Development of Capacitive Electrodes
Ana Teresa Ramalho Cordeiro
Thesis to obtain the Master of Science Degree in
Biomedical Engineering
Supervisors:
Professor Jorge Manuel Ferreira Morgado
Professor Maria Isabel de Sousa Rocha
Examination Committee
Chairperson:
Supervisor:
Professor João Pedro Estrela Rodrigues Conde
Professor Jorge Manuel Ferreira Morgado
Member of the Committee:
Professor Francisco André Corrêa Alegria
September 2014
ii
Acknowledgments
I would like to express my sincere gratitude to my supervisor Professor Jorge Morgado for the useful
comments and remarks, his enormous patience and motivation through the learning process of this
master thesis, as well as to Professor Isabel Rocha. Furthermore, I would like to thank Professor Raúl
Martins for his guidance, enthusiasm, immense knowledge as well for the continous support on the way.
Also, I would like to thank my lab mates who willingly shared their precious time whenever I needed
some help. I would like to thank to my family and friends who have always supported me, whether when
making fun of my stressful personality or just by being there.
Thank you
iii
iv
Resumo
O presente trabalho aborda a implementação de um processo para a fabricação de eléctrodos capacitivos.
O procedimento usa moldes de óxido de alumı́nio (OA) nanoporosos preenchidos com nı́quel. Após
a remoção do molde de alumina, os eléctrodos de nı́quel tem a superfı́cie coberta com nanofios. Por
cima, dois materiais dieléctricos, álcool polivinı́lico e dióxido de titânio, foram depositados para aumentar
ainda mais a sua capacidade.
No desenvolvimento dos moldes de alumina nanoporosa, os estudos iniciais focaram-se no controle
da morfologia. De seguida, os eléctrodos nanoestruturados (Eléctrodos I) foram caracterizados em
termos da capacidade eléctrica. O maior aumento da capacidade foi para os eléctrodos fabricados com
um molde de OA poroso obtido através de dupla anodização com corrente constante de 70 mA. Este
aumento variou com a frequência, sugerindo a existência de um sistema de dispersivo. O aumento foi
de 25.29 % e 32.73 % para 100 Hz e 1 kHz, respectivamente, em relação a um eléctrodo de referência
(a 1 mm).
Um segundo tipo de eléctrodos (Eléctrodos II) foi desenvolvido por um processo mais simples que
consiste num filme de uma dispersão de nanotubos de carbono no polı́mero condutor PEDOT:PSS num
contacto de alumı́nio. O objectivo era verificar se esta abordagem resultava no aumento da rugosidade
para que a área superficial aumentasse, uma vez que seria mais simples do que o processo anterior
usado para Eléctrodos I. A sua capacidade foi medida. Não se verificou um aumento significativo,
sugerindo que o conteúdo de nanotubos de carbono era demasiado baixo.
Palavras-chave: Eléctrodos capacitivos, alumina anodizada porosa
v
vi
Abstract
The present work addresses the implementation of a procedure for the manufacture of capacitive electrodes.
That procedure used anodized aluminum oxide (AAO) porous templates with an electrochemical nickel
filler. After removal from the alumina template, the nickel electrodes have the surface covered with
nanowires. On the top of this, two dielectric materials, polyvinyl alcohol and titanium dioxide, were
deposited to further increase their capacitance.
In the development of the anodized porous alumina templates, the studies were initially focused on the
morphology control. Afterwards, nanostructured electrodes (Electrodes I) were prepared and characterized in terms of capacitance. The highest increase of the capacitance value was found for electrodes
fabricated with a nanoporous AAO template obtained through double-step anodization at constant current of 70 mA. This capacitance increase varied with the frequency, suggesting the existence of a dispersive system, and the increase was a 25.29 % and 32.73 % increase for 100 Hz and 1 kHz, respectively,
with respect to a reference electrode, at the minimum distance.
Another type of electrodes (Electrode II) was developed by a simpler process consisting in depositing
a dispersion of carbon nanotubes in a conducting polymer PEDOT:PSS on an aluminum contact. The
objective was to verify if this approach resulted in yielded roughness so that the surface area would
significantly increase, since it would be simpler than the development process used for Electrodes I.
Their electric capacitance was measured. It was not found a significant increase of the capacitance
suggesting that the carbon nanotubes content was too low.
Keywords: Capacitive electrodes, anodized aluminum oxide
vii
viii
Contents
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
iii
Resumo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
v
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
vii
List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
xi
List of Figures
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
xv
Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1
Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1
1 Motivation
1
2 Theoretical Foundations and Literature Review
5
2.1 Anodic Aluminum Oxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5
2.1.1 Aluminum anodization structures . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6
2.1.2 AAO Electrochemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6
2.1.3 Kinetics of Anodic Aluminum Oxide Formation . . . . . . . . . . . . . . . . . . . .
12
2.1.4 Double-step Anodization of Aluminum . . . . . . . . . . . . . . . . . . . . . . . . .
17
2.1.5 Pre-textured Porous AAO vs Self-organized Porous AAO templates
. . . . . . . .
19
2.2 Filling of the AAO template - Electrodeposition . . . . . . . . . . . . . . . . . . . . . . . .
20
2.2.1 Nickel electrodeposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
21
2.3 High-k Dielectrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
22
3 Experimental Procedures: methods and materials
23
3.1 Electrodes I - Manufacture of the AAO templates by aluminum anodization . . . . . . . . .
24
3.1.1 Annealing Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
24
3.1.2 Degreasing Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
25
3.1.3 Electropolishing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
25
3.1.4 Anodization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
26
3.1.5 Pore-widening and barrier layer destruction . . . . . . . . . . . . . . . . . . . . . .
31
3.2 Electrodes I - Electrodeposition of nickel to fill the templates . . . . . . . . . . . . . . . . .
32
3.3 Electrodes I - Cleaning and assembling . . . . . . . . . . . . . . . . . . . . . . . . . . . .
33
3.3.1 Metallic contact and Polymeric layer . . . . . . . . . . . . . . . . . . . . . . . . . .
34
3.3.2 Cleaning of the remaining aluminum substrate and aluminum oxide
35
ix
. . . . . . . .
3.4 Electrodes I - Deposition of the dielectric (spin coating) . . . . . . . . . . . . . . . . . . . .
36
3.5 Atomic Force Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
38
3.6 Electrodes II - Evaporation of the aluminum contact . . . . . . . . . . . . . . . . . . . . . .
38
3.7 Electrodes II - Deposition of the PEDOT:PSS and Multiwall Carbon Nanotubes composite
film . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
39
3.8 Electrodes II - Deposition of the dielectric (spin coating) . . . . . . . . . . . . . . . . . . .
40
3.9 Conductance and Capacitance measurements . . . . . . . . . . . . . . . . . . . . . . . .
40
4 Results and Discussion
45
4.1 Anodization Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
46
4.1.1 Nanoporous structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
46
4.1.2 Pore Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
49
4.1.3 Mean pore diameter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
54
4.1.4 Surface Area and Pore Length . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
57
4.2 Electrodeposition Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
58
4.3 Conductance and Capacitance Results . . . . . . . . . . . . . . . . . . . . . . . . . . . .
58
4.3.1 Dielectric material influence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
58
4.3.2 Electrodes I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
61
4.3.3 Electrodes II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
64
5 Conclusions
67
5.1 Achievements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
67
5.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
69
Bibliography
74
x
List of Tables
3.1 Summary of the procedures involved in the manufacture of the AAO templates by aluminum anodization.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
24
4.1 Capacitance values for Electrodes E at 100 Hz. . . . . . . . . . . . . . . . . . . . . . . . .
59
4.2 Capacitance values for Electrodes E at 1 kHz. . . . . . . . . . . . . . . . . . . . . . . . . .
60
4.3 Capacitance values for Electrodes I at 100 Hz. . . . . . . . . . . . . . . . . . . . . . . . .
62
4.4 Capacitance values for Electrodes I at 1 kHz. . . . . . . . . . . . . . . . . . . . . . . . . .
62
4.5 Area increase (in %) with respect to the area of I.base flat electrode, for Electrodes I at
100 Hz. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
63
4.6 Area increase (in %) with respect to the area of I.base flat electrode, for Electrodes I at 1
kHz. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
64
4.7 Capacitance values for Electrodes II at 100 Hz. . . . . . . . . . . . . . . . . . . . . . . . .
64
4.8 Capacitance values for Electrodes II at 1 kHz. . . . . . . . . . . . . . . . . . . . . . . . . .
64
xi
xii
List of Figures
1.1 Proposed design for the electrode surface structure to be developed. . . . . . . . . . . . .
1
1.2 Surface area compararison. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2
2.1 Structural types of aluminum oxide. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6
2.2 Anodization setup. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7
2.3 Anodic reaction of aluminum at different current-voltage conditions. . . . . . . . . . . . . .
7
2.4 Aluminum pitting corrosion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8
2.5 Macroscopic and nanoscopic view of porous AAO. . . . . . . . . . . . . . . . . . . . . . .
8
2.6 Voltage effects on the AAO nanopores. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
10
2.7 Non-porous AAO layer formation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13
2.8 Hexagonal porous AAO template. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13
2.9 Porous oxide barrier layer region. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14
2.10 Porous AAO layer formation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
15
2.11 Top and bottom surface of AAO template formed by a one-step anodization. . . . . . . . .
18
2.12 Double anodization summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
18
2.13 Hexagonal honeycomb-like organized AAO pore array. . . . . . . . . . . . . . . . . . . . .
19
2.14 Indentations on aluminum susbtrate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
19
2.15 Pore-widening process. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
21
3.1 Substrate AlLowPur. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
25
3.2 Electropolishing solution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
26
3.3 Anodization scheme. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
27
3.4 Anodization setup. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
27
3.5 Anodization electrolyte. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
28
3.6 Anodization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
28
3.7 Cleanning solution 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
29
3.8 Anodized substrate AlSigma. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
29
3.9 Cleanning solution 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
30
3.10 Anodized (completely) substrate AlCookFoil. . . . . . . . . . . . . . . . . . . . . . . . . .
30
3.11 Anodized substrate AlCookFoil. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
31
3.12 Anodized substrate AlLowPur. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
32
xiii
3.13 Pore-widening solution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
32
3.14 Scheme of the electrodepostion setup. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
33
3.15 Electrodepostion nickel bath. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
34
3.16 Nickel electrodepostion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
34
3.17 Electrodeposited layer scheme. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
35
3.18 Electrodeposited substrate AlLowPur. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
35
3.19 Assembling of the electrode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
35
3.20 Electrodes I.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
36
3.21 PVA solution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
37
3.22 Titanium isopropoxide solution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
38
3.23 PEDOT:PSS + MWCN dispersion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
39
3.24 Electrodes II. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
40
3.25 Conductivity measurement scheme. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
41
3.26 Capacitance measurement scheme. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
41
3.27 Equivalent circuit. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
42
4.1 Porous AAO template, bottom surface, fabricated through a one-step anodization of substrate AlSigma at 70 mA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
46
4.2 3D AFM Image: Porous AAO template, bottom surface, obtained by a one-step anodization process of substrate AlSigma at 70 mA. . . . . . . . . . . . . . . . . . . . . . . . . . .
47
4.3 Porous AAO template, top surface, prepared via double-step anodization of substrate
AlCookFoil at 20 mA.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
48
4.4 Enhanced Honeycomb six dot nanodimples structure. . . . . . . . . . . . . . . . . . . . .
48
4.5 3D Image: Porous AAO template, top surface, via double-step anodization of Substrate
AlCookFoil at 20 mA.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
49
4.6 Nanodimples study. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
49
4.7 Three area’s scale - Porous AAO template, bottom surface, via one-step anodization of
Substrate AlSigma at 70 mA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
50
4.8 Pore density analysis: Porous AAO template, bottom surface, via one-step anodization of
Substrate AlSigma at 70 mA, larger area. . . . . . . . . . . . . . . . . . . . . . . . . . . .
51
4.9 Pore density analysis: Porous AAO template, bottom surface, via one-step anodization of
Substrate AlSigma at 70 mA, medium and smaller area. . . . . . . . . . . . . . . . . . . .
51
4.10 Three area scales - Porous AAO template, top surface, prepared via double-step anodization of Substrate AlCookFoil at 20 mA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
53
4.11 Pore density analysis: Porous AAO template, top surface, prepared via double-step anodization of Substrate AlCookFoil at 20 mA, larger area. . . . . . . . . . . . . . . . . . . .
53
4.12 Pore density analysis: Porous AAO template, top surface, prepared via double-step anodization of Substrate AlCookFoil at 20 mA, medium and smaller area.
. . . . . . . . . .
54
4.13 Nano-pore top structure scheme. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
55
xiv
4.14 Mean pore radius analysis: Porous AAO template, bottom surface, prepared via one-step
anodization of Substrate AlSigma at 70 mA. . . . . . . . . . . . . . . . . . . . . . . . . . .
55
4.15 Graphic for the nano-pore radius distribution: Porous AAO template, bottom surface, prepared via one-step anodization of Substrate AlSigma at 70 mA. . . . . . . . . . . . . . . .
56
4.16 Mean pore radius analysis: Porous AAO template, top surface, via double-step anodization of Substrate AlCookFoil at 20 mA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
56
4.17 Nano-pore hole radius distribution: Porous AAO template, top surface, prepared via doublestep anodization of Substrate AlCookFoil at 20 mA.
. . . . . . . . . . . . . . . . . . . . .
57
4.18 Nickel nanowires in commercial AAO template. . . . . . . . . . . . . . . . . . . . . . . . .
58
4.19 Nickel nanowires on the nickel smooth layer. . . . . . . . . . . . . . . . . . . . . . . . . . .
59
4.20 Electrodes I layered structure scheme. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
59
4.21 Extra electrodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
60
4.22 Capacitance values for Electrodes E, at different distances. . . . . . . . . . . . . . . . . .
61
4.23 Capacitance values for Electrodes I, at different distances. . . . . . . . . . . . . . . . . . .
63
xv
xvi
Chapter 1
Motivation
tab The aim of Biomedical Engineering is to improve the daily life with the development of new technologies. By applying engineering principles to medicine and biology its purpose is to provide better design
and problem solving skills to healthcare.
The objective of this work is to develop a manufacturing technique for a proposed design (Figure
1.1) to improve the electric capacitance of electrodes destined to measure electrocardiographic signals.
Biopotential measurements with capacitive electrodes do not need any direct contact between electrode
and skin, which saves the time required to expose and prepare the contact area when measuring with
conductive electrodes. Furthermore, with the improvement of the electric capacitance, one would be
able to measure signals at higher distances and this opens up innumerous possibilities. For instance,
in daily life, if one is required to drive for long periods of time, the implementation of this type of electrodes in the back of the sits could be helpful to monitor the driver’s fatigue for safety purposes, or if an
individual is required to handle heavy machinery, to avoid accidents related with sleep deprivation, the
implementation of such electrodes would also be beneficial for monitoring his state.
One of the factors to increase the electric capacitance is the increase of the surface area (the electric
capacitance is proportional to the surface area). Thereby, by taking inspiration from the human body,
where in the smaller intestine a large surface area is also required to better absorb the nutrients, the
proposed electrode design consists in villi-like structures, which maximize the surface area.
Figure 1.1: Proposed design for the electrode surface structure to be developed.
For the same flat surface, the narrower the villi are, the higher the yielded total surface will be (see
Figure 1.2). For that reason, the following proposed solution resorted to nanotechnological techniques
1
to achieve it.
Figure 1.2: Surface area compararison. In the images, ∆s is the flat surface area where the villi are.
Altough for both images, ∆s is the same, the effective surface area (in red) is higher for the first image,
where the vilosite structures are narrower, than for the second.
To develop new specialized miniaturized devices, nanostructures, such as nanoparticles, nanotubes,
nanowires and nanopores, are on demand. Nanoparticles and nanowires are especially important due
to their innumerous applications on electronic and optics components, biosensors, among others.
Nanomaterials are materials with basic structural units, grains, particles, tubes, spheres, fibers or
other constituent components, having dimensions on the nanometric range. They can be made from a
wide range of materials such as metals, ceramics, polymers and composites.
To obtain specific nanostructures several methods can be employed. These can be distinguished
between top-down and bottom-up approaches.
The top-down side makes use of conventional microfabrication methods, with specialized tools, to
shape such structures. Techniques such as electron beam lithography, ion beam lithography, x-ray
lithography, among others, allow us to engineer nanostructured materials, with few nanometers resolution. However, the high cost of the necessary lab equipment makes these techniques of limited use for
low-cost products [Picraux, 2013].
On the other hand, the bottom-up side makes use of the chemical and physical properties of molecules
and atoms to assemble and form specific systems and structures. These approaches allow us to achieve
large scale ordering of self-assembled nanostructures at a low cost, therefore being much more attractive [Picraux, 2013].
