Ti4+ and Zr4+ forms
as presented by Eqs. (4-8) and (4-9).
The ionic radius of F- is smaller in comparison to that of O2-. Thus, more fluoride ions
will be available at the interface of the titanium surface and the oxide layer since this
will be a diffusion controlled process relative to the larger size of O2- [58].
Figure 4.3 shows an illustration of the electrochemical cell employed for fabricating
TiO2 nanotubes layer.
Fig. 4.3 (a) Illustration of an electrochemical cell that indicates the electrolyte ions
species, (b) SEM image of top view of nanoporous TiO2-ZrO2-ZrTiO4 fabricated on a
Ti50Zr alloy after anodisation for 15 min in 0.1 wt % NH4F, 5V, and c) SEM image of
top view of TiO2-ZrO2-ZrTiO4 nanotubes fabricated on Ti50Zr alloy after anodisation
for 2.75 h in 0.3 wt % NH4F, 5V
77
The F- concentration in the electrolyte, nanotube inner diameter, outer diameter and wall
thickness are denoted as CF-, Di, Do, and Wt here after respectively. The Di, Do and Wt of
the nanotubes anodised in the electrolyte with CF- increasing from 0.3 to 0.5 wt % were
measured and are quantitatively presented in Fig. 4.4.
78
79
Fig. 4.4 Histograms of as-formed and annealed TiO2-ZrO2-ZrTiO4 nanotube parameters anodised at 20 V for 2 h for different concentrations of fluorine anion
(F-): (a,d) inner diameter (Di), (b,e) outer diameter (Do), and (c,f) wall thickness (Wt), Note: The nanotube size distribution graphs were generated from100
nanotubes on different positions for each of three samples (300 measurements)
80
During anodisation, a nanoporous oxide layer was observed growing on the surface of
the Ti50Zr alloy after 15 min in an electrolyte with F- concentrations ranging from 0.1 -
0.5 wt % at an applied potential of 20 V. Although nanotubes grew underneath this
nanoporous layer after 2 h as shown in Fig. 4.2(b), the nanoporous layer existed when
the F- concentration was low; i.e., about 0.1 - 0.2 wt %. The mean Di of the nanotubes
was 40.3 ± 12.6 nm, the mean Do was 66.6 ± 14.6 nm, and the mean Wt was 10.0 ±
2.4 nm when the Ti50Zr alloy was anodised in an electrolyte with a CF- = 0.1 wt %.
When the CF- increased to 0.2 wt %, the mean Di increased to 44.2 ± 13.0 nm, the mean
Do increased to 72.7 ± 13.0 nm and the mean Wt decreased to 9.2 ± 1.8 nm.
As shown in Fig. 4.4, when the CF- increased from 0.3 to 0.5 wt %, the percentage of
nanotubes with both Di < 40 nm and 40 < Di < 80 nm increased; similarly, the
percentage of nanotubes with both Do < 60 nm and 60 < Do < 120 nm increased; and
also the number of nanotubes with Wt < 10 nm increased. The percentage of nanotubes
with Wt ranging from 10 to 20 nm did not show an obvious change. The mean Di, Do
and Wt achieved for CP-Ti in the same condition showed similar trends. It can be seen
that the distribution of both Di and Do increased with an increase in CF-, while the
distribution of Wt decreased slightly when CF- increased. These nanotubes were open at
the top and closed at the bottom. After 2.75 h anodisation, almost all the nanotubes were
separated (Fig. 4.1(e)). When the time of anodisation was extended to 6.25 h, the top
features of the nanotubes were damaged with randomly dispersed 2 µm observable
holes (Fig. 4.1(f)).
The different distribution of Di, Do and Wt of the TiO2-ZrO2-ZrTiO4 nanotubes is related
to the different chemical and electrochemical properties of titanium and zirconium. The
standard electrode potential of Zr→Zr4+ is 1.553 V, less than that of Ti→Ti4+, 2.132 V
[236]. Thus zirconium competes with titanium during anodisation, and will oxidised
first under the same electrochemical conditions. Also the standard enthalpies of
formation of ZrO2 and TiO2 are different, that is, -264.199 and -228.360 gram calories
per mole, respectively [237]. Hence nanotubes of two different heights formed on the
surface of the Ti50Zr alloy. Such a pattern caused different roughness parameters in
comparison to the single phase of titania nanotubes formed on pure titanium. The
nanotubes formed on the α phase of Ti50Zr exhibit smaller Di, Do and Wt. This
morphological difference arises because the α phase contains more zirconium compared
81
to the β phase of the alloy, and the solubility equilibrium constant of (NH4)2ZrF6 is
higher than that of (NH4)2TiF6. In terms of CP-Ti, although there was still a nanoporous
layer on the surface of nanotubes with large pore size that caused some difficulties for
measuring the actual distribution of the size of the nanotubes, the inner and outer
diameter increased specifically when CF- increased.
Figure 4.5 shows the Di, Do and Wt of TiO2-ZrO2-ZrTiO4 nanotubes anodised in an
electrolyte with a constant CF- of 0.5 wt % for 2 h when the applied potential varied
from 5 to 20 V. On anodisation of Ti50Zr with the applied potential increased from 5 to
20 V at a constant CF- of 0.5 wt % for 2 h, the percentage of nanotubes with Di < 20 nm
decreased; but the number of nanotubes with 20 < Di < 40 and 40 < Di < 60 nm
increased; similarly, the number of nanotubes with Do <40 nm decreased but the
number of nanotubes with both with 40 < Do < 80 and 80 < Do < 120 nm increased;
however, the number of nanotubes with Wt both < 5 nm and 5 < Wt < 15 nm increased.
82
83
Fig. 4.5 Histograms of TiO2-ZrO2-ZrTiO4 nanotube parameters anodised in 0.5 wt % NH4F electrolyte for 2 h at different applied potentials: (a, b) inner
diameter (Di) as-formed and after annealing, (c, d) outer diameter (Do) as-formed and after annealing, (e, f) wall thickness (Wt) as-formed and after annealing,
respectively
84
There was a nanoporous layer on the top of the nanotubular surface when the titanium
alloy was anodised at the applied potential of 5 V. It is assumed that increasing the
applied potential provides a larger anodic current for nanotube formation and leads to an
increase in inner and outer nanotube diameter [58, 110]. It can be concluded that
increasing applied potential did not affect the Wt of nanotubes. The mean sizes of as-
formed TiO2-ZrO2-ZrTiO4 nanotubes on Ti50Zr were smaller than those of as-formed
single phase TiO2 nanotubes on pure titanium. Figs. 4.6 and 4.7 demonstrate the mean
nanotube size of Ti50Zr and CP-Ti with changing CF- and applied potential. As
abovementioned, it was impossible to measure Di, Do and Wt of fabricated TiO2
nanotubes on CP-Ti anodised at 5 V because a nanoporous layer covered the entire
nanotubular surface.
Fig. 4.6 Mean nanotube size of Ti50Zr and CP-Ti as a function of CF-: (a) Di for As-
formed, (b) Di for Annealed (c) Do for as-formed, (d) Do for Annealed, (e) Wt for as-
formed and (f) Wt for Annealed
15
25
35
45
55
65
0.2 0.3 0.4 0.5 0.6
Mea
n D
i(n
m)
CF- (wt%)
Ti50Zr-As FormedCP-Ti-As Formed
30405060708090
100110
0.2 0.3 0.4 0.5 0.6
Mea
n D
o (n
m)
CF- (wt%)
Ti50Zr-As FormedCP-Ti-As Formed
c)
5
7
9
11
13
15
17
19
0.2 0.3 0.4 0.5 0.6
Mea
n W
t (n
m)
CF- (wt%)
Ti50Zr-As FormedCP-Ti-As Formed
e)
15
25
35
45
55
65
0.2 0.3 0.4 0.5 0.6
Mea
n D
i(n
m)
CF- (wt%)
Ti50Zr-AnnealedCP-Ti-Annealed
b)
30405060708090
100110
0.2 0.3 0.4 0.5 0.6
Mea
n D
o (n
m)
CF- (wt%)
Ti50Zr-AnnealedCP-Ti-Annealed
5
7
9
11
13
15
17
19
0.2 0.3 0.4 0.5 0.6
Mea
n W
t (n
m)
CF- (wt%)
Ti50Zr-AnnealedCP-Ti-Annealed
f)
a)
d)
85
Fig. 4.7 Mean nanotube size of Ti50Zr and CP-Ti as a function of applied potential: (a)
Di for as-formed, (b) Di for Annealed (c) Do for as-formed, (d) Do for Annealed, (e) Wt
for as-formed and (f) Wt for Annealed
The effects of applied potential and anodisation time on nanotube length were evaluated
as shown in Fig. 4.8. The nanotubes length increased with an increase of time of
anodisation. The length also increased with an increase of applied potential. The effect
of anodisation time was more significant than applied potential on nanotube growth.
0
10
20
30
40
50
60
70
0 5 10 15 20 25
Mea
n D
i (nm
)
Applied Potential (V)
Ti50Zr-As FormedCP-Ti-As Formed
a)
0
20
40
60
80
100
0 5 10 15 20 25
Mea
n D
o(n
m)
Applied Potential (V)
Ti50Zr-As FormedCP-Ti-As Formed
c)
02468
1012141618
0 5 10 15 20 25
Mea
n W
t(n
m)
Applied Potential (V)
Ti50Zr-As FormedCP-Ti_As Formed
e)
0
10
20
30
40
50
60
70
0 5 10 15 20 25
Mea
n D
i(nm
)
Applied Potential (V)
Ti50Zr-AnnealedCP-Ti-Annealed
b)
0
20
40
60
80
100
120
0 5 10 15 20 25
Mea
n D
o(n
m)
Applied Potential (V)
Ti50Zr-AnnealedCP-Ti-Annealed
d)
02468
1012141618
0 5 10 15 20 25
Mea
n W
t(n
m)
Applied Potential (V)
Ti50Zr-AnnealedCP-Ti-Annealed
f)
86
Fig. 4.8 The effect on nanotube length by: a) anodisation time and b) applied potential
Another group of TiO2-ZrO2-ZrTiO4 nanotubes were fabricated when the applied
potential increased from 20 to 35 V in an electrolyte with a constant CF- of 0.5 wt % for
1 h. These applied potential was chosen to investigate the trend of nanotube sizes
distribution in higher applied potential. Fig. 4.9 shows SEM images of nanotubular
layer which fabricated at applied potential from 20 to 35V. Fig. 4.9 (a), (c), (e) and (g)
are before annealing and Fig. 4.9 (b), (d), (f) and (h) shows TiO2-ZrO2-ZrTiO4
nanotubes after annealing at 500 °C for 3 h. TiO2-ZrO2-ZrTiO4 nanotubes were the
result of anodisation and dissolution of oxide layer in presence of fluorine anions.
87
Fig. 4.9 SEM images of nanotubular layer fabricated at a) 20V - as-formed, b) 20V -
annealed, c) 25V - as-formed, d) 25 V - annealed, e) 30 V - as-formed, f) 30V -
annealed, g) 35 V - as-formed and h) 35 V - annealed
88
The growth rate of TiO2 was reported 2.5 nm/V [238] and it was 2.9 nm//V [239] for
ZrO2. Therefore an adherent layer of TiO2-ZrO2-ZrTiO4 grew on the surface which had
higher thickness where ZrO2 was exist because of its higher oxidation rate. These layer
consisted of two parts; substrate and oxide layer which is very compact and protective
and oxide layer and electrolyte which is porous. These pores are suitable places for
starting of tube growth because of trapping of F- and formation of dissoluble complex of
[TiF6]2- and [ZrF6]2- with different reaction rate. The difference between oxidation and
dissolution rate of titanium and zirconium caused to fabrication of different inner and
outer diameter and wall thickness of nanotubes as well as different nanotube length
which caused different roughness. Fig. 4.10 show distribution of Di, Do and Wt which
were as a function of applied potential.
89
90
Fig. 4.10 Histogram and distribution of Di, Do and Wt of as-formed and annealed TiO2-ZrO2-ZrTiO4 nanotubes fabricated at a,b) 20 V, c,d) 25 V, e,f) 30 V and
g,h) 35 V
91
The highest percentage in Di at 20 V was between 28 < Di < 52 nm. The greatest
frequency increased with respect to potentials of 25, 30 and 35 V to corresponding
values of 42 < Di < 76 nm, 41 < Di < 87 and 56 < Di < 108 nm. A similar increase in the
highest percentage of Do nanotubes at 20 V was from 60 < Do < 88 nm to 85 < Do < 119
nm, 91 < Do < 143 nm and 111 < Do < 161 nm, respectively when the applied potential
increased to 25, 30 and 35 V. The difference in Wt was between 2 to 6 nm by increasing
the applied potential starting from 9 < Wt < 17 nm at 20 V and reaching 17 < Wt < 31
nm at 35 V.
The distribution of Di, Do and Wt for the TiO2-ZrO2-ZrTiO4 nanotubes fabricated on
Ti50Zr alloy resulted from the different oxidation and soluble complex formation of Ti
and Zr with F- ions [240]. The frequency of smaller sized Di and Do decreased on
increasing the applied potential; thus enabling the formation of larger inner and outer
diameters due to the higher anodic current. The broad distribution of Di and Do at 30
and 35 V shows a better condition for oxidation of Ti and Zr because of high current
flow while the dissolution rate of (NH4)2[TiF6] and (NH4)2[ZrF6] remained constant.
In order to find a better statistical distribution of nanotube sizes, Weibull distribution
was studied. There was not obvious difference between normal distribution and Weibull
distribution as detailed in Table 4.1 and Fig. 4.11 for Di of as-formed and Fig. 4.12 for
Di, Do and Wt of annealed TiO2-ZrO2-ZrTiO4 as an example.
92
Table 4.1 The distribution of Di, Do and Wt according to normal and Weibull distribution methods
Applied
Potential (V)
As-fromed Annealed
Di (nm) Do (nm) Wt (nm) Di (nm) Do (nm) Wt (nm)
Normal Weibull Normal Weibull Normal Weibull Normal Weibull Normal Weibull Normal Weibull
20 24<Di<50 Di<41±3 53<Do<89 Do<78±4 9<Wt<15 Wt<13±4 28<Di<52 Di<41±3 60<Do<89 Do<80±5 1<Wt<14 Wt<14±3
25 5<Di<82 Di<70±4 92<Do<126 Do<116±7 12<Wt<25 Wt<21±3 42<Di<76 Di<65±3 85<Do<120 Do<110±6 1<Wt<8 Wt<19±4
30 56<Di<97 Di<84±4 96<Do<142 Do<128±6 13<Wt<15 Wt<21±3 41<Di<88 Di<72±3 91<Do<144 Do<128±5 1<Wt<23 Wt<24±3
35 61<Di<101 Di<89±4 106<Do<157 Do<142±6 16<Wt<27 Wt<24±4 56<Di<108 Di<91±3 111<Do<161 Do<146±6 1<Wt<25 Wt<26±4
93
Fig. 4.11 Histogram and fitted normal and Weibull distribution of Di of as-formed
nanotubular TiO2-ZrO2-ZrTiO4 fabricated at: a) 20 V, b) 25 V, c) 30 V, d) 35 V and e)
normal distribution of all four conditions
94
Fig. 4.12 Histogram and fitted normal and Weibull distribution of Di, Do and Wt of
annealed nanotubular TiO2-ZrO2-ZrTiO4 fabricated at a) 10 V, b) 15 V, c) 20 V
These distributions of nanotubes sizes affected the physical and chemical properties of
nanotubular surface.
The EDS results of the as-formed TiO2-ZrO2-ZrTiO4 nanotubes showed that there was a
layer of TiO2 and ZrO2. Fig. 4.13(b) shows the EDS result measured at the bottom of
the nanotubular layer. The oxygen and fluorine concentrations were higher than those
at the top of the nanotubular layer as shown in Fig. 4.13(a).
95
Fig. 4.13 EDS analysis for the (a) top and (b) bottom of TiO2-ZrO2-ZrTiO4 nanotubes
formed in 0.5 wt % NH4F at 20 V after 2 h
4.3.2 Surface roughness of the TiO2-ZrO2-ZrTiO4 nanotubular surface
As described in Chapter 3 there are three types of parameters for characterising surface
topography: (i) the amplitude parameters, (ii) the spacing parameters, and (iii) hybrid
parameters. These parameters can be determined with a 3D-Profilometer that is coupled
with SurfVision software. Amplitude parameters such as Mean Roughness (Sa) and Root
Mean Square (RMS) Roughness (Sq) that measure the vertical characteristics of surface
deviations have been used in this chapter and evaluated over the complete 3D surface.
These parameters represent an overall description of the surface textures that enable the
differentiation of peaks, valleys, and the spacing of surface textures [210]. The
roughness amplitude parameters of Sa and Sq increased with increasing both the CF-
from 0.3 to 0.5 wt % and the applied potential from 5 to 20 V (Figs. 4.14(b) and
4.14(d)). As it can be seen in Fig. 4.14(b), the difference between Sa of nanotubular
layer fabricated in CF- = 0.4 and CF
- = 0.5 wt % was higher than the difference between
Sa of nanotubular layer fabricated in CF- = 0.3 and CF
- = 0.4 wt % according to the
different chemical and electrochemical reaction rates of titanium and zirconium in the
alloy. The roughness parameters of nanotubular layer formed at the applied potential 20
to 35 V were correspondingly presented and discussed in Chapters 6 and 7.
96
Fig. 4.14 The mean roughness (Sa) and the mean water contact angle (W.C.A.) of the
nanotubular surfaces of Ti50Zr alloy as a function of: (a), (b) F- concentration and (c),
(d) applied potential, respectively
Fig. 4.15 The mean roughness (Sa) and the mean water contact angle (W.C.A.) of the
nanotubular surfaces of CP-Ti as a function of: (a), (b) F- concentration and (c), (d)
applied potential, respectively
-2
3
8
13
18
23
0.2 0.3 0.4 0.5 0.6
Mea
n W
.C.A
(θ°)
CF- (wt%)
Ti50Zr
W.C.A-As formed
W.C.A-Annealed
a)
0.2
0.7
1.2
1.7
2.2
2.7
3.2
3.7
0.2 0.3 0.4 0.5 0.6
Mea
n S a
(µm
)
CF- (wt%)
Ti50Zr
Sa-As formedSa-Annealed
b)
-5
0
5
10
15
20
25
0 5 10 15 20 25
Mea
n W
.C.A
(θ°)
Applied Potencial (V)
Ti50Zr
W.C.A-As formed
W.C.A-Annealed
c)
0.2
0.7
1.2
1.7
2.2
2.7
3.2
3.7
0 5 10 15 20 25
Mea
n S a
(µm
)
Applied Potential (V)
Ti50Zr
Sa-As formed
Sa-Annealed
d)
5
10
15
20
25
30
35
40
0.2 0.3 0.4 0.5 0.6
Mea
n W
.C.A
(θ°)
CF- (wt%)
CP-Ti
W.C.A-As FormedW.C.A-Annealed
a)
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0.2 0.3 0.4 0.5 0.6
Mea
n S a
(µm
)
CF- (wt%)
CP-Ti
Sa-As FormedSa-Annealed
b)
0
5
10
15
20
25
30
35
40
0 5 10 15 20 25
Mea
n W
.C.A
(θ°)
Applied Potencial (V)
CP-Ti
W.C.A-As FormedW.C.A-Annealed
c)
0
0.2
0.4
0.6
0.8
1
1.2
0 5 10 15 20 25
Mea
n S a
(µm
)
Applied Potential (V)
CP-Ti
Sa-As FormedSa-Annealed
d)
97
The roughness parameter of Sa of TiO2 nanotubes on the surface of CP-Ti increased
with an increase of CF- but revealed lower value than nanotubular layer of TiO2-ZrO2-
ZrTiO4.The Sa of TiO2 nanotubes also increased when the applied potential increased
from 10 to 20V with lower value than TiO2-ZrO2-ZrTiO4 nanotubes.
The nanotubular layers of TiO2-ZrO2-ZrTiO4 grown on the surface of Ti50Zr were
nanoscale in the horizontal dimension. However, the layers exhibited different heights
depending on the microstructure (i.e., different phases of α and β) of the surface due to
the different chemical and electrochemical reaction rates of titanium and zirconium in
the alloy. The surface roughness parameters were on the micrometer scale and displayed
higher values than on CP-Ti before anodisation. The degree of symmetry of the surface
heights about the mean plane is represented by Sskw, the Skewness of a 3D surface
texture, and it is the third central moment of the profile amplitude probability density
function [210]. The sign of Sskw indicates the dominant nature of topography. For
example, Sskw > 0 implies high peaks; whereas Sskw < 0 indicates valley structures such
as deep scratches [210]. Almost the entire as-formed TiO2-ZrO2-ZrTiO4 nanotubular
surface exhibited negative Sskw or a lesser value in comparison to that of the as-formed
TiO2 nanotubes (Table 4.2).
The nature of the height distribution is indicated by Sku, i.e., the kurtosis of the 3D
surface texture that is the fourth central moment of the profile amplitude probability
density function and describes the sharpness of distribution [210]. Sku is 3.0 when the
surface heights are normally distributed. For Sku < 3.0, the distribution is platykurtoic,
indicating a few high peaks and low valleys. For Sku > 3.0, the distribution is
leptokurtoic and indicates many high peaks and low valleys. Almost all of the as-
formed TiO2-ZrO2-ZrTiO4 nanotubular surfaces were platykurtoic and all of the as-
formed TiO2 nanotubular surfaces were leptokurtoic (Table 4.2); thereby reinforcing the
changing natures of the topography dependent of the forming mechanism.
The surface index (SI) is calculated using Surfvision software by the following equation:
SI = Sp/SL (4-10)
where SP is projected surface area that is the total exposed three-dimensional surface
area being analysed, including peaks and valleys and SL is the lateral surface area that is
measured in the lateral direction. The surface index SI increased significantly due to the
98
increase in the nanotube size and the surface roughness. As-formed TiO2-ZrO2-ZrTiO4
nanotubes displayed a higher SI than that of the as-formed TiO2 nanotubes. The surface
area and volume index, roughness amplitude parameters of Sskw and Sku of the
nanotubular surfaces of Ti50Zr anodised under various conditions are listed in Table
4.2.
99
Table 4.2 Surface area index SI and roughness amplitude parameters of Sskw and Sku of the nanotubular surfaces of Ti50Zr and CP-Ti anodised under
various conditions, Note: Each data point is an average of five measurements
Anodisation conditions
As-formed Annealed
Surface area index
SI Sskw Sku Surface area index Sskw Sku
0.3 - 20V Ti50Zr 4.96 ± 0.73 -1.79 ± 0.43 6.64 ± 2.18 4.34 ± 0.22 -0.23 ± 0.08 21.05 ± 2.81
CP-Ti 1.57 ± 0.07 1.34 ± 0.59 8.11 ± 6.08 1.47 ± 0.05 -0.85 ± 0.41 3.84 ± 1.35
0.4 - 20V Ti50Zr 8.05 ± 1.59 -2.98 ± 0.66 16.85 ± 6.36 5.58 ± 1.13 -2.52 ± 1.13 39.44 ± 3.85
CP-Ti 1.80 ± 0.06 0.67 ± 0.22 3.40 ± 0.27 2.12 ± 0.06 -0.18 ± 0.18 2.49 ± 0.19
0.5 - 20V Ti50Zr 13.18 ± 0.38 0.61 ± 0.02 2.11 ± 0.03 3.12 ± 0.06 -1.43 ± 0.12 16.60 ± 1.6
CP-Ti 1.71 ± 0.07 -0.67 ± 0.12 3.30 ± 0.12 1.49 ± 0.01 -0.80 ± 0.03 2.97 ± 0.07
0.5 - 15V Ti50Zr 10.99 ± 2.04 0.25 ± 0.37 2.08 ± 0.28 4.156 ± 0.3 -0.93 ± 0.09 17.03 ± 3.49
CP-Ti 2.20 ± 0.13 0.83 ± 0.25 3.38 ± 0.69 2.61 ± 0.13 0.27 ± 0.29 2.37 ± 0.33
0.5 - 10V Ti50Zr 10.69 ± 6.56 0.06 ± 0.34 2.03 ± 0.06 2.66 ± 0.30 -2.02 ± 0.18 12.41 ± 2.5
CP-Ti 1.31 ± 0.03 0.33 ± 0.19 3.68 ± 0.09 1.29 ± 0.02 0.01 ± 0.07 2.75 ± 0.03
0.5 - 5V Ti50Zr 5.29 ± 2.53 0.85 ± 0.45 3.86 ± 2.26 2.53 ± 0.07 -1.15 ± 0.15 10.30 ± 2.33
CP-Ti* - - - - - -
*All the surface was covered by a nanoporous layer
100
4.3.3 Hydrophilic properties of the nanotubular surfaces
Hydrophobicity and hydrophilicity are the terms that are defined by the wetting
properties of a surface by water [241]. Accepted definitions generally consider water
contact angle as follows: (i) superhydrophobic: static water contact angle > 150 º, (ii)
hydrophobic: water contact angle between 90 º and 150 º, (iii) hydrophilic: water
contact angle between 10 º and 90 º, and (iv) superhydrophilic: water contact angle less
than 10 º [242]. Parameters such as type of the material, surface roughness,
heterogeneities of the surface, chemical composition and presence of contamination; as
well as parameters related to the ambient conditions, can influence the wettability of a
surface [241]. The nanotubular surfaces of TiO2-ZrO2-ZrTiO4 nanotubes anodised in all
conditions were hydrophilic. The hydrophilicity of the anodised nanotubular surface did
not change obviously on increasing CF-, but it increased with an increase in the applied
potential. Figures 4.14 and 4.15 show the mean roughness (Sa) and mean water contact
angle (W.C.A.) of the as-formed TiO2-ZrO2-ZrTiO4 nanotubular surfaces, annealed
TiO2-ZrO2-ZrTiO4 nanotubular surfaces, and the as-formed TiO2 and annealed TiO2
naotubular surfaces as a function of the CF- and the applied potential.
The wettability of the nanotubular surface increased when roughness parameters
increased. It is noticeable that the wettability of nanotubular surface on Ti50Zr was
higher than that of the nanotubular surface on CP-Ti. The wettability properties of TiO2-
ZrO2-ZrTiO4 nanotubes formed at the applied potential of 20 to 35 V for 1 h were
presented and discussed in Chapters 6 and 7 relatively.
The surface energy of the nanotubular surface was calculated according to the Owens-
Wendt (Equ. 3-1). Almost all of the as-formed TiO2-ZrO2-ZrTiO4 nanotubular surfaces
exhibited a higher surface energy than those of the as-formed TiO2 nanotubular
surfaces. The surface energies of the as-formed TiO2-ZrO2-ZrTiO4 and annealed TiO2-
ZrO2-ZrTiO4 nanotubular surfaces; as well as those of the as-formed TiO2 and annealed
TiO2 nanotubular surfaces were measured and are detailed in Table 4.3.
101
Table 4.3 Calculated surface energy of TiO2-ZrO2-ZrTiO4 nanotubular surfaces fabricated on Ti50Zr and TiO2 nanotubular surfaces on CP-Ti, Note:
Each data point is an average of five measurements
Anodisation conditions
As-formed Annealed
Contact angle
()
sd
(mJm-2)
sp
(mJm-2)
s = sd + s
p
(mJm-2)
Contact angle
()
sd
(mJm-2)
sp
(mJm-2)
s = sd + s
p
(mJm-2)
0.3 - 20V Ti50Zr 20.50 ± 0.40 15.98 52.67 68.65 2.70 ± 0.17 14.93 58.71 73.64
CP-Ti 28.68 ± 10.08 22.45 41.87 64.32 22.50 ± 4.82 12.95 55.56 68.51
0.4 - 20V Ti50Zr 18.40 ± 1.04 17.88 51.39 69.27 2.00 ± 1.73 14.15 59.77 73.92
CP-Ti 22.97 ± 8.06 0.58 86.36 86.94 15.30 ± 3.75 13.57 57.84 71.41
0.5 - 20V Ti50Zr 16.03 ± 0.45 16.93 53.37 70.30 1.67 ± 1.44 15.44 58.10 73.54
CP-Ti 24.47 ± 0.45 14.39 52.65 67.04 14.97 ± 5.25 3.33 77.92 81.25
0.5 - 15V Ti50Zr 17.40 ± 0.53 17.89 51.77 69.66 2.40 ± 3.17 14.70 59.03 73.73
CP-Ti 22.00 ± 1.70 11.00 58.58 69.58 19.87 ± 2.70 7.52 65.58 73.10
0.5 - 10V Ti50Zr 19.33 ± 0.61 15.62 53.62 69.24 3.33 ± 0.29 14.42 59.33 73.75
CP-Ti 35.9 ± 11.46 17.44 42.19 59.63 26.67 ± 0.49 12.27 54.19 66.46
0.5 - 5V* Ti50Zr 20.11 ± 1.76 15.99 52.83 68.82 3.67 ± 0.58 14.98 58.57 73.55
CP-Ti* - - - - - - - -
*All the surface was covered by a nanoporous layer
102
4.3.4 Mechanical properties of the TiO2-ZrO2-ZrTiO4 nanotubes
The nanomechanical characteristics of the various as-formed and annealed TiO2-ZrO2-
ZrTiO4 nanotubes were measured using a nanoindentation to obtain the nano hardness
and Young’s modulus. Fig. 4.16 shows the curves of the load-unload forces versus the
nanoindentation depths of the as-formed and annealed TiO2-ZrO2-ZrTiO4 nanotubes
fabricated at different applied potentials.
Fig. 4.16 Loading-unloading forces versus the nanoindentation depths of a) as-formed
TiO2-ZrO2-ZrTiO4 and b) annealed TiO2-ZrO2-ZrTiO4 fabricated at the applied potential
20 to 35 V
It is worth noting that the nano Young’s modulus and hardness can be determined
instantaneously as a function of depth. Using the method described in Chapter 3, the
hardness, reduced elastic modulus and elastic modulus were calculated and presented in
Fig. 4.17.
0
0.002
0.004
0.006
0.008
0.01
0.012
-5E-10 0 5E-10 1E-09 1.5E-09 2E-09 2.5E-09
Load
(N)
Displacement (m)
As Formed nanotubes
20V
25V
35V
(a)
0
0.002
0.004
0.006
0.008
0.01
0.012
-5E-10 0 5E-10 1E-09 1.5E-09 2E-09
Load
(N)
Displacement (m)
Annealed Nanotubes
20V
25V
35V
(b)
103
Fig. 4.17 The nano mechanical properties of: a) hardness, b) reduced elastic Modulus
and c) elastic Modulus of TiO2-ZrO2-ZrTiO4 nanotubes fabricated at different applied
potential
The hardness of nanotubular layer of TiO2-ZrO2-ZrTiO4 was maximum 211.8 and 336.2
MPa before and after annealing which were lower than its respective bare metal which
104
revealed 4.38 GPa. The elastic modulus of the nanotubular layer of TiO2-ZrO2-ZrTiO4
was maximum 149.9 and 248.7 GPa before and after annealing which were higher than
its respective bare metal which revealed 102.9 GPa. It can be concluded that the
porosity of the nanotubes influenced the hardness and the ceramic properties of
nanotubular layer influenced elasticity of the bare metal. By increasing the applied
potential and consequently the nanotube size the hardness and elasticity of nanotubes
decreased.
4.3.5 Effect of annealing on the TiO2-ZrO2-ZrTiO4 nanotubes
TiO2 and ZrTiO4 nanotubes with an amorphous structure and ZrO2 with an
orthorhombic structure were fabricated via anodisation under the conditions of 0.5 wt %
NH4F, 20 V for 2 h. The amorphous TiO2 and ZrTiO4 nanotubes transformed into (i)
tetragonal anatase TiO2, (ii) a mixture of orthorhombic TiO2 and ZrO2 similar to
srilankite, and (iii) orthorhombic ZrTiO4 after annealing at 500 ºC for 3 h; as indicated
by the XRD patterns in Fig. 4.18. Varghese et al. [23] reported that amorphous TiO2
nanotubes transformed to anatase and rutile preferably along the length and the
curvatures of the nanotubes rather than wall thickness by annealing. A crystalline cubic
phase of ZrO2 accompanied by the phase of TiZrO4 or ZrTiO4 formed on the surface of
Ti50Zr alloy via anodisation [112, 132].
105
Fig. 4.18 XRD patterns of the nanotube samples fabricated on Ti50Zr (α and β phases)
via anodisation. (a) as-formed amorphous TiO2 and ZrTiO4 and orthorhombic ZrO2; (b)
annealed at 500 °C for 3 h tetragonal anatase, srilankite (a mixture of orthorhombic
TiO2 and ZrO2) and orthorhombic ZrTiO4
After annealing at 500 °C for 3 h some changes in inner and outer diameter and wall
thickness of the nanotubes formed on Ti50Zr via anodisation were observed which can
explain the formation of crystalline phases of TiO2-ZrO2-ZrTiO4. Annealing also caused
significant diversity in the roughness parameters and hydrophilic properties of the
nanotubular surfaces fabricated under different CF- and applied potential conditions.
After annealing the values of Sa and Sq decreased for the nanotubular surface fabricated
at each applied potential of 5 to 20 V. The surface parameters increased when the
applied potential increased from 5 to 10 V, but decreased when the applied potential
increased from 15 to 20 V. Almost all the annealed TiO2-ZrO2-ZrTiO4 nanotubular
surfaces exhibited more negative Sskw values compared to the annealed TiO2
nanotubular surfaces, although both exhibited cracks after annealing that would have
106
influenced the results. After annealing, the value of Sku of the TiO2-ZrO2-ZrTiO4
nanotubular surfaces increased but those for TiO2 nanotubular surfaces decreased.
Annealed TiO2-ZrO2-ZrTiO4 nanotubular surfaces exhibited a higher surface index than
the annealed TiO2 nanotubular surfaces.
The hydrophilicity of the TiO2-ZrO2-ZrTiO4 nanotubular surfaces on Ti50Zr increased
significantly after annealing for both anodising conditions; i.e., for increasing CF- and
applied potential. That is, the hydrophilicity increased with increases in both the Di and
Do. After annealing, almost all the TiO2-ZrO2-ZrTiO4 nanotubular surfaces anodised
under all the conditions exhibited superhydrophilic properties. In contrast, the TiO2
nanotubular surfaces fabricated on CP-Ti showed hydrophilic properties. In general, the
surface energy of the annealed TiO2-ZrO2-ZrTiO4 nanotubular surfaces was higher than
those of annealed TiO2 nanotubular surfaces and the as-formed TiO2-ZrO2-ZrTiO4
nanotubular surfaces. This result was reflected in lower water contact angles.
The hardness and elasticity of nanotubular layer increased by annealing although
increasing the porosity resulted to a decrease of hardness and elasticity.
X-ray photoelectron spectroscopy (XPS) was used to evaluate the chemistry of the
nanotubes. The XPS spectra of as-formed and annealed TiO2-ZrO2-ZrTiO4 are shown in
Fig. 4.19.
Fig. 4.19 XPS spectra for the as-formed and the annealed TiO2-ZrO2-ZrTiO4 nanotubes.
(a) O1s, (b) Ti 2p, and (c) Zr 3d
0
5000
10000
15000
20000
25000
30000
35000
40000
45000
522524526528530532534536538540
Inte
nsity
( a.
u.)
Binding energy (eV)
a)
Annealed
As Formed
O 1s
0
5000
10000
15000
20000
25000
30000
450452454456458460462464466468470472
Inte
nsity
(a.u
.)
Binding energy (eV)
b)
Annealed
As Formed
Ti 2p
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
174176178180182184186188190192194
Inte
nsity
(a.u
.)
Binding energy (eV)
Annealed
As Formed
c)Zr 3d
107
The nanotubular surface consisted of 48.55 at % O, 13.43 at % Ti, 8.64 at % Zr for the
as-formed nanotubes, and 51.23 at % O, 14.57 at % Ti and 8.35 at % Zr after annealing.
The remaining atomic percentages arose from carbon contamination and remnant
fluorine. The F residue after annealing was 0.15 at % in comparison to 3.23 at % for the
as-formed nanotubes. The Ti 2p3/2 and Zr 3d5/2 peaks have binding energies of 458.7 -
458.8 and 182.5 -182.7 eV, respectively, which represent the fully oxidised titanium ion
in its Ti4+ and zirconium ion in its Zr4+ state. A small splitting of the O 1s spectrum at
532.27 (as-formed) and 532.21 eV (annealed) showed the existence of ZrTiO4. This is
because of the different coordinations of oxide ions (Zr-O-Zr, Ti-O-Ti and Zr-O-Ti)
[243] as well as shifting to lower binding energy (Zr 3d5/2) due to phase transformation
during annealing.
