"Anodic aluminum oxide processing, characterization

and application to DNA hybridization electrical detection"

Moreno Hagelsieb, Luis

ABSTRACT

Metal oxides have recently come under study thanks to their physical and electrical properties for different
applications such as MOS devices, i.e. substituting the silicon oxide with a high-k material, or as MIM
(Metal Insulator Metal) capacitors, to increase capacitance per unit area and circuit integration. One oxide
of interest in this field is aluminum oxide since it features good electrical insulation and high dielectric
constant. In-depth studies are presented here on the use of non-porous anodic Al2O3. Major physical and
electrical parameters have been obtained, i.e. dielectric constant, stress, deformation, resistance, surface
quality. Constant, low anodizing current density results in a denser oxide, with a thickness of around 100
nm. Performances such as capacitance, breakdown voltage, etc. can be improved when compared to other
Al2O3 obtained by other methods. Results are also comparable to other high-k oxides. Fair performance is
maintained for temperature raised up to 200°C, which opens new possible applications. Its mechanical and
physical properties make it candidate in biological and MEMS devices. DNA re-association or hybridization
is the underlining principle of DNA sensors. Different electrical Al structures protected by a thin anodic
Al2O3 are tested. Interdigitated capacitors, the most promising electrical structure, were selected and
process characterization performed. Three electrical extraction procedures are performed on the same
device lying on a passivated silicon substrate: inter-electrodes capacitance, the self-resonance frequency,
and the e...

CITE THIS VERSION

Moreno Hagelsieb, Luis. Anodic aluminum oxide processing, characterization and application to DNA
hybridization electrical detection. Prom. : Flandre, Denis http://hdl.handle.net/2078.1/5024

Le dépôt institutionnel DIAL est destiné au dépôt DIAL is an institutional repository for the deposit
et à la diffusion de documents scientifiques and dissemination of scientific documents from
émanant des membres de l'UCLouvain. Toute UCLouvain members. Usage of this document
utilisation de ce document à des fins lucratives for profit or commercial purposes is stricly
ou commerciales est strictement interdite. prohibited. User agrees to respect copyright
L'utilisateur s'engage à respecter les droits about this document, mainly text integrity and
d'auteur liés à ce document, principalement le source mention. Full content of copyright policy
droit à l'intégrité de l'œuvre et le droit à la is available at Copyright policy
paternité. La politique complète de copyright est
disponible sur la page Copyright policy

Available at: http://hdl.handle.net/2078.1/5024 [Downloaded 2024/04/25 at 18:49:08 ]

 CHAPTER I

INTRODUCTION
Oxides form a broad subject area of research and technology development. For more
than 40 years, silicon oxide and low permittivity dielectrics have been used in
microelectronic fabrication. Metal oxides have come more recently under study thanks
to their physical and electrical properties for different applications such as MOS
devices, i.e. substituting the silicon oxide with a high-K material, or as MIM (Metal
Insulator Metal) capacitors, to increase capacitance while achieving devices size
reduction and circuit integration increase. One oxide of interest in this field is aluminum
oxide since it features good electrical insulation properties as well as a high dielectric
constant and good mechanical properties of great interest for MEMS fabrications.
Anodization has been used for obtaining aluminum oxides that thanks to their
hydrophilic property have been under study for applications in humidity sensors and for
the cross-linking of DNA, proteins, etc., for chemical and biological sensors in
development that could be eventually used for research and medical applications, as the
origin of the present work.

In this introduction, we will briefly review the basics on metallic oxides, the state of the
art of the anodic aluminum oxide in the micro electronic world, MIM basic
requirements and the state of the art so far in the biosensing electrical detections to
finish with a short review of the content of this thesis.

1.1.- Metallic Oxides

Oxides form a broad subject area of research and technology development, which
encompasses different disciplines such as materials science, solid-state chemistry,
physics etc.

For more than 30 years, silicon dioxide (SiO2) has been the versatile dielectric barrier
material of choice for the semiconductor device industry. It has provided adequate high-
capacitance gate insulation at the front end of the line and sufficiently low-k, crosstalk-
free insulation between interconnect wiring levels at the back end. Now, due to scaled-
down chip components, more densely populated chip real estate, and the introduction of
low-resistance interconnect materials like copper, the industry’s familiar, reliable SiO2
“fencing” is no longer good enough for either application.

There are stronger demands for advanced components that feature large data storage,
high-speed, and/or multi-functions. However, such demands rise up problems that
traditional semiconductor electronics cannot solve. For example, the low dielectric
constant of silica, used as the insulating material for the polysilicon gate electrode,
imposes serious limitations for downsizing. Such a limitation can be overcome if we
will be able to use high dielectric (high-K) oxide materials as the insulator.

Other materials have been developed with lower dielectric constants for interconnect
applications and higher dielectric constants for gate or capacitor applications [1-4].
However, integrating these materials into the wafer and device has brought many

1
 CHAPTER I - INTRODUCTION

challenges that have delayed their introduction. As devices continue to shrink, even
lower- or higher-k materials are needed, posing even more integration challenges.

Ideas are to substitute the silicon oxide with another oxide, especially metallic ones
which show higher dielectric constant values, or making a two oxide layer combining
both the silicon and the metallic oxides.

In addition, metallic oxides are materials recently in study thanks to their physical and
electrical properties for different applications such as conductors, semiconductors,
insulators, piezoelectrics, ferroelectrics, ferromagnetics, antiferromagnetics,
superconductivity, CMR materials, HTSC, and nonlinear optical materials. Therefore,
thin films of these oxide materials have a high potential for device applications. A large
number of materials are under investigation to be employed in future devices, for
example in electric field-effect devices, superconducting junctions, in magnetic tunnel
junctions, as exchange bias layers in GMR heads, and in tunable high-frequency
devices.

The potential of these materials for device applications is excellent, but some of the key
factors controlling the physics, for instance the doping level and the structural order, are
often difficult to control. Moreover, the compatibility of different oxides in terms of
interface structure and electronic properties is a poorly understood issue, as are the
effects of (substrate-induced) strain. In all cases, structural and electronic properties
depend on deposition method and growth conditions, which have to be well understood
and controlled. Advanced devices and fine-tuning of the electronic properties of these
materials require further research in these areas.

Aluminum oxide is currently under study for applications such as MOS and MIM
devices, due to its higher dielectric constant compared to the silicon oxide, competing
against many other materials and especially other metallic oxides like hafnium oxides,
tantalum oxides and many others (see Table 1.1). The second variable in consideration
is their electrical resistance, their purpose being to obtain a good capacitance density
while reducing leakage.

Some typical methods for the microelectronics fabrication of metallic oxides are:

• Evaporation. This involves two basic processes: a hot source material evaporates
and condenses on the substrate. Evaporated atoms that collide with foreign
particles may react with them; for instance, if aluminum is deposited in the
presence of oxygen, it will form aluminum oxide. Those particles also reduce the
amount of vapor that reaches the substrate, which makes the thickness difficult
to control.
• PECVD. Process that utilizes a plasma to enhance chemical reaction rates of the
precursors. PECVD processing allows deposition at lower temperatures, which
is often critical in the manufacture of semiconductors. Metallic oxides can be
deposited or a metal surface oxidized by reacting with an oxygen plasma. As it
produces a not so even coverage and oxidation is slow, it is better for thin
thicknesses.
• Anodization. Electrochemical metal oxidation can be done at room temperature.
Process with fast evolving resulting in a low cost system. In the other hand, it
could present more thickness variation.

2
 CHAPTER I - INTRODUCTION

• ALD: used to fabricate ultra thin and conformal thin film structures for many
semiconductor and thin film device applications. A unique attribute of ALD is
that it uses sequential self-limiting surface reactions to achieve control of film
growth in the monolayer or sub-monolayer thickness regime. It is a very precise
method, with more even or conformal coverage that works even in the back of
suspended structures. But, it is a high cost deposition method.