Among the bottom-up techniques, one easy and simple technique to obtain the desired structure
mentioned before is to use anodized aluminum oxide (AAO) porous membranes. The process of fabrication of the AAO templates constitues a bottom-up technique since it makes use of aluminum chemical
and physical characteristics when anodized at certain conditions, exposed in the next chapters.
Porous AAO has been used in the past in fields ranging from membrane reactors to bioactive surfaces
for tissue engineering ([Basile and Gallucci, 2010], [Parkinson et al., 2009] and [Fliniaux et al., 2005]).
The auto-assembling character of the nano-porous AAO templates is what makes it so simple and
attractive.
Once the desired templates or masks, depending on the technique, are obtained, they can be filled
with different materials. The choice of the material depends on the application of the device, ranging
2
from organic materials to metals. Various techniques such as electro-deposition, polymerization, sol-gel
and chemical vapor deposition (CVD) have been used to fill porous templates ([Fliniaux et al., 2005], [Ali
and Maqbool, 2013], [Kim, 2012] and [Kasi et al., 2013]).
The electric capacitance is proportional not only to the surface area of the capacitor electrodes, but
also to the dielectric constant of the medium between the plates. The capacitive electrodes will form a
capacitor system with the individual’s body, and will have as a dielectric medium everything that stands
in between (air, clothes). Therefore, to increase even further their capacitance, the dielectric constant
of the medium must be increased, and for that high dielectric constant materials (High-k dielectrics) can
be deposited on top of the electrodes.
In this thesis it is proposed the application of the self-assembling nano-porous AAO template technique, at room temperature, in association with a direct current nickel electrodeposition technique to fill
the template, in order to achieve the desired high surface area nanostructured design (Figure 1.1). As
mentioned above, to increase further the electric capacitance, two dielectric materials (polyvinyl alcohol
and titanium dioxide) with relatively high dielectric constant were deposited on the electrodes.
The chemical and electrochemical methods used are simple, with low equipment cost and the materials are not expensive.
3
4
Chapter 2
Theoretical Foundations and
Literature Review
tab Based on the proposed approach to manufacture the desired nanostructure, this chapter starts with
an overview of the physical principles, as well as of the process to obtain nanoporous AAO to serve as
a template.
Next, to obtain the desired nanostructure, the nanoporous AAO template needs to be filled and for
that a nickel electrodeposition process was used. There is a brief explanation of the process.
In third, a brief overview of dielectric materials is made.
2.1
Anodic Aluminum Oxide
tab For several years, in order to protect aluminum components against corrosion, several industries
have made use of the anodization process to create an anodic oxide layer on the metal.
Besides the application of the nonporous AAO layer, there have also been applications of the porous
AAO layer as an effective surface finish or as a pre-treatment for further surface processing, for instance,
for staining the aluminum with protective inks [Sato, 1991].
The anodization of aluminum, under certain conditions, allows the fabrication of porous AAO membranes, obtaining a highly ordered array of cylindrical shaped pores which can serve as templates in
nanofabrication.
In 1998, Masuda et al. [Masuda et al., 1998], were first able to produce a highly ordered hexagonal
honeycomb-like pore structure with a cell size of 500 nm using a two-step anodization process.
With further studies, by several researchers (described below), it was found that the pore diameters,
periodicity and density distribution can be controlled through specific macroscopic anodization conditions. Due to this highly controllable character, there has been a wide interest in using AAO layers in the
field of nanotechnology.
5
2.1.1
Aluminum anodization structures
tab Alumina is the most common form of aluminum oxide (Al2 O3 ). Since the aluminum molar mass is
26.98 g/mol and the oxygen molar mass is 16 g/mol, about 52.9% of alumina’s weight is aluminum (
%aluminum = [(2 × 26.98)/(2 × 26.98 + 3 × 16)] × 100% = 52.92%).
Aluminum naturally reacts with the oxygen present in the atmosphere creating a thin oxide layer
which serves as a protective barrier. In 1935, Verwey and others [Poinern et al., 2011], showed that the
anodic film reacts further with the environment, resulting in hydration of the outer surface of the barrier
film. Thus, the natural oxide structure consists of a hydrated porous layer growing on top of a thin smooth
inert dense layer, the barrier layer.
Considering this, depending on the anodization conditions on the lab, anodic aluminum oxide can be
produced in one of two very specific structures: it can either exist as a nonporous barrier layer or as a
porous oxide structure as represented in Figure 2.1.
Figure 2.1: Structural types of aluminum oxide. From left to right, the first image represents the nonporous barrier type oxide and the second image represents the porous oxide. In both images, A is the
aluminum substrate, B is the inner oxide layer and C is the outer oxide layer.
The porous nanostructure is electrically insulating, optically transparent or semi-transparent, chemically stable, bio-inert and a biocompatible material [Brüggemann, 2013].
As mentioned before, the highly ordered and defined nano-architecture constitutes an excellent template for nanostructures. The template fabrication is possible due to the extensive studies made on the
electro-chemical modulation of aluminum anodization in order to obtain specific surface characteristics.
This process is usually performed at few degrees Celsius, requiring a cooling mechanism in order to
lower the temperature ([Parkinson et al., 2009], [Mat, 2011] and [Na et al., 2010]), and in an acidic
media [Araoyinbo et al., 2012].
2.1.2
AAO Electrochemistry
tab The AAO layer parameters, such as barrier layer thickness, pore diameter and pore height are
directly dependent upon the electrochemical conditions of the anodization process.
The basic anodization setup is depicted in Figure 2.2.
In the anodization process, what occurs is an electrolytic passivation process in which the natural
surface oxide layer on a metal is thickened. The metal to be treated constitutes the anode electrode of the
setup (Figure 2.2). The anodization process forces the reaction between the anode’s metallic cations
6
Figure 2.2: Anodization setup. Very simple scheme of the anodization process; the working electrode is
the anode, and it is the material to be anodized and the counter electrode is the cathode. Depending on
the metal to be anodized, the voltage, electrolyte and temperature may vary.
(formed upon metal oxidation) with the oxygen anions present in the electrolyte, with a consequent
reduction at the cathode (it can be the reduction of other species, such as H+ ). The oxide barrier formed
at the anode is extremely stable to further corrosion.
In section 2.1.3 ahead, the aluminum anodization process, in which the aluminum functions as the
anode, is explained in detail.
In the anodization of aluminum, different sets of current-voltage conditions result in different AAO
layer structures, and, of course, this depends, also, on the electrolyte conditions. Figure 2.3 shows how
the obtained morphologies depend on both voltage and current.
Figure 2.3: Anodic reaction of aluminum at different current-voltage conditions. Figure based on image
from reference [Poinern et al., 2011].
At low voltages and high currents, pitting corrosion (Figure 2.4) occurs. With an increase of the
7
voltage with lower current, an electro-polishing effect takes place.
Figure 2.4: Aluminum pitting corrosion. Encircled in the image, there are some regions which suffered
pitting in the initial experiments of this thesis.
As the voltage is further increased, a porous layer will form, leading to a current decrease, and, at a
certain value, the target porous nanostructure is achieved (Figure 2.5 b).
(a) Macroscopic view of porous AAO.
(b) Nanoscopic view of porous AAO.
Figure 2.5: Macroscopic and nanoscopic view of porous AAO. Image a is of one of the test samples
produced in this thesis. The yellow grey circle is the anodized region, and was anodized at a constant
voltage of 60 V. Image b is the AFM (Atomic Force Microscopy) image of a sample anodized at constant
voltage of 40 V, from reference [Li et al., 1998].
In the extreme, with low currents and high voltages, a thick layer of smooth aluminum oxide layer is
formed.
Materials
tab The initial composition and mechanical characteristics of the aluminum substrate have been shown
to be significantly important on the resulting morphology of the nanostructure.
8
Any physical defect resulting from an electrochemical, chemical, mechanical or thermal process will
influence the obtained nanostructure due to the fact that pore nucleation is a combination of both random
nucleation and nucleation biased by the surface defects, those being preferential pore nucleation sites
([Mat, 2011] and [Vorozhtsova et al., 2011]).
Besides the physical/mechanical aspects of the substrate, the purity of the aluminum will also influence the resulting nanostructure because the presence of alloying elements in the substrate reduces
the rate of growth of the forming oxide layer and influences its structure during the anodization process
([Poinern et al., 2011] and [Domaska et al., 2013]).
In addition, it may exist already a natural oxide layer at the surface of the metal, which influences
also the anodization process.
For these reasons, all surface pre-anodization treatments have a significant impact on the selfordering of the pore structures.
The usual process to prepare the aluminum substrate for anodization starts with its degreasing in
acetone, or a similar solvent ([Parkinson et al., 2009], [Mat, 2011], [Sulka et al., 2002] and [Ho et al.,
2011]).
The following step is an electrochemical polishing of the aluminum surface. This helps to get rid of
the natural oxide layer, and also may also help to reduce the surface roughness. The electrochemical
polishing is usually done in a solution of perchloric acid and ethanol (1:4), for a few minutes, followed
by extensive rising with water ([Araoyinbo et al., 2012], [Vorozhtsova et al., 2011], [Ho et al., 2011] and
[Lee et al., 2007]).
Next, the aluminum substrates are subjected to an annealing process. This step helps the release
of mechanical stresses that the substrate may have. This annealing process is done at a temperature
close to 400◦ C, which is around two thirds of the aluminum melting point, for several hours ([Poinern
et al., 2011], [Vorozhtsova et al., 2011] and [Oliveira, 2009]).
Electrolytes
tab As will be explained further ahead, the manufacture of porous AAO templates is done in acidic
electrolytes. The three main electrolytes used are aqueous solutions of sulfuric acid ([Vorozhtsova et al.,
2011], [Sulka et al., 2002], [Lee et al., 2007], [Oliveira, 2009], [Kim et al., 2013] and [Jessensky et al.,
1998]), oxalic acid ([Mat, 2011], [Vorozhtsova et al., 2011], [Kim et al., 2013] and [Jessensky et al.,
1998]) or phosphoric acid ([Vorozhtsova et al., 2011] and [Kim et al., 2013]).
Depending on the wanted characteristics for the nanopore structure, the applied voltage, temperature, concentration of the electrolyte and time of anodization need to be carefully chosen for each
electrolyte. This dependency is mainly attributed to two factors: conductivity and pH of the electrolyte
[Samantilleke et al., 2012].
In the case of the electrolyte conductivity, for instance, if one is using an electrolyte based on a
solution of sulfuric acid, because of its high conductivity, the application of high potentials leads to the
breakdown of the oxide layer [Samantilleke et al., 2012].
9
Concerning the pH value, it helps to determine the size of the nanopores. With a lower pH, the
potential threshold for field-enhanced dissolution (explained ahead in section 2.1.3) at the pore tip will
be lower, leading to a smaller pore size [Samantilleke et al., 2012]. However, the pH is not the decisive
factor on pore size. This can be perceived in the work of several researchers, showing that, although the
pH is a relevant factor, it is not the main one. It is the voltage that is central in determining the pore size.
For instance, Wang et al. performed studies using oxalic acid (at voltages of 15 and 40 V) and sulfuric
acid (at voltages of 10 and 20 V) electrolytes [Na et al., 2010]. From the images presented (Figure 2.6),
it can be roughly observed that higher voltages lead to both larger pore diameter and pore interspacing:
with oxalic acid (0.3M), when the voltage was raised from 15 V to 40 V, the pore diameter apparently
increased from 20 nm to about 50 nm; with sulfuric acid (0.3 M), when the voltage was raised from 10 V
to 20 V, the pore diameter increased from 15 nm to roughly 25 nm.
(a)
(b)
(c)
(d)
Figure 2.6: Voltage effects on the AAO nanopores. FE-SEM (Field Emission - Scanning Electron Microscopy) images of the AAO templates fabricated in (a) 0.3M sulfuric acid at 20V, (b) in 0.3M oxalic
acid at 40V, (c) in 0.3M sulfuric acid at 10V and, finally, (d) in 0.3M oxalic acid at 15V. All images from
reference [Na et al., 2010].
Parkinson et al. managed to produce different AAO porous templates, for application in cell cultures,
using the three main acidic electrolytes [Parkinson et al., 2009]. All the anodizations were performed with
a standard double-step anodization procedure (explained ahead in section 2.1.4). For an electrolyte of
10
oxalic acid (0.3 M, aq.), at 30 V the pore size was of 60 nm (anodization times of 7+14 h), while at 60 V
the pore size was 125 nm (anodization times of 4+4 h). Once again, it was shown that higher voltages
result in larger pore sizes. For a sulfuric acid electrolyte (0.3 M, aq.), at 24 V the pore size was 41 nm
(anodization times of 5+5 h). They also performed two other anodizations using phosphoric acid, one at
2.5 M, at 60 V (anodization times of 5+7 h) yielding a pore size between 180 and 200 nm, and other at
0.3 M concentration and 130 V (anodization times of 5+7 h) yielding a pore size of, approximately, 500
nm. From these last two anodizations, it is clear that the voltage has a larger influence in pore size than
pH.
If the aim is to obtain larger pore sizes, even if the electrolyte is not the most significant factor
to be taken into account to achieve this goal, some researchers resort to combinations of the acidic
electrolyte with other substances in order to allow the anodization procedure at higher voltages. For
instance, a phosphoric acid/methanol/water mixture has been used as the electrolyte for an anodization
at a constant potential of 195 V at -4◦ C. For such high voltage, a low temperature was required to avoid
AAO layer breakdown. The addition of methanol to the electrolyte was intended to help dissipate the
heat from the reaction, without freezing the electrolyte [Samantilleke et al., 2012].
Besides the three main electrolytes, other researchers have used other acids. There are reports on
the use of citric, malic, chromic, glycolic ([Samantilleke et al., 2012] and [Sulka, 2008]), boric, tartaric
[Naieff and Rashid] and malonic [Lee et al., 2007] acids.
As mentioned before, the electrolyte pH, although not being the main factor that influences the size
of the pores, seems to be essential for the formation of pores, or, in other words, essential to start the
nucleation process (explained ahead in section 2.1.3). Studies made have shown that for neutral or
alkaline electrolytes the AAO layer formed is non porous, consisting instead of a smooth barrier layer
on the aluminum surface. A production of nanoporous AAO layer in alkaline electrolyte was reported by
Araoyinbo et al. However, in order to achieve this, a special circuit had to be made. In general, alkaline
electrolytes produce nonporous AAO layers [Araoyinbo et al., 2012].
The anodization process is usually performed at a few degrees Celsius. The electrolyte temperature
influences the thickness and mechanical properties of the oxide formed. For higher temperatures (above
50◦ C), a thin, soft and non-protective porous oxide layer is formed, while at low temperatures (between
0 ◦ C and 5◦ C), the resulting porous oxide layer is more compact and hard [Aerts et al., 2007].
Finally, besides the voltage, pH, conductivity and temperature, each electrolyte can specifically influence the obtained nanostructure. Some studies show that for constant voltage, temperature and
concentration, the porous AAO layers anodized in oxalic acid have more uniform nano-channels, fewer
embodied anions and superior hexagonal self-ordering than membranes anodized with sulfuric acid,
and these last ones were found to have lower flexibility, hardness and abrasion resistance [Belwalkar
et al., 2008].
11
Voltage Conditions
tab The voltage conditions, in an acidic medium, are related to the architecture of the pores formed.
In short, the higher the voltage the higher the electric field strength is and that makes the pores larger
and with a better arrangement and uniformity (explained ahead in section 2.1.3) [Domaska et al., 2013].
However, depending on the electrolyte concentration, composition and temperature, there is a maximum
voltage that can be reached before breakdown occurs [Poinern et al., 2011].
Because of the regularity of the pore structure, many researchers have developed models that can
roughly predict the pore diameter and the inter-pore distance under potentiostatic conditions. For instance, Xu et al. explained the migration velocity of the ions under the electric field during the growth
and dissolution process of the forming porous oxide layer [Zhou, 2011]. Also, Ono et al. established a
linear correlation between pore diameter and voltage for citric, malonic, oxalic, phosphoric, sulfuric and
tartaric acids [Sulka, 2008].
2.1.3
Kinetics of Anodic Aluminum Oxide Formation
tab With the progress of nanotechnology, a considerable effort was put in the understanding of the
mechanism underlying the formation of the nanoporous AAO membrane and its dependence on several
macroscopic parameters.
Non-Porous AAO Layer (barrier type layer) Formation
tab For the non-porous AAO type of layer, or barrier type layer, the electrolyte used is a neutral or alkaline
solution, with a pH greater than 7 [Poinern et al., 2011]. Electrolytes based on ammonium adipate [Oh
and Chi, 2000], sodium borate, sodium chromate and sodium hydrogen phosphate solutions, among
others, are usually used. The AAO layer formed under these conditions is generally thin and non-porous
[Poinern et al., 2011].