4.4 Conclusions
Nanotubes formed on Ti50Zr alloy via anodisation have potential biomedical
applications such as promoting cell interaction; biomedical coatings; and drug delivery
and release of other payloads. The different phase components of Ti50Zr alloy led to the
formation of TiO2-ZrO2-ZrTiO4 nanotubes that are mostly microscopically needlelike
and separated from each other by an interspace. A distribution of nanotubes with
different sizes including inner diameter Di, outer diameter Do and wall thickness Wt,
resulting from the different zirconium contents in the α and β phases of the alloy, was
observed. The effects of the F- concentration (CF-) in the electrolyte, the applied
potential on the nanotube characteristics were revealed. The conclusions are as follows:
1. Because of the different chemical and electrochemical reaction rates of titanium
and zirconium with O2- and F- in the electrolytes, an increase in the CF- led to an
increase in the number of nanotubes with smaller Di, Do and Wt.
2. Increasing the applied potential has the consequence of increasing the anodic
current. The number of nanotubes with smaller Di and Do decreased; the number
of nanotubes with larger Di and Do increased; and the Wt of the nanotubes
fabricated at 5, 10, 15 and 20 V increased slightly. After annealing the mean of
Di and Do increased at higher voltage (e.g. 15 and 20 V) and Wt did not change
much.
108
3. The mean Di, Do and Wt of nanotubes fabricated on the surface of Ti50Zr
increased with an increase of applied potential from 20 to 35 V. Annealing
resulted in decreasing Di and Do; and Wt did not change.
4. The mean Di and Do of nanotubes fabricated on the surface of CP-Ti also
increased with an increase of applied potential from 20 to 35 V. But the Wt did
not show obvious change. Annealing resulted in decreasing Di and Do; and Wt
did not change.
5. According to little difference between normal distribution and Weibull
distribution it can be said that the normal distribution was an appropriate method
to represent the distribution of nanotube sizes of TiO2-ZrO2-ZrTiO4.
6. The Mean Roughness (Sa) and the Root Mean Square (RMS) Roughness (Sq) of
the nanotubular surfaces increased on increasing both the CF- and the applied
potential.
7. The hydrophilic properties of the anodised TiO2-ZrO2-ZrTiO4 nanotubular
surface of Ti50Zr increased with an increase in the CF- and the applied potential,
which is related to changes in the distribution of the nanotube size. The
hydrophilic properties of these surfaces increased when the roughness
parameters of the surface increased. The surface energy followed the same
increasing trend of hydrophilic properties. The TiO2-ZrO2-ZrTiO4 nanotubular
surfaces of Ti50Zr exhibited higher surface energy than those of TiO2
nanotubular surfaces on CP-Ti.
8. The surface area index SI increased with increasing CF- and applied potential,
which is linked to changes in the nanotube size distribution.
9. The hardness of bare metal decreased by the porosity of fabricated nanotubular
layer via anodisation while the elasticity increased by ceramic properties of the
oxide layer. Porosity affected the trend of hardness and elasticity of TiO2-ZrO2-
ZrTiO4 nanotubes.
10. Annealing changed the phase structure, surface roughness, hydrophilic property
and surface energy; i.e., (i) the amorphous nanotubes changed to a mixture of
anatase, orthorhombic TiO2 and ZrO2, and orthorhombic ZrTiO4, (ii) Sa and Sq
decreased, (iii) the hydrophilicity and surface energy increased, and (iv)
hardness and elasticity of nanotubular layer increased.
109
11. The mixed nanotubes of TiO2-ZrO2-ZrTiO4 on Ti50Zr alloy with a variety of
nanotube size distribution and other related properties such as hydrophilicity and
roughness are more promising for biomedical applications than the 100 % TiO2
nanotubes on CP-Ti.
110
Chapter 5 Nanotubes formed in non-aqueous electrolyte
Abstract
A high surface area of nanotubular layer of metal oxide fabricated via anodisation on
the surface of biocompatible titanium and titanium alloys offers high potential for
effective conditions for implant applications. In this chapter, the morphology, and the
physical and chemical characteristics of four groups of TiO2-ZrO2-ZrTiO4 nanotubes
that were fabricated via anodisation in a non-aqueous electrolyte were investigated.
Their nanoscale dimensional characteristics (i.e., inner diameter Di, outer diameter Do
and wall thicknesses Wt) were evaluated. The microstructure, morphology, hydrophilic
and mechanical properties of the nanotubes were characterised using scanning electron
microscopy (SEM), 3D-Profilometry, goniometer and nanoindentation. The applied
potential during anodisation and the water content of the non-aqueous electrolyte
influenced the shape and the profile of the titanium and zirconium oxide layer, resulting
in different nanoscale characteristics for the nanotubes. The oxidation and dissolution
rates were competing, resulting in different surface roughness parameters although these
rates are low in a non-aqueous electrolyte.
5.1 Introduction
Compared to other surface treatments such as hydrothermal [244], seeded growth [245]
and template-assisted deposition [246] for fabricating TiO2 nanotubes, anodisation is
simpler and reproducible. These nanotubes are more uniform and their characteristics
and morphology can be more easily controlled. It has been almost a decade when first
nanotubular layer of titanium dioxide fabricated via anodisation [10]. Several
generations of TiO2 nanotubes were formed by changing anodisation conditions [57, 61,
65, 66]. Apart from the applied potential which mostly influences the inner and outer
diameter and wall thickness of nanotubes [61] and the anodisation time which mostly
influences the length of nanotubes [62], the type of the electrolyte influences the
morphology of the nanotubular layer [73, 80, 247, 248]. In order to obtain more uniform
nanotubes with high aspect ratio several attempts has been undertaken through changing
the conditions of the electrolyte [67, 72, 79, 110, 132].
One attempt was to tailor the electrochemical conditions to enhance the aspect ratio of
the self-organised TiO2 nanotubes [67]. In this study, the pH profile within the tube was
111
constructed to modified the geometry of the TiO2 nanotubes using anodic oxidation in
buffered electrolyte (NH4F and (NH4)2SO4 or NaF and Na2SO4). Although F- is one of
the essential ions to fabricate nanotubes it reduces the length of the nanotubes up to
several hundred nanometer in an acidic pH with the same reason (dissolution of oxide
layer). The thickness of nanotubular layer was 2.5 µm with nm inner diameter in
buffered electrolyte [67]. Another attempt was tailoring the anodisation condition by
using non-aqueous electrolyte and smoother and longer nanotubes were fabricated in
this condition [79].
In a polar electrolyte such as ethylene glycol the oxide chemical dissolution is limited to
oxygen amount of its water (H2O) content [68]. As a result, fabricating a nanotubular
layer with the same thickness as a nanotubular layer formed in an aqueous electrolyte
takes a longer time. Furthermore, the viscosity, whether stirring [76] and the H2O
content of the electrolyte influence the formation of the nanotubes. It has been reported
that nanotubes with a diameter of 20 to 160 nm and a maximum height of 45 µm [68] or
higher (>150 µm) [77] were obtained at an applied potential ranging from 30 to 120 V
using an organic electrolyte such as glycerol [76, 80], ethylene glycol [68, 78] and
acetic acid [79]. When NH4F in glycerol was used as the electrolyte the TiO2 nanotubes
were smooth and not connected by outside walls [80]. The diffusion constant which is
the key factor of anodisation is relatively low in such an electrolyte. This has an
influence on the pH value of the pore tip and consequently on the current transient and
dissolution rate of ions.
This chapter describes how changes in the water content (H2O) of a non-aqueous
electrolyte and applied potential influenced the characteristics including the tube
diameter, roughness and hydrophilic and mechanical properties of a nanotubular TiO2-
ZrO2-ZrTiO4 layer. The nanotubular layers of TiO2-ZrO2-ZrTiO4 were fabricated with
mean Di (i) 29 ± 5, (ii) 35 ± 5, (iii) 34 ± 8 nm and (iv) 42 ± 10; mean Do (i) 51 ± 5 nm,
(ii) 61 ± 7, (iii) 59 ± 11 and (iv) 78 ± 10 nm; and mean Wt (i) 12 ± 1, (ii) 16 ± 1, (iii) 11
± 1 and (iv) 18 ± 1 nm, respectively. The roughness and water contact angle of different
nanoscale morphologies were measured and their surface energies were calculated.
112
5.2 Materials and methods
The substrates for fabrication nanotubular layer was Ti50Zr alloy (wt % hereafter) discs
with 8 mm diameter and 2 mm thickness cut from a cast ingot by electrical discharge
machining (EDM) and titanium foils (10×10×0.05 mm). Prior to anodisation the disc
samples were first polished to a 1 µm finish with diamond paste. All the samples were
cleaned by sonification successively in methanol, isopropanol, acetone and ethanol for
15 min each, afterwards rinsed with deionised water and dried in a stream of nitrogen
gas.
The Ti50Zr discs were exposed to an electrolyte through spot-welded to titanium foil
and contacted to the anode of an electrochemical cell. The plate shape platinum counter
electrode with a 1 cm2 surface area was placed 4 cm from the samples in a two-
electrode configuration using a DC power supply. The electrolyte was composed of
0.5 wt % of NH4F with the addition of 5 and 10 wt % of water (H2O) in ethylene glycol
(Sigma-Aldrich reagent grades). After the electrochemical treatment the sample surfaces
were rinsed with acetone for 5 min and followed by drying in air.
Two applied potentials, i.e., 20 and 30 V were used for each electrolyte composition to
create different sizes and morphologies of nanotubes. The anodised TiO2-ZrO2-ZrTiO4
nanotubes were post-treated by immersion in 0.5 M HCl at room temperature for 1 min,
and then rinsed with deionised water for 1 min to dissolve the remained upper
nanoporous layer. The samples were then heat-treated at 500 ˚C for 2 h at the heating
rate of 2 ˚C/min in a conventional muffle furnace (Nabertherm LT15/13/P330;
Nabertherm GmbH, Lilienthal, Germany) to crystallise the amorphous nanotubular
layer. The inner diameter, outer diameter and wall thickness of the anodised nanotubes
were denoted Di, Do and Wt, respectively. Each data point is an average of 100
measurements. on different positions for each of three samples (300 measurements).
A field-emission scanning electron microscope (FESEM, ZEISS SUPRA 40 VP)
equipped with energy dispersive X-ray spectroscopy (EDS) was used for
microstructural characterisation. A 3D-Profilometer (Bruker, Contour GT-K1; Bruker
Pte Ltd, Singapore) that was coupled with SurfVision software (Veeco Instruments Inc.;
Plainview, NY, USA) was used to evaluate the amplitude parameters of the anodised
TiO2-ZrO2-ZrTiO4 nanotubes. The vertical characteristics of surface deviations,
113
amplitude parameters over the complete 3D surface [210], were measured and presented
by Mean Roughness (Sa) and Root Mean Square (RMS) Roughness (Sq). The symmetry
and nature of the surface heights represented by Sskw, the skewness of a 3D surface
texture, and Sku, the kurtosis of the 3D surface texture, are the third and fourth central
moment of the profile amplitude probability, respectively [210]. The water contact
angle was measured using a goniometer (NRL C.A. Goniometer, Ramé-hart, Inc.;
Succasunna, NJ-USA) and the surface energy was calculated based on the Owens-
Wendt (OW) method [211], given by:
(1 + cos θ )γL=2(√γLd γS
d+√γLp γS
p) (3 − 1)
where L indicates the liquid surface tension (water = 72.8 mJm-2, glycerol = 63.4 mJm-
2); Ld and S
d are the liquid and solid dispersive component (water = 21.8 mJm-2,
glycerol = 37.0 mJm-2). The variables Lp and S
p are indicative of the liquid and solid
polar components (water = 51.0 mJm-2, glycerol = 26.4 mJm-2) and S is the sum of Sd
and Sp [212].
5.3 Results and discussion
5.3.1 Formation and characterisation of TiO2-ZrO2-ZrTiO4 nanotubes
Titania-zirconia-zirconium titanate (TiO2-ZrO2-ZrTiO4) nanotubes were fabricated on
the surface of Ti50Zr alloy via anodisation in a non-aqueous electrolyte (ethylene
glycol) which was consisted of 5 and 10 wt % of water (H2O) and 0.5 wt % NH4F. The
potential of 20 and 30 V were applied to samples for each composition of electrolyte for
90 min. The effect of the water content in the electrolyte on nanotubes size,
morphology, hydrophilic and mechanical properties were investigated. Fig. 5.1 shows
the SEM images of the top view of the fabricated nanotubes on the Ti50Zr alloy in non-
aqueous electrolyte.
114
Fig. 5.1 SEM images of top view of nanotubes fabricated at: a) 5 wt % H2O at 20 V, b)
10 wt % H2O at 20V, c) 5 wt % H2O at 30 V and d) 10 wt % H2O at 30V in ethylene
glycol
There were a distribution of size of Di, Do and Wt of nanotubes (Fig. 5.2).
115
116
Fig. 5.2 Histograms of TiO2-ZrO2-ZrTiO4 nanotube parameters anodised in 0.5 wt % NH4F and ethylene glycol consisting of 5 and 10 wt % H2O for 90 min at
20 and 30V: (a, d) inner diameter (Di) of as-formed and annealed nanotubes, (b, e) outer diameter (Do) of as-formed and annealed nanotubes, (c, f) wall
thickness (Wt) of as-formed and annealed, respectively
117
According to normal distribution the highest percentage of nanotube were evaluated and
presented in Table 5.1. To find better statistical distribution of nanotube sizes, Weibull
distribution was also studied. Fig. 5.3 shows that the difference between normal
distribution and Weibull distribution. It can be seen that the Weibull distribution can be
negligible and normal distribution can be an acceptable method to present.
118
Table 5.1 The distribution of Di, Do and Wt according to normal and Weibull distribution methods
Water Content
(wt%) and
Applied Potential
(V)
As-fromed Annealed
Di (nm) Do (nm) Wt (nm) Di (nm) Do (nm) Wt (nm)
Normal Weibull Normal Weibull Normal Weibull Normal Weibull Normal Weibull Normal Weibull
5- 20 24<Di<34 Di<31±6 46<Do<56 Do<54±11 11<Wt<12 Wt<12±24 16<Di<27 Di<24±4 42<Do<55 Do<51±8 12<Wt<15 Wt<14±12
10 - 20 29<Di<40 Di<37±6 53<Do<68 Do<64±8 15<Wt<17 Wt<16±31 25<Di<37 Di<33±6 50<Do<62 Do<59±10 16<Wt<20 Wt<19±12
5 - 30 25<Di<42 Di<37±4 48<Do<69 Do<63±6 10<Wt<12 Wt<11±8 22<Di<33 Di<29±5 50<Do<65 Do<61±7 14<Wt<18 Wt<17±9
10 - 30 33<Di<52 Di<46±4 68<Do<87 Do<82±8 17<Wt<19 Wt<18±23 31<Di<43 Di<40±7 68<Do<87 Do<82±8 18<Wt<20 Wt<19±16
119
120
Fig. 5.3 Histogram and fitted normal and Weibull distribution of Di, Do and Wt of as-formed and annealed nanotubular TiO2-ZrO2-ZrTiO4 fabricated at a,e) 5 wt
% H2O, 20 V, b, f) 10 wt % H2O, 20 V c, g) 5 wt % H2O, 30 V, d,h) 10 wt % H2O, 30 V
121
The walls of neighbouring nanotubes were not separated, although increasing water
content helped showing the needle-like microstructure of the underneath alloy.
Figure 5.4 presents the mean size of Di, Do and Wt of the fabricated nanotubues before
(Fig. 5.4(a)) and after annealing (Fig. 5.4(b)). The Di, Do and Wt of fabricated TiO2-
ZrO2-ZrTiO4 nanotubues increased when the water content of the organic electrolyte
increased from 5 to 10 wt % at the same applied potential. Increasing the applied
potential from 20 V to 30 V resulted in an increase in Di and Do but Wt did not increase
for each composition of electrolyte. After annealing Di of nanotubes decreased while the
change of Do was negligible. The thickness of the tube walls increased after annealing.
The Di, Do and Wt of the as-formed TiO2 nanotubes shown in Fig. 5.4(c) increased when
the water content of the organic electrolyte increased from 5 to 10 wt % at the same
applied potential. It can be observed that increasing the applied potential from 20 to 30
V resulted in an increase in Di, Do and Wt of the as-formed TiO2 for each composition of
electrolyte. The mean sized of nanotubes fabricated on the surface of CP-Ti were larger
than nanotubes fabricated on the surface of Ti50Zr. After annealing (Fig. 5.4(b) and (d))
Di of TiO2-ZrO2-ZrTiO4 and TiO2 nanotubes decreased, Do remained almost constant
and Wt increased. The crystallisation of TiO2-ZrO2-ZrTiO4 and TiO2 nanotubes occurred
at walls and grew towards the centre of nanotubes. The nanotubular surface fabricated
under non-aqueous electrolyte conditions showed a brittle layer after annealing which
was not suitable for further investigation.
122
Fig. 5.4 Mean nanotube sizes of Di, Do and Wt: a,c) as-formed and b,d) annealed TiO2-
ZrO2-ZrTiO4 and TiO2 nanotubes fabricated in two different water contents and two
different applied potentials in ethylene glycol
Once a constant potential, i.e. 20 V was applied to the Ti50Zr disk during anodisation, it
was oxidised accordingly and a layer of TiO2-ZrO2-ZrTiO4 formed at the interface of
the substrate and electrolyte. It was described before [249] that the surface of this layer
randomly broke due to the formation of [TiF6]2- and [ZrF6]2- which are soluble
complexes in water. These breakdown sites were the places where the nanotubes and the
interspace between nanotubes started to grow as non-uniform nanoporous layer. The
Ti4+ and Zr4+ which were the production of oxidation attended at interface of substrate
and oxide layer at the bottom of these pores in every step of the process. The F- ions
which have higher diffusion rate than O2- reached at the bottom of the pores and formed
the soluble complex and took their places to the new F- ions and left the bottom of the
pores to the top because of diffusion flux. These oxidation and dissolution which were
equilibrated continued up to the end of anodisation time. In all these processes there was
always competition between Ti and Zr to be oxidised and making soluble complex. It
0
20
40
60
80
100
120
5%-20 10%-20 5%-30 10%-30
Nan
otub
e si
ze (n
m)
Water content (wt%)and applied potential (V)
Ti50Zr
Wt-As Formed
Di-As Formed
Do-As Formed
a)
0
20
40
60
80
100
120
5%-20 10%-20 5%-30 10%-30
Nan
otub
e si
ze (n
m)
Water content (wt%)and applied potential (V)
Ti50Zr
Wt-Annealed
Di-Annealed
Do-Annealed
b)
0
20
40
60
80
100
120
5%-20 10%-20 5%-30 10%-30
Nan
otub
e si
ze (n
m)
Water content (wt%)and applied potential (V)
CP-Ti
Wt-As Formed
Di-As Formed
Do-As Formed
c)
0
20
40
60
80
100
120
5%-20 10%-20 5%-30 10%-30N
anot
ube
size
(nm
)Water content (wt%)and applied potential (V)
CP-Ti
Wt-Annealed
Di-Annealed
Do-Annealed
d)
123
was Zr that was oxidised and formed the complex first because the standard electrode
potential of Zr→Zr4+ and Ti→Ti4+ are 1.553 V and 2.132 V [236] and the standard
enthalpies of formation of ZrO2 and TiO2 are -264.199 and -228.360 gram calories per
mole, respectively [237]. In aqueous electrolyte this completion resulted in obvious
different nanotube size and height on the α and β phases of the Ti50Zr alloy. But in non-
aqueous electrolyte the rate of these reactions are low and depend on the water content
of the electrolyte. Although ethylene glycol is a polar solvent, its polarity (79) is less
than water (100) and also has lower dielectric constant (37.7) than water (79.7).
Therefore, both oxidation rate and dissolution rate are limited and Ti and Zr have the
same chance to be oxidised to Ti4+ and Zr4+ and dissolve as [TiF6]2- and [ZrF6]2-. In this
case although there was difference in nanotube size and height but they were not largely
distributed and the nanotubues were not separated.
A layer of TiO2 and ZrO2 was observed at the bottom of the as-formed nanotubular
layer of TiO2-ZrO2-ZrTiO4 which has shown in Fig. 5.5 according to the EDS results.
The oxygen and fluorine weight percent of this layer were lower than those at the top of
the nanotubular layer.
Fig. 5.5 EDS analysis for: (a) top and (b) bottom of TiO2-ZrO2-ZrTiO4 nanotubes
formed in 0.5 wt % NH4F, 5 wt % H2O in ethylene glycol at 20 V after 90 min
Figure 5.6 shows the effect of applied potential and anodisation time on nanotube
length. The nanotubes length increased with an increase of anodisation time up to 20 h.
The length also increased when the applied potential increased from 20 to 35 V in the
same electrolyte. The time of anodisation exhibited more significant effect on nanotube
124
growth than applied potential. The length of nanotubes also increased when the water
content of the electrolyte increased from 5 to 10 wt % at the same applied potential and
anodisation time.
Fig. 5. 6 The effect on TiO2-ZrO2-ZrTiO4 nanotubes length formed in organic
electrolyte by: a) anodisation time and b) applied potential
Fabricated TiO2 and ZrTiO4 nanotubes revealed amorphous structure and ZrO2 revealed
orthorhombic structure after anodising in non-aqueous electrolyte. After annealing at
500 °C for 2 h, the amorphous TiO2 and ZrTiO4 nanotubes transformed into a mixture
of tetragonal anatase and rutile TiO2, a mixture of orthorhombic TiO2 and ZrO2 similar
to srilankite and orthorhombic ZrTiO4. Fig. 5.7 presents the XRD patterns of as-formed
and annealed nanotubular layer fabricated in non-aqueous electrolyte.
125
Fig. 5.7 XRD patterns of the nanotube samples fabricated on Ti50Zr (α and β phases)
via anodisation in non-aqueous electrolyte. (a) as-formed amorphous TiO2 and ZrTiO4
and orthorhombic ZrO2; (b) annealed at 500 °C for 2 h tetragonal anatase, rutile,
srilankite (a mixture of orthorhombic TiO2 and ZrO2) and orthorhombic ZrTiO4
5.3.2 Surface roughness and hydrophilic property of the TiO2-ZrO2-ZrTiO4
nanotubular surface
Figure 5.8 shows roughness parameters of Sa and Sq measured for the as-formed
nanotubular surfaces of TiO2-ZrO2-ZrTiO4 (Fig. 5.8(a)) and TiO2 (Fig. 5.8(c)). These
amplitude parameters (Sa and Sq) of TiO2-ZrO2-ZrTiO4 nanotubes increased with an
increase of water content for each applied potential. Increasing the applied potential for
each composition of electrolyte resulted in increasing of Sa and Sq. The Sa and Sq did
126
not change for TiO2 nanotubes when the water content of the electrolyte increased. The
amplitude parameters of Sa and Sq increased when the applied potential increased for
each composition of electrolyte. The nanotubular layer of TiO2 revealed lower
roughness than the nanotubular layer of TiO2-ZrO2-ZrTiO4.
Fig. 5.8 a), c) The mean roughness (Sa), (Sq) and b), d) the mean water contact angle
(W.C.A.) of the nanotubular surfaces of TiO2-ZrO2-ZrTiO4 and TiO2 respectively, Note:
Each data point is an average of five measurements
The TiO2-ZrO2-ZrTiO4 nanotubes fabricated in the electrolyte with 5 wt % H2O at 20 V
with mean Di = 29 nm exhibited Sskw ≤ 0 and Sku ≥ 3, indicating uniform symmetry
normally distributed of surface heights, whilst the other 3 nanotubular surfaces
exhibited Sskw < 0 and Sku > 3, indicating leptokurtoic distribution with many high peaks
and valleys. The TiO2 nanotubes fabricated in the electrolyte with 10 wt % H2O at 30 V
with mean Di = 52 nm exhibited Sskw ≈ 0 and Sku ≈ 3, indicating uniform symmetry
normally distributed of surface heights, while the other 3 nanotubular surfaces exhibited
Sskw > 0 and Sku > 3, indicating leptokurtoic distribution with many high peaks and
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
29 (5 wt%H2O-20V) 35 (10 wt% H2O-20V) 34 (5 wt%H2O-30V) 42 (10 wt%H2O-30V)
Mea
n Sa
(µm
)
Mean Di (nm)
Roughness-Ti50Zr Sa-As Formed
Sq-As Formeda)
0
2
4
6
8
10
12
14
16
29 (5 wt%H2O-20V) 35 (10 wt% H2O-20V) 34 (5 wt%H2O-30V) 42 (10 wt%H2O-30V)
Mea
n W
.C.A
. (θ°
)Mean Di (nm)
Wettability-Ti50Zr W.C.A-As Formedb)
0
0.1
0.2
0.3
0.4
0.5
0.6
32 (5 wt%H2O-20V) 37 (10 wt% H2O-20V) 42 (5 wt%H2O-30V) 52 (10 wt%H2O-30V)
Mea
n Sa
(µm
)
Mean Di (nm)
Roughness-CP-Ti Sa-As Formed
Sq-As Formedc)
02468
1012141618
32 (5 wt%H2O-20V) 37 (10 wt% H2O-20V) 42 (5 wt%H2O-30V) 52 (10 wt%H2O-30V)
Mea
n W
.C.A
. (θ°
)
Mean Di (nm)
Wettability-CP-Ti W.C.A-As Formedd)
127
valleys. Upper layer of nanotubes were cracked after annealing at 500 °C for 2 h and it
was impossible to measure the roughness parameter and water contact angle.
The surface index (SI) is calculated using Surfvision software using the following
equation:
SI = Sp/SL (5-1)
The total exposed three-dimensional surface area being analysed (SP) is the projected
surface area, including peaks and valleys and SL is the lateral surface area that is
measured in the lateral direction.
Table 5.2 listed the surface index with other amplitude parameters of the TiO2-ZrO2-
ZrTiO4 and TiO2 nanotubes fabricated using four different water contents and applied
potentials. The nanotubular layer of TiO2-ZrO2-ZrTiO4 revealed higher surface area
index (SI) than nanotubular layer of TiO2 because of different height of TiO2 and ZrO2
nanotubes resulted of their different oxidation and dissolution rates.
128
Table 5.2 Surface area index SI and roughness amplitude parameters of Sskw and Sku of
the nanotubular surfaces of Ti50Zr and CP-Ti anodised under various conditions, Note:
Each data point is an average of five measurements
Water Content (wt%) and Applied
Potential (V)
As-formed
Surface area index SI Sskw Sku
5 - 20 Ti50Zr 2.42 ± 0.95 -0.74 ± -0.57 4.56 ± 2.46
CP-Ti 1.14 ± 0.01 0.44 ± 0.10 5.60 ± 0.28
10 - 20 Ti50Zr 4.32 ± 0.11 -4.69 ± 0.18 31.94 ± 1.88
CP-Ti 1.37 ± 0.04 1.36 ± 0.16 7.36 ± 1.11
5 - 30 Ti50Zr 4.12 ± 0.96 -5.71 ± 2.15 49.38 ± 35.30
CP-Ti 1.82 ± 0.09 0.80 ± 0.30 3.58 ± 0.67
10 - 30 Ti50Zr 4.52 ± 0.92 -4.67 ± 0.53 31.75 ± 7.34
CP-Ti 1.76 ± 0.03 0.16 ± 0.09 2.95 ± 0.19
129
According to the generally accepted definitions for wetting properties [242], both TiO2-
ZrO2-ZrTiO4 and TiO2 nanotubular layer were hydrophilic (10° < θW < 90°).
Hydrophilic property of TiO2-ZrO2-ZrTiO4 and TiO2 nanotubes increased with an
increase in the nanotube sizes (Fig. 5.8 (b) and (d)). Hydrophilic property of TiO2-
ZrO2-ZrTiO4 and TiO2 nanotubes increased with an increase of roughness parameters
which were the result of increasing of water content and applied potential. The
difference between water contact angle on the nanotubular surfaces of TiO2-ZrO2-
ZrTiO4 and TiO2 was not vary although nanotubular layer of TiO2 exhibited lower
roughness parameters than TiO2-ZrO2-ZrTiO4 due to their larger nanotube sizes.
The surface energies of the as-formed TiO2-ZrO2-ZrTiO4 and TiO2 nanotubular surfaces
calculated using the Owens-Wendt (OW) method [211] are listed in Table 5.3. All of
the TiO2-ZrO2-ZrTiO4 nanotubular surfaces exhibited a higher surface energy than those
of the as-formed TiO2 nanotubular surfaces. The different heights of TiO2-ZrO2-ZrTiO4
nanotubular surface depending on the microstructure (i.e., different phases of α and β)
of Ti50Zr alloy resulted in the roughness parameters of 500 > Sa and Sq > 1500 nm
which was due to the different chemical and electrochemical reaction rates of titanium
and zirconium in the alloy although their rates were lower than what was observed for
nanotubular layer fabricated in aqueous electrolyte [240].
130
Table 5.3 Calculated surface energy of as-formed and annealed dimpled surface TiO2-
ZrO2-ZrTiO4 nanotubular surfaces fabricated on Ti50Zr, Note: Each data point is an
average of five measurements
Anodisation conditions
As-formed
Contact angle
()
sd
(mJm-2)
sp
(mJm-2)
s = sd + s
p
(mJm-2)
5 wt%H2O-20V Ti50Zr 14.3 ± 0.6 13.27 58.59 71.86
CP-Ti 16.5 ± 0.4 12.93 58.27 71.20
10 wt%H2O-20V Ti50Zr 10.0 ± 0.4 12.90 60.34 73.24
CP-Ti 11.0 ± 0.7 12.40 60.81 73.21
5 wt% H2O-30V Ti50Zr 11.0 ± 0.2 10.30 64.04 74.34
CP-Ti 12.5 ± 0.3 15.04 56.81 71.86
10 wt%H2O-30V Ti50Zr 9.0 ± 0.1 13.37 59.91 73.28
CP-Ti 10.1 ± 0.3 13.15 59.97 73.12
131
5.3.3 Mechanical properties
To evaluate the nanomechanical characteristics of the various as-formed and annealed
TiO2-ZrO2-ZrTiO4 nanotubes a nanoindentation was used to measure the nano hardness
and Young’s modulus. Fig. 5.9 shows the curves of the load-unload forces versus the
nanoindentation depths of the as-formed and annealed TiO2-ZrO2-ZrTiO4 nanotubes
fabricated at different water content in an organic electrolyte (ethylene glycol) and
applied potentials.
Fig. 5.9 Loading-unloading forces versus the nanoindentation depths of a) as-formed
TiO2-ZrO2-ZrTiO4 and b) annealed TiO2-ZrO2-ZrTiO4 fabricated at 5 and 10 wt % H2O
and the applied potential 20 and 30 V
The nano Young’s modulus and hardness can be determined instantaneously as a
function of depth. The hardness, reduced elastic modulus and elastic modulus were
calculated using the method described in Chapter 3 and presented in Fig. 5.10.
0
0.002
0.004
0.006
0.008
0.01
0.012
0 5E-10 1E-09 1.5E-09 2E-09
Load
(N)
Displacement (m)
As Formed nanotubes
5%-20V
5%-30V
10%-30V
(a
0
0.002
0.004
0.006
0.008
0.01
0.012
-5E-10 0 5E-10 1E-09 1.5E-09
Load
(N)
Displacement (m)
Annealed Nanotubes
5%-20V
5%-30V
10%-30V
(b)
132
Fig. 5.10 The nano mechanical properties of of TiO2-ZrO2-ZrTiO4 nanotubes fabricated
at different water content and applied potential: a) hardness, b) reduced elastic Modulus
and c) elastic Modulus
133
The hardness of nanotubular layer of TiO2-ZrO2-ZrTiO4 was maximum 330.2 and
4019.0 MPa before and after annealing which was lower than its respective bare metal
which revealed 4.38 GPa. The elastic modulus of the nanotubular layer of TiO2-ZrO2-
ZrTiO4 was maximum 185.9 GPa before annealing which was higher than its respective
bare metal which revealed 102.86 GPa; and 84.4 GPa after annealing which was lower
than its respective bare metal. It can be concluded that the porosity of the nanotubes
influenced the hardness and the ceramic properties of nanotubular layer influenced the
elasticity of the bare metal. By increasing the applied potential and consequently the
nanotube size the hardness and elasticity of nanotubes decreased. Increasing the water
content of the organic electrolyte (ethylene glycol) resulted in increasing of hardness
and elasticity of the nanotubular layer.
5.4 Conclusions
The effects of water content of the non-aqueous electrolyte and the applied potential on
the nanospacing, morphology and physical, chemical and mechanical properties of as-
formed and annealed TiO2-ZrO2-ZrTiO4 nanotubular surfaces were investigated. The
key conclusions are as follows:
1. The mean inner diameter (Di) and outer diameter (Do) of TiO2-ZrO2-ZrTiO4
nanotubes fabricated in the electrolytes of ethylene glycol containing 5 and 10
wt % H2O increased with an increase in the applied potential of 20 to 30 V from
29 ± 5 and 51 ± 5 nm, and 35 ± 5 and 61 ± 7 nm and to 34 ± 8 and 59 ± 11 nm,
and 42 ± 10 and 78 ± 10 nm; but the Wt did not show obvious change. Increasing
water content resulted in an increase in Di, Do and Wt.
2. The Di and Do of TiO2 nanotubes fabricated in the electrolytes of ethylene glycol
containing 5 and 10 wt % H2O increased with an increase in the applied
potential of 20 to 30 V but the Wt did not show obvious change. Increasing water
content resulted in an increase in Di, Do and Wt. TiO2 nanotubes revealed higher
Di, Do and Wt than TiO2-ZrO2-ZrTiO4 nanotubes.
3. After annealing at 500 °C for 2 h the mean Di of both TiO2-ZrO2-ZrTiO4 and
TiO2 decreased, the mean Wt increased but the mean Do did not changed. It was
found that the top layer of nanotubes was brittle and cracked, and it was
removed for further investigation.
134
4. There was a little difference between normal distribution and Weibull
distribution of nanotube sizes. The normal distribution proves an appropriate
method in evaluating the distribution of nanotube sizes.
5. Increasing water content and applied potential resulted in an increase in Mean
Roughness (Sa), Root Mean Square (RMS) Roughness (Sq) and hydrophilic
properties and a decrease in the surface energy of TiO2-ZrO2-ZrTiO4 nanotubes.
TiO2 nanotubes exhibited lower Sa, Sq and almost the same hydrophilic
properties and surface energy in comparison to TiO2-ZrO2-ZrTiO4 nanotubes.
6. The surface area index SI of TiO2-ZrO2-ZrTiO4 nanotubes increased with
increasing water content and applied potential, which is linked to changes in the
nanotube size distribution and surface roughness. TiO2 nanotubes revealed lower
surface area index SI than that of TiO2-ZrO2-ZrTiO4 nanotubes.
7. Annealing changed the phase structure, Di, Do and Wt of nanotubes. The
amorphous nanotubes changed to a mixture of tetragonal anatase and rutile,
orthorhombic TiO2 and ZrO2, and orthorhombic ZrTiO4. Hardness and elasticity
of nanotubular layer increased after annealing.
8. The hardness of the nanotubular layer is lower than that of the bare metal due to
the porosity of the fabricated nanotubues via anodisation; and the elasticity of
the nanotubes is lower than that of the bare metal due to the ceramic properties
of the hard and brittle oxide layer. The hardness and elasticity of TiO2-ZrO2-
ZrTiO4 nanotubes were also influenced by the porosity of the nanotubular layer.
9. The mixed nanotubes of TiO2-ZrO2-ZrTiO4 on Ti50Zr alloy fabricated in the
non-aqueous electrolyte with a variety of nanotube size distribution and other
related properties such as hydrophilicity and roughness are more promising for
biomedical applications than the 100 % TiO2 nanotubes on CP-Ti.
135
Chapter 6 Bioactivity of nanotubes fabricated in aqueous and non-
aqueous electrolytes
Abstract
The chemical behaviour and topography of implant surfaces can be enhanced by
designing the biomaterial surface properties with regard to their influences on the
bioactivity and cell behaviour. Fabrication of metal oxide nanotubes on metallic
biomaterials, especially titanium alloys such as Ti50Zr via anodisation, could enhance
the growth of hydroxyapatite (HA), which is a necessary step to provide a bioactive
surface for the attachment of bone cells to attach. In this chapter, eight groups of TiO2-
ZrO2-ZrTiO4 and TiO2 nanotubes with a diversity of nanoscale dimensional
characteristics (i.e., inner diameter Di, outer diameter Do and wall thicknesses Wt) were
fabricated via anodisation in aqueous and non-aqueous electrolytes. The as-formed and
annealed nanotubes were characterised using scanning electron microscopy (SEM), X-
ray diffraction (XRD) and 3D-Profilometry. The different nanoscale characteristics of
the nanotubes and the different roughness parameters of the nanotubular surfaces were
obtained by adjusting the applied potential during anodisation, which influenced the
oxidation rates of titanium and zirconium. The bioactivity of a nanotubular layers of
TiO2-ZrO2-ZrTiO4 and TiO2 were assessed by immersion in a modified simulated body
fluid (m-SBF). The results revealed that the nanotubes were more effective in inducing
hydroxyapatite (HA) growth in contrast to the just anodised surface and the bare metal.
