conductive polymers for corrosion protection : a critical ...€¦ · conductive polymers for...

Conductive polymers for corrosion protection:

a critical investigation.

Adam Michalik

Thesis submitted for the degree of

Doktor der Naturwissenschaften (Dr. rer. nat.)

at the

Faculty of Chemistry and Biochemistry

in

Ruhr-Bochum-University

· 2009 ·

iii

To Dorota and Dominik

v

ACKNOWLEDGMENTS

I would like to thank my supervisor Prof. M. Stratmann for introducing me to the

field of Electrochemistry and for the support during my studies. I also would like to

express my gratitude to Dr. M. Rohwerder for his extraordinary engagement in my

research. Particularly I would like to thank for help in the analysis of the experimental

data and for the correction of the following thesis. Thanks also to Dr H. Asteman

for the preparation of the iron oxide layers on platinum substrates and to Sonnur

Isik-Uppenkamp for assistance in SKP measurements.

I would like to thank to Herve Ehahoun for support and encouragement during my

studies.

Adam Michalik

Duesseldorf

April 26, 2009

vii

ABSTRACT

The potential of conducting polymer coatings for corrosion protection is a topic of current

controversy. In general, the efficacy of conducting polymers very much depends on how they

are applied and on the conditions of the corrosion experiment, i.e. depending on the exact

conditions a conducting polymer may have excellent protection capability or may lead to a

disastrously enhanced corrosive attack. A number of possible protection mechanisms based

on different electrochemical effects are proposed. The three most discussed mechanisms of

corrosion protection by conducting polymer coatings are investigated in this thesis. Two of

them are the ennobling mechanism based on the assumption that the high potential of the

polymer is able to keep the metal surface in the passive state and the possible capability to

smear out the sites of oxygen reduction. Using a special model sample set up it could be

shown that electrochemically deposited polypyrrole films can indeed protect small pin hole

like defects from corrosion. In the case of immersion into chloride free solution the polypyrrole

is able to maintain the defect in the passive state for a few hours while, however, in chloride

containing solution the passivitiy breaks down after few seconds. The size of the defect

significantly determines the degree of possible protection. Polypyrrole showed the ability

to recharge itself after disconnection from the defect, so it could provide a real longterm

protection under cyclic climatic conditions. In the presence of a larger defect, however, the

positive effect of the conducting polymer turns into a negative one and fast delamination is

observed.

Experiments carried out in 18O2 containing atmosphere gave answer to the question concern-

ing the oxygen reduction site. Sputter profiles measured using ToF-SIMS indicate a presence

of 18O− and 18OH− ions created during the delamination just on the top of the polypyrrole

layer. That excludes the possibility of the high conductive polymer like pure polypyrrole to

smear out the sites of oxygen reduction.

Ion transport through the polymer matrix which is essential for the “intelligent” release of

corrosion inhibitors was investigated by the reduction and the delamination of ICPs in oxygen

free and oxygen containing atmospheres. The obtained results showed that the incorporation

of cations is the dominating process during the delamination of conductive coating. This is

due to the fact that even the slightest initial cation uptake into the coating turns it finally over

the distances relevant here into a cation permselective polymer. This finding is important, in

fact a real break-through, for the design of corrosion protection coatings based on conducting

polymers that will really work! Macroscopic percolation networks of the conducting polymer

in the coating must be prevented!

viii

Hence, coatings have to prepared with only microscopic percolation networks. The investi-

gations carried out on such coatings revealed a novel protection mechanism. It was observed

that separated particles of conductive polymer which were distributed on the metal-polymer

interface created a so called “protection zone” around their border. It is proposed that the

above effect is caused by the polarization of the metal/(non-conducting matrix)polymer in-

terface by the conductive particle. When the delamination front approaches the conducive

particle and the galvanic coupling between the particle and defect is established then the

reduction of the polymer will slowly start. The reduction of the polymer is accompanied by

the ionic current follows along the polymer/metal interface. Due to the ionic resistance of the

interface this current induces an increase of the potential at the interface, or at least prevents

that it is pulled down by the advancing delamination. This entails a decrease of the rate of

the oxygen reduction at the very delamination front and thus a decrease of the progress of

the delamination front.

ix

ACRONYMS

2D - Two dimensional

3D - Three dimensional

AD - Analog to Digital

AFM - Atomic Force Microscope

APS - Ammonium Peroxydisulfate

ASCII - American Standard Code for Information Interchange

CCC - Chromate Conversion Coatings

CP - Cell potential

CV - Cyclic Voltammogram

DC - Direct Current

DMF - 2,5-Dimethylfuran

DMSO - Dimethyl sulfoxide

EQCM - Electrochemical Quartz Crystal Microbalance

ICP - Intrinsically Conductive Polymers

ITO - Indium Tin Oxide

LED - Light Emitting Diode

KP - Kelvin Probe

OPC - Open cell potential

PC - Personal computer

Py - Pyrrole momomer

PPy - Polypyrrole

PANI - Polyaniline

PVB - Polyvinyl Butyral

PVD - Physical Vapor Deposition

RPM - Rotation Per Minute

SEM - Scanning Electron Microscope

SHE - Saturated Hydrogen Electrode

SKP - Scaning Kelvin Probe

THF - Tetrahydrofuran

ToS - p-Toluol sulfonate

ToF SIMS - Time of Flight Secondary Ion Mass Spectrometry

XPS - X-ray Photoelectron Spectrometry

Contents

1 The state of the art of corrosion protection by conducting polymers. 1

1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.2 Principles of corrosion of iron. . . . . . . . . . . . . . . . . . . . . . . . 2

1.3 Protection against corrosion . . . . . . . . . . . . . . . . . . . . . . . . 3

1.3.1 Conversion coatings . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.3.2 Organic coatings . . . . . . . . . . . . . . . . . . . . . . . . . . 6

1.3.3 Corrosion of coated metals . . . . . . . . . . . . . . . . . . . . . 7

1.4 Conductive polymers for corrosion protection. . . . . . . . . . . . . . . 9

1.4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

1.4.2 Polymerization methods . . . . . . . . . . . . . . . . . . . . . . 14

1.4.3 Inhibition of corrosion by conductive polymer. . . . . . . . . . . 17

2 Scanning Kelvin Probe 21

2.1 The basics of potential measurement by the Kelvin Probe technique . . 21

2.2 SKP measurement over conductive coatings . . . . . . . . . . . . . . . 23

2.3 The hardware realization . . . . . . . . . . . . . . . . . . . . . . . . . . 24

2.4 The control software . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

2.4.1 Structure of the program . . . . . . . . . . . . . . . . . . . . . . 27

2.4.2 Configuration of the program . . . . . . . . . . . . . . . . . . . 27

xi

xii CONTENTS

2.4.3 Panel ”Scan“ . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

2.4.4 Subprogram “Look” . . . . . . . . . . . . . . . . . . . . . . . . 33

2.4.5 Selected subroutines and tasks . . . . . . . . . . . . . . . . . . . 34

3 The preparation and characterization of the polypyrrole films 37

3.1 Electrodeposition of polypyrrole films . . . . . . . . . . . . . . . . . . . 37

3.1.1 Substrate preparation . . . . . . . . . . . . . . . . . . . . . . . 38

3.1.2 Deposition of the polypyrrole layer . . . . . . . . . . . . . . . . 38

3.2 Characterization of the polypyrrole layers . . . . . . . . . . . . . . . . . 41

3.2.1 Electrochemical characterization of the PPy . . . . . . . . . . . 41

3.2.2 XPS analysis of the obtained PPy layers . . . . . . . . . . . . . 45

3.2.3 Morphology of the PPy film . . . . . . . . . . . . . . . . . . . . 50

4 The passivation of the defect by ICP. 53

4.1 Ennobling of the defect by ICP’s . . . . . . . . . . . . . . . . . . . . . 54

4.1.1 Experimental setup . . . . . . . . . . . . . . . . . . . . . . . . . 56

4.1.2 Influence of the type of the electrolyte for the passivation . . . . 58

4.1.3 Recharging of the polymer during wet/dry cycles . . . . . . . . 59

4.2 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

5 The shift of the oxygen reduction site 63

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

5.2 Experimental setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

5.3 Oxygen reduction dislocation . . . . . . . . . . . . . . . . . . . . . . . 66

5.4 Durability of the polypyrrole in the alkaline environment . . . . . . . . 68

5.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

CONTENTS xiii

6 The ion transport through the conductive coating 73

6.1 Oxygen and polymer reduction during delamination of conductive coat-

ings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

6.2 Cation incorporation versus anion release . . . . . . . . . . . . . . . . . 81

6.2.1 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

7 Non continuous ICP films: on the role of the polarization of the

metal/non-conductive polymer interface by ICP patterns. 95

7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

7.1.1 Interface polarization by conductive polymer islands at the in-

terface metal/non-conductive polymer. . . . . . . . . . . . . . . 97

7.2 Delamination test of an isolated PANI dot applied on a chromium sub-

strate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

7.2.1 Investigation of the delamination performance of non-conducting

polymer on chromium. . . . . . . . . . . . . . . . . . . . . . . . 101

7.2.2 Polarization of the chromium-polymer interface. . . . . . . . . . 102

7.3 Interface polarization by a PANI dot on the iron. . . . . . . . . . . . . 105

7.4 Delamination test of the PANI dot on gold. . . . . . . . . . . . . . . . 108

7.5 Patterns of the conductive coating. . . . . . . . . . . . . . . . . . . . . 112

7.5.1 Preparation of the polyaniline pattern. . . . . . . . . . . . . . . 113

7.5.2 Investigation of the delamination properties of the polyaniline

patterns with different spacing between conductive pigments. . . 115

7.5.3 Examination of the delamination properties of the polyaniline

patterns with increased volume of the conductive polymer. . . . 118

7.6 On the diffusion at polymer—metal interfaces: an investigation on polymer—

iron. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126

7.6.1 Diffusion of anions and cations. . . . . . . . . . . . . . . . . . . 131

7.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134

xiv CONTENTS

8 Main conclusions and outlook 137

Chapter 1

The state of the art of corrosion

protection by conducting polymers.

1.1 Introduction

The discovery of metal melting and processing brought humanity from the Stone Age

into other, more advanced levels of the civilization. According to archaeologists these

are called the Bronze Age and the Iron Age due to the types of metals which were

commonly used for making tools, weapons, jewelry and many other objects used in the

everyday life. Since that time the corrosion of metals which led to the gradual decay

and final breakage of metal object attracted humans attention. However the lack of

knowledge of the structure of matter and of understanding of reaction mechanisms

restrained people from discovering and applying efficient methods of the corrosion

protection. The great progress in Natural Sciences which begun in the 18th century

and still continues opened new perspectives in the corrosion field. In the late 1700s the

Italian scientist Luigi Galvani published the work “De Viribus Electricitatis in Motu

Musculari Commentarius”1. In this essay Galvani made a bridge between chemical

reactions and electricity. He showed that nerves and muscles can by activated by

an electrical impulse. This work can be treated as corner-stone of electrochemistry.

Further development of this branch of science brought a complete theory which also

explains the phenomenon of metal corrosion.

1Latin for Commentary on the Effect of Electricity on Muscular Motion

2 1. The state of the art of corrosion protection by conducting polymers.

1.2 Principles of corrosion of iron.

Iron is believed to be the fourth most abundant element in Earth’s crust. Due to its

mechanical properties, relatively easy processing and widespread occurrence it became

as base element of many types of steels and alloys the most applied construction metal.

Therefore iron is a good example to show the basic mechanism of the atmospheric

corrosion. According to Evans [1,2] the corrosion of iron is an electrochemical process

which occurs when a surface of iron is covered by an aqueous moisture film. In such

conditions two reaction occurs: anodic iron oxidation and cathodic oxygen reduction.

They result in a dissolution of iron and a subsequent creation the iron hydroxide. The

whole process is presented by the following equations:

2Fe ⇀↽ 2Fe2+ + 4e− (1.1)

O2 + 2H2O + 4e− ⇀↽ 4OH− (1.2)

2Fe2+ + 3H2O +1

2O2 ⇀↽ 2FeOOH + 4H+ (1.3)

Usually FeOOH does not provide good protection of the underlying metal when exposed

to the atmosphere . This is due to SO2 which is present in the air as a pollution. This

leads to formation of H2SO4 which dissolves iron oxide hydroxide and opens pores in

its structure [3], which at later stages may be filled up by FeSO4. In this solution iron

dissolves as Fe2+ and partially oxidizes to Fe3+ and then precipitates as a magnetite

(Fe3O4). Due to the fact that magnetite is semiconductive electrons produced during

the dissolution of iron can be the transfered to the FeOOH/Fe3O4 interface. At this

location FeOOH transforms to magentite according to the equation:

Fe ⇀↽ Fe2+ + 2ee− (1.4)

8FeOOH + Fe2+ + 2e− ⇀↽ 3Fe3O4 + 4H2O (1.5)

(1.6)

On the other hand at sites with good air entry the magnetite reoxidizes to FeOOH.

3Fe3O4 +3

4O2 +

9

2⇀↽ 9FeOOH (1.7)

Summarizing, the corrosion of iron proceeds as a result of the establishment of the

electrochemical cell Fe |FeSO4|FeOOH . The corrosion process is accelerated at higher

content of SO2 in the air. Rusting of the steel in the absence of SO2 quickly decreases

when the surface is coated with a closed layer of FeOOH contrary to the case when the

1.3. Protection against corrosion 3

corrosion experiment is performed in SO2 containing atmosphere. In such conditions

the corrosion progresses with a constant rate [6].

1.3 Protection against corrosion

Considering that metal corrodes due to the reaction with aggressive species which are

present in the electrolyte an obvious method for corrosion protection would be to block

the access of these species to the metal surface. This can be done in various ways. The

first method relies on the strategy that corrosion products can create the dense layer

which will inhibit a transport of ions, water, oxygen and other species which take part

in the corrosion process to the metal surface. However, in many cases the corrosion

products just create a very porous layer e.g. rust which easily can be penetrated by

electrolyte. The properties of the protective layer can be improved by alloying, i.e.

by adding additional elements to the metal melt during its production. It has been

shown that products of the corrosion of iron alloyed with Cr, Ni, Cu, Si form denser

layers which effectively slow down further progress of corrosion. However, alloying is

a very expensive process which can significantly influence the price of the final metal

product. Additionally alloying influences other properties of the metal e.g. mechanical

properties. Another way is to coat the metal surface with a layer of substances which

will not corrode or corrode at small rates. In the following sections two most popular

kinds of coatings will be discussed: conversion coatings and organic coatings.

1.3.1 Conversion coatings

Conversion coatings are formed by a transformation of the metal surface due to chemical

or electrochemical reactions. Conversion coatings are mainly used as a pretreatment for

subsequently applied organic coatings. In this case, besides their function for corrosion

protection they are supposed to increase the adhesion of the later applied coatings.

Also they can be applied as a stand-alone corrosion protection. There are two general

kinds of conversion coatings used today: phosphate- or chromate based coatings. The

first one found an application as a protective layer for ferrous metals and alloys. The

other one is commonly applied on light metals such as e.g. aluminum. However, due


to their potentially carcinogenic toxicity they are now stepwise abolished, which has

lead to intense research for possible replacements.

Phosphate based conversion coatings

The idea of phosphating of metal surfaces was proposed for the first time by W.A Ross

[4] in 1869. His method consisted in a treatment of the steel in boiling phosphoric acid.

However, due to violence of this reaction it did not get many supporters. Significant

improvement of this process was proposed by Coslett [4] in 1906. To the phosphate

acid he added ferrous phosphate which decreased the reaction temperature. Since that

time this method started to be commonly used in steel industry and got known as

“Coslettizing”. With small modification this method was used till 1917 when Parker

proposed to add Zn and Mn to the solution which decreased the process time to 10

min. Further modification resulting in a subsequent decrease of the process time was

achieved by adding copper, nickel an finally Ti colloids to the solution [5]. In general

the phosphating process is performed in four steps:

1. Cleaning in an alkaline solution in order to remove residuals of oil which is often

added as a lubricant during metal production. The oil may provide temporary

corrosion protection in the period between fabrication and painting. In the case of

zinc and aluminum alloys the treatment in alkaline solution results in roughening

of the surface due to dissolution of zinc or aluminum oxide respectively.

2. Activation of the surface is aimed to create nucleation centers for the phosphate

crystals prepared in the third step. It is realized by an immersion of the metal

surface in a solution containing Ti colloids which adsorb to the surface thus

creating nucleation centers. The number of adsorbed particles decides about

amount and size of further phosphate crystals.

3. Phosphating in most cases is prepared in solutions containing either zinc phos-

phates or iron phosphates. The phosphation of steel proceeds according to fol-

lowing equations:

• In the case of zinc phosphates at first


the dissolution of iron from the substrate takes place (see. eq. 1.8). Due to

the cathodic part of this reaction the pH locally increases. When it reach

sufficiently high value zink phosphate precipitates on the surface according

to equation 1.10.

Step 1: Fe ⇀↽ Fe2+ + 2e− anodic reaction (1.8)

2H+ + 2e− ⇀↽ H2 cathodic reaction (1.9)

Step 2: 3Zn(H2PO4)2 ⇀↽ Zn3(PO4)2 + 4H2PO−

4 + 4H+ (1.10)

• When iron phosphates are used the phosphation proceeds in a different way.

At first iron oxide is created which later is transformed to iron phosphate.

Step 1: 4Fe + 3O2 ⇀↽ 2Fe2O3 (1.11)

Step 2: 2H2PO−

4 + 3FeO ⇀↽ Fe3(PO4)2 + H2O + 2OH− (1.12)

3Fe + 4H2PO−

4⇀↽ Fe3(PO4)2 + 2HPO2−

4 + 3H2(1.13)

4. Optional posttreatment is aimed to passivate the metal surface also between

the phosphonate crystals. Due to environmental reasons nowadays chromium-

free posttreatments such as ZrF2−6 , TiF2−

6 or Cu2+ are used nowadays for this

purpose.

Chromate based conversion coatings

A very effective protection against corrosion can be provided by chromate conver-

sion coatings (CCC). First coatings of this type were applied at the beginning of 20th

century, however the chromium-chromate coatings which show an excellent corrosion

resistance were developed in 1950 [7–10]. Since that time this kind of coating gained

a predominant market position. The superior corrosion protection provided by CCCs

is due to the complex mechanism of the corrosion inhibition. In general it consists of

three components. Firstly the coating which consists of an insoluble components acts

as a barrier against the aggressive species present in the environment preventing their

attack to the metal surface [11]. On the other side the CCC coating can be consid-

ered as a bipolar membrane [12, 13]. This means that the outer layer of the coating is

charged negatively and cation selective, while inner layer is positively charged thus an-

ion selective. Transport of ions through such membrane is significantly limited, hence

the surface reactivity reduced. The last component is known as a “self healing” effect


which is an ability to protect small mechanical or chemical defects in the coating. The

mechanism of “self-healing” proceeds as following. When the defect in the coating is

created and filled by aqueous electrolyte the Cr(VI) can be released from the CCC due

to hydrolysis of Cr(III)–O–Cr(VI) bonds in the porous Cr(OH)3 network of CCC [14].