The non-porous AAO layer formation process starts with a stage where a constant current density is
maintained and, therefore, the electric field strength is also constant across the oxide barrier layer. As
the layer grows, the electrical resistance increases, and so, the voltage increases in order to maintain
the galvanostatic conditions, until the formation voltage is reached. While this is happening, the oxide
continues to grow by the migration of Al3+ cations into the electrolyte and the inward motion of O2− and
OH− anions, with the voltage remaining constant. The oxide formation occurs both at the metal/oxide
and oxide/electrolyte interfaces, but majorly at the metal/oxide interface [Poinern et al., 2011].
As time passes on, a stationary state is reached in which the layer growth has decreased to a point
where it is equal to the rate of dissolution. The efficiency of the current responsible for building the oxide
layer is high, close to 100% [Li et al., 1998].
The higher the formation voltage the thicker the AAO barrier layer will be, in other words, the thickness
of the AAO barrier layer is directly proportional to the applied voltage [Poinern et al., 2011].
12
Figure 2.7: Non-porous AAO layer formation. Layer A is the aluminum substrate, layer B is the
metal/oxide interface, layer C is the oxide and layer D is the oxide/electrolyte interface.
The result of this process is a smooth, non-porous AAO barrier layer, with the oxide filling any minor
surface imperfections, and being stable, hard, electrically insulating and resistant to any further reactions.
Nano-Porous AAO Layer Formation
tab The nano-porous AAO layer’s structure consists on an self-assembled hexagonal honeycomb-like
pore array, as it can be seen in Figure 2.8. This arrangement has been reported by several researchers
([Mat, 2011], [Lee et al., 2007], [Jessensky et al., 1998], [Yin et al., 2007] and [Zhang et al., 2010]).
Figure 2.8: Hexagonal porous AAO template. SEM (scanning electron microscope) image of the top
surface of the AAO pore array from reference [Yin et al., 2007].
Two distinct regions can be identified on the nano-porous AAO layer (see Figure 2.9).
One of the regions is enhanced in the inset on Figure 2.9: it is a thin non porous oxide layer, of
constant thickness, adjacent to the metal substrate on one side and, on the other side, it is connected
to the pore region. This region is called the barrier layer and basically is the closed bottoms of the oxide
13
Figure 2.9: Porous oxide barrier layer region. In the porous AAO can be distinguished the region with an
open pore, and a thin region (enhanced in the inset) of barrier layer oxide which constitutes the bottom
of the pores. Image from reference [Nielsen et al., 2010].
pores.
The other region is constituted by the rest of the porous structures themselves.
The non-porous AAO is less brittle, harder and more wear resistant than the nano-porous AAO.
Consequently, electrolytes composed of less concentrated acids, because their pH value is closer to the
pH value formation of the harder non-porous AAO, tend to form harder and more resistant nano-porous
AAO.
There are several models proposed for the pore formation. However, the exact mechanism of pore
nucleation and growth steps is not yet quite clear.
One accepted model explains pore nucleation through an electric field assisted local chemical dissolution at the oxide/electrolyte interface, with consequent oxide generation at the metal/oxide interface
[Poinern et al., 2011].
In contrast to the barrier-type AAO layer, the electrolytes used to fabricate the nano-porous AAO
needs to have a pH lower than 5, and some examples were mentioned above.
The theory is that electrolytes with a pH greater than 5 produce a barrier layer type AAO because the
Al3+ cations released from the metal substrate are retained in the oxide layer. However, for electrolytes
with a pH less than 5, the flow of Al3+ cations into the electrolyte is higher, and, in some regions, the
formation of new oxide at the oxide/electrolyte interface is unstable, leading to thinner oxide in random
spots. In these places, with thinner oxide, the electric field increases, with consequent increase of the
current. In regions of high current flow there is an increased dissolution rate, and it produces a large
numbers of hemispherical depressions which correspond to the pore nucleation sites. The formation of
these hemispherical shape depressions aggravates the dissolution process by concentrating the electrical field, increasing even more the dissolution rate at those points. This is the field-assisted dissolution
model, which is based on random variations of the electric field, and on the places where the electric
field increases, the current increases, and in regions of high current flow there is an increased dissolution
rate ([Poinern et al., 2011] and [O’sullivan and Wood, 1970]).
14
Figure 2.10: Porous AAO layer formation. In A there is a initial thin smooth oxide layer grows; B due
to pH lower than 5, the oxide is unstable, and, in some spots, dissolutio occurs, and the electric field
increases, with consequent increase of the dissolutian rate, and hemispherical nanodepressions start to
appear; C the nanodepressions have the efect of concentrating even more the electric field (red arrows),
increasing even more the dissolution rate; D through this positive feedback mechanism, the nanopores
start to grow.
Considering this, the key mechanism that affects the type of nanostructure obtained during anodization is the degree of uniformity of the electric field over the metal surface. When compared to the
non-porous AAO layer formation, where the electrolyte pH is greater than 5 and the electric field is fairly
constant over the entire surface, making a smooth planar oxide, for the porous AAO layer formation, it is
the regional variations of electric field due to the low pH that allow the formation of the nanopores.
As in the case of the barrier-type AAO layer, for the porous AAO layer formation, oxide production
also occurs at both the metal/oxide and oxide/electrolyte interfaces.
The process of AAO formation involves a series of oxi-reduction reactions [Poinern et al., 2011]. At
the metal/oxide interface, Al3+ cations migrate from the metal to form the oxide layer. The O2− anions,
at the oxide/electrolyte interface, travel into the oxide layer. Close to 70% of the Al3+ cations and the
O2− anions contribute to the formation of the barrier oxide layer.
2Al + 3O2− →
− Al2 O3 + 6e− .
(2.1)
At the oxide/electrolyte interface, Al3+ cations migrate and with O2− anions, form the oxide layer. The
formation rate at this interface is much smaller than that at the metal/oxide interface.
2Al3+ + 3H2 O →
− Al2 O3 + 6H +
(2.2)
The form of the oxygen-containing anions is uncertain. They could either be O2− or OH− . At the
oxide/electrolyte interface, part of the Al2 O3 is dissolved into the electrolyte. The oxide anions move
across the oxide layer, driven by electric field, contributing to the formation of Al2 O3 at the metal/oxidel
interface. However, these oxide anions are not enough for the continuously growing oxide layer. A
significant amount of anions must be supplied from dissociation of water at the oxide/electrolyte interface
acording to one of the two following equations [Su and Zhou].
15
H2 O →
− 2H + + O2−
(2.3)
H2 O →
− H + + OH −
(2.4)
In the later case, aluminium hydroxide forms as an intermediate phase and it will decompose very
quickly and no detectable hydroxide layer appears.
The rest of the Al3+ cations (30 %) are dissolved into the electrolyte and it is this dissolution that
allows the formation of the porous layer (dissolution of the anodic aluminum).
2Al →
− 2Al3+ + 6e−
(2.5)
Meanwhile, at the cathode side, hydrogen production occurs.
6H + + 6e− →
− 3H2
(2.6)
The overall aluminum anodization equation is
2Al + 3H2 O →
− Al2 O3 + 3H2
(2.7)
During the growth of the porous oxide, the thin barrier layer region of the nano-porous AAO layer is
constantly regenerating.
Since the dissolution is dependent on the electric field strength, the pore diameter it is also directly
dependent on the electric field produced by the anodizing voltage.
Aluminum oxidation occurs over the entire pore base resulting in oxide growth perpendicular to the
surface. Also, the neighboring pores help to prevent growth in any other direction.
Finally, dissolution of the oxide on the pore walls is, essentially, non-existent because the electric
field strength there is too small to make any significant contribution to the flow of ions.
During oxide formation, volume expansion occurs because alumina is less dense than metallic aluminum. The hemispherical pore base is also influenced by the forces resulting from this volume expansion. The mechanical stresses intervene in the self-ordering of the pores. At the beginning of the
anodization process, pores nucleate randomly all over the aluminum surface. As the time increases, the
pore patterns become more ordered, with some self-adjustment of the pore configurations taking place,
until a single, long-range ordered pore structure is achieved [Ono et al., 2005]. Eventually, a densely
packed hexagonal array of ordered pores is obtained.
When the final pore organization is achieved, the time factor will only affect the length of the pore
walls. The walls grow linearly with time, with the pore diameter and arrangement remaining the same
([Lima et al., 2009] and [Lee et al., 2006]).
16
The pore wall, besides amorphous alumina, may contain anions from the electrolyte, small quantities
of water and nano-crystallites. It can also have some voids due to possible oxygen evolution during
oxide formation ([Poinern et al., 2011] and [Brown et al., 2012]).
Overall, this pore nucleation model, based on the field-enhanced dissolution, also explains why the
pore nucleation sites are sensitive to the surface topography and imperfections such as impurities, pits,
scratches, grain boundaries, among others. All this imperfections induce electric field variations at their
locations causing biased pore nucleation.
Additionally, chemical composition also affects the nanostructure [Domaska et al., 2013]. For instance, Zaraska et al. showed that the presence of alloying elements in an aluminum alloy influences
AAO features such as porosity, barrier layer thickness, pore diameter and pore density of the oxide layer
[Poinern et al., 2011].
Besides this model, other models have been proposed to explain the pore nucleation mechanism.
One of the models, for instance, presented by Patermarakis et al. suggested that pore nucleation results
from spontaneous recrystallization of the unstable lattice of oxide at the metal/oxide interface to a more
stable denser nano-crystalline. The resulting recrystallization causes ruptures in the oxide surface and
produces regions of rarefied oxide between nanocrystallites, suggesting that this is the location where
pore nuclei will form [Patermarakis, 2009].
2.1.4
Double-step Anodization of Aluminum
tab The pore nucleation process is quite random. For certain nanotechnological applications, a certain
degree of precision and regularity is required.
As mentioned before, in 1998, Masuda et al. developed an anodization procedure able to achieve a
self-assembled highly ordered hexagonal AAO structure [Masuda et al., 1998].
The method studied by Matsuda involved a first long time anodization to form an highly ordered
pore array at the metal/oxide interface. With a first inspection, the organization of the pores at the
oxide/electrolyte interface reflects the initial random nucleation sites (see Figure 2.11 a). Only when the
oxide layer is separated from the aluminum substrate, it is possible to observe the resultant organized
nano-indentations on the metal substrate, or, on the other hand, to see the organized pore bottoms (see
Figure 2.11 b) which contrast with the initial organization of the pores at the top of the layer.
The nano indentations left in the metal, after removing the first oxide layer, will serve as preferential
sites for pore nucleation for a subsequent anodization. This double-step anodization procedure is what
Masuda and Satoh developed and studied. With the removal of the first oxide layer, it is possible to
achieve an extremely organized pore array with packed hexagonal configuration of straight and parallel
pore channels [Poinern et al., 2011].
One aspect that corroborates the theory that the nano indentation serve as pore nucleation sites
for the second anodization is the fact that, through current density profiles analysis, it can be observed
17
Figure 2.11: Top and bottom surface of AAO template formed by a one-step anodization. In image a,
SEM image of the top surface of aluminum anodized at 195 V, on a one-step anodization; in image b,
SEM image of the bottom surface of the same aluminum substrate. Images from reference [International,
2013].
Figure 2.12: Double anodization summary. In the first anodization, the pores nucleate quite randomly,
taking a long time to self-organize; at the end of the first anodization, the oxide is removed, leaving
an organized nanoindentation on the metal substrate; in the second anodization, pores nucleate in the
previous indentations. Since in the second anodization the pores are organized from the begining, it
does not take as much time as the first one.
that the first step takes much more time to achieve a steady state oxide growth rate than the second
step. In other words, the nano indentation sites provide a fast ordered initiation of pore nucleation in the
subsequent anodization step and, therefore, the steady state oxide growth is achieved in a much shorter
time [Poinern et al., 2011].
Research into a three-step anodization process has been done and was concluded that the process
did not show any significant improvement in terms of the ordering of the pores over a double-step
anodization process [Sulka et al., 2002].
With this double-step anodization, an extremely organized hexagonal honeycomb-like pore array can
be obtained.
One of the key identifying features of the organization of the pore array is that, at the pore surface,
six dot nanodimples that accentuate its hexagonal honeycomb-like form can be observed ([Zhang et al.,
2010] and [Nishinaga et al., 2013]).
These nanodimples are the result of the oxidation of the nanoindentations left on the aluminum
substrate from the first anodization.
At the end of the first anodization, when the pore organization is achieved, the hemispherical bottom
pores are all next to each other (Figure 2.14 a), and, after removal of this first oxide layer, the aluminum
18
Figure 2.13: Hexagonal honeycomb-like organized AAO pore array. Image of the pore array resulting
from an aluminum substrate being anodized at 40 V in 0.3 M oxalic acid (from reference [Zhang et al.,
2010]). The hexagonal organization of the pores is quite clear, as well as the six dot nanodimples
present at each corner of the hexagonal form (group of the six dot nanodimples enhanced with white
circles in the image).
(a) Bottom of the oxide layer.
(b) Aluminum substrate
Figure 2.14: Indentations on aluminum susbtrate. Image a is the bottom of the oxide layer; image b is
the negative indentation left from the bottom of the oxide pores from image a on the aluminum substrate.
In image a, some points where the nanopores did not completely arrange themselves (enhanced with
the white circles) are shown, and that is reflected on image b with spikes of the aluminum subtrate. Both
images are from reference [Lee et al., 2007].
substrate has those nanoindentations which are the negative of the removed oxide pores bottom surface (Figure 2.14 b). When anodized a second time, the pattern of these nanoindentations remains,
constituting the top surface of the second oxide layer.
2.1.5
Pre-textured Porous AAO vs Self-organized Porous AAO templates
tab The key requirement to obtain an organized packed pore array is the existence of pre-organized
sites where pore nucleation can occur. These sites can be either self-achieved or created through pretexturing.
19
Pre-textured Porous AAO templates
Several techniques have been developed in order to control these initial pore nucleation sites. For instance, the aluminum substrate surface can be pre-textured or nano-imprinted using a mask or template,
using methods like the pre-texturization of the substrate with the tip from a scanning probe microscope,
with electron beam lithography, interference lithography and with ion beam lithography. It can also be
pre-textured with a procedure which involves nanosphere lithography using a master print template with
planar ordered arrays of nanometer-sized spheres as lithography masks. The master print template can
be made of silicon carbide, silicon nitride, nickel or polydimethylsiloxane (PDMS) ([Brüggemann, 2013]
and [Huamanrayme Bustamante, 2012]).
However, these techniques require the use of sophisticated and expensive equipment and they are
limited to small master print template areas. This makes the working area usually smaller than a few
square millimeters.
Self-organized Porous AAO templates
These templates are fabricated using the two step anodization process described earlier. It is the
main process used by the scientific community. The periodic nanoindentations formed during the first
anodization serve as pore nucleation sites for the second anodization. This self-arrangement characteristic allow to use this technique through several square centimeters and this, associated with the
simplicity of the method, is what makes it so much more interesting from an industrial point of view.
2.2
Filling of the AAO template - Electrodeposition
tab Among the several techniques that can be used to fill nano-porous AAO templates, electrodepostion
represents a simple, low-cost and high throughput technique to create the desired nanostructures [Jeong
et al., 2007].
Electrodeposition, or electroplating, is a process in which occurs the reduction of metal cations dissolved in an electrolyte, by the application of an electrical current, in order to obtain a coherent metal
coating on an electrode, the cathode.
The setup for the electroplating is basically the same as the setup for the anodization process (Figure 2.2). However, whereas initially the aluminum substrate is the anode and the aluminum ions are
dissolved into the electrolyte, or react with the oxygen ions from the electrolyte, now, the aluminum substrate (with the oxide template attached) will serve as the cathode, since the objective is to fill the oxide
template attached to it.
Since the oxide template is electrically isulating, before electrofilling the template, some treatments
have to be done to destroy the oxide at the bottom of the nanopores (barrier layer of the porous AAO
layer), exposing a clear aluminum substrate at the bottom of the pores, that serve as the cathode, so
that the metal cations dissolved in the electrolyte can be deposited in the pores.
20
The treatment that is usually carried out consists on the immersion of the AAO template, attached to
the remaining aluminum substrate, in a aqueous solution of 5% phosphoric acid for a few minutes, as
many researchers already did ([Fliniaux et al., 2005], [Mat, 2011] and [Zhang et al., 2010]). Since the
pore bottoms barrier layer is much thinner that the rest of the porous region of the AAO layer, they will
dissolve, and the only modifications to the rest of the oxide layer is the widening of the pores.
Figure 2.15: Pore-widening process. Schematics of the structure before and after the pore-widening
process. In the After Pore-widening image the thin barrier layer in the bottom of the pores is dissolved,
exposing the aluminum substrate underneath.