The Di, roughness, hydrophilic property and crystalline phase of the nanotubes
demonstrated an effective influence on inducing HA.
6.1 Introduction
The mineralised collagen fibril is the basic building block which is common in a family
of materials with different structural motif known as bone [29]. The plate shaped
carbonated apatite or dahllite (Ca5∙(PO4∙CO3)3(OH)) stands in a framework of these
collagen fibrils [29, 250]. Although dahllite displays a hexagonal crystal symmetry, the
crystals of bone are plate-shaped; probably because plate shaped octacalcium phosphate
is a transition phase. The average length and width of this crystal are 50 × 25 nm with a
uniform thickness of 1.5 - 4 nm and it has a highly ordered surface [29]. The organised
structure of bone which is hierarchically complex has two major functions: a
136
mechanical function and a metabolic function. The metabolic function of bone is a
reservoir for calcium and phosphate [250]. For example remodelling process in hard
tissue, which is guided by bone cells and proteins can also occur on surfaces of
bioactive implants by the biomineralisation of apatite [251]. Bioactive implant materials
can be mineralised into apatite by means of body fluid (in vivo) or simulated body fluid
(in vitro) when there is an interface to integrate spontaneously with living tissue [251].
Mineralisation of crystalline apatite occurs due to the electrostatic interaction of the
functional groups of the bioactive surface with the calcium and phosphate ions in the
fluid through a transition phase of amorphous calcium phosphate [251].
Recent research has focused on binary and tertiary alloys of titanium with
biocompatible alloying metals such as tantalum, niobium and zirconium for application
as implants [27, 46, 93-95, 230, 252, 253]. Ti50Zr, with a fine martensitic structure is
one of these new biocompatible alloys And, in comparison to CP-Ti, it exhibits 2.5
times higher hardness and ~ 2.3 - 3 times higher tensile strength [230]. Anodisation is a
surface modification method that can improve the biocompatibility and bioactivity of
metal implants. A layer of metal oxide nanotubes forms when the electrochemical
oxidation is undertaken in an electrolyte (aqueous or non-aqueous) consisting of
fluorine anions [11]. Titanium dioxide, fabricated both in aqueous and non-aqueous
electrolyte, exhibited bioactivity [11, 16, 234]. Hydroxyapatite can be deposited or
mineralised on the surface of metal implant using physical and chemical techniques,
such as plasma spray [182], sputtering [183, 184], electron-beam physical vapour
deposition (EB-PVD) [185] and electrodeposition [186]. However, a method that is
technically simpler; i.e., immersion in a simulated body fluid (SBF) such as Hank’s
solution [22, 189], is also feasible. Some attempts have been undertaken to achieve the
Ca/P ratio of stoichiometric synthetic hydroxyapatite (Ca10(PO4)6(OH)2) formed on a
nanotubular layer of TiO2 using the immersion method [189, 254] is 1.67. Although the
fabrication of a nanotubular layer on Ti50Zr alloy has been reported, its bioactivity in
inducing hydroxyapatite needs to be investigated.
In this chapter, eight different nanotubular layers of TiO2-ZrO2-ZrTiO4 and TiO2 were
fabricated with different inner diameter (Di), outer diameter (Do) and wall thickness
(Wt) dimensions and physical and chemical properties. Hydroxyapatite was mineralised
on the surface of (i) TiO2-ZrO2-ZrTiO4 nanotubes, (ii) the anodised surface (flat
137
surface), and (iii) bare Ti50Zr alloy and compared to (i) TiO2 nanotubes, (ii) the
anodised surface (flat surface), and (iii) bare CP-Ti to evaluate their degree of
bioactivity. The effect of nanotube size, the roughness and surface chemistry of the
nanotubular layer on inducing HA mineralisation was elucidated.
6.2 Materials and methods
Ti50Zr alloy was prepared by casting and then discs 8 mm diameter and 2 mm thickness
were fabricated by electric discharge machining. After polishing to a 1 μm diamond
finish, the samples were ultrasonically cleaned for 15 minutes each in methanol,
isopropanol, acetone and ethanol prior to rinsing with pure water and drying in a
nitrogen stream. CP-Ti (Baoji Boxin Metal Materials Co., Ltd.) foils (10×10 mm) were
degreased and washed in the same manner. The cleaned discs and foils were divided
into four groups. TiO2-ZrO2-ZrTiO4 and TiO2 nanotubes were fabricated on the surfaces
of the first group of samples using a DC power supply with two-electrode configuration
in 1 M (NH4)2SO4 and 0.5 wt % NH4F (Sigma-Aldrich reagent grades) at different
applied potentials. The electrolyte for second group of samples was ethylene glycol
contained 0.5 wt % NH4F and 5 and 10 wt % pure water (H2O). Compact TiO2-ZrO2-
ZrTiO4 and TiO2 films without nanotubular structure were produced on the surfaces of
the third group of samples by anodisation in 1 M (NH4)2SO4 at 20 V using the same
power supply. After anodisation the samples were washed with deionised water for 5
minutes and dried with nitrogen stream. The samples which were treated in non-
aqueous electrolyte were rinsed with acetone then were post-treated by immersion in 0.5
M HCl at room temperature for 1 min and finally rinsed with deionised water for 1 min
to dissolve the remained upper nanoporous layer. The forth group of samples was the
bare metals without any surface modification. Half of the electrochemically produced
surface layers were annealed at 500 ˚C for 3 h in air in a muffle furnace (Nabertherm
LT15/13/P330) at a heating rate of 2 ˚C min−1.
The sample surfaces, after anodisation and soaking in m-SBF [213], were analysed with
a field-emission scanning electron microscope (FESEM, ZEISS SUPRA 40 VP)
equipped with an energy-dispersive X-ray spectrometry (EDX). The crystalline phases
were determined by X-ray diffraction using Cu Kα incident radiation at 40 kV and
40 mA with a scanning step size of 0.02 ° from 2θ equal of 10 to 90 °.
138
The peaks, valleys, and the spacings of the surface textures were characterised by using
the roughness parameters including Mean Roughness (Sa) and Root Mean Square
(RMS) Roughness (Sq), the Skewness (Sskw) and the kurtosis (Sku) of a 3D surface
texture. These surface roughness parameters were measured using a 3D-Profilometer
(Bruker, Contour GT-K1; Bruker Pte Ltd, Singapore) coupled with SurfVision software
(Veeco Instruments Inc.; Plainview, NY, USA) [210]. The Sskw represents the degree of
symmetry of the surface heights about the mean plane and Sku indicates the nature of the
height distribution. It is notable that a Sskw > 0 implies high peaks; whereas Sskw < 0
indicates valley structures such as deep scratches. When Sku is 3.0, the surface heights
are normally distributed, Sku < 3.0, indicating a few high peaks and low valleys
(platykurtoic) and Sku > 3.0, indicating many high peaks and low valleys (leptokurtoic)
[210].
The water contact angle measurements were undertaken using a goniometer (NRL C.A.
Goniometer, Ramé-hart, Inc.; Succasunna, NJ-USA). The surface energy was calculated
using the Owens-Wendt (OW) method [211], as follows:
(1 + cos θ )γL=2(√γLd γS
d+√γLp γS
p) (3 − 1)
where L indicates the liquid surface tension (water=72.8 mJm-2, glycerol=63.4 mJm-2);
Ld and S
d are the liquid and solid dispersive component (water=21.8 mJm-2,
glycerol=37.0 mJm-2). The variables Lp and S
p indicate the liquid and solid polar
components (water=51.0 mJm-2, glycerol=26.4 mJm-2) and S is the sum of Sd and S
p
[212].
The samples were soaked in modified simulated body fluid (m-SBF) to evaluate the in
vitro bioactivity and assess their ability for inducing apatite formation. All three groups
of Ti50Zr and CP-Ti samples were soaked in m-SBF for 3 weeks. The m-SBF solutions
were prepared according to reference [213] with the composition and the preparation
method presented in Ref. [214]. The m-SBF was buffered with 2-(4-(2-hydroxyethyl)-1-
piperazinyl) ethane sulfonic acid (HEPES) (17.892 g HEPES dissolved in 100 ml 0.2 M
NaOH) and 1 M NaOH at pH 7.4 at 37˚C. m-SBF with ionic concentrations equal to
human blood plasma was prepared by dissolving reagent grade chemicals with the
composition 5.403 g NaCl, 0.504 g NaHCO3, 0.426 g Na2CO3, 0.225 g KCl, 0.230 g
139
K2HPO4 - 3H2O, 0.311 g MgCl2 - 6H2O, 0.293 g CaCl2, and 0.072 g Na2SO4in 1 litre
deionised water. After soaking, the samples were rinsed with pure water and dried at
room temperature. The change of pH value during incubation was measured using
pH/mV & Ion/pH meter Series (Oakton, Eutech instruments Pte Ltd, USA). The atomic
percentage of calcium to phosphate ratio of HA is calculated using EDS results.
6.3 Results and discussion
6.3.1 Dimensional, physical and chemical properties of TiO2-ZrO2-ZrTiO4
nanotubes fabricated in aqueous electrolyte
The nanotubular layer of TiO2-ZrO2-ZrTiO4 were fabricated in 0.5 wt % NH4F, 1 M
(NH4)2SO4 in pure water for 1 h at an applied potential of 20, 25, 30 and 35 V. Figure
4.9 in Chapter 4 shows the top view of as-formed and annealed nanotubes. The mean Di
of as-formed nanotubular layer of TiO2-ZrO2-ZrTiO4 increased from 37 ± 13 nm to 64 ±
18, 76 ± 20 and 81 ± 20 nm with an increase in applied potential from 20 to 25, 30 and
35 V. The mean Di of nanotubular layer of TiO2 also increased from 43 ± 10 nm to 52 ±
6, 65 ± 9 and 68 ± 6 nm with an increase of applied potential from 20 to 25, 30 and 35
V. Fig. 6.1 shows the mean size of nanotubular layer of TiO2-ZrO2-ZrTiO4 and TiO2.
140
Fig. 6.1 Mean nanotube size as a function of applied potential: (a) Ti50Zr - as-formed,
(b) Ti50Zr - annealed (c) CP-Ti - as-formed, (d) CP-Ti - annealed
The mean roughness (Sa) and the water contact angle (W.C.A.) of the nanotubular
surfaces as a function of the applied potential of the as-formed and annealed TiO2-ZrO2-
ZrTiO4 and TiO2 nanotubes anodised on Ti50Zr and CP-Ti are presented in Fig. 6.2. It
can be seen that the mean Sa of as-formed TiO2-ZrO2-ZrTiO4 nanotubes decreased with
an increase in Di of nanotubes and the Di increased with increasing applied potential.
Although the Sa of the annealed nanotubular surfaces slightly increased with the
increasing applied potential, they were smoother than those of as-formed TiO2-ZrO2-
ZrTiO4 nanotubular surfaces. The same tendency was observed for TiO2 nanotubes
although they revealed lower roughness amplitude parameters. After annealing the Sa of
the TiO2 nanotubes slightly increased or remained constant. The Sskw and Sku results for
the TiO2-ZrO2-ZrTiO4 nanotubes under all conditions were leptokutoic, which implies
0
20
40
60
80
100
120
140
160
180
20 25 30 35
Mea
n si
ze (n
m)
Applied Potential (V)
Ti50Zr-As Formed
Wt
Di
Do
a)
0
20
40
60
80
100
120
140
160
180
20 25 30 35
Mea
n si
ze (n
m)
Applied Potential (V)
Ti50Zr-Annealed
Wt
Di
Do
b)
0
20
40
60
80
100
120
140
160
180
20 25 30 35
Mea
n si
ze (n
m)
Applied Potential (V)
CP-Ti-As Formed
Wt
Di
Do
c)
0
20
40
60
80
100
120
140
160
180
20 25 30 35
Mea
n si
ze (n
m)
Applied Potential (V)
CP-Ti-Annealed
Wt
Di
Do
d)
141
that the surface exhibited mostly valleys as would be reflected by the needle-like
microstructure of the alloy. The same tendency was revealed after annealing except with
slightly lower values. The result of Sskw and Sku of the surface of TiO2 nanotubes
indicated that they displayed a normal distribution of peaks and valleys that were
similar to a flat surface and the bare metals of Ti50Zr and CP-Ti. After annealing the
Sskw and Sku of the nanotubular surfaces showed the same tendency with slightly lower
values.
Fig. 6.2 The mean roughness (Sa) of the nanotubular surfaces as a function of the
applied potential: (a) Ti50Zr alloy, (c) CP-Ti; and the mean water contact angle
(W.C.A.) of the nanotubular surfaces as a function of the applied potential: (b) Ti50Zr
alloy and (d) CP-Ti, Note: Each data point is an average of five measurements
The surface index of TiO2-ZrO2-ZrTiO4 nanotubes, which was calculated by dividing
the surface area of the nanotubular layer to the lateral surface area, decreased with an
increase in applied potential and Di, and the surface area index of TiO2-ZrO2-ZrTiO4
nanotubes was (i) a maximum 10.4 and a minimum of 3.5 times higher than TiO2
0
0.5
1
1.5
2
2.5
3
3.5
20 25 30 35
Mea
n Sa
(µm
)
Applied Potential (V)
Roughness of Ti50ZrAs Formed
Annealed
05
1015202530354045
20 25 30 35
Mea
n W
.C.A
(θ°)
Applied Potential (V)
Wettability of Ti50Zr
As Formed
Annealed
b)
0
0.5
1
1.5
2
2.5
3
3.5
20 25 30 35
Mea
n Sa
(µm
)
Applied Potential (V)
Roughness of CP-Ti As Formed
Annealed
c)
0
5
10
15
20
25
30
35
40
45
20 25 30 35
Mea
n W
.C.A
. (θ°
)
Applied Potential (V)
Wettability of CP-Ti As Formed
Annealed
d)
a)
142
nanotubes and (ii) a maximum of 17.6 and a minimum of 5.7 times higher than those of
the flat surface and bare Ti50Zr (Table 6.1).
143
Table 6.1 Surface area index SI and roughness amplitude parameters of Sq, Sskw and Sku of the nanotubular surfaces of Ti50Zr and CP-Ti anodised
under various conditions, Note: Each data point is an average of five measurements
Anodisation conditions
As-formed Annealed
Surface area index
SI Sq Sskw Sku
Surface area index
SI Sq Sskw Sku
20V Ti50Zr 17.80 ± 2.25 3.94 ± 0.19 -1.6 ± 0.08 5.26 ± 0.34 3.88 ± 1.02 0.87 ± 0.12 -1.91 ± 0.51 20.22 ± 8.27
CP-Ti 1.71 ± 0.07 0.72 ± 0.03 -0.67 ± 0.12 3.30 ± 0.12 1.49 ± 0.01 0.74 ± 0.06 -0.80 ± 0.03 2.97 ± 0.07
25V Ti50Zr 8.56 ± 1.75 2.29 ± 0.46 -2.28 ± 1.42 12.88 ± 9.26 2.91 ± 0.59 0.85 ± 0.17 -1.74 ± 1.51 17.53 ± 13.66
CP-Ti 1.35 ± 0.07 0.33 ± 0.04 -0.31 ± 0.17 3.43 ± 0.95 1.29 ± 0.01 0.34 ± 0.07 0.01 ± 0.09 2.95 ± 0.38
30V Ti50Zr 8.62 ± 1.30 1.75 ± 0.41 -2.62 ± 1.20 33.24 ± 34.43 4.18 ± 0.60 0.95 ± 0.10 -0.81 ± 0.63 9.51 ± 5.6
CP-Ti 1.35 ± 0.02 0.31 ± 0.01 0.51 ± 0.13 3.88 ± 0.13 1.30 ± 0.01 0.30 ± 0.00 -0.11 ± 0.14 3.09 ± 0.10
35V Ti50Zr 5.82 ± 1.27 1.45 ± 0.29 -3.29 ± 1.27 30.1 ± 17.35 3.97 ± 0.47 0.94 ± 0.06 -0.87 ±0.36 16.34 ± 16.25
CP-Ti 1.67 ± 0.02 0.47 ± 0.01 -0.45 ± 0.06 2.48 ± 0.08 1.66 ± 0.04 0.58 ± 0.03 0.56 ± 0.07 2.71 ± 0.07
Anodised in (NH4)2SO4 Ti50Zr 1.01 ± 0.00 0.06 ± 0.01 0.12 ± 0.35 9.34 ± 4.72 1.04 ± 0.02 0.18 ± 0.08 0.41 ± 1.2 5.53 ± 5.18
CP-Ti 1.39 ± 0.02 0.62 ± 0.17 0.19 ± 0.05 2.31 ± 0.23 1.45 ± 0.03 0.29 ± 0.01 -0.26 ± 0.15 3.86 ± 0.21
Bare Metal Ti50Zr 1.00 ± 0.00 0.05 ± 0.01 0.20 ± 0.58 9.64 ± 3.48 - - - -
CP-Ti 0.18 ± 0.00 -0.23 ± 0.00 4.9 ± 0.00 - - - - -
144
The water contact angle of TiO2-ZrO2-ZrTiO4 nanotubes decreased with an increase of
Di for both the as-formed and the annealed nanotubes. The values of water contact angle
on the surface of TiO2-ZrO2-ZrTiO4 after annealing were lower than those of the as-
formed nanotubes. The nanotubular layer of TiO2-ZrO2-ZrTiO4 exhibited super-
hydrophilic properties under almost all conditions except for the as-formed TiO2-ZrO2-
ZrTiO4 nanotubes fabricated at an applied potential of 20 V with Di = 37 nm which
exhibited hydrophilic properties. The water contact angle of TiO2 nanotubes decreased
with an increase of Di for both the as-formed and the annealed nanotubes and they
revealed lower values after annealing. The TiO2 nanotubular layer exhibited hydrophilic
properties in all conditions. The flat surface and bare metal surface of Ti50Zr and CP-Ti
revealed hydrophilic properties with water contact angles that were at least 2 times
higher than their nanotubular counterpart layers. It can be seen that for both the as-
formed TiO2-ZrO2-ZrTiO4 and TiO2 nanotubes the water contact angle increased with
an increase of Sa while it decreased on the annealed nanotubular layer. It can be
concluded that increasing Di and the crystalline phase of the nanotubular layer
influenced the hydrophilic properties more than the surface roughness. The surface
energy of TiO2-ZrO2-ZrTiO4, TiO2 nanotubes, flat surface and bare Ti50Zr alloy and
CP-Ti are listed in Table 6.2.
145
Table 6.2 Calculated surface energies of TiO2-ZrO2-ZrTiO4 nanotubular surfaces fabricated on Ti50Zr and TiO2 nanotubular surfaces on CP-Ti, Note:
Each data point is an average of five measurements
Anodisation conditions
As-formed Annealed
Contact angle
()
sd
(mJm-2)
sp
(mJm-2)
s = sd + s
p
(mJm-2)
Contact angle
()
sd
(mJm-2)
sp
(mJm-2)
s = sd + s
p
(mJm-2)
20V Ti50Zr 20.50 ± 0.4 15.98 52.67 68.65 2.7 ± 0.2 14.93 58.71 73.64
CP-Ti 28.7 ± 10.1 22.45 41.87 64.32 22.5 ± 4.8 12.95 55.56 68.51
25V Ti50Zr 9.5 ± 1.3 15.01 57.63 72.64 6.1 ± 0.3 14.69 58.66 73.35
CP-Ti 30.0 ± 1.0 18.14 45.24 63.38 13.5 ± 0.9 9.25 65.04 74.29
30V Ti50Zr 9.2 ± 0.3 15.45 57.13 72.58 5.9 ± 1.7 15.22 58.00 73.22
CP-Ti 25.3 ± 5.1 13.98 52.69 66.68 13.3 ± 2.0 13.33 58.83 72.16
35V Ti50Zr 6.5 ± 2.3 14.05 59.45 73.50 4.9 ± 0.7 14.20 59.47 73.67
CP-Ti 18.0 ± 3.0 13.62 56.74 70.36 13.3 ± 1.8 10.16 63.57 73.73
Anodised in (NH4)2SO4 Ti50Zr 47.7 ± 1.5 0.19 67.89 68.09 20.7 ± 1.5 3.59 74.42 78.01
CP-Ti 64.8 ± 3.8 7.29 30.25 37.54 47.2 ± 1.2 26.36 27.1 53.46
Bare Metal Ti50Zr 42.2 ± 1.8 14.56 40.68 55.24 - - - -
CP-Ti 54.5 ± 4.7 15.37 30.17 45.55 - - - -
146
6.3.2 Bioactivity of TiO2-ZrO2-ZrTiO4 and TiO2 nanotubes fabricated in aqueous
electrolyte
A dense and uniform layer of the bonelike apatite was formed on the surface, as shown
in Figs. 6.3 and 6.4, when the TiO2-ZrO2-ZrTiO4 and TiO2 nanotubes were soaked in m-
SBF. It has been reported that for various TiO2 layers, an anatase or a rutile structure of
TiO2 nanotubes were more favourable for apatite formation than an amorphous structure
[48, 254]. Therefore, some samples were annealed at 500 °C for 3 h to obtain crystalline
phases. The surface of the as-formed and annealed TiO2-ZrO2-ZrTiO4, TiO2 nanotubes
and flat surface were covered by nanobristle spheres (Fig. 6.3(f)) and a layer of
nanoflakes (Figs. 6.3(h) and 6.4(h)) after 3 weeks immersion in m-SB, respectively.
However, no HA precipitation was observed on bare metal samples after 3 weeks of
immersion and on flat surface within 3 weeks immersion in m-SBF.
147
Fig. 6.3 Mineralisation of HA on the nanotubular surface of TiO2-ZrO2-ZrTiO4
fabricated at a), b) 20 V (Di = 37 nm)-as-formed, c), d) 20V (Di = 40 nm)- annealed, e),
f) 25 V (Di = 64 nm)-as-formed, g), h) 25V (Di = 59 nm)-annealed, i), j) 30 V (Di = 76
nm)-as-formed, k), l) 30V (Di = 64 nm)-annealed, m), n) 35 V (Di = 81 nm)-as-formed
and o), p) 35V (Di = 82 nm)-annealed
148
Fig. 6.4 Mineralisation of HA on the nanotubular surface of TiO2 fabricated at a), b) 20
V (Di = 43 nm)-as-formed, c), d) 20V (Di = 51 nm)-annealed, e), f) 25 V (Di = 52 nm)-
as-formed, g), h) 25V (Di = 45 nm)-annealed, i), j) 30 V (Di = 65 nm)-as-formed, k), l)
30V (Di = 53 nm)-annealed, m), n) 35 V (Di = 68 nm)-as-formed and o), p) 35V (Di =
63 nm)-annealed
149
The maximum measured thickness of the HA layers mineralised on the surface of TiO2-
ZrO2-ZrTiO4 and TiO2 were 2.7 and 5.3 µm, respectively as shown in Fig. 6.5.
Fig. 6.5 SEM images of cross section of mineralised HA layers on the surface of a)
TiO2-ZrO2-ZrTiO4 nanotubes and b) TiO2 nanotubes fabricated in aqueous electrolyte
The EDS spectra of the above mentioned samples detected Ca, P and O peaks, which
are the essential ingredients of hydroxyapatite, after immersion in m-SBF for 3 weeks.
The XRD patterns of the TiO2-ZrO2-ZrTiO4 and TiO2 nanotubes before and after
immersion in m-SBF are shown in Fig. 6.6. The annealed specimen revealed the anatase
peak of titania and the orthorhombic phase for a mixture of titania, zirconia and
zirconium titanate prior immersion in m-SBF. After immersion in m-SBF for 3 weeks,
some hydroxyapatite (HA) peaks appear in the XRD patterns.
150
Fig. 6.6 XRD patterns of: a) TiO2-ZrO2-ZrTiO4 nanotubes anodised on Ti50Zr (α and β phases) and annealed at 500°C for 3 h before and after
immersion in m-SBF for 3 weeks, showing tetragonal anatase, srilankite (a mixture of orthorhombic TiO2 and ZrO2) and orthorhombic ZrTiO4
with mineralised hydroxyapatite, b) TiO2 nanotubes anodised on CP-Ti (α phase) and annealed at 500 °C for 3 h before and after immersion in
m-SBF for 3 weeks, showing tetragonal anatase with mineralised hydroxyapatite; and c) hydroxyapatite
151
Figure 6.7 shows the atomic percentage of the calcium to phosphate ratio (Ca/P) of HA
that was mineralised on the surfaces of the TiO2-ZrO2-ZrTiO4 and TiO2 nanotubes. The
Ca/P ratio of HA precipitated on the nanotubular layer of TiO2-ZrO2-ZrTiO4 increased
with an increase in Di for both the as-formed and annealed nanotubes. It can be
observed that the Ca/P ratio of HA increased with an increase of Sa for the as-formed
TiO2-ZrO2-ZrTiO4 nanotubes but it decreased with an increase of Sa for the annealed
samples. On the other hand, the Ca/P ratio increased with a decrease of water contact
angle for both the as-formed and annealed nanotubular layers. The Ca/P ratio of HA
increased after annealing for each nanotubular layer fabricated under the same
conditions regardless of the difference in the inner diameters (Di) of the nanotubes.
Fig. 6.7 The percentage of atomic ratio of Ca/P of mineralised HA on four different
nanotubular surfaces of a) Ti50Zr and b) CP-Ti
The Ca/P ratio of HA precipitated on the nanotubular layer of TiO2 decreased with an
increase of Di for both the as-formed and annealed nanotubes. Increasing Sa of both as-
formed and annealed TiO2 nanotubes resulted in an increase of the Ca/P ratio of HA;
although it decreased with a decrease of water contact angle. The Ca/P ratio of HA
remained constant after annealing for each nanotubular layer fabricated under the same
conditions regardless of different inner diameters. A HA layer formed on the flat surface
of the oxide layer on Ti50Zr and CP-Ti. The measured Ca/P ratios of HA were (i) 1.27
before and after annealing for Ti50Zr, and (ii) 1.29 before annealing and 1.22 after
1.21.251.3
1.351.4
1.451.5
1.551.6
1.651.7
20 25 30 35
Ca/
P (a
t %)
Applied Potential (V)
Ti50Zr As Formed
Annealed
1.21.25
1.31.35
1.41.45
1.51.55
1.61.65
1.7
20 25 30 35
Ca/
P (a
t %)
Applied Potential (V)
CP-Ti As Formed
Annealedb)a)
152
annealing for CP-Ti; the values of which were lower than their corresponding
nanotubular layers.
The reactions on the nanotubular layer subjected to the alkaline aqueous solution of m-
SBF are as follows:
TiO2.nH2O + OH- → HTiO3-.nH2O (6-1)
ZrO2.nH2O + OH- → HZrO3-.nH2O (6-2)
ZrTiO4. nH2O + OH- → HZrTiO5-. nH2O (6-3)
The isoelectric point of TiO2 is reported at a pH of 3.9 and the corresponding value for
ZrO2 is 5.5 [255, 256]. Both of these values are lower than the pH of m-SBF (7.4) and
indicate the formation of a negative layer of Ti-OH and Zr-OH. The negatively charged
surface absorbs the cations in the solution such as Na+, K+, Mg2+ and Ca2+ by Coulomb
force. According to our observation a patch like of salt crystals of the abovementioned
cations and existing anions such as Cl-, HCO3-, CO3
2-, HPO42- SO4
2- formed as a
transition state for the formation of the composition of calcium phosphate (Fig. 6.8).
These crystals were in equilibrium with their ions in the solution as they are not
sparingly soluble salts. A composition of calcium phosphate forms on the positively
charged surface (Ca rich layer) due to electrostatic interaction with the phosphate ion in
the m-SBF. Then the layer that is negatively charged absorbs more Ca2+ to form
amorphous apatite. With increasing the immersion time, the amorphous apatite
eventually transforms to a crystalline structure that has lower solubility in neutral and
basic solutions in comparison to other structural form of calcium phosphate. Sodium,
magnesium, chloride and carbonate ions may become incorporated into an apatite
similar to the compositional and structural features of bone mineral since the m-SBF
mimics the blood plasma not only in composition but also in concentration [251].
Octacalcium phosphate phase (Ca8H2(PO4)6·5H2O; OCP) is the first formed Ca-P
crystal nuclei in the simulated body fluid [257]. Thermodynamically OCP tends to
convert to a calcium deficient HA although it remains in the apatite lattice with a plate
like morphology. Fig. 6.9 schematically illustrates the process of bone-like apatite
formation on nanotubular surface in m-SBF.
153
Fig. 6.8 SEM images showing the transition of soluble salt crystal of present ions of m-
SBF into HA on the nanotubular surface of TiO2-ZrO2-ZrTiO4 with increasing
immersion time in m-SBF
Fig. 6.9 Schematic presentations of the process of bone-like apatite formation on
nanotubular surface in m-SBF
154
The local pH change up to 3 weeks of mineralisation of HA is shown in Fig. 6.10. The
pH value increased while negative hydrolysed nanotubular surface formed. During the
absorption of Ca2+ ions the changes of pH value were negligible. Then the pH value
increased as result of the exchange of OH- with PO43-. While the OH- consumed to form
hydroxyapatite [Ca10(PO4)6(OH)2] the pH value decreased and remaining constant after
the HA formation was completed.
Fig. 6.10 The pH value of static m-SBF as a function of soaking time for nanotubular
TiO2-ZrO2 -ZrTiO4 fabricated at 20, 25, 30 and 35V
Considering the observation for changing the Ca/P ratio of HA mineralised on the
surface of TiO2-ZrO2-ZrTiO4 nanotubes it can be deduced that there was a direct
relation between Di, hydrophilic properties and crystalline phase of nanotubes rather
than the roughness parameters for inducing the formation and growth of HA on
nanotubular surface. In terms of TiO2 nanotubes, the hydrophilic properties and
roughness parameter revealed a more significant influence on the formation and growth
155
of HA than the Di and the crystalline phase of the nanotubular layer. The role of
morphological features that give rise to high roughness lies in the creation of the initial
reaction sites for apatite nucleation and formation.
The nanomechanical characteristics of the mineralised HA on the surface of TiO2-ZrO2-
ZrTiO4 nanotubes were measured using a nanoindentation. The curve of the load-
unload forces versus the nanoindentation depths of the HA is shown in Fig. 6.11. The
nano-Young’s modulus was 23.9 GPa and the nano hardness was 87.0 MPa for the
nanotubular layer of TiO2-ZrO2-ZrTiO4. The average elastic modulus for the osteons
and the interstitial lamellae in the longitude direction was reported as 24.7 GPa and 30.1
GPa. The hardness was reported to be varied among the microstructure components in
the range of 0.41 - 0.89 GPa [258].
Fig. 6.11 Loading-unloading forces versus the nanoindentation depths of HA on
nanotubular surface of TiO2-ZrO2-ZrTiO4 fabricated at the applied potential of 35 V
6.3.3 Dimensional, physical and chemical properties of TiO2-ZrO2-ZrTiO4
nanotubes fabricated in non-aqueous electrolyte
The nanotubular layer of TiO2-ZrO2-ZrTiO4 were fabricated in 0.5 wt % NH4F, 5 and
10 wt % pure water (H2O) in ethylene glycol at 20 and 30 V. Figure 5.1 in Chapter 5
0
0.002
0.004
0.006
0.008
0.01
0.012
0 5E-10 1E-09 1.5E-09 2E-09 2.5E-09
Load
(N)
Displacement (m)
156
shows the top view of the as-formed and Fig. 6.12 shows the dimpled surface of
annealed nanotubes. The mean Di of as-formed nanotubular layer of TiO2-ZrO2-ZrTiO4
increased from 29 ± 5 nm to 35 ± 5 with an increase of water content from 5 to 10 wt %
at 20 V and increased from 34 ± 8 to 42 ± 10 nm at 30 V. The mean Di of nanotubular
layer of TiO2 also increased from 32 ± 5 nm to 37 ± 5 with an increase of water content
from 5 to 10 wt % at 20 V and increased from 42 ± 5 to 52 ± 5 nm at 30 V. These
nanotubular layer exhibited a brittle layer after annealing therefore, the cracked upper
layer was removed by ultrasound and remaining dimpled surface used for bioactivity
investigation.
Fig. 6.12 Dimpled surface of TiO2-ZrO2-ZrTiO4 nanotubes after annealing and
removing the upper cracked layer a) 5 wt % H2O - 20 V, b) 10 wt % H2O - 20 V, c) 5 wt
% H2O - 30 V and d) 10 wt % H2O - 30 V
Figure 5.8 in Chapter 5 shows the roughness parameters of Sa and Sq measured for the
as-formed nanotubular surfaces and Fig. 6.13 shows the roughness parameters of Sa and
157
Sq measured for the annealed nanotubular layer with a dimpled surface. These
amplitude parameters (Sa and Sq) as fully explained in Chapters 5 and 7 increased with
an increase of water content for each applied potential and with an increase of the
applied potential for each composition of electrolyte. The hydrophilic property of
nanotubes increased with an increase of the nanotube sizes and roughness parameters
that were the result of increasing the water content and applied potential. The nanotubes
fabricated in the electrolyte with 5 wt % H2O at 20 V with mean Di = 29 nm exhibited
Sskw ≈ 0 and Sku ≈ 3 which indicats uniform symmetry and a normal distribution of
surface heights. The other 3 nanotubular surfaces exhibited a more distinct leptokurtoic
distribution with many high peaks and valleys. The majority of the 4 dimpled surfaces
exhibited a uniform symmetry with normal distribution in surface heights. Table 5.2 in
Chapter 5 presents these amplitude parameters plus the surface index (SI). The
hydrophilic property of nanotubes increased with an increase of the nanotube sizes and
roughness parameters that were the result of increasing the water content and applied
potential. The surface energies of the as-formed TiO2-ZrO2-ZrTiO4 and annealed
dimpled surface TiO2-ZrO2-ZrTiO4 nanotubular surfaces calculated using the Owens-
Wendt (OW) method [211] are listed in Table 5.3 (Chapter 5).
Fig. 6.13 a) The mean roughness (Sa), (Sq) and the mean water contact angle (W.C.A.)
of the nanotubular and dimpled surface of TiO2-ZrO2-ZrTiO4, respectively, Note: Each
data point is an average of five measurements
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
25 (5 wt%H2O-20V) 29 (10 wt% H2O-20V) 45 (5 wt%H2O-30V) 49 (10 wt%H2O-30V)
Mea
n Sa
(µm
)
Mean Di (nm)
Roughness Sa-Annealed-Dimple
Sq-Annealed-Dimple
a)
0123456789
10
25 (5 wt%H2O-20V) 29 (10 wt% H2O-20V) 45 (5 wt%H2O-30V) 49 (10 wt%H2O-30V)
Mea
n W
.C.A
. (θ°
)
Mean Di (nm)
WettabilityW.C.A-Annealed-Dimpleb)
158
6.3.4 Bioactivity of TiO2-ZrO2-ZrTiO4 and TiO2 nanotubes fabricated in non-
aqueous electrolyte
The dimpled surface after removing the upper cracked layer of annealed TiO2-ZrO2-
ZrTiO4 nanotubes with 4 different nanotube sizes was investigated for their ability to
induce hydroxyapatite after immersion in m-SBF for 3 weeks. SEM images of
mineralised hydroxyapatite on the as-formed TiO2-ZrO2-ZrTiO4 nanotubes and dimpled
surface are shown in Fig. 6.14 and SEM images of mineralised hydroxyapatite on the
as-formed and annealed TiO2 nanotubes are shown in Fig. 6.15.
159
Fig. 6.14 SEM images of mineralisation of HA on the as-formed TiO2-ZrO2-ZrTiO4
nanotubes and dimpled surface fabricated in non-aqueous electrolyte: a, b) and c, d)
high and low magnification-5 wt % H2O-20V, e, f) and g, h) high and low
magnification-10 wt % H2O-20V, i, j) and k, l) high and low magnification-5 wt %
H2O-30V and m, n) and o, p) high and low magnification-10 wt % H2O-30V
160
Fig. 6.15 SEM images of mineralisation of HA on the as-formed and annealed TiO2
nanotubes fabricated in non-aqueous electrolyte: a, b) and c, d) high and low
magnification-5 wt % H2O-20V, e, f) and g, h) high and low magnification-10 wt %
H2O-20V, i, j) and k, l) high and low magnification-5 wt % H2O-30V and m, n) and o,
p) high and low magnification-10 wt % H2O-30V
161
The maximum thickness of the HA layers which mineralised on the surface of
nanotubular TiO2-ZrO2-ZrTiO4 was 9.6 µm. The maximum thickness of the HA layers
which mineralised on the surface of TiO2 nanotubes was 7.1 µm (Fig. 6.16).