Table 1.1.- Dielectric constant for most Popular Metallic

Oxides in Study
Oxide Permittivity
Al2O3 10
HfO2 20
ZrO2 23
Y2 O3 15
Ta2O5 12 – 23
TiO2 30 - 150

Anodizing process has many possibilities to yield different physical and electrical
characteristics, based on the chemicals used to obtain it, the oxidation speed,
temperature, etc. Dense or porous layers can be achieved. Dense aluminum oxide is
preferred to reduce leakage and increase capacitance density; porous to increase the
total surface area, or even rough surfaces to keep the protection of underlying materials
while still leaving more useful area for sensing. Chemicals can also influence the
presence of a more or less hydrophilic surface, corresponding to hydroxyl groups (OH)
used in some biological applications. Long anodizing time can also affect the properties
of the layer, in some cases it can change from an amorphous oxide to a crystalline one,
the later providing higher dielectric constant but also higher leakage.

3
 CHAPTER I - INTRODUCTION

1.2.- Anodic Aluminum Oxide Current and in

Development Applications

A few of the more important applications of anodizing are described below, as well as
new nanotechnology applications that may grow in importance.

a) Clear anodize.
Clear anodize usually means sulfuric acid anodizing followed by hot water seal where
the electrolytic process is done above room temperatures (military specification titled
Anodic Coatings for Aluminum and Aluminum Alloys MIL-A-8625, Type II). This is
the most widely used anodized coating. It is used on some aluminum alloys as the
surface finish for automotive trim. It is also the surface for commercial
photolithography plates. The photoemulsion adheres to the anodize coating, and the
printing pattern is made by selective dissolution of emulsion. Printing ink adheres to the
emulsion and water adheres to exposed oxide. The oxide surface is wear resistant and
stands the rigors of high-speed printing presses [5, 6, 7].

Immersing the anodized coating in a dye solution before sealing creates an attractive
colored surface for consumer products.

b) Hard anodize.
Hard coat or hard anodizing is similar to standard anodizing, in as much as it also
creates a film of aluminum oxide on the part. It is produced at colder temperatures and
provides a surface that is harder, smoother, and more wear and corrosion resistant than
standard anodizing (MIL-A-8625, Type III). Hard coat anodized films can also be
grown to greater thickness, typically .002 inches or more. This produces a coating with
larger cells and smaller diameter pores (see structure in section 1.3.2) compared to clear
anodizing (temperature, voltage and other process parameters are selected based on the
electrolyte being used in order to meet the “hard” limit values provided in the military
standard). The coating is extremely hard and durable and is used for engineering
applications such as bearing surfaces [5, 6]. Hard coat finishes are typically darker in
appearance than standard anodizing depending on the coating, thickness and alloy.

A thin phosphoric acid anodize coating is used as an adhesive bonding primer coat on
aircraft and aerospace alloy sheets. This is an excellent surface for the epoxy adhesive
and also improves corrosion resistance. Chromic acid anodize is used for optimum
corrosion resistance for severe applications, such as aerospace and military use, but
because chrome, in some form, is a carcinogen, this process is being phased out.

c) Architectural applications
Architectural applications for anodized aluminum include door and window trim and
exterior structural panels. These surfaces must be stable for many years under harsh
atmospheric conditions. Neither clear nor dyed coatings are satisfactory. Coatings
ranging in color from gold to dark bronze are made by "integral coloring" that is
achieved by using certain organic acids in the anodizing bath. The organic anions
become incorporated in the oxide and cause it to darken. Similar hues are produced by a
two-step anodizing process in which an AC anodizing follows the DC sulfuric acid step.
In the AC step, a metal, usually tin or nickel, is deposited at the bottom of the pores.

4
 CHAPTER I - INTRODUCTION

The metal deposit changes the optical properties of the coating and the thickness of the
metal deposit is controlled to produce the desired color by optical interference. These
coatings have superior long term stability compared with integral colored coatings [6, 7,
8].

d) Nanotechnology applications
The importance of surface texture in the development of pores has recently been
dramatically demonstrated, and the results open new applications for anodized
structures. A hexagonal array of nanoscale (on the order of billionth of a meter)
depressions was impressed on an aluminum surface using a silicon carbide die
fabricated using electron beam lithography. Feature interval was 70-500 nm, feature
depth was 200 nm, and feature width was of similar magnitude. Anodization in oxalic
acid solution produced perfectly ordered arrays of pores corresponding to a patterned
texture, where the oxide grows on the patterned surface [5]. Adjustment of process
conditions produces precisely ordered pore arrays with dimensions suitable for use as 2-
D (two-dimensional) photonic crystals in the visible wavelength.

Pores can be used as templates to make structures such as nanowires and nanotubes.
The preparation of nanowires of different materials has recently attracted wide attention
in view of possible future applications in areas such as photocatalysis, electrochemistry,
battery research and enzyme immobilization [9]. There are different concepts of using
the pores in nanoporous templates for preparing nanomaterials [10, 11]. To make
nanowires the pores are filled with a metal or other material, by cathodic deposition or
electroless deposition. Tubes are made by coating the pore walls; addition of functional
groups to the tube interior wall creates tubular nanoreactors [12]. Two step aluminum
porous anodization process is used to fabricate the nanowires under low frequency.
Nanowires are formed by depositing the desired material over the aluminum pores and
then eliminating the porous alumina using orto-phosphoric acid solutions leaving only
the formed nanowires that will have the pore size of the alumina structure [13]. This
technology is still in development.

e) Electrolytic capacitors
The main commercial application of the anodic barrier non-porous oxide is the
dielectric film in electrolytic capacitors. These capacitors have high energy density,
wide voltage range (3 volt to 600 volt) and are relatively low cost. These capacitors are
of two types: wet and solid. Wet type uses an organic electrolyte as electrical contact
between oxide dielectric and a current collector. Most aluminum electrolytic capacitors
are of this type and contain a roll of two aluminum foils – one foil with the oxide
dielectric – separated by a paper spacer impregnated with the organic electrolyte. To
enhance the capacitance per unit volume, the surface area of metal substrate is increased
before deposition of the anodic oxide dielectric. Aluminum foil is etched and tantalum
or niobium powder is sintered to make a high surface area pellet. The high specific area,
thin dielectric layer, and relatively high dielectric constant (about 8-10 for aluminum
oxide and 25 for tantalum oxide) combine to make capacitors with higher energy
density than with any other type of dielectric, such as polymer film or ceramic [6].

f) Humidity Sensors
Based on its hydrophilic property of the aluminum oxide, many works have been done
in developing humidity sensors. Anodization is carried out in an acidic solution, like
sulfuric and phosphoric acids. As previously explained, it is possible to form very good
porous structures as shown in figure 1.4. Porous and walls dimensions can be controlled

5
 CHAPTER I - INTRODUCTION

by the voltage and the acid used for their preparation. At the end of the process the
structures are immersed in boiled distilled water for a few minutes and then gold
electrodes are put in contact to complete the circuit [14, 15, 16]. Significant relationship
capacitance-humidity values are obtained, the response to humidity in the ambient
shows dependence on the pore size (i.e. voltage used for constructing them). Long
exposure time and high humidity can lead to a degradation of the sensor. Periodic
heating is recommended to maintain their reliability [16]. This technology is already in
use.