The oxide-free aluminum substrate at the bottom of each pore, after the barrier layer dissolution and
pore-widening, acts as a cathode, allowing the growth of nanowires (shape of the pore) made from the
substance present in the electrolyte bath, that will be electrodeposited.
2.2.1
Nickel electrodeposition
tab Among the range of materials that can be electrodeposited, from metals to bio-molecular materials
[Samantilleke et al., 2012], nickel is the metal whose electrodeposition is one of the easiest processes.
Nickel electroplating (or electrodeposition) is a simple process, in which a direct current is made to
flow between two electrodes immersed in a conductive, aqueous solution of nickel salts. The direct
current causes an oxidation at the anode, such as the oxidation of a metal with dissolution of ions or the
oxidation of water to produce oxygen and, in parallel, the deposition of nickel present in the bath at the
cathode. Specifically, the nickel in the bath solution consists of divalent, positively charged ions (Ni2+ ),
which, with the flow of the current, react with two electrons, each being converted to atomic nickel that
is deposited at the cathode surface [Di Bari, 2000].
Several researchers already applied the nickel electrodeposition process to fill AAO templates ([Ali
and Maqbool, 2013], [Ho et al., 2011], [Jeong et al., 2007], [Lee, 2012] and [Nielsch et al., 2000]). For instance, Cortes et al. [Cortés et al., 2009] studied the influence of electrodeposition potential, pore size,
pH, composition, and temperature of the electrolytic bath on the structure of nickel nanowires arrays
electrodeposited into anodic alumina oxide porous membranes. Results showed that the electrodeposition potential controls the growth of nickel nanowires along some preferential crystallographic planes.
For instance, at -0.90 V (vs. Ag/AgCl), single crystalline nanowires with a strong orientation along the
(111) plane direction were obtained. High temperatures and a moderately acid pH solution contributed
to improve the single crystalline character of nanowires.
Normally, a Watts plating solution (described further ahead, in section 3) is used as the plating bath
([Jeong et al., 2007], [Nielsch et al., 2000] and [Cortés et al., 2009]).
21
2.3
High-k Dielectrics
tab Dielectrics with a dielectric constant higher than that of silicon nitride (dielectric constant, k, around
7) are classified as high dielectric constant materials. High-k dielectric materials have enormous potential for use in electronic devices. Silicon-based dielectrics (SiOx Ny ) have been widely used in the
manufacturing of silicon based semiconductor devices ([Singh and Ulrich, 1999] and [Brezeanu et al.]).
There are several materials with high dielectric constant, which can be divided into inorganic and
organic dielectric materials. Among the organic materials, polyvinyl alcohol (PVA) is one of the most
used because of its flexibility and, also, because of the value of its dielectric constant (k=10), which
is higher than that of most organic materials. Besides, although there are a few polymers that have
a higher dielectric constant, like CYPEL (cyanopulluane, k=18.5), they usually have a more complex
fabrication method, translating in higher costs, when compared with PVA. Other organic dielectric materials are, for instance, polyester (k=3), polyvinylpyrrolidone (PVP, k close to 4), polymethyl methacrylate
(PMMA, k approximately 3.5) and other polymers whose dielectric constant is below 3. The deposition
method for these materials is usually spin-coating. On the other hand, inorganic dielectrics usually have
higher dielectric constant than organic dielectrics and are mostly oxides. Some of the inorganic dielectric materials are, for instance, hafnium oxide (HfO2 , dielectric constant usually around 11, but higher
values around 20 have been reported), titanium oxide (TiO2 , dielectric constant usually around 40, but
depending on the treatments, it can reach 100), aluminum oxide (Al2 O3 , k approximately 9), zirconium
oxide (ZrO2 , k close to 25), among others. The main deposition methods are sol-gel based methods
from solutions and sputtering [Ortiz et al., 2009].
22
Chapter 3
Experimental Procedures: methods
and materials
tab The work of this thesis consists on the development of large surface area electrodes with subsequent
deposition of a high constant dielectric material, in order to increase electrodes capacitance.
Two types of electrodes were developed:
Electrodes I - main type of electrode developed. It was manufactured based on nano-porous AAO
templates;
Electrodes II - secondary type of electrodes developed by a simpler process, which involved composite films of PEDOT:PSS (polymer described on section 3.7) with Multi-Wall Carbon Nanotubes.
The fabrication of Electrodes I involved four steps:
Electrodes I - Manufacture of the AAO templates by aluminum anodization;
Electrodes I - Electrodeposition of nickel to fill the templates;
Electrodes I - Cleaning and assembling;
Electrodes I - Deposition of the dielectric (spin coating).
In the stage Electrodes I - Manufacture of the AAO templates by aluminum anodization, surface
topography images were taken using an Atomic Force Microscope (AFM).
The second type of electrodes fabricated (Electrodes II), to be used as comparison with the main
AAO based electrodes, consisted on a PEDOT:PSS and Multi-Wall Carbon Nanotubes composite film
deposited on an aluminum metallic contact, with subsequent deposition of a dielectric.
The development of these electrodes consisted on three parts:
Electrodes II - Evaporation of the aluminum contact;
Electrodes II - Deposition of the PEDOT:PSS and Multiwall Carbon Nanotubes composite film (drop
cast);
Electrodes II - Deposition of the dielectric (spin coating).
23
At the end, the conductance of Electrodes I was measured. The capacitance of both Electrodes I
and II was measured in a capacitor-like setup which consisted on an irradiator plate and the respective
electrode.
3.1
Electrodes I - Manufacture of the AAO templates by aluminum
anodization
tab Initial experiments were done on a high purity aluminum substrate (thickness 0.5 mm, 99.999%
purity, Sigma-Aldrich) to test the process, following the choice of innumerous researchers whose procedure always involved the use of high purity aluminum substrates ([Fliniaux et al., 2005], [Lee et al.,
2007], [Kim et al., 2013], [Yin et al., 2007] and [Lee et al., 2006]).
However, as the goal was to develop a low cost method to manufacture the electrodes, further works
were carried out with commercial aluminum cooking foil (Freshmate Aluminum Foil, thickness 20 micrometers, 99.95% purity, of the brand Vileda Professional) and of aluminum plates (thickness 0.5 mm,
99.95% purity, from Francisco Soares, Lda) with lower degree of purity but purchased at a much lower
price.
Therefore, for the anodization process, these three types of substrates (AlSigma - high purity aluminum substrate, from Sigma Aldrich; AlCookFoil - commercial aluminum cooking foil of the brand Vileda
Professional; AlLowPur - lower purity aluminum plates from Francisco Soares, Lda) went through preand pos-anodization treatments. The following table summarizes these treatments.
Substrates
Procedures
Annealing
Degreasing
Electropolishing
Anodization
Pore-widening and barrier layer destruction
AlSigma
X
X
X
X
—
AlCookFoil
X
X
—
X
X
AlLowPur
X
X
X
X
X
Table 3.1: Summary of the procedures involved in the manufacture of the AAO templates by aluminum
anodization.
3.1.1
Annealing Process
tab The substrates AlSigma and AlLowPur were cut into 2 × 3cm2 pieces (Figure 3.1). Substrate
AlCookFoil was cut into 9×4cm2 strips, which were, after the annealing process, attached to microscope
slides for handling.
The annealing process was performed in a standard laboratory hot plate, at its maximum of 400 ◦ C
for 7h, in air.
24
There were no macroscopically visible changes in substrates AlCookFoil after such annealing process, but substrates AlSigma and AlLowPur appeared to be slightly shinier.
Figure 3.1: Substrate AlLowPur: after the annealing process.
3.1.2
Degreasing Process
tab All substrates were sonicated, on a standard ultrasound machine, in isopropanol (gas chromatography 99.5%, Scharlau Chemie S.A.) for 20 min to clean the substrates surface. Afterwards, they were
carefully dried under nitrogen gas from a pressure gun. There were no visible changes on the substrates.
3.1.3
Electropolishing
tab This step was only performed on substrates AlSigma and AlLowPur because, due to its small thickness, substrate AlCookFoil was always easily ruptured.
Electropolishing is an electrochemical polishing to remove material from a metallic surface. The
setup is the same as that used in the anodization process (see Figure 3.3).
To achieve a successful electropolishing of a rough surface, the surface prominences must dissolve
faster than the recesses. Usually, electrolytes made of concentrated acid solutions, with high viscosity or
with mixtures of perchlorates are used. As mentioned before, many authors used a solution of perchloric
acid and ethanol (1:4 v/v) ([Araoyinbo et al., 2012], [Vorozhtsova et al., 2011], [Ho et al., 2011] and [Lee
et al., 2007]) and so this was the choice made in this work.
To prepare the electrolyte (Figure 3.2), 20 ml of perchloric acid (perchloric acid 70%, AnalR, sp.gr1.70,
BDH Chemicals Ltd) were mixed with 80 ml of ethanol (Ethanol absolute for UV, IR, HPLC 99.9 %, PanReac AppliChem).
The electropolishing process was performed for 8 min, at 5 V at room temperature (23 ◦ C), followed
by extensive rising with distilled water. These conditions were enough to polish the small area of the
exposed substrate before pitting was observed [Zhou et al., 2011].
After anodization, the substrates appear to be much smoother and shinier.
25
Figure 3.2: Electropolishing solution. The solution was prepared by mixing 20 ml of perchloric acid
(perchloric acid 70%, AnalR, sp.gr1.70, BDH Chemicals Ltd) with 80 ml of ethanol (Ethanol absolute for
UV, IR, HPLC 99.9 %, PanReac AppliChem).
3.1.4
Anodization
tab The usual setup for anodizing aluminum substrates is shown in Figure 2.2. However, throughout the
initials trials on the process, it was found that the use of this setup resulted in a non-uniform anodization
of the aluminum substrate, with consequent rupture of the topmost region of the substrate before the
remaining part could be anodized. For this reason, it was decided to anodize the aluminum using the
setup shown in Figure 3.3, which provides a good support to the growing oxide template.
With this setup, only a circular region of the aluminum surface is exposed to the electrolyte. The
aluminum that isn’t exposed to the electrolyte is not anodized and, so, it provides a rigid frame for the
handling of the porous AAO template. The acrylic plates which hold the aluminum substrate in place
also provide a rigid support.
Between the tube and the aluminum substrate, an O-ring prevents electrolyte leakages and also
defines the shape of the anodized surface (Figure 3.4).
The substrate surface to be anodized is a circle with a diameter of 1.6 cm (inner-diameter of the
O-ring), and the distance between the substrate and the counter electrode (cathode C, see Figure 3.3),
which was a high purity platinum disc, was 3.5 cm.
In order to do the nickel electrodeposition on the template, it is required to have a counter electrode
in the back of the template. By controlling the anodization time, it was possible to obtain a porous AAO
template supported on the remainder aluminum substrate which, with this setup, was not exposed to the
electrolyte and can act as the counter electrode in the electrodeposition.
26
Figure 3.3: Anodization scheme. Setup chosen to perform the anodization, so that the growing oxide
had a good support.
(a) O-ring.
(b) Anodization setup.
Figure 3.4: Anodization setup. Anodization was performed acording to the scheme shown in Figure 3.3,
but an aluminum plate was added in between the substrate bottom surface and the acrylic bottom plate
to provide an easy electrical contact. As we can see from the image, the area exposed to the electrolyte
is the area that the 1.6 cm inner-diameter O-ring allows. The distance d (see Figure 3.3) is 3.5 cm, and
the counter electrode (cathode C, see Figure 3.3) is a high purity platinum disc.
The selected electrolyte for the anodization of all the substrates was an aqueous solution of oxalic
acid 0.3 M. The solution (Figure 3.5) was prepared by dissolving 37.821 g of dihydrated oxalic acid
(H2 C2 O4 .2H2 O, from MERCK) in distilled water up to a final volume of 1 L.
The anodization was performed at room temperature, and, in order to prevent the increase of temperature, due to the fact that throughout the anodization process heat is produced, the setup was placed
in ice, as it can be seen in Figure 3.6.
Substrates AlSigma
This type of substrate was used in the initial tests where only a one-step anodization was performed.
27
Figure 3.5: Anodization electrolyte. Aqueous solution of oxalic acid 0.3 M prepared with 37.821 g of
dihydrated oxalic acid (H2 C2 O4 .2H2 O, from MERCK) in 1 L of distilled water.
Figure 3.6: Anodization. The anodization setup (from Figure 3.4) was placed inside a blue rubber
container, on an ice bath to keep it at low temperature.
Initially, anodization with a constant voltage of 50 V was performed and it was observed that large
variations of the current took place throughout the process. For this reason, it was experimented to
anodize the substrate at a constant current of 70 mA (the value was chosen because, despite the large
variations, the current in the previous experiment was around that value). For this current, the voltage
was maintained fairly constant throughout the process (with small variations of 1 or 2 V). The anodization
process was performed during 1 h.
Afterwards, a solution of 19 wt% of chloridric acid (HCl) with 0.2 M of copper chloride (CuCl2 ) was
prepared by diluting 58 ml of HCl (33 %) into 50 ml of distilled water. Then, 2.689 g of CuCl2 were added
to the previous solution, under constant stirring (Figure 3.7).
With this solution, the remaining aluminum substrate behind the anodized area was removed revealing the see-through aluminum template made from substrate AlSigma (Figure 3.8).
28
Figure 3.7: Cleanning solution 1. The solution consisted in 19 wt% of chloridric acid (HCl) with 0.2 M of
copper chloride (CuCl2 ). This solution dissolves metallic aluminum.
(a) Top side.
(b) Bottom side.
Figure 3.8: Anodized substrate AlSigma. Image a is the top side of the anodized substrate AlSigma;
image b is the bottom side of the same substrate. In some places there is still a small amount of the
aluminum substrate behind the anodized section, but we can clearly see the anodized template.
Substrate AlCookFoil
Substrate AlCookFoil, it was submited to a double-step anodization procedure. Initially, the anodizations were performed at 50 V, and the current varied in time as observed for the substrate AlSigma.
Therefore, the subsequent anodizations were carried out under constant current, with the voltage remaining more or less constant throughout the process.
With substrates AlCookFoil, several double-step anodizations were performed at different current
29
values, with a minimum of 20 mA and a maximum of 70 mA. The same current value was used in the
two anodization steps of each double-step anodization process.
Since these substrates were prepared using a double-step anodization, an aqueous solution of
NaOH 0.3 M was prepared in order to remove the oxide layer formed during the first anodization, before
performing the second anodization. To prepare the solution, 1.2 g of NaOH (97%, Sigma-Aldrich) were
dissolved in 100 ml of distilled water (Figure 3.9).
Figure 3.9: Cleanning solution 2. Aqueous solution of NaOH 0.3 M prepared with 1.2 g of NaOH (97%,
Sigma-Aldrich) dissolved in 100 ml of distilled water.
Initially, one-step anodizations, one at 20 mA and other at 70 mA, were performed until the substrates
were completely anodized (current became zero). The oxide layers obtained were see-through indicating
the formation of the oxide across the complete thickness of aluminum in the exposed area of substrates
AlCookFoil (Figure 3.10).
Figure 3.10: Anodized (completely) substrate AlCookFoil. Substrate AlCookFoil is see-through since it
was completely anodized at a constant current of 20 mA.
30
With this, it was possible to approximately established, by error and trial, two anodization times to
perform the double-step anodizations. It was decided that the double-step anodizations times would be
5 min. (first anodization step), followed by oxide removal in the 0.3 M NaOH solution during 3 min., and
finally 8 min. for the second anodization step.
These anodization times allow the formation of a porous AAO template with a remaining aluminum
substrate (Figure 3.11), under the oxide area, that will act as a counter electrode for the subsequent
nickel electrodeposition.
Figure 3.11: Anodized substrate AlCookFoil. Substrate AlCookFoil after double-step anodization using
5+8 min. The substrate is not see-through because it still has an aluminum layer under the anodized
area, which will act as counter electrode for subsequent electrodepostion.
Substrate AlLowPur
After analyzing the topography resulting from the anodizations with substrates AlSigma and AlCookFoil (further ahead in section 4), it was decided to perform double-step anodizations at constant current
of 20 mA and 70 mA on substrate AlLowPur. A relatively constant voltage was observed during the
process.
For these substrates, it was decided to use 1.5 h for the first anodization process and 1 h for the
second, since the substrate is fairly thick. In between, the oxide was removed by immersing the substrate
in the 0.3 M NaOH solution for 10 min.
After the anodizations, all substrates were rinsed thoroughly with distilled water.
3.1.5
Pore-widening and barrier layer destruction
tab Before the nickel electrodeposition into the porous AAO template, the barrier layer region in the
porous AAO oxide layer next to the metal needs to be removed in order to do the electrodeposition,
since this barrier is an insulator. Furthermore, a certain degree of pore widening is desired to facilitate
the electrodeposition bath to penetrate into the pores. To carry out this process, the substrates were
immersed in a aqueous solution of 5 wt% phosphoric acid (H3 PO4 ).