Fig. 6.16 SEM images of cross section of mineralised HA layers on the surface of a)
TiO2-ZrO2-ZrTiO4 nanotubes and b) TiO2 nanotubes fabricated in non-aqueous
electrolyte
The peaks of Ca, P and O peaks, the essential ingredient of hydroxyapatite, were
detected by the EDS spectra of all above mentioned samples. The XRD patterns of the
TiO2-ZrO2-ZrTiO4 nanotubes fabricated in non-aqueous electrolyte before and after
immersion in m-SBF are shown in Fig. 6.17. The anatase peak of titania and
orthorhombic phase of the mixture of titania, zirconia and zirconium titanate of the
annealed specimen before immersion in m-SBF are visible. Some hydroxyapatite (HA)
peaks appeared after immersion in m-SBF for 3 weeks, in the XRD patterns.
162
Fig. 6.17 XRD patterns of the nanotube samples fabricated on Ti50Zr (α and β phases)
via anodisation in non-aqueous electrolyte, annealed dimpled surface (at 500 °C for 2 h)
of tetragonal anatase, rutile, srilankite
The atomic ratio of calcium to phosphate calculated using EDS results is shown in Fig.
6.18. If a biomaterial implant after implantation in vivo is encapsulated by fibrous
tissues it would be isolated from the surrounding bone. Formation of a carbonated
apatite layer on the surface of a biomaterial implant similar to hydroxyapatite (HA) of
bone composition is an ideal characteristic for healing mechanism. The Ti-OH and Zr-
OH groups located on the surface of TiO2-ZrO2-ZrTiO4 nanotubes are favoured sites for
apatite nucleation [56, 251, 254, 257]. The Ca2+ ions firstly are absorbed onto the
hydrolysed nanotubular oxide surface by Coulomb attraction forces. The surface would
be positively charged therefore the existing phosphate groups inside the SBF are
adsorbed to the surface and calcium phosphate forms. The stoichiometric Ca/P (at %) of
163
hydroxyapatite [Ca10(PO4)6(OH)2] is 1.67. According to the obtained Ca/P ratio of
eight different nanotubular surfaces, as-formed nanotubular surface with Di = 29 and 35
nm had higher value of Ca/P ratio than the as-formed nanotubular surface with Di = 34
and 42 nm. This may indicate the effect of nanospacing and roughness parameter and
hydrophilic properties although their differences are not considerable. The dimpled
surface of annealed TiO2-ZrO2-ZrTiO4 nanotubes with Di = 29 and 49 nm had higher
value of Ca/P ratio than Di = 25 and 45 nm because of their larger diameter and lower
Sa.
Fig. 6.18 The percentage of atomic ratio of Ca/P of mineralised HA on a) as-formed and
annealed dimpled surface of TiO2-ZrO2-ZrTiO4 nanotubes and b) as-formed and
annealed TiO2 nanotubes fabricated in non-aqueous electrolyte
Fig. 6.19 shows local pH changes up to 3 weeks of mineralisation of HA. Firstly the pH
value increased due to the formation of negative hydrolysed nanotubular surface. The
changes of pH value were negligible when the absorption of Ca2+ ions was undertaken.
The absorption of PO43- as a result of exchanging with OH- increased the pH value. The
pH value decreased when the OH- consumed to form hydroxyapatite [Ca10(PO4)6(OH)2]
and it remained constant after the HA formation completed.
1
1.1
1.2
1.3
1.4
1.5
1.6
1.7
5%-20 10%-20 5%-30 10%-30
Ca/
P (a
t %)
Water content (wt%)and applied potential (V)
Ti50Zr As FormedAnnealed
a)
1
1.1
1.2
1.3
1.4
1.5
1.6
1.7
5%-20 10%-20 5%-30 10%-30
Ca/
P (a
t %)
Water content (wt%)and applied potential (V)
CP-Ti As FormedAnnealed
b)
164
Fig. 6.19 The pH value of static m-SBF as a function of soaking time for nanotubular
TiO2-ZrO2-ZrTiO4 in a non-aqueous electrolyte with 5 and 10 wt % H2O at 20 and 30 V
The nanomechanical characteristics of the mineralised HA on the surface of TiO2-ZrO2-
ZrTiO4 nanotubes were measured using a nanoindentation. Fig. 6.20 shows the curve of
the load-unload forces versus the nanoindentation depths of the HA. The Young’s
modulus was 49.9 GPa and the nano hardness was 599.7 MPa. It has been reported that
the average elastic modulus for the osteons and the interstitial lamellae in the longitude
direction was 24.7 GPa and 30.1 GPa. As mentioned before the hardness was reported
to be varied among the microstructure components in the range of 0.41 - 0.89 GPa
[258].
165
Fig. 6.20 Loading-unloading forces versus the nanoindentation depths of HA on
nanotubular surface of TiO2-ZrO2-ZrTiO4 fabricated at 10 wt % H2O and the applied
potential of 30 V
6.4 Conclusions
Nanotubular surfaces of TiO2-ZrO2-ZrTiO4 and TiO2 were fabricated via anodisation on
Ti50Zr and CP-Ti. The formation of hydroxyapatite on the TiO2-ZrO2-ZrTiO4 and TiO2
nanotubular surfaces in comparison to flat anodised surface were evaluated. The main
conclusions are as follows.
1. Increasing the applied potential resulted in an increase in Di, leading to (i)
decreasing roughness parameters of Sa and Sq, (ii) increasing hydrophilic
properties and (iii) increasing the surface energy of TiO2-ZrO2-ZrTiO4
nanotubes. The mean Sa and Sq of nanotubular surface decreased,
hydrophilic properties and surface energy increased after annealing. The
same tendency was observed for TiO2 nanotubes on the surface of CP-Ti,
although their hydrophilic properties and surface energy were lower than
those for TiO2-ZrO2-ZrTiO4 nanotubes.
2. All the nanotubular surfaces of TiO2-ZrO2-ZrTiO4 fabricated in aqueous
electrolyte were bioactive and the size, shape, hydrophilic properties and
crystalline phase of the surface layers affected the Ca/P atomic ratio of the
0
0.002
0.004
0.006
0.008
0.01
0.012
0 2E-10 4E-10 6E-10 8E-10 1E-09
Load
(N)
Displacement (m)
166
deposited HA during SBF immersion. All the nanotubular surfaces of TiO2
fabricated in aqueous electrolyte were bioactive and the growth of HA was
influenced by the hydrophilic properties and roughness parameter rather than
the Di and crystalline phase of the nanotubular layer.
3. The nanoscale topography and chemistry of the nanotubular surfaces
affected the mineralisation of hydroxyapatite while immersion in SBF. The
bioactivity of the nanotubular surfaces was clearly enhanced compared with
the flat anodised surfaces and the bare metals of Ti50Zr and pure Ti.
4. Increasing of water content and applied potential resulted in an increase in
mean Di, Sa and Sq of TiO2-ZrO2-ZrTiO4 fabricated in non-aqueous
electrolyte. The hydrophilic properties of the nanotubular surface increased
with an increase in the nanotube sizes and roughness parameters.
5. All the nano size surfaces of the as-formed and annealed dimpled surface of
TiO2-ZrO2-ZrTiO4 and TiO2 nanotubes were bioactive and the size, shape,
roughness parameters and hydrophilic properties of the nanotubular layers
affected the atomic ratio of Ca/P after immersion in the m-SBF for 3 weeks.
167
Chapter 7 Cell response of nanotubes formed in both aqueous
electrolyte and non-aqueous electrolyte
Abstract
Studies of biomaterial surfaces and their influence on cell behaviour provide insights
concerning the design of surface physicochemical and topography properties of implant
materials. Fabrication of biocompatible metal oxide nanotubes on metallic biomaterials,
especially titanium alloys such as Ti50Zr via anodisation, alters the surface chemistry as
well as surface topography of the alloy. In this chapter, different groups of TiO2-ZrO2-
ZrTiO4 nanotubes that exhibit diverse nanoscale dimensional characteristics (i.e., inner
diameter Di, outer diameter Do and wall thicknesses Wt) and surface chemistry were
fabricated via anodisation in both aqueous and non-aqueous electrolytes. Two groups of
TiO2-ZrO2-ZrTiO4 nanotubes were post treated to assess their biocompatibility. The
nanotubes were annealed and characterised using scanning electron microscopy (SEM)
and 3D-Profilemetry. The potential applied during anodisation influenced the oxidation
rate of titanium and zirconium; thereby resulting in different nanoscale characteristics
for the nanotubes. Both the different oxidation and dissolution rates led to changes in
the surface roughness parameters. The in vitro cell response to the nanotubes with
different nanoscale dimensional characteristics was assessed using osteoblast cells
(SaOS2). Results of the MTS assay indicated that the nanotubes fabricated in aqueous
electrolyte with inner diameter (Di) ~ 18 nm exhibited the highest percentage of cell
adhesion of 65.2 %. This result can be compared to (i) 64.4 % cell adhesion at Di ~
30 nm, (ii) 41.0 % cell adhesion at Di ~ 40 nm, (iii) 25.9 % cell adhesion at Di ~ 59 nm,
(iv) 33.1 % at Di ~ 64 nm, and (v) 33.5 % at Di ~ 82 nm. The nanotubes with Di ~
59 nm exhibited the greatest roughness parameter of Sa (Mean Roughness), leading to
the lowest ability to interlock with SaOS2 cells. Post-treating on nanotubular surface by
making it cleaner and cover it with hydroxyapatite (HA) resulted in an increase in cell
adhesion notably to 74.8 % and 59.7 %, respectively. Results of the MTS assay
indicated that the percentage of cell adhesion on the nanotubes fabricated in non-
aqueous electrolyte was critically affected by the nanoscale topographical parameters
including the tube inner diameter (Di), the tube wall thickness (Wt), the amplitude
roughness (Sa) and the spacing roughness (Sm) of the nanotubular surface. The cell
adhesion was essentially promoted to 84.9 % on nanotubes with small inner diameter of
168
25 nm, or 80.3 % on nanotubes with large wall thickness of 34 nm due to the
accelerated integrin clustering and focal contacts formation. A nanotubular surface with
low spacing roughness of 33 nm3/nm2 led to a low cell adhesion of 61.0 %; similarly a
nanotubular surface with high amplitude roughness of 1.03 µm showed a cell adhesion
of 61.5 % even the inner diameters (29 nm) and the wall thicknesses (24 nm) of the
nanotubes were within the critical borderlines for cells to survive and thrive.
7.1 Introduction
The application of orthopeadic or dental implants is effectual when they form a solid
connective boundary between the bone tissue and the surface of the material in the
absence of a fibrous tissue interface [15]. This attribute is related to the molecules
involved in the adhesion of bone cells as well as the morphology and physico-chemical
properties of the implant surface. The morphology and capacity for proliferation and
differentiation of bone cells are affected by the quality of this adhesion [15]. The
biocompatible materials that are also known as ‘hybrid materials’ or ‘osteoconductive
structural materials’ can adsorb biomolecules and water from body fluid or culture
medium in vivo or in vitro [15] because cells are not equipped to interact directly with
biomaterials [34]. There are several stages regarding cell adhesion, which are known as
short-term and long-term adhesion. These phases of cell adhesion to an implant surface
involve different forces that are documented [15, 34, 137]. Surface chemistry and
surface energy influence short term adhesion; whereas topography affects long-term
adhesion as well as the cell morphology, proliferation, differentiation and migration [15,
259].
Chemical surface treatment is a surface modification method that increases
biocompatibility and bioactivity of biomaterial surfaces, especially that of metals.
Anodic oxidation, which is an electrochemical surface treatment, can be used to
fabricate a nanotubular or nanoporous layer of metal oxides on the surface of titanium
(Ti), zirconium (Zr), tantalum (Ta), niobium (Nb) and hafnium (Hf). Titanium and
binary, tertiary and higher titanium alloys especially alloyed with Zr, Ta and Nb
elements have been used extensively for orthopeadic implants due to their favourable
mechanical properties and biocompatibility, which is related to their ability to form a
protective oxide surface in air. A range of nanoscale dimensional characteristics (i.e.,
inner diameter, outer diameter and wall thickness) have been fabricated by altering
169
electrochemical conditions such the type and concentration of electrolyte, applied
potential and anodisation time [11, 234, 260]. Several studies [11, 165, 167, 177, 180]
have shown that the bioactivity of nanotubes is a function of their fabrication method
and that the bioactivity may increase by post treatment such as annealing.
The effect of macro, micro and nano sized topography on cell adhesion has been
reported [261-263]. Contact guidance of cells defines the alignment of cells that is
affected by the substrate surface morphology. The importance of the roughness
amplitude (height) parameter, space and hybrid parameters on cell adhesion has been
investigated [137, 264]. Surface chemistry plays an important role in cell adhesion
because it influences short term adhesion, which is a step that involves adsorption and
rearrangement of proteins [137]. It has been reported that cells adhere better on
hydrophilic surfaces, on surfaces with higher surface energy, and surfaces with
functional groups such as OH- and NH2- [137]. TiO2 nanotubes, as fabricated by an
electrochemical surface treatment, have exhibited an improved surface modification due
to (i) changes in the surface topography, such as the roughness and organisation of the
surface, and (ii) changes in surface chemistry that effect an increase in hydrophilic
properties and surface energy [180, 240, 265].
The investigation of different sizes of nanotubes indicated that cellular activity
increased when the length scale was less than 30 nm with a maximum activity at 15 nm.
A tube diameter larger than 50 nm decreased cellular activity [17]. The surface occupied
by the head of integrins (which are trans-membrane receptors that bind to specific
amino acids of proteins to form focal contact) has been reported to be around 10 nm
[137]. Consideration of these dimensions indicates that 15 nm can be an optimum
spacing for intergrin clustering and focal contact formation [17]. This larger length scale
was suitable for both bone forming cells and bone resorbing cells [266]. In another
study of human mesenchymal stem cells (hMSC), these cells adhered and grew with
minimum differentiation on small nanotubes of <50 nm. On the other hand, on larger
diameter nanotubes, i.e., > 50 nm, the hMSC cells became elongated and were guided to
differentiate into osteoblast cells. This result agreed with the general belief that cells
incline to differentiate into a specific lineage when they are forced to adapt under stress
[21]. There is also a report on an up-regulated level of ALP (alkaline phosphatase)
activity and osteoblast (MC3T3-E1 mouse) elongation on annealed TiO2 nanotubes;
170
which demonstrated a higher cell adhesion on 100 nm nanotubes than on 30 - 70 nm
nanotubes [20].
It has been shown that titanium zirconium alloy (TiZr) is a biocompatible titanium alloy
[252]. Therefore, studies have investigated the biocompatibility of TiZr alloy by
growing nanotubes on its surface via anodisation and rapid breakdown anodisation
(RBA) [151, 162, 267]. Various nanoarchitectures were fabricated using two-step
anodisation in an non-aqueous electrolyte to obtain nanopattern scaffold for cell growth
[268]. The influence of surface energy of titanium-zirconium alloy on osteoblast cell
functions was reported [27], therefore it would be a worthwhile study to investigate the
effect of the size, the roughness and wettability of nanotubes on the growth of bone
cells. The effect of size and the roughness of nanotubes have been discussed in Chapters
4 and 5. In this chapter TiO2-ZrO2-ZrTiO4 nanotubular layers were fabricated in
aqueous electrolyte with mean Di (i) 18 ± 6, (ii) 30 ± 10, (iii) 40 ± 12, (iv) 59 ± 17, (v)
64 ± 23 and (vi) 82 ± 26 nm; Do equal to (i) 38 ± 7, (ii) 63 ± 10, (iii) 74 ± 14, (iv) 102 ±
17, (v) 117 ± 26 and (vi) 136 ± 25; and Wt equal to (i) 7 ± 3, (ii) 11 ± 4, (iii) 13 ± 4, (iv)
17 ± 5, (v) 21 ± 7 and (vi) 24 ± 7 nm, respectively. The effect of some post surface
modification for improving the cell adhesion was assessed. Another group of TiO2-
ZrO2-ZrTiO4 nanotubes with dimpled surface were fabricated in non-aqueous
electrolyte by removing the upper layer of annealed nanotubes. The roughness and
water contact angle were measured and their surface energies calculated. SaOS2 cells
were seeded onto the TiO2-ZrO2-ZrTiO4 nanotubular surfaces and the effects of the
distribution of nanotube size, roughness parameters and surface energy of the
nanotubular surface on the cellular activity were comprehensively investigated.
7.2 Materials and methods
Ti50Zr alloy (wt % hereafter) discs with 8 mm diameter and 2 mm thickness were
electrical discharge machined (EDM) from a cast ingot. The disc samples were polished
to a 1 µm finish with diamond paste. Subsequently the discs were degreased by
sonification in, successively, methanol, isopropanol, acetone and ethanol for 15 min
each; i.e., a total wash time of 1 hour. The final preparation involved a wash with
deionised water and drying in a stream of nitrogen gas.
171
The Ti50Zr discs were spot-welded to titanium foil to enable the construction of the
anode for an electrochemical cell. The samples were placed 4 cm from the platinum
counter electrode with a 1 cm2 surface area in a two-electrode configuration using a DC
power supply. One group of TiO2-ZrO2-ZrTiO4 nanotubes were fabricated via
anodisation at room temperature with the electrolyte composed of 1 M (NH4)2SO4 with
the addition of 0.5 wt % of NH4F (Sigma-Aldrich reagent grades). After the
electrochemical treatment the samples were rinsed in deionised water for 5 min and
dried with a nitrogen gas stream.
Six applied potentials, i.e., 10, 15, 20, 25, 30 and 35 V, were used to create different
sizes and morphologies of nanotubes. The TiO2-ZrO2-ZrTiO4 nanotubes were pre-
treated by immersion in 0.5 M NaOH at 50 ˚C for 2 min, and then rinsed with deionised
water to increase biomineralisation. The samples were then annealed at 500 ˚C for 3 h at
the heating rate of 2 ˚C/min in a conventional muffle furnace (Nabertherm
LT15/13/P330; Nabertherm GmbH, Lilienthal, Germany). To reduce any contamination
of remaining chemicals from the electrolyte some samples were soaked at 70 °C for 14
days in pure water with periodic replacements (three times per day). To increase
osseointegration one series of samples were soaked in modified simulated body fluid
(m-SBFand incubated at 37 ˚C for 3 weeks. The ion compositions of the m-SBF [213],
which have an ion composition nearly equal to that of blood plasma, were adjusted by
dissolving 5.403 g NaCl, 0.504 g NaHCO3, 0.426 g Na2CO3, 0.225 g KCl, 0.230 g
K2HPO4 - 3H2O, 0.311 g MgCl2 - 6H2O, 0.293 g CaCl2, and 0.072 g Na2SO4 in
deionised water as previously reported [213]. The m-SBF was buffered at pH 7.4 at
37 ˚C using 2-(4-(2-hydroxyethyl)-1-piperazinyl) ethane sulfonic acid (aka HEPES) and
1 M NaOH. An amount of 17.892 g HEPES has dissolved in 100 ml 0.2 M NaOH to
adjust the pH. The ability of the samples to form apatite was evaluated in a static SBF
environment. After removing the samples from m-SBF they were rinsed with deionised
water and dried at room temperature for 24 h.
Another group of TiO2-ZrO2-ZrTiO4 nanotubues were fabricated in non-aqueous
electrolyte. The electrolyte was composed of 0.5 wt % of NH4F with the addition of 5
and 10 wt % of H2O in ethylene glycol (Sigma-Aldrich reagent grades). After the
electrochemical treatment the sample surfaces were rinsed with acetone for 5 min and
followed by drying in air. Two applied potentials, i.e., 20 and 30 V were used for each
172
electrolyte composition to create different sizes and morphologies of nanotubes. The
TiO2-ZrO2-ZrTiO4 nanotubes were post-treated by immersion in 0.5 M HCl at room
temperature for 1 min, then rinsed with deionised water for 1 min to dissolve the
remained upper nanoporous layer. The samples were then heat-treated at 500 ˚C for 2 h
at the heating rate of 2 ˚C/min in a conventional muffle furnace (Nabertherm
LT15/13/P330; Nabertherm GmbH, Lilienthal, Germany) to crystallise the amorphous
nanotubular layer. The brittle and cracked nanotubular layer after annealing were
removed by sonification in pure water for 15 min and dried in a stream of nitrogen gas.
To enhance the bioactivity the nanotubular layer was chemically treated by immersing
in 0.5 M NaOH solution at 50 °C for up to 2 min then rinsed with pure water for 2 min
and dried in a stream of nitrogen gas. The inner diameter, outer diameter and wall
thickness of the anodised nanotubes were denoted Di, Do and Wt, respectively. Each data
point is an average of 100 measurements on different positions for each of three samples
(300 measurements).
A field-emission scanning electron microscope (FESEM, ZEISS SUPRA 40 VP) was
used for metallographic characterisation. The amplitude parameters and space
parameters was determined with a 3D-Profilometer (Bruker, Contour GT-K1; Bruker
Pte Ltd, Singapore) that was coupled with SurfVision software (Veeco Instruments Inc.;
Plainview, NY, USA). Mean Roughness (Sa) and Root Mean Square (RMS) Roughness
(Sq), amplitude parameters, which measure the vertical characteristics of surface
deviations, were evaluated over the complete 3D surface [210]. Sskw, the skewness of a
3D surface texture, represents the degree of symmetry of the surface heights about the
mean plane, and it is the third central moment of the profile amplitude probability
density function [210]. The kurtosis of the 3D surface texture, Sku, indicates the nature
of the height distribution that is the fourth central moment of the profile amplitude
probability density function and describes the sharpness of the distribution [210]. The
spacing parameters such as Sm (Surface Material Volume), Sc (Core Void Volume) and
Sv (Surface Void Volume) that represent the volume of material or space provided by
the surface relative to the cross sectional area of the measurement were evaluated [210].
A goniometer (NRL C.A. Goniometer, Ramé-hart, Inc.; Succasunna, NJ-USA) was
used to measure the water contact angle. The surface energy was calculated based on
the Owens-Wendt (OW) method [211], given by:
173
(1 + cos θ )γL=2(√γLd γS
d+√γLp γS
p) (3 − 1)
where L indicates the liquid surface tension (water=72.8 mJm-2, glycerol=63.4 mJm-2);
Ld and S
d are the liquid and solid dispersive component (water=21.8 mJm-2,
glycerol=37.0 mJm-2). The variables Lp and S
p are indicative of the liquid and solid
polar components (water=51.0 mJm-2, glycerol=26.4 mJm-2) and S is the sum of Sd
and Sp [212].
All samples for the cell culture studies were sterilised in a muffle furnace at 180 ˚C for
3 h. The samples were placed in a well in the cell culture plate. SaOS2 cells (Barwon
Biomedical Research, Geelong Hospital, Victoria, Australia) were seeded on the surface
of the TiO2-ZrO2-ZrTiO4 samples with different inner, outer diameter and wall
thickness. The seeded cell density was 5×103 cells per well (100 mm2 area). A MTS
assay was used to measure the in vitro proliferation of the SaOS2 cells. The cell
morphology was observed by means of SEM after cell culturing. The cells were
dehydrated for the SEM observation by immersion in a buffer solution in which the
ethanol concentrations were increased progressively every 10 min to 60, 70, 80, 90 and
100 %. This procedure was followed by chemical drying using hexamethyldisilazane
(HMDS, Sigma - Aldrich, Australia) for 10 min. A gold layer was deposited on the
samples prior to SEM observation.
The cell morphology also was observed using a confocal microscopy (Leica SP5, Leica
Microsystems, Germany). The cell-seeded samples after cell culture were fixed in
paraformaldehyde, and then permeabilised with triton-X 100 in phosphate-buffered
saline (PBS) (Sigma - Aldrich, Australia) for 10 min each at room temperature for the
confocal microscopy observation. Then the samples were stained with 1 % phalloidin
and 40-6-diamidino-2-phenylindole for 40 min at room temperature. Between each of
the steps three washes by PBS were included. If storage of stained samples was required
they were stored in PBS at 4 °C and the observations were conducted within a week of
staining. Significant differences in the cell number were analysed using one-way
ANOVA (ρ < 0.05).
174
7.3 Results and discussion
7.3.1 Dimensional, physical and chemical properties of TiO2-ZrO2-ZrTiO4
nanotubes fabricated in aqueous electrolyte
The SEM images of the TiO2-ZrO2-ZrTiO4 nanotubes fabricated on Ti50Zr alloy under
different applied potentials and then annealed at 500 oC for 3 h are shown in Fig. 7.1.
The top view of the TiO2-ZrO2-ZrTiO4 nanotubes fabricated in 0.5 wt % NH4F at 10 V
for 1 h is shown in Fig. 7.1(a). In this study, the F- concentration in the electrolyte,
nanotube inner diameter, outer diameter and wall thickness are denoted as CF-, Di, Do,
and Wt, respectively.
Fig. 7.1 SEM images of annealed TiO2-ZrO2-ZrTiO4 nanotubes fabricated at a) 10 V, b)
15 V, c) 20 V, and d) 25 V, e) 30 V and f) 35 V
175
The mechanism of growing nanotubes on the surface of titanium, which is the result of
competition between oxidation of the metal to TiO2 and dissolution of TiO2 to form
[TiF6]2-, is elucidated in prior chapters [58, 249]. The formation of TiO2-ZrO2-ZrTiO4
nanotubes was presented previously [240]. The principal oxidation and reduction
reactions occur as follows:
Ti Ti4+ + 4 e- (7-1)
Zr Zr4+ + 4 e- (7-2)
4H+ + 4e- 2H2 (7-3)
The formation of a soluble complex of Ti4+ and Zr4+ with the ligand of F- takes place as:
Ti4+ + Zr4+ + 12 F-
[TiF6]2- + [ZrF6]2- (7-4)
Using anodisation in the electrolyte of 0.5 wt % of NH4F, titania-zirconia-zirconium
titanate (TiO2-ZrO2-ZrTiO4) nanotubes were fabricated on the surface of Ti50Zr alloy at
an applied potential in the range of 10 - 35 V for 1 h. The original grain structure of the
alloy was reflected in the subsequent growth of the nanotubes with an orientation based
on the needle-like morphology of the underlying alloy [240].
Figure 7.2 shows the mean Di, Do and Wt of annealed TiO2-ZrO2-ZrTiO4 nanotubes
anodised in an electrolyte with a constant CF- of 0.5 wt % and where the applied
potential increased from 10 to 35 V for 1 h. The mean Di of nanotubes increased with
respect to potentials of 10, 15, 20, 25, 30 and 35 V. A similar increase in the mean of Do
nanotubes at 10 V was from nm to 53 < Do < 73 and increased upto 111 < Do < 161 nm,
respectively when the applied potential increased to 15, 20, 25, 30 and 35 V. The
difference in Wt was between 12 to 14 nm by increasing the applied potential starting
10 V and reaching to 35 V.
176
Fig. 7.2 Mean nanotube size of Di, Do and Wt for annealed TiO2-ZrO2-ZrTiO4 nanotubes
anodised on Ti50Zr as a function of applied potential
The distribution of Di, Do and Wt for Ti50Zr alloy resulted from the different oxidation
and soluble complex formation of Ti and Zr with F- ions [240]. The frequency of
smaller sized Di and Do decreased on increasing the applied potential; thus enabling the
formation of larger inner and outer diameters due to the higher anodic current. The
broad distribution of Di and Do at 30 and 35 V shows a better condition for oxidation of
Ti and Zr because of high current flow while the dissolution rate of (NH4)2[TiF6] and
(NH4)2[ZrF6] remained constant.
Figure 7.3 shows the effect of changing nanotube size distribution on surface roughness
amplitude parameters. The values of Sa and Sq increased with an increase in applied
potential up to 25 V (Di = 59 nm). They revealed lower values at 30 and 35 V although
the mean Di increased. The texture of a surface, which is defined as surface roughness,
affects the character of the surface when it is contrasted to another surface of different
texture. There are several classical parameters used to evaluate surface roughnesses.
Amplitude parameters that are used to measure the vertical surface deviation
0
20
40
60
80
100
120
140
160
180
10 15 20 25 30 35
Mea
n na
notu
be si
zes (
nm)
Applied Potential (V)
WtDiDo
177
characteristics are one of the three roughness parameters [210]. The value of Sa
increased with increasing applied potential due to the availability of a larger anodic
current and its effect on oxidation reaction and consequently on dissolution.
Fig. 7.3 Illustration of changing roughness parameters of Sa and Sq via changing Di of
TiO2-ZrO2-ZrTiO4 nanotubes anodised on Ti50Zr
Ti50Zr alloy consists of two phases; α and β that contain a different amount of Zr. The β
phase consists of higher amount of Zr than α phase. The standard enthalpies of
formation of ZrO2 and TiO2 are different, and also the solubility equilibrium constant of
(NH4)2[ZrF6] is higher than that of (NH4)2[TiF6]. Therefore, two different heights of
TiO2 nanotubes and ZrO2 nanotubes with interspace formed [237]. The reduction of Sa
for nanotubular layers fabricated at 30 and 35 V can be explained by examining the
equilibrium state between the oxidation and dissolution reactions that was achieved at a
higher current density. Thus, the kinetics of oxidation in comparison to dissolution was
higher in the same period of time, which resulted in larger Di and Do and lower Sa and
Sq. The inference is that there was more physio-chemical activity inside the tubes than
along the length dimension.
0
0.5
1
1.5
2
2.5
18 30 40 59 64 82
Rou
ghne
ss a
mpi
tude
par
amet
ers -
(µm
)
Mean of Di (nm)
Sa
Sq
178
Consider a line, or ruler, that is located parallel to and intersecting a roughened surface.
This line will have intervals of intersection with the 2-D surface profile. The ratio of the
intersection lengths with the original measurement is defined as the bearing area. As the
ruler is moved more deep into the surface profile then the bearing area will increase and
the bearing area curve can be computed. The amount of material contained in the
surface peaks accounts for 0 to 10 % of the bearing area ratio and is represented by the
parameter Sm. In an alike manner, the volume that the surface would support from 10 -
80 % of the bearing ratio is Sc, and the relevant parameter from 80 - 100 % of the
bearing ratio is Sv. The dimensions of Sm, Sc, and Sv are nm3/nm2 [269]. Table 7.1
details the surface area index (SI) that is calculated by dividing the projected surface
area, which is the total exposed three-dimensional surface area being analysed, to the
surface area measured in the lateral direction, as well as the spacing parameters of Sm,
Sc and Sv. The nanotubular layer with Di = 64 nm revealed lowest Sm and the nanotubes
layer with Di = 59 nm revealed highest Sm.
179
Table 7.1 Surface index and spacing parameters of TiO2-ZrO2-ZrTiO4 nanotubes
fabricated in different applied potential
Anodisation Condition
(Mean size) SI Sm
(nm3/nm2)
Sc
(nm3/nm2)
Sv
(nm3/nm2)
10V (Di = 18 nm) 2.2 ± 0.3 34 ± 10 556 ± 106 153 ± 35
15V (Di = 30 nm) 2.5 ± 0.5 35 ± 10 820 ± 200 95 ± 33
20V (Di = 40 nm) 3.0 ± 0.6 35 ± 9 874 ± 204 124 ± 36
25V (Di = 59 nm) 5.3 ± 0.8 43 ± 26 1435 ± 223 231 ± 73
30V (Di = 64 nm) 2.9 ± 0.5 28 ± 4 711 ± 126 75 ± 13
35V (Di = 82 nm) 3.9 ± 0.4 30 ± 2 909 ± 95 100 ± 9
180
The dominant nature of topography is indicated by Sskw. High peaks are implied when
Sskw > 0; whereas Sskw < 0 is indicative of valley-like features such as deep scratches
[210]. The negative Sskw values show that the samples with the nanotubular layer
fabricated under different applied potentials exhibited low valleys rather than high
peaks reflected of their interspace. The nanotubular layer fabricated at 25 V revealed a
higher negative skewness than those nanotubes fabricated at other applied potential.
The sharpness distribution of the sample surface is described by Sku. For normally
distributed surfaces Sku is 3.0. When Sku < 3.0, the distribution is platykurtoic,
indicating a few high peaks and low valleys. When Sku > 3.0, the distribution is
leptokurtoic and indicates many high peaks and low valleys [210]. All the TiO2-ZrO2-
ZrTiO4 nanotubes fabricated at applied potentials from 15 to 35 V exhibited
leptokurtoic structures due to possible influence of their interspace although they were
near mesokurtic distributions when they formed at 10, 30 and 35 V. The nanotubular
layer fabricated at 25 V revealed a higher Sku value, indicating many peaks and valleys
that agree with other roughness parameters such as Sa and Sq in this study. The reason
for these roughness distribution responses relates to a higher physio-chemical activity
inside pores rather than at the end of nanotubes. The increased pore activity arises due
to the availability of anodic current at the higher applied potential while the dissolution
rate of (NH4)2[TiF6] and (NH4)2[ZrF6] remained constant.
7.3.2 Cell adhesion and spreading on TiO2-ZrO2-ZrTiO4 nanotubes fabricated in
aqueous electrolyte
In this part the cell adhesion on the surface of highly ordered nanotubular layer
fabricated at applied potentials from 10 to 35 V with a distribution of different Di, Do
and Wt was investigated. These distributions and interspace, which were related to the
different phase distributions of α and β phases and the needle like microstructure in
Ti50Zr alloy, played a key role in affecting the cellular response. SEM observations of
SaOS2 cells cultured on samples were carried out after 24 h of incubation. The cell
shape and spreading behaviour with formation of cytoplasmic extensions and filopodia
of the SaOS2 cells adhering on top of the various sizes of TiO2-ZrO2-ZrTiO4 nanotubes
were different. Fig. 7.4(e) shows the extension of filopodia, the exploring tools of the
cell, towards neighbouring spaces to find a more suitable place for their further
development. It was found that the cells were more flattened without any particular
181
orientation on the top of nanotubes with 12 < Di < 24 nm, 20 < Di < 40 nm, 28 < Di <
52 nm and 56 < Di < 108 nm than on top of nanotubes with 42 < Di < 76 nm and 41 < Di
< 87 nm, on which they appeared quasi round and non-spreading. The morphology of
the underneath nanotubular layer was visible through the extension part of cells at
higher magnification as shown in Fig. 7.4(f) and 7.5(f). Cell filopodia travelled between
nanotubes and the filopodia ends served as anchor points that allowed probing of the
local environment for proteins so that more stable focal contacts could be established.
Fig. 7.6 illustrates the possible movement of filopodia on nanotubular surface to make
the stable adhesion sites.
182
Fig. 7.4 SEM images of SaOS2 cells cultured after 24 h on three different pore sizes of
TiO2-ZrO2-ZrTiO4 nanotubes fabricated at a) 10 V- low magnification, b) 10 V- higher
magnification, c) 15 V- low magnification, d) 15 V- higher magnification, e) 20 V- low
magnification and f) 20 V higher magnification
183
Fig. 7.5 SEM images of SaOS2 cells cultured after 24 h on three different pore sizes of
TiO2-ZrO2-ZrTiO4 nanotubes fabricated at a) 25 V- low magnification, b) 25 V- higher
magnification, c) 30 V- low magnification, d) 30 V- higher magnification, e) 35 V- low
magnification and f) 35 V higher magnification
184
Fig. 7.6 Schematic illustration of exploring tools of an osteoblast cell on nanotubular
surface
The osteoblast cells which form new bone, and osteoclast cells which degrade bone
similar to other cells (except blood cells), adhere to extra cellular matrix, ECM [264]. In
vitro, within a few seconds, biomaterial surfaces adsorb water, ions and biomolecules
such as proteins from the culture medium. The cell then senses the foreign surface
through ECM and this adsorbed layer. The cell adheres via several phases [137]. The
biomaterial links to cell in the early phase, or attachment phase, by physicochemical
interactions such as ionic forces, van der Waals forces, etc. The later phase, or adhesion
phase, involves different biomolecules such as ECM proteins, cell membrane proteins,
cytoskeleton proteins, cell surface receptors and integrins [15, 137]. Integrins that act as
an interface between the intracellular and extracellular component are heterodimer,
which have different sub-units that bind to specific amino acids. Through this
interaction, they transmit a signal from the ECM to the nucleus to regulate many
cellular functions such as cell adhesion, motility, shape, growth and differentiation [137,
185
264]. Focal contacts are adhesion sites between tissue cultured cells and the substrate
surface that form when integrins are clustered because of interaction with the ECM-
adsorbed biomolecules [15, 264]. An interface of cell/matrix/substrate forms via ECM
protein synthesised by cells [137] when the cytoskeleton is recognised to allow the cell
to spread on the substrate [264]. Fig. 7.7 shows a schematic illustration of cells and
focal contacts which included integrins on the surface of nanotubular layer.