6
 CHAPTER I - INTRODUCTION

1.3.- Anodization
An oxide film can be grown on certain metals – aluminum, niobium, tantalum, titanium,
tungsten, zirconium – by an electrochemical process called anodizing. For each of these
metals, process conditions promote growth of a thin, dense, barrier oxide of uniform
thickness. The thickness of this layer and its properties vary greatly depending on the
metal, with only the aluminum and tantalum (and recently niobium) films being of
present substantial commercial and technological importance as capacitor dielectrics.
Aluminum is unique among these metals in that, in addition to the thin barrier oxide,
anodizing aluminum alloys in certain acidic electrolytes produces a thick oxide coating,
containing a high density of microscopic pores. This coating has diverse and important
applications including architectural finishes, prevention of corrosion of automobile and
aerospace structures, and electrical insulation. In fact, it is this porous coating that is
most often thought of as the product of anodizing. Both barrier (non-porous) and porous
oxides can be grown on aluminum.

Bath electrolytes are selected in which the oxide is insoluble, or dissolves at a slower
rate than it deposits, so that an adherent oxide layer grows. The bath composition is the
primary determinant of whether the film will be barrier or porous. Barrier oxide grows
in near neutral solutions in which aluminum oxide is hardly soluble, most commonly
ammonium borate, phosphate, or tartrate compositions. Porous oxide grows in acid
electrolytes in which oxide can not only be deposited but also dissolves. The most
widely used bath is diluted sulfuric acid, typically about 1 M. Other baths used for
particular applications are made with oxalic acid or phosphoric acid [6, 7, 8].

Metals that can be anodized also react readily (oxidize) with oxygen in air, so that under
ambient conditions the surface is always covered with a thin oxide film. The details of
film structure and composition depend on the history of exposure to the ambient
atmosphere but aluminum in contact to the air always forms a natural protective layer
barrier oxide layer next to the metal that is 2-3 nm thick. This keeps the metal isolated
from the atmospheric agents, giving it stabilization of the surface so that no further
reactions with its environment occur. It is an excellent electrical insulator too, including
also protection against corrosion. But this is sometimes not good enough for the
electronics applications, so another well controlled process must be used in order to
form an accurately known thickness. This can be achieved by using the anodic oxidation
or anodization.

At the anode, where the anodic oxidation will take place, this insulator layer formed
during the process will start to give resistance to the continuity of the oxidation. The
formation of this layer will depend on the action of the diffusion through this new layer.

In our case we will study in depth the non-porous anodization process.

1.3.1.- Barrier or non-porous oxides

When an aluminum piece is covered with its native barrier oxide, no significant current
flows until the voltage is raised to between 1 and 2 volt. This barrier oxide supports an
electric field (Volts/thickness) on the order of 1 V/ nm, a very high field indeed. If this

7
 CHAPTER I - INTRODUCTION

piece is put at the anode in an electrolytic cell, e.g. borate electrolyte, the voltage across
the oxide can be increased, without initiating current flow, until the field in the oxide is
large enough to drive aluminum and oxygen ions through the oxide. The current through
the oxide is an ionic current and these ions react to form the oxide layer. The process of
high-field ionic conduction is central to anodization. Oxide anions move inward to react
with aluminum at the metal/oxide interface to form oxide. At the cathode, the circuit is
completed by the reduction of hydrogen ions to hydrogen gas. The process is illustrated
in Figure 1.1.

The rate at which the oxide thickens is proportional to the current density (A/cm2). The
field in the oxide does not change with oxide thickness, and has only a small
dependence on current density and temperature. While the oxide thickens, the voltage
across the oxide increases proportionally. At room temperature the thickness/voltage
ratio is close to 1.2 nm/V.

Figure 1.1.- Sketch illustrating ion transport through the oxide film [5].

The anodization process is an electrochemical method, which is done by immersing the

two electrodes in a container that is filled up with an acid or base solution, i.e. the
electrolyte. The sample that will be oxidized is placed in the positive electrode, i.e. the
anode. At the negative one, i.e. the cathode, the reference electrode will complete the
circuit and distribute the current. Main reactions occurring on each side are:

Cathode (-) 6H+ + 6e-
3H2↑ (1.1)

Anode (+) 2Al + 3O-2
Al2O3 + 6e- (1.2)

The reference wafer must be of a conductive material. The preference usually goes to
something neutral/non-reactive for the electrolyte and the anodization process.
Homogeneity of the layer through the surface of the wafer can be achieved by using a
reference cathode which covers the area being processed.

Both, the area to be anodized and the metallic one of the reference wafer must be face to
face. The solution level must be sufficient to cover the area to be anodized.

8
 CHAPTER I - INTRODUCTION

Table 1.2.- Non-Porous Aluminum Oxide Typical Properties Values

[www.goodfellow.com, http://www.matweb.com]

Mechanical Units of Measure Value

Density gm/cc 3.89
Porosity % 0
Color — ivory
Flexural Strength MPa 379
Elastic Modulus GPa 375
Shear Modulus GPa 152
Bulk Modulus GPa 228
Poisson’s Ratio — 0.22
Compressive Strength MPa 2600
2
Hardness Kg/mm 1440
Fracture Toughness KIC MPa•m1/2 4
Maximum Use Temperature
°C 1750
(no load)

Thermal Units of Measure Value

Thermal Conductivity W/m°K 35
–6
Coefficient of Thermal Expansion 10 /°C 8.4
Specific Heat J/Kg•°K 880

Electrical Units of Measure Value

Dielectric Strength ac-kv/mm 16.9
Dielectric Constant @ 1 MHz 8 - 10
Dissipation Factor @ 1 kHz 0.0002
Loss Tangent @ 1 kHz —
Volume Resistivity ohm•cm >1014

In anodization, the thickness can be controlled. It is affected by the treatment conditions

and thin or thick layers can be formed. The non-porous aluminum oxide layer has very
interesting technological properties, like:

• Good transparency in the visual spectra, so that its thickness can be measured by
non destructive methods like ellipsometry.
• Chemical inertia from a pH from 5 to 10, giving it an excellent protection
against many corrosive agents.
• Average thermal conductivity and high melting point, which allows its use for
high temperature processes (max 300°C).
• Hardness around 7 and 8 in Mohs numbers, elastic modulus of 5x104 daN/mm2.
• Good electrical insulation properties with a breakdown voltage between 20 and
40 V/µm, which rest in good level even at high temperatures, much better than
other insulators.
• Dielectric constant around 8 to 10.

9
 CHAPTER I - INTRODUCTION

These properties demonstrate why this oxide has a good potential for applications in
MIM, MIS, MEMS and other devices.

Non-porous, barrier oxide formed on clean aluminum at room temperature has an

amorphous (non-crystalline) structure; Figure 1.2 shows a cross section of an
amorphous 220 nm oxide grown at 10 mA/cm2 in 0.16 M ammonium tartrate at 20°C
[17]. As expected for an amorphous structure, there are no distinguishing structural
features. The micrograph in Figure 1.2 was obtained using a transmission electron
microscope.

Figure 1.2.- Cross section of an amorphous aluminum oxide structure formed by

anodization [17].

Modifying the initial native oxide in certain ways, for example, by heating in air at high
temperature, and then anodizing at elevated temperature produces a barrier film that is
crystalline [18], see Fig. 1.3. The amorphous oxide that grows during a brief exposure to
air at high temperature is only slightly thicker than the usual ambient temperature film,
but it probably contains "seeds" of the crystalline oxide that has been identified in films
grown over long times at temperatures above 550°C. During subsequent anodization,
i.e. at 70°C, these seeds promote growth of a uniform crystalline phase. To move
aluminum and oxygen ions through the crystalline oxide requires a higher field than in
amorphous oxide, so a thinner crystalline oxide supports the same voltage as a thicker
amorphous oxide. Crystalline oxide is advantageous as a capacitor dielectric, because a
thinner dielectric results in a higher capacitance [18].

Figure 1.3.- Top view (left) and cross section (right) of crystalline barrier oxide
[18].