31
Figure 3.12: Anodized substrate AlLowPur after a double-step anodization using 1.5 h+1 h anodization
times. The substrate is not see-through because it still retains an aluminum layer under the anodized
area, which will act as counter electrode for subsequent electrodeposition.
The aqueous solution of 5 wt% of H3 PO4 (Sigma-Aldrich) was prepared by dissolving 2.5 g of H3 PO4
into 47 ml of distilled water (Figure 3.13).
Figure 3.13: Pore-widening solution. Aqueous solution of 5 wt% of H3 PO4 (Sigma-Aldrich) was prepared
by dissolving 2.5 g of H3 PO4 into 47.5 ml of distilled water.
This process was done at 40 ◦ C for 3 min, for substrates AlCookFoil, and for 8 min for substrates
AlLowPur. At the end, the substrates were extensively rinsed with distilled water.
3.2
Electrodes I - Electrodeposition of nickel to fill the templates
tab Electrodeposition was only performed on susbtrates AlCookFoil and AlLowPur since substrate AlSigma is too expensive in view of the applications of these electrodes.
32
To perform electrodeposition, the setup is very similar to the anodization setup however the polarity
of the voltage is reversed, as it can be seen in Figure 3.14.
Figure 3.14: Scheme of the electrodepostion setup.
In the electrodeposition, it is the aluminum substrate (with the open porous AAO template on top)
that serves as the cathode and the platinum disk serves as anode, allowing the deposition of nickel from
the electrolyte bath.
The nickel bath (known as Watts bath) chosen to perform the nickel electrodeposition was prepared
with the following concentrations: nickel sulfate hexahydrate (NiSO4 .6H2 O) at 300 g/l, nickel chloride
hexahydrate (NiCl2 .6H2 O) at 45 g/l, boric acid (H3 BO3 ) at 45 g/l and, finally, NaOH was added to adjust
the pH to 4 (Figure 3.15).
The electrodeposition (Figure 3.16) was carried out at 65 ◦ C with a constant current of 10 mA (the
voltage was around 2 V) until a smooth nickel layer was observed at the surface (see scheme in Figure
3.17 and a picture of the obtained substrate in Figure 3.18), at which point, the voltage decreased rapidly
from the 2 V value. The process took approximately 1 h for templates made from substrate AlCookFoil
and 2 h for templates made from substrate AlLowPur.
During the electrodepostion the bath was periodically renewed, without stopping the process.
3.3
Electrodes I - Cleaning and assembling
tab The large surface area electrode will consist on deposited metallic nickel composed by the smooth
layer supporting the array of nickel nanowires. This has to be separated from the porous AAO template
and remaining aluminum substrate, which served as counter electrode. To make it easier to handle,
before removing the aluminum substrate and the oxide, a metallic contact button was attached and a
supportive polymeric layer was deposited.
33
Figure 3.15: Electrodeposition nickel bath. Watts bath prepared with the following concentrations: nickel
sulfate hexahydrate (NiSO4 .6H2 O) at 300 g/l, nickel chloride hexahydrate (NiCl2 .6H2 O) at 45 g/l, boric
acid (H3 BO3 ) at 45 g/l. NaOH was added to adjust the pH to 4.
Figure 3.16: Nickel electrodeposition. Electrodeposition of nickel onto the open porous AAO template,
at 65 ◦ C at a constant current of 10 mA (voltage was around 2 V).
3.3.1
Metallic contact and Polymeric layer
tab First, a metallic contact button was glued to the smooth nickel surface with silver conductive epoxy
adhesive paste. Next, to provide extra support to the nickel layer once it is freed, a polydimethylsiloxane
(PDMS) layer was deposited all over the smooth nickel and the metallic contact button surfaces, leaving
just the button tip to provide the electric contact (Figure 3.19).
The PDMS (Sylgard 184 silicon elastomer kit from Dow Corning) solution was prepared by mixing a
34
Figure 3.17: Electrodeposited layer scheme. The deposited nickel layer has two distinct regions, one
is a smooth layer and the other a nanowired layer (which fills the AAO pores). Electrodeposition was
stopped when the smooth layer formed (see Figure 3.18).
Figure 3.18: Electrodeposited substrate AlLowPur. Open porous AAO template, made from substrate
AlLowPur, filled with nickel after electrodeposition. The circular deposited nickel layer has a more yellow
tone than the aluminum substrate. In the aluminum substrate it is possible to see the trace of the O-ring
around the nickel layer.
silicon elastomer and a curing agent in a 10:1 w/w proportion. Before deposition, the solution was placed
under vacuum for 30 min to release any air bubbles it might had. After deposition on the electrode, it
was cured at 65 ◦ C for 3 h in a standard laboratory oven.
Figure 3.19: Assembling of the electrode. On the left it is the aluminum substrate with nickel electrodeposited template where the metallic contact is glued on, and a PDMS layer is deposited on top for extra
support; on top-right it is shown the metallic buton contact and on down-right a scheme of the transversal
cut of the layers.
3.3.2
Cleaning of the remaining aluminum substrate and aluminum oxide
tab Once the support for the nickel electrode was made, the cleaning of the remaining aluminum substrate and oxide template proceeded in order to expose the nickel nanowires supported by the nickel
smooth layer.
35
To dissolve both the metallic aluminum and the aluminum oxide a more concentrated NaOH (3 M)
solution was prepared by dissolving approximately 120 g of NaOH into 100 ml of distilled water.
The exposed nanowired nickel surface appeared black due to the nickel nanowires, contrarily to the
light yellowish grey smooth nickel counter surface of the electrode.
(a) Electrodes I - front side.
(b) Electrodes I - back side.
(c) Electrodes I - transversal cut.
Figure 3.20: Electrodes I. Image a shows the black nickel nanowired front surface of the electrode;
image b shows the smooth nickel yellowish grey back surface of the electrode; image c is a scheme of
Electrodes I cross-section.
In the cleaning process with NaOH, the electrodes made with substrates AlCookFoil (commercial
aluminum cooking foil) curled, even with the PDMS layer supporting it, making it impossible for subsequent dielectric deposition and further use.
3.4
Electrodes I - Deposition of the dielectric (spin coating)
tab The following step was the deposition of the dielectric material on the nickel nanowired surface of the
electrode. Two dielectric materials were chosen: polyvinyl alcohol (PVA) and titanium dioxide (TiO2 ). In
36
order to allow the deposition by spin coating, an aqueous solution of PVA and a sol-gel solution of TiO2
precursor were used.
The first dielectric (Figure 3.21) prepared was a 6 wt% aqueous PVA solution in which 300 mg
of PVA (average Mw 124,000 - 186,000, 98-99% hydrolyzed, Sigma-Aldrich) were dissolved, under
constant magnetic stirring, at 80 ◦ C, into 5 ml of distilled water.
Figure 3.21: PVA solution. The first dielectric. It consisted on a 6 wt% aqueous PVA solution in which
300 mg of PVA (average Mw 124,000 - 186,000, 98-99% hydrolyzed, Sigma-Aldrich) were dissolved,
under constant magnetic stirring, at 80 ◦ C, into 5 ml of distilled water.
The second dielectric (Figure 3.22) was prepared by a sol-gel process, using a TiO2 precursor
solution prepared according to literature [Shinen et al.]. To prepare it, 5 ml of acetic acid (CH3 COOH,
acetic acid glacial, Scharlau) were slowly added to 50 ml of ethanol under continous stirring. Finally, 6.3
ml of titanium isopropoxide (TIP, Ti[OCH(CH3 )2 ]4 , Sigma-Aldrich) were slowly added to the solution and
stirred for more 2 min. All the steps were performed at room temperature.
The two dielectrics were deposited onto the electrodes by spin coating at 1800 rpm, for 30 s, in a
spin coater.
Spin coating allow us to prepare uniform thin films on flat substrates. In this process an excess
amount of a solution is applied on the substrate, for instance, with a syringe, and then, the substrate is
rotated at high speed in order to spread the fluid by centrifugal force. Rotation is continued while the
fluid spins off the edges of the substrate and dries. The final film thickness is defined by the rotational
speed and the solution concentration.
37
Figure 3.22: Titanium isopropoxide solution. The second dielectric precursor. It consisted on a solution
of 50 ml of ethanol, 5 ml of acetic acid (CH3 COOH, acetic acid glacial, Scharlau) and 6.3 ml of titanium
isopropoxide (TIP, Ti[OCH(CH3 )2 ]4 , Sigma-Aldrich).
3.5
Atomic Force Microscopy
tab The atomic force microscope (AFM) is a particular kind of scanning probe microscope (SPM). These
are designed to measure local properties, such as height, friction, magnetism, with a probe.
AFM operates by measuring the force between a probe and the sample. Usually, the probe is a
sharp tip, which is a 3-6 µm tall pyramid with 15-40 nm end radius. As the cantilever suffers deflections
an optical lever operates by reflecting a laser beam on the back of it. The reflected laser beam strikes a
position-sensitive photo-detector, indicating the angular deflections of the cantilever [Zang, 2013].
From the samples prepared, in the phase Electrodes I - Manufacture of the AAO templates by aluminum anodization, AFM images were taken. It was possible to confirm that the nano-porous AAO
template was obtained and the images acquired are shown below, in section 4.
The AFM images were analyzed using the free software Gwyddion, which is a data visualization and
analysis program for SPM images, such as AFM.
3.6
Electrodes II - Evaporation of the aluminum contact
tab The first step in the production of this second type of electrodes was the evaporation of an aluminum
film on a Polyethylene terephthalate (PET) substrate, in order to form the electrode contact.
First, the PET substrates were cut into 2 x 2 cm2 squares, and were sonicated in isopropanol for 5
min. Then, a tape mask was placed, leaving only a circle with a 1.2 cm diameter exposed.
38
The masked substrates were placed in a vacuum evaporator. The evaporator was left for 1.5 h to
achieve vacuum, and then, aluminum was evaporated for 20 minutes from aluminum wire.
3.7
Electrodes II - Deposition of the PEDOT:PSS and Multiwall Carbon Nanotubes composite film
tab Next, a small hole in the center of the aluminum circle was made, and conductive silver epoxy
paste was placed on the other side of the substrate, covering the hole, ensuring that an electric contact
between the aluminum circle and the silver dot on the back of the PET substrate is made.
A dispersion of 0.015 g of Multi-wall Carbon Nanotubes (MWCN, diameter betwen 110 nm and
170 nm, length between 5 µm and 9 µm, Sigma-Aldrich) and 3 ml of ethylene glycol (C2 H6 O2 , anhydrous, 99.8%, Sigma-Aldrich), with 0.003 g of SDS (sodium dodecyl sulfate, 98%, Sigma-Aldrich) surfactant, was made and sonicated for 1 h. Part of this dispersion was mixed with PEDOT:PSS (poly(3,4ethylenedioxythiophene):polystyrene sulfonate (1.3 wt % dispersion in H2 O; composition: PEDOT content, 0.5 wt. % and PSS content, 0.8 wt. %; conductive grade from Sigma-Aldrich) in a proportion of 10
wt% of PEDOT:PSS, and left to be magnetically stirred for 12 h.
Figure 3.23: PEDOT:PSS + MWCN dispersion.
PEDOT:PSS is a mixture of two polymers, being PEDOT a conjugated polymer. Polystyrene Sulfonic
acid is mixed with PEDOT, doping it (thereby increasing its conductivity). The blend can be dispersed in
water. Together the macromolecules form a macromolecular salt. Films of PEDOT:PSS are conducting
with high flexibility [Groenendaal et al., 2000]. By adding MWCN, the conductivity is improved. Also,
the nanotubes provide irregularities on the surface of a deposited composite film, thereby increasing the
surface area.
39
The dispersion was drop casted into the aluminum circles, with the tape masks still on. Then, the
substrates were let to dry and then the tape masks were removed, living an aluminum circle covered by
a PEDOT:PSS and carbon nanotubes composite film (Figure 3.24).
Figure 3.24: Electrodes II. The circle is the aluminum electrode covered by the PEDOT:PSS + MWCN
composite film.
3.8
Electrodes II - Deposition of the dielectric (spin coating)
tab Due to the fact that the PEDOT:PSS films are soluble in water based solutions, only the titanium
dioxide precursor sol-gel based solution was deposited onto these substrates.
The substrates were placed on the spin coater, and the TIP sol-gel solution was spin coated for 30
seconds, at 1800 rpm.
3.9
Conductance and Capacitance measurements
tab The conductance of the electrodes was measured indirectly by a handheld LCR meter (from Agilent)
on the R (resistance) mode. The measurement was done simply by placing one of the contacts on the
electrode metallic contact, and the other in several points of the nanowired surface of the electrode. The
resistance measured was the total resistance of the series Electrode Resistance + Apparatus Resistance
(Figure 3.25). The Apparatus Resistance was measured afterwards, simply by connecting the + and terminals, and substracted.
Since the conductance (Siemens) is the inverse of the resistance (Ohms), it can be obtained by a
simple calculation.
To measure the capacitance, a setup (Figure 3.26) was assembled which consisted on a copper
paper glued on a cardboard with an area (30 x 15 cm2 ) much larger than the electrode apparent surface
40
Figure 3.25: Conductivity measurement scheme. The resistances Re and Ra are in serie, so to measure
Re, we need to subtract the Ra resistance from the measurement done on the electrode.
area. This size difference is to ensure that the electric field lines created between the plate and the
electrode are parallel to each other.
The plate and each electrode were placed parallel to each other, operating as the two plates of
a capacitor. They were connected to the handheld LCR meter, now in the C (capacitance) mode, to
measure it.
The capacitance measurements were done for the frequencies of 100 Hz and 1 kHz, at the distances
of 1 mm, 5 mm and 10 mm.
Figure 3.26: Capacitance measurement scheme. The electrode and the parallel bigger plate form a
capacitor whose capacitance can be measured by the handheld LRC meter. The electrode has a guard
placed on its back to protect from parasitic electric fields; the guard is not connected to the electrode,
just to the non conductive PDMS film that the electrode has.
41
To prevent interference from other electric fields, the electrode was placed in a copper paper quilted
styrofoam. The metallic contact button of the electrode was not in contact with the copper paper, only
the non conductive PDMS layer of the electrode touched the copper paper. This copper paper quilted
styrofoam constituted a guard for the electrode. The guard was connected to the mass to nullify any
electric field lines between the back surface and sides of the electrode and any radiator close by (either
from the bigger plate or from other apparatus), restricting the electrode’s electric surface area to its
nanowired surface.
The equivalent circuit of the system is depicted in Figure 3.27.
Figure 3.27: Equivalent circuit. The branch with the Ve voltage drop represents the plate-electrode
system, where the resistance parallel to the capacitor is the resistance for the total signal loss (external
and internal). The branch capacitor with the Vg voltage drop represents the guard circuit. The voltage
drop at nodes 1 and 2 is mesured and subtracted to obtain Ve.
In the circuit, the branch with the Ve voltage drop represents the plate-electrode system capacitor,
where the resistance parallel to the capacitor is the resistance for the total signal loss (external and
internal). In order to obtain the voltage drop at impedance Ze (see Figure 3.26), the voltage drop at the
nodes 1 and 2 (Figure 3.27) is measured, to the mass, and the resulting difference yields Ve.
In the circuit, the branch that comes out of the node 2 and has a capacitor connected to the mass
representes the guard circuit, allowing us to obtain the voltage drop Ve without interference of parasitic
field lines.
This voltage is proportional to the impedance Ze, which in turn, is proportional to the electric reactance (capacitive), as it follows in equation 3.1
Z = R + jX
(3.1)
Z is the impedance (Ohms), R is the resistance (Ohms) and X is the reactance (Ohms).
For a capacitive reactance
X=−
1
1
=−
wC
2πf C
C is the capacitance (Farads) and f is the frequency (Hertz).
Finally, the capacitance depends on the folowing parameters.
42
(3.2)
C=
A
d
(3.3)
With = 0 r , in which 0 ≈ 8.854 × 10−12 F m−1 is the vacuum electric constant and r is the relative
static permittivity (dielectric constant) of the material between the plates (for vacuum, r = 1). A is the
area of overlap of the two plates and d is the distance between them.
The conductivity measurements were carried out only on Electrodes I. Both the capacitance of Electrodes I and Electrodes II were measured.
All measurements results are presented in section 4.
43
44
Chapter 4
Results and Discussion
tab In this chapter, we present first the AFM analysis of the nanostructure of the anodized aluminum
from substrates AlSigma and AlCookFoil.
Then, to test the dielectric material, one smooth extra type of electrodes (Electrodes E) was made:
E.1 - smooth electrode with PVA dielectric on top;
E.2 - smooth electrode with titanium oxide.