186
Fig. 7.7 Proposed illustration of the position of integrins on a) Di = 18 ± 6 nm best
nanospacing, b) rough surface of Di = 59 ± 17nm and c) Wt = 24 ± 7 nm most existing
of proteins
Regardless of other properties of the surface and the population of proteins, since the
surface occupancy by the head of an integrin is approximately 8 - 12 nm in size, they
187
were better clustered and flattened on nanospacing of 12 < Di < 24 nm and formed
stable focal contacts. Fig. 7.7 (a) and (c) shows how cells behave when the Di of
nanotube increased to adhere themselves on the nanotubular surface. This is in
accordance to the proposed optimum nanospacing (< 30 nm) reported by other
researchers [17, 19]. Although nanotubular layer with 56 < Di < 108 nm was not in the
range of optimum nanospacing, the range of Wt of 17 < Wt < 31 nm where the
population of proteins congregated onto the wall tops of nanotubes due to the large
empty space of big nanotubes on the surface results in the occurrence of the flat
morphology of cells. A higher aggregation and agglomeration of proteins on the surface
of TiO2 nanotubes in comparison to flat Ti (with a TiO2 native oxide layer) has been
reported previously [21]. The cell behavior on the surface of the nanotubular layer of
TiO2-ZrO2-ZrTiO4 with 42 < Di < 76 nm and 41 < Di < 87 nm was also affected by
other properties of the nanotubular surface such as roughness parameters.
For a better understanding of the quantity of spreading of cells on different substrates,
ImageJ software was used to measure the length of filopodia and the surface area of
spreading cells and the results are presented in Fig. 7.8(a) and 7.8(b). The longest
filopodia grew on the nanotube surfaces of mean Di = 40 nm after 1 and 7 days. The
shortest filopodia grew on nanotubes with mean Di = 18 nm after 1 day and mean Di =
30 nm after 7 days. The highest surface area was occupied by cells on the nanotubes
with mean Di = 40 nm after 1 day and mean Di = 59 nm after 7 days. The lowest surface
area was occupied by cells on the nanotubes with mean Di = 64 nm after 1 day and
mean Di = 30 nm after 7 days. The quasi round and non-spreading cells exhibit higher
peaks in profilometry measurement than flattened and spread cell. Therefore, using the
Sp, which is the largest peak height value from the reference surface within the defined
sampling area, the heights of selected cells were measured, as shown in Fig. 7.8(c). The
highest cells grew on the nanotubes with mean Di = 59 nm after 1 day and mean Di =
64 nm after 7 days. The lowest cells grew on the nanotubes with mean Di = 64 nm after
1 day and mean Di = 40 nm after 7 days. These graphs show results in accordance to the
above mentioned behaviour of cells on nanotubular surfaces. The cells exhibited shorter
filopodia, lower surface area and lower cell height on nanotubular surface with desirable
nanosize under 30 nm such as Di where the head of integrins were fitted great and Wt
where sufficient adsorbed protein existed for performing the requiring contacts than
188
nanotubular surface with nanosize upper 30 nm. The exceptions exist when other
surface parameter influenced the cell behaviour.
189
Fig. 7.8 Illustration of a) the length of filopodia, b) surface area and c) height of SaOS2
cells after 1 day and 7 days grown on different nanotube sizes (ρ < 0.05)
0
1
2
3
4
5
6
7
8
9
10
18 30 40 59 64 82
Filo
podi
a le
ngth
(µm
)
Mean of Di (nm)
Length of Filopodia
1 Day
7 Days
a)
-2000
0
2000
4000
6000
8000
10000
12000
18 30 40 59 64 82
Surf
ace
Are
a (µ
m2 )
Mean of Di (nm)
Surface area of elangated cells
1 Day
7 Days
b)
-10
0
10
20
30
40
50
60
18 30 40 59 64 81
Sp (µ
m)
Mean of Di (nm)
Heigth of Cells
1 Day
7 Days
c)
190
SaOS2 cells on nanotubular layer with 28 < Di < 52 nm and 56 < Di < 108 nm were
spread wider with lamellipods than on nanotubular layer with 12 < Di < 24 nm and 20 <
Di < 40 nm; thereby demonstrating a high level of cell migration. SaOS2 cells on
nanotubular layers of 42 < Di < 76 nm and 41 < Di < 87 nm were less spread than the
above mentioned nanotube size and exhibited dorsal ruffles; thereby demonstrating a
high cellular activity reflecting in their height (Sp) but they were relatively sparse on the
surface after 7 days. The SEM images of the cell growth after 7 days as shown in Fig.
7.9 and 7.10 indicate the effect of nanospacing and surface chemistry on cell behaviour.
TiO2-ZrO2-ZrTiO4 nanotubes with 41 < Di < 87 nm were lower in surface energy (Table
6.2) in comparison to other nanotubular layer with Di > 30 nm, which might influence
protein adsorption and their rearrangement on the nanotubular layer [15]. They also
demonstrated lower spacing roughness parameters than other nanotubular layers in this
study, which would potentially disrupt the formation of focal adhesion.
191
Fig. 7.9 SEM images of SaOS2 cells cultured for 7 days on the four different pore sizes
of TiO2-ZrO2-ZrTiO4 nanotubes fabricated at a) 10 V- low magnification, b) 10 V-
higher magnification, c) 15 V- low magnification, d) 15 V- higher magnification, e) 20
V- low magnification and f) 20 V higher magnification
192
Fig. 7.10 SEM images of SaOS2 cells cultured for 7 days on the four different pore
sizes of TiO2-ZrO2-ZrTiO4 nanotubes fabricated at a) 25 V- low magnification, b) 25 V-
higher magnification, c) 30 V- low magnification, d) 30 V- higher magnification, e) 35
V- low magnification and f) 35 V higher magnification
The percentage of attached cells on six surface morphologies and the percentage of total
cells were obtained using MTS assay and are presented in Fig. 7.11. Among the six
nanotubular layers, the highest percentage of attached cells was 65.2 ± 9.4 % on the
nanotubes with mean Di of 18 nm. The percentage of cells attached to nanotubes with
193
higher mean Di of 30 and 40 nm decreased to 64.4 ± 5.7 % and 41.0 ± 2.6 %.
Considering previous work [17] it was expected that the percentage of attached cells
decreases with an increase in the distribution of Di of the nanotubes. In terms of
nanotubular layers with 42 < Di < 76 nm, which revealed lower cell attachment than
other, the roughness of the surface might also play an important role. It has been
reported that when the height and spacing between peaks of a surface increased, the
cells were not able to adhere as they adhere on the top of peaks of smoother surface
[137, 264]. This can be the reason for the lowest cell attachment on the nanotubular
surface with 42 < Di < 76 nm due to its highest Sa. The percentage of cells attached to
nanotubes with higher mean Di of 64 and 82 nm increased to 33.1 ± 2.2 % and 33.5 ±
4.1 % but they were still less than that on nanotubes with smaller mean Di. This result
agrees with the hypothesis that the human osteoblasts spread more intimately on lower
roughness amplitude surfaces than on rough ones when the topography is on the
nanoscale [262]. Two reasons may describe the lower cell attachment on nanotubular
surface with 41 < Di < 87 nm than 56 < Di < 108 nm: its lower Sm and surface energy
although the Sa did not reveal a large difference. In the contrary of expectation, the
nanotubular layer with 56 < Di < 108 nm did not reveal the lowest cell attachment
because of the large space of Di which prevents anchoring of the head of integrins. The
Wt of this nanotubular layer was 17 < Wt < 31 nm which was in the range of proposed
optimum nanospacing where the proteins can be adsorbed by the fluid. The mean
differences obtained by a one-way ANOVA test were significant at the level of 0.05
between Di of 40 ± 12 and 59 ± 17 nm. This result indicates the synergetic effects of a
variety of properties of the nanotubular surface as well as the important morphological
metric of the optimal nanospacing.
194
Fig. 7.11 The percentage of cells attached to the surface of different pore sizes of TiO2-
ZrO2-ZrTiO4 nanotubes after 7 days of cell seeding using MTS assay (ρ < 0.05)
The ratio of length to width (L/W) of cell nucleus and actin stress fibre on nanotubular
surface was measured by ImageJ software using confocal microscopy images after 7
days (Fig. 7.12). When the Di of TiO2-ZrO2-ZrTiO4 nanotubes increased from 18 to 30
and 40 nm, the L/W ratio of nuclei (~ 1.5) did not change considerably. It increased on
the nanotubular surface with Di = 59 nm then decreased to ~ 1.1 on Di = 64 and 82 nm.
The trend of the L/W ratios for the actin stress fibres was similar (~ 2.0) with the
maximum of 2.9 on Di = 59 nm. This result again confirms the before mentioned cell
behaviour against changing Di and roughness parameters of nanotubular layer.
0
20
40
60
80
100
120
140
18 30 40 59 64 82
Perc
enta
ge o
f cel
l adh
esio
n (%
)
Mean Di (nm)
Cell Adhesion
Total Cell
195
Fig. 7.12 Confocal images of stained SaOS2 cells presenting nuclei (blue) and actin
stress fibres (red) after 7 days on TiO2-ZrO2-ZrTiO4 nanotubes with a) Di = 18 nm, b)
Di =30 nm, c) Di =40 nm, d) Di = 59 nm, e) Di = 64 nm and f) Di = 82 nm
Figures 7.13 (a) and (b) show the SEM images of SaOS2 cells adhered on the surface of
annealed TiO2-ZrO2-ZrTiO4 nanotubes (Di = 40 nm) which were cleaned for two weeks
in 70 °C pure water while the pure water replaced three times per day. This cleaning
process was undertaken to minimise the remaining contamination of the electrolyte
during anodisation. The cells were flattened on the surface and they did not exhibit any
particular orientation. Fig. 7.14 shows the length of filopodia and the surface area of
spreading cells which were measured using SEM images by ImageJ software. On the
cleaner surface the filopodia did not elongated as much as those on the nanotubular
surface before cleaning after 1 day and 7 days. The surface area which occupied by cells
on the cleaner nanotubular surface was less than those on the nanotubular surface before
cleaning after 1 day and 7 days. The height of grown cells on cleaner surface was less
than the height of grown cells on the nanotubular surface before cleaning after 1 day
196
and 7 days. These result shows that the cleaner surface offered an acceptable condition
for cells to adhere and grow without much attempt to migrate to find better surface.
Fig. 7.13 SEM images of SaOS2 cells cultured for 24 h on the TiO2-ZrO2-ZrTiO4
nanotubes (Di = 40 nm) a) with reduced contamination - low magnification, b) with
reduced contamination - higher magnification, c) with post-treated by HA - low
magnification and d) with post-treated by HA - higher magnification
The nanotubular surface which was covered by HA after soaking in m-SBF for three
weeks revealed a convenient condition for SaOS2 cells to adhere. The length of
filopodia on HA surface was shorter than those on nanotubular surface without HA. The
cells did not expand on HA surface as much as they expand on nanotubular surface
before soaking in m-SBF. The grown cells on HA surface showed higher height than the
grown cells on the nanotubular surface without HA after 1 day and 7 days because of
the nature and height of HA. It can be concluded that changing the chemistry of the
197
surface served the cell to form better focal contact and decreasing the intention to
migrate.
198
Fig. 7.14 Illustration of a) the length of filopodia, b) surface area and c) height of
SaOS2 cells after 1 day and 7 days grown on different post-treated nanotubular surface
with the same Di
0
1
2
3
4
5
6
7
8
9
10
40 Cleaner Surface HA
Filo
podi
a le
ngth
(µm
)
Mean of Di (nm)
Length of Filopodia
1 Day
7 Days
a)
0
200
400
600
800
1000
1200
40 Cleaner Surface HA
Surf
ace
Are
a (µ
m2 )
Mean of Di (nm)
Surface area of elangated cells
1 Day
7 Days
b)
0
1
2
3
4
5
6
7
8
9
40 Cleaner Surface HA
Sp (µ
m)
Mean of Di (nm)
Heigth of Cells
1 Day
7 Days
c)
199
Fig. 7.15 illustrates the percentage of attached cells on nanotubular surface which was
cleaned in 70 °C pure water for two weeks to reduce the surface contamination and
nanotubular surface which was coated with HA by soaking in m-SBF in comparison to
nanotubular layer without these post treatment with same Di and the percentage of total
cells were obtained using MTS assay. The percentage of cells attached to nanotubes
with cleaner surface was 78.4 ± 7.7 % which was (almost two times) higher than the
nanotubular surface before cleaning (41.0 ± 2.6 %). The percentage of cells attached to
nanotubes covered with HA increased to 59.7 ± 13.6 %.
Fig. 7.15 The percentage of cells attached to the different post-treated nanotubular
surface of TiO2-ZrO2-ZrTiO4 with the same Di after 7 days of cell seeding using MTS
assay
It can be seen that the result obtained here was similar to the previous studies which
indicated that osteoblast proliferation was significantly greater on nanophase HA than
on conventional formulation of the same ceramic [270]. It has also been reported that
the growth of MG63 osteoblastic cells on electrodeposited CaP surface was more
significant than untreated titanium surface in a direct comparison [271]. Denser and
smaller crystalline size of HA on TiO2 nanotubes increased adhesion of MG63
0
20
40
60
80
100
120
140
40 Clean Surface HA
Perc
enta
ge o
f cel
l adh
esio
n (%
)
Mean Di (nm)
Cell Adhesion
Total Cell
200
osteoblastic cells due to increase in protein activity of such a layer [25]. Figures 7.16
and 7.17 show the SEM and confocal images of cells which were grown on the surface
of post-treated TiO2-ZrO2-ZrTiO4 nanotubes after 7 days, respectively.
Fig. 7.16 SEM images of SaOS2 cells cultured for 7 days on the TiO2-ZrO2-ZrTiO4
nanotubes (Di = 40 nm) a) with reduced contamination - low magnification, b) with
reduced contamination - higher magnification, c) with post-treated by HA - low
magnification and d) with post-treated by HA - higher magnification
201
Fig. 7.17 Confocal images of stained SaOS2 cells presenting nuclei (blue) and actin
stress fibres (red) after 7 days on TiO2-ZrO2-ZrTiO4 nanotubes a) with Di = 40 nm
before post-treating b) reduced contamination, c) with post-treated by HA
The ratio of length to width (L/W) of cell nucleus and actin stress fibre on nanotubular
surface was measured by ImageJ software using confocal microscopy images (Fig.
7.17) after 7 days. On cleaner surface of TiO2-ZrO2-ZrTiO4 nanotubes and covered with
HA, the L/W ratio of nuclei (~ 1.5) did not change considerably, however the L/W
ratios of the actin stress fibres increased to ~ 3.2 and ~ 2.2 on the cleaner nanotubular
surface and covered with HA, respectively. These results again interpret the before
mentioned behaviour of cells against changing of chemistry of nanotubular layer.
7.3.3 Dimensional, physical and chemical properties of TiO2-ZrO2-ZrTiO4
nanotubes fabricated in non-aqueous electrolyte
Figure 6.12 in Chapter 6 shows SEM images of top view of annealed TiO2-ZrO2-ZrTiO4
nanotubues fabricated in ethylene glycol with 5 to 10 wt % water (H2O) content and the
applied potential from 20 V to 30 V. As the nanotubular surface fabricated under non-
aqueous electrolyte conditions exhibited a brittle layer after annealing and it was not
suitable for further investigation, the cracked upper layer was removed by ultrasound
and the remaining dimpled surface used for cell culture. The mean sizes of Di, Do and
Wt of the dimpled surface of nanotubes is presented in Fig. 7.18. The Di, Do and Wt of
the surface with a dimpled surface increased when the water content of the organic
electrolyte increased from 5 to 10 wt %. Increasing the applied potential from 20 V to
202
30 V resulted in an increase in Di, Do and Wt for each electrolyte. The dimpled surface
of the nanotubular TiO2-ZrO2-ZrTiO4 layer was of larger nanoscale, revealing the bottle
shape of nanotubes that were larger at the bottom with a slightly thicker Wt.
Fig. 7.18 Mean nanotube size of Di, Do and Wt of dimpled surface after annealing and
removing the upper cracked layer fabricated in two different water contents and two
different applied potentials in ethylene glycol, Note: The nanotube size distribution
graphs were generated from 100 nanotubes on different positions for each of three
samples (300 measurements)
Figure 6.13 in Chapter 6 shows roughness parameters of Sa and Sq measured for the
annealed nanotubular layer with a dimpled surface. These amplitude parameters (Sa and
Sq) decreased with an increase of water content for each applied potential. Increasing
the applied potential for each composition of electrolyte resulted in increasing Sa and Sq.
It can be observed that the majority of the 4 dimpled surfaces, after removal of the
upper cracked layer of nanotubes, exhibited a uniform symmetry with normal
distribution in surface heights since Sskw ≈ 0 and Sku ≈ 3. Table 7.2 details the surface
area index (SI), the spacing parameters of Sm, Sc and Sv and surface energy of dimpled
surface of TiO2-ZrO2-ZrTiO4 nanotubes. The hydrophilic property of nanotubes also
0
20
40
60
80
100
120
140
5%-20 10%-20 5%-30 10%-30
Nan
otub
e si
ze (n
m)
Water content (wt%)and applied potential (V)
Wt-Annealed-Dimple
Di-Annealed-Dimple
Do-Annealed-Dimple
203
increased with an increase of the nanotube sizes and roughness parameters that were the
result of increasing the water content and applied potential.
204
Table 7.2 Surface index, spacing parameters and calculated surface energy of TiO2-ZrO2-ZrTiO4 nanotubes with dimpled surface fabricated in
different water contents and applied potentials, Note: Each data point is an average of five measurements
Anodisation Condition (Mean size) SI Sm
(nm3/nm2)
Sc
(nm3/nm2)
Sv
(nm3/nm2)
Contact angle
()
sd
(mJm-2)
sp
(mJm-2)
s = sd + s
p
(mJm-2)
5 wt% H2O-20V (Di = 25 nm) 1.5 ± 0.1 62 ± 45 1074 ± 170 81 ± 16 9.0 ± 0.2 13.37 59.91 73.28
10 wt% H2O-20V (Di = 29 nm) 1.7 ± 0.0 33 ± 7 867 ± 62 67 ± 2 7.8 ± 0.4 15.16 57.77 72.93
5 wt% H2O-30V (Di = 45 nm) 1.7 ± 0.1 57 ± 28 1581 ± 682 110 ± 40 7.0 ± 0.7 13.85 59.64 73.49
10 wt% H2O-30V (Di = 49 nm) 1.7 ± 0.0 44 ± 6 1055 ± 95 111 ± 35 8.0 ± 0.2 15.20 57.69 72.89
205
7.3.4 Cell adhesion and spreading on TiO2-ZrO2-ZrTiO4 nanotubes fabricated in
non-aqueous electrolyte
The dimpled surfaces that were revealed after removing the upper cracked layer of
annealed TiO2-ZrO2-ZrTiO4 nanotubes were investigated for their induction of adherent
SaOS2 cells. Fig. 7.19 shows the formation of SaOS2 cell cytoplasmic extensions that
were widely spread with the extension of sheet-like protrusions, known as lamellipodia,
on all of the four dimpled nanotubular surfaces. The SaOS2 cells did not fabricate
finger-like protrusions of the plasma membrane, known as filopodia, to explore
neighboring spaces that might be more suitable locations for attachment and spreading.
Fig. 7.19 SEM images of SaOS2 cells cultured for 7 days on the four different pore
sizes of TiO2-ZrO2-ZrTiO4 nanotubes fabricated at a) 5 wt % H2O - 20 V, b) 10 wt %
H2O - 20 V, c) 5 wt % H2O - 30 V, d) 10 wt % H2O - 30 V
206
The quantity and growth of cell spreading on the dimpled surfaces was quantitatively
assessed by application of ImageJ software. The ratio of length to width (L/W) of cell
nuclei and actin stress fibre (cytoplasmic protein) which is one of three types of filament
that physically joins the cell membrane and nuclear membrane and controls the
architecture and movement of cells [264, 272], was determined by using confocal
microscopy as well as the surface area of the actin stress fibre (SAS.F) after culture for 7
days. When the Di of the dimpled surface increased from 25 to 29 and 45 nm, the L/W
ratio of nuclei (~ 1.7) did not change considerably. The trend of the L/W ratios for the
actin stress fibres was similar (~ 3.1). The highest percentage of the surface area
measurements for the actin stress fibres were 646 < SAS.F. < 1230 µm2 on a dimpled
surface of Di = 25 nm. These values decreased to 562 < SAS.F. < 862 µm2 for Di =
29 nm. At Di = 45 nm for the dimpled surface, the highest percentage of the surface area
measurements of the actin stress fibre was 634 < SAS.F. < 948 µm2. This decreased to
330 < SAS.F. < 891 µm2 for the nanotubular surface with Di = 49 nm.
207
Fig. 7.20 Confocal images of stained SaOS2 cells presenting nuclei (blue) and actin
stress fibres (red) after 7 days on dimpled surface with a) Di = 25 nm, b) Di =29 nm, c)
Di =45 nm and Di =49 nm
Table 7.3 presents the percentage of attached cells and total cell numbers, as measured
by the MTS assay, on the four surface morphologies of different Di. The highest
percentage of attached cells was 84.9 ± 7.4 % on the nanotubes with mean Di of 25 nm
and the lowest one was 61.0 ± 7.1 % on the nanotubes with mean Di of 29 nm. The
percentage of cells attached to nanotubes with higher mean Di of 45 and 49 nm
increased to 61.5 ± 8.1 % and 80.3 ± 4.4 %, respectively; but these were still less than
that on the nanotubes with the lesser mean Di of 25 nm.
208
Table 7.3 The percentage of cells attached to the surface of different pore sizes and
roughness parameters of TiO2-ZrO2-ZrTiO4 nanotubes fabricated in organic electrolyte
after 7 days of cell seeding using MTS assay (ρ < 0.05)
Anodisation
conditions Cell adhesion (%) Di (nm) Wt (nm) Sa (µm) Sm (nm3/nm2)
5 wt%H2O-20V 84.9 ± 7.4 25 ± 5 27 ± 2 0.72 ± 0.12 62 ± 45
10 wt%H2O-20V 61.0 ± 7.1 29 ± 4 30 ± 2 0.55 ± 0.01 33 ± 7
5 wt% H2O-30V 61.5 ± 8.1 45 ± 6 24 ± 2 1.03 ± 0.28 57 ± 28
10 wt%H2O-30V 80.3 ± 4.4 49 ± 8 34 ± 3 0.73 ± 0.04 44 ± 6
209
The biological fluid functionalises the biomaterial implants in vivo, or the proteins
available in the culture media from the serum functionalise the biomaterial implants in
vitro, by providing protein and water that can be adsorbed onto the substrate. Therefore,
cells are now located in an environment that facilitates attachment since this is not
possible for the as-received, “naked” surface of biomaterial implant [15]. Extracellular
matrix (ECM), which is an intricate arrangement of glycoproteins, collagens,
proteoglycans and growth factors, has different roles. ECM acts as a physical scaffold
for cell attachment and the organisation of cellular structures. The ECM not only
provides a ligand for integrins expressed on the cell surface but also dictates the type of
integrin found within focal adhesions to which cells adhere [262, 273].
The most common biological molecules that can act as cell attachment molecules are
the integrins. These heterodimeric proteins are constituted of two subunits, α and β
[157]. There would be several phases for attachment when a cell studded with microvilli
perches on a surface. It has been reported that adhesion was not sufficient for survival;
that is, cells also need to spread on a surface to survive [274]. In a fraction of a second
the cell flattens due to non-specific electrostatic forces, such as van der Waals, and there
is the passive formation of ligand-receptor bonds. Then the cell aligns and the contact
area is widened [17, 259, 274]. Later, after several hours, the spreading phase proceeds
depending on the cell metabolism and the occurrence of clustering of integrins at the
anchoring sites. Focal adhesions are adhesion sites within 10 and 15 nm distance from
the substrate where they are formed by clustered receptors (integrins) using cytoplasmic
proteins [157, 264]. The head of integrins is approximately 8 - 12 nm. Thus, their
clustering has been proposed and proved to be enhanced when the optimum
nanospacing is less than 30 nm [19]. Therefore, they were initially clustered aptly on the
dimpled surface of 20 < Di < 30 nm; then on the dimpled surface of 31 < Wt < 37 nm,
due to an optimum nanospacing and similar roughnesses. However, the condition for
clustering of integrins on the dimpled surface of 25 < Di < 33 nm was not provided, as
might be expected due to optimal nanospacing, because of its low roughness spacing
parameter that could potentially disrupt the formation of focal adhesion. In the case of
the dimpled surface of 22 < Wt < 26 nm, the high roughness parameter of Sa = 1.03 µm
prevents formation of clusters.
210
These results agree with the fact that long-term adhesion is correlated to topography.
The hybrid roughness parameters, as measured in the current work, are especially
important since they affect the synthesis of ECM proteins by the cell themselves and the
cell-cell contacts. These phenomenological factors represent the strength of the cell-
material interface [264]. The mean differences obtained by a one-way ANOVA test
were significant at the 0.05 level between Di of 25 ± 5, 49 ± 8 nm and Di of 29 ± 4, 45 ±
6 nm. This result indicates the synergic effects of other properties of the dimpled
surface as well as the important morphological metric of the optimal nanospacing.
7.4 Conclusions
The fabrication of nanotubular layers on the surface of Ti50Zr under four different
electrochemical conditions was investigated. The influence of the characteristics of the
nanotubular surfaces on the cell behaviour of osteoblasts was evaluated. Conclusions
are as follows:
1. TiO2-ZrO2-ZrTiO4 nanotubes with six different distributions of inner
diameter (Di), outer diameter (Do) and wall thickness (Wt) were fabricated
via anodisation under different applied potentials in the aqueous electrolyte.
The annealed nanotubes exhibited a mean inner diameter (Di) of 18 ± 6, 30 ±
10, 40 ± 12, 59 ± 17, 64 ± 23 and 82 ± 26 nm when anodised at applied
potentials of 10, 15, 20, 25, 30 and 35 V, respectively.
2. The TiO2-ZrO2-ZrTiO4 nanotubes with Di = 59 ± 17 nm showed a roughness
with higher amplitude parameters of Sa and Sq, in comparison with
nanotubular surfaces that were fabricated at 10,15, 20, 35 and 35 V. These
parameters decreased when the applied potential increased to 30 and 35 V
because of the changes in the oxidation and dissolution rates.
3. The lowest density of SaOS2 cells occurred on the nanotubular surface with
59 ± 17 nm because of its high roughness amplitude parameters. Then next
lowest density evolved on the nanotubular surface with 64 ± 23 nm because
of its lowest surface energy and lowest roughness spacing parameters.
4. The highest density of SaOS2 cells was on the nanotubular surface with 18 ±
6 nm and the second highest density appeared on the nanotubular surface
with 30 ± 10 nm because of the optimum nanospacing.
211
5. SaOS2 cells exhibited the longest filopodia on the nanotubular surface with
Di = 40 ± 12 nm after culture for 1 and 7 days and the lowest height after
culture for 7 days. SaOS2 cells revealed the shortest filopodia and lowest
surface area and lowest height on the nanotubular surface with Di = 64 ±
23 nm after culture for 1 day and highest height of cell growth after culture
for 7 days.
6. The percentage of cells attached to nanotubes with cleaner nanotubular
surface and covered with HA increased to 78.4 ± 7.7 % and 59.7 ± 13.6 %
which was promising in comparison to the nanotubular surface before these
post-treatment (41.0 ± 2.6 %).
7. The upper nanotubular layer fabricated in ethylene glycol with 5 and 10 wt
% of water at 20 and 30 V after annealing was brittle and cracked. Therefore
it was removed to reveal a dimpled surface for bone cell adhesion study. The
nanotubes on the dimpled surface exhibited larger sizes of Di, Do and Wt than
the upper layer. The mean Di of nanotubes under the condition of 5 and 10
wt % of water anodised at 20 and 30 V were 25 ± 5, 29 ± 4, 45 ± 6 and 49 ±
8 nm, respectively.
8. The mean Sa and Sq of the dimpled surface decreased with an increase in
water content and increased with an increase in applied potential during
anodisation.
9. It was found that the optimal nanospacing was either Di = 25 ± 5 nm or Wt =
34 ± 3 nm for a maximum adhesion of bone cells (84.9 % and 80.3 %
respectively).
10. A lower percentage of cell attachment was observed on the dimpled surface
with the higher roughness of Sa = 1.03 µm (~ 61.5 %) and lower space
roughness of Sm = 33 nm3/nm2 (~ 61.0 %).
11. The surface area of actin stress fibres of SaOS2 cells was higher on optimal
nanospacing surfaces for the nanotubular surfaces with Di = 25 ± 5 nm.
212
Chapter 8 Nanoporous and nanotubular metal oxide layers on
biocompatible metals of Ta, Nb and Zr and their potential
applications
Abstract
Biocompatible metals such as titanium (Ti), zirconium (Zr), niobium (Nb) and tantalum
(Ta) that confer a stable oxide layer on their surfaces are commonly used as implant
materials or alloying elements for titanium-based implants due to their exceptional high
corrosion resistance and excellent biocompatibility. These metal oxide layers are highly
protective and stable and their surface properties can be tuned through creating a
nanoporous or a nanotubular layer with controlled topographical characteristics at the
nanometer scale using electrochemical surface modification. In this chapter, four kinds
of metal oxide nanostructures of tantala (Ta2O5), zirconia (ZrO2), titania (TiO2) and
niobia (Nb2O5) were fabricated via anodisation. The characteristics of these metal
oxides at nanoscale were characterised and their potential applications were discussed.
The tantala (Ta2O5) and the niobia (Nb2O5) exhibited a nanoporous structure and
zirconia (ZrO2) and titania (TiO2) a nanotubular structure. The nanosize distribution,
topography and both the physical and chemical properties of the nanolayers and their
bioactivities, as measured by the formation of hydroxyapatite (HA), were investigated
using scanning electron microscope combined with energy dispersive spectroscopy
(SEM-EDS), X-ray diffraction (XRD) analysis and 3D-Profilometry. The surface
energy of the nanoporous and nanotubular surfaces and its influence on the
hydroxyapatite formation in simulated body fluid (SBF) were evaluated. Observations
indicated that a Ta2O5 nanotube fabricated on the surface of tantalum were present for
several seconds prior to transforming to a nanoporous layer. The nanoporous Ta2O5
exhibited an irregular porous structure, high roughness and high surface energy as
compared to bare tantalum metal; and the highest bioactivity after annealing among the
four kinds of nanoporous structures. The amorphous nanoporous Nb2O5 showed a
uniform porous structure and low roughness, but no bioactivity before annealing.
Hydroxyapatite formation occurred only on the annealed nanoporous Nb2O5 with a
base-centred monoclinic structure. Fabrication of a uniform layer of ZrO2 nanotubes
required a higher applied potential as compared to that for TiO2 nanotubes with the
same inner diameters. Overall, the nanoporous and nanotubular layers of Ta2O5, Nb2O5,
213
ZrO2 and TiO2 imparted new functionality to the bare metal surfaces in the form of
enhanced bioactivity which is promising for their application as metal implants.
8.1 Introduction
Metals and alloys that are used as biomaterials also have a range of applications in
industry and medicine due to their excellent mechanical, physical and chemical
properties. Extensive studies have been carried out on titanium that has been coated by
its natural oxide layer formed in air. Other biocompatible metals and their respective
oxide layers are also of interest. For example, the protective oxide layer that forms
naturally on a tantalum surface or that is fabricated on the surface of other metals by
surface treatments such as chemical vapour deposition [275], electro deposition [123,
124] and sol gel [276] methods have other applications. Tantalum pentoxide, Ta2O5, is
used as a protective coating for chemical equipment due to its excellent corrosion
resistance [275]. A thin film of Ta2O5, formed by radio frequency (RF) sputtering of
SiO2-Ta2O5 films, has been used in optical devices such as transmission media for
integrated optical circuits [277]. Ta2O5 has demonstrated modest dielectric properties
with an ability to form a high quality thin film [278] for the miniaturisation of the
capacitive component of memory devices. The biocompatibility of the tantalum
pentoxide layer suggests tantalum as a good candidate for bioengineering implant
applications [116]. However, tantalum metal is more dense (ρTa = 16.6 g.cm-3) [114]
and more expensive than titanium (ρTi = 4.5 g.cm-3) [115]; therefore, researchers have
taken an alternative processing route to form a thin layer of tantalum onto the surface of
titanium and other metal implants. By anodisation a nanoporous Ta2O5 can be formed
on the metal in an acidic electrolyte containing HF [104, 117, 279]. The organic
electrolytes were also used for anodising a nanoporous oxide layer at an applied
potential of 10 to 40 V [121, 122].
The application of niobium pentoxide (Nb2O5), which can be produced via different
methods such as reactive sputtering [126], sol-gel processes [127], templating
techniques [128] and anodic oxidation [51, 129], has been reported in applications for
gas sensors [280], catalysts [281], both optical and electrochromic devices [126] and
biocompatible prostheses [127]. Niobium pentoxide (Nb2O5) is also a N type
semiconducting oxide with an electrical conductance in the presence of oxidising or
reducing gases that is reversible; thus, it has been investigated as an oxygen and
214
ammonia sensor material [280]. The niobium pentoxide is distinguished with
characteristics of displaying promoter and supporter effects in Fischer-Tropsch
synthesis, acidic and redox properties, and photosensitivity. The layered structure of
Nb2O5 can act as a catalyst under different chemical reactions due to its surface
properties and structure [281]. As well, the optical transmission of Nb2O5 changes
rapidly and reversibly from the near UV to near IR when H+ and Li+ ions are placed in
its layer lattice and, therefore, enable Nb2O5 to be used in solid state electrochemical
devices [282]. There are studies on the formation of a nanoporous layer of Nb2O5 by
controlling the effects of mixed electrolytes, applied potential and anodisation time.
These nanoporous layers displayed different range of pore sizes and thickness formed
mostly in an acidic electrolyte [51, 130, 283].
An important application for zirconium dioxide (ZrO2) is as an industrial catalyst,
especially with its use as an acid catalyst and for NOx reduction because it is more
stable under hydrothermal conditions in comparison to zeolites. This metal oxide also
exhibits large pores and a flexible mix in metal oxide composition [224, 225]. The
electrochemical deposition of a zirconium thin film on the surface of stainless steel has
been reported for these applications [284]. Zirconia, ZrO2, can have optoelectronic and
biomedical applications due to its high mechanical, chemical, and thermal stability
[226]. The investigation of optoelectronic applications of ZrO2 nanotubes has proceeded
via photoluminescence and cathodoluminescence measurements and the as-forme ZrO2
nanotubes were crystalline and were observed to demonstrate bright luminescence
[226]. The formation of self-organised sponge-like porous ZrO2 [53] and nanotube
oxide layer with different nanoscale sizes and thicknesses [132] has been described in
detail along with the effect of changing the condition of electrolyte, applied potential
and time of anodisation [102, 132-135].
The composition of the substantial mineral form of hard tissue such as bone and dentin
is apatite, which is secreted by bone cells on the implant surface during the process of
attachment. The bioactive materials which contained Si-OH, Ti-OH, Zr-OH, Nb-OH
and Ta-OH have been reported to exhibit the ability to induce apatite formation [37].
The effect of Ta-OH groups on the surface of alkaline treated bioactive tantalum metal
to form bone-like apatite was investigated and the Ca/P ratio of 1.59 was reported for
crystalline apatite after immersing in SBF [285]. The nanoporous niobium oxides which
215
prepared by sol-gel method and coated on 316LSS enhanced hydroxyapatite formation
[286]. Nanosheets and nanofibre-like surface formed on niobium surface by
hydrothermal alkaline treatment significantly influenced its apatite inducing ability
[287]. Amorphous nanoporous bioactive sodium niobate hydrogel layer formed on the
surface of niobium by alkaline treatment induced the deposition of a CaP layer during
soaking in SBF [288]. As-formed ZrO2 nanotubes have been reported to induce apatite
formation [289] and annealing improved its bioactivity [290]. Using pre-treatment such
as effective dipping treatment improved bioactivity of as-formed ZrO2 nanotube with
diameter 35 to 80 nm [291]. Although a few studies have investigated the bioactivity of
ZrO2 nanotubes, there is still a lack of research in this aspect for nanoporous Nb2O5 and
Ta2O5 which is the object of this chapter.
In this chapter the size of the pores of the as-formed and annealed Ta2O5, Nb2O5 and
ZrO2 nanoporous and nanotubular layer fabricated on the surfaces of tantalum, niobium
and zirconium and different electrolyte compositions of other studies is presented. Their
hydrophilic properties and surface energies are investigated and linked to changes of
their roughness and bioactivities. These properties are important for the biomedical
applications of these biocompatible metals. However there are insufficient studies in
this aspect, which is a prime motivation for the current work.