Then it becomes reduced to Cr(III) hydroxide and precipitates in the defect. The con-

tribution of this effect to the overall corrosion behavior of CCC was not clear, however

recent measurements confirmed the importance of this effect [11].

Unfortunately CCC have a significant disadvantage. It is their potent toxicity and

cancerigenic influence on human body [15–17]. Therefore a large afford is invested to

exchange this coatings by others which are environmentally friendly.

1.3.2 Organic coatings

Another type of coatings commonly used for the corrosion protection are organic coat-

ings. In general they are recognized to be a barrier which separates the metal surface

form corrosive environment. However, every coating shows some ability for water [19],

oxygen [20] and ion transport [21], hence the protection mechanism can not be consid-

ered only as a simple blocking of access of above species. At least of similar importance

is that at the interface of the organic caoting and the substrate the electrochemical re-

action kinetics are significantly suppressed [74]. Organic coatings may also play a role

of a storage of corrosion inhibitors which can be released from the coating due to leach-

ing and actively protect the metal surface. In most cases they are added as pigments to

the primer coating which has a direct contact to the metal. This ensures that leaching

substances can quickly reach the metal surface. They can protect the surface either by

a sacrificial effect or by inhibiting properties. The sacrificial protection depends on a

dissolution of pigment particles instead of the protected metal. This protection mech-

anism is mostly used in order to protect ferrous substrates. It is realized by addition of

small zinc particles to the paint. Inhibiting properties of the coating can be attained

by addition of substances which are able to reduce the rates of cathodic or anodic

reactions. This can be achieved by various methods: by passivation the metal surface,

creation of insoluble precipitates on metal surface, anodic protection or by neutralizing

acidic species.


1.3.3 Corrosion of coated metals

The corrosion of coated metals in most cases initializes at defects in the coating. At

these sites the metal surface is exposed directly to the atmosphere. Very often defects

are created due to mechanical shock, chemical attack of aggressive species or as a result

of aging. Also they can be created during the production of the coating for example

as a result of inaccurate cleaning of the surface prior painting. At these sites water

and various ions present in the atmosphere may reach the bare metal surface and

initialize the corrosion process. The mechanism of the corrosion in the defect is the

same like in the case of uncoated metal. However, further progress of corrosion leads

to the disbondment of the coating from the metal substrate. There are two general

mechanisms which lead to the coating delamination: cathodic and anodic delamination.

Cathodic delamination

The cathodic delamination occurs on ferrous metals and other alloys featuring con-

ductive passive layers which were directly coated with organic coatings, i.e. without

further conversion coating. This process leads to the breakage of the bonding between

the metal surface and the coating. Due to this effect the coating looses its protective

properties. The cathodic delamination leads to creation of blisters or to flaking off the

lacquer from the metal surface.

Figure 1.1: Scheme of the mechanism of the cathodic delamination


The electrochemical model describing the fundamental mechanism of cathodic delami-

nation of coated steal was given by Stratmann et.al. [23–25] and is illustrated in figure

1.1. The cathodic delamination initializes at defect sites in the coating. In humid

atmosphere such places are condensation centers. Therefore shortly after exposure to

humid air the water condenses in defects. Due to impurities which are present on

the substrate and in the air the water becomes an electrolyte hence the corrosion can

start as described in section 1.2. Oxygen can be relatively easily transported through

the polymer layer. Therefore cathodic oxygen reduction can occur in places under

the lacquer just at the edge of defect where the contact with electrolyte is ensured.

Hence a galvanic element is formed, as the anodic reaction is inhibited at the intact

interface. At the anodic defect sites the pH of the solution is locally decreased while

at cathodic sites it becomes alkaline. This changes of the pH stabilizes both reactions.

The disbondement of the coating is caused by the cathodic reaction at the interface.

Due to the oxygen reduction radicals are created. These aggressive species attack and

destroy the polymer structure thusly causing a decrease of its adhesion. The reaction

are accompanied by a migration of cations from the defect toward the cathodic site.

This movement is necessary to ensure a charge neutrality and to close the electrical

circuit between corroding defect and delaminating interface (electrons flow through the

metal from the defect to the delamination sites). For a weak interface, the migration

of ions determines the rate of the delamination progress [24].

Anodic delamination

Anodic delamination is initiated in similar way as the cathodic delamination. At

defects local electrodes are created and separated. However, the disbondment of the

coating is caused by the dissolution of the substrate which leads to the creation of

thin crevice along the metal/polymer interface. The cathode usually remains in the

vicinity of the defect where an oxygen access is easy. This type of the delamination

appears often on conversion layer coated galvanized steal where in order to protect

the iron in the defect the sacrificial dissolution of zinc under the coating takes place.

Also aluminum substrates are susceptible for this kind of delamination for the case of

conversion coatings.

1.4. Conductive polymers for corrosion protection. 9

1.4 Conductive polymers for corrosion protection.

1.4.1 Introduction

Since the discovery of intrinsically conducting polymers (ICPs) in the late 1970 by

Heeger, MacDiarmid and Shirakawa, for which they were awarded with the Nobel

prize [26–28], the unique combination of physical and chemical properties of ICPs has

drawn the attention of scientists and engineers from many different fields of research and

they were studied for various application possibilities. The major feature which made

the ICPs so promising is that they posses both: electronic properties of semiconductors

and the processing advantages of conventional polymers.

ICPs used as modified electrodes found an application in electroanalytical, biosensing

and drug delivery devices. They can work as an conductometric transducer which can

measure the changes of the charges produced during enzymatic conversion [29, 30]. In

1990, Burroughes and co-workers at the University of Cambridge discovered that a

semiconducting polymer can also exhibit electroluminescence [31, 32]. This discovery

opened a new field of the ICP application as light emitting diodes. Such polymer LED

devices consist of several layers of polymer materials sandwiched between a metal-

lic cathode and an inorganic or organic anode, such as indium tin oxide (ITO) or

a conducting polymer [33]. ICPs also are used as gas sensors. For this application

the advantage of of conducting polymers compared to inorganic materials used until

now are their diversity, their easy synthesis and particularly, their sensitivity at room

temperature [34]. Another possible application of ICPs is as actuators. During the

reduction/oxidation processes the volume of conducting polymers changes due to in-

corporation or expulsion of ions and solvent. Hence the changes of a polymer volume

might be used as a source of the direct motion. The great advantage of this kind of

actuators is that due to the ’muscle-like’ nature they are more suitable for biomimetic

propulsions. They might be used in micro scale for cubic millimeter large mobile micro-

(a) (b)

Figure 1.2: Different forms of polyacetylene, a) cis-transoid, b) trans-transoid.


robots [35] as well as in the large scale. For example they are used to bend a foil which

is intended for use in the propeller blade of a novel underwater vehicle [36]. Moreover

ICPs a intensively studied for modern regenerative fuel cells [37]. The wide scope of

the above mentioned applications of ICPs shows how important this group of polymers

is for modern and innovative technology.

Polyacetylene was the first conductive polymer discovered in 1970’s. In the reduced

state this polymer shows only semiconductive properties but after treatment in an io-

dine vapor its conductivity increased more than fifteen order of magnitude and reach

a value in the range 104-106S/cm2 which is comparable that to metals [38]. The spec-

troscopic measurements confirmed that polyacetylene undergoes oxidation which leads

to the transformation of the neutral polymer chains into polycarbocations. In order

to maintain the charge neutrality of the polymer matrix the above reaction is accom-

panied by a simultaneous incorporation of anions between the polymer matrix. Poly-

acetylene can exist in two isomeric forms: cis-transoid and trans-transoid (see fig. 1.2).

The second form is thermodynamically stable. Polyacetylene can be obtained by the

(a) Polyacetylene (b) Polytiophene (c) Polypyrrole

(d) Polyphurane (e) Polyaniline (f) Polyazulene

(g) Polyindole (h) Poly para-phenylene

Figure 1.3: Idealized structrure of ICPs: a)polyacetylene, b)polythiophene,

c)polypyrrole, d)polyaniline


Ziegler-Natta polymerization of acetylene. However, application of this polymer meets

serious difficulties due to very low durability when exposed to air. Its conductivity de-

creases rapidly due to formation of carbonyl, hydroxyl and epoxide groups, which lead

to the destruction of the conjugated structure. An improvement of the polyacetylene

stability in air was achieved by the modification of the polymerization process proposed

by Naarmann [40]. However, problems with stability of the polyacetylene still remain

as a significant disadvantage of this polymer.

Due to extensive research in the area of conductive polymers another group of con-

ductive polymers was discovered. Heterocyclic conjugated polymers: polypyrrole [41],

polyfuran [42] and polythiophene [42–45] can be conveniently prepared by electro-

chemical or chemical oxidation of pyrrole, furan and thiophene respectively. Also

other aromatic systems such as aniline [46, 47], azulene [48, 49], indoline [42], para-

phenylene [50, 51] were found to undergo a polymerization resulting with conductive

polymer (see fig. 1.3).

Electronic conductivity of ICPs

In the following section a short overview on the mechanism of the electronic conduc-

tivity of ICPs will be presented. The polyacetylene will be taken as an example to

illustrate principles of the conduction mechanism in conducting polymers, because of

the simplicity of its structure. The polyacetylene chain consists of single and double

bonds which are situated in alternative sequence. As shown in figures 1.4a and 1.4b the

polyacytylene has a degenerated ground state. In other words there is more than one

equi-energetic resonance structure. When both structures are present in a single poly-

meric chain, a defect results where the two structures meet (see fig. 1.5a). This defect

is called a soliton and consists of a single unpaired electron, however the overall charge

still equals zero. Hence the above soliton does not carry any charge and therefore it

can be called as neutral. By controlled addition of p-doping anions which consume free

(a) (b)

Figure 1.4: Degenerated states in polyacetylene with reversed order of alternating

bonds.


electrons a positive soliton can be created (see fig. 1.5b). The n-doping of polyacetylene

will result in a negative soliton which is schematically presented in fig. 1.5c. Neutral

and charged solitions are stable. Their stability is achieved by spreading the charge

over several monomer units [52]. The presence of solitons in the chain influences the

band structure of the polymer. Due to them a new energy level is created in the middle

of the band gap which can accommodate zero, one or two electrons per solition. When

the doping level is sufficiently high the soliton states may interact between each other

and form a soliton band. The band structure of polyacetylene is presented in figure 1.6.

(a)

(b)

(c)

Figure 1.5: Solitons in the polyacetyelene chain, a)neutral b)positive c)negative

Solitons may arise as a result of thermal isomerization of cis-polyacetylene to the trans

structure. Solitons may move along the polymer chain by successive alternation of

neighbor single bonds. They can propagate as a wave which is comparable to a moving

electron or hole in semiconductors. Solitons may exchange electrons between neighbor

chains according to “intersoliton hoping” mechanism which is explained in the follow-

ing. Let’s assume the situation that two polymer chains are placed close together and

in every chain exists one soliton. In the first chain there is an unlocalized neutral

soliton which may move along the chain. In the second one there is a positive soli-

ton which is localized by a doping anion. When the moving soliton approaches the

localized one the free electron from the first chain might hop to the localized charge of

the other chain. Two neutral solitons which are present in a single chain may recom-

bine. That results in the elimination of soliton defects in the chain. A charged soliton

together with a neutral one can form an energetically preferred state called polaron

which causes two new states in the band gap: a bonding and an anti-bonding. For

instance in polypyrrole the polaron states are symmetrically located about 0.5 eV from

the band edges. Chemically a polaron can be considered as a radical cation. Polarons


Figure 1.6: Band structure of the polymer with the soliton states

Figure 1.7: The band structure of polymer with bipolaron states

may also recombine and create a bipolaron which is a doubly charged defect. Polarons

as well as bipolarons are delocalized over several monomer units. When the polymer is

highly doped interacting bipolarons can create a band and antiband within the band

gap.

Polymers does not have a degenerated ground state. It is an example for the so called

conductive heterocyclic polymers. In the reduced form it exists as a chain of aromatic

rings connected with long bonds (see fig. 1.8). Another form of PPy shows an alternat-

ing bond sequence called the quinoid form (see fig. 1.8). However, this form is unstable

and immidiately transforms to the aromatic one, i.e. these states do not have the same

energy. Hence, no degeneration. In such structure formation of two single solitons

is not energetically favorable. Therefore polaron defects are created. At sufficiently

high doping levels they recombine into bipolarons. Hence polarons and bipolarons are

responsible for the electronic conductivity in polypyrrole.


Figure 1.8: Forms of the polyprrole chain.

1.4.2 Polymerization methods

Electrochemical polymerization

Heterocyclic conjugated polymers can be prepared by electrochemical polymerization.

Due to many advantages of this method over chemical polymerization approaches it

became a convenient way for the preparation of conductive polymers for research. It

allows a direct polymerization of the polymer on conductive substrates eliminating

problems concerned with the application of the insoluble (or difficult to solve) poly-

mer. The obtained polymers are already in the oxidized state, hence further oxidation

to the conductive form is not necessary. The polymers can be directly doped with

various anions and the doping level can be adjusted by the proper choice of the prepa-

ration conditions. Additionally, the thickness of the prepared polymer film can be well

controlled during the polymerization process. Moreover, it is very convenient to char-

acterize the polyrrole layers by electrochemical methods such as cyclic voltammometry

or electrochemical impedance spectroscopy.

The process of electrochemical polymerization was intensively studied during the last

decades. Despite of that, some parts of the polymerization process are still the subject


Figure 1.9: The mechanism of the polypyrrole polymerization.

of discussion. There is an agreement between scientists that the polymerization process

is initiated by the one electron oxidation of the monomer to a radical cation. Further

growth of the polymer may proceed in two scenarios. According to some authors [53]

two radical cations recombine and form a dihydro dimer dication, which subsequently

deprotonates and forms a dimer. The dimer formed in such manner has a lower oxida-

tion potential than the monomer. Therefore it undergoes oxidation and couples with

another radical cation. The other possibility was proposed by Wei et. al. According to

this theory, the propagation of the polymer chain progresses due to elecrophilic addi-

tion of a neutral monomer molecule to the radical cation [55,56]. The potential which

is necessary to oxidize an oligomer is less positive than potential needed for oxidation

of the corresponding monomer. Therefore simultaneously with the polymerization the

oxidation of ICP chain proceeds. Finally, both mechanisms of the polymerization lead

to polymers in the doped state. Figure 1.9 illustrates the first one of the two proposed

mechanisms for the polymerization. According to the above reaction PPy layers can be

deposited on various metallic substrates. However, many oxidizable metals are not sta-

ble at the anodic potentials which are applied during the polymer deposition. Hence,


the preparation of ICP films on metals like iron, zinc is difficult due to the dissolution

of the metal during the polymerization. In order to hinder the substrate dissolution

those substrates have to be passivated prior to polymerization. It the case of iron and

mild steel this can be done by a treatment of the substrate in 10% nitric acid [61].

The obtained passive layer does not prevent the polymerization but just hinders the

dissolution of the substrate. Hence, a galvanostatic polymerization of PPy in solutions

containg salts such as: K2C2O4, KNO3, Na2SO4 resulted in strongly adhering films.

Another method of the passivation of not nobel metals was proposed by Lacaze et.

al. [63,64] He showed that during the polymerization of PANI or PPy on iron in oxalic

acid solution a thin layer Fe(II)-oxalate precipitates on the metal surface. According

to Iroh this layer dissolves just before the polypyrrole formation [62], thus the PPy is

deposited directly on iron surface.

The properties of the obtained polymers may significantly vary with the chosen prepa-

ration conditions. According to Ouyang (see tab. 1.1) the conductivity of PPy films

prepared in different solvents and electrolytes may vary in the range between 1 ·10−6−79S/cm2 [57]. The type of solution is one of many factors which influence the conduc-

tivity PPy layers. In general conductive PPy layer can be obtained from organic or

aqueous solutions, however the most conductive polymers can be obtain when prepared

in water based electrolytes [57] of the tosylate salt. In the case of polymerization in

organic solvents good conductivity might by obtained using acetonitrile with small ad-

dition of water (around 2-5%) due to a proton scavenging effect described in [66]. The

conductivity of the polymer depends also on other factors such as: anion basicity [67],

solution temperature, doping anion concentration [57]. Therefore even slight changes

of the preparation conditions may significantly influence the properties of the polymer.

Chemical polymerization

Another method of obtaining conductive polymers is chemical polymerization. Con-

trary to the electodeposition this allows to prepare the ICP on various type of substrates

such as: glass, metals and metal oxides, polymers, textiles. Additionally this method

is less costly comparing to the electropolymerization, therefore it is preferable for the

ICP production on the industrial scale. The mechanism of the polymerization is simi-

lar to that one described in the former section. The only difference is that the anodic


Doping anion

SolventPPy(BF−

4 ) PPy(ClO−

4 ) PPy(NO−

3 ) PPy(TsO−)

DMSO 7 · 10−6 1 · 1010−6

DMF 1 · 10−4 5 · 10−4 3 · 10−6 8 · 10−3

THF 31 0.9

H2O 8.4 34 2 79

CH3NO2 69 56

Table 1.1: Conductivity of the pylypyrrole prepared in various solvents and electrolytes

[57].

electrode is replaced by a chemical oxidizer. The chemical polymerization can be pre-

formed either in the bulk of the solution or directly on the surface of the substrate. In

the case of the first method the monomer is dissolved in the solution which contains

strong chemical oxidants such as ammonium peroxydisulfate (APS) [58], ferric ions,

permanganate or dichromate anions [59], or hydrogen peroxide [60]. This oxidants are

able to oxidize a monomer to a radical cation which subsequently reacts with another

monomer and thusly oligomers and polymers are created. Finally the insoluble poly-

mers precipitate as solids. However conducting polymers in general are poorly or non

soluble, therefore the obtained precipitate is difficult to process, in particular it is dif-

ficult to apply as a coating. In order to coat a surface with a layer of the conductive

polymer another method can be utilized. The surface of the substrate which is sup-

posed to be coated has to be enriched either by the monomer or by an oxidizing agent.

Then it has to be treated with the oxidizer or monomer respectively. The advantage

of this method is that polymerization occurs only at the desired surface. On the other

side the above method can be applied only to materials which can be modified by either

a layer of an oxidizing agent or a layer of a monomer.