A smooth electrode without any dielectric was prepared for comparison (E.base), and the capacitance of all these electrodes is analysed.
Next, based on the results from the AFM analysis of the anodized substrates AlSigma and AlCookFoil, a process for the manufacture of Electrodes I from substrate AlLowPur was developped (as explained above). Four types of Electrodes I were assembled:
I.a) - prepared with constant current of 70 mA without dielectric;
I.b) - prepared with constant current of 20 mA without dielectric;
I.c) - prepared with constant current of 70 mA with the PVA dielectric;
I.d) - prepared with constant current of 70 mA with the titanium oxide sol-gel dielectric.
Also, a smooth electrode was made to serve as base for comparison, without dielectric (I.base).
Their conductance and capacitance were measured and analysed.
For Electrodes II, no AFM measurements were made yet, and one type of electrode was prepared:
II.a) - PEDOT:PSS + MWCN composite film with the titanium oxide sol-gel dielectric.
Also, a smooth electrode, with a film of PEDOT:PSS with no MWCN and with no dielectric, was made
to serve as reference for comparison (II.base). Then their capacitance was measured and analysed.
45
4.1
Anodization Results
4.1.1
Nanoporous structure
tab Substrates AlSigma
Substrates AlSigma were characterized by AFM. Images of a porous AAO template manufactured
with a one-step anodization at constant current of 70 mA were obtained.
The AFM images of the surface were acquired without any surface pos-anodization treatments, besides the cleaning of the remainder of the aluminum substrate with the 19 wt% HCl with 0.2M of CuCl2
solution, in order to observe the bottom of the nanopores.
(a)
(b)
Figure 4.1: Porous AAO template, bottom surface, through one-step anodization of Substrate AlSigma
at 70 mA. Two AFM images of distinct regions on the bottom surface of the porous AAO oxide template.
The template was obtained through aluminum (Substrate AlSigma) one-step anodization at constant
current of 70 mA. It can be observed the closed pore bottoms of the porous oxide layer. Both images
cover a 1 x 1 µm2 region.
Both images from Figure 4.1 were taken on two regions of 1 x 1 µm2 of the bottom surface of the
porous AAO template.
We did not succeed in obtaining images from the top surface of substrate AlSigma, due to the fact
that the template did not pass through any pos-anodization treatments, and, therefore a characteristic
thin layer of oxide debris probably remained on top (over the pores), making it impossible to stabilize the
AFM image.
Figure 4.1 shows the closed pore bottoms with a hemispherical shape, as expected. It is also clear
the non-uniformity of pore diameters, reflected on the different sizes of the pore bottoms, which is characteristic of a one-step anodization. However, it can be noticed that the bottoms of the smaller diameter
pores do not reach the same depth as the larger, more organized and uniform pores, suggesting that
the self-organization was still in progress when the process was stopped (1 h). For this reason, in subsequent experiments, performed with substrate AlLowPur, the first anodization time was increased to
1.5 h.
46
Figure 4.2: 3D AFM Image: Porous AAO template, bottom surface, obtained by a one-step anodization
process of substrate AlSigma at 70 mA. 3D AFM image of the same region from the second image from
Figure 4.1.
Figure 4.2 is a 3D image of the same region shown in the second image from Figure 4.1. This
figure clearly shows the depth difference between the bottom of the smaller diameter pores and the
bigger diameter pores.
Susbstrates AlCookFoil
For substrates AlCookFoil, AFM images of the porous AAO template manufactured through a doublestep anodization at constant current of 20 mA were obtained.
In order to allow image acquisition through AFM, the pore bottoms needed to be closed. Usually they
are closed by the barrier layer at the pore bottom (Figure 4.1), however, since the target was to observe
the top surface of the porous layer, with the pos-anodization treatment (immersion in the 5 wt% H3 PO4
solution) which removes the oxide debris layer, widens the pores, but also removes the pore bottom
barrier layer, the remaining aluminum substrate could not be removed. Therefore, the bottom surface of
the oxide could not be observed, only the top surface.
Figure 4.3 shows an AFM image of the top surface of the porous AAO template manufactured
through a double-step anodization, at constant current of 20 mA, of the substrate AlCookFoil.
Although the AFM image shows many defects and artifacts because it was very difficult to stabilize
the image, in this figure it can already be observed an organized hexagonal honeycomb-like structure,
typically obtained for the porous AAO membranes formed through a double-step anodization procedure.
In image b of Figure 4.3, the hexagonal charater of the structure is highlighted.
The distinctive six dot nanodimples that characterize the hexagonal honeycom-like structure, as referred in the literature [Nishinaga et al., 2013], are clearly observed, which is also a feature charateristic
of the double-step anodization, as mentioned above.
47
(a)
(b)
Figure 4.3: Porous AAO template, top surface, prepared via double-step anodization of substrate AlCookFoil at 20 mA. AFM images of the same region of the top of the porous AAO oxide template. The
hexagonal honeycomb-like organization of the pores is highlighted.
Figure 4.4: Enhanced Honeycomb six dot nanodimples structure. Elargement of one part of the honeycomb like structure of the porous AAO template depicted in Figure 4.3. Scale bar of 100 nm.
In Figure 4.5, the porous AAO top surface can be seen from a 3D perspective, providing a better
view of the height of the nanodimples.
The height of the nanodimples was further analyzed for the region depicted in the Figure 4.3.
To determine the height of the nanodimples, in relation to the middle point between two consecutive
nanodimples, a series of profiles (first image of Figure 4.6) were traced with a Gwyddion tool which then
allows to plot the surface topography along the straight line that constitutes the profile.
Each line profile crossed two consecutive nanodimples, as well as the space in between. The resultant topographic profiles are shown in image b of Figure 4.6.
Through an analysis of the graphics presented, using MatLab, the average height for the nanodimples
was calculated to be, approximately, 8.1 nm. This value is not representative for the entire surface area
of the electrode due to the small region in which the study was made. It only serves as a reference.
Using the same tools, the average distance between the maximum points of two consecutive nanodimples was also calculated, yielding a value of, approximately, 45.4 nm. Again, this only serves as a
reference value.
48
Figure 4.5: 3D Image: Porous AAO template, top surface, via double-step anodization of Substrate
AlCookFoil at 20 mA. 3D AFM image of the same image in the Figure 4.3.
(a)
(b)
Figure 4.6: Nanodimples study. Image a: nanodimple’s line profiles drawn on Figure 4.3, so that each
profile passes through two consecutive nanodimples. Each line profile is numbered in the image; image
b: topographic fluctuation along each line profile.
4.1.2
Pore Density
tab Substrates AlSigma
For pore density analysis, three AFM images which cover different dimensions of the surface were
considered, in order to obtain a more reliable approximation of the pore density.
The images in Figure 4.7 cover 5 x 5 µm2 , 2 x 2 µm2 and 1 x 1 µm2 areas of the bottom surface,
respectively for images a, b and c.
One major artifact, especially in image a of Figure 4.7, is a sort of image dragging, where, instead of
49
(a) Larger area: 5 x 5 µm2 or 25 µm2 .
(b) Medium area: 2 x 2 µm2 or 4 µm2 .
(c) Smaller area:1 x 1 µm2 or 1 µm2 .
Figure 4.7: Three area’s scale - Porous AAO template, bottom surface, via one-step anodization of
Substrate AlSigma at 70 mA. Bottom surface of the porous AAO template, via one-step anodization of
Substrate AlSigma at 70 mA, seen at three diferent scales (5 x 5 µm2 , 2 x 2 µm2 and 1 x 1 µm2 areas,
respectively).
small dots (nanopores bottoms), it appears that there are tubes stretching out from each dot. Nonetheless, the image was used to help estimate the pore density.
Figure 4.8 is the result of the analysis of image a of Figure 4.7 with Gwyddion algorithms and
analysis tools.
In this figure it is shown the individual pore bottoms enhanced. Some appear elongated due to
the dragging artifact in the image. This image artifact hides other possible pores. Nonetheless, using
Gwyddion analysis tools, it was possible to identify most of the pores in this area, counting up to a total
of 392 individual pores in an area of 25 µm2 . This is an underestimate of the pore density because,
besides the fact that the dragging artifact hides some possible pores, in the image, some pore bottoms
were not identified by the Gwyddion algorithms for grain selection.
The same analysis was applied to images b and c from Figure 4.7. The results are presented in
Figure 4.9.
In both these cases, the individual pore bottoms identified are no longer in a tube like shape (shape
acquired in the presence of the dragging artifact), making the count more reliable.
50
Figure 4.8: Pore density analysis: Porous AAO template, bottom surface, via one-step anodization of
Substrate AlSigma at 70 mA, larger area. The red domains in the image are the resultant identification
of the pores by the Gwyddion algorithms. The image treated is the first image from Figure 4.7.
(a)
(b)
Figure 4.9: Pore density analysis: Porous AAO template, bottom surface, via one-step anodization of
Substrate AlSigma at 70 mA, medium and smaller area. The red domains in the image are the resultant
identification of the pores by the Gwyddion algorithms. The images are the treatement of, respectively,
images b and c from Figure 4.7.
In image a of Figure 4.9, for a 4 µm2 area, 81 pore bottoms were counted, and in image b, for an
area of 1 µm2 , 24 pore bottoms were counted.
From this last result, a pore density of 24 pores/µm2 , for a porous AAO template fabricated by a
one-step anodization at constant current of 70 mA is obtained. This means that for a 4 µm2 area, 88
individual pores were expected. This is more or less the case counted for the first image of Figure 4.9,
51
with only an approximate difference of 9%, making it a good estimate. The small difference may be due
to the fact that some small pores were not identified by the analysis algorithms, or by the fact that the
porous template resultant from this one-step anodization is not extremely uniform.
However, for an area of 25 µm2 , it is expected around 600 individual pores. Compared to the 392
pores previously mentioned for Figure 4.8, this constitutes a difference of 35 %, confirming in fact that
an underestimate was made previously for Figure 4.8. Considering this, the pore count from image a
from Figure 4.7 (or Figure 4.8) was not taken into account to make the following average estimate for
pore density.
The average estimate for pore density of 22pores/µm2 was calculated from the 81pores/4µm2 ≈
20pores/µm2 for image b from Figure 4.7 (or image a from Figure 4.9) and the 24pores/µm2 from
image c from Figure 4.7 (or image b from Figure 4.9).
Substrates AlCookFoil
For pore density analysis of the templates using substrate AlCookFoil, also, three images, which
cover different dimensions of the surface, were considered.
In the Figure 4.10, a surface area of 2.7 x 2.7 µm2 , 1.2 x 1.2 µm2 , and and 1 x 1 µm2 is covered,
respectively for images a, b and c.
In image a (Figure 4.10), besides some image artifacts, the extended hexagonal honeycomb-like
pattern of the porous AAO template can be observed.
Again, using Gwydion analysis algorithms, it was possible to count the number of pores. In an area
of 7.29 µm2 (2.7 x 2.7 µm2 ), 318 pores were counted (Figure 4.11 is the result from the analysis of
image a of Figure 4.10).
The images from Figure 4.12 are the result of the application of Gwyddion algorithms and analysis
tools, just like before, to images b and c shown in Figure 4.10.
For the area of 1.44 µm2 (image a from Figure 4.12), 56 pore bottoms were counted, and for the
area of 1 µm2 (image b from Figure 4.12), 43 pore bottoms were counted.
From the last result, pore density of 43 pores/µm2 , for a porous AAO template fabricated by a doublestep anodization at constant current of 20 mA is obtained. This means that for a 1.44 µm2 area, 62
individual pores were expected. This closes to the 56 counted for image a of Figure 4.12, with only
an approximate difference of 9.7%, making it a good estimate. Also, for an area of 7.29 µm2 , with this
estimate it is expected, approximately, 314 individual pores, which is pretty close to the 318 counted
pores from Figure 4.11, with only a 1.3% difference.
In this case, for substrate AlCookFoil, there was no significant artifact like the previous image dragging that interfered with the count of the pores in the figures, and, therefore , the three images were
taken into account in the estimate average pore density calculated for this substrate.
52
(a) Larger area: 2.7 x 2.7 µm2 or 7.29 µm2 .
(b) Medium area: 1.2 x 1.2 µm2 or 1.44 µm2 .
(c) Smaller area: 1 x 1 µm2 or 1 µm2 .
Figure 4.10: Three area scales - Porous AAO template, top surface, via double-step anodization of
Substrate AlCookFoil at 20 mA. Top surface of the porous AAO template, via double-step anodization of
Substrate AlCookFoil at 70 mA, seen at three diferent scales(2.7 x 2.7 µm2 , 1.2 x 1.2 µm2 , and 1 x 1
µm2 , respectively).
Figure 4.11: Pore density analysis: Porous AAO template, top surface, via double-step anodization of
Substrate AlCookFoil at 20 mA, larger area. The red domains in the image correspond to the pores
identified by the Gwyddion algorithms. The image treated is the first image of Figure 4.10.
53
(a) Medium area: 1.2 x 1.2 µm2 or 1.44 µm2 .
(b) Smaller area: 1 x 1 µm2 or 1 µm2 .
Figure 4.12: Pore density analysis: Porous AAO template, top surface, via double-step anodization of
Substrate AlCookFoil at 20 mA, median and smaller area. The red domains in the images are the
resultant identification of the pores by the Gwyddion algorithms. The images are the treatment of,
respectively, images b and c from Figure 4.10.
The average estimate for pore density of 41pores/µm2 was calculated from the 318pores/7.29µm2 ≈
43pores/µm2 for image a from Figure 4.10 (or Figure 4.11), from the 56pores/1.44µm2 ≈ 38pores/µm2
for image b from Figure 4.10 (or image a from Figure 4.12) and from the 43pores/µm2 from image c
from Figure 4.10 (or image b from Figure 4.12).
Comparing the pore density results for a porous AAO template fabricated by a one-step anodization
at constant current of 70 mA (22pores/µm2 ) with the results for a porous AAO template fabricated by
a double-step anodization at constant current of 20 mA (41pores/µm2 ), this confirms previous reports
in literature (explained in the introduction), that for higher voltages, or in this case higher current which
obviously correspond to an higher voltage, larger pore diameters are obtained, that is the pore density
will be lower than the one obtained for smaller voltages (currents).
4.1.3
Mean pore diameter
tab Resorting again to Gwyddion algorithms and analysis tools and MatLab, it was possible to get an
estimate for the mean pore diameter.
For the previous section, in order to count the number of pores, using Gwyddion analysis tools,
certain parameters were defined such that, the red domains created by the algorithm (algorithm mask),
that can be observed in Figures 4.8, 4.9, 4.11 and 4.12, cover not only the hole of the nanopore
structure, but also part of the pore walls (this can be clearly seen in the images from the Figure 4.12).
However, to be able to obtain an estimate of the mean pore diameter, a new algorithm mask, with new
54
refined parameters that try to adjust each spot of the mask only to the hole of the nano-pore, was
specified.
Figure 4.13: Nano-pore top structure scheme.
Substrate AlSigma
For substrate AlSigma, since there are only images of the bottom side of the porous AAO template,
it is only possible to have an estimate of the nano-pore (nano-pore hole plus nano-pore walls) diameter.
Nonetheless, the new Gwyddion algorithm mask was applied to an image acquired from the porous
AAO templates resulting from the anodization of substrates AlSigma (one-step anodization at constant
current of 70 mA).
Figure 4.14: Mean pore radius analysis: Porous AAO template, bottom surface, via one-step anodization
of Substrate AlSigma at 70 mA. The purple domains are mask resultant from the new parameters.
The image was analyzed with reference to the maximum inscribed disc radius possible for each
nano-pore. This parameter, when applied to a non-circular area, gives an underestimate of the area,
because it is only based on the largest inscribed disc possible inside the area, never totally covering it.
However, it still yields a good estimate of the radius value.
The radius distribution obtained for the area shown in Figure 4.14 is shown in Figure 4.15.
55
Figure 4.15: Graphic for the nano-pore radius distribution. Porous AAO template, bottom surface, via
one-step anodization of Substrate AlSigma at 70 mA.
According to this, the maximum value for the radius possible for the nano-pores in that area is,
approximately, 121.6 nm, the mode is 95.6 nm and the average value is, excluding zero values, approximately, 81.8 nm.
The value obtained is a very rough estimate because we are not measuring directly on the nanopore hole, and only on its counter surface, and the algorithm tries to adjust a circular shape to the pore
by deficit. It is also necessary to consider that this estimate was done over a very small region of the
electrodes surface, and so only serves as a reference.
Substrate AlCookFoil
The new Gwyddion algorithm mask was applied to an image acquired from the porous AAO templates resultant from the anodization of substrates AlCookFoil (double-step anodization at a constant
current of 20 mA).