8.2 Materials and methods
Tantalum (10×10×0.1 mm), niobium (10×10×0.05 mm), zirconium (10×10×0.05 mm)
and titanium (10×10×0.05 mm) foils (Baoji Boxin Metal Materials Co. Ltd., Shaanxi,
China) were degreased by sonification in methanol, isopropanol, acetone and ethanol
for 15 min each progressively. The substrates were then washed with deionised water
and dried using a nitrogen gas stream. A two-electrode configuration was employed
using a DC power supply with a 1 cm2 platinum plate acting as the counter electrode
and placed 4 cm from the working electrode. Electrochemical experiments were
performed at room temperature with the electrolyte composed of 1 M H2SO4 +
3.3 wt % NH4F for tantalum and niobium at 20 V for 120 min and 16 min, respectively;
1 M (NH4)2SO4 + 0.3 wt % NH4F (with the addition of H2SO4 to attain pH 5) for
zirconium at 30 V for 95 min and 1M H2SO4 + 0.5 wt % NH4F for titanium at 10, 15
and 20V for 120 min. The electrolytes were prepared from reagent grade chemicals
(Sigma Aldridge) and pure water. After the electrochemical treatment the samples were
216
rinsed with deionised water for 5 min and dried with a nitrogen stream. Annealing of
the samples was performed in air at 290 °C for 10 min in a conventional muffle furnace
(Nabertherm LT15/13/P330; Nabertherm GmbH, Lilienthal, Germany).
Microstructural characterisation of the samples was carried out using a field-emission
scanning electron microscope (FESEM, ZEISS SUPRA 40 VP). Phase characterisation
was performed by means of X-ray Diffraction (XRD, Bruker D8; Bruker Pte Ltd,
Singapore) using Cu Kα incident radiation at 40 kV and 40 mA with a 2θ scanning step
size of 0.02 ° from 10 to 90 °.
Roughness parameters were measured using a 3D-Profilometer (Bruker, Contour GT-
K1; Bruker Pte Ltd, Singapore) and analysed using the SurfVision software (Veeco
Instruments Inc.; Plainview, NY, USA). Mean Roughness (Sa) and Root Mean Square
(RMS) Roughness (Sq) were studied along with some other amplitude parameters
(skewness (Sskw) and kurtosis (Sku)). These parameters were described in Chapter 3.
The water contact angle was measured using a goniometer (NRL C.A. Goniometer,
Ramé-hart, Inc.; Succasunna, NJ-USA). Surface energy was calculated based on
Owens-Wendt (OW) method [211], as described in Chapter 3.
Bioactivity assessments were carried out by soaking the as-formed and annealed metal
oxides samples in simulated body fluid (SBF) and incubating at 37 ˚C for up to 3 weeks.
A modified (m-SBF) [213] with an ion composition nearly equal to blood plasma was
prepared by dissolving 5.403 g NaCl, 0.504 g NaHCO3, 0.426 g Na2CO3, 0.225 g KCl,
0.230 g K2HPO4 - 3H2O, 0.311 g MgCl2 - 6H2O, 0.293 g CaCl2, and 0.072 g Na2SO4 in
pure water as previously reported [213]. The m-SBF was buffered at pH 7.4 at 37 ˚C
using 2-(4-(2-hydroxyethyl)-1-piperazinyl) ethane sulfonic acid (aka HEPES) and 1 M
NaOH. An amount of 17.892 g HEPES has dissolved in 100 ml 0.2 M NaOH. The
ability of the pretreated samples to form apatite was evaluated in a static SBF
environment. The samples were removed after incubation for 3 weeks in m-SBF; then
rinsed with deionised water and dried at room temperature for 24 h. The atomic
percentage of calcium to phosphate ratio is calculated using EDS results.
The nano size distribution measurements were generated from 100 nano-pores and
nanotubes at different positions for each sample. An average of five readings per sample
217
was acquired for the roughness parameters, water contact and surface energy
measurements.
8.3 Results and discussion
8.3.1 Process conditions and formation mechanism of nanoporous/nanotubular
metal oxides
The application of a suitable voltage to a metal surface produces a Mn+ ion that can
dissolve in an electrolyte or form an insoluble metal oxide layer by reacting with
existing O2- and forming an oxide layer on the metal surface. However, partial
dissolution is also possible due to the composition and condition of the electrolyte [11]
and this processing conditions needs to be controlled since this influences the formation
of the nanolayers of metal oxides.
The oxidation reaction at the interface of the metal surface and electrolyte at a constant
applied potential is given by:
M Mn+ + ne- (M = Ta, Nb, Zr and Ti) (8-1)
As well:
xH2O xO2- + 2xH+ (8-2)
Then,
2Mn+ + nO2- M2On (8-3)
The reduction reaction at the surface of counter electrode (normally platinum) in this
condition is given by:
2nH+ + 2ne- nH2 (8-4)
In the presence of a fluorine anion, chemical dissolution occurs, i.e.:
Ta5+ + 7F-[TaF7]2- (8-5)
Nb5+ + 7F-[NbF7]2- (8-6)
Zr4+ + 6F-[ZrF6]2- (8-7)
Ti4+ + 6F- [TiF6]2- (8-8)
218
Nanoporous Ta2O5 can be formed by anodisation [119, 279] and under specific
conditions, such as high viscosity of the concentrated acid, nanotubes can be grown in 5
- 20 s [120]. The diameter of the pores of Ta2O5 were 2 to 35 nm in cases where the
oxide layer was 350 - 400 nm thick [104, 117, 279], although the height of the
nanoporous layer has been reported to be up to 16 µm [121, 122]. A porous Nb2O5 layer
with pore diameter of 20 - 30 nm and layer thickness of ~ 500 nm [51] can be fabricated
on the surface of niobium by anodisation in the presence of F- ions. This layer system
has been investigated for use as a capacitor [292]. Nanotubular ZrO2 can be obtained by
anodic oxidation of zirconium in an electrolyte consisting of the fluorine anion by
means of a process that is similar to other valve metals [53]. The anodising conditions
can be improved by employing a buffered electrolyte to obtain ZrO2 nanotubes of ~
50 nm diameter and layer thickness of ~ 17 µm [102, 132].
8.3.2 The dynamics of the anodisation process for tantala (Ta2O5), niobia (Nb2O5)
and zirconia (ZrO2)
Figure 8.1(a) shows a top view of a nanoporous Ta2O5 that exhibits irregularly
distributed pores with a diameter range of 35 - 65 nm. The thickness of this layer was
1.24 µm, taking note that the sub-structure revealed multiple layers (Fig. 8.1(b)), as
reported previously [122], where every layer exhibited a closed bottom. Fig. 8.1(c)
shows the cross section of the multiple layers and the arrow indicates Ta2O5 nanotubes
with a length of nearly 500 nm that formed initially after several seconds during
anodisation. These long nanotubes transformed into nanoporosity after several seconds
during anodisation and exhibited similar morphological features to other layers. A top
view of the multiple layers can be observed in Fig. 8.1(d).
219
Fig. 8.1 SEM images of nanoporous Ta2O5: (a) top view (b) the bottom view and the
cross section of multilayer, (c) shorted life nanotubes (arrow), (d) the top view of
multilayer formed on a Ta after anodisation for 120 min in 1M H2SO4 + 3.3 wt %
NH4F, 20V, SEM image of nanoporous Nb2O5 (e) top view and (f) cross section formed
on niobium after anodisation for 16 min in 1M H2SO4 + 3.3 wt % NH4F, 20V and SEM
image of ZrO2 nanotubes (g) top view, (h) cross section and (i) bottom view formed on
a zirconium after anodisation for 95 min in 1M (NH4)2SO4 + 0.3 wt % NH4F, pH=5 and
20V
When anodisation starts, a compact oxide layer forms on the surface of tantalum for a
finite thickness according to the decay characterisitics of the current. Tantalum ions,
Ta5+, arrive at the interface of the oxide layer and electrolyte and become soluble in the
electrolyte by forming [TaF7]2-. Thus, pores formed at the interface between the oxide
220
layer and electrolyte. These pores continued to develop during anodisation because they
preferred sites that trap the necessary ions. Then, the growth rate of the metal oxide
nanoporous and nanotubes may be determined by the diffusion rate of F- ions and the
soluble metal complex [58].
Although the formation mechanism for the nanoporosity of tantalum and other valve
metals is uncertain, the nanoporosity could be related to the molecular dimension of
[TaF7]2- and the formation kinetics of Ta2O5. The standard electrode potential of
Ta→Ta5+ is 0.75 V, which is less than that of Ti→Ti4+, 2.132 V [236]. Thus Ta5+
oxidises quickly. In addition, the standard enthalpies of formation for Ta2O5 and TiO2
are -492.790 and -228.360 gram calories per mole, respectively [237]. It can be
concluded that the oxidisation process and the formation of a Ta2O5 layer on tantalum
are faster than the processes on the titanium surface.
The mechanism and kinetics for the oxide layer formation can be calculated on the basis
of the ionic radii for Ta5+ and F- on a first principles basis. In this instance, the physical
and chemical environment around the [TaF7]2- at the pore sites, which affect the
molecular size of [TaF7]2-, are ignored. The structure of the complex has been reported
to be a two pentagonal pyramid. The volume of the circumscribed sphere of this
structure can be calculated as 26.094×10-30 m3, which is smaller than the octahedral
volume of [TiF6]2- = 29.182×10-30 m3. It can be expected that the diffusion of [TaF7]2- is
faster than [TiF6]2- and that the pore growth does not only occur solely at the bottom,
but also at the walls, thereby resulting in pores that merge and agglomerate. Over a
period of several seconds the pore was bottle-shaped similar to that of nanotubes with a
closed bottom where the diameter of tubes near the bottom was greater than that at the
tube necks. Gaps are also created at the bottom of the layer when the pores merged,
which permits the electrolyte to become available underneath the nanotubes. Thus, a
new layer of nanotubes can form under the same mechanism, thereby resulting in the
formation of multilayers of nanoporous layer over time. The nanotubes dissolve
according to the dissolution rate of the oxide layer in the presence of the fluorine anions
(F-).
The nanoporous Nb2O5 (as shown in Fig. 8.1(e)) exhibited nearly uniform pore
diameters of 22 - 42 nm. The thickness of this layer was ~ 235 nm (Fig. 8.1(f)). The
same formation mechanism can be hypothesised for nanoporous Nb2O5 layer formation
221
as has been described for the formation of the nanoporous Ta2O5 layer. Niobium (Nb)
and Zr are in the same period and adjacent columns in periodic table of the elements.
The standard electrode potential of Nb→Nb5+ is 0.644 V, which is less than that of
Zr→Zr4+, 1.553 V [236] and Nb is, therefore, oxidised rapidly. In addition, the standard
enthalpy of formation for Nb2O5 is -458.640 gram calories per mole; whereas the
enthalpy of formation is -264.199 gram calories per mole for ZrO2 [237].
The processes of oxidisation and formation of the oxide layer on the niobium surface
are rapider than those on the zirconium surface. Consider, for example, that the physical
and chemical environment around the [NbF7]2- ion within the pore structure are ignored.
The volume of the ion complexes can then be calculated on the basis of the ionic radii
for Nb5+ and F- ions and assuming a lattice structure of a two pentagonal pyramid. The
volume of the circumscribed sphere of the [NbF7]2- ion is calculated as 32.02×10-30 m3,
which is smaller than the volume of the octahedral [ZrF6]2- ion which is 36.08×10-30 m3.
It can be expected, therefore, that the diffusion of a [NbF7]2- ion within a pore is rapider
than for a [ZrF6]2- ion. The growth of the pore does not arise only at the bottom but also
at the walls, which will result in the merging of pores.
Figure 8.1(g-i) show the fabricated ZrO2 nanotubes that exhibit (i) uniform inner pore
diameters of 21 - 35 nm, (ii) an outer diameter of 54 - 68 nm, and (iii) wall thickness of
10 - 16 nm. The thickness of this layer was about 24.4 µm. As previously explained
[240] the standard electrode potential of Zr→Zr4+ is 1.553 V is lower than that of
Ti→Ti4+, 2.132 V [236] which results in rapid oxidation. In addition, the standard
enthalpy of formation for ZrO2 is -264.199 gram calories per mole whereas that for
TiO2 is -228.360 gram calories per mole [237]. The process of oxidation and the
formation of the oxide layer on the surface of zirconium are faster than those for the
titanium surface. However, the formation of the [ZrF6]2- ion takes more time than for the
[TiF6]2- ion because the solubility equilibrium constant of (NH4)2ZrF6 is higher than that
of (NH4)2TiF6. Based on the hypothesis described previously, the volume of the
circumscribed sphere of the octahedral structure of [ZrF6]2- can be calculated as
36.08×10-30 m3 and 29.18×10-30 m3 for [TiF6]2- which causes a lower diffusion rate
inside the pore. It has been reported that nanotubes grow longer with a smaller diameter
when the dissolution rate is slower compared to the oxidation process [249].
222
8.3.3 Physical characteristics of nanoporous Ta2O5 and Nb2O5 and nanotubular
ZrO2 layers
Figure 8.2(a) demonstrates the influence of the formation of nanoporous Ta2O5 and the
TiO2 nanotubes on the Sa with respect to the bare metals. The nanoporous layer of
Ta2O5 revealed a lower surface roughness than the TiO2 nanotubes with almost the same
distribution of inner diameter (Di) when the concentration of fluoride ion (CF- hereafter)
was 0.5 wt % at 20 V. The roughness similarity arose because the TiO2 nanotubes were
separated from each other whereas the nanoporous Ta2O5 formed a continuous layer.
The surface area index (SI) may be calculated by dividing the projected surface area,
viz. the total exposed three-dimensional surface area being analysed, including peaks
and valleys, to the surface area measured in the lateral direction. The volume index (VI)
may be calculated by dividing the natural volume, viz. the amount of liquid that it
would take to submerge the dataset to its highest point, to the normal volume that is
measured in the lateral of nanoporous layer. The SI and VI values of nanoporous Ta2O5
are shown in Table 8.1 with the corresponding roughness amplitude parameters and
verify that the Ta2O5 nanoporous layer did not change significantly the surface area and
roughness of the bare metal. Although the distribution of nanoporous Ta2O5 was not
uniform, there were no high peaks and low valleys according to the measured Sskw
which is close to zero. The peaks and valleys exhibited a platykurtoic distribution that
was almost uniform due to the Sku being near to 3.
223
Fig. 8.2 Demonstration of (a,c,e) changes of roughness (Sa) and (b,d,f) water contact
angle of bare tantalum, niobium, zirconium and titanium, after fabrication of
nanotubular and nanoporous layer, and subsequently annealing, respectively, Note:
Each data point is an average of five measurements
224
Table 8.1 Surface area and volume index and roughness amplitude parameters of nanoporous Ta2O5, Nb2O5 and nanotube ZrO2, Note: Each
data point is an average of five measurements
Surface area Index
(SI)
Volume Index
(VI) Sa () Sq () Sskw Sku
Nanoporous Ta2O5 -as-formed 1.34 ± 0.01 18393 ± 86 0.27 ± 0.01 0.35 ± 0.01 0.60 ± 0.23 4.82 ± 0.52
Nanoporous Ta2O5 -Annealed (10 min at 290°C) 1.27 ± 0.01 1838 ± 132 0.25 ± 0.01 0.32 ± 0.01 0.72 ± 0.03 3.90 ± 0.17
Bare tantalum foil (0.1 mm) 1.06 ± 0.00 18229 ± 0 0.23 ± 0.00 0.29 ± 0.00 0.06 ± 0.00 2.79 ± 0.00
Nanoporous Nb2O5 -as-formed 1.50 ± 0.00 18398 ± 134 0.26 ± 0.00 0.34 ± 0.00 0.64 ± 0.03 3.78 ± 0.21
Nanoporous Nb2O5 -annealed (10 min at 290°C) 1.38 ± 0.00 18357 ± 144 0.26 ± 0.01 0.33 ± 0.01 0.39 ± 0.02 3.28 ± 0.05
Bare niobium foil (0.05 mm) 1.04 ± 0.00 18327 ± 0 0.31 ± 0.00 0.38 ± 0.00 0.22 ± 0.00 2.68 ± 0.00
ZrO2 Nanotube--as-formed 2.44 ± 0.30 18422 ± 12 0.56 ± 0.05 0.71 ± 0.05 -0.86 ± 0.018 3.75 ± 0.49
ZrO2 Nanotube-annealed (10 min at 290°C) 2.94 ± 0.15 18428 ± 79 0.61 ± 0.01 0.76 ± 0.00 -0.57 ± 0.20 3.30 ± 0.35
Bare zirconium foil (0.05 mm) 1.03 ± 0.00 18531 ± 0.00 0.12 ± 0.00 0.20 ± 0.00 -4.27 ± 0.00 39.31 ± 0.00
225
Figure 8.2(b) shows the change of hydrophilic properties of bare tantalum and titanium
after fabrication of the nanoporous and nanotubular layers on their surfaces. The water
contact angle measurement of a surface represents the wetting properties of the surface.
In the literature, a superhydrophobic surface is defined when the water contact angle is
more than 150˚ and hydrophobic when it is 90˚ < θw < 150˚ [241]. A surface has
superhydrophilic properties when θw < 10˚ and has hydrophilic properties with a water
contact angle 10˚ < θw < 90˚ [241]. The as-formed nanoporous Ta2O5 exhibited
hydrophilic properties that were similar to the TiO2 nanotubes with the same
distribution of inner diameters, Di (under the conditions of CF- = 0.5 wt %, 20 V). Fig.
8.2(a) and (b) also suggest a direct link between roughness and water contact angle.
When the roughness increased there was a decrease in the water contact angle. The
calculated surface energy for nanoporous Ta2O5 increased after anodisation and also
after annealing due to the crystallisation. The water contact angle and surface energy
have been detailed in Table 8.2 for nanoporous Ta2O5.
226
Table 8.2 Water contact angle and surface energy of as-formed and annealed nanoporous Ta2O5, Nb2O5 and nanotube ZrO2 in comparison to its
bare metal, Note: Each data point is an average of five measurements
Contact Angle
(˚)
sd
mJm-2
sp
mJm-2
s = sd + s
p
mJm-2
Nanoporous Ta2O5 -as-formed 26.0 ± 5.0 11.8 55.2 67.0
Nanoporous Ta2O5 -annealed (10 min at 290°C) 12.1 ± 1.6 15.2 56.7 71.9
Bare tantalum foil (0.1 mm) 47.9 ± 1.2 5.1 49.5 54.6
Nanoporous Nb2O5 -as-formed 25.0 ± 7.0 11.02 56.94 67.96
Nanoporous Nb2O5 -annealed (10 min at 290°C) 15.5 ± 1.4 18.11 52.20 70.31
Bare niobium foil (0.05 mm) 61.0 ± 0.0 6.44 34.94 41.38
ZrO2 Nanotube-as-formed 17.0 ± 1.0 11.56 60.04 71.60
ZrO2 Nanotube-annealed (10 min at 290°C) 8.5 ± 1.0 12.84 60.77 73.61
Bare zirconium foil (0.05 mm) 58.7 ± 1.3 5.82 38.042 43.87
227
Figure 8.2(c) indicates the increments in roughness for the nanoporous Nb2O5 layer and
nanotubes while exhibiting almost the same distribution of Di (under the conditions of
CF- = 0.5 wt %, 15 V) with respect to the corresponding bare metals. The roughness of
the titanium increased upon the creation and separation of TiO2 nanotubes that give rise
to gaps in the surface architecture, whereas the roughness of the niobium decreased
upon the growth of nanoporous Nb2O5. The Sskw of the Nb2O5 nanoporous surface is
higher than zero, indicating the presence of high peaks and low valleys with a
leptokurtoic distribution. The surface area and volume indices of nanoporous Nb2O5 are
shown in Table 8.1, as well as the corresponding roughness amplitude parameters. Fig.
8.2(d) shows the changes of hydrophilic properties of the bare niobium and titanium
after fabrication of the nanoporous and nanotubular layers on their surfaces. The
nanoporous surface revealed a water contact angle in the range of almost the same
distribution of Di (under the conditions of CF- = 0.5 wt %, 15 V) for TiO2 nanotubes that
showed hydrophilic properties. This data also implies that there is a link between
roughness and the water contact angle. The surface energy of this layer was higher than
bare niobium. The water contact angle and surface energy are listed in Table 8.2 for the
nanoporous Nb2O5.
Figure 8.2(e) shows the effect of ZrO2 and TiO2 nanotube formation on the roughness,
Sa, of the bare metals. The presence of the nanotubular ZrO2 has increased the
roughness of the bare metal and also revealed a higher roughness than the nanotubular
layer on the TiO2 nanotubes surface with almost the same distribution of Di (under the
conditions of CF- = 0.5 wt %, 10 V). The Sskw of the ZrO2 nanotubular layer is lower
than zero, which signifies deep valleys such as scratches with a leptokurtoic
distribution. The surface area and volume indices of ZrO2 nanotube is shown in Table
8.1 along with their roughness amplitude parameters.
228
The change of the hydrophilic properties of bare zirconium and titanium after
fabrication of naotubular layer on their surfaces is shown in Fig. 8.2(f). Although both
nanotubular layers fabricated on the zirconium and titanium surfaces are hydrophilic,
the ZrO2 nanotubes exhibited a lower water contact angle in comparison to TiO2 with
nearly the same distribution of Di (CF- = 0.5 wt %, 10 V) which led to a high surface
energy. The water contact angle and surface energy is detailed in Table 8.2 for ZrO2
nanotubes. A direct link between roughness and the water contact angle can be observed
in this figure.
Nanoporous Ta2O5 with a mixture of amorphous and hexagonal phase was fabricated on
tantalum via anodisation using an electrolyte of 1 M H2SO4 + 3.3 wt % NH4F at an
applied potential of 20 V for 120 min. The amorphous nanoporous Ta2O5 transformed
into hexagonal Ta2O5 after annealing at 290 ºC for 10 min, as indicated by the XRD
patterns in Fig. 8.3(a). After annealing the pore size decreased to the range of 23 -
49 nm as a result of crystallisation. Annealing increased the hydrophilic properties of
the nanoporous Ta2O5 layer as much as TiO2 nanotubes because of their different
porosity due to their structures.
229
Fig. 8.3 XRD patterns of (a) the nanoporous Ta2O5 and bare tantalum foil, (b) the
nanoporous Nb2O5 and bare niobium foil, (c) the ZrO2 nanotubes and bare zirconium
foil and (d) the TiO2 nanotubes and bare titanium foil
Nanoporous Nb2O5 with a mixture of amorphous and monoclinic phase was fabricated
on niobium via anodisation using an electrolyte of 1 M H2SO4 + 3.3 wt % NH4F at an
applied potential of 20 V for 16 min. The amorphous nanoporous Nb2O5 transformed
into base-centred monoclinic Nb2O5 after annealing at 290 ºC for 10 min, as indicated
by the XRD patterns in Fig. 8.3(b). The pore size, after annealing, was irregular and laid
in the range of 26 - 60 nm. The pore size was affected by the crystallisation process
during the heat treatment. Annealing did not influence the roughness of the nanoporous
Nb2O5 but decreased the water contact angle of the layer.
Cubic ZrO2 nanotubes were fabricated on zirconium via anodisation using an electrolyte
of 1 M (NH4)2SO4 + 0.3 wt % NH4F (with the addition of H2SO4 to attain a pH=5) at an
230
applied potential of 30 V for 95 min, as indicated by the XRD patterns in Fig. 8.3(c).
After annealing, the inner diameter of the nanotubes laid in the range of 20 - 36 nm,
whilst the nanotube outer diameter (Do) was 48 - 68 nm and the wall thickness (Wt) was
in the range of 9 - 11 nm. The roughness (Sa) and hydrophilic properties of ZrO2
nanotubes increased after annealing. Fig. 8.4(a) shows the pore sizes of nanoporous
Ta2O5 and Nb2O5, compared to TiO2 nanotube with nearly the same Di. Fig. 8.4(b), (c)
and (d) show the Di, Do and Wt of ZrO2 nanotubes compared to TiO2 nanotube with
nearly the same pore sizes as a visual illustration.
Fig. 8.4 Illustration of pore size of the nanoporous and nanotubular layers
8.3.4 Bioactivity of nanoporous and nanotubular metal oxide layers
Bioinert implant materials upon implantation in vivo are encapsulated by fibrous tissues
that isolate them from the surrounding bone. This is not the ideal healing mechanism,
whereas a bioactive material is preferred which bonds to living bone by forming a
carbonated apatite layer on their surfaces similar to hydroxyapatite (HA) of bone
231
composition [56]. The bioactivity of the biocompatible metals with the as-formed and
annealed nanoporous and nanotubular metal oxides was assessed by immersion in the
m-SBF for a period up to 3 weeks. The response of the oxide layers to the SBF
immersion were observed after 1 day, 1 week, 2 weeks and 3 weeks. After 1 day
immersion, no growth of HA was perceived on the surfaces of the as-formed and
annealed nanoporous Ta2O5 and Nb2O5 and the nanotubular TiO2 and ZrO2. However
after 3 weeks immersion as shown in Fig. 8.5, 8.6 and 8.7 HA were deposited onto the
surfaces of the nanoporous and nanotubular layers.
Fig. 8.5 (a) low and (b) high magnifications of HA on as-formed nanoporous Ta2O5; (c)
low and (d) high magnifications of HA on annealed nanoporous Ta2O5
232
Fig. 8.6 SEM images of HA on a) as-formed and b) annealed nanoporous Nb2O5
Fig. 8.7 (a) low and (b) high magnifications of HA on as-formed nanotubular ZrO2; (c,)
low and (d) high magnifications of HA on annealed nanotubular ZrO2
233
The thickness of the HA layers mineralised on the surface of nanoporous Ta2O5,
nanoporous Nb2O5, nanotubular ZrO2 and TiO2 were 6.9, 3.7, 3.3 and 3.8 µm
respectively as shown in Fig. 8.8.
Fig. 8.8 SEM images of cross section of mineralised HA layers on the surface of a)
nanoporous Ta2O5, b) nanoporous Nb2O5, c) nanotubular ZrO2 and d) nanotubular TiO2
The atomic ratio of calcium to phosphate calculated using EDS results after 3 weeks is
indicated in Fig. 8.9.
234
Fig. 8.9 Bioactivity of biocompatible nanoporous and nanotubular oxide metals after 3
weeks in m-SBF at 37 ˚C
The M-OH groups located on the surface of biocompatible metal oxides are favoured
sites for apatite nucleation [56, 251, 254, 257]. First, Ca2+ ions are absorbed onto the
hydrolysed nanoporous and nanotubular oxide surface by Coulomb attraction forces.
Then, existing phosphate groups inside the SBF are adsorbed to the positively charged
surface, resulting in the formation of calcium phosphate. The stoichiometric Ca/P (at %)
of octacalcium phosphate [Ca8H2(PO4)6×5H2O], tricalcium phosphate [Ca3(PO4)2] and
hydroxyapatite [Ca10(PO4)6(OH)2] are 1.33, 1.5 and 1.67, respectively. Fig. 8.10 shows
local pH changes up to 3 weeks of mineralisation of HA. When negative hydrolysed
nanotubular surface formed the pH value increased. When the absorption of Ca2+ ions
was accomplished the changes of pH value were negligible. The exchange of OH- with
PO43- as a result of their absorption increased the pH value. By consuming the OH- to
form hydroxyapatite [Ca10(PO4)6(OH)2] the pH value decreased and remaining constant
after the HA formation completed.
235
Fig. 8.10 pH value of static m-SBF as a function of soaking time for: a) nanoporous
Ta2O5 and Nb2O5 and b) nanotubular TiO2 and ZrO2
According to the obtained Ca/P ratio for the metal oxides, annealed nanoporous Ta2O5
had a high value of Ca/P ratio which may indicate the presence of a mixture of calcium
and phosphate abovementioned compositions. Other nanoporous and nanotubular metal
oxide layers induce calcium phosphate and need more time or pre-treatment to induce
crystalline hydroxyapatite. Ta2O5 reaches an isoelectric point at a pH of 2.7 - 3.0, which
can be compared to Nb2O5 that attains an isoelectric point at a pH of 4.0 [255, 256].
Thus, the Ta2O5 surface, compared to the Nb2O5 surface, becomes more negatively
charged in SBF at a pH of 7.4. Therefore there is an appropriate chemical environment
for a higher Ca/P atomic ratio to be attained for Ta2O5 in comparison to Nb2O5. In
addition, the isoelectric point of TiO2 has been reported at a pH of 3.9 and the
corresponding value for ZrO2 is 5.5 [255, 256], which would be expected to lead to a
higher Ca/P ratio for TiO2.
8.4 Conclusions
Nanoporous oxide layers on the surfaces of tantalum and niobium and nanotubular
layers on the surfaces of zirconium and titanium were fabricated via anodisation. The
bioactivity of the nanoporous and nanotubular layers were evaluated. The prime
conclusions are as follows.
1. The pore size distributions of the nanoporous tantala (Ta2O5) were not uniform
in comparison to TiO2 nanotubes. The pore size distribution of the nanoporous
niobia (Nb2O5) was almost uniform in comparison to the TiO2 nanotubes. Higher
236
applied potential was needed in fabricating ZrO2 nanotubes with the same
distribution of inner diameter than TiO2 nanotubes during anodisation.
2. The nanoporous Ta2O5 layer and nanotubular ZrO2 and TiO2 layer exhibited a
higher roughness than their respective bare metals but the nanoporous Nb2O5
layer exhibited a lower roughness than its bare metal.
3. The nanoporous layers of Ta2O5 and Nb2O5 and the nanotubular layers of ZrO2
and TiO2 revealed an increase in hydrophilic property and surface energy
compared to their respective bare metals.
4. The hydrophilic property and the surface energy of the nanoporous layers of
Ta2O5 and Nb2O5 and the nanotubular layers of ZrO2 and TiO2 were increased
after annealing.
5. After annealing the pore size of nanoporous Ta2O5 decreased whereas the pore
size of nanoporous Nb2O5 increased. This is due to their different crystalline
phases before and after annealing which possess different lattice parameters. The
inner diameter of ZrO2 nanotubes after annealing did not show any obvious
change because the crystalline phase did not changed but annealing led to an
increase in the inner diameter of anatase TiO2.
6. As-formed Ta2O5 and ZrO2 exhibited a good bioactivity similar to TiO2.
Annealed Ta2O5 showed a high level of bioactivity which is promising for
biomedical applications. The as-formed amorphous nanoporous Nb2O5 did not
show bioactivity; but the annealed nanoporous Nb2O5 with a base-centred
monoclinic structure was bioactive.
237
Chapter 9 Conclusions
9.1 Introduction
The motivation and principal objectives of this thesis were the investigation of the
bioactivity and biocompatibility of Ti50Zr alloy with a nanotubular surface fabricated
via anodisation. The nanotubular layer which is the result of electrochemical surface
modification played an important role in endowing new surface functionalities of metal
implants. The nanotubular layer fabricated in this study revealed an interesting
morphology. This morphology which was a reflection of the needle-like microstructure
of the alloy exhibited a distribution of inner diameter (Di), outer diameter (Do) and wall
thickness (Wt) of nanotubes and a range of surface roughness parameters due to
different heights of nanotubes. This attractive structure needed to be considered while
their bioactivity and biocompatibility were investigated.
9.2 Major findings
Initially two different morphologies of TiO2-ZrO2-ZrTiO4 nanotubes were fabricated in
two different types of electrolyte; aqueous and non-aqueous electrolyte. In Chapter 4,
firstly the effect of fluorine (F-) anion concentration (CF-) in the aqueous electrolyte on
the nanotube parameters at constant applied potential and anodisation time was
investigated. An increase in the CF- from 0.3 to 0.5 wt % led to an increase in the
number of nanotubes with smaller Di, Do and Wt. Because the chemical and
electrochemical reaction rate of Zr was higher than Ti as a result Zr reacted faster than
Ti. In a same condition the nanotube size of ZrO2 was smaller than TiO2 and the
nanotube length was longer. The mean Di and Do of TiO2 nanotubes on the surface of
CP-Ti increased specifically when CF- increased. Then the effects of the applied
potential on the nanotube parameters at constant F- ion concentration in electrolyte and
the anodisation time were investigated. Conclusions are as follows:
1. Increasing the applied potential from 5 to 35 V resulted in boarder Di and Do. It
means that the number of nanotubes with smaller Di and Do decreased or the
number of nanotubes with larger Di and Do increased due to the increase in the
anodic current. The Wt of nanotubes increased slightly with an increase in the
applied potential from 5 to 35 V.
238
2. Increasing applied potential from 5 to 35 V resulted in an increase in mean Di
and Do of TiO2 nanotubes on the surface of CP-Ti. The extent of increasing the
mean Wt was not as much as the mean Di and Do.
3. At lower applied potential (5 to 20 V) the mean Di and Do of TiO2-ZrO2-ZrTiO4
nanotubes were lower than those of TiO2 nanotubes in contrast to higher applied
potential.
The effect of water content of 5 and 10 wt % in a non-aqueous electrolyte (ethylene
glycol) which contained 0.5 wt % NH4F were investigated on fabrication of TiO2-ZrO2-
ZrTiO4 nanotubes in Chapter 5. The effect of microstructure of Ti50Zr alloy on the
orientation of nanotubes was more observable when the water content increased. The
walls of nanotubes were not separated when they were fabricated in non-aqueous
electrolyte. The key conclusions are as follows:
1. Increasing the water content from 5 to 10 wt % in electrolyte resulted in an
increase in the number of nanotubes with larger Di, Do and Wt or a decrease in
the number of nanotubes with smaller Di, Do and Wt due to the increased rate of
chemical and electrochemical reactions.
2. Increasing the applied potential from 20 to 30 V during anodisation with the
same water content in the electrolyte resulted in boarder distribution of Di and
Do. It means that the number of nanotubes with smaller Di and Do decreased or
the number of nanotubes with larger Di and Do increased due to the increased
anodic current. The extent of increasing of the mean Wt was not as much as the
mean Di and Do.
3. Increasing the applied potential from 20 to 30 V resulted in an increase in the
mean Di and Do of TiO2 nanotubes on the surface of CP-Ti but the mean Wt did
not change obviously. Increasing the water content from 5 to 10 wt % in the
electrolyte at the same applied potential resulted in an increase in the mean Di,
Do and Wt.
4. The mean Di, Do and Wt of TiO2 nanotubes were higher than those of TiO2-
ZrO2-ZrTiO4 nanotubes.
5. The mean Di and Do of nanotubes fabricated in non-aqueous electrolyte were
smaller than the mean Di and Do of nanotubes fabricated in aqueous electrolyte.
239
The distribution of nanotube sizes fabricated in both aqueous and non-aqueous
electrolytes were analysed using normal and Weibull distribution methods. The results
indicated that the normal distribution was an appropriate method to represent the
distribution of nanotube sizes of TiO2-ZrO2-ZrTiO4.
The Mean Roughness (Sa) and the Root Mean Square (RMS) Roughness (Sq) of the
nanotubular layer of TiO2-ZrO2-ZrTiO4 were evaluated using 3D-Profilometry
(Chapters 4, 5, 6 and 7). As it mentioned before because of different oxidation and
dissolution rates of the elements of the alloy; Ti and Zr, the nanotubes not only
exhibited different nanotube sizes but also exhibited different lengths, resulting in
different amplitude and space roughness parameters. The key conclusions are as
follows:
1. Increasing the CF- in the aqueous electrolyte resulted in an increase in the Sa
and Sq.
2. Increasing the applied potential in the aqueous electrolyte from 5 to 20V
resulted in an increase in the Sa and Sq but these roughness parameters
decreased when the applied potential increased from 20 to 35 V due to the
increased anodic current.
3. Increasing the water content from 5 to 10 wt % in the non-aqueous
electrolyte at the same applied potential resulted in an increase in Sa and Sq.
4. Increasing the applied potential from 20 to 30 V in the non-aqueous
electrolyte with the same water content resulted in an increase in Sa and Sq.
5. The Sa and Sq of the nanotubular layer of TiO2 on the surface of CP-Ti
increased with an increase in the CF- in the aqueous electrolyte and it
increased when the applied potential increased from 10 to 20 V then
decreased when the applied potential increased from 20 to 30V.
6. The Sa and Sq of the nanotubular layer of TiO2 on the surface of CP-Ti
increased with an increase in water content in the non-aqueous electrolyte
and applied potential.
7. The value of Sa and Sq of the nanotubular layer of TiO2-ZrO2-ZrTiO4 were
higher than those of the nanotubular layer of TiO2 anodised under the same
condition.