10
 CHAPTER I - INTRODUCTION

1.3.2.- Porous oxides

Porous aluminum oxides are most commonly grown in dilute sulfuric acid, typically 10
weight percent concentration, but there are also commercial processes using phosphoric
acid, chromic acid, oxalic acid, and mixtures of inorganic and organic acids. A feature
common to these anodizing baths is the ability to retain a relatively high concentration
of aluminum in solution. This is essential, because a large fraction of the aluminum that
is oxidized is not retained in the film, but passes into solution. For example, for
anodization in sulfuric acid, about 60% of the oxidized aluminum is in the film and the
remainder is found in solution. Porous films 100 µm thick can easily be made – this is
100 times thicker than the thickest barrier film. Unlike barrier films, a high voltage is
not needed to make a thick porous film because of the unique structure of these films.

In this case, with aluminum dissolution into the electrolyte during porous film
formation, the anodic reaction is:

2Al
2Al+3 + 6e- (1.3)

The reaction at the cathode is hydrogen gas evolution:

6H+ + 6e-
3H2↑ (1.4)

An idealized sketch of film structure is shown in Figure 1.4. The oxide has a cellular
structure with a central pore in each cell. It is expected to yield uniform hexagonal cells,
but most anodization conditions produce films with more disorder, with a distribution of
cell sizes and pore diameters [5]. Cell and pore dimensions depend on bath composition,
temperature, and voltage, but the end result is always an extremely high density of fine
pores. The cell diameter is in the range 50-300 nm and the pore diameter is typically 1/3
to 1/2 of the cell diameter. For example, film thickness of 20 to 50 µm and 40 to 60 nm
diameter cells with 20 nm pores are typical for coatings grown in sulfuric acid.

A thin barrier oxide is at the base of each pore, in the vicinity of the metal/oxide
interface. Each pore in a structure is centered over a scallop-shaped depression in the
metal. At the base of each pore is anodic oxide and pores are separated by oxide
comprising the cell walls. Cell geometry concentrates current through the oxide at the
pore base. The shortest path between metal and electrolyte is within this zone, where the
field is uniform and at its highest value. This is a critical feature for development of the
porous structure. Ions move by high field conduction: aluminum cations reach the pore
surface and pass into solution, and oxide deposition is confined to the metal/oxide
interface at the pore base. As aluminum metal is oxidized, the metal/oxide interface
moves into the metal. What had been barrier oxide at the periphery of the pore base is
no longer penetrated by the field and becomes part of the cell wall. The cell wall and
pore increase in height, that is, the film thickens, while cell and pore diameters remain
fixed. Because the barrier oxide thickness remains constant, the cell voltage and current
remain nearly constant as the film thickens.

11
 CHAPTER I - INTRODUCTION

Figure 1.4.- Idealized structure of anodic porous aluminum oxide [5].

It is not apparent from this description of steady state film growth how the pores and
cells initiate. The metal surface has some degree of roughness. During early growth, the
film on ridges and protuberances becomes thicker than in the depressions. It appears
that ions move through oxide more easily at these locations. This may be because of
higher film stress, impurities, or oxide flaws, and so the current concentrates at these
locations. This is a temporary situation and as the oxide acquires more uniform
properties, the current shifts towards the thinner oxide in depressions. Because of the
concave geometry that has developed, there is a slightly higher electric field in
depressions, and field-assisted dissolution promotes local oxide thinning and current
concentration. This initiates pores, and the pore size, density, and distribution adjust
until steady state prevails. For commercial processes most of the adjustment typically
occurs within the first minute [17].

The usefulness of these films would be quite limited if it were not possible to close the
pores after growth is completed. This step is called "sealing," and is most commonly
done by making the anodized coating react with hot water:

Al2O3 + 3H2O
2AlOOH*H2O (1.5)

The reaction product fills the pores and makes an impermeable anodized layer that is
stable under a wide range of atmospheric and environmental conditions.

12
 CHAPTER I - INTRODUCTION

1.4.- MIM Capacitors

Dielectric materials with high permittivity are being investigated for integrating
packaging, as well as for gate insulators in MOS devices [19] and for MIM capacitors
[20]. Second application main requirements are taken into consideration for the
electrical characterization in this thesis.

Under the powerful market pressure for smaller, smarter (and cheaper) mobile handsets
for wireless communications, notably, a strong demand is put onto microelectronic
technologies to come up with solutions allowing for the integration of other elements
than just transistors. A key factor is the integration of electrical passive elements such as
coils, resistors and capacitors, as this approach provides the capability for removing
large numbers of discrete components by integrating them into the interconnection
structure. This technique is known as "embedded passives”.

The most popular integration substrate solutions today include:

• Ceramic
• Low Temperature Co-fired Ceramic – LTCC
• Glass / Quartz
• Silicon
• Flex substrate
• Laminate FR4 substrates

The most common among those are ceramic, LTCC and Silicon.

Technologies using a step-by-step deposition of alternating dielectric and metal thin

film layers on top of a substrate are collectively referred to as Thin Film On Substrate
(TFOS). Substrates used for TFOS technologies are: ceramic, glass and Si. Some
academic efforts have recently focused on depositing thin films on flex and laminate
substrates.

Embedded passive approaches are best suited in “passive-hungry” applications, such as:

• High-frequency part of modulator-demodulator circuits for telecommunications

(RF front-end)
• Filtering and biasing networks in signal bus conditioning

For 2 GHz+ applications, key factors in order to obtain High-Q components are:

• Tight geometric and process tolerance,

• Lot uniformity & lot-to-lot reproducibility.

Therefore, at system-level cost parity, the technology with tighter tolerance will win for
HF applications. These are TFOS technologies applied to glass or semiconductor
wafers, where tolerance is a function of (very precise) photolithographic steps. This
subset of TFOS technologies is also referred to as Thin Film on Wafer (TFOW).

13
 CHAPTER I - INTRODUCTION

Table 1.3 - Embedded capacitors - technology comparison

Freq
C range Area
Capacitors Tolerance Range
(pF) (mm2)
(GHz)
Laminate 2 – 350 20 % Up to 5 0.5 – 50
Cu 2 – 200 20 % Up to 50 0.5 – 50
On Substrate Ceramic
Ag 2 – 200 20 % Up to 80 0.5 – 50
TFOW 0.1 – 3000 5% Up to 50+ 0.003 – 4
Silicon 0.1 – 30 5% Up to 10 0.001 – 0.1
On Chip
GaAs 0.1 – 15 5% Up to 100 0.001 – 0.1
Discret 0.5 – 1 E6 5 – 30 % Up to 2.4 0.2 - 25

Table 1.4 – Integrated Passive Material Trends

Material Who’s Using It
CMD, IRC, Bourns, Vishay, KOA,
Tantalum Nitride
Philips, Intarsia
Hafnium Diboride CMD, KOA
CMD, IRC, Bourns, Vishay, KOA,
Silicon Oxide
Philips
Silicon Nitride CMD
Aluminum Oxide Intarsia, CTS
Liquid Polyamide CMD
Lead Magnesium Niobate (PMN) IBM Deutschland, Nat Semi
Parylene Film Dow, Foster, Miller
Polyamide/BT CMD, Du Pont
Tantalum Pentoxide Arkansas Micro
Lead-Zirconium-Titanium Oxide
CMD
(PZT)
Polymer Encapsulated Ferrite CMD
99% Alumina CMD
High Purity Glass CMD
CA Nitride CMD
Specialty Polymers Dow

Table 1.3 illustrates the typical performance of integrated capacitive elements for some
of the most popular technologies. In Table 1.4, most important passive materials trends
in the industry are cited.