Figure 4.16: Mean pore radius analysis: Porous AAO template, top surface, via double-step anodization
of Substrate AlCookFoil at 20 mA. The purple domains are the mask resultant from the new parameters.
56
Once again, the analysis consisted in the measurement of the maximum inscribed disc radius possible in each nano-pore hole.
The radius distribution obtained for the area is shown in Figure 4.17.
Figure 4.17: Graphic for the nano-pore hole radius distribution. Porous AAO template, top surface, via
double-step anodization of Substrate AlCookFoil at 20 mA.
According to this, the maximum (circular) nano-pore hole radius value possible in that area is, approximately, 47.9 nm, the mode is 35.4 nm and the average value is, approximately, 28.4 nm. As expected,
these values are smaller than those obtained at 70 mA.
Again, the average value obtained is a rough estimate as explained above.
4.1.4
Surface Area and Pore Length
tab Substrate AlSigma
Since it was not possible to acquire an AFM top surface image of substrate emphAlSigma, it is not
possible to make an estimate of the superficial area that is gained with the pores nor of the pore length.
Substrate AlCookFoil
With substrate AlCookFoil, it was possible to acquire images of the top surface. Using Gwyddion
analysis tools, for that same image (Figure 4.16), it was possible to calculate the inner pore surface
added, which amounts to 2.42 µm2 . This represents roughly a 33% increase of surface area in relation
to the same geometrical flat region.
Using the defined purple mask, an average pore’s end depth (depth of the bottom of the pores)
value was obtained, with a standard deviation of 23.0%, and an average pore’s start depth (depth at
wich each individual pore hole begins) value was, also, obtained, with a standard deviation of 8.24%.
Subtracting these two values, an average pore length of 17.9 nm was established. A possible, quite
obvious, explanation for such small pore length, compared to the hundreths of nanometers presented
in literature, is the fact that the anodization times established before for substrate AlCookFoil are far
too short to yield a considerable pore length. Nonetheless, these times were enough to established the
organized honeycomb hexagonal like structure for substrate AlCookFoil.
57
4.2
Electrodeposition Results
tab Attempts were made to obtain AFM images of the electrodeposited nanowired nickel surfaces. However, we could not obtain a stable image.
One evidence that there are in fact nanowires on the surface is its color. Smooth layers of electrodeposited nickel have a grey tone (in the present case, a yellowish grey tone). But, when nickel nanowires
are manufactured, they have a dark black color, as it can be observed in Figure 4.18.
Figure 4.18: Nickel nanowires in commercial AAO template. The black circle is the commercial AAO
template filled with electrodeposited nickel (previous to nickel electrodeposition, the AAO template was
see through). Image from reference [Whitsett et al., 2012].
Figure 4.18 refers to a commercial porous AAO template filled with electrodeposited nickel, from
reference [Whitsett et al., 2012]. As it can be seen, the previous see through template acquired black
color due to the nickel nanowires. This is what we observed.
In fact, in some initial trials, where the development of the procedure was still being established, it
was possible to remove, in some areas, the dark black nickel nanowires from the remaining metallic
surface, which, on the other hand, was a metallic shiny grey, as it can be seen in Figure 4.19.
Therefore, although there is no AFM images of the electrodeposition result, this observation suggests
that, in fact, an complete electrodeposition occurred acording to the scheme of Figure 3.17, resulting in
an electrode’s structure as shown in the Figure 4.20.
4.3
4.3.1
Conductance and Capacitance Results
Dielectric material influence
tab Smooth electrodes were prepared to conclude about the dielectric’s hability to increase the electrodes capacitance:
E.1 - smooth electrode with PVA dielectric film deposited on top;
E.2 - smooth electrode with titanium oxide dielectric prepared via sol-gel from the TIP precursor.
E.base - smooth, dielectric-less electrode prepared for comparison.
58
Figure 4.19: Nickel nanowires on the nickel smooth layer. Manufactured nickel structure consistent with
the structure from Figure 4.20. In the image, the grey smooth metallic shiny surface can be seen in the
left side of the circle, where the nanowired layer has been scraped off.
Figure 4.20: Electrodes I layered structure scheme. Electrodes I most likely have the structured layers
shown in the scheme.
Electrodes E (Figure 4.21) consisted on a circular copper contact (conductive adhesive copper
paper; 8 mm diameter) with a thin copper strap to provide the electric connection (3 x 10 mm2 ), which
amounts to a total area of, approximately, 80.265 mm2 . Around the measuring area, there is an electric
field guard made from the same material.
The dielectrics were deposited by spin-coating for 30 seconds, at 1800 rpm, as described above.
The capacitance measurements were done at the three distances (1 mm, 5 mm and 10 mm) between
each of the electrodes and the parallel counter plate (see scheme in Figure 3.26), and for each distance
at the two distinct frequencies (100 Hz and 1 kHz).
Tables 4.1 and 4.2 show the measured capacitance values.
Electrode
E.base
E.1
E.2
1 mm
2.8 pF
3.5 pF
2.9 pF
Distance
5 mm
2.4 pF
3.1 pF
2.4 pF
10 mm
2.1 pF
2.8 pF
2.1 pF
Table 4.1: Capacitance values for Electrodes E at 100 Hz.
Results on these tables show that, at both frequencies, the capacitance of electrodes E.2 is similar
to that of electrodes E.base, being higher for electrodes E.1. That is PVA seems to improve more the
electrodes capacitance than titanium oxide. This is not what was expected and at variance with what is
59
(a) Front side.
(b) Back side.
Figure 4.21: Extra electrodes. Image a shows a circular electrode with a strap to make an electric contact
made from adhesive copper paper glued to a square of glass. Also, without touching the electrode area,
the rest of the glass is covered with the same copper paper to provide an electric guard to the electrode.
Image b is the back of the electrode.
Electrode
E.base
E.1
E.2
1 mm
4.10 pF
4.75 pF
4.09 pF
Distance
5 mm
3.57 pF
4.33 pF
3.72 pF
10 mm
3.51 pF
4.09 pF
3.60 pF
Table 4.2: Capacitance values for Electrodes E at 1 kHz.
reported in the literature, where TiO2 has a dielectric constant usually around 40 and PVA has a value
of 10 for the dielectric constant, as mentioned in section 2.3.
The fact that the PVA dielectric yields higher capacitance values than titanium oxide might rely on
the deposition method (spin coating), although, as TiO2 was prepared by sol-gel, the obtained material
maybe different from regular TiO2 . For both solutions, the same deposition conditions were applied;
however, it was visible that the titanium oxide sol-gel solution was much more fluid than the PVA solution.
When using spin coating, the more viscous the solution is, for the same spin deposition speed, the higher
the thickness of the films will be. Therefore, it maybe that the deposition spin speed for titanium oxide
sol-gel solution was too high to leave enough dielectric material to support a measurable increase of
the dielectric constant. This could also explain the fact that, for the lower frequency, the capacitance
measured for electrode E.2 is practically the same as the one measured for E.base.
According to equation 3.3, with the increase of the distance d, the capacitance should decrease,
and this is verified for all electrodes, at both frequencies (Figure 4.22).
60
(a) 100 Hz.
(b) 1 kHz.
Figure 4.22: Capacitance values for Electrodes E, at different distances. Image a is the variance of the
Electrodes E’s capacitance at 100 Hz; image b is the variance of the Electrodes E’s capacitance at 1
kHz.
Along the increasing distance, for the frequency of 100 Hz, the capacitance of electrodes E.1 has
an average increase of 29 % in relation to the E.base capacitance, with a standard deviation of 12 %.
For frequency 1 kHz, along the increasing distance, the capacitance of electrodes E.1 has an average
increase of 18 % in relation to the E.base capacitance, with a standard deviation of 14 %. This is contrary
to what was expected (for both frequencies the percentual increase of the capacitance value should be
the same). The difference in the increase between the two frequencies seems to indicate the there might
be other components in the equivalent circuit presented before. Any conductive surface nearby, from
the metallic frames of the operation table to the operator itself may induce dispersions dependent of the
frequency, leading to the different percentual increases. Also, the measuring apparatus, associated with
the previous reason, may have different sensibility for each frequency measured.
4.3.2
Electrodes I
tab Five types of Electrodes I were assembled, all made from substrate AlLowPur :
I.a) - prepared with constant current of 70 mA without dielectric;
I.b) - prepared with constant current of 20 mA without dielectric;
I.c) - prepared with constant current of 70 mA with the PVA dielectric;
I.d) - prepared with constant current of 70 mA with the titanium oxide dielectric.
I.base - smooth electrode without dielectric was made to serve as reference.
All the electrodes had the same appearance as the electrode in Figure 3.20.
The electrodes were manufactured by the process described before. It is the anodization setup
(particulary, the O-Ring) that determines the electrode’s area. Any deformation of the shape of the ORing during the setup of the process of manufacture leads to the same deformation of the electrode
shape. But, since the manufacture process was carefully done, the electrodes appear to be uniform, all
with the same shape, with a flat geometric area of 201.062 mm2 .
61
Conductance
tab To quantify electrodes electrical conductance, using the setup from Figure 3.25, the electrical resistance between the tip of the metallic contact button and a random point on the nanowired surface
of the electrode was measured. The consistent value obtained at several points was 0.2 Ω, which is a
small acceptable value for the resistance. Since the inverse quantity of the electrical resistance is the
electrical conductance, the corresponding conductance measured is 5 S.
Surface area and Capacitance
tab The capacitance measurements were, once again, done at three distances (1 mm, 5 mm and 10
mm) between the electrodes and the parallel counter plate (Figure 3.26), and each distance at two
distinct frequencies (100 Hz and 1 kHz).
The following tables show the measured capacitance results.
Electrode
I.base
I.a
I.b
I.c
I.d
1 mm
5.5 pF
7.3 pF
7.0 pF
6.0 pF
5.8 pF
Distance
5 mm
4.0 pF
4.9 pF
4.7 pF
4.0 pF
4.4 pF
10 mm
3.9 pF
4.0 pF
4.1 pF
3.7 pF
4.0 pF
Table 4.3: Capacitance values for Electrodes I at 100 Hz.
Electrode
I.base
I.a
I.b
I.c
I.d
1 mm
6.96 pF
8.72 pF
8.53 pF
7.41 pF
7.11 pF
Distance
5 mm
5.61 pF
5.89 pF
5.87 pF
5.64 pF
5.84 pF
10 mm
5.20 pF
5.50 pF
5.57 pF
5.22 pF
5.36 pF
Table 4.4: Capacitance values for Electrodes I at 1 kHz.
Figure 4.23 shows the capacitance of Electrodes I as a function of the distance.
As it can be observed, electrodes I.a and I.b have higher capacitance than the others. This two
electrodes, besides I.base, were the ones in which the dielectric film was not deposited. This is contrary
to what was expected because the high-k dielectric material should improve the capacitance measurement of the electrodes on which it was deposited on (electrodes I.c and I.d), from the equation 3.3. The
medium between the large plate and the electrodes with the dielectric on top (electrodes I.c and I.d) is
composed by air and the thin layer of dielectric, and therefore should have a higher dielectric constant
than a medium composed of only air (medium for electrodes without dielectric, electrode I.a and I.b).
One possible explanation for the higher value of electrodes I.a and I.b is that the dielectric deposition
method used may not have been the most appropriate. The spinning motion that is responsible to
spread the dielectric film, associated with the dielectric motion itself, may have caused the nanowires
62
(a) 100 Hz.
(b) 1 kHz.
Figure 4.23: Capacitance values for Electrodes I, at different distances. Image a shows the capacitance
of Electrodes I measured at 100 Hz; image b shows the capacitance of Electrodes I measured at 1 kHz.
to collapse, decreasing the active surface area. However, further studies are required, namely a SEM
analysis of the electrodes surface to determine if indeed the surface area was affected by the dielectric
deposition.
Nevertheless, for all electrodes, as before, with the increase of the distance, the capacitance value
decreases.
The highest capacitance value, for both frequencies, corresponds to the electrode I.a, prepared with
constant current of 70 mA and without dielectric film, having, at the minimum distance, a capacitance
increase of 25.29% and 32.73% for 100 Hz and 1 kHz, respectively, with respect to the value measured
for I.base. Once angain, as previously for Electrodes E, a difference in the capacitance increase between
the two frequencies is verifyed, and, as before, this may be due to the presence of the conductive
surfaces that induce dispersions dependent of the frequency associated with the fact that the measuring
apparatus may have different sensibility for each frequency measured.
Considering the air’s permittivity to be, approximately, 8.854×10−12 F/m, from the capacitance equation (equation 3.3) an estimate of the increase in the electrodes I.a and I.b (electrodes without dielectric)
surface area can be made, in relation to electrode I.base. These estimates of the area increase (in %)
are shown in tables 4.5 and 4.6.
Electrode
I.a
I.b
1 mm
32.7%
27.3%
Distance
5 mm
22.5%
17.5%
10 mm
2.6%
5.1%
Table 4.5: Area increase (in %) with respect to the area of I.base flat electrode, for Electrodes I at 100
Hz.
From the values on tables 4.5 and 4.6, for the same frequency, the estimated percentual area
increase should have been maintained along the different distances, and that does not happen. We do
63
Electrode
I.a
I.b
1 mm
25.3%
22.6%
Distance
5 mm
5.0%
4.6%
10 mm
5.8%
7.1%
Table 4.6: Area increase (in %) with respect to the area of I.base flat electrode, for Electrodes I at 1 kHz.
not have at present an explanation for this phenomenon since the measuring system was the same as
for the measurement of the previous type of electrodes (Electrodes E).
Also, at distance of 1 mm, the percentual area increase for both type of electrodes is in the same
range of the measured (roughly) surface area increase for substrate AlCookFoil in section 4.1.4 of about
33 %. It was expected that the area’s increase would be higher than the 33 % because the anodization
time was much longer for substrate AlLowPur than for substrate AlCookFoil, but that doesn’t happen.
A possible explanation is the fact that the difference in the type of substrates affects each anodization
rate, or maybe the oxide dissolution rate, resulting in similar thickness porous AAO on both susbstrates
AlCookFoil and AlLowPur even though there is a large difference in the anodization times.
4.3.3
Electrodes II
tab Electrodes II, have the structure:
II.a) - PEDOT:PSS + MWCN composite film with the titanium oxide dielectric.
A smooth reference electrode, with a film of PEDOT:PSS without MWCN and without dielectric
(II.base) was prepared.
Capacitance
tab The capacitance measurements were carried out, as above, at the three distances (1 mm, 5 mm
and 10 mm) between the electrodes and the parallel counter plate (Figure 3.26), and each distance at
the two distinct frequencies (100 Hz and 1 kHz).
Tables 4.7 and 4.8 show the measured values for Electrodes II.
Electrode
II.base
II.a
1 mm
4.2 pF
4.3 pF
Distance
5 mm
3.5 pF
3.8 pF
10 mm
3.3 pF
3.5 pF
Table 4.7: Capacitance values for Electrodes II at 100 Hz.
Electrode
II.base
II.a
1 mm
5.76 pF
5.25 pF
Distance
5 mm
5.17 pF
4.67 pF
10 mm
4.96 pF
4.50 pF
Table 4.8: Capacitance values for Electrodes II at 1 kHz.
64
From the results, we can see that some capacitance values for II.a are actually smaller than those
for II.base. Also, a significant difference in the capacitance values associated to the use of the two
electrodes appears to exist, being the differences always smaller than 10%. This may be due to the
fact that the nanotubes present in the film did not have a significant impact on the surface area. Further
studies are required, namely by AFM, to confirm the surface roughness increase of the PEDOT:PSS
films upon addition of the MWCN.
65
66
Chapter 5
Conclusions
tab In this work, two procedures to manufacture capacitive electrodes based on a proposed design were
presented. The main process studied resorted to the manufacture of porous anodic aluminum oxide
templates, and subsequent nickel filling. With this procedure, it is possible to obtain a highly organized
nanowired surface on the electrodes, increasing the surface area. The second, alternative, method
resorted to PEDOT:PSS with Multi-Wall Carbon Nanotubes composite films so that the nanotubes would
provide a certain surface roughness and thereby increasing the electrodes capacitance. This method,
although it does not provide an organized structure for the electrodes surface like the previous method
(which is not required), represents a good alternative for the manufacture of the desired electrodes.
The methods and materials involved in both types of electrodes were not expensive.
5.1
Achievements
tab Starting with the manufacture of Electrodes I, from the first porous AAO templates from aluminum
substrates AlSigma and AlCookFoil it was possible to verify by AFM studies the formation of the selfassembled porous template, particularly of the honeycom-like structure easily identified on substrates
AlCookFoil, created by the double-step anodization.