1.4.3 Inhibition of corrosion by conductive polymer.

Conductive polymers as materials which already showed some anti corrosion behavior

became a natural candidate for the further research. However, numerous investigations


reported by the literature lead often to contradicting conclusions about the performance

of this polymer class as corrosion inhibitors [68–70]. This is probably due to the

lack of knowledge of the basic electrochemical mechanisms involved in the substrate

protection by these polymers. The three few models of the possible corrosion protection

mechanism which are proposed in the literature. The three most investigated ones will

be described below.

One of the most extensively studied models is so-called ”ennobling mechanism”. It is

based on the assumption that the conductive polymer acts as an oxidizer and main-

tains metal in the passivity domain. This mechanism could induce the oxidation of the

free metal surface at small defects in the passive layer. But while some publications

report that this mechanism works, others report that this is only true in chloride free

solutions [72,76,77]. Wessling claimed that the ennobling mechanism also improves the

passivity of the oxide layer at the metal/polymer interface through oxidation by the

applied conducting redox polymer, thusly inhibiting electrochemically driven delamina-

tion (see e.g. [73]). However, for example for iron or steel the oxide layer at the buried

interface is passive anyway at the high pH prevailing during delamination [23–25, 74].

Hence, it is not astonishing that this effect is negligible [78].

An alternative mechanism is that, the electrons produced during metal oxidation at

the defect area can go into the polymer and dislocate the oxygen reduction process

from the metal/polymer interface. This would hinder coating disbondment caused by

interfacial oxygen radicals and/or hydroxide ions [69, 71].

The typical polymeric coatings used as a corrosion protection for non noble metals

act as a barrier for ions and corrosion products. The coating blocks their movement

and does not allow to establish a galvanic coupling between local anodes and cath-

odes. So, the ion free coatings which will not enhance ionic currents are highly recom-

mended for corrosion protection. However, in the case of conductive coatings the good

ionic conductivity could be an advantage. The ”self healing” mechanism proposed by

Kendig [75] and which is considerably discussed in literature [146], is based on the

assumption that doping anions with corrosion inhibiting properties inside the polymer

matrix are released during the reduction of the polymer and migrate to the corroding

defect. Here the inhibitor anion could significantly decrease of the corrosion rate. The

conductive polymer would act as a storage for corrosion inhibitors which supplies them

immediately just after the corroding defect appears. The efficacy of the inhibition is


strongly dependent from the concentration of the inhibitor. Therefore the transport of

the inhibitor must be fast enough to ensure proper concentration.

The mechanisms of the corrosion protection presented above are based on unique prop-

erties of the conductive polymers. In the case of the “ennobling mechanism” the charge

stored in the polymer is used for the passivation of the metal surface while for the

mechanism of the shift of the oxygen reduction site the electronic conductivity of the

coating is the most important property. On the other side the “self healing” mecha-

nism depends on the ionic conductivity of the coating and its ability to release anions

form the polymer matrix. But all these mechanisms may contribute simultaneously to

the substrate protection. It has to be pointed that many properties of the ICPs such

as ionic and electronic conductivity, type of doping anion etc. can be varied during

the polymerization process, thus it is possible to adjust properties of the coating as

required by certain mechanism. However, there is a serious danger that properties

demanded for one mechanism are unwanted in the case of the other one. Therefore in

order to estimate the usability of conductive polymers as corrosion protective coatings

all aspects of their electrochemical behavior have to be taken into account.

In the following chapters the validity of the above mechanisms will be studied. The

optimal properties of the polymer which is supposed to be used as a corrosion protection

coating will be proposed with respect to all investigated mechanisms.

Chapter 2

Scanning Kelvin Probe

2.1 The basics of potential measurement by the

Kelvin Probe technique

The Kelvin Probe (KP) is an instrument which provides a non-contact and a non-

destructive measurement of the work function of conducting and semiconducting ma-

terials. The principles of the Kelvin Probe technique was first proposed by Lord Kelvin

in 1862. The idea behind the Kelvin Probe is based on the effect that under equilib-

rium conditions different metals equalize their Fermi levels when they are connected.

This effect is schematically presented in figure 2.1. When metals are disconnected

(see fig. 2.1a) each of them posses its own Fermi level and both metals are electrostat-

ically neutral. After the connection (see fig. 2.1b) electrons flow from the metal with

higher Fermi level to the metal with lower Fermi level. That results in changes of both

levels. When the equilibrium is reached the Fermi levels in both materials are equal.

Due to the transfer of electrons metals get charged. The metal which initially had lower

Fermi level gets a negative charge, while the other one gets a positive charge. That

results in a Volta potential difference ∆Ψrefsam which is proportional to the difference in

work functions according to the following equation.

∆Ψrefsam = Ψsam − Ψref = −(Φsam − Φref)/e (2.1)

Assuming that the work function of one metal is known, the ∆Ψrefsam can be utilized for

directly measuring of the work function of the other metal. Kelvin proposed that the

Volta potential difference can be measured, if the investigated metals form a capacitor

22 2. Scanning Kelvin Probe

(a) (b)

Figure 2.1: Position of Fermi levels for a) not connected metals, b) connected metals.

Figure 2.2: Scheme of the Kelvin Probe

with a modulated capacity (see fig. 2.2). If one of the capacitor plates vibrates with a

frequency ω the capacity changes periodically according to the equation.

C = ǫA

d0 + ∆d · sin(ωt)(2.2)

where A is the area of the capacitor plates, d0 is the average distance between them

and ∆d is the amplitude of the vibration. This entails a change of the charge which

is stored on the capacitor plates causing the flow of the alternate current. Considering

that

q = C · ∆Ψrefsam and i =

dq

dt(2.3)

2.2. SKP measurement over conductive coatings 23

then

i = ∆Ψrefsam · ǫAω · ∆d

d20

· cos(ωt) (2.4)

The alternating current flowing between both plates is proportional to the Volta po-

tential difference. However, this current reaches the magnitude of 10−12A and its

measurement procured significant difficulties due to not sufficient development of sig-

nal acquisition and processing techniques. A large progress in these technologies in the

early 20th century cprovided the means to build a reliable KP instrument. Nowadays

Scanning Kelvin Probes (SKP) which allows the mapping of the potential over the

sample are commonly used in surface analysis. They offer a high sensitivity of 1-3 meV

and a very good spatial resolution of about 100 µm and better.

A new application of the Kelvin Probe was proposed by Stratmann et.al. [79–81]. He

suggested to use the SKP for the investigation of the corrosion potential at buried

metal/polymer interfaces. He proved that changes in the electrode potential of the

substrate/polymer interface occurring during delamination areequivalent to changes in

the Volta-potential difference ∆ΨRefPol measured between the SKP reference probe and

the polymer surface. The corrosion potential Ecorr can be expressed as following:

Ecorr = ∆ΨRefPol + const (2.5)

Hence, the Kelvin Probe can be utilized for the investigation of the local changes of the

corrosion potential under the coating. Since this invention SKPs became extensively

used in corrosion studies. This non invasive method gave to scientists a unique pos-

sibility of the observation of the corrosion potential evolution during corrosion tests.

Especially the SKP was utilized in studies of the cathodic delamination phenomena.

Results obtained using this equipment helped to formulate the theory which explains

this effect [79–81].

2.2 SKP measurement over conductive coatings

The SKP can be also be used for an investigation of the delamination of conductive

coatings from metallic substrates. However, the interpretation of SKP profiles is more

complex compared to the case of profiles measured during the delamination of insulat-

ing coatings. ICPs posses their own Fermi level. Hence, the measured Volta potential

difference is proportional to the difference in work function between the polymer and


the reference tip. However, this is true only when the polymer is in oxidized state i.e. it

is conductive. If the polymer gets reduced e.g. by polarization to a sufficiently negative

potential, then it becomes insulating and the SKP measures the potential difference

between metal substrate and the SKP tip.

2.3 The hardware realization

There are two general approaches for the practical realization of the Kelvin Probe. The

first method proposed by Stratmann et.al. is based on the idea of a compensation of

the Volta potential difference ∆Ψrefsam by an additionally applied potential Uapp [79–81].

When the applied poten‘ tial Uapp balances the ∆Ψrefsam, then the alternating current

between the sample and reference will not flow. This situation is presented in figure

2.3. Thus, by a variation of the Uapp and a simultaneous monitoring of the current flow

the ∆Ψrefsam can be determined. This procedure is called a “nulling procedure”. The

nulling of the SKP signal can be realized by an electronic circuit which is schematically

presented in figure 2.4. The integration loop is intended to balance the ∆Ψrefsam by the

Uapp. It is realized as following. A vibration of the SKP tip is caused by a sinusoidal

signal which comes from the generator built into the Lock In amplifier. The current

flowing between the tip and the sample is converted and amplified to a voltage signal

by the preamplifier. The amplification of this signal is proportional to the difference

(a) (b)

Figure 2.3: Compensation of the Volta potential difference by applied external potential

2.3. The hardware realization 25

Figure 2.4: Scheme of the electronic circuit used for the nulling procedure.

between applied potential Uapp and the Volta potential difference ∆Ψrefsam. The amplified

signal goes to the Lock In amplifier which demodulates from the signal the component

at the same frequency as the signal generated for the vibration of the tip. Then its

amplitude is expressed as a DC voltage which is given to the input of the integrator.

That results in the change of the Uapp. The Uapp will increase as long as it does not

match to measured potential ∆Ψrefsam . When the output of the amplifier equals to

zero no potential will be measured by the Lock In and no signal will be given to the

integrator. Therefore the Uapp will stabilize at the value which matches ∆Ψrefsam.

Another approach to the measurement of the contact Volta potential difference was

proposed Baikie [82]. His idea is to measure the ∆Ψrefsam in off-null conditions. He

suggested to apply a series of different Uapp and measure the signal at the output of

the preamplifier. Then it is possible to find the value of Uapp which balances the ∆Ψrefsam

using a linear approximation. This method of the ∆Ψrefsam measurement additionally

allows for the automatic control of the distance between the SKP tip and the sample

surface [83].

The laboratories in the Max-Planck-Institute in Duesseldorf are equipped with SKP

which operate in the nulling mode. The apparatus is presented in figure 2.5.


Figure 2.5: Scanning Kelvin Probe at Max-Planck-Institute in Duesseldorf

2.4 The control software

In this thesis the SKP was intensively used to perform delamination tests of conductive

coatings. In order to automatize the measurement process a special software dedicated

for control of the SKP apparatus was written. Programing was done in the HPVee

6.0 graphical programing language. The program was designed to fulfill the following

requirements:

• Acquire signals from the SKP electronics and from other devices monitoring

parameters of the experiment.

• Control the position of the sample and the position of the SKP tip. Allow the

user to change both positions from the software level.

• Allow user to measure the potential over one point of the sample for certain

period of time.

2.4. The control software 27

• Allow user to perform a mapping of the potential over the desired area. The

scanned region should have a rectangular shape. Its size should be adjustable by

user.

• Allow user to automatically perform a sequence of measurements.

• Measured values should be stored in ASCI format.

• Measured data should be visualized in 2D or 3D graphs.

• Graphs should be exportable to various graphical formats.

• Program should have an intuitive structure and a nice appearance.

2.4.1 Structure of the program

The program consists of a main panel (see fig. 2.6) which is followed by four sub-panels:

• Configuration - this panel allows to enter parameters which are essential for the

proper operation of the program.

• Scan - is devoted to the control of motors, acquisition of measurement data,

control of the experimental parameters

• Look - sub program which allows a simple manipulation of the measured data.

• Info - presents short information about the program.

The above panel is shown in figure 2.7. The above subprogams can be activated by

pressing corresponding buttons which are placed on the “Main Panel“.

2.4.2 Configuration of the program

All parameters which are essential for the proper operation of the program are stored

in the configuration file. Additionally the configuration file contains parameters which

can be edited by the user, e.g. the default file path for the data storage, calibration

factor. Every user can create his own configuration file. This can be done using the

Configuration Panel of the Control program or by editing the configuration file in a text


Figure 2.6: The main panel of the Control Software for SKP at Max-Planck-Institute

in Duesseldorf.

editor. The path to the actually used configuration file is stored in the file “init.cfg“

which should be placed in the same directory as the main program.

2.4.3 Panel ”Scan“

Panel ”Scan“ (see fig. 2.7) is intended to provide to the user a full control of the SKP

apparatus during the preparatory and execution phase of the experiment. It consists

of three general sections:

• Motor section - responsible for the control of SKP motors.

• Experiment section - this section allows to set all parameters of the experiment.

• Data visualization section - consist of charts and indicators which present mea-

sured data.

Motor section

The motor section is devoted to manage the movement of the sample and its positioning

before and during the experiment. The screenshot of this section is presented in the

figure 2.8. There are radio buttons placed on the upper left side of the section. They


Figure 2.7: Front panel of the control program for the Scanning Kelvin Probe.


allow a to chose a motor which is responsible for the movement in certain axis. Motors

X,Y are responsible for the movement in horizontal plane while motor Z adjusts the

height of the SKP tip. The actual status of motors is indicated by LEDs which are

placed beside the radio buttons. Green color of the LED and description ”stand“

indicate that the motor is ready to perform a step. When the motor is moving the

LED turns red and the description is changed to ”moving“. The size of the single step

can be adjusted by moving the slider of the “Step Size” control. The speed of the

motor can be chosen using the “Speed” control. By pressing the button “Direction”

the direction of the motion for the chosen axis can be changed, e.g. for the X axis “+”

means a motion to the right, while “-” means a motion to the left. For the Y axis “+”

and “-” mean motions forward and backward respectively. For the Z axis “+” and “-”

mean motions down and up. Button stop allows to immediately stops the motion of any

motor. Steps can be made in the single or continuous mode. In the single mode (button

“make step”) after pressing the button only one step is preformed. In the continuous

mode steps are performed in sequence until the button “stop” is pressed. The last

button in this section is called “Ref/Zero”. Pressing of this button results in the reset

of the current position and setting the position counters to zero for all motors. During

experiment the “Motor section” is not active. Therefore any changes of parameters

or any attempts to perform a step will be ignored. The section is reactivated after

stopping the experiment.

Figure 2.8: Motor section in the SKP control program, panel “Scan”.


Experiment section

The Experiment section (see fig. 2.9) allows the user to set all parameters of the mea-

surement. The first parameter which has to be set is a name for the file in which the

measured data will be stored. It can be done after pressing the button “File selection”.

After that appears a standard dialog window which allows to choose a filename. Short

description of the experiment (it should not exceed 255 characters) can be entered af-

ter pressing the button “File Info”. The experiment section also contains a calibration

button. The calibration procedure will be described later. There are the following

types of experiments allowed by the SKP Control program:

1. Point Measurement. The experiment consists of the continuous monitoring of

the potential over one point of the sample. The measurement is performed for

a desired period of time which can be adjusted by setting a parameter “Point

measure. time” which represents the duration of the experiment in seconds.

2. Surface Scan. During this procedure the SKP maps the potential distribution

over a chosen area. The measured area always has a rectangular shape with the

one exception when 0 steps in the Y axis are chosen. In this case just a line

along the X axis is scanned. The size of the mapped area as well as the number

of measurement points are chosen by setting two parameters for axes X and Y:

“Scan Range” which is expressed in µm and “No. Steps”. These values are used

for the calculation of the step size for each axis. The calculation is performed

after pressing a button named “Set Par” and the resulting size of steps is shown

on indicators on the panel. It is possible to measure a series of potential maps

separated by a time delay which is set by the parameter “Scan delay” expressed

in seconds.

The selection of the desired experiment can be done by marking the proper option on

the radio button which is located in the middle of the Experiment section. Parameters

located in the lower left part of the section control the way the data acquisition from

additional equipment, such as the HEKA potentiostat, KHEITLEY picoamperometer,

oxygen sensor, humidity and temperature sensors. The parameter “Curr. Range”

allows to set current ranges for potentiostat and picoamperomenter. Change of ranges

has to be confirmed by pressing corresponding buttons “Set Curr Pot” for potentiostat

and “Set Electr. Curr” for Picoamperometer. Parameters “Measur. Interval” and


Figure 2.9: SKP control software, “Experiment” section, panel “Scan”

“Measur. time” refers to the delay between readings and to the total time when all

sensors are monitored. All parameters entered in the “Experiment” section have to be

confirmed by pressing a button “Set Par”.

Data presentation

All measured data are presented on the panel by means of charts and indicators. A 2D

chart in the upper right corner of the panel shows the measured potential (y axis) vs.

time or position of the tip (x axis), depending on the type of performed experiment.

In the case of Point Measurement it represents time, while for other experiments it

shows the position of the tip. The chart is updated after every single measurement.

Additionally, data obtained during Surface Scan are shown on the 3D chart which

appears in the separate window when at least three lines are measured. In order to

make the 3D chart more readable the potential is additionally indicated by a color

scale. Appearance of the chart including colors of axes and background, font type and

size, axes range, can be modified by the user. The 3D chart can be exported to various

graphical formats using the option “Save” from the menu of the chart window.

The chart present in the low right corner of the front panel (see fig. 2.7) is devoted to

present readings form:

• HEKA potentiostat - potential, polarization current


Figure 2.10: Data presentation, 3D graph window.

• Picoamperometer - corrosion current

• Oxygen sensor - oxygen content in the SKP chamber expressed as a voltage in

the range of 0-5 V. Maximum value corresponds to the amount of oxygen which

is present in the air.

• Humidity Sensor - Relative humidity inside the SKP chamber expressed in %

• Temperature Sensor - Temperature inside the SKP chamber expressed in ◦C

Readings from the above gauges are also shown on indicators.

2.4.4 Subprogram “Look”

The sub-program “Look” is a tool which helps the user to crate 2D and 3D charts from

the data saved in ASCI files and in “fig” files. Program is able to treat the data which

were obtained during the surface scan experiment. It allow the user to:

• recalibrate a saved measurement

• create charts from a selected area of the scan

• flatten the slope of the measurement which results from the misalignment of the

sample during the experiment


• detect the position of the delamination front and analyse its progress with time

on the basis of a series of measurements

• adjust graphical details of charts e.g. font type, font size, axis thickness, color

scales, etc.

• export obtained charts to various graphical formats

The sub program “Look” exists also as a stand alone application which can be used

for the data treatment on other computers.

2.4.5 Selected subroutines and tasks

Communication with hardware

The main task of the control program is to acquire the data from the following mea-

surement devices:

• SKP integrator - measures Volta potential difference.

• Oxygen sensor - measures the oxygen content inside the SKP chamber.

• Humidity and temperature sensor - measures the temperature and the relative

humidity inside the SKP chamber.

• HECA potentiostat - allows a polarization of investigated samples, measures

polarization currents and polarization potential.