We can see that Al2O3-based capacitors are already in use. Intarsia used very large glass
panels as substrate material, whereas CTS Corp. uses LTCC or ceramic substrates.

The need to reduce MIM capacitors size without capacitance reduction suggests the use
of high-k materials. Previous investigations have concerned silicon oxide, silicon
nitride, tantalum oxide and hafnium oxide, the last two showing higher capacitance but
also higher leakage. For hafnium oxides, a stacked layer of aluminum oxide has been
proposed to improve the results [21].

14
 CHAPTER I - INTRODUCTION

From all this information the critical performance for this application is:

1. Capacitance per unit area. Today’s most common range is 0.5-1 nF/mm2.
2. Frequency of operation up to the GHz range.
3. Industry standard temperature stability range: -55°C to 125°C.
4. Breakdown voltage (= dielectric strength) above 20 V/µm for ceramic-based
technologies, and above 1 MV/cm for “Thin Film On Substrates” technologies
(TFOS) traditional applications.

In this work the electrical characteristics of aluminum oxide MIM capacitors obtained
by Al anodization will be studied. Previous studies on anodized aluminum with
different anodizing solutions have been reported [22], showing capacitance densities of
0.53 nF/mm2 for thicknesses around 145 nm, but frequency, temperature and breakdown
voltage performance were not fully addressed.

Our work will also introduce a different anodizing electrolyte solution (i.e. boric
acid/ammonium hydroxide/ethylene glycol) and compare the full electrical results to the
literature, as well as to proper experiments on ALD HfO2 and Al2O3 structures.
Furthermore, our aim is to extend the temperature range above 150°C to address the
needs of harsh environments.

15
 CHAPTER I - INTRODUCTION

1.5.- DNA Hybridization Detection

Another interesting application for our aluminum oxide is based on its hydrophilic
behavior. Most biological methods include as first step the functionalization, i.e. a
process which starts with a silane reaction with the surface functional groups of a
substrate. The surface conditions are critical here, mainly its roughness, the available
groups for the reaction, usually superficial hydroxyl groups (OH), and its resistance to
chemicals and by-products from the different biological process steps.

A particular case is DNA (deoxyribonucleic acid) re-association by base pairing or

hybridization which forms the underlining principle of DNA microarrays. These bring
new opportunities in human health monitoring as micro miniaturization of analytical
methods expands [23]. Microelectronics were used mainly in the biochips constructions
up to now for addressing capture probes on specific plots. The use of microelectronics
for nucleic acid hybridization detection remains one of the major challenges at the
moment.

Up to now, microarray detection is mostly performed optically using fluorescent labeled

DNA probes, e.g. by an external scanner or CCD camera, the drawback being the low
sensitivity and the high cost of such equipments. Other methods for the detection of
specific DNA sequences under development based on silicon (Si) chips include
impedance measurement [24], DNA charge detection, quartz crystal microbalance,
evanescent wave [25], fluorescence photodiodes [26], photocurrent spectroscopy, etc.

The use of standard integrated circuit materials, dominated by CMOS (complementary

metal-oxide semiconductor) processes, is being widely studied and could be useful
given its low cost.

DNA chips were first developed using glass (i.e. glass slides or capillaries), plastic or
ceramic substrates. Glass and plastic substrates are still the leading technology in DNA
chips due to their low price, the availability of flexible high-throughput manufacturing
methods and biocompatibility.

With the widespread of microscope slides a variety of methods to immobilize DNA

probes have been used: (a) cross-linking of DNA onto polylysine coated glass slides
[27], (b) hydrophobic interaction between DNA and the substrate [28], (c) covalent
attachment of DNA possessing a terminal thiol group to functional alkanethiol based
monolayers [29], (d) binding of amino-modified oligonucleotides to aldehyde-activated
glass [30] or to glass surface modified with epoxide [31], and (e) tethering of thiol-
modified or disulfide-modified DNA onto aminosilanes using heterobifunctional cross-
linkers [32]. But to date, little detailed information is available about the optimization of
DNA immobilization on silicon (Si) wafer materials. The well-known polypyrol-
electropolymerisation method requires conductive surface (gold, platinum, etc.), which
is not typically available in low-cost standard microelectronic process [33]. Recent
studies show the covalent immobilization of thiol-terminated oligonucleotides on self-
assembled monolayers on silicon oxide (SiO2) [34, 35]. But the authors indicated the
need for a thorough pretreatment of the oxidized silicon wafers [34]. Due to the strong
oxidizing property of the solution used for hydration, only noble metals like gold (Au)

16
 CHAPTER I - INTRODUCTION

and platinum (Pt), but not aluminum (Al), could be used to realize the electronic
devices.

The use of a protective layer is imperative to overcome this problem and restrain the use
of CMOS compatible materials, in our case aluminum electrodes; this will change the
electrical method for detecting the presence of the hybridized DNA from resistive to
capacitive.

Several electrical detection methods have indeed been proposed and can be regrouped
into two main categories: AC current or capacitance measurements, and DC current or
resistance measurements.

Typical examples of the first category include Lillis et al [36, 37] works on electrolyte
insulator semiconductor structures, which measure the capacitance change between a
silicon substrate and a reference electrode in contact with the electrolyte. A maximum
variation of about 3% is reported after the hybridization of 0.1 nano-moles or 1 µg of
DNA fragments [36, 37]. Guiducci et al [38] also demonstrated the possibility of
detecting nano-molar DNA concentration through capacitance measurement at the
electrode/solution interface of a two gold-electrode cell.

In the second category, Hoffman et al [39] presented a sensor based on immobilization

of DNA probe molecules on an array of interdigitated gold lines, and measurement of a
DC current signal resulting from an electrochemical redox cycling process. The typical
limit of detection by redox processes corresponds to the micro-molar range only, with
some exceptions as will be seen below.

The best results reported to date have been obtained combining resistance detection with
the silver enhancement technique. Silver precipitation was initially proposed as an
alternative to the fluorescence assay for DNA hybridization detection on standard glass
slides, since the silver produces an opaque layer on the surface that can be easily
detected in colorimetry, by the absorption of light or by its scattering [40]. In Xue et al
[41] and Park et al [42] approaches, an oligonucleotide-modified gold nanoparticle
probe is used to label DNA samples that match with probes fixed on a silicon dioxide
layer, in a gap area between two gold electrodes. Afterward, silver enhancement is
performed to fulfill the gaps between the gold particles and the electrodes and the
decrease of resistance between the electrodes is monitored. In Xue’s case, the gap area
is 16x50 µm2 and their sensitivity is around 10 pM after 10-15 minutes silver
enhancement [41]. Park et al use a gap of 20 µm between the electrodes and claim they
can detect DNA concentrations as low as 0.5 pM after 35 minutes of silver precipitation
enhancement [42].

Based on their labeling, detection methods can also be classified, i.e. non-labeled vs.
labeled.

17
 CHAPTER I - INTRODUCTION

Table 1.5.- Recent works comparison data from different approaches in DNA
hybridization electrical detection on silicon technologies.