Substrates AlSigma underwent a one-step anodization at constant current of 70 mA and substrate
AlCookFoil underwent a double-step anodization at constant current of 20 mA, yielding an approximate
of 22pores/µm2 and 41pores/µm2 respectively. This indirectly confirms literature reports, according to
which higher voltages (higher current) lead to larger pore diameters. If the pore diameters are larger,
than, for the same area, less pores will fit, and so the pore density will be lower than the one obtained
for smaller voltages (smaller currents). The analysis of the pore diameter also confirmed this, having the
substrates yielded an average value of 163.6 nm and 56.8 nm for pore diameter for substrates AlSgma
and subtrates AlCookFoil respectively.
Due to the template characteristics explained before, an estimate of the AlSigma nanoporous template surface area increase was not possible to make, but for the templates made from substrate AlCookFoil (double-step anodization at constant current of 20 mA) there was an estimated increase of
67
33% of the surface area.
From the results of the capacitance measurements of Electrodes I (four types were fabricated: 1)
prepared with a porous template obtained through double-anodization with constant current of 70 mA
without dielectric; 2) prepared with a porous template obtained through double-anodization with constant current of 20 mA without dielectric; 3) prepared with a porous template obtained through doubleanodization with constant current of 70 mA with the PVA dielectric; 4) prepared with a porous template
obtained through double-anodization with constant current of 70 mA with the titanium oxide dielectricted)
made from substrate AlLowPur, it was observed that the capacitance increase varies with the frequency.
Such phenomenum may be due to the dispersive influence of any conductive surface nearby, from the
metallic frames of the operation table to the operator itself, which may induce dispersions dependent of
the frequency. Also, the measuring apparatus, associated with the previous reason, may have different
sensibility for each frequency measured.
Also, for electrodes I.a (70 mA, no dielectric) and I.b (20 mA, no dielectric), the calculated percentual
area increase, at each frequency, should have been maintained with the varying distance, and that does
not occur. A possible explanation for this observation is not yet clear. Additional studies are required.
Considering the capacitance values obtained at 1 mm distance, due to the longer anodization times,
the percentual area increase for both types of electrodes was expected to be higher than the 33 %
measured surface area increase for substrate AlCookFoil in section 4.1.4. That does not happen,
although it’s close. One possible explanation, as mentioned before, is the fact that the type of substrates
may affect its anodization rate and/or maybe the oxide dissolution rate, resulting in similar thickness
porous AAO on both susbstrates AlCookFoil and AlLowPur even though there is a large difference in
the anodization times. However, to verify this, other studies involving AFM or SEM imaging of the porous
templates fabricated from substrate AlLowPur are required.
Regarding the dielectric material to be deposited on top of the electrodes, from the initial measurements done on the flat electrodes E, it was found that PVA yielded better results than titatinum dioxide.
As mentioned before, since this finding does not agree with literature reports, further studies on the process to obtain the titanium dioxide dielectric, or alternatives for this dielectric deposition method, should
be done.
The results obtained for the PVA dielectric on flat electrodes E evidenced the same behaviour, in
which the capacitance increase varyed with the frequency, as in Electrodes I probably for the same
reasons. An average capacitance increase of 29 % and 18 % was observed for the frequencies of 100
Hz and 1 kHz, respectively.
For the nanowire based elecrodes (Electrodes I), when the dielectric was applied the capacitance
measurements yield very poor results with respect to the same electrodes without dielectric, which was
quite unexpected. A possible cause for this result is that the deposition method used (spin coating)
may have caused the colapse of the nanowire structures, decreasing the surface area. However, further
studies, for instance the analysis of the electrodes surface before and after dielectric deposition, by SEM
imaging, are necessary.
68
Finally, regarding Electrodes II, their purpose was simply to assess if such simple method (rough
film prepared by spin coating) could yield results close enough to the ones obtained for the porous AAO
templates that would justify change on the manufacture method, but the results so far obtained failed to
show that. One reason that could justify the obtained results for Electrodes II is, possibly, the low weight
percentage of nanotubes used in the composite films, which was not enough to cause a significant increase of the surface (roughness) area. Further studies, namely by AFM analysis of the composite films
surface, would be quite profitable since this approach to increase the electrodes surface area is quite
simple, and would constitute an excelent process to the manufacture of high surface capacitive electrodes. For instance, maybe by increasing the carbon nanotubes content in the dispersion, a significant
increase of the capacitance value could be achieved.
5.2
Future Work
tab As future work, before improving the electrode’s characteristics, a better, more accurate system for
measuring the capacitance should be established, and the capacitance measurements should be done
at higher distances and frequencies, to fully understand the behaviour of both electrodes. Also, in future
experiments, to better examine the nanowired structure of Electrodes I surface, and also the topography of Electrodes II surface, additional characterization methods, such as SEM (scanning electron
microscopy) imaging, should be explored.
Further, in order to provide flexibility to the electrodes, allowing its use in other applications, filing
of the porous AAO templates with polymers with flexibility close to that of PDMS should be explored
(although in case of PDMS, since this is not a conductive polymer could not serve as electrode).
Regarding the electrodes, other deposition methods for the dielectric material should be investigated
and also other methods to obtain the titanium dioxide dielectric. Further studies on a composite dielectric
of PVA with nanoparticles of titanium dioxide may be advantageous, since the PVA dielectric is able
to maintain a certain flexibility (which, in the future works may be achievable for the electrodes), by
associating the titanium dioxide particles with PVA, a material with a higher dielectric constant than PVA
may be obtained, with a similar flexibility to that of PVA.
69
70
Bibliography
Fabrication of commercially viable anodic aluminum oxide membranes by anodizing commercial aluminum cooking foils. Nanotechnology Materials and Devices Conference (NMDC), IEEE, October
2011.
Effect of nanostructured surfaces on conductive ink printing. LOPE-C, 2012.
Vertically aligned nickel nanowire/epoxy composite for electrical and thermal conducting material. 62nd
Electronic Components and Technology Conference, IEEE, May 2012.
T. Aerts, T. Dimogerontakis, I. De Graeve, J. Fransaer, and H. Terryn. Influence of the anodizing temperature on the porosity and the mechanical properties of the porous anodic oxide film. Surface and
Coatings Technology, 201(16-17):7310–7317, May 2007.
G. Ali and M. Maqbool. Fabrication of cobalt-nickel binary nanowires in a highly ordered alumina template via ac electrodeposition. Nanoscale research letters, 8(1):352, January 2013.
A. O. Araoyinbo, A. Rahmat, M. N. Derman, and K. R. Ahmad. Room temperature anodization of
aluminum and the effect of the electrochemical cell in the formation of porous alumina films from acid
and alkaline electrolytes. Adv. Mat. Lett, 3(3):273–278, July 2012.
A. Basile and F. Gallucci. Membranes for Membrane Reactors. Wiley Online Library, 2010.
A. Belwalkar, E. Grasing, W. V. Geertruyden, Z. Huang, and W. Z. Misiolek. Effect of processing parameters on pore structure and thickness of anodic aluminum oxide (aao) tubular membranes. Journal of
Membrane Science, 319(1-2):192–198, July 2008.
G. Brezeanu, M. Brezeanu, and F. Bernea. High-k dielectrics in nano&microelectronics.
I. Brown, J. Wu, and G. Smith. Ceramic materials - progress in modern ceramics. chapter 10, pages
191–226. InTech, April 2012.
D. Brüggemann. Nanoporous aluminium oxide membranes as cell interfaces. Journal of Nanomaterials,
2013(32), January 2013.
A. Cortés, G. Riveros, J. L. Palma, J. C. Denardin, R. E. Marotti, E. a. Dalchiele, and H. Gómez. Singlecrystal growth of nickel nanowires: Influence of deposition conditions on structural and magnetic
properties. Journal of Nanoscience and Nanotechnology, 9(3):1992–2000, March 2009.
71
J. Di Bari. Electrodeposition of nickel. In M. Schlesinger and M. Paunovic, editors, Modern Electroplating,
chapter 3, pages 79–114. John Wiley and Sons, 2000.
M. Domaska, M. Norek, W. J. Stepniowski, and B. Budner. Fabrication of high quality anodic aluminum
oxide (aao) on low purity aluminum - a comparative study with the aao produced on high purity aluminum. Electrochimica Acta, 105:424–432, August 2013.
O. Fliniaux, C. Elie-Caille, J. Pantigny, and C. Bourdillon. Production of microporous aluminum oxide
electrodes as supports for tethered lipid bilayers of large surface area. Electrochemistry Communications, 7(7):697–702, July 2005.
L. Groenendaal, F. Jonas, D. Freitag, H. Pielartzik, and J. R. Reynolds.
Poly (3, 4-
ethylenedioxythiophene) and its derivatives: past, present, and future. Advanced Materials, 12(7):
481–494, 2000.
J. H. Ho, H. K. Cheol, H. J. Yun, H. B. Suc, B. M. Hak, and H. C. Jin. Effect of annealing on the magnetic
properties of ni nanowires prepared by using an anodized aluminum oxide template. Journal of the
Korean Physical Society, 58(31):654–658, March 2011.
R. Huamanrayme Bustamante.
Fabricação de nanoestruturas de alumina anódica porosa e suas
aplicações na sı́ntese de nanomateriais. PhD thesis, Universidade de São Paulo, June 2012.
P. International. Anodizing of aluminium, January 2013.
J.-H. Jeong, S.-H. Kim, Y. Choi, and S.-S. Kim. Microstructure of nanopores in aao templates favoring the
growth of nickel nanowires by electrodeposition. Physica Status Solidi, 4(12):4429–4432, December
2007.
O. Jessensky, F. Muller, and U. Gosele. Self-organized formation of hexagonal pore arrays in anodic
alumina. Applied Physics Letters, 72(10):1173–1175, January 1998.
J. K. Kasi, A. K. Kasi, M. Bokhari, and N. Afzulpurkar. Synthesis of unique structures of carbon nanotube
at anodic aluminum oxide template. Applied Mechanics and Materials, 421:319–323, September
2013.
M. Kim, Y.-C. Ha, T. N. Nguyen, H. Y. Choi, and D. Kim. Extended self-ordering regime in hard anodization and its application to make asymmetric aao membranes for large pitch-distance nanostructures.
Nanotechnology, 24(50), December 2013.
W. Lee, R. Ji, U. Gösele, and K. Nielsch. Fast fabrication of long-range ordered porous alumina membranes by hard anodization. Nature Materials, 5(9):741–747, September 2006.
W. Lee, K. Nielsch, and U. Gösele. Self-ordering behavior of nanoporous anodic aluminum oxide (aao)
in malonic acid anodization. Nanotechnology, 18(47):1–8, October 2007.
F. Li, L. Zhang, and R. Metzger. On the growth of highly ordered pores in anodized aluminum oxide.
Chemistry of Materials, 10(9):2470–2480, August 1998.
72
L. C. D. Lima, M. P. D. Albuquerque, M. M. G. D. Macedo, and R. A. Simão. Morphological characterization of the surface of nano-porous aluminum oxide by afm and laws of scale. NANO: Brief Reporst
and Reviews, 4(3):157–164, June 2009.
H. Masuda, K. Yada, and A. Osaka. Self-ordering of cell configuration of anodic porous alumina with
large-size pores in phosphoric acid solution. Japanese Journal of Applied Physics, 37(11A):L1340,
November 1998.
W. Na, Z. Wen-Di, X. Ji-Peng, M. Bin, Z. Zong-Zhi, J. Qing-Yuan, E. Bunte, J. Hüpkes, and H. P. Bochem.
Fabrication of anodic aluminum oxide templates with small interpore distances. Chinese Physics
Letters, 27(6), June 2010.
T. M. Naieff and K. H. Rashid. Comparative study for anodizing aluminum alloy 1060 by different types
of electrolytes solutions. In first scientific conference on modern technologies in oil and gas refineryChemical Eng. Dept., University of technology and midland refineries company, pages 25–27.
K. Nielsch, F. Müller, a. P. Li, and U. Gösele. Uniform nickel deposition into ordered alumina pores by
pulsed electrodeposition. Advanced Materials, 12(8):582–586, April 2000.
P. Nielsen, J. Beermann, O. Albrektsen, S. Hassing, P. Morgen, and S. I. Bozhevolnyi. Two-photon luminescence microscopy of large-area gold nanostructures on templates of anodized aluminum. Optics
Express, 18(16):17040–17052, 2010.
O. Nishinaga, T. Kikuchi, S. Natsui, and R. O. Suzuki. Rapid fabrication of self-ordered porous alumina
with 10-/sub-10-nm-scale nanostructures by selenic acid anodizing. Scientific Reports, 3(2748):1–6,
January 2013.
H.-J. Oh and C.-S. Chi. The electrochemical behaviors of barrier-type anodic films on aluminum in
ammonium adipate solution. Bull. Korean Chem. Soc, 21(2):193–199, August 2000.
C. P. D. Oliveira. Eletrodeposição de nanoestruturas de cobalto em alumina anódica porosa e sua
caracterização magnética. Master’s thesis, Universidade Federal de São Carlos, 2009.
S. Ono, M. Saito, and H. Asoh. Self-ordering of anodic porous alumina formed in organic acid electrolytes. Electrochimica Acta, 51(5):827–833, November 2005.
R. P. Ortiz, A. Facchetti, and T. J. Marks. High-k organic, inorganic, and hybrid dielectrics for low-voltage
organic field-effect transistors. Chemical Reviews, 110(1):205–239, October 2009.
J. O’sullivan and G. Wood. The morphology and mechanism of formation of porous anodic films on aluminium. Proceedings of the Royal Society of London. Series A, Mathematical and Physical Sciences,
317:511–543, 1970.
L. G. Parkinson, N. L. Giles, K. F. Adcroft, M. W. Fear, F. M. Wood, and G. E. Poinern. The potential of
nanoporous anodic aluminium oxide membranes to influence skin wound repair. Tissue engineering.,
15(12):3753–3763, December 2009.
73
G. Patermarakis. The origin of nucleation and development of porous nanostructure of anodic alumina
films. Journal of Electroanalytical Chemistry, 635(1):39–50, October 2009.
S. T. Picraux. Nanotechnology. http//www.britannica.com/EBchecked/topic/962484/nanotechnology/236454/Bottomup-approach, October 2013.
G. E. J. Poinern, N. Ali, and D. Fawcett. Progress in nano-engineered anodic aluminum oxide membrane
development. Materials, 4(3):487–526, February 2011.
A. P. Samantilleke, J. A. Carneiro, P. Alpuim, V. M. Teixeira, and T. T. T. Thuy. Nanoporous alumina
templates: anodisation and mechanical characterisation. In Nanotechnology, volume 4. JN Govil,
2012.
T. Sato. Mechanism of electrolytic coloring of anodized aluminum. Plating and surface finishing, 3:
70–72, March 1991.
M. H. Shinen, K. I. Ajeel, and F. A. Rasin. Preparation of titanium dioxide (tio2) via the sol-gel process.
R. Singh and R. K. Ulrich. High and low dielectric constant materials. The Electrochemical Society,
pages 26–30, 1999.
Z. Su and W. Zhou. Porous anodic metal oxides.
G. D. Sulka.
Highly ordered anodic porous alumina formation by self-organized anodizing.
In
A. Eftekhari, editor, Nanostructured Materials in Electrochemistry, chapter 1, pages 1–116. John Wiley
& Sons, January 2008.
G. D. Sulka, S. Stroobants, V. Moshchalkov, G. Borghs, and J.-P. Celis. Synthesis of well-ordered
nanopores by anodizing aluminum foils in sulfuric acid. Journal of The Electrochemical Society, 149
(7):D97–D103, May 2002.
M. Vorozhtsova, J. Drbohlavova, and J. Hubalek. Chemical microsensors with ordered nanostructures.
In O. Minin, editor, Microsensors. InTech, 2011.
J. Whitsett, S. M. Condren, and G. Lisenskyl. Synthesis of Nickel Nanowires, June 2012.
A. Yin, M. Tzolov, D. Cardimona, L. Guo, and J. Xu. Fabrication of highly ordered anodic aluminium
oxide templates on silicon substrates. Circuits, Devices & Systems, 1(3):205–209, June 2007.
L. Zang. Lecture 10: Basics of Atomic Force Microscope (AFM), October 2013.
J. Zhang, J. E. Kielbasa, and D. L. Carroll. Controllable fabrication of porous alumina templates for
nanostructures synthesis. Materials Chemistry and Physics, 122(1):295–300, July 2010.
F. Zhou. Growth mechanism of porous anodic films on aluminium. PhD thesis, University of Manchester,
2011.
T. Zhou, M. Zheng, L. Ma, Z. He, M. Li, C. Li, and W. Shen. Electronic Supplementary Information (ESI)
Size control of CuInSe2 nanotube arrays via nanochannel-confined galvanic displacement. Journal of
Materials Chemistry, pages 1 – 6, 2011.
74