• Keithley Picoamperometer - measures the current flowing from the defect to the

delaminating sample

The above mentioned devices are equipped with analog outputs providing a voltage

which represents the measured data. This voltage can be used for the acquisition of

experimental parameters such as currents or potentials. This can be done using an

AD converter which will convert the voltage to the digital value which by subsequent

scaling corresponds to the actual value. Factors used for scaling are given by the

manufacturers of the devices. The conversion of the analog signal was done be means

of the AD converter ICPCON I-7017. Communication between the computer and


the AD converter was performed by a serial port RS 232. The parameters of the

communication protocol were supplied by the manufacturer of this device.

Another device which has to communicate with the PC is the ISIC motor controller.

It is responsible for the movement of all motors and positioning of the sample and

SKP tip. A communication with the controller was realized via a RS 232 port using

commands supplied by the manufacturer of the motor controller.

Steady-state potential detection

As described above the Volta potential difference is measured by balancing it with an

external voltage. This procedure is realized by the integration loop. The time necessary

for the equalization of both potentials Ψrefsam and Uapp depends the time constant of the

integrator and from the difference between them. Hence the time necessary to get

a reading from the SKP electronics may vary. For example if the SKP tip is moved

form the part of the sample which has a potential V1 to the another part of the

sample which has a significantly different potential V2, then the balancing voltage will

gradually change from U1 until it reaches a value U2 sufficient for nulling. In this

situation the time of nulling will be long. If V1 and V2 are close, then U1 quickly

will reach U2, hence nulling time will be short. Therefore during the mapping of the

potential the software has to recognize if after every step of the motors the nulling

procedure is over and then record measured potential. This is achieved by checking a

few subsequent measurements of the applied Uapp potential. If the difference between

them is small that means the procedure is finished. Thusly, the program collects six

subsequent measurements of the potential V1, V2, V3, V4, V5, V6. Using following

equations the parameter ∆V is calculated.

Va = V1 + V2 + V3 Vb = V4 + V5 + V6 (2.6)

∆V = |Va − Vb| (2.7)

If this parameter is smaller than 3 mV than it is assumed that the nulling procedure is

accomplished, hence the average value of all six measurements can be taken as a result

of the measurement.


SKP Calibration

Calibration of the SKP proceeds according to the method proposed by Stratmann

[79–81]. From the software side it consists in pressing the “Calibration” button in the

“Experiment” section when the SKP tip is positioned over the electrochemical couple

which is used as a calibration standard (usually Cu/Cu2+ sat.). The calibration factor

is expressed in mV vs. SHE can be saved and recalled from the configuration file.

Data storage

The measured data are saved to the file which is selected at the beginning of the

experiment. Saving to the file is performed after each measurement. For all types

of experiments the data are saved in ASCII format. Additionally for “Surface Scan”

the file with the extension “fig” is created. This file contains measured data which are

stored according to the Matlab data format and can be opened with Matlab compatible

applications. When a series of experiments is performed, the data obtained for each of

them are saved automatically to separate files. Each filename is created by adding the

ordinal number of the experiment to the filename which was chosen by the user before

the start of the sequence of experiments.

Chapter 3

The preparation and

characterization of the polypyrrole

films

3.1 Electrodeposition of polypyrrole films

As described in the section 1.4.2 the polypyrrole might be obtained in two ways: by

chemical and by electrochemical polymerization. Of the many advantages of the elec-

trochemical polymerization some deserve to be underlined in more detail. The great

value of the electrochemical polymerization is that it allows a precise control of the film

thickness during the polymerization process. For application as corrosion protecting

coatings the conductive polymer has to be applicable on a metal surface as coating.

In the case of ICPs prepared chemically this task is not trivial. Due to the very weak

solubility of these polymers (especially polypyrrole) the common methods of coating,

such as spreading of a solution of the ICP followed by the evaporation of the solvent

cannot be applied. Using the electropolymerization route the polymer films are de-

posited directly on the metal surface, so the above difficulties do not arise. Therefore,

electrochemical polymerization was chosen as a best method for the preparation of

the conductive coatings for the corrosion studies. A lot of research has been done

concerning the electrochemical polymerization of polypyrrole [84, 86, 88]. Mostly the

polypyrrole is deposited on substrates which have a cylindrical geometry for example

on wires. This geometry ensures the homogeneous growth of the polypyrrole layers due

38 3. The preparation and characterization of the polypyrrole films

to the lack of geometric nucleation centers. However, considering that the ICPs are

supposed to replace the conventional coatings which are in most cases applied on large

steel sheets the performance of the ICP should be investigated in the same geometry.

Therefore all experiments aiming on investigation the corrosion resistance of ICPs were

done on flat PPy layers.

3.1.1 Substrate preparation

In order to avoid chemical interaction of the PPy with the substrate gold was chosen

as substrate for the deposition of the polypyrrole. In order to obtain smooth and ultra-

flat surfaces the PPy was deposited on the gold films evaporated on glass substrates.

Glass plates with lateral dimensions of 10 x 12 mm and thickness of 1 mm were bought

from Berliner Glass. Before the evaporation glass substrates were cleaned in a freshly

prepared mixture which consisted of 30% of H2O2 (concentration of 30%) and 70% of

concentarted sulfuric acid (concentration of 96%). Substrates were immeresed in this

mixture for 10 min. Than they were rinsed first with Milipore water and then with

pure ethanol (pureness of 99.9%). After that substrates were dryied under nitrogren

flow. The evaporation was done using a Leybold PVD system which was additionally

equipped with electron-beam as well as thermal deposition sources. The pressure in

the chamber during the deposition of the films was 10−4 mbar or below. The thickness

of the prepared layers was monitored in situ by quartz microbalance. Due to the very

weak adhesion of gold to the glass the evaporation of the films had to be done in two

steps. In the first step 4 nm a layer of chromium was prepared by e-beam assisted

evaporation. This layer served as adhesion promoter for the gold film to the glass.

Subsequently, directly following the Cr layer deposition the gold film was prepared

using the thermal vapor deposition method. The thickness of the gold layer was about

300 nm.

3.1.2 Deposition of the polypyrrole layer

The elechtrochemical polymerization and other electrochemical experiments were per-

formed in a standard three electrode cell, consisting of a reference electrode and a gold

sheet employed as a counter electrode (see fig. 3.1). Instead of the commonly used

Ag/AgCl electrode a Hg/Hg2SO4 electrode was used. This excluded the possible leak-

3.1. Electrodeposition of polypyrrole films 39

Figure 3.1: The setup for the electrochemical experiments.

ing of chloride anions from the electrode and their further incorporation to the polymer

matrix. The substrate was mounted in a teflon holder and acted as a working electrode.

In this work, all potentials showed in the diagrams are referred to the SHE electrode. A

HEKA potentiostat/galvanostat controlled by self-programmed software was used for

all electrochemical experiments. The polypyrrole (PPy) layer was prepared from 0.1 M

pyrrole containing aqueous solution. As supporting electrolyte 0.1 M 4-toluol sulfonate

sodium salt solution was employed. It was found that the polypyrrole layers grew ho-

mogeneously, when the pH of the solution was adjusted to 4. This was done using

4-toluol sulfonate acid. Before experiments, the solution was de-aerated with nitrogen

for 20 min. During experiments circulation of the nitrogen in the electrochemical cell

was maintained. After polymerization, samples were rinsed with water and dried in

a nitrogen flow. Pyrrole monomer, sodium tosylate salt and 4-toluol sulfonate acid

were purchased from Merck. Tetrabutylammonium tosylate salt was purchased from

Fluka. All reagents were used without further purification. Solutions were prepared

using de-ionized water with resistivity of below 1 µS. The polypyrrole layers which

were obtained under potentiostatic or the potentiodynamic control showed a signifi-

cant variation of the color at different parts of the sample. At the edges of sample the


E / V vs. SHE

0,86

0,88

0,9

0,92

0,94

0,96

t / s0 50 100 150 200

Figure 3.2: Potential during the deposition of the polypyrrole.

film was much darker than in the middle. This indicated differences in the thickness

of the PPy layer. The polypyrrole started to grow at the edges of the sample which

acted as specific geometric nucleation centers. It was found that the homogeneity of

the layer could be improved using the galvanostatic control. Figure 3.2 shows the po-

larization potential which was recorded during the deposition of the polypyrrole layer

using the galvanostatic mode at a current density of 0.3 mA/cm2. Due to the small

number of nucleation centers at the beginning of the experiment (mostly edges of the

sample) the polarization potential had to rise up to 960mV. At this potential other

nucleation centers were created also in the other parts of the sample. This resulted in

the subsequent drop of the potential. Finally the polarization potential stabilized at

905 mV. This indicates that the the whole sample was coated with a thin layer of the

polymer. Hence, the following growth of the polymer could proceeded homogeneously

in all parts of the sample. Further polymerization resulted with a very slight increase

of the polarization potential. This is caused by the increasing resistivity of the poly-

mer layer due to its enlarging thickness. Also a decrease of the content of the pyrrole

monomer in the solution might contribute to that.

3.2. Characterization of the polypyrrole layers 41

3.2 Characterization of the polypyrrole layers

3.2.1 Electrochemical characterization of the PPy

One of the most used methods for the investigation of the electrochemical proper-

ties of conductive polymers is cyclic voltammometry. Many fruitful information can

be obtained from voltammometric curves, such as: nature of the ion exchange dur-

ing the reduction/oxidation cycles [85], redox capacity [86], properties of the polymer

structure [87] and mobility of ions in the film [89]. In the following section the cyclic

voltammometry was utilized to check the properties of the obtained polymer film.

Reduction and oxidation of the polymer which happen during repetitive cycling of

the polymer change the structure of and the charge stored in the polymer matrix. In

order to maintain the electrical neutrality of the polymer the above processes have to be

accompanied by ion exchange between polymer and electrolyte, and hence by migration

of ions within the polymer. The reduction of the polymer might proceed according to

two scenarios: anions stored in the polymer can be released from the polymer matrix

to the solution or cations present in the electrolyte can be incorporated. Similarly the

oxidation of the polymer might proceed by an expulsion of former incorporated cations

or by an incorporation of anions. Above reactions are described by the equations:

PPy+A− + e− → PPy0 + A−, anion expulsion (3.1)

PPy+A− + e− + K+ → PPy0A−K+, cation incorporation (3.2)

Additionally both reactions may occur in parallel i.e. a partial incorporation and a

partial expulsion of ions [91]. The type of the ion exchange depends on the mobility of

ions present in the polymer and in the electrolyte. If the supporting electrolyte consists

of large and immobile cations than migration of anions will be promoted. But if the

electrolyte contains small cations, e.g. sodium or potassium, than cation incorporation

may take place, even if the doping anion is also small and mobile. The passive corrosion

protection provided by paints is based on the inhibition of ion migration and diffusion

between local cathodes and anodes, i.e. a low ionic conductivity of this coatings is

desired. In the case of a conductive polymer (corrosion inhibiting) anions present in

the polymer matrix might play a protective role but incorporation of cations is to

be expected to decrease the protective properties of the coating. Therefore special

attention has to be payed during the cycling of the conductive polymer in order to


(a) (b)

Figure 3.3: Structural formula of: a) tetrabutylammonium cation, b) tosylate anion.

detect possible incorporation of cations. When cation incorporation was to be avoided,

e.g. in order to investigate the anion exchange without influence of cation exchange,

tetrabutylammonium tosylate salt was chosen for the cyclovoltammetic measurements.

As shown in figure fig. 3.3 the large tetrabutylammonium cation consists of four carbon

branches, thus it is large and hence very immobile and cannot be incorporated into

the polymer matrix during the re-oxidation of the polymer [88]. Hence the whole ion

exchange can be realized only by the incorporation and expulsion of the tosylate anion.

The concentration of the supporting electrolyte was 0.1 M. The pH of the solution was

adjusted to 4 using the corresponding acid. The open circuit potential (OPC) which

was measured just before the cycling was chosen as a starting potential for cycling. The

polarization potential was decreased by -1100 mV from the starting potential. When

the applied potential achieved one of the border values of the scan range the direction

of the scanning was changed to opposite one. Cycling was performed with a scanning

speed of 10 mV/s.

Figure 3.4 shows a curve obtained after cycling of an 1 µm thick polypyrrole film. The

first cycle showed a large cathodic current peak which is located at the potential of

-70 mV and a current plateau located at -600 mV. According to Li the doping anions

can be placed between the PPy plain layers or within a PPy layer [85, 92]. In absence

of small cations which could be incorporated during the first reduction peak, the peak

indicates the release of anions from the sites located between PPy layers. The current

plateau at more negative potential points a slow expulsion of anions form sites within

the PPy layer. As it can be seen in fig. 3.4 the shape of the voltammogram for the

next cycles differs significantly form the first one. A stable cyclovoltammogram was


achieved after 3 cycles. This effect is well known for PPy cycled in organic and water

electrolytes [85, 86, 88, 90]. However, in organic solutions it is even more pronounced.

Heinze et. al. observed that during the electropolymerization a special structure of

counter-anions and the solvent in the polypyrrole develops. This structure is irreversible

changed after the first reduction [93]. Wernet and Wegner suggested that the layered

structure of the polypyrrole disappears after the cycling [94]. Beck et. al. suggested

that the degree of the disorder of a PPy film increased due to the first reduction [95].

Despite of all published hypotheses the origin of this effect is still not clear. In the case

of the steady state cycles the first reduction peak appears at the potential of 60 mV

which is more positive comparing to the case of the first cycle. The peak is wider,

which suggest a more sluggish release of ions. The current plateau does not occur in

the second reduction cycle. This can be an effect of the above mentioned changes in the

structure of the PPy. On basis of the obtained curves the ratio of the charge released in

�/�/�

��

��

�

��

��

�/�/�/��/��

��

Figure 3.4: Cyclic voltammogram of the PPy layer performed in 0.1M tetrabutylam-

monium tosylate solution at scanning speed of 10mV/s

the first reduction cycle QE(0) to the charge released in the steady state discharge QE

can be calculated. The QE(0)/QE ratio equals to 7.72. This indicates that only a small

amount of the anions which were expelled during the reduction can be incorporated

again into the polymer matrix during its reoxidation. Another parameter describing


the redox properties of the polymer is the redox capacity which is a measure for the

mole ratio of charge per monomer and can be calculated by the following equations [97]:

y0 = 2QE(0)/γ(Q0 − QE(0)), for the first reduction (3.3)

y = 2QE/γ(Q0 − QE), for the steady state reduction (3.4)

Assuming a current efficiency of γ = 100% for the electropolymerization (charge Q(0)),

redox capacities range from y0 = 5.96 · 10−4 and y = 8 · 10−5. The loss of the redox

capacity can be caused by dificulties in the incorporation of tosylate anions after their

first expulsion.

In order to investigate the behavior of the PPy film under conditions similar to real

corrosion conditions, the PPy layer was cycled in solution conting small ions which

could be incorporated during the cycling. As an electrolyte the 1 M KCL solution

was used. Figure 3.5 shows the cyclovoltammogram obtained for an 1 µm thick layer.

Similarly to the former case the first cycle differs significantly from the next ones. The

large reduction peak appears at -250 mV. It is accompanied by the shoulder at the

negative potential site. During the following cycles its position shifts gradually in the

negative direction and finally it stabilizes at the potential of -325 mV. The shoulder

visible on the first cycle disappeared completely. The observed reduction and oxidation

currents are much higher as in the case when the PPy was cycled in the presence of

tetrabutylammonium cations and the tosylate anions. This indicates that the small

potassium cations and chloride anions might be easily incorporated and released form

the polymer. The gradual evolution of the cyclovoltammogram shape suggests that

during the cycling a progressive exchange of anions and cations takes place until the

equilibrium between incorporated tosylate, chloride and sodium ions is achieved. The

redox capacity calculated for the first reduction cycle equals to y0 = 1.4 · 10−3 while in

the case of the steady state reduction it equals to y = 3·10−3. The increase of the redox

capacity after few cycles can be attributed to the partial exchange of immobile tosylate

anions by mobile chloride anions. Every PPy sample used in corrosion experiments was

cycled in the 0.1 M tetrabutylammoniu tosylate solution after the preparation. Thus,

the amount of the active charge centers in the polypyrrole coatings was known for all

individual experiments.


�/�/�

��

��

��

��

��

�

��

��

�/�/�/��/��

��

Figure 3.5: Cyclic voltammogram of the PPy layer performed in 0.1M NaCl solution

at speed of 10mV/s

3.2.2 XPS analysis of the obtained PPy layers

The first commercial XPS instrument was deliver to the market by Hewlett-Packard in

1954. Since that time the XPS became a common method for the analysis of the surface

composition of metals, non-conductive polymers and semiconductors. In 1983 Plfuger

and coworkers [98] showed that it can be successfully applied for the characterization of

conductive polymers, in particular polypyrrole. The XPS method allows to determine

many interesting parameters of PPy such as the doping level expressed as N+/N or

S/N ratios [99], degree of branching and crosslinking structure [100–102], valence band

structure [98] and the structural changes caused by chemical modification [101]. XPS

is considered a non-destructive method, however there are publications which report

that the polypyrrole or complex doping anions decompose due to the X radiation [103].

But these effects appear after long time exposure to the X radiation, only. Hence

XPS used for short measurement time can be successfully applied for the polymer

characterization. The following section will aim on the measurement of the doping

level and the characterization of structure of the obtained polypyrrole layer using the

XPS method.


Figure 3.6: A typical XPS survey spectrum of the polypyrrole layer after cycling in 0.1

tetrabutylammonium tosylate solution.

The XPS measurements were performed using a PHI Quantum 2000. As X-ray source,

the monochromatic Al Kα with photon energy of 1486.6 was employed. The X-ray

power was 23.3 W and the diameter of the X-ray beam was 100 µm. In order to

avoid possible charging of the specimen the neutralizer was applied with a current

of 10 µA and voltage of 1.0 V. Samples were mounted on the sample holder using

metal screws which ensured an electric contact between the sample holder and the

polymer. Measurements were started when the pressure in the chamber reached level

of 1·10−8 mbar. The sample stage was aligned to be perpendicular to the incident

X-ray beam and to be at 45 angle to the analyzer.

Prior to the XPS measurement the polyprrole sample was cycled in 0.1 M solution of

the tetrabutylammonium tosylate salt as described in the former section. In order to

verify the composition of the obtained PPy and to localize possible impurities at first,

always a survey measurement was performed. Four sweeps in the range of 0 – 1200 eV

were done. The scanning interval was 0.4 eV. Figure 3.6 shows a typical XPS spectrum.