No of
Ref. Transducer Detection limit Reported results
Bases

1 x 5 mm IDE | Z | impedance
[51] 10 µm width or spacing, 20 82.2 µM modulus
Pt + 50 % @ 10 Hz
1.5 µm x 60 nm gap Not 30% capacitance
[43] 0.5 µM
4 mm Poly-Si over Si cited change
200 kHz frequency
150 x 150 µm² FBAR
[44] 24 50 µM shift (Ref. freq. 2
Pt + ZnO
GHz)
60 mV Vth absolute
20 x 500 µm² EIS 1 g/L
[52] 1000 shift
10 nm SiO2 (~ 800 µM)*
(60 min hybridizing)
ISFET length: 32 µm
25 % current shift
ISFET width: 140 µm 1 pg/L
[45] 15 (full match
63 nm SiO2 (~ 0.1 pM)*
vs point mutation)
+ 30 nm Si3N4
0.2 mV Vth diff. shift
50 x 50 µm² EIS
[53] 12 2 nM (after 10 min
2 nm SiO2 + PLL
hybridizing)
| Z | impedance
8 x 4 mm² EIS modulus
[54] 16 3 µM
Direct Si-C bonds + 20% @ 100 kHz
(20 min hybridizing)
5 mV Vth absolute
ISFET length: 1, 1.5 µm ± 4.1011/ cm²
shift
[55] ISFET width: 8, 16 µm 19 (~ 630
(~ 60 min
SiO2: 8, 10, 12 nm fmole/cm²)*
hybridizing)
ℑ(Z) imag. imp. + 200
1 mg/L (~ 150
1 cm² EIS Ω
[46] 20 nM)*
15 nm SiO2 @ 100 kHz (~ 60
1011 to 1012 /cm²
min)
∅ 100 µm patch ± 1 nA DC flow
[49] 15 0.1 pM / enzyme
Pt coating (1 min readout)
20 µm gap 200 MΩ → 100 kΩ
[42] 27 0.5 pM / Au + Ag
5 nm Ti + 60 nm Au (35 min enhancement)
* based on approximate conversion data

First, direct detection of non-labeled DNA can be based, e.g. on passive electrodes such
as micron-scaled interdigitated fingers or nanogaps [43], electromechanical resonators
such as FBAR [44], and field-effect devices such as ISFET and “Electrolyte Insulator
Semiconductor”, EIS structures [45, 46]. The nanogap design seems promising thanks

18
 CHAPTER I - INTRODUCTION

to its full CMOS compatibility, whereas FBAR resonators still need optimization. An
abundant literature exists about field-effect devices, which feature high sensitivities
(around one DNA pico mole per square centimeter) [47] but are usually quite affected
by environment noise.

(b)
(a)

(c)
(d)
Figure 1.5.- Example of different devices from literature in the DNA hybridization
electrical detection, a) Direct impedance sensing [43], b) Electrochemistry using
CombiMatrix Pt-coated chips [49], c) Field-effect transducing [53], and d)
Conductimetry [42]

On the other hand lays the detection of labeled DNA. Electrochemical sensing of redox
labels offers the lowest detection limit and appears applicable inside a CMOS post-
process, but has not been further exploited yet. A reason could be the difficulty to
achieve very high detection sensitivities, with a few exceptions achieving the nM levels
[48], or the pM range [49]. Excellent results have also been demonstrated using silver
precipitation on hybridized DNA labeled by colloidal gold nano particles in order to
form conductive bridges between pairs of electrodes, as shown with the DC
measurements examples explained before [39 - 42]. Their main drawback, apart the
electrodes cost since made of noble metals, concerns their measurement range up to the
giga-Ohm before treatment, which might impede correct calibration or might easily
trigger false detection.

19
 CHAPTER I - INTRODUCTION

So, up to now, capacitance methods were mainly limited to low electrical signal
variations and high DNA concentrations. Secondly, all reports use gold or noble metal
electrodes to avoid surface oxidation and degradation in biological solutions and enable
good ohmic contact with the electrolyte or with the silver precipitates. Such metals are
usually not compatible with standard low-cost silicon circuits fabrication processes and
resistance measurements after silver precipitation still present further hazards and
limitations like electrode-silver interface contact impedance suffering from parasitic
effects, such as “double-layer” like phenomenon [50]. Since gold catalyzes silver
precipitation, large electrodes gaps are required to avoid measuring low resistance in the
absence of hybridization or false readings resulting from silver non-specific
precipitation due to very long revelation times for low DNA concentrations. The
conduction path arising between two electrodes can also take any arbitrary shape,
depending on the exact location of the labeled DNA pairs, resulting in a wide statistical
distribution of the resistance measured after some time.

Most techniques arriving to detection in the range below the pico mol target
concentrations use structures made of non CMOS compatible materials or with
complicated fabrication (see Fig. 1.5). DNA strands are usually of short base numbers,
around 15 to 30 bp, to keep within the double layer and avoid contortion influence.
Most popular techniques in development focus on field effect devices, which can make
the detection in solution.

20
 CHAPTER I - INTRODUCTION

1.6.- Summary and Highlights of this Work

1.6.1.- Objectives

The first objective of this work is to obtain the non-porous anodic aluminum oxide main
mechanical and electrical properties, in order to understand and find its possible
applications into the micro and nano electronics world, in devices such as MIM and
MIS capacitors as well as MEMS. Properties like interface characteristics, leakage
performance, capacitance, resistance to temperature and voltage, surface quality
(roughness, defects), etc. are addressed here, as well as the understanding on the
anodizing method influence on these properties.

Secondly, based on the aluminum oxide surface chemical and structural quality, in
depth development is done for its use into the DNA hybridization electrical detection
device manufacturing. Our device objective is to match current optical existing methods
performance in terms of detection limit and reliability, so that to be considered as an
alternative. With regards to fabrication, the device in development must be simple and
compatible with standard CMOS process to assure its possible integration into ICs at the
lowest fabrication cost as possible. As a sensor, it must be competitive vs. the state of
the art in electronical DNA hybridization detection methods under development.

1.6.2.- Work Content

In the second chapter of this work, a deep study of the properties of Aluminum oxide
prepared by anodization is carried out to obtain its main physical, mechanical and
electrical characteristics, taking into account the possibility of applications such as
sensors, capacitors, MOS devices (high-K dielectrics), and MEMS structures.

Differently to literature, we thoroughly investigate the fabrication of the anodic

aluminum oxide at low current density, using boric acid and ammonium hydroxide in
ethylene glycol as electrolyte, to obtain a denser oxide for thicknesses between 50 and
200 nm. Furthermore, extensive electrical characterizations of relevant figures-of-merit
(i.e. capacitance per unit area, permittivity, leakage, capacitance variation with
temperature or TCC, capacitance variation with voltage or VCC, and breakdown
voltage) are proved compatible with TFOS MIM applications and better than those of
aluminum oxides formed by deposition techniques such as ALD (atomic layer
deposition). The performance test is extended at high temperature going up to 300°C,
envisioning new possible applications under harsh environments.

Being CMOS compatible is important for the possibility of integrating the sensor or
device with the system that contains the circuit it will work with, introducing only a few
processing steps, and hence lowering the cost.

With regards to DNA sensors, the first step, detailed on chapter 3, is to evaluate how
different materials and surfaces used in a CMOS process behave when attaching the
DNA strands: i.e. silicon oxides grown with different methods, different combinations
of silicon oxide, silicon nitrides and polysilicon layers, metals and metal oxides. Two
reliability criteria have to be evaluated, binding of DNA strands and resistance to the
chemicals involved in the process. Our first original approach is focused on thermally

21
 CHAPTER I - INTRODUCTION

grown SiO2 for DNA binding, on Aluminum for sensing electrodes and on Aluminum
Oxide for passivation.

In chapter 4, our technique aims at solving the problems discussed in section 1.5 for on-
chip DNA hybridization electrical detection by using different aluminum constructed
structures protected by aluminum oxide formed by anodization. At low frequency,
capacitive sensing is done after silver enhancement, between small interdigitated
electrodes. At high frequency, in GHz range, structures like spiral and meander
inductors, two level inductors and interdigitated capacitors, are evaluated for exploring
how the silver grains present after enhancement affect their resonance frequency.

Therefrom the best option/device is selected, i.e. interdigitated capacitors, process

characterization is done (technological and biological), and further analysis carried out
on these structures to explore all the electrical parameters that could be affected after the
biological process: capacitance between fingers at low frequency AC measurement,
how this change in capacitance changes its resonance frequency, at high frequency, and
finally capacitance from the electrodes to the substrate, i.e. MOS CV curve.