240
8. The values of Sa and Sq of the nanotubular layer of TiO2-ZrO2-ZrTiO4
fabricated in aqueous electrolyte were higher than those of the nanotubular
layer of TiO2-ZrO2-ZrTiO4 anodised in the non-aqueous electrolyte.
In Chapters 4 and 5, it was found that there was a direct link between roughness
parameter and wettability or hydrophilic properties of nanotubular layers of TiO2-ZrO2-
ZrTiO4 fabricated in both aqueous and non-aqueous electrolyte. The key conclusions are
as follows:
1. Increasing the CF- in the aqueous electrolyte resulted in an increase in the
hydrophilic properties of TiO2-ZrO2-ZrTiO4 nanotubes.
2. Increasing the applied potential in the aqueous electrolyte from 5 to 30 V
resulted in an increase in the hydrophilic properties of TiO2-ZrO2-ZrTiO4
nanotubes.
3. The hydrophilic properties of TiO2 nanotubes on the surface of CP-Ti increased
with an increase of CF – in aqueous electrolyte from 0.3 to 0.5 wt % and an
increase of the applied potential from 5 to 30 V, and did not change when the
applied potential increased from 30V to 35V.
4. The hydrophilic properties of TiO2-ZrO2-ZrTiO4 nanotubes fabricated in
aqueous electrolyte were higher than those of the TiO2 nanotubes anodised under
the same applied potential and it increased with an increase in mean Di and
roughness parameters.
5. Increasing the water content from 5 to 10 wt % in the non-aqueous electrolyte at
the same applied potential resulted in an increase in the hydrophilic properties of
TiO2-ZrO2-ZrTiO4 nanotubes.
6. Increasing the applied potential from 20 to 30 V in the non-aqueous electrolyte
with the same water content resulted in an increase in the hydrophilic properties
of TiO2-ZrO2-ZrTiO4 nanotubes.
7. The hydrophilic properties of TiO2 nanotubes on the surface of CP-Ti increased
with an increase of water content in non-aqueous electrolyte from 5 to 10 wt %
and an increase of the applied potential from 20 to 30 V.
8. The hydrophilic properties of TiO2-ZrO2-ZrTiO4 nanotubes fabricated in non-
aqueous electrolyte were almost the same in comparison to the hydrophilic
properties of TiO2 nanotubes anodised under the same condition.
241
9. The hydrophilic properties of TiO2-ZrO2-ZrTiO4 nanotubes fabricated in
aqueous electrolyte were higher than those of the TiO2-ZrO2-ZrTiO4 nanotubes
fabricated in non-aqueous electrolyte probably because of their different mean
mean Di and roughness parameters of Sa.
There was a direct link between the hydrophilic properties and the calculated surface
energy of nanotubular layer of TiO2-ZrO2-ZrTiO4 fabricated in aqueous electrolyte. The
surface energy of the nanotubular layer increased when the hydrophilic properties
increased. Such a link was not observed for the nanotubular layer of TiO2-ZrO2-ZrTiO4
fabricated in non-aqueous electrolyte due to the possibility of remaining of the organic
electrolyte inside the nanotubes. The surface energies of the nanotubular layers of TiO2-
ZrO2-ZrTiO4 were higher than those of the nanotubular layers of TiO2 fabricated in
aqueous electrolyte under the same applied potential and anodisation time. The surface
energies of the nanotubular layers of TiO2-ZrO2-ZrTiO4 fabricated in aqueous
electrolyte were lower than those of the nanotubular layers of TiO2-ZrO2-ZrTiO4
fabricated in non-aqueous electrolyte under the same applied potential and anodisation
time.
The hardness and elastic modulus of both nanotubular layers of TiO2-ZrO2-ZrTiO4
fabricated in aqueous and non-aqueous electrolyte were measured using nanoindenter,
Chapters 4 and 5. The porosity of both nanotubular layers of TiO2-ZrO2-ZrTiO4
fabricated in aqueous and non-aqueous electrolyte caused a decrease in the hardness of
the Ti50Zr alloy surface. However the ceramic properties of the metal oxide layer
fabricated in aqueous electrolyte caused an increase in the elasticity of the alloy surface.
The hard and brittle nanotubular layer fabricated in non-aqueous electrolyte exhibited
lower elastic modulus than the bare alloy. The nano-hardness and elasticity were
influenced by the values of the porosity (nanotube sizes) for both nanotubular layers.
In Chapters 4 and 5, the nanotubular layer of TiO2-ZrO2-ZrTiO4 and TiO2 on the surface
of Ti50Zr and CP-Ti fabricated in aqueous and non-aqueous electrolyte were
amorphous. After annealing crystalline phases formed and their physical, chemical and
mechanical properties changed. Conclusions are as follows:
1. The amorphous TiO2-ZrO2-ZrTiO4 nanotubes fabricated in aqueous electrolyte
changed to a mixture of anatase, orthorhombic TiO2 and ZrO2, and orthorhombic
242
ZrTiO4. The amorphous TiO2 nanotubes changed to tetragonal anatase after
annealing.
2. The mean Di and Do and Wt of TiO2-ZrO2-ZrTiO4 nanotubes anodised at lower
applied potential decreased after annealing; but increased when they were
anodised at higher applied potential. The mean Wt did not show an obvious
change. The mean Di and Do of TiO2 nanotubes anodised at higher applied
potential increased after annealing and the mean Wt of TiO2 nanotubes did not
exhibit a remarkable change.
3. The mean Sa and Sq of TiO2-ZrO2-ZrTiO4 and TiO2 nanotubes decreased and
their hydrophilic properties, surface energy increased after annealing; whilst the
hardness and elasticity of the nanotubular layer of TiO2-ZrO2-ZrTiO4 increased.
4. The amorphous TiO2-ZrO2-ZrTiO4 nanotubes fabricated in non-aqueous
electrolyte changed to a mixture of tetragonal anatase and rutile, orthorhombic
TiO2 and ZrO2, and orthorhombic ZrTiO4 after annealing.
5. The mean Di and Do and Wt of TiO2-ZrO2-ZrTiO4 nanotubes changed after
annealing. The mean Di of both TiO2-ZrO2-ZrTiO4 and TiO2 nanotubes
decreased, the mean Wt increased and the mean Do did not change.
6. The top layer of nanotubes fabricated in non-aqueous electrolyte was brittle and
cracked after annealing. Therefore it was removed and the underneath dimpled
surface of nanotubes was used for further investigation.
The bioactivity of TiO2-ZrO2-ZrTiO4 and TiO2 nanotubes fabricated in aqueous and
non-aqueous electrolyte was evaluated by immersing them in m-SBF and the Ca/P ratio
(at %) was measured using EDS. The nanotubular layer of TiO2-ZrO2-ZrTiO4 and TiO2
fabricated in both aqueous and non-aqueous electrolytes with a variety range of
nanotube sizes were bioactive and the size, the shape, hydrophilic properties and
crystalline phase of the nanotubes affected the Ca/P ratio of HA.
The cell behaviour of osteoblast (SaOS2) on the surface of TiO2-ZrO2-ZrTiO4
nanotubes fabricated in aqueous and non-aqueous electrolyte with a variety of nanotube
sizes and shapes with different physical and chemical properties was evaluated. It was
found that besides the Di of nanotubes the topography of the nanotubular surface and its
chemical properties affect the density of cell attachment. The highest percentage of cell
attachment of 65.2 % was on the surface of nanotubular layer with Di = 18 ± 6 nm.
243
Although it decreased when the Di increased to 59 ± 17 nm as expected (the lowest
percentage of cell attachment of 25.9 %), the percentage of cell attachment increased on
the surface of TiO2-ZrO2-ZrTiO4 nanotubes with Di = 64 ± 23 nm and Wt = 21 ± 7 and
Di = 82 ± 26 nm and Wt = 24 ± 7 nm, respectively. It can be concluded that the high cell
attachment was on the surface of nanotubes with either Di or Wt in the range of optimum
nanospacing and increasing the roughness parameter of the surface distracted the
process of cell attachment. Two post treatments; cleaning and coating with HA which
were undertaken on the nanotubular layer with Di = 40 ± 12 nm resulted in an increase
of ~ 1.5 to 2 times of the percentage of cell attachment. Almost the same result was
observed when the osteoblast cells seeded on the surface of TiO2-ZrO2-ZrTiO4
nanotubes fabricated in non-aqueous electrolyte. Although the morphology and
topography of nanotubular layer fabricated in aqueous electrolyte are different from
nanotubular layer fabricated in non-aqueous electrolyte two surfaces with almost the
same Sa and Sm were compared. The percentage of attached cells on dimpled
nanotubular surface fabricated in non-aqueous electrolyte was higher than that of the
nanotubular layer fabricated in aqueous electrolyte with smaller Di and higher
hydrophilic properties and surface energy. It can be concluded that the non-separated
wall of the nanotubes may receive more protein adsorption and serves as a more suitable
nano-pattern surface for bone cell attachment.
It can be concluded from the study presented here that the mixed nanotubes of TiO2-
ZrO2-ZrTiO4 on Ti50Zr alloy with a variety of nanotube size distribution and
morphology and other related properties such as roughness, hydrophilic and bioactivity
along with other excellent properties of the substrate (bare alloy) are more promising
and have potential for biomedical application than 100 % TiO2 nanotubes on CP-Ti.
The nanoporous and nanotubular layer of metal oxides fabricated on the surface of the
most commonly used titanium alloying element of Ta, Nb and Zr were processed and
their physical, chemical and bioactive properties in comparison to TiO2 nanotubes with
almost the same Di were evaluated. The nanoporous Ta2O5 with non-uniform pore size
exhibited a decrease of the roughness and nanoporous Nb2O5 with uniform pore size
exhibited an increase of the roughness in comparison to their respective bare metal. The
nanotubular layer of ZrO2 with almost the same Di as nanotubular layer of TiO2 was
fabricated at a higher applied potential and exhibited higher roughness than its
244
respective bare metal. The hydrophilic properties and surface energy of these
nanoporous and nanotubular layer was higher than their bare metal and increased after
annealing and their crystalline phase changed. Similar to TiO2 nanotube, as-formed
nanoporous Ta2O5 and ZrO2 nanotubes revealed a good bioactivity but as-formed
nanoporous Nb2O5 were not bioactive. All these nanoporous and nanotubular layers
were bioactive after annealing and annealed nanoporous Ta2O5 exhibited a high level of
bioactivity. It can be concluded that these alloying elements are promising for
biomaterial application alone or accompany with other biomaterials.
9.3 Recommendation for future work
Some area of this study remained still to be identified and future research can be
explored as follows:
1. There was a direct link between the roughness parameters of the surface and its
wettability. If the water contact angle of nanotubular surface with same
distribution of Di and different roughness parameters, and same roughness
parameter and different distribution of Di were measured, a modelling study
could be undertaken to clarify this relationship. With this model an optimum
nanotubular layer for enhanced bone cell attachment could be fabricated by
controlling the anodisation parameters.
2. Although this study attempted to interpret the hydrophilic property, bioactivity
and bone cell attachment by studying the roughness amplitude and spacing
parameters and surface area index (SI), evaluating the nanoporosity of the
nanotubular layer with different Di could provide a more reliable explanation.
3. More fundamental study is needed to modify the mechanical properties of
nanotubular layer fabricated in non-aqueous electrolyte after annealing.
4. There is a debate about optimum nanospacing for cell adhesion. There were
conclusions that Di ≤ 30 nm of as-formed nanotubes or Di ≤ 100 nm of annealed
nanotubes was the optimum nanospacing. Therefore cell attachment study with
the same distribution of Di and roughness parameter could be undertaken on as-
formed and annealed nanotubes to clarify this debate.
5. The effect of applying post treatment on the nanotubes anodised in non-aqueous
electrolyte such as removing contaminations of the nanotubular layer or
245
deposition of an HA layer via immersion in SBF on cell adhesion could be
undertaken.
6. More fundamental knowledge is needed to explain the different relationship
between water contact angle and surface energy of the nanotubular layer
fabricated in aqueous and non-aqueous electrolyte and to modify it.
7. The nanotubular layer fabricated in both electrolytes can be coated by HA
electrodeposition and the effect of this layer on improving bone cell attachment
could be studied and compared with the mineralised HA by soaking in m-SBF.
The physical, chemical and mechanical properties of these final layers can be
measured and compared.
246
References
[1] Williams DF. On the nature of biomaterials. Biomaterials 2009;30:5897-909. [2] Navarro M, Michiardi A, Castaño O, Planell JA. Biomaterials in orthopaedics. Journal
of The Royal Society Interface 2008;5:1137-58. [3] Niinomi M. Mechanical biocompatibilities of titanium alloys for biomedical
applications. Journal of the Mechanical Behavior of Biomedical Materials 2008;1:30-42.
[4] Niinomi M. Metallic biomaterials. Journal of Artificial Organs 2008;11:105-10. [5] Long M, Rack HJ. Titanium alloys in total joint replacement - A materials science
perspective. Biomaterials 1998;19:1621-39. [6] Duan K, Wang R. Surface modifications of bone implants through wet chemistry.
Journal of Materials Chemistry 2006;16:2309-21. [7] Liu X, Chu P, Ding C. Surface modification of titanium, titanium alloys, and related
materials for biomedical applications. Materials Science and Engineering: R: Reports 2004;47:49-121.
[8] Hsu HC, Wu SC, Sung YC, Ho WF. The structure and mechanical properties of as-cast Zr-Ti alloys. Journal of Alloys and Compounds 2009;488:279-83.
[9] Wang X, Li Y, Xiong J, Hodgson PD, Wen Ce. Porous TiNbZr alloy scaffolds for biomedical applications. Acta Biomaterialia 2009;5:3616-24.
[10] Zwilling V, Aucouturier M, Darque-Ceretti E. Anodic oxidation of titanium and TA6V alloy in chromic media. An electrochemical approach. Electrochimica Acta 1999;45:921–9.
[11] Roy P, Berger S, Schmuki P. TiO2 Nanotubes: Synthesis and Applications. Angewandte Chemie International Edition 2011;50:2904-39.
[12] Liu X, Chu PK, Ding C. Surface nano-functionalization of biomaterials. Materials Science and Engineering: R: Reports 2010;70:275-302.
[13] Gulati K, Simovic S, Losic D. Highly ordered titania (TiO2) nanotube arrays fabricated by electrochemical self-ordering process toward development of implantable drug delivery devices with triggered drug release. Canberra, ACT2010. p. 159-60.
[14] Lavenus S, Ricquier JC, Louarn G, Layrolle P. Cell interaction with nanopatterned surface of implants. Nanomedicine 2010;5:937-47.
[15] Anselme K. Osteoblast adhesion on biomaterials. Biomaterials 2000;21:667-81. [16] Minagar S, Wang J, Berndt CC, Ivanova EP, Wen C. Cell response of anodized
nanotubes on titanium and titanium alloys. Journal of Biomedical Materials Research - Part A 2013;101 A:2726-39.
[17] Park J, Bauer S, Von Der Mark K, Schmuki P. Nanosize and vitality: TiO2 nanotube diameter directs cell fate. Nano Letters 2007;7:1686-91.
[18] Von Wilmowsky C, Bauer S, Lutz R, Meisel M, Neukam FW, Toyoshima T, Schmuki P, Nkenke E, Schlegel KA. In Vivo Evaluation of Anodic TiO2 Nanotubes; An Experimental Study in the Pig. Journal of Biomedical Materials Research - Part B Applied Biomaterials 2009;89:165-71.
[19] Park J, Bauer S, Schlegel KA, Neukam FW, von der Mark K, Schmuki P. TiO2 Nanotube Surfaces: 15 nm-An Optimal Length Scale of Surface Topography for Cell Adhesion and Differentiation. Small 2009;5:666-71.
247
[20] Brammer KS, Oh S, Cobb CJ, Bjursten LM, Heyde Hvd, Jin S. Improved bone-forming functionality on diameter-controlled TiO2 nanotube surface. Acta Biomaterialia 2009;5:3215-23.
[21] Oh S, Brammer KS, Li YSJ, Teng D, Engler AJ, Chien S, Jin S. Stem cell fate dictated solely by altered nanotube dimension. Proceedings of the National Academy of Sciences 2009;106:2130-5.
[22] Roguska A, Pisarek M, Andrzejczuk M, Dolata M, Lewandowska M, Janik-Czachor M. Characterization of a calcium phosphate-TiO2 nanotube composite layer for biomedical applications. Materials Science and Engineering C 2011;31:906-14.
[23] Varghese OK, Gong D, Paulose M, Grimes CA, Dickey EC. Crystallization and high-temperature structural stability of titanium oxide nanotube arrays. Journal of Materials Research 2003;18:156-65.
[24] Bai Y, Park IS, Park HH, Lee MH, Bae TS, Duncan W, Swain M. The effect of annealing temperatures on surface properties, hydroxyapatite growth and cell behaviors of TiO2 nanotubes. Surface and Interface Analysis 2011;43:998-1005.
[25] Narayanan R, Lee HJ, Kwon TY, Kim KH. Anodic TiO2 nanotubes from stirred baths: Hydroxyapatite growth & osteoblast responses. Materials Chemistry and Physics 2011;125:510-7.
[26] Sista S, Nouri A, Li Y, Wen C, Hodgson PD, Pande G. Cell biological responses of osteoblasts on anodized nanotubular surface of a titanium-zirconium alloy. Journal of Biomedical Materials Research - Part A 2013;101:3416-30.
[27] Sista S, Wen Ce, Hodgson PD, Pande G. The influence of surface energy of titanium-zirconium alloy on osteoblast cell functions in vitro. Journal of Biomedical Materials Research Part A 2011;97A:27-36.
[28] Grigorescu S, Ungureanu C, Kirchgeorg R, Schmuki P, Demetrescu I. Various sized nanotubes on TiZr for antibacterial surfaces. Applied Surface Science 2013;270:190-6.
[29] Weiner S, Wagner HD. The material bone: Structure-mechanical function relations. Annual Review of Materials Science 1998;28:271-98.
[30] Uddin MH, Matsumoto T, Okazaki M, Nakahira A, T S. Biomimetic fabrication of apatite related biomaterials. In: M A, editor. Biomimetics, learning from nature. Vukovar, Croatia: In-Teh; 2010. p. 289-303.
[31] Mour M, Das D, Winkler T, Hoenig E, Mielke G, Morlock MM, Schilling AF. Advances in Porous Biomaterials for Dental and Orthopaedic Applications. Materials 2010;3:2947-74.
[32] Rho JY, Kuhn-Spearing L, Zioupos P. Mechanical properties and the hierarchical structure of bone. Medical Engineering and Physics 1998;20:92-102.
[33] Wheeless CR. Wheeless' textbook of orthopaedics. [S.l.]: C.R. Wheeless, M.D.; 1996. [34] Anselme K. Biomaterials and interface with bone. Osteoporosis International
2011;22:2037-42. [35] Mark K, Park J, Bauer S, Schmuki P. Nanoscale engineering of biomimetic surfaces:
cues from the extracellular matrix. Cell and Tissue Research 2009;339:131-53. [36] Yu WQ, Jiang XQ, Zhang FQ, Xu L. The effect of anatase TiO2 nanotube layers on
MC3T3-E1 preosteoblast adhesion, proliferation, and differentiation. Journal of Biomedical Materials Research - Part A 2010;94:1012-22.
[37] Kokubo T, Kim H-M, Kawashita M. Novel bioactive materials with different mechanical properties. Biomaterials 2003;24:2161-75.
248
[38] Hench LL, Polak JM. Third-generation biomedical materials. Science 2002;295:1014+6-7.
[39] Steinemann SG. Titanium - The material of choice? Periodontology 2000 1998;17:7-21.
[40] Niinomi M. Mechanical properties of biomedical titanium alloys. Materials Science and Engineering A 1998;243:231-6.
[41] Zhou YL, Niinomi M, Akahori T. Effects of Ta content on Young's modulus and tensile properties of binary Ti-Ta alloys for biomedical applications. Materials Science and Engineering A 2004;371:283-90.
[42] Kikuchi M, Takahashi M, Okuno O. Mechanical properties and grindability of dental cast Ti-Nb alloys. Dental Materials Journal 2003;22:328-42.
[43] Niespodziana K, Jurczyk K, Jurczyk M. The synthesis of titanium alloys for biomedical applications. Reviews on Advanced Materials Science 2008;18:236-40.
[44] Elias LM, Schneider SG, Schneider S, Silva HM, Malvisi F. Microstructural and mechanical characterization of biomedical Ti–Nb–Zr(–Ta) alloys. Materials Science and Engineering: A 2006;432:108-12.
[45] Okazaki Y. A new Ti-15Zr-4Nb-4Ta alloy for medical applications. Current Opinion in Solid State and Materials Science 2001;5:45-53.
[46] Okazaki Y, Rao S, Ito Y, Tateishi T. Corrosion resistance, mechanical properties: corrosion fatigue strength and cytocompatibility of new Ti alloys without A1 and V. Biomaterials 1998;19:1197-215.
[47] Lausmaa J. Surface spectroscopic characterization of titanium implant materials. Journal of Electron Spectroscopy and Related Phenomena 1996;81:343-61.
[48] Tsuchiya H, Macak JM, Müller L, Kunze J, Müller F, Greil P, Virtanen S, Schmuki P. Hydroxyapatite growth on anodic TiO2 nanotubes. Journal of Biomedical Materials Research Part A 2006;77A:534-41.
[49] Sul YT, Johansson CB, Jeong Y, Albrektsson T. The electrochemical oxide growth behaviour on titanium in acid and alkaline electrolytes. Medical Engineering and Physics 2001;23:329-46.
[50] Masuda H, Hasegwa F, Ono S. Self-ordering of cell arrangement of anodic porous alumina formed in sulfuric acid solution. Journal of the Electrochemical Society 1997;144:L127-L30.
[51] Sieber I, Hildebrand H, Friedrich A, Schmuki P. Formation of self-organized niobium porous oxide on niobium. Electrochemistry Communications 2005;7:97-100.
[52] Sieber I, Kannan B, Schmuki P. Self-Assembled Porous Tantalum Oxide Prepared in H2SO4/HF Electrolytes. Electrochemical and Solid-State Letters 2005;8:J10-J2.
[53] Tsuchiya H, Schmuki P. Thick self-organized porous zirconium oxide formed in H2SO4/NH4F electrolytes. Electrochemistry Communications 2004;6:1131-4.
[54] Oh S, Daraio C, Chen L-H, Pisanic TR, Fiñones RR, Jin S. Significantly accelerated osteoblast cell growth on aligned TiO2 nanotubes. Journal of Biomedical Materials Research Part A 2006;78A:97-103.
[55] Sibert ME. Electrochemical Oxidation of Titanium Surfaces. Journal of the Electrochemical Society 1963;110:65-72.
[56] Kokubo T. Apatite formation on surfaces of ceramics, metals and polymers in body environment. Acta Materialia 1998;46:2519-27.
249
[57] Fojt J, Moravec H, Joska L. Nanostructuring of titanium for medical applications. Nanocon 2010, 2nd International Conference. Olomouc, Czech Republic, EU: Tanger Ltd 2010.
[58] Macak JM, Tsuchiya H, Ghicov A, Yasuda K, Hahn R, Bauer S, Schmuki P. TiO2 nanotubes: Self-organized electrochemical formation, properties and applications. Current Opinion in Solid State and Materials Science 2007;11:3-18.
[59] Sun L, Wang X, Li M, Zhang S, Wang Q. Anodic titania nanotubes grown on titanium tubular electrodes. Langmuir 2014;30:2835-41.
[60] Sánchez-Tovar R, Lee K, García-Antón J, Schmuki P. Formation of anodic TiO2 nanotube or nanosponge morphology determined by the electrolyte hydrodynamic conditions. Electrochemistry Communications 2013;26:1-4.
[61] Bauer S, Kleber S, Schmuki P. TiO2 nanotubes: Tailoring the geometry in H3PO4/HF electrolytes. Electrochemistry Communications 2006;8:1321-5.
[62] Beranek R, Hildebrand H, Schmuki P. Self-organized porous titanium oxide prepared in H2SO4/HF electrolytes. Electrochemical and Solid-State Letters 2003;6:B12-B4.
[63] Neupane MP, Park IS, Lee MH, Bae TS, Watari F. Influence of heat treatment on morphological changes of nano-structured titanium oxide formed by anodic oxidation of titanium in acidic fluoride solution. Bio-Medical Materials and Engineering 2009;19:77-83.
[64] Oh S, Finones R, Daraio C, Chen L, Jin S. Growth of nano-scale hydroxyapatite using chemically treated titanium oxide nanotubes. Biomaterials 2005;26:4938-43.
[65] Demetrescu I, Ionita D, Pirvu C, Portan D. Present and Future Trends in TiO2 Nanotubes Elaboration, Characterization and Potential Applications. Molecular Crystals and Liquid Crystals 2010;521:195-203.
[66] Zhao J, Wang X, Chen R, Li L. Fabrication of titanium oxide nanotube arrays by anodic oxidation. Solid State Communications 2005;134:705-10.
[67] Macak JM, Tsuchiya H, Schmuki P. High-Aspect-Ratio TiO2 Nanotubes by Anodization of Titanium. Angewandte Chemie International Edition 2005;44:2100-2.
[68] Wan J, Yan X, Ding J, Wang M, Hu K. Self-organized highly ordered TiO2 nanotubes in organic aqueous system. Materials Characterization 2009;60:1534-40.
[69] Lee W-J, Alhoshan M, Smyrl WH. Titanium Dioxide Nanotube Arrays Fabricated by Anodizing Processes. Journal of the Electrochemical Society 2006;153:B499-B505.
[70] Popat KC, Eltgroth M, LaTempa TJ, Grimes CA, Desai TA. Decreased Staphylococcus epidermis adhesion and increased osteoblast functionality on antibiotic-loaded titania nanotubes. Biomaterials 2007;28:4880-8.
[71] Ghicov A, Tsuchiya H, Macak JM, Schmuki P. Titanium oxide nanotubes prepared in phosphate electrolytes. Electrochemistry Communications 2005;7:505-9.
[72] MacAk JM, Sirotna K, Schmuki P. Self-organized porous titanium oxide prepared in Na2SO4/NaF electrolytes. Electrochimica Acta 2005;50:3679-84.
[73] Macak JM, Taveira LV, Tsuchiya H, Sirotna K, Macak J, Schmuki P. Influence of different fluoride containing electrolytes on the formation of self-organized titania nanotubes by Ti anodization. Journal of Electroceramics 2006;16:29-34.
[74] Feng B, Chu X, Chen J, Wang J, Lu X, Weng J. Hydroxyapatite coating on titanium surface with titania nanotube layer and its bond strength to substrate. Journal of Porous Materials 2009;17:453-8.
250
[75] Crawford G, Chawla N, Das K, Bose S, Bandyopadhyay A. Microstructure and deformation behavior of biocompatible TiO2 nanotubes on titanium substrate. Acta Biomaterialia 2007;3:359-67.
[76] Narayanan R, Kwon T-Y, Kim K-H. TiO2 nanotubes from stirred glycerol/NH4F electrolyte: Roughness, wetting behavior and adhesion for implant applications. Materials Chemistry and Physics 2009;117:460-4.
[77] Vasilev K, Poh Z, Kant K, Chan J, Michelmore A, Losic D. Tailoring the surface functionalities of titania nanotube arrays. Biomaterials 2010;31:532-40.
[78] Albu SP, Ghicov A, Macak JM, Schmuki P. 250 µm long anodic TiO2 nanotubes with hexagonal self-ordering. physica status solidi (RRL) – Rapid Research Letters 2007;1:R65-R7.
[79] Tsuchiya H, Macak JM, Taveira L, Balaur E, Ghicov A, Sirotna K, Schmuki P. Self-organized TiO2 nanotubes prepared in ammonium fluoride containing acetic acid electrolytes. Electrochemistry Communications 2005;7:576-80.
[80] Macak JM, Tsuchiya H, Taveira L, Aldabergerova S, Schmuki P. Smooth Anodic TiO2 Nanotubes. Angewandte Chemie International Edition 2005;44:7463-5.
[81] Song YY, Schmidt-Stein F, Bauer S, Schmuki P. Amphiphilic TiO2 nanotube arrays: An actively controllable drug delivery system. Journal of the American Chemical Society 2009;131:4230-2.
[82] Cai Q, Yang L, Yu Y. Investigations on the self-organized growth of TiO2 nanotube arrays by anodic oxidization. Thin Solid Films 2006;515:1802-6.
[83] Cai Q, Paulose M, Varghese OK, Grimes CA. The Effect of Electrolyte Composition on the Fabrication of Self-Organized Titanium Oxide Nanotube Arrays by Anodic Oxidation. Journal of Materials Research 2011;20:230-6.
[84] Xin Y CP. Nanostructured Titania coatings for Biological Applications: Fabrication and Characterization. In: Z S, editor. Biological and Biomedical Coatings Handbook : Processing and Characterization. Hoboken: CRC Press; 2011. p. 203-36.
[85] Crawford GA, Chawla N, Ringnalda J. Processing and microstructure characterization of a novel porous hierarchical TiO2 structure. Journal of Materials Research 2009;24:1683-7.
[86] Crawford GA, Chawla N. Porous hierarchical TiO2 nanostructures: Processing and microstructure relationships. Acta Materialia 2009;57:854-67.
[87] Chawla GACaN. Biocompatible TiO2 nanotube coatings fabricated by anodic oxidation of Ti. In: RL Karlinsey, editor. Recent Developments in Advanced Medical and Dental Materials Using Electrochemical Methodologies. Kerala, India: Research Signpost; 2009. p. 177-99.
[88] Crawford GA, Chawla N. Tailoring TiO2 nanotube growth during anodic oxidation by crystallographic orientation of Ti. Scripta Materialia 2009;60:874-7.
[89] Crawford GA, Chawla N, Houston JE. Nanomechanics of biocompatible TiO2 nanotubes by Interfacial Force Microscopy (IFM). Journal of the Mechanical Behavior of Biomedical Materials 2009;2:580-7.
[90] Moon SK, Kwon JS, Uhm SH, Lee EJ, Gu HJ, Eom TG, Kim KN. Biological evaluation of micro-nano patterned implant formed by anodic oxidation. Current Applied Physics 2014;14:S183-S7.
[91] Bauer S, Park J, Faltenbacher J, Berger S, Von Der Mark K, Schmuki P. Size selective behavior of mesenchymal stem cells on ZrO2 and TiO2 nanotube arrays. Integrative Biology 2009;1:525-32.
251
[92] S. L. Wrought Titanium and Titanium Alloys, Properties and Selection: Nonferrous Alloys and Special-Purpose Materials. ASM Handbook: ASM International; 1990. p. 592–633.
[93] Fukuda A, Takemoto M, Saito T, Fujibayashi S, Neo M, Yamaguchi S, Kizuki T, Matsushita T, Niinomi M, Kokubo T, Nakamura T. Bone bonding bioactivity of Ti metal and Ti-Zr-Nb-Ta alloys with Ca ions incorporated on their surfaces by simple chemical and heat treatments. Acta Biomaterialia 2011;7:1379-86.
[94] Nakada H, Numata Y, Sakae T, Okazaki Y, Tanimoto Y, Tamaki H, Katou T, Ookubo A, Kobayashi K, LeGeros RZ. Comparison of bone mineral density and area of newly formed bone around Ti-15%Zr-4%Nb-4%Ta alloy and Ti-6%A1-4%V alloy implants. Journal of Hard Tissue Biology 2008;17:99-108.
[95] Choe H-C, Saji VS, Ko Y-M. Mechanical properties and corrosion resistance of low rigidity quaternary titanium alloy for biomedical applications. Transactions of Nonferrous Metals Society of China 2009;19:862-5.
[96] Li SJ, Yang R, Li S, Hao YL, Cui YY, Niinomi M, Guo ZX. Wear characteristics of Ti-Nb-Ta-Zr and Ti-6Al-4V alloys for biomedical applications. Wear 2004;257:869-76.
[97] Tsuchiya H, Macak JM, Ghicov A, Schmuki P. Self-Organization of Anodic Nanotubes on Two Size Scales. Small 2006;2:888-91.
[98] Tsuchiya H, Macak JM, Ghicov A, Tang YC, Fujimoto S, Niinomi M, Noda T, Schmuki P. Nanotube oxide coating on Ti-29Nb-13Ta-4.6Zr alloy prepared by self-organizing anodization. Electrochimica Acta 2006;52:94-101.
[99] Tsuchiya H, Macak JM, Ghicov A, Schmuki P. Anodic oxide nanotubes on Ti alloys. ECS Transactions 2007;3:365-74.
[100] Macak JM, Tsuchiya H, Taveira L, Ghicov A, Schmuki P. Self-organized nanotubular oxide layers on Ti-6Al-7Nb and Ti-6Al-4V formed by anodization in NH4F solutions. Journal of Biomedical Materials Research Part A 2005;75A:928-33.
[101] Feng XJ, Macak JM, Albu SP, Schmuki P. Electrochemical formation of self-organized anodic nanotube coating on Ti-28Zr-8Nb biomedical alloy surface. Acta Biomaterialia 2008;4:318-23.
[102] Tsuchiya H, Macak JM, Taveira L, Schmuki P. Formation of self-organized zirconia nanostructure. 208th Meeting of the Electrochemical Society. Los Angeles, CA: ECS Transactions; 2005. p. 351-7.
[103] Choi J, Lim JH, Lee J, Kim KJ. Porous niobium oxide films prepared by anodization-annealing-anodization. Nanotechnology 2007;18:055603 (6pp).
[104] Sieber IV, Schmuki P. Porous Tantalum Oxide Prepared by Electrochemical Anodic Oxidation. Journal of the Electrochemical Society 2005;152:C639-C44.
[105] Tsuchiya H, Nakata J, Fujimoto S, Berger S, Schmuki P. Anodic porous and tubular oxide layers on Ti alloys. 3 ed. Honolulu, HI2008. p. 359-67.
[106] Tsuchiya H, Akaki T, Nakata J, Terada D, Tsuji N, Koizumi Y, Minamino Y, Schmuki P, Fujimoto S. Anodic oxide nanotube layers on Ti-Ta alloys: Substrate composition, microstructure and self-organization on two-size scales. Corrosion Science 2009;51:1528-33.
[107] Tsuchiya H, Akaki T, Nakata J, Terada D, Tsuji N, Koizumi Y, Minamino Y, Schmuki P, Fujimoto S. Metallurgical aspects on the formation of self-organized anodic oxide nanotube layers. Electrochimica Acta 2009;54:5155-62.
252
[108] Wang Y, Tao J, Wang L, He P, Wang T. HA coating on titanium with nanotubular anodized TiO2 intermediate layer via electrochemical deposition. Transactions of Nonferrous Metals Society of China 2008;18:631-5.
[109] Das K, Bose S, Bandyopadhyay A. TiO2 nanotubes on Ti: Influence of nanoscale morphology on bone cell-materials interaction. Journal of Biomedical Materials Research Part A 2009;90A:225-37.
[110] Yasuda K, Schmuki P. Control of morphology and composition of self-organized zirconium titanate nanotubes formed in (NH4)2SO4/NH4F electrolytes. Electrochimica Acta 2007;52:4053-61.
[111] Yasuda K, Schmuki P. Electrochemical formation of self-organized zirconium titanate nanotube multilayers. Electrochemistry Communications 2007;9:615-9.
[112] Yasuda K, Schmuki P. Formation of self-organized zirconium titanate nanotube layers by alloy anodization. Advanced Materials 2007;19:1757-60.
[113] Ghicov A, Aldabergenova S, Tsuchyia H, Schmuki P. TiO2-Nb2O5 nanotubes with electrochemically tunable morphologies. Angewandte Chemie - International Edition 2006;45:6993-6.
[114] M. S. Tantalum-Properties of Pure Metals, Properties and Selection: Nonferrous Alloys and Special-Purpose Materials. ASM Handbook: ASM International; 1990. p. 1099–201.
[115] WS. L. Titanium-Properties of Pure Metals, Properties and Selection: Nonferrous Alloys and Special-Purpose Materials. ASM Handbook: ASM International 1990. p. 1099–201.
[116] Levine BR, Sporer S, Poggie RA, Della Valle CJ, Jacobs JJ. Experimental and clinical performance of porous tantalum in orthopedic surgery. Biomaterials 2006;27:4671-81.
[117] Sieber I, Hildebrand H, Friedrich A, Schmuki P. Initiation of tantalum oxide pores grown on tantalum by potentiodynamic anodic oxidation. Journal of Electroceramics 2006;16:35-9.
[118] Arnould C, Delhalle J, Mekhalif Z. Electrodeposition from ionic liquid of 2D ordered Ta2O5 on titanium substrate through a polystyrene template. Journal of the Electrochemical Society 2009;156:K186-K90.
[119] Q. Lu a SM, Skeldon P, Thompson GE, Masheder D, Habazaki H, Shimizu K. Anodic film growth on tantalum in dilute phosphoric acid solution at 20 and 85 °C. Electrochimica Acta 2002;47:2761-7.