As it could be expected, four peaks can be recognized on the obtained spectrum. They

can be attributed to: S 2p, S 1s, C 1s, N 1s and O 1s in the increasing order of the

binding energy. The two last peaks can be attributed to O and N KLL Auger lines. The

absence of the sodium signal indicates that all residuals of the electrolyte which was


used for the polymerization were removed either by cleaning after the polymerization

or by further cycling in the tetrabutylammonium tosylate solution.

(a) (b)

(c) (d)

Figure 3.7: Deconvolution of selected peaks obtained from detailed XPS measurement,

a) C 1s peak, b) N 1s peak, c) O 1s peak, d) S 2p peak. Details on the assignment of

the different species: see text.

A detailed analysis of the relative concentration of elements, doping level and the

structure of the polypyrrole was performed on basis of separate measurements for all

peaks. The C 1s, O 1s, N 1s and S 2p XPS spectra were collected with a resolution

of 0.1 eV. The shape of smaller peaks (e.g. sulfur) could be significantly deformed

by high noise levels due to low signal to noise ratio. Therefore a various number of

measurement sweeps were taken for different elements. For the carbon peak which is

the largest one only 6 sweeps were collected. The N 1s and O 1s peak were measured


with 7 and 8 sweeps respectively. For the S 2p peak which is the smallest one (and its

shape can be significantly influenced by the noise) 16 sweeps were taken. The amount

of the investigated elements can be directly obtained form the area of the appropriate

peak. However measured areas have to be corrected by the sensitivity factors which

are characteristic for the used detector. According to the ESCA 2000 manufacturer

the following factors were applied: 0.314 for C 1s, 0.499 for N 1s, 0.733 for O 1s and

0.717 for S 2p [104]. The shift in binding energies of C 1s, O 1s, N 1s and S 2p peaks

was corrected by setting the main C 1s peak at 284.6 eV [105]. The background was

subtracted using the Shirley method. All spectra were deconvoluted into Gaussian-

Lorentz components. The full widths at half-maximum (FWHM) were kept the same

for all components in any particular envelope.

Figure 3.7 shows the detailed spectra for investigated peaks. The deconvolution of the

C 1s spectra (see fig. 3.7a) led to four components with relative ratios of 1 : 0.22 : 0.07 : 0.03

in increasing order of BE. The main peak at 284.6 eV (corrected from the raw data:

285.1 eV) can be attributed to the β-carbons (C–C) in the PPy ring [103] and the

tosylate anion [106]. The second peak located at 286.3 eV is due to the C-OH bond-

ing [106], C=N [107] and =C–NH•+ (polaron) defects [106]. Also C-S bonds from the

tosylate anion contribute to this signal [108, 109]. The third peak at 288 eV is caused

by C=O and –C=N+ (bipolaron) defects [108]. The last peak at 290.9 eV can be

attributed to the π − π∗ satellite in the PPy ring [103].

In the case of the O 1s spectrum (see fig. 3.7c) two peaks can be found. The oxygen

peak consists of two components at 531.3 eV and 532.9 eV with relative intensities of

1 : 0.22 respectively. The first one can be assigned to the sulfate group C–S in tosylate

anion and C=O groups in PPy [110]. The other one is due to the C–OH groups and

bound water [106, 111].

Turning to the sulfur line, the S 2p doublet is visible at 167.6 eV and 168.9 eV (see

fig. 3.7d). This is assigned to oxidized form of sulfur (SIV ) of the R–SO3 anion [110].

The ratio of the relative intensities is 1 : 0.61. The N1s peak consists of four components

with relative ratios of 0.11 : 1 : 0.2 : 0.12 in increasing order of binding energy. The main

peak which is present at 399.8 eV represents the neutral N in the pyrrole ring (–NH–).

According to some authors [112–114] the shoulder at lower energy of 397.9 eV comes

form imine-like nitrogen =N– and may be considered as a defect in the polymer [115].

Two additional peaks can be recognized at higher energies of 401.2 eV and 403 eV.


C 1s O 1s

Peak position /eV 284.6 286.3 288 290.9 531.1 532.7

Peak composition/% 74.7 16.7 5.9 2.7 81.4 18.6

Chemical composition /% 58.2 11.6 3.2 1.1 8.9 2.0

N 1s S 2p

Peak position /eV 399.7 397.6 401.1 402.4 167.6 168.9

Envelope conc. /% 69.5 8.1 14.1 8.3 62 38

Spectrum conc. /% 8.5 1.1 1.6 0.9 1.8 1.1

Table 3.1: Relative intensities of components of each envelope expressed as a percentage

of the total signal.

The first one is caused by polaron charge carrier species –NH•+ [116]. The other one is

related to bipolaron carrier species –NH+. The obtained XPS results allow to determine

the doping level in terms of ratios of S/N which equals to 0.24. This value can be

confirmed by the Nδ+/N. Considering that the Nδ+ was taken as a sum of polaron and

bipolaron signal the Nδ+/N ratio equaled to 0.22. That means that there is one doping

anion per 4 to 5 pyrrole monomers which is in good agreement with published data

where the doping level vary in the range 0.2–0.4 [115]. The slight difference between

both ratios suggests that not all positive charges are compensated by tosylate anions.

The O531/Stotal ratio equals to 3.1 which means that almost all oxygen attributed to

the binding energy of 531 eV comes form the sulfate group of the tosylate anion, as it

should ideally be the case. The small amount of excess oxygen can be attributed to

C=O bonds. The chemical composition of polypyrrole can be verified using calculated

values of Ctotal/Stotal, Ctotal/Ntotal and Ctotal/Ototal. The evaluation of above ratios

was performed according to the general formula C4H3N(C7H7O3S)n assuming that n

is a doping level expressed as Stotal/Ntotal ratio. Measured and calculated values are

presented in table 3.2. In all cases the amount of measured carbon slightly exceeds

the theoretical value. This can be due to some tetrabutylammonium residuals present

on the polymer surface. The tetrabutylammonium molecule consists of 16 atoms of

carbon therefore even very small amount of this molecules might significantly increase

the carbon signal.


Ratios Ctotal/Stotal O531/Stotal Stotal /Ntotal N+/Ntotal Ctotal/O531 Ctotal/Ntotal

Measured 25.8 3.1 0.24 0.22 8.3 6.1

Theor. 23.8 3 — — 8 5.7

Table 3.2: Ratios of measured elements. Theoretical values were calculated from the

formula C4H3N(C7H7O3S)n assuming that doping level equals to Stotal/Ntotal.

3.2.3 Morphology of the PPy film

The morphology of the PPy layers was investigated using images obtained by Scanning

Electron Microscope (SEM) and Atomic Force Microscope (AFM). Figure 3.8 shows

two images of PPy samples made using SEM. The first image (fig. 3.8a represents the

surface of a PPy film with a thickness of 58 nm. The other figure shows the image (fig.

3.8b) of an 1µm thick PPy layer.

(a) (b)

Figure 3.8: Pictures of the PPy layers obtained using Scanning Electron Microscope,

a) thin layer b) thick layer

In both cases a nodular structure of the surface is to be seen. This type of the structure

is typical for this polymer as is confirmed in the literature [117–120]. However, the size

of the PPy aggregates differs significantly. For the 58 nm thick sample the diameter

of aggregates vary in the range 50–100 nm while for the thicker PPy layer it can reach

even 500 nm.


Figure 3.9: The AFM image of the PPy sample, thickncess of the PPy layer of 1µm.

The AFM image measured on a sample with the thickness of 90 nm confirm the results

obtained by SEM (see fig. 3.9). The size of the PPy aggregates varies in the range 100–

150 nm. The average roughness of the PPy layer was calculated using the Nanoscope

software and it equals to 1.757 nm.

Chapter 4

The passivation of the defect by

ICP.

As proposed by Mendoli in the early 80’s, conductive polymers could be utilized for

the corrosion protection. His idea was based on the assumption that the conductive

polymer could provide an anodic galvanic protection, acting as an oxidizer to main-

tain the metal in the passivity domain and prohibit its dissolution. At first this so

called “ennobling mechanism” got many supporters among the corrosion scientists.

However, later more and more critical opinions about the usability of conductive poly-

mers for corrosion protection appeared. Till today this subject is still a matter of a

dispute. Many contradictory results concerning performance of conductive polymers

were published recently. These inconsistent results are most likely due to the fact that

most experiments were performed under very special conditions. Some publications

report that this mechanism works, others report that this is only true in chloride free

solutions [71–73, 84]. Others claim that the ennobling mechanism also improves the

passivity of the oxide layer at the metal/polymer interface through oxidation by the

applied conducting redox polymer, thusly inhibiting electrochemically driven delamina-

tion (see e.g. [73]). However, for example for iron or steel the oxide layer at the buried

interface is passive anyway at the high pH prevailing during delamination [23–25, 74].

Hence the further passivation of the metal surface will not have any significant influence

to the delamination process, which was also shown by McMurray et.al. [121]. Hence,

just the ennobling of the interface between the metal and the polymer can not provide

sufficient protection against delamination. Nevertheless, if the effect of the ennobling

54 4. The passivation of the defect by ICP.

of the metal is not limited only to the metal/polymer interface but extends to areas of

the metal which are uncoated (e.g defects) than possible passivation of these defects

could significantly inhibit corrosion.

Therefore the following chapter will focus on an investigation of possible corrosion

inhibition by the ennobling of defects.

4.1 Ennobling of the defect by ICP’s

When the metal is coated by a conductive polymer the delamination process initialized

by the corroding metal in a defect down to the metal might be inhibited by the metal

passivation due to electrochemical coupling with the conducting polymer. The idea of

the ennobling of the defect by the conductive polymer is presented in fig. 4.1. The

model sample used here for the investigation of the defect is similar to samples used

for the typical delamination test. It consists of the metal substrate (e.g. iron) which is

coated with a polymer. However in this case it will be a conductive polymer. At one side

of the sample the coating was removed, thus the defect was created. Then the defect

was filled with an electrolyte which ensured a galvanic coupling between the conductive

polymer and the metal at the defect. Shortly after injection of the electrolyte to the

defect corrosion of the metal starts. The dissolution of the metal entails the cathodic

shift of the potential which initializes the reduction of the conductive polymer. Thusly

the positive charge stored in the polymer matrix can be transfered to the defect. Figure

4.2 shows the electrochemical behavior of an iron electrode under different potentials.

There are two regions marked in the figure. The first one is pointed by a red rectangle

Figure 4.1: The idea of the defect protection by the conductive polymer

4.1. Ennobling of the defect by ICP’s 55

Figure 4.2: Electrochemical behavior of the iron electrode.

which indicates the range of potential where the active dissolution of an iron takes

place. The other region marked by a green rectangle points the zone where the surface

of the iron electrode is passivated. In this region the iron oxide which is present on the

surface constitutes a protective layer which inhibits dissolution of the metal. Therefore

in order to protect the iron against a dissolution its potential has to be kept in this

passivity domain. This can be done by supplying an anodic current which will shift the

potential of the iron electrode from the dissolution region over the critical passivating

current density into the passivity domain.

This current can be provided by the conductive polymer as a consequence of its reduc-

tion:

Py+ + e− → Py (4.1)

The kinetics of the reduction of the polymer might decide about success or failure of

the protection. As it is shown in figure 4.2 both regions are separated by a peak of the

current. Hence, the current density which is necessary for the passivation of the iron

surface has to be larger than the critical current density (indicated in the figure). That

means that during the reduction of the conductive polymer the charge has to be released

quickly enough so sufficient current could be created. Once the metal surface has been

successfully passivated, depending on the environment a smaller or larger current will

be necessary to maintain passivation. Therefore the total charge which can be released

from the polymer will decide about the durability of the protection. Additionally it

has to be noticed that only that part of the polymer which has a galvanic coupling to

the defect will contribute to the passivation process. This active part of the polymer is

placed just in the neigbourhood of the defect and it is marked in figure 4.1 by a light


Figure 4.3: Protection of pin holes by ICP induced passivation.

green color. Hence the amount of the charge potentially available for the oxidation of

the metal is limited. This means that in the case of large defects the active part of the

polymer might be to small and it could not be able to supply a sufficient current for

the passivation. Therefore there is a higher probability that proposed mechanism will

work in the case of pinholes rather than in the case of large defect. This is schematicaly

presented in fig. 4.3. A relatively large ratio of active polymer (galvanically coupled to

the defect) to defect area gives a chance that supplied current will be large enough to

passivate the defect. The stability of the protective oxide also depends on the type of

the electrolyte in the defect. Very aggressive species such as chloride might attack the

passive layer and lead to a higher passive current density and hence to the brake down

of the protection. In the following section the effect of the ennobling of the defect in

the chloride containing and chloride free electrolytes will be examined.

4.1.1 Experimental setup

.

The following experiment was focused on an observation of the passivation of the de-

fect by the conductive polymer. A special attention was adressed to the observation

of the open circuit potential (OCP) potentials of the defect and the polymer and the

current which flowed between them. In the case of the delamination experiment a pre-

cise measurement of the potential over the polymer is not possible using electochemical

methods. Due to a small volume of the electrolyte and a short distance between the

defect and the polymer the measurement of the potential over the polymer (even using

microelectrodes) will always be affected by the influence of the potential of the defect.


Figure 4.4: The experimental setup for the investiagtion of the “ennobling effect“.

Additionally defect and polymer both are situated on the same sample. Therefore the

measurement of the electric current which flows between them is not possible. Hence,

the exact measurement of both potentials and the current demanded a special experi-

mental setup which is shown in figure 4.4. The defect and the polymer were placed on

separate substrates and immeresed in the electrolyte. The electrical contact between

them was realized by an additional wiring. That allowed an independent measurement

of both OCPs. Also the current flowing between the defect and the polymer could

be recorded. Another advantage of such setup is that using the switch the defect and

polymer can be connected at a certain moment which is convenient for the observer.

Hence, the electrochemical behavior of defect and the polymer could be observed at

the very moment when the galvanic coupling was established. The defect was simu-

lated by an iron wire with a diameter of 0.5 mm, sealed in epoxy resin and polished

with grinding paper up to 1000 (the resulting area of the defect was about 0.2mm2).

As a substrate for the conductive polymer a gold layer evaporated on the glass plate

was used. This excluded any electrochemical interaction between the polymer and the

substrate. Therefore the reduction of the polymer could be caused only by the iron

dissolution. Polypyrrole was chosen as the conductive polymer for the experiment. Its

open circuit potential (OPC) is about +400 mV vs. SHE. That potential should be

high enough to protect iron against corrosion. The detailed procedure of the polymer-


ization was described in the chapter 3. The area of the polymer which covered the

gold substrate was 4 cm2. The defect was connected to the polymer by closing the

switch, while the open cell potentials of the iron and the polypyrrole were monitored

using two separate Ag/AgCl reference electrodes. The current which flows between the

polypyyrole and the corroding iron was measured by an amperometer. Experiments

were performed in two different electrolytes: 0.1M K2SO4 solution with pH adjusted

to 4 and 0.1M KCl with native pH of 6. At the beginning of the experiment the switch

was set in a position “off“ which separated the defect from the polymer. When the

iron defect was immersed in the electrolyte its CP started to decrease indicating the

beginning of the corrosion process. When the potential of iron reached a value of -450

mV vs. SHE indicating an active corrosion the switch was turned on and the electrical

connection between the polymer and defect was established.

4.1.2 Influence of the type of the electrolyte for the passiva-

tion

Figure 4.5a shows the results of the passivation experiment obtained in the K2SO4

solution. Shortly after immersion of the iron into the electrolyte the defect reached a

corrosion potential of -450 mV vs. SHE. Then the iron and the polymer were connected,

which resulted the polypyrrole pulling up the potential of the small corroding iron

electrode to the potential of the polypyrrole, where the iron is passive. The small

current continuing to flow between the polymer and the iron during its passive phase

indicates that the a certain current density is required to keep it passive, i.e. to

continuously re-passivate active sites.

Obviously, the continuous formation of small active sites in the passive oxide layer on

the iron requires steady repassivation currents provided by reduction of the polypyrrole.

Over the time, this causes a steady decrease in the potential of the polypyrrole, signaling

its steady discharge, i.e. its oxidation power being consumed, and a steady decrease

in the potential of the iron electrode, signaling its becoming less and less passive,

also obvious from the slightly increasing current flow between polypyrrole and iron.

Increase of the current with time accompanies the decay of the defect potential. After

400 min, when the potential decreased to 200 mV vs. SHE , the oxidation power of

the polypyrrole breaks down and a sudden and massive discharge of the polymer takes

place, caused by the reactivation of the defect. The potential of the defect reached


stable value of -150 mV vs. SHE which indicates an actively corroding iron electrode.

Now the (partly) reduced polypyrrole acts as an additional site for oxygen reduction,

i.e. now due to the high cathodic currents an increased iron corrosion takes place.

Figure 4.5b shows the same experiment performed in chloride containing solution. In

this case obviously the iron electrode cannot be polarized even far enough for even

just a seemingly passivation, as the passive layer is destroyed very fast as a result of

strong attack by the chloride ions. The polypyrrole layer was not able to maintain the

iron defect even near the passive state and the polypyrrole coating was fully discharged

almost immediately after connection with the defect.

4.1.3 Recharging of the polymer during wet/dry cycles

In figure 4.6a the results of a series of consecutive connections and disconnections of

the iron electrode with the polypyrrole coating are presented. In this experiment an

iron wire with diameter of 0.065 mm and length of 5 mm (resulting in an area of 1mm2)

was used for simulating the defect. Initially, about 8 min after connection of the defect

and the polypyrrole a sharp current peak marked the break-down of the protection

(the defect is much larger than in the one used for the experiments shown in fig. 4.5).

After the break down, the iron electrode and the polypyrrole were disconnected from

each other. Then the open cell potential of the polypyrrole increased until it reached

E /

V v

s. S

HE

�0.6

�0.4

�0.2

0

0.2

0.4

0.6

I / mA

0

0.05

0.1

0.15

0.2

0.25

t / min0 100 200 300 400 500

PotentialCurrent

(a)

E /

V v

s. S

HE

�0.4

�0.3

�0.2

�0.1

0

0.1

0.2

0.3

0.4

I / mA

0

0.05

0.1

0.15

0.2

0.25

t / s0 20 40 60 80 100

PotentialCurrent

(b)

Figure 4.5: Immersion test in: a) K2SO4 at pH4 b) chloride containing electrolyte at

pH6


the starting value (fig. 4.6b), i.e. the ocp of fully oxidized polypyrrole. As no other

oxidant is present in these experiments, this indicates that the reduced polypyrrole was

fully re-oxidized by atmospheric oxygen.