Finally our original biosensor is evaluated for the DNA hybridization detection of TP53
(tumor suppressor) mutations test, an important application in medical research and
diagnosis, where positive and negative samples and control groups are included
allowing the discrimination of all species found in different samples.

The final goal of this project is to enable the combination of all studied techniques in a
multi sensor implementation, targeting e.g. the simultaneous detection of DNA by both
electrical measurement with top contact electrodes and optical response of an
underlying photo-diode. The idea paves the way to enhance the response and sensitivity
of the detection by combining multiple signals.

22
 CHAPTER I - INTRODUCTION

REFERENCES

[1] S. Chang, J. Lee, H. Shin, “Gate-Induced Drain Leakage Currents in Metal

Oxide Semiconductor Field Effect Transistors with High-k Dielectric”, Jpn. J.
Appl. Phys. Vol. 41, pp. 4432–4435, 2002.
[2] M.F. Li, H.Y. Yu, Y.T. Hou, J.F. Kang, X.P. Wang, C. Shen, C. Ren, Y.C.
Yeo, C.X. Zhu, D.S.H.Chan, A. Chin, D.L. Kwong, “Selected Topics on HfO2
Gate Dielectrics for future ULSI CMOS Devices”, (invited) ICSICT 2004,
Beijing, 2004.
[3] L. Kang, K. Onishi, Y. Jeon, B.H. Lee, C. Kang,W.-J. Qi, R. Nieh, S. Gopalan,
R. Choi, J.C. Lee, “MOSFET Devices with Polysilicon on Single-Layer HfO2
High-K Dielectrics”, IEDM, 35-38, 2000.
[4] J. Robertson, “Band Structures and band offsets of high K dielectrics on Si”,
Applied Surface Science 190, pp. 2-10, 2002.
[5] H. Asoh, K. Nishio, M. Nakao, T. Tamamura, H. Masuda, “Conditions for
Fabrication of Ideally Ordered Anodic Porous Alumina Using Pretextured Al”,
Journal of The Electrochemical Society, Vol. 148, pp B152-B156, 2001.
[6] G. E. Thompson, G. C. Wood, “Anodic Films on Aluminum”, in Treatise on
Materials Science and Technology, Vol. 23, Corrosion: Aqueous Processes and
Passive Films, pp 205-329, J. C. Scully (editor), Academic Press, New York,
1983.
[7] S. Wernick, R. Pinner, P.G. Sheasby, “The Surface Treatment and Finishing of
Aluminum and Its Alloys (5th edition)”, ASM International, Metals Park,
Ohio, USA, 1987.
[8] H. Uchi, T. Kanno, R. S. Alwitt, “Structural Features of Crystalline Anodic
Alumina Films”, Journal of The Electrochemical Society, Vol. 148, pp B17-
B23, 2001.
[9] P. Forrer, F. Schlottig, H. Siegenthaler, M. Textor, “Electrochemical
preparation and surface properties of gold nanowire arrays formed by the
template technique”, J. of Appl. Electrochem. 30, pp. 533-541, 2000.
[10] C.R. Martin, “Nanomaterials - A Membrane-Based Synthetic Approach”,
Science 266, pp 1961-1966, 1994.
[11] D. Routkevitch, A.A. Tager, J. Haruyama, D. Almawlawi, M. Moskovites,
J.M. Xu: “Nonlithographic Nanowire Arrays: Fabrication, Physics, and Device
Applications” IEEE Trans. Electron Devices 43, 1646-1658, 1996.
[12] T. Kyotani, L. Tsai, A. Tomita, “Preparation of ultrafine carbon tubes in
nanochannels of an anodic aluminum oxide film”, Chemistry of Materials, Vol.
8, p 2109-2113, 1996.
[13] Z.Wang, Y.-K. Su, H.-L. Li, “AFM study of gold nanowire array
electrodeposited within anodic aluminum oxide template”, Appl. Phys. A 74,
pp. 563–565, 2002.
[14] G. Sberveglieri, R. Anchisini, R. Murri, C. Ercoli, N. Pinto, “An Al2O3 sensor
for low humidity content: characterization by impedance spectroscopy”,
Sensors and Actuators B 32, pp 1-5, 1996.
[15] R.K. Nahar, V.K. Khanna, “Ionic doping and inversion of the characteristic of
thin film porous Al2O3 humidity sensor”, Sensors and Actuators B 46, pp 35-
41, 1998.
[16] R.K. Nahar, “Study of the performance degradation of thin film aluminum
oxide sensor at high humidity”, Sensors and Actuators B 63, pp 49-54, 2000.

23
 CHAPTER I - INTRODUCTION

[17] R.C. Furneaux, G.E. Thompson, G.C. Wood, “The application of

ultramicrotomy to the electronoptical study of surface films on aluminium”,
Corrosion Science, Vol. 18, pp. 853-881, 1978.
[18] R. S. Alwitt, H. Takei, “Crystalline Aluminum Oxide Films”, Thin Films
Science and Technology, Vol. 4, Passivity of Metals and Semiconductors, pp
741-746, M. Froment (editor), Elsevier, New York, 1983.
[19] G.D. Wilk, R.M. Wallace, J.M. Anthony, “High-K gate dielectrics: current
status and materials properties considerations”, J. Appl Phys; 89 (10):5243,
2001
[20] H. Hu, C. Zhu, Y.F. Lu, B.J. Cho, W.K. Choi. A high performance MIM
capacitor using HfO2 dielectrics. IEEE Electron Dev Lett 23 (9):514, 2002.
[21] F. Mondon, S. Blonkowski. Electrical characterisation and reliability of HfO2
and Al2O3-HfO2 MIM capacitors. Microelectronics Reliability 43, pp. 1259-
1266, 2003.
[22] H.J. Oh, Y. Jeong, S.J. Suh, Y.J. Kim, C.S. Chi, “Electrochemical
characteristics of alumina dielectric layers”, Journal of Physics and Chemistry
of solids 64, pp 2219-2225, 2003.
[23] E. Souteyrand, J.P. Cloarec, J.R. Martin, M. Cabrera, M. Bras, J.P. Chauvet, V.
Dugas, F. Bessueille, “Use of microtechnology for DNA chips
implementation”, Applied Surface Science, vol. 164, pp. 246-251, 2000.
[24] J.P. Cloarec, N. Deligianis, J. R. Martin, I. Lawrence, E. Souteyrand, C.
Polychronakos, M.F. Lawrence, “Immobilization of homooligonucleotide
probe layers onto Si/SiO2 substrates: characterization by electrochemical
impedance measurements and radiolabelling”, Biosens. Bioelectron., Vol. 17,
pp. 405-412, 2002.
[25] F.F. Bier, F. Kleinjung, F.W. Scheller, “Real-time measurement of nucleic-acid
hybridization using evanescent-wave sensors: steps towards the genosensor”,
Sens. Actuat. B, vol. 38-39, pp. 78-82, 1997.
[26] M. Xue, W. Xu, R. Lenigk, M. Carles, N.J. Sucher, N.Y. Ip, M. Chan,
“Integrated Photodiode DNA Identification System for Genetic Chip
Applications”, Extended Abstract of 2001 International Conference on Solid
State Devices and Materials (SSDM 2001), Tokyo, Japan, pp. 4562-4567,
2001.
[27] M. Schena, D. Shalon, R.W. Davis, P.O. Brown, “Quantitative monitoring of
gene expression patterns with a complementary DNA microarrays”, Science,
vol. 270, pp. 467-470, 1995.
[28] J.F. Allemand, D. Bensimon, L. Jullien, A. Bensimon, V. Croquette, “pH-
dependent specific binding and combing of DNA”, Biophys. J., vol. 73, pp.
2064-2070, 1997.
[29] B.A. Cavic, M.E. McGovern, R. Nisman, M. Thompson, “High surface density
immobilization of oligonucleotide on silicon”, The Analyst, Vol. 126, pp. 485-
490, 2001.
[30] N. Zammatteo, L. Jeanmart, S. Hamels, S. Courtois, P. Louette, L. Hevesi, J.
Remacle, “Comparison Between Different Strategies of Covalent Attachment
of DNA to Glass Surfaces to Build DNA Microarrays”, Anal. Biochem., vol.
280, pp. 143-150, 2000.
[31] J.B. Lamture, K.L. Beatie, B.E. Burke, H.D. Eggero, D.J. Ehrlich, R. Fowler,
“Direct detection of nucleic acid hybridization on the surface of a charge
couple device”, Nuc. Acids Res., vol. 22, pp. 2121-2125, 1994.