[120] El-Sayed HA, Birss VI. Controlled interconversion of nanoarray of Ta dimples and high aspect ratio Ta oxide nanotubes. Nano Letters 2009;9:1350-5.
[121] Wei W, MacAk JM, Shrestha NK, Schmuki P. Thick self-ordered nanoporous Ta2O5 films with long-range lateral order. Journal of the Electrochemical Society 2009;156:K104-K9.
[122] Wei W, Macak JM, Schmuki P. High aspect ratio ordered nanoporous Ta2O5 films by anodization of Ta. Electrochemistry Communications 2008;10:428-32.
[123] Arnould C, Delhalle J, Mekhalif Z. Multifunctional hybrid coating on titanium towards hydroxyapatite growth: Electrodeposition of tantalum and its molecular functionalization with organophosphonic acids films. Electrochimica Acta 2008;53:5632-8.
[124] Arnould C, Delhalle J, Mekhalif Z. Corrigendum to "Multifunctional hybrid coating on titanium towards hydroxyapatite growth: Electrodeposition of tantalum and its
253
molecular functionalization with organophosphonic acids films" [Electrochim. Acta 53 (2008) 5632-5638] (DOI:10.1016/j.electacta.2008.03.003). Electrochimica Acta 2009;54:2402.
[125] Arnould C, Volcke C, Lamarque C, Thiry PA, Delhalle J, Mekhalif Z. Titanium modified with layer-by-layer sol-gel tantalum oxide and an organodiphosphonic acid: A coating for hydroxyapatite growth. Journal of Colloid and Interface Science 2009;336:497-503.
[126] Aagard RL. Optical waveguide characteristics of reactive dc-sputtered niobium pentoxide films. Applied Physics Letters 1975;27:605-7.
[127] Velten D, Eisenbarth E, Schanne N, Breme J. Biocompatible Nb2O5 thin films prepared by means of the sol-gel process. Journal of Materials Science: Materials in Medicine 2004;15:457-61.
[128] Mozalev A, Sakairi M, Saeki I, Takahashi H. Nucleation and growth of the nanostructured anodic oxides on tantalum and niobium under the porous alumina film. Electrochimica Acta 2003;48:3155-70.
[129] Lu Q, Hashimoto T, Skeldon P, Thompson GE, Habazaki H, Shimizu K. Nanoporous Anodic Niobium Oxide Formed in Phosphate/Glycerol Electrolyte. Electrochemical and Solid-State Letters 2005;8:B17-B20.
[130] Choi J, Lim JH, Lee SC, Chang JH, Kim KJ, Cho MA. Porous niobium oxide films prepared by anodization in HF/H3PO4. Electrochimica Acta 2006;51:5502-7.
[131] Tsuchiya H, Macak JM, Sieber I, Schmuki P. Self-Organized High-Aspect-Ratio Nanoporous Zirconium Oxides Prepared by Electrochemical Anodization. Small 2005;1:722-5.
[132] Tsuchiya H, Macak JM, Taveira L, Schmuki P. Fabrication and characterization of smooth high aspect ratio zirconia nanotubes. Chemical Physics Letters 2005;410:188-91.
[133] Berger S, Faltenbacher J, Bauer S, Schmuki P. Enhanced self-ordering of anodic ZrO2 nanotubes in inorganic and organic electrolytes using two-step anodization. Physica Status Solidi - Rapid Research Letters 2008;2:102-4.
[134] Berger S, Jakubka F, Schmuki P. Formation of hexagonally ordered nanoporous anodic zirconia. Electrochemistry Communications 2008;10:1916-9.
[135] Shin Y, Lee S. A freestanding membrane of highly ordered anodic ZrO2 nanotube arrays. Nanotechnology 2009;20:105301 (5pp).
[136] Lee W-J, Smyrl WH. Zirconium Oxide Nanotubes Synthesized via Direct Electrochemical Anodization. Electrochemical and Solid-State Letters 2005;8:B7-B9.
[137] Anselme K, Ploux L, Ponche A. Cell/Material Interfaces: Influence of Surface Chemistry and Surface Topography on Cell Adhesion. Journal of Adhesion Science and Technology 2010;24:831-52.
[138] Finke B, Luethen F, Schroeder K, Mueller PD, Bergemann C, Frant M, Ohl A, Nebe BJ. The effect of positively charged plasma polymerization on initial osteoblastic focal adhesion on titanium surfaces. Biomaterials 2007;28:4521-34.
[139] Lotz MM, Burdsal CA, Erickson HP, McClay DR. Cell adhesion to fibronectin and tenascin: Quantitative measurements of initial binding and subsequent strengthening response. Journal of Cell Biology 1989;109:1795-805.
[140] Bausch AR, Ziemann F, Boulbitch AA, Jacobson K, Sackmann E. Local measurements of viscoelastic parameters of adherent cell surfaces by magnetic bead microrheometry. Biophysical Journal 1998;75:2038-49.
254
[141] Yamamoto A, Mishima S, Maruyama N, Sumita M. A new technique for direct measurement of the shear force necessary to detach a cell from a material. Biomaterials 1998;19:871-9.
[142] Bowers VM, Fischer LR, Francis GW, Williams KL. A micromechanical technique for monitoring cell-substrate adhesiveness: measurements of the strength of red blood cell adhesion to glass and polymer test surfaces. Journal of Biomedical Materials Research 1989;23:1453-73.
[143] Truskey GA, Proulx TL. Relationship between 3T3 cell spreading and the strength of adhesion on glass and silane surfaces. Biomaterials 1993;14:243-54.
[144] Mege JL, Capo C, Benoliel AM, Bongrand P. Determination of binding strength and kinetics of binding initiation. A model study made on the adhesive properties of P388D1 macrophage-like cells. Cell Biophysics 1986;8:141-60.
[145] Arvidsson A, Franke-Stenport V, Andersson M, Kjellin P, Sul Y-T, Wennerberg A. Formation of calcium phosphates on titanium implants with four different bioactive surface preparations. An in vitro study. Journal of Materials Science: Materials in Medicine 2007;18:1945-54.
[146] Bruckert F, Weidenhaupt M. Applications of Micro- and Nano-technology to Study Cell Adhesion to Material Surfaces. Journal of Adhesion Science and Technology 2010;24:2127-40.
[147] Kabaso D, Gongadze E, Perutková Š, Matschegewski C, Kralj-Iglič V, Beck U, van Rienen U, Iglič A. Mechanics and electrostatics of the interactions between osteoblasts and titanium surface. Computer Methods in Biomechanics and Biomedical Engineering 2011;14:469-82.
[148] Li B, Li Y, Li J, Fu X, Li C, Wang H, Liu S, Guo L, Xin S, Liang C, Li H. Improvement of biological properties of titanium by anodic oxidation and ultraviolet irradiation. Applied Surface Science 2014;307:202-8.
[149] Decuzzi P, Ferrari M. Modulating cellular adhesion through nanotopography. Biomaterials 2010;31:173-9.
[150] Gentile F, Tirinato L, Battista E, Causa F, Liberale C, di Fabrizio EM, Decuzzi P. Cells preferentially grow on rough substrates. Biomaterials 2010;31:7205-12.
[151] Kim S, Lim J, Lee S, Nam S, Kang H, Choi J. Anodically nanostructured titanium oxides for implant applications. Electrochimica Acta 2008;53:4846-51.
[152] Engler AJ, Sen S, Sweeney HL, Discher DE. Matrix Elasticity Directs Stem Cell Lineage Specification. Cell 2006;126:677-89.
[153] Pelham Jr RJ, Wang YL. Cell locomotion and focal adhesions are regulated by substrate flexibility. Proceedings of the National Academy of Sciences of the United States of America 1997;94:13661-5.
[154] Kim TI, Jang JH, Kim HW, Knowles JC, Ku Y. Biomimetic approach to dental implants. Current Pharmaceutical Design 2008;14:2201-11.
[155] Wang HB, Dembo M, Wang YL. Substrate flexibility regulates growth and apoptosis of normal but not transformed cells. American Journal of Physiology - Cell Physiology 2000;279:C1345-C50.
[156] Chen G, Wen X, Zhang N. Corrosion resistance and ion dissolution of titanium with different surface microroughness. Bio-Medical Materials and Engineering 1998;8:61-74.
255
[157] Anselme K, Davidson P, Popa AM, Giazzon M, Liley M, Ploux L. The interaction of cells and bacteria with surfaces structured at the nanometre scale. Acta Biomaterialia 2010;6:3824-46.
[158] Jung YL, Donahue HJ. Cell sensing and response to micro- and nanostructured surfaces produced by chemical and topographic patterning. Tissue Engineering 2007;13:1879-91.
[159] Lord MS, Foss M, Besenbacher F. Influence of nanoscale surface topography on protein adsorption and cellular response. Nano Today 2010;5:66-78.
[160] Li JR, Shi L, Deng Z, Lo SH, Liu GY. Nanostructures of designed geometry and functionality enable regulation of cellular signaling processes. Biochemistry 2012;51:5876-93.
[161] Yim EKF, Leong KW. Significance of synthetic nanostructures in dictating cellular response. Nanomedicine: Nanotechnology, Biology and Medicine 2005;1:10-21.
[162] Kim WG, Choe HC. Nanostructure and corrosion behaviors of nanotube formed Ti-Zr alloy. Transactions of Nonferrous Metals Society of China (English Edition) 2009;19:1005-8.
[163] Jeong Y-H, Lee K, Choe H-C, Ko Y-M, Brantley WA. Nanotube formation and morphology change of Ti alloys containing Hf for dental materials use. Thin Solid Films 2009;517:5365-9.
[164] Jang SH, Choe HC, Ko YM, Brantley WA. Electrochemical characteristics of nanotubes formed on Ti-Nb alloys. Thin Solid Films 2009;517:5038-43.
[165] Capellato P, Smith BS, Popat KC, Claro APRA. Fibroblast functionality on novel Ti30Ta nanotube array. Materials Science and Engineering C 2012;32:2060-7.
[166] Ding D, Ning C, Huang L, Jin F, Hao Y, Bai S, Li Y, Li M, Mao D. Anodic fabrication and bioactivity of Nb-doped TiO2 nanotubes. Nanotechnology 2009;20:305103 (6pp).
[167] Choe HC. Nanotubular surface and morphology of Ti-binary and Ti-ternary alloys for biocompatibility. Thin Solid Films 2011;519:4652-7.
[168] Li Z, Ning C, Ding D, Liu H, Huang L. Biological properties of Ti-Nb-Zr-O nanostructures grown on Ti35Nb5Zr alloy. Journal of Nanomaterials 2012;2012:834042-9.
[169] Mîndroiu M, Pirvu C, Ion R, Demetrescu I. Comparing performance of nanoarchitectures fabricated by Ti6Al7Nb anodizing in two kinds of electrolytes. Electrochimica Acta 2010;56:193-202.
[170] Mei S, Zhao L, Wang W, Ma Q, Zhang Y. Biomimetic titanium alloy with sparsely distributed nanotubes could enhance osteoblast functions. Advanced Engineering Materials 2012;14:B166-B74.
[171] Oh S, Brammer KS, Li YSJ, Teng D, Engler AJ, Chien S, Jin S. Reply to von der Mark et al.: Looking further into the effects of nanotube dimension on stem cell fate. Proceedings of the National Academy of Sciences of the United States of America 2009;106.
[172] Oh S, Jin S. Titanium oxide nanotubes with controlled morphology for enhanced bone growth. Materials Science and Engineering: C 2006;26:1301-6.
[173] Wang N, Li H, Lü W, Li J, Wang J, Zhang Z, Liu Y. Effects of TiO2 nanotubes with different diameters on gene expression and osseointegration of implants in minipigs. Biomaterials 2011;32:6900-11.
256
[174] Li Y, Ding D, Ning C, Bai S, Huang L, Li M, Mao D. Thermal stability and in vitro bioactivity of Ti-Al-V-O nanostructures fabricated on Ti6Al4V alloy. Nanotechnology 2009;20:065708 (6pp).
[175] Choe HC, Kim WG, Jeong YH. Surface characteristics of HA coated Ti-30Ta-xZr and Ti-30Nb-xZr alloys after nanotube formation. Surface and Coatings Technology 2010;205:S305-S11.
[176] Jeong YH, Kim WG, Park GH, Choe HC, Ko YM. Surface characteristics of HA coated Ti-Hf binary alloys after nanotube formation. Transactions of Nonferrous Metals Society of China (English Edition) 2009;19:852-6.
[177] Yu WQ, Zhang YL, Jiang XQ, Zhang FQ. In vitro behavior of MC3T3-E1 preosteoblast with different annealing temperature titania nanotubes. Oral Diseases 2010;16:624-30.
[178] Mazǎre A, Voicu G, Truscǎ R, Ioniţǎ D. Heat treatment of TiO2 nanotubes, a way to significantly change their behaviour. UPB Scientific Bulletin, Series B: Chemistry and Materials Science 2011;73:97-108.
[179] Aquino JM, Rocha-Filho RC, Bocchi N, Biaggio SR. Microwave-assisted crystallization into anatase of amorphous TiO2 nanotubes electrochemically grown on a Ti substrate. Materials Letters 2014;126:52-4.
[180] Shin DH, Shokuhfar T, Choi CK, Lee SH, Friedrich C. Wettability changes of TiO2 nanotube surfaces. Nanotechnology 2011;22:315704 (7pp).
[181] Hamlekhan A, Butt A, Patel S, Royhman D, Takoudis C, Sukotjo C, Yuan J, Jursich G, Mathew MT, Hendrickson W, Virdi A, Shokuhfar T. Fabrication of anti-aging TiO2 nanotubes on biomedical Ti alloys. PLoS ONE 2014;9:e96213.
[182] Khor KA, Dong ZL, Quek CH, Cheang P. Microstructure investigation of plasma sprayed HA/Ti6Al4V composites by TEM. Materials Science and Engineering A 2000;281:221-8.
[183] Kim WG, Choe HC. Surface characteristics of hydroxyapatite/titanium composite layer on the Ti-35Ta-xZr surface by RF and DC sputtering. Thin Solid Films 2011;519:7045-9.
[184] Lee K, Choe HC, Kim BH, Ko YM. The biocompatibility of HA thin films deposition on anodized titanium alloys. Surface and Coatings Technology 2010;205:S267-S70.
[185] Choe HC. Photofunctionalization of EB-PVD HA-coated nano-pore surface of Ti-30Nb-xZr alloy for dental implants. Surface and Coatings Technology 2013;228:S470-S6.
[186] Ma J, Liang CH, Kong LB, Wang C. Colloidal characterization and electrophoretic deposition of hydroxyapatite on titanium substrate. Journal of Materials Science: Materials in Medicine 2003;14:797-801.
[187] Parcharoen Y, Kajitvichyanukul P, Sirivisoot S, Termsuksawad P. Hydroxyapatite electrodeposition on anodized titanium nanotubes for orthopedic applications. Applied Surface Science 2014;311:54-61.
[188] Chen J, Zhang Z, Ouyang J, Chen X, Xu Z, Sun X. Bioactivity and osteogenic cell response of TiO2 nanotubes coupled with nanoscale calcium phosphate via ultrasonification-assisted electrochemical deposition. Applied Surface Science 2014;305:24-32.
257
[189] Kodama A, Bauer S, Komatsu A, Asoh H, Ono S, Schmuki P. Bioactivation of titanium surfaces using coatings of TiO2 nanotubes rapidly pre-loaded with synthetic hydroxyapatite. Acta Biomaterialia 2009;5:2322-30.
[190] Kim EJ, Jeong YH, Choe HC, Brantley WA. Surface phenomena of HA/TiN coatings on the nanotubular-structured beta Ti-29Nb-5Zr alloy for biomaterials. Applied Surface Science 2012;258:2083-7.
[191] Jeong YH, Choe HC, Eun SW. Hydroxyapatite coating on the Ti-35Nb-xZr alloy by electron beam-physical vapor deposition. Thin Solid Films 2011;519:7050-6.
[192] Raja KS, Misra M, Paramguru K. Deposition of calcium phosphate coating on nanotubular anodized titanium. Materials Letters 2005;59:2137-41.
[193] Kar A, Raja KS, Misra M. Electrodeposition of hydroxyapatite onto nanotubular TiO2 for implant applications. Surface and Coatings Technology 2006;201:3723-31.
[194] Wang LN, Luo JL. Formation of hydroxyapatite coating on anodic titanium dioxide nanotubes via an efficient dipping treatment. Metallurgical and Materials Transactions A: Physical Metallurgy and Materials Science 2011;42:3255-64.
[195] Zhao L, Mei S, Chu PK, Zhang Y, Wu Z. The influence of hierarchical hybrid micro/nano-textured titanium surface with titania nanotubes on osteoblast functions. Biomaterials 2010;31:5072-82.
[196] Xiao J, Zhou H, Zhao L, Sun Y, Guan S, Liu B, Kong L. The effect of hierarchical micro/nanosurface titanium implant on osseointegration in ovariectomized sheep. Osteoporosis International 2011;22:1907-13.
[197] Sul YT. Electrochemical growth behavior, surface properties, and enhanced in vivo bone response of Tio2 nanotubes on microstructured surfaces of blasted, screw-shaped titanium implants. International Journal of Nanomedicine 2010;5:87-100.
[198] Von Wilmowsky C, Bauer S, Roedl S, Neukam FW, Schmuki P, Schlegel KA. The diameter of anodic TiO2 nanotubes affects bone formation and correlates with the bone morphogenetic protein-2 expression in vivo. Clinical Oral Implants Research 2012;23:359-66.
[199] Popat KC, Leoni L, Grimes CA, Desai TA. Influence of engineered titania nanotubular surfaces on bone cells. Biomaterials 2007;28:3188-97.
[200] Bjursten LM, Rasmusson L, Oh S, Smith GC, Brammer KS, Jin S. Titanium dioxide nanotubes enhance bone bonding in vivo. Journal of Biomedical Materials Research - Part A 2010;92:1218-24.
[201] Peremarch CPJ, Tanoira RP, Arenas MA, Matykina E, Conde A, De Damborenea JJ, Barrena EG, Esteban J. Bacterial adherence to anodized titanium alloy. Journal of Physics: Conference Series 2010;252.
[202] Yu WQ, Jiang XQ, Xu L, Zhao YF, Zhang FQ, Cao X. Osteogenic gene expression of canine bone marrow stromal cell and bacterial adhesion on titanium with different nanotubes. Journal of Biomedical Materials Research - Part B Applied Biomaterials 2011;99 B:207-16.
[203] Qi YM, Ma BH, Geng YJ, Deng JY, Cui CX. Nano-surface modifacation of biomedical β Titanium alloy for dental implant and its antibacterial property. Advanced Materials Research 2014;904: 142-5.
[204] Grigorescu S, Ungureanu C, Kirchgeorg R, Schmuki P, Demetrescu I. Various sized nanotubes on TiZr for antibacterial surfaces. Applied Surface Science 2013;270:190-6.
258
[205] Zhao L, Wang H, Huo K, Cui L, Zhang W, Ni H, Zhang Y, Wu Z, Chu PK. Antibacterial nano-structured titania coating incorporated with silver nanoparticles. Biomaterials 2011;32:5706-16.
[206] Lin C, Tang P, Zhang W, Wang Y, Zhang B, Wang H, Zhang L. Effect of superhydrophobic surface of titanium on staphylococcus aureus adhesion. Journal of Nanomaterials 2011;2011: 178921 (8 pp).
[207] Ercan B, Taylor E, Alpaslan E, Webster TJ. Diameter of titanium nanotubes influences anti-bacterial efficacy. Nanotechnology 2011;22:295102 (11pp).
[208] Mendenhall W, Sincich T. Continuous random variables. In: S Yagan, editor. Statistics for engineering and the sciences. 5th ed. ed. New Jersey, USA: Prentice Hall; 2007. p. 168-210.
[209] Rinne H. Definition and properties of the WEIBULL distribution. The Weibull Distribution: A Handbook. United Kingdom: Chapman and Hall/CRC; 2008. p. 27-97.
[210] Gadelmawla ES, Koura MM, Maksoud TMA, Elewa IM, Soliman HH. Roughness parameters. Journal of Materials Processing Technology 2002;123:133-45.
[211] Owens DK, Wendt RC. Estimation of the surface free energy of polymers. Journal of Applied Polymer Science 1969;13:1741-7.
[212] Dann JR. Forces involved in the adhesive process. I. Critical surface tensions of polymeric solids as determined with polar liquids. Journal of Colloid and Interface Science 1970;32:302-20.
[213] Oyane A, Onuma K, Ito A, Kim HM, Kokubo T, Nakamura T. Formation and growth of clusters in conventional and new kinds of simulated body fluids. Journal of Biomedical Materials Research - Part A 2003;64:339-48.
[214] Kokubo T, Takadama H. How useful is SBF in predicting in vivo bone bioactivity? Biomaterials 2006;27:2907-15.
[215] Fischer-Cripps AC. Critical review of analysis and interpretation of nanoindentation test data. Surface and Coatings Technology 2006;200:4153-65.
[216] Oliver WC, Pharr GM. Measurement of hardness and elastic modulus by instrumented indentation: Advances in understanding and refinements to methodology. Journal of Materials Research 2004;19:3-20.
[217] Li H, Khor KA, Cheang P. Impact formation and microstructure characterization of thermal sprayed hydroxyapatite/titania composite coatings. Biomaterials 2003;24:949-57.
[218] Halka M, Nordstrom B. Transition Metals. New York: Facts On File; 2010. [219] Campbell FC. Elements of Metallurgy and Engineering Alloys. Materials Park, OH,
USA: ASM International; 2008. [220] Park HH, Park IS, Kim KS, Jeon WY, Park BK, Kim HS, Bae TS, Lee MH.
Bioactive and electrochemical characterization of TiO2 nanotubes on titanium via anodic oxidation. Electrochimica Acta 2010;55:6109-14.
[221] Oh S, Brammer KS, Moon K-S, Bae J-M, Jin S. Influence of sterilization methods on cell behavior and functionality of osteoblasts cultured on TiO2 nanotubes. Materials Science and Engineering: C 2011;31:873-9.
[222] H. Nielsen R, H. Schlewitz J, Nielsen H, Updated by S. Zirconium and Zirconium Compounds. Kirk-Othmer Encyclopedia of Chemical Technology: John Wiley & Sons, Inc.; 2000.
259
[223] Webster RT, Albany TWC. Zirconium-Properties of Pure Metals, Properties and Selection: Nonferrous Alloys and Special-Purpose Materials. ASM Handbook: ASM International 1990. p. 1099–201.
[224] Yamaguchi T. Application of ZrO2 as a catalyst and a catalyst support. Catalysis Today 1994;20:199-217.
[225] Bethke KA, Kung MC, Yang B, Shah M, Alt D, Li C, Kung HH. Metal oxide catalysts for lean NOx reduction. Catalysis Today 1995;26:169-83.
[226] Hahn R, Berger S, Schmuki P. Bright visible luminescence of self-organized ZrO2 nanotubes. Journal of Solid State Electrochemistry 2010;14:285-8.
[227] Agarwal P, Paramasivam I, Shrestha NK, Schmuki P. MoO3 in self-organized TiO2 nanotubes for enhanced photocatalytic activity. Chemistry - An Asian Journal 2010;5:66-9.
[228] Paramasivam I, Nah YC, Das C, Shrestha NK, Schmuki P. WO3/TiO2 nanotubes with strongly enhanced photocatalytic activity. Chemistry - A European Journal 2010;16:8993-7.
[229] Liu G, Lu H, Chen Z, Li F, Wang L, Watts J, Lu GQ, Cheng HM. Ti-Zr-O nanotube arrays with controlled morphology, crystal structure and optical properties. Journal of Nanoscience and Nanotechnology 2009;9:6501-10.
[230] Kobayashi E, Matsumoto S, Doi H, Yoneyama T, Hamanaka H. Mechanical properties of the binary titanium-zirconium alloys and their potential for biomedical materials. Journal of Biomedical Materials Research 1995;29:943-50.
[231] Li Y, Wong C, Xiong J, Hodgson P, Wen C. In vitro cytotoxicity of binary TI alloys for bone implants. Gold Coast, QLD2009. p. 295-8.
[232] Kim WG, Choe HC, Ko YM, Brantley WA. Electrochemical characteristics of nanotube formed Ti-Zr alloy. Quebec City, QC2008. p. 462-5.
[233] Kim WG, Choe HC, Ko YM, Brantley WA. Nanotube morphology changes for Ti-Zr alloys as Zr content increases. Thin Solid Films 2009;517:5033-7.
[234] Minagar S, Berndt CC, Wang J, Ivanova E, Wen C. A review of the application of anodization for the fabrication of nanotubes on metal implant surfaces. Acta Biomaterialia 2012;8:2875-88.
[235] Wang Y, Wen C, Hodgson P, Li Y. Biocompatibility of TiO2 nanotubes with different topographies. Journal of Biomedical Materials Research - Part A 2013.
[236] Chemical Properties of Materials. In: CRC Materials Science and Engineering Handbook, ed. J. F. S. a. W. Alexander, CRC Press 2000, Boca Raton, 2000, ch. 9, tbl. 309.
[237] Thermodynamic and Kinetic Data. In: CRC Materials Science and Engineering Handbook, ed. J. F. S. a. W. Alexander, CRC Press 2000, Boca Raton, 2000, ch. 4, tbl. 67.
[238] Taveira LV, Macák JM, Tsuchiya H, Dick LFP, Schmuki P. Initiation and growth of self-organized TiO2 nanotubes anodically formed in NH4F/(NH4)2SO4 electrolytes. Journal of the Electrochemical Society 2005;152:B405-B10.
[239] Madou MJ, Kinoshita K. Electrochemical measurements on metal oxide electrodes-I. Zirconium dioxide. Electrochimica Acta 1984;29:411-7.
[240] Minagar S, Berndt CC, Gengenbach T, Wen C. Fabrication and characterization of TiO2-ZrO2-ZrTiO4 nanotubes on TiZr alloy manufactured via anodization. Journal of Materials Chemistry B 2014;2:71-83.
260
[241] EL-010 SASNZC. Guidance on the measurement of wettability of insulator surfaces. GPO Box 5420, Sydney, NSW 2001, Australia: Standard Australia; 2005. p. 15.
[242] Webb HK, Hasan J, Truong VK, Crawford RJ, Ivanova EP. Nature inspired structured surfaces for biomedical applications. Current Medicinal Chemistry 2011;18:3367-75.
[243] Ikawa H, Yamada T, Kojima K, Matsumoto S. X-Ray Photoelectron-Spectroscopy Study of High-Temperature and Low-Temperature Forms of Zirconium Titanate. Journal of the American Ceramic Society 1991;74:1459-62.
[244] Tsai CC, Teng H. Regulation of the physical characteristics of titania nanotube aggregates synthesized from hydrothermal treatment. Chemistry of Materials 2004;16:4352-8.
[245] Tian ZR, Voigt JA, Liu J, McKenzie B, Xu H. Large oriented arrays and continuous films of TiO2-based nanotubes. Journal of the American Chemical Society 2003;125:12384-5.
[246] Sander MS, Côté MJ, Gu W, Kile BM, Tripp CP. Template-assisted fabrication of dense, aligned arrays of titania nanotubes with well-controlled dimensions on substrates. Advanced Materials 2004;16:2052-7.
[247] Macak J, Schmuki P. Anodic growth of self-organized anodic TiO2 nanotubes in viscous electrolytes. Electrochimica Acta 2006;52:1258-64.
[248] Kim D, Schmidt-Stein F, Hahn R, Schmuki P. Gravity assisted growth of self-organized anodic oxide nanotubes on titanium. Electrochemistry Communications 2008;10:1082-6.
[249] Yasuda K, Macak JM, Berger S, Ghicov A, Schmuki P. Mechanistic aspects of the self-organization process for oxide nanotube formation on valve metals. Journal of the Electrochemical Society 2007;154:C472-C8.
[250] Weiner S, Traub W, Wagner HD. Lamellar bone: Structure-function relations. Journal of Structural Biology 1999;126:241-55.
[251] Kim HM, Himeno T, Kawashita M, Kokubo T, Nakamura T. The mechanism of biomineralization of bone-like apatite on synthetic hydroxyapatite: an in vitro assessment. Journal of The Royal Society Interface 2004;1:17-22.
[252] Chen X, Nouri A, Li Y, Lin J, Hodgson PD, Wen Ce. Effect of surface roughness of Ti, Zr, and TiZr on apatite precipitation from simulated body fluid. Biotechnology and Bioengineering 2008;101:378-87.
[253] Michelle Grandin H, Berner S, Dard M. A review of Titanium Zirconium (TiZr) alloys for use in endosseous dental implants. Materials 2012;5:1348-60.
[254] Kunze J, Muller L, Macak J, Greil P, Schmuki P, Muller F. Time-dependent growth of biomimetic apatite on anodic TiO2 nanotubes. Electrochimica Acta 2008;53:6995-7003.
[255] Points of Zero Charge. Chemical Properties of Material Surfaces: CRC Press; 2001. p. 731-44.
[256] Surface Charging in Absence of Strongly Adsorbing Species. Chemical Properties of Material Surfaces: CRC Press; 2001. p. 65-309.
[257] Ma Q, Li M, Hu Z, Chen Q, Hu W. Enhancement of the bioactivity of titanium oxide nanotubes by precalcification. Materials Letters 2008;62:3035-8.
[258] Wang XJ, Chen XB, Hodgson PD, Wen CE. Elastic modulus and hardness of cortical and trabecular bovine bone measured by nanoindentation. Transactions of Nonferrous Metals Society of China (English Edition) 2006;16:s744-s8.
261
[259] Anselme K, Bigerelle M. Statistical demonstration of the relative effect of surface chemistry and roughness on human osteoblast short-term adhesion. Journal of Materials Science: Materials in Medicine 2006;17:471-9.
[260] Zhang W, Xi Z, Li G, Wang Q, Tang H, Liu Y, Zhao Y, Jiang L. Highly ordered coaxial bimodal nanotube arrays prepared by self-organizing anodization on Ti alloy. Small 2009;5:1742-6.
[261] Popat KC, Chalvanichkul KI, Barnes GL, Latempa Jr TJ, Grimes CA, Desai TA. Osteogenic differentiation of marrow stromal cells cultured on nanoporous alumina surfaces. Journal of Biomedical Materials Research - Part A 2007;80:955-64.
[262] Anselme K, Bigerelle M. Topography effects of pure titanium substrates on human osteoblast long-term adhesion. Acta Biomaterialia 2005;1:211-22.
[263] Anselme K, Linez P, Bigerelle M, Le Maguer D, Le Maguer A, Hardouin P, Hildebrand HF, Iost A, Leroy JM. The relative influence of the topography and chemistry of TiAl6V4 surfaces on osteoblastic cell behaviour. Biomaterials 2000;21:1567-77.
[264] Anselme K, Ponche A, Bigerelle M. Relative influence of surface topography and surface chemistry on cell response to bone implant materials. Part 2: Biological aspects. Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine 2010;224:1487-507.
[265] Balaur E, Macak JM, Tsuchiya H, Schmuki P. Wetting behaviour of layers of TiO2 nanotubes with different diameters. Journal of Materials Chemistry 2005;15:4488.
[266] Park J, Bauer S, Schlegel KA, Neukam FW, Von Mark KD, Schmuki P. TiO2 nanotube surfaces: 15 nm - an optimal length scale of surface topography for cell adhesion and differentiation. Small 2009;5:666-71.
[267] Jha H, Hahn R, Schmuki P. Ultrafast oxide nanotube formation on TiNb, TiZr and TiTa alloys by rapid breakdown anodization. Electrochimica Acta 2010;55:8883-7.
[268] Grigorescu S, Pruna V, Titorencu I, Jinga VV, Mazare A, Schmuki P, Demetrescu I. The two step nanotube formation on TiZr as scaffolds for cell growth. Bioelectrochemistry 2014;98:39-45.
[269] Inc. VI. Vision for profiler. 4.20 ed. 100 Sunnyside Blvd., Woodbury, NY 11797: Veeco Instruments Inc.; 2009.
[270] Webster TJ, Ergun C, Doremus RH, Siegel RW, Bizios R. Enhanced functions of osteoblasts on nanophase ceramics. Biomaterials 2000;21:1803-10.
[271] Peng P, Kumar S, Voelcker NH, Szili E, Smart RSC, Griesser HJ. Thin calcium phosphate coatings on titanium by electrochemical deposition in modified simulated body fluid. Journal of Biomedical Materials Research - Part A 2006;76:347-55.
[272] Anselme K, Bigerelle M. Role of materials surface topography on mammalian cell response. International Materials Reviews 2011;56:243-66.
[273] Burridge K, Chrzanowska-Wodnicka M. Focal adhesions, contractility, and signaling. Annual Review of Cell and Developmental Biology; 1996. p. 463-519.
[274] Pierres A, Benoliel AM, Bongrand P. Cell fitting to adhesive surfaces: A prerequisite to firm attachment andsubsequent events. European Cells and Materials 2002;3:31-45.
[275] Ishikawa M, Sugimoto K. Multiple layer protective coating composed of Ta2O5, Cr2O3 and Al2O3 produced by the MO-CVD technique. Corrosion Engineering 1989;38:619-30.
262
[276] Arnould C, Volcke C, Lamarque C, Thiry PA, Delhalle J, Mekhalif Z. Titanium modified with layer-by-layer sol–gel tantalum oxide and an organodiphosphonic acid: A coating for hydroxyapatite growth. Journal of Colloid and Interface Science 2009;336:497-503.
[277] Terui H, Kobayashi M. Refractive-index-adjustable SiO2-Ta2O5 films for integrated optical circuits. Applied Physics Letters 1978;32:666-8.
[278] Cava RF, Peck WF, Krajewski JJ. Enhancement of the dielectric-constant of Ta2O5 through substitution with TiO2. Nature 1995;377:215-7.
[279] Sieber I, Kannan B, Schmuki P. Self-assembled porous tantalum oxide prepared in H2SO4/HF electrolytes. Electrochemical and Solid State Letters 2005;8:J10-J2.
[280] Chambon L, Maleysson C, Pauly A, Germain JP, Demarne V, Grisel A. Investigation, for NH3 gas sensing applications, of the Nb2O5 semiconducting oxide in the presence of interferent species such as oxygen and humidity. Sensors and Actuators B: Chemical 1997;45:107-14.
[281] Tanabe K, Okazaki S. Various reactions catalyzed by niobium compounds and materials. Applied Catalysis A, General 1995;133:191-218.
[282] Pawlicka A, Atik M, Aegerter MA. Synthesis of multicolor Nb2O5 coatings for electrochromic devices. Thin Solid Films 1997;301:236-41.
[283] Choi J, Lim JH, Lee J, Kim KJ. Porous niobium oxide films prepared by anodization–annealing–anodization. Nanotechnology 2007;18:055603 (6pp).
[284] Stefanov P, Stoychev D, Stoycheva M, Ikonomov J, Marinova T. XPS and SEM characterization of zirconia thin films prepared by electrochemical deposition. Surface and Interface Analysis 2000;30:628-31.
[285] Kim HM, Kaneko H, Kokubo T, Miyazaki T, Nakamura T. Mechanism of apatite formation on bioactive tantalum metal in simulated body fluid. Key Engineering Materials 2003; 240-242:11-14.
[286] Nagarajan S, Raman V, Rajendran N. Synthesis and electrochemical characterization of porous niobium oxide coated 316L SS for orthopedic applications. Materials Chemistry and Physics 2010;119:363-6.
[287] Wang XJ, Xiong JY, Li YC, Hodgson PD, Wen CE. Apatite formation on nano-structured titanium and niobium surface. Materials Science Forum 2009;614:85-92.
[288] Godley R, Starosvetsky D, Gotman I. Bonelike apatite formation on niobium metal treated in aqueous NaOH. Journal of Materials Science: Materials in Medicine 2004;15:1073-7.
[289] Guo L, Zhao J, Wang X, Xu R, Lu Z, Li Y. Bioactivity of zirconia nanotube arrays fabricated by electrochemical anodization. Materials Science and Engineering C 2009;29:1174-7.
[290] Guo L, Zhao J, Wang X, Xu X, Liu H, Li Y. Structure and bioactivity of zirconia nanotube arrays fabricated by anodization. International Journal of Applied Ceramic Technology 2009;6:636-41.
[291] Wang LN, Adams A, Luo JL. Enhancement of the capability of hydroxyapatite formation on Zr with anodic ZrO2 nanotubular arrays via an effective dipping pretreatment. Journal of Biomedical Materials Research - Part B Applied Biomaterials 2011;99 B:291-301.
[292] Yoo JE, Choi J. Electrochemical surface enlargement of a niobium foil for electrolytic capacitor applications. Electrochemistry Communications 2011;13:298-301.