E /

V v

s. A

g/A

gCl

�0.6

�0.4

�0.2

0

0.2

I / mA

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

t / min0 50 100 150

(a)

E /

V v

s. A

g/A

gCl

�0.6

�0.4

�0.2

0

0.2

X Axis Title0 4 8

t / min40 44 48

t / min72 76 80

X Axis Title108 112

I / mA

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

X Axis Title180 184 188

(b)

Figure 4.6: Cyclic discharging and recharging of the polypyrrole by connecting and

disconnecting it with the iron, lower graph: enlarged snapshot showing details of the

iron protection and break-down. Red curve shows the galvanic current, blue one the

potential of the polypyrrole, green one of the iron.

Even after many consecutive discharge and recharge cycles, the polypyrrole layer

showed a reproducible ability to recharge itself after every discharge caused by the


E / V vs.

Ag

/Ag

Cl

-1

-0,8

-0,6

-0,4

-0,2

0

0,2

0,4

t / s0 500 1 000 1 500 2 000 2 500 3 000 3 500 4 000

Figure 4.7: Galvanostatic discharge of the polypyrrole layer in 1M KCl solution at

the current density of 1 µA/cm. The polymer thickness was 1 µm. In the case of

contact with iron, the current does not remain constant, but increases drastically at

the potential when the iron becomes active, i.e. the break down point will be at a

different potential.

connection to the actively corroding iron. These results show that for very small

pinholes and under cyclic climatic conditions with many wet/dry (corresponding to

active/passive) transitions a polypyrrole based coating might provide real long-term

protection. The polypyrrole layer is capable of passivating the small pin-hole defects

for a certain time, depending on the ratio of active polymer to active iron areas. The

low currents required for keeping the iron passive cause a steady decrease in potential

of the polypyrrole (see fig. 4.7), as the ratio of oxidized to reduced states in the poly-

mer decreases. According to the Nernst equation, this should occur by logarithm of

this ratio. Hence, despite of the initially small currents potential changes quite quickly

at the beginning, then the decrease slows down a bit until at a certain discharge value

it suddenly breaks down. This breakdown is shown in fig. 4.7, where the development

of potential of the polypyrrole during galvanostatic reduction at a current density of 1

µm is shown. Coupled with a defect which becomes active below a certain potential,

the break down can start earlier. But if the conditions change from active to passive

before this critical break down occurs, the polypyrrole may be recharged by the at-

mospheric oxygen and when the conditions change again it is again capable to keep


the pin-holes passive and no corrosion will occur. However, while conducting redox

polymers may show a certain capability to inhibit corrosion in small pinhole sized de-

fects in the coating, especially in the immersed case where the whole coating area is

active and can provide the necessary current densities required for the initial defect

passivation, the performance is completely different in the presence of a large defect,

especially in the case of a non-immersed sample with an electrolyte covered defect,

typical for atmospheric corrosion conditions. Here, only the coating in the vicinity of

the defect is actively involved in the reaction. Obviously, only for the case of extremely

small pinholes a current density sufficient for passivation may be reached. For larger

defects, quick reduction of the polymer and its delamination is anticipated, as will be

discussed in more detail later.

4.2 Conclusion

The results presented here show that conducting polymer coatings may provide an

efficient passivation of small defects, especially in the immersed state where the whole

coating can contribute to the passivation currents. For the non-immersed state only

extremely small defects may be passivated, but in the case of cyclic wet/dry conditions

a chance of recharging of the coating should exist, which could provide a real long-

term protection. However, the presence of chloride ions in the solution is detrimental

already for quite small defects. In the presence of larger defects, the coating is not

capable to passivate it and quick coating reduction is observed. Since this reduction

proceeds faster than the typical delamination observed for most non-conductive organic

coatings, this means that in this case conductive coatings are clearly of disadvantage.

Chapter 5

The shift of the oxygen reduction

site

5.1 Introduction

Oxygen reduction at the polymer/metal interface plays a key role in the delamination

of non-conductive coatings. During oxygen reduction aggressive species such as radicals

and OH− are formed. A high concentration of these aggressive species is a main reason

which leads to a disbondment of the coating. Therefore, most anti-corrosion strategies

are aimed to eliminate the oxygen reduction, e.g. by insulating conversion layers, or

at least to reduce consequences of the high concentration of aggressive species at the

interface, e.g. by use of superior coatings. As it was shown in the former chapter

passivation of the defect is also an effective way (indeed the preferable way). The

application of conductive polymers as a coating opens new possibilities of protection

which do not exist for insulating coatings. One of them was proposed by Kinlen [69]

and it is schematically presented in fig. 5.1. The inhibition of the corrosion might be

possible thanks to the conductivity of ICPs. Due to the conductivity of the ICP coating

the site of oxygen reduction could be dislocated from the interface metal/polymer (fig.

5.1a) into the bulk of the polymer, i.e. be smeared out (fig. 5.1b). The thusly reduced

local intensity of aggressive species should lead to a longer lifetime of the system.

Of course radicals created inside the polymer will attack and finally destroy the poly-

mer matrix. But due to the large volume of the polymer this process should take much

64 5. The shift of the oxygen reduction site

(a) (b) (c)

Figure 5.1: The idea of the dislocation of the site of the oxygen reduction, a) delami-

antion of the insulating or slightly conductive coating. Reduction of the oxygen at

the metal-polymer interface leads to the delaminaiton of the coating, b) smearing out

of the oxygen reduction. The large volume of the polymer allow to neutralize large

amount of radicals and OH−, c) highly conductive coating. The site of the oxygne

reduction was shifted to the ICP/Top-Coat interface and results with the delamination

of the Top-Coat.

more time compared to the time which is necessary for the delamination of the insu-

lating coating. Hence, the degradation of the interface should be significantly slowed

down (fig.5.1b). However, the whole success of the so called “smearing out” mechanism

relies on the assumption that electrons will be consumed inside the polymer. If not

than another scenario which is presented in fig.5.1c could happen. Electrons could be

transferred across the conductive polymer to its surface. Then the oxygen reduction

which will appear on the surface of the polymer and a high concentration of radicals

on the top of the conductive coating will be created. Considering the fact that ICP

are supposed to be used as a primer coating and should be covered with the top coat

the delamination of the top coat from the conductive primer would occur. Hence, the

position of the oxygen reduction would have to be controlled in a very exact manner. It

can be assumed that the location of the reduction of the oxygen depends on the conduc-

tivity of the polymer. For instance in the case of the highly conductive polymer a very

high mobility of holes (ICPs are p-conductive) will assure fast transport of the charge

through the polymer. It might happen that oxygen reduction will occur predominantly

on the top of the ICP film. Hence, the targeted effect seems to be obtainable rather

5.2. Experimental setup 65

for less conductive polymers. Low conductivity should make it more likely that the

oxygen reduction occurs inside the polymer. On the other side too low conductivity

of the polymer will result in the ordinary delamination, as oxygen reduction will take

place at the internal metal/polymer interface as for a non-conducting coating. Hence,

the conductivity of the coating has to fit into a certain range. In order to delimit this

range a method for the exact localization of the oxygen reduction site is necessary. In

the following section a method will be presented which allows to localize the position

of this reaction and an investigation of the delamination of the polypyrrole layer.

5.2 Experimental setup

The aim of the following experiment is to find out where the oxygen reduction happens

during the delamination of the conductive coating. The OH− anions which are created

during the reduction of the oxygen can be used as a indicator of the reaction. Therefore

the depth profile of the concentration of OH− anions inside the conductive polymer will

give an answer where the investigated reaction happens. This profile can be obtained

using the ToF-SIMS spectrometry. However, during ToF-SIMS a high intensity of OH−

anions can be created due to the fragmentation of residual water molecules. This is

caused by the gallium ions which bombard the polymer surface during the measure-

ment. Therefore, in order to obtain meaningful results the OH− created during the

measurement has to be distinguished from those ones created during the delamination.

This can be done by use of rare isotopes. When the delamination experiment is per-

formed in an atmosphere containing heavy oxygen (18O2) than 18OH− anions will be

created. These anion might be easily recognized by the ToF SIMS spectrometry.

In fig. 5.2 the experimental setup is presented. The sample consisted of a glass sub-

strate coated with a thin gold layer (c.a. 100nm). The polypyrrole layer was deposited

on a part of the substrate by means of electrochemical polymerization. The part of

the sample not coated by the polymer was destined for the defect. In order to obtain

a low potential in the defect small pieces of iron were placed in the this area. The

purity of iron was 0.9999%. After filling the defect with the electrolyte the corrosion

of the iron starts and its potential decreases, entailing a decrease of the potential of

the gold surface exposed at the defect edge. A thusly prepared sample was placed

inside a gas tight container and 18O2 was introduced. Then the 1M KCl solution was


Figure 5.2: Experimental set-up for investigating a possible dislocation of the site of

the oxygen reduction.

delivered to the defect through the small fithrough in the container. The corroding

iron pulled down the potential of the gold and initialized a delamination process. The

progress of the delamination was indicated by the change of the color of the polypyrrole

which could be seen even with the bare eye. When the delamination had progressed

c.a. 4mm the experiment was stopped. The container was opend and the electrolyte

removed from the defect. Subsequently the dry sample was meaured by the ToF-SIMS

sepectrometer.

5.3 Oxygen reduction dislocation

In figure 5.3 a result of the ToF-SIMS measurements is presented. The ToF SIMS

spectrometer allowed to perform a measurement over a square area with a size which

could be varied in the range from 25 x 25 µm up to 250 x 250 µm. The size of

100x100 µm was recognized as suitable for this measurement. It allowed to make a

mapping of the delaminated and not-delaminated regions simultaneously. Using this

opportunity two maps of the distribution of the 18O2 and 18OH− were obtained and

they are presented in figures 5.3a and 5.3b respectively. The brighter points indicate

a presence of investigated ions. Both maps look very similar. Two regions which

have different content of 18OH− and 18O2 anions can be recognized. Both regions

are separated by sharp border where the amount of measured ions rapidly decreases.

5.3. Oxygen reduction dislocation 67

(a) (b)

Figure 5.3: Results of the ToF SIMS measurements of the polypyrrole layer which was

delaminated in the 18O2 containing atmosphere, a) distribution of 18O2, b) distribution

of 18OH−.

The region which shows the larger amount of detected ions is attributed to the region

where the coating was delaminated. The other one represents the intact polymer. The

amount of measured 18O2 anions seems to be slightly larger than the amount of 18OH−

anions. This means that some of 18O2 atoms had to be adsorbed at the surface of the

polymer. Also some fragmentation of the 18OH− anions contributed to the increased

signal of 18O2.

These measurements seem to suggest that the reduction of the oxygen has occurred

mainly at the top surface of the polymer, i.e. that the conductivity of the applied

polypyrrole layer was already too high. However, it is still not clear if the entire

reduction of the oxygen was shifted to the top of the surface or if some part of the oxygen

was reduced inside the polymer layer. Therefore two depth profiles were performed for

both regions. However, the ToF-SIMS measurement is not a quantitative one i.e. the

intensity of ions does not always reflect their abundance on measured surface. Measured

ions have to be sputtered from the surface and then registered by the detector. Thus

the number of detected ions is influenced by the strength of the bonding to the surface

and their environment; this is the so-called matrix effect. A first step for taking this

68 5. The shift of the oxygen reduction siteCounts

0

50

100

150

200

250

300

d / nm0 10 20 30 40 50

18OH-

SO-2

(a)

Counts

0

500

1 000

1 500

2 000

2 500

3 000

3 500

d / nm0 10 20 30 40 50

18OH-

SO-2

(b)

Figure 5.4: ToF-SIMS depth profiles, a) not delaminated area, b) delaminated area.

into account is to normalise the signals to the signal of a reference ion. In this case

the signal for the SO−

2 ion was chosen as a reference measurement. The SO−

2 ion is

a fragment of a tosylate ion which was used as a dopant during the polymerization

of the polypyrrole. Considering the fact the this ion is not mobile its amount should

be the same in the delaminated and not-delaminated area. Results of the performed

profiling are shown in figure 5.4. The first profile was measured at the region where the

polymer was intact (5.4a). At the surface of the intact area the ratio of18OH−

SO−

2

equals

to 0.65 while in the case of the delaminated area the same ratio equals to 3.4. i.e.18OH− anions in the delaminated area seem to have a significantly higher signal. After

first few nm of sputtering the amount of the 18OH− decreased significantly. However,

it can be seen that the OH−/SO−

2 ratio remains higher for the reduced ppy than for

the intact one. That indicates that the oxygen reduction was mainly shifted from the

metal-polymer interface to the top surface of the polymer, although some bulk oxygen

reduction should have occurred, too.

5.4 Durability of the polypyrrole in the alkaline en-

vironment

The idea of the corrosion protection by the “smear out” effect is based on the assump-

tion that due to a low conductivity of the PPy the degradation of the bulk of the

conductive coating due to the OH− anions and radicals will progress slower than the

5.4. Durability of the polypyrrole in the alkaline environment 69

delamination of the insulating coating caused by the breakage of the metal-polymer

bonding would do. However, if the ICP easily degrades in alkaline environment, then

the whole benefit rising form the shifting of the oxygen reduction site from the in-

terface to the bulk will be lost. Therefore the stability of the conductive coating in

the alkaline environment is a key issue which will decide about the performance of

the protection. The following discussion will focus on a potential risk of the polymer

degradation caused by alkaline environment. The degradation of the polypyrrole in

the aqueous media is a widely studied subject [130, 133]. It manifests in the decrease

of the conductivity, adhesion and by swelling of the film. It might proceed in two

ways: the first one is an electrochemical proccess which consists in the anodic overox-

idation. The other one is purely chemical and it is caused by the nucleophilic attack

of water molecules and OH− anions to radical cationic sites. The mechanism of the

irreversible change of the polypyrrole structure due to its overoxidation was presented

by Beck [132]. He proposed the following scheme of the reaction:

Figure 5.5: Mechanism of the overoxidation of polypyrrole.

At first the polaron is oxidized to the bipolaron and then it reacts with nuclophiles

and a pyrrolinone is created. However the observed mass balance during the overoxi-

dation corresponds only by 70-90% of what is predicted according to the above model.

Chirstensen and Hamnet [133] suggested that it may be connected with the CO2 emmi-

sion caused by the polypyrrole degradation and its related swelling. They observed that

end groups of the polymer chain are initialy maximaly oxidized. When the swelling of

the film slows, the C=O groups starts to oxidize to CO2. At the end of the overoxidation

experiment conjugated segments of reduced length were still present in overoxidized

film. Also some parts of the coating still presented good adhesion to the substrate.

Therefore it is concluded that overoxidation takes place at a relativly small number

of locations. Considering above Christensen prosed the scheme of oxidation of the

polypyrrole chain which is shown in fig.5.6.

As observed Christensen and Hamnet [133] the above reaction appears at anodic po-

tentials. Polarization of the polypyrrole layer in 1M NaClO4 aqueous solution showed


Figure 5.6: Degradation of the polypyrole chain due to the overoxidation.

that at potentials higher than 0.7mV (vs.SHE) this overoxidation took place. Hence

under the natural environmental conditions when the polypyrrole is not polarized this

degradation should not occur. However, at such condition another chemical degrada-

tion process may take place. Results obtained by Wernet and furhter confirmed by

Beck [131] show that the substitution in the pyrrole rings happens in alkaline solution

under the open-cell conditions. The IR spectrum of the polypyrrole layer obtained after

15 min of the immersion in 0.1M NaOH showed no carbonyl band but very pronounced

bands which correspond to substituted products. The following equation presents the

scheme of the reaction of the nucleophile attack at radical sites in polypyrrole.

Figure 5.7: The scheme of the polypyrrole reduction in the alkaline enviroment without

polarization.

The above consideration shows that polypyrrole is sensitive to high pH and might degra-

date in such conditions. Even if the conductivity of the polypyrrole would be adjusted

in a way which assures the “smearing out” effect the quick loss of the conductivity

caused by nucleophile attack would shift the oxygen reduction back to polymer-metal

interface. Hence the delamination would be initilized. Figure 5.8 shows the picture

of the polypyrrole sample after the delamination experiment performed in humid air

atmosphere. The defect was placed at the left side of the sample, hence the delamina-

tion front progressed from the left edge toward the right side. The strong change of

5.5. Conclusion 71

Figure 5.8: Delamination of the polypyrrole coating in oxygen containing atmosphere.

The defect was place at the left site of the sample. Irreversible change of the PPy color

is visible in on the left side of the sample (PPy turned transparent). Intact PPy is

marked by the dark area on the right side of the sample.

the color of the polypyrrole coating is visible at places where the coating was delami-

nating. PPy reversibly changes its color during the reduction/oxidation performed in

the nitrogen atmosphere. This effect was observed during the cycling of the PPy layer

(see chapter 3), but in the case of delamination experiment the change of the color

is irreversible. The PPy layer remains transparent after delamination even when the

sample is not polarized any more. This indicates a permanent changes in the polymer

structure which are most likely due to nucleophile attack.

5.5 Conclusion

The mechanism of smearing out the oxygen reduction sites does not provide an advan-

tage for corrosion protection. The conductivity needs to be quite low to achieve such

a smearing out, but even then a significant reduction activity seems to take place on

the polypyrrole surface. Moreover, since the conductive polymer is quickly reduced by

the corrosion in the defect, thus becoming non-conductive (which is further enhanced

by the observed pH-induced degradation), the initial shift in the oxygen reduction site

while the polymer is still conductive is in fact of no importance. Since the coating is


delaminated where it is fully reduced, it is expected that the observed oxygen reduc-

tion at the surface of the conductive coating takes place only at very early stages of

its reduction, i.e. at the front where the delamination curves are the steepest. When

the coating is fully reduced, oxygen reduction should again take place at the metal-

polymer (sputter profiles were limited to the top most 100 nm and did not reach down

to the buried interface) and cause the observed delamination, which seems to directly

follow the coating reduction (see following chapter). The polypyrrole showed very weak

resistance to the alkaline conditions caused by the oxygen reduction. Hence, although

it was not tried to tailor the conductivity of the PPy in a way that a real smear out

would occur, it is assumed that due to the reduction and the consequent degradation

by high pH generally a quick break-down would occur in any case. The permantent

change of the color of the coating after the experiment indicates irreversible changes

in the polypyrrole structure.

Chapter 6

The ion transport through the

conductive coating

The investigations presented in the former chapters confirmed that conductive polymers

can provide some protection against corrosion according to some of the mechanisms

proposed in literature, but only under very special conditions. The efficiency of the

protection seems to be very sensitive to the conditions of the experiment, hence the

application of conductive polymers as an inhibitor of the corrosion in real systems which

are exposed to variable atmospheric conditions is questionable. One main problem that

was observed in all experiments discussed so far is that the ICP gets quickly reduced

when in contact with a actively corroding defect. Passivation of the defect just by the

oxidation power of the ICP itself will occur only for very small defects and even the it

usually will not last very long.