24
 CHAPTER I - INTRODUCTION

[32] L.A. Chrisey, G.U. Lee, C.E. O'Ferrall, “Covalent Attachment of Synthetic
DNA to Self-assembled Monolayer Films”, Nuc. Acids Res., vol. 24, pp. 3031-
3039, 1996.
[33] T. Livache, A. Roget, E. Dejean, C. Barthet, G. Bidan, R. Téoule, “Preparation
of a DNA matrix via an electrochemically directed copolymerization of pyrrole
and oligonucleotides bearing a pyrrole group”, Nuc. Acids Res., vol. 22, pp.
2915-2921, 1994.
[34] R. Lenigk, M. Carles, N.Y. Ip, N.J. Sucher, “Surface Characterization of a
Silicon-chip-based DNA Microarray”, Langmuir, vol. 17, pp. 2497-2501,
2001.
[35] J.P. Cloarec, J.R. Martin, C. Polychronakos, I. Lawrence, M.F. Lawrence, E.
Souteyrand, “Functionalization of Si/SiO2 Substrates with Homo-
oligonucleotides for a DNA Biosensor”, Sens. Actuat. B, vol. 58, pp. 394-398,
1999.
[36] B. Lillis, E. Hurley, S. Galvin, A. Mathewson, H. Berney, “A Novel, High
Surface Area, Capacitance Based Silicon Sensor for DNA Hybridisation
Detection”, The 16th European Conference on Solid-State Transducers,
Prague, Czech Republic, September15-18, 2002.
[37] H. Berney, J. West, E. Haefele, J. Alderman, W. Lane, J.K. Collins, “A DNA
diagnostic biosensor: development, characterization and performance” Sensors
and Actuators B 68, pp. 100-108, 2000.
[38] C. Guiducci, C. Stagni, G. Zuccheri, A. Bogliolo, L. Benini, B. Samorì, B.
Riccò. “A Biosensor for Direct Detection of DNA Sequences Based on
Capacitance Measurement”. ESSDERC 2002, pp. 479-482.
[39] F. Hofmann, A. Frey, B. Holzapfl, M. Schienle, C. Paulus, P. Schindler-Bauer,
R. Hintsche, E. Nebling, J. Albers, W. Gumbrecht, R. Thewes, “Passive DNA
Sensor with Gold Electrodes Fabricated in a CMOS Backend Process”.
ESSDERC 2002.
[40] I. Alexandre, S. Hamels, S. Dufour, J. Collet, N. Zammatteo, F. De
Longueville, J.L. Gala, J. Remacle. “Colorimetric silver detection of DNA
microarrays”, Anal. Biochem. 295, pp. 1-8, 2001.
[41] M. Xue, J. Li, Z. Lu, P.K. Ko, M. Chan, “Array-Based Electrical Detector of
Integrated DNA Identification System for Genetic Chip Applications”,
ESSDERC 2002.
[42] S.J. Park, T.A. Taton, C.A. Mirkin, “Array-Based Electrical Detection of DNA
with Nanoparticle Probes”, Science 295, pp. 1503-1506, 2002.
[43] M. Yi, K.-H. Jeong, L.P. Lee, “Theoretical and experimental study towards a
nanogap biosensor”, Biosens. Bioelectron, 20 pp. 1320-1326, 2005.
[44] R. Gabl, H.-D. Feucht, H. Zeininger, G. Eckstein, M. Schreiter, R. Primig, D.
Pitzer, W. Wersing, “First results on label-free detection of DNA and protein
molecules using a novel integrated sensor technology based on gravimetric
detection principles”, Biosens. Bioelectron. 19, pp. 615-620, 2004.
[45] F.K. Perkins, S.J. Fertig, K.A. Brown, D. McCarthy, L.M. Tender, M.C.
Peckerar, “An Active Microelectronic Transducer for Enabling Label-Free
Miniaturized Chemical Sensors”, IEDM 00, IEEE Electr. Dev. Meeting, pp.
407-410, 2000.
[46] A. Macanovic, C. Marquette, C. Polychronakos, M.F. Lawrence, “Impedance-
based detection of DNA sequences using a silicon transducer with PNA as the
probe layer”, Nucleic Acids Res. 32 (2) e20 online publication, 2004.
[47] A. Poghossian, A. Cherstvy, S. Ingebrandt, A. Offenhausser, M.J. Schoning.,
“Possibilities and limitations of label-free detection of DNA hybridization with

25
 CHAPTER I - INTRODUCTION

field-effect-based devices” Sensors and Actuators B 111-112 pp. 470-480,

2005.
[48] E. Nebling, T. Grunwald, J. Albers, P. Schäfer, R. Hintsche, “Electrical
Detection of Viral DNA Using Ultramicroelectrode Arrays”, Anal. Chem. 76,
pp. 689-696, 2004.
[49] K. Dill, D.D. Montgomery, A.L. Ghindilis, K.R. Schwarzkopf, “Immunoassays
and sequence-specific DNA detection on a microchip using enzyme amplified
electrochemical detection”, J. Biochem. Biophys. Methods 59 pp.181-187,
2004.
[50] C. Schyberg, “Caractérisation Electrique de Structures Si/SiO2 Nues et
Greffées de Molécules Organiques et Biologiques en Vue de l’Elaboration d’un
Immunocapteur”, Thèse de l’Institut National des Sciences Appliquées de
Lyon, France, 1996.
[51] M. Gheorghe, A. Guiseppi-Elie, “Electrical frequency dependent
characterization of DNA hybridization”, Biosens. Bioelectron. 19, p. 95-102,
2003.
[52] E. Souteyrand, J.P. Cloarec, J.R. Martin, C. Wilson, I. Lawrence, S. Mikkelsen,
M.F. Lawrence, “Direct Detection of the Hybridization of Synthetic Homo-
Oligomer DNA Sequences by Field Effect”, J. Phys. Chem. B 101, 15, p. 2980-
2985, 1997.
[53] J. Fritz, E.B. Cooper, S. Gaudet, P.K. Sorger, S.R. Manalis, “Electronic
detection of DNA by its intrinsic molecular charge”, Proc. Nat. Acad. Sci.
USA 99, 22 p. 14142-14146, 2002.
[54] W. Cai, J.R. Peck, D.W. Van der Weide, R.J. Hamers, “Direct electrical
detection of hybridisation at DNA-modified silicon surfaces”, Biosens.
Bioelectron, 19, 1013–1019, 2004.
[55] F. Uslu, S. Ingebrandt, D. Mayer, S. Böcker-Meffert, M. Odenthal, A.
Offenhäusser, “Labelfree fully electronic nucleic acid detection system based
on a field-effect transistor device”, Biosens. Bioelectron. 19, p. 1723-1731,
2004.

26