A very common way of increasing the resistance of coating to the corrosion consists in

adding corrosion inhibiting pigments to the paints and other organic coatings applied on

metallic surfaces. The most efficient pigments are those containing chromates (usually

in form of strontium chromate), but because of their toxic and carcinogenic nature

their use has to be progressively decreased. But nearly all powerful inhibitors may

have detrimental effects on environment, when released in substantial amounts. Since

in nearly all pigments the release of inhibitors depends upon leaching, coatings need

to be highly pigmented to ensure a sufficient presence of inhibitors over years, and,

of course, inhibitors are constantly released into the environment, even when they

are not needed, thusly presenting a permanent environmental problem. Hence, novel

74 6. The ion transport through the conductive coating

Figure 6.1: Schematic of the “intelligent self-healing coating”. The doping anions

stored in the polymer matrix are released from the coating due to the reduction of

the polymer caused by electrons produced at the corroding defect. Released anions

passivate the surface of the corroding metal.

approaches are desperately sought for. The new concept of the corrosion protection

which could be served by the ICP was proposed by Barisci et.al [135]. He proposed the

idea of an “intelligent self-healing coating” based on conductive redox polymers which

is schematically presented in figure 6.1.

During the polymerization of the conductive coating anions are incorporated to the

polymer matrix. They neutralize the positive charge located at the oxidized parts of

the polymeric chain and thus ensure the electroneutrality of the ICP. However, if an

active defect in the coating is created (i.e. a defect going through to the metal which

starts to corrode), the oxidized parts of the polymer are reduced by electrons released in

consequence of the corrosion of the defect. Then the polymer looses its positive charges,

hence the ions that provided charge neutrality of the polymer are not more necessary.

The electric current caused by moving electrons has to be accompanied by migration

of ions which closes the electrical circuit. This migration could be realized in two ways.

One possibility is that cations are incorporated from the defect, where they compensate

the negative charge of the remaining anions. The other option is that the anions present

in the polymer matrix are expelled from the polymer. This means anions expelled form

the polymer due to its reduction could cumulate at the defect. If the released anions

posses an ability to passivate the defect, then the corrosion process might be slowed or

completely inhibited. Summarizing, the conductive coating could act as a reservoir for

corrosion inhibitors which would be released as a result of corrosion induced reduction

75

of the conductive polymer and may even cure “injures” in the coating. Kinlen et.al. [69]

and Kendig et al. [75] showed that in the case of polyaniline anions can be released,

most likely by a corrosion-induced increase in pH caused by oxygen reduction at the

conducting polymer. Dopant release as a definitive consequence of electrochemical ICP

reduction induced by decrease in potential, which always accompanies delamination

originating at a defect, was shown by Paliwoda et.al. [145]. Unlike for the case of

standard corrosion pigments no uncontrolled leaching is observed in these cases, as the

release of the anions is triggered by the corrosion activity at the defect. Especially

the correlated interfacial potential changes are an effective and very precise trigger

for an intelligent release of corrosion inhibitors stored as dopants in the conducting

polymer [146] ensuring a very case selective, really intelligent release for the case of

polypyrrole, while for PANI most likely the correlated increase in pH will cause the

release of the anions [75].

As reduction is linked to ion transport, because either anions have to be released or

cations be incorporated, the electrochemical activity and of course the desired release

of the anions will be linked to the electronic and especially the ionic conductivity of the

polymer. Usually, electrochemical studies on the release and incorporation properties

of ions from and into conducting polymers are carried out on coatings of thicknesses

in the range of a few micrometers [137]. However in realistic corrosion scenarios the

ions have to be transported across much larger distances i.e over 100 micrometers or

more. As shown by Paliwoda [145] in such conditions the incorporation of cations

was observed instead of the release of ions. Additionally, a quick delamination of the

coating from the defect site was observed. Thus the subject of the transport of ions

over large distances across the conductive polymer seems to be very important for

understanding how the ICP coating will behave.

It has to be pointed out that the reduction of the polymer is a competitive reaction

to the reduction of the oxygen. On the other side the conductive polymer after its

reduction becomes insulating thus further oxygen reduction will then occur at the

metal-polymer interface. Hence fast reduction of the ICP will entail the development

and subsequent progress of the delamination front. Therefore in order to understand

the process of the delamination of the conductive coating it is necessary to investigate

the influence of the reduction of the polymer as well as the reduction of the oxygen.


6.1 Oxygen and polymer reduction during delami-

nation of conductive coatings.

Due to the fact that both reactions, oxygen and polymer reduction, can occur during

the delamination of the conductive coating it is very difficult to study the individual

role of both reactions to the overall effect. However, this can be examined, if both

reaction are separated during the experiment. This can be realized using the experi-

mental setup which is schematically presented in fig. 6.2. Similarly to the experiment

presented in chapter 5 in order to avoid electrochemical interaction of the conductive

polymer and the metallic substrate a gold layer deposited on glass plate was used as

a substrate. The polypyrrole layer was deposited electrochemically according to the

procedure which is described in detail in chapter 4. In order to stabilize the thin

polypyrrole coatings the samples were additionally coated with a ca. 5 µm layer of

MS-Top clear coat (Glasurit). This is a two-component lacquer without any additives,

pigments or corrosion inhibitors. The basic formulation of this model coating is similar

to typical automobile repair coatings. It is a polyfunctional epoxy-ester, cured with

polyamidoamines and containing groups of –OH, –COOH, –CONH and –NH. The coat-

ing was prepared by applying one droplet of Glasurit to a Speedline Technologie P6700

spincoater at a rotation speed of 2000 rpm (rotation per minute) for 20 s.

Afterwords the samples were dried for half an hour at 60 ◦C and then stored in a

desiccator. Before the experiment the sample was mounted in the setup presented

in figure 6.2. Due to the choice of an inert substrate, i.e. gold, an artificial defect

was simulated using potentiostatic polarization. In earlier experiments, the defect

Figure 6.2: Setup for the investigation of the contribution of oxygen and polymer

reduction to the delamination of the conductive coating.

6.1. Oxygen and polymer reduction during delamination of conductive coatings. 77

was simulated by pieces of iron. However, the potential of corroding iron in nitrogen

atmosphere (about -800 mV SHE) was observed to be much lower than in air. In order

to ensure the same defect potential under both conditions potentiostatic control was

chosen. An Ag/AgCl microelectrode was used as reference electrode [156]. A platinum

sheet served as counter electrode and a gold wire, placed 0.5 mm from the sample,

was polarized with the potentiostat at a potential of -660 mV versus Ag/AgCl, thusly

simulating a defect of corroding iron. Indeed, this was not necessary as at the edge of

the sample gold was exposed to the electrolyte, acting by its polarisation as defect. An

additional advantage coming out from the potentiostatic polarisation of the substrate

is that the current flowing from the defect to the sample could be recorded. For the

measurement of the cathodic current to the sample a Keithley Picoammeter 485 was

used at zero resistance. The cathodic delamination of the coating can occur only in

humid atmosphere, therefore gases which were supposed to be introduced into the SKP

chamber passed three bottles filled with water. Thanks to that the relative humidity

inside the chamber was maintained at the level of 95%.

The reduction/delamination of the conductive coating was performed in two different

atmospheres. At the beginning of the experiment the SKP chamber was filled with

nitrogen. Thus only the reduction of the polymer accompanied with ion exchange was

possible. The electrolyte was purged with nitrogen for 15 min and then injected into

the defect. Then the desired potential of the defect was adjusted using the potentiostat.

In a second step, after about 2-3 h, when about 2-2.5 mm of the conducting polymer

were reduced in the nitrogen atmosphere, oxygen was introduced which additionally

resulted in the reduction of oxygen. The progress of the reduction/delamination was

observed by the Scanning Kelvin Prove which recorded maps of the potential over

conductive polymer.

In fig. 6.3 the evolution of the potentials as monitored by SKP is shown for the case of

four polypyrrole films of different thickness electrodeposited on gold and covered by the

unpigmented transparent topcoat (MS top coat, see above). The potential measured

on the intact polypyrrole film (not affected by the clear coat) is about 350 mV vs. SHE

(±50 mV), as this potential which is equivalent to the position of the Fermi level [138]

in the polypyrrole, referenced versus SHE, depends extremely sensitively on even the

slightest deviations during the polypyrrole film preparation). Starting at the border to

the artificial defect the polypyrrole is reduced to the potential adjusted in the defect,

i.e. -500 mV vs. SHE. Figure 6.3a shows the results obtained on the sample with a


0 1000 2000 3000 4000−500

−400

−300

−200

−100

0

100

200

300

400

d / µm

E /

mV

(a)

0 1000 2000 3000 4000−500

−400

−300

−200

−100

0

100

200

300

400

d / µm E

/ m

V(b)

0 1000 2000 3000 4000−500

−400

−300

−200

−100

0

100

200

300

400

d / µm

E /

mV

(c)

0 1000 2000 3000 4000−500

−400

−300

−200

−100

0

100

200

300

400

d / µm

E /

mV

(d)

Figure 6.3: Successive SKP profiles (time interval 30 min between profiles) measured

by SKP, first during reduction in oxygen-free nitrogen atmosphere (solid lines), then

during delamination in air (dotted lines). For sample (d) air was exchanged back to

nitrogen at the end of the experiment. The samples differ in the thickness of the

tosylate doped polypyrrole film: (a) 0.271 C/cm2 deposition charge (about 1.1µm),

(b) 0.103 C/cm2 deposition charge (about 0.4µm), (c) 0.0501 C/cm2 deposition charge

(about 0.2µm), and (d) 0.022 C/cm2 deposition charge (about 0.1µm).

thickness of about 1.1 µm. The first five scans were measured in nitrogen atmosphere

and show the reduction of the polypyrrole which indeed progressed as a moving front.

The potential profile in the reduced area shows a slight slope which is due to only a

small Ohmic drop iR caused by the ion transport through the coating and/or along the

interface (the horizontal profile or even slight upwards bend near the border between

6.1. Oxygen and polymer reduction during delamination of conductive coatings. 79

d /

μm

500

1 000

1 500

2 000

2 500

3 000

3 500

4 000

4 500

t / min0 200 400 600 800

270 mC/cm2

103 mC/cm2

50 mC/cm2

22 mC/cm2

(a)

i /

μA

0

0,5

1

1,5

2

2,5

3

3,5

t / min0 100 200 300 400 500 600 700 800

271 mC/cm2

103 mC/cm2

50 mC/cm2

22 mC/cm2

(b)

Figure 6.4: (a)Position of the reduction/delamination front vs. time for the four sam-

ples. (b) Galvanic current between artificial defect and sample vs. time. The change

to oxygen containing atmosphere results in a sudden and steep increase in the current

(see b) and for the three thinner films in a significant slow down of the progress of the

front; in fact, for these three films the progress comes even basically to a stop for a

period of about 3 h (between 200 and 380 min).

d = 100 and 0 µm is due to an electronic artifact of the SKP), which indicates a

very low resistance R for the ion flow required for keeping charge neutrality. After the

introduction of the oxygen (dashed red curves) no significant change in the progress of

the front occurs, only a very slight decrease of the speed of the reduction front can be

noticed. However, for thinner samples a different behavior was observed. Figures 6.3b,

6.3c, 6.3d show results obtained for samples of about 0.4, 0.2 and 0.1 µm thickness. For

all thinner samples the first part of the experiment performed in nitrogen atmosphere

proceeds quite similar to the about 1 µm thick film, but a higher iR drop is observed,

i.e. the slope in the potential profile in the reduced region is much higher than for the

thick coating. Even more significant, after switching to oxygen containing atmosphere

all three thinner samples show a significant slow down in the progress of the front, for

a while the progress of the front even nearly stops (see fig. 6.4a).

Also, already in the curves for the 400 nm film shown in fig. 6.3b an instantaneous

jump to more positive potentials is observed in the area where the polypyrrole film was

reduced in nitrogen atmosphere. This jump increases with decreasing film thickness

(fig. 6.3b, 6.3c, 6.3d). At the same time the measured Galvanic current between

artificial defect and the sample also makes a jump to significantly higher currents

80 6. The ion transport through the conductive coatingE / m

V

-600

-400

-200

0

200

400

d / μm0 500 1000 1500 2000 2500

Reduction at -500mV in nitrogen

Directly after change to +300mV

(a)E /

mV

-600

-400

-200

0

200

400

600

d / μm0 500 1000 1500 2000

Delamination at -500mV

Directly after change to +300mV

(b)

Figure 6.5: (a) A 1µm polypyrrole film is first reduced in nitrogen atmosphere and

then the potential in the defect is switched to a higher value (e.g. +300 mV). The

potentials in the reduced area measured directly after the switching remain low. (b) If

the same experiment is performed in air the potentials go up immediately, indicating

delamination.

(see fig. 6.4b). After a while the potentials in the area where the potential jump

occurred slightly decrease. More notably the slope in the potential profile, indicating

a substantial iR drop, gradually decreases and further progress of the moving front

is observed, but slower than before in nitrogen. While in nitrogen atmosphere the

currents scale quite well with the film thickness, after switching to oxygen atmosphere

the currents and also the shape of the current curves over time directly after the change

in atmosphere are pretty much the same (see fig. 6.4b). If the polypyrrole film is first

reduced in nitrogen atmosphere and then the potential in the defect is switched to a

higher value (e.g. +300 mV), the potentials in the reduced area measured directly after

the switching remain low (see fig. 6.5a). If the same experiment is performed in air the

potentials go up immediately (see fig. 6.5b). This means that in the latter case there

is a good Ohmic contact between defect and region which has shown the low potentials

just before the switching, which indicates that the interface is delaminated in this

area, because the electrolyte film which forms at the delaminated interface provides

direct Ohmic contact to the defect (see e.g. [74]). The coating just reduced in nitrogen

atmosphere on the other hand is obviously not delaminated.

6.2. Cation incorporation versus anion release 81

I / A

-2x10-4

-1x10-4

0

10-4

2x10-4

3x10-4

4x10-4

E / V vs. Ag/AgCl-1,0 -0,8 -0,6 -0,4 -0,2 0,0 0,2 0,4

Figure 6.6: Cyclic voltammograms of polypyrrole doped with tosylate (0.1 C/cm2

deposition charge) in 0.1M tetrabutylammonium chloride. Scan speed: 10 mV/s.

6.2 Cation incorporation versus anion release

Since Ohmic drops in the potential profile indicate low ion mobility, additional experi-

ments were carried out in order to investigate, what kind of ions dominated the process,

i.e. to what extend cation incorporation and to what extend anion expulsion occur dur-

ing the polymer reduction. For the ion transport studies in addition to the tosylate

doped polypyrrole also samples doped with the smaller chloride ion were prepared, so

as to increase the anion release properties of the coating. Chloride doped polypyrrole

layers can be obtained either by direct polymerization in electrolyte containing chloride

anions or by anion exchange from polypyrrole originally doped with other anions, such

as the tosylate. The first method ensures that really only chloride will be present in

the polymer matrix, but as described in the literature the polymer itself may have a

completely different structure than the tosylate doped one [154]. Therefore, the an-

ion exchange from tosylate doped polypyrrole was chosen. In the case of the tosylate

anion this will not guarantee the complete anion exchange, but the polymer matrix

should remain similar and enable a comparison of the tosylate and the chloride doped

polypyrroles; as the smaller chloride anions take the places of the larger tosylate a

much improved anion transport should result this way. The ion exchange was per-

formed using the same electrochemical set-up as in the case of the polymerization. In


order to avoid cation incorporation a 0.1M tetrabutylammonium chloride salt solution

was used, where the ultra large cation cannot be incorporated. Before the experiment

the solution was purged with nitrogen for 15 min. After immersion in the solution the

samples were cycled 10 times between potentials of -1100 and 200mV vs. Ag/AgCl at

a scanning speed of 10 mV/s (see fig. 6.6). XPS spectra of the sample before and after

cycling are shown in figure 6.7.

Chloride was indeed incorporated into the polymer, but a small sulfur peak indicates

that the anion exchange was not complete and that some tosylate anions (S2O2−8 ) still

remained inside the polymer. Before the ion exchange the sulfur atomic concentration

in the coating was about 2.3%. After ion exchange the chloride and sulphur concentra-

tion were 1.7% and 0.49%, respectively. This shows that after cycling almost the same

doping level was reached (2.19% compared to 2.3%) This is in accordance with the

much higher currents occurring during the chloride exchange shown in fig. 6.6, which

demonstrates that nearly all the anions in the coating were exchanged, as correspond-

ing curves for the other films of different thickness show that the exchanged charge

scales with film thickness, and that a correspondingly high mobility was there, which

is in contrast to the much lower mobility of the larger tosylate anion (the currents

during cycling in tosylate were much lower, stemming from a surface layer only). Sim-

ilar experiments with a small cation in solution, such as in KCl, show even somewhat

higher currents and exchange seems to occur by both anions and cations, but with a

predominance of the cations.

In fig. 6.8a the potential profile during reduction in nitrogen atmosphere is shown for

the same type of polypyrrole, i.e. polypyrrole doped with tosylate, but now with tetra-

butylammonium chloride as electrolyte in the defect. The big organic cation cannot be

incorporated into the conducting polymer and hence the reduction can only occur, if

the tosylate anion is released. As can be seen, although near the defect at the beginning

the polypyrrole slowly starts to get reduced, after a few hours not much further change

is observed and no real progress of the reduction front takes place. However, immedi-

ately after injection of KCl containing solution into the defect a fast reduction of the

polypyrrole sets in, propagating into the intact polymer. Obviously, the expulsion of

the tosylate anion does occur only in a small, about 300 µ wide region near to the defect

(in fig. 6.8a) (the front stops at d = 200µm, and the d = 0 position of the scans is an-

other 100 µm away from the position of the border to the defect). No further progress

is observed, and the polypyrrole even near the defect does not get fully reduced (only


CPS

0

1000

2000

3000

4000

5000

6000

7000

8000

Binding Energy / eV350 300 250 200 150

C 1s

S 2s S 2p

(a)

CPS

0

1000

2000

3000

4000

5000

6000

7000

8000

Binding Energy / eV350 300 250 200 150

C 1s

S 2pCl 2pCl 2s

(b)

Figure 6.7: XPS spectra in the range between 150 and 350 eV binding energy of (a)

polypyrrole doped with tosylate and (b) with chloride (after anion exchange by cycling

in tetrabutylammonium chloride).

down to about -100 to 0mV vs. SHE). The small potassium cation, however, seems to

be able to easily move through the polypyrrole (as also fig. 6.3 shows). This means

that for the reduction curves shown in fig. 6.3 the charge neutrality was kept nearly

exclusively by cation incorporation into the conducting polymer. This is not surprising

as already from the CVs shown in fig. 6.6 it is obvious, that reduction/oxidation based

on sole release/incorporation of tosylate results only in an electrochemical activity of a

surface layer of the polypyrrole film, while it was observed in similar experiments that

reduction/oxidation based on additional incorporation/release of potassium shows the

full activity of the whole polypyrrole film. This is also the case for CVs in tertrabuty-

lammonium chloride containing electrolyte, as it was possible to exchange nearly all

tosylate for chloride (in agreement with Reynolds et al. [139] and figs. 6.6 and 6.7).

Hence, the same experiments as carried out for polypyrrole doped by tosylate were

also performeld for polypyrrole doped by chloride, prepared by exchanging the tosy-

late by chloride. Hence, for this case chloride expulsion should occur. And indeed, as

can be seen in fig. 6.8b, a slow progress of the reduction front can be observed, i.e.

the reduction can proceed based on solely chloride expulsion for charge compensation.

However, the progress is extremely slow, much slower than for the case of incorporation

of the small potassium cation. Interestingly, if after a certain while, when the front has

progressed over some distance (about 600µm, i.e. d = 500 + 100µm), KCl is injected

to the defect, no increase in the velocity of the reduction front is observed (see fig.


0 1000 2000 3000 4000−500

−400

−300

−200

−100

0

100

200

300

400

d / µm

E /

mV

(a)

0 500 1000 1500 2000−500

−400

−300

−200

−100

0

100

200

300

d / µmE

/ m

V(b)

Figure 6.8: Potential profiles during reduction in nitrogen. (a) For a polypyrrole film

doped with tosylate and with the big organic tetrabutylammonium cation in the defect

(solid lines). The progress is very slow, the polypyrrole does not get fully reduced even

directly at the border to the defect (potentials stay above -100 mV). After injection of

KCl into the defect a fast reduction front progresses into the polypyrrol (dotted lines).

Time intervals for the case of terabutylammonium in the defect: 3 h, after injection of

KCl: 15 min! (b) For a polypyrrole film doped with chloride (exchanged) and again

with 0.1M tetrabutytlammonium chloride in the defect (solid lines). Time interval

between successive profiles: 3 h. The progress is slow, but steady. Full reduction is

observed. After injection of KCl into the defect no increase in velocity is observed

(dotted lines, still 3 h time between scans).

6.8a). If the experiment is started with the small potassium cation in the defect, then

a much faster progress of the reduction front is observed (see fig. 6.8a).

Obviously, even if the anion is quite mobile cation incorporation is favoured. In fig. 6.9b

it is shown that the cation mobility in polypyrrole depends on the size of the cation.

Due to its higher charge density the small Li+ cation is surrounded by a much bulkier

hydration shell than the potassium cation, which reduces its mobility [24]. Indeed, the

Li+ cation incorporation is obviously much slower, than the potassium incorporation,

resulting in much slower progress of the reduction front. In fig. 6.10 an XPS analy-

sis of both poylypyrrole doped with tosylate and doped with exchanged chloride that

was reduced in a delamination set-up in the presence of the small potassium cation

in the defect is shown. In both cases a significant potassium concentration can be


d /

μm

0

500

1 000

1 500

2 000

2 500

3 000

3 500

t / min0 500 1 000 1 500 2 000 2 500 3 000

KCl at the defect

Tetrabutylammonium

chloride at the defect

(a)

d /

μm

500

1 000

1 500

2 000

2 500

3 000

t / min0 200 400 600 800 1 000

LiO4 at the defect

KCl at the defect

(b)

Figure 6.9: Comparison of the progress of the reduction front vs. time for chloride

doped polypyrrole in the presence of a small cation in the defect (black circles) and for

the case of the big tetrabutylammonium cation in the electrolyte (red triangles), data

from fig. 6.8b). The injection of KCl has no effect on the slow curve started with only

the big cation in the defect. (b) The effect of the cation size on reduction velocity for

polypyrrole doped with tosylate.

measured, further confirming cation incorporation as the dominating charge neutral-

ization mechanism during polypyrrole reduction for both cases, even for the case of the

mobile chloride anion as dopant, although some chloride has been clearly expulsed as

the chloride peak is lower than before the reduction of the polypyrrole (see figs. 6.7b

and 6.10b).

Why is the progress of the moving front so much affected by the change from nitrogen

to oxygen for the case of the thinner coatings and not for the thicker one? Why does

cation incorporation dominate, and why does initial enforcement of anion expulsion

prevent later fast cation incorporation? While during the polypyrrole reduction in

nitrogen atmosphere the currents steadily decrease as the reduction front proceeds

into the coating, following roughly a√

t behavior characteristic for diffusion-controlled

processes, in oxygen atmosphere the currents steadily increase with the progress of

the moving front (fig. 6.4b). During reduction in nitrogen the only reaction is the

polypyrrole reduction. For this an ion flow of either cations from the defect or anions

into the defect or a mixture of both has to compensate the electronic reduction current.

As the amount of the polypyrrole to be reduced is directly proportional to the film

thickness, the currents should directly scale with the film thickness (what they roughly

86 6. The ion transport through the conductive coatingCPS

0

2000

4000

6000

8000

1e+04

Binding Energy / eV350 300 250 200

C 1s

K 2p

S 2sS 2p

(a)

CPS

0

2 000

4 000

6 000

8 000

1e+04

1,2e+04

1,4e+04

Binding Energy / eV350 300 250 200

C 1s

K 2p

S 2pCl 2p

(b)

Figure 6.10: XPS spectra between 160 and 350 eV binding energy after the reduction

(in the delamination set-up, with KCl in the defect) of (a) tosylate doped and (b)

chloride doped polypyrrole. The charge neutralization mainly occurs via incorporation

of the small cation.

do). If the reaction is controlled by the diffusion of ions, the progress of the reduction

front will steadily decrease, following a√

t law, as observed. This way, the reduction

current decreases in the same way as the reduction front moves further away from the

defect, because due to this diffusion control per unit time a correspondingly decreasing

volume of polypyrrole is reduced. As in all four cases the oxygen was introduced

after a similar time, i.e. when a similarly extended stretch of reduced polypyrrole was

reached, the observed similar shape of the current over time curves directly after oxygen

introduction suggests that at first oxygen reduction predominantly takes place at the

interface between the already reduced polypyrrole and the metal. As a consequence

delamination of the film from the gold should occur. The results shown in fig. 6.5

support this. Obviously, in the presence of oxygen delamination and film reduction

occur simultaneously, or at least delamination closely follows reduction. Hence, the

different delamination behaviour of the 1.1 µm thick polypyrrole coating, which is

characteristic also for thick coatings of polyaniline, compared to the thinner coatings

has to do with the fact that in nitrogen atmosphere, where all coatings behave the same,

the reduction reaction takes place within the bulk, while in the oxygen containing

atmosphere the overwhelming part of the electrochemical reaction, i.e. the oxygen

reduction, takes place at the delaminated interface metal/polymer. Since the area of the

interface between reduced polypyrrole and metal steadily increases, even if the velocity


�/�/��

�

�

�

�

�

�/�/��

��2��

��2��

Figure 6.11: Current vs. progress of the reduction/delamination front for the about

1.1µm and for the about 0.4µm thick polypyrrole film. In the nitrogen atmosphere

only polypyrrole reduction takes place, resulting in about three times higher currents

for the thicker sample (not at the beginning, but at later stages). After change to

air oxygen reduction at the interface and delamination of the films occur. While the

thick coating shows immediately after the gas exchange a nearly linear dependence on

delaminated area, for the thinner one this takes a while.

of the front is slowly decreasing because of the diffusion control, still steadily increasing

currents result. This become directly obvious from fig. 6.11, where for the about 1.1

µm and the about 400 nm thick film the galvanic currents between defect and coating

are plotted versus the progress of the reduction/delamination front. After switching to

oxygen atmosphere for the thick film nearly immediately a linear dependence between

delaminated area and current is established, while for the thinner coating it takes a

while before the linear dependence is reached. This is most likely due to the fact that

the thinner coating cannot provide the high cation transport required for the much

higher reduction currents caused by the oxygen reduction. Hence, in the case of the

thin coating the interface needs to delaminate first before the reduction/delamination

can further proceed, because it needs the good ion transport along the delaminated

interface. For the thick coating this is not necessary, because of the good ion transport

though the bulk (see also fig. 6.12 for explanation).


The increase in current with newly delaminated area shows that the oxygen reduction

takes place at the delaminated area. However, the absolute values of the current

density for oxygen reduction in the area already reduced during reduction in nitrogen

is different for both films (for the case of the thick one it is about 40% lower than for the

thinner one). This most likely is due to differences in the exposed gold area at the defect

border (due e.g. to partial mechanical film delamination at the edge of the sample),

which results in different initial oxygen reduction currents just after switching (different

offset currents). This is also assumed to be the explanation for the slight deviation of

the proportionality of the reduction currents with film thickness. As discussed above,

during reduction in nitrogen the currents should be directly proportional to the film

thickness. This is roughly the case as can be seen from fig. 6.4b. However, at the

beginning of the reduction experiments, the reduction currents for the 400 nm film are

not that much lower than for the 1.1µm film (see fig. 6.11). But the proportionality

is established after some time of reduction. Only for the thinnest coating the currents

are still much too high. This maybe due to the well known fact that the innermost

polypyrrole of an electrodeposited polypyrrole film has a different structure from that in

the outer layers, see e.g. [140,157]. Concluding, the fact that for the reduction process in

nitrogen atmosphere the velocity is nearly the same for all films, while the velocity of the

front significantly decreases for thinner coatings in the oxygen containing atmosphere,

where much higher currents have to be provided, suggests that for the thicker coating

the major part of the ion transport occurs over the bulk, that is within the polypyrrole,

while for decreasing films thickness transport along the interface becomes more and

more important. Since ion diffusion/migration along an intact interface is much slower

than along the delaminated interface, this also explains why for the thinner coatings

the jump of the potential in the already reduced area towards more positive potentials

is observed and why the further progress of the front stops for a moment (see fig. 6.4a).

This is simply due to the fact that the much higher ion flow required for supporting

oxygen reduction at such negative potentials as -400 to -500mV vs. SHE cannot be

supported by the thin polypyrrole films, i.e. the interface needs to be delaminated first.

Hence, the potential jumps to more positive values first, because the oxygen reduc-

tion charges the interface positively until the oxygen reduction kinetics are sufficiently

slowed down to rates which can be provided by the ion transport. These lower rates

result in slower delamination. During this slow delamination the potentials decrease,

but as the interfacial resistance seems still to be higher than through the bulk of the


(a) (b)

Figure 6.12: Schematic explanation of the role of cation transport in thick coatings

and interfacial transport at the delaminated interface for thin coatings. (a) In nitrogen

(no delamination), (b) in air (delamination). For the thick coating delamination does

not increase ion transport significantly.

thick film a pronounced slope remains for all thinner coatings (see also [23, 74]). In

nitrogen atmosphere the reduction current scales directly with the film thickness, just

the same way as the conductivity does (as this also increases with the thickness). This

explains why all films behave the same. As shown in fig. 6.3d for the thinner films

a switch back from air to nitrogen atmosphere results in an immediate lifting of the

Ohmic drop, because the oxygen reduction falls away and thusly the involved currents

drop to much smaller values; the ion flow has just to provide the current flow required

for the polypyrrole reduction. This also results in a significant increase of the veloc-

ity of the moving front. In fact, it is even higher than before switching to oxygen,

which will most likely be due to the already achieved delamination of the interface

before switching back to nitrogen and the corresponding increase in overall ionic con-

ductivity. Recapitulating, these findings show that for the investigated tosylate doped

polypyrrole films the ion transport through the bulk of a thick film is more effective

than that one along the delaminated interface and that for thick continuous coatings

the overall delamination velocity is thusly very high and determined by the fast cation

transport properties of the polypyrrole. As a consequence of the above, the conclu-

sion to be drawn is: the lower the amount of polypyrrole, the better. This means

that polypyrrole would be unsuitable for corrosion protection, if the risk exists that

large defects occur which cannot be inhibited. For polyaniline a similar behaviour is


observed, and in fact it is assumed that this is generally the case for all redox active

conducting polymers.

Why is that the case? What is the reason for the fast cation and the slow anion trans-

port? Generally, the mobility of dopants is connected to their size, i.e. small anions are

mobile and large anions are less mobile. However, the size of an ion is usually not the

only factor influencing its mobility (see e.g. [141, 142]). Depending on the polypyrrole

polymerization conditions (see e.g. [91, 134]), the thickness of the polymer [134] and

the ions present in the electrolyte solution [90, 158] anion and/or cation exchange is

observed in the Ppy films. Polypyrrole films doped by Cl−, ClO−

4 , BF−

4 , SO2−4 and

NO2−3 under most conditions predominantly expel the anions during electrochemical

reduction. Nearly exclusive anion exchange was observed in the case of polypyrrole

doped with ClO−

4 in aqueous solution of CsCl as well as NaCl [143]. This indicates

that cations with smaller ionic radius, but significantly hydrated, as well as cations

with larger ionic radius, but less hydrated, were not incorporated into this polymer. In

the case of polypyrrole doped with ClO−

4 and PF−

6 a significant dependence between

the rate of electropolymerization and the mobility of anions in the film was observed.

The lower the polymerization rate the more compact films were obtained, resulting in

a suppressed mobility of the incorporated anions. In this way polypyrrole films with

small but immobilized anions were obtained, which show pronounced cation incorpo-

ration during reduction [88, 142, 144]. Polypyrrole doped with BF−

4 ions shows mostly

anion release during reduction. However, also in this case under certain conditions, i.e.

at advanced states of reduction, cation incorporation is observed [91]. According to the

hypothesis presented by the authors, cations are incorporated in order to neutralize

the charge of trapped, i.e. immobile, BF−

4 ions deep inside the bulk of the polymer. A

similar hypothesis postulating the existence of two forms of anions in the polypyrrole

film, socalled “free” and “bound” anions, was proposed by Vorotyntsev et al. [141].

Similarly, for the case of polypyrrole doped with tosylate, Naoi et al. observed that at

low levels of reduction mostly anions are released, but when the polymer is reduced

further predominantly cation incorporation contributes to the charge compensation re-

action [91]. This observation is consistent with results presented by Zhou et al. who

observed that during the reduction of tosylate doped polypyrrole substantial amounts

of Na+ or Cs+ cations were incorporated [143]. For polypyrrole doped with chloride

Reynolds et al. concluded from the EQCM measurements that anion exchange pre-

dominates during electrochemical reduction in NaF, NaCl and NaClO4 solutions [139].


(a) (b) (c)

Figure 6.13: Schematic explanation of (a) the high mobility for cations in the presence

of anions remaining in the polymer. (b) If anions are quantitatively expelled from the

conducting polymer during reduction then the fast cation hopping is inhibited in the

ion-free zone (c).

To our knowledge claimed exclusive exchange of either anions or cations has so far

never proven to be really of 100% exclusivity, with the exception of experiments where

exceptionally large anions or cations are used so that there incorporation is definitely

impossible. With this exception, the incorporation of 1 or 2% of the other kind of

ions cannot be excluded in the standard immersion electrochemistry. This point is im-

portant for answering the question why cation incorporation is always the dominating

effect, and why it has no effect if the small cation is injected at an advanced stage

of the polypyrrole reduction, i.e. after already an extended stretch of film has been

reduced (see fig. 6.10a).

This is explained in the model illustrated in fig. 6.13: At the beginning of the reduction

of the conducting polymer both anion expulsion as well as cation incorporation occur,

supposedly in the same way as they would do during a standard electrochemical reduc-

tion experiment. However, while in case of the immersed sample the transport has to

occur just over the thickness of the conducting polymer film, usually not more than a

few 10µm, in the delamination set-up the reduction proceeds laterally from the border

with the electrolyte into the oxidized coating. Hence, the distance across which either

transport of anions or cations has to take place is quickly one to two orders of magni-

tude larger. Now, even for a conducting polymer doped with a mobile anion, showing

during standard electrochemical reduction predominant anion release behavior even in

the presence of small cations in the electrolyte, some cation incorporation will occur

in addition to the anion release. Hence, some negative charge will be immobilized in


the polymer, neutralized by cations. This is the scenario of a typical cation transport

membrane, where cations can hop from fixed negative charge to fixed negative charge.

With each anion remaining in the polymer the mobility of the cations will increase, i.e.

the further the reduction front progresses into the conducting polymer, the higher the

contribution of cation incorporation to the charge neutralization. Finally, predominant

cation incorporation is observed. On the other hand, when all anions were expelled

from the conducting polymer during reduction (by use of a very large organic cation

that cannot be incorporated at all, see fig. 6.8), also the mobility of small cations is

very low in the reduced polymer, because the fast hopping from negative charge to

negative charge cannot take place anymore. This is clearly shown by the fact that the

progress of the reduction front remains unaffected when the KCl was injected after the

reduction already had substantially advanced from the border with the electrolyte (see

fig. 6.8). In this case the cations have to cross conducting polymer free of anions. As

no effect at all is observed, in this case anion expulsion remains the dominating process.

However, if the KCl is injected when the reduction is not fully complete yet even in the

polymer near the electrolyte, such as was the case for the tosylate doped polypyrrole

film shown fig. 6.9a, then still enough anions are left in the polymer for allowing the

mechanism of fast cation incorporation to step into action. It is assumed that this

mechanism is of general validity, i.e. it will occur not only for polypyrrole films, but

also for other conducting redox polymers, such as polyaniline or polythiophene. In all

these cases the charge neutrality has to be achieved by ion transport. Depending on

the exchange properties of the polymer it just may take a different distance from the

defect before the cation incorporation will dominate the process. This also explains

why no positive corrosion protection was observed for polypyrrole electrodeposited on

iron and doped with molybdate, an iron corrosion inhibitor, when delamination attack

occurred starting from a defect covered by KCl or NaCl containing electrolyte [146],

i.e. a set-up similar to the one used here and which is characteristic for atmospheric

corrosion situations, while corrosion protection was observed when the whole sample

was immersed in electrolyte [137]. One difference is the amount of active surface in

contact with the electrolyte as explained in fig. 1 in Section 2. But molybdate is a

very effective corrosion inhibitor on iron, so its release should significantly reduce the

passivation current density already at low concentrations [145]. The main problem is

that in the first case no release of the inhibitor anion was observed, in the latter case

the release occurs. The key difference between the two situations is just the length

scale over which reduction of the polymer has to occur. Since the active surface in


contact with the electrolyte is much smaller in the case of the delamination set-up, for

the release of the same amount of inhibitor the reduction front has to proceed over

several 100 micrometers as compared to one micron (film thickness) in the immersion

case. Based on the above discussed and the findings reported in [137] it has to be

assumed that at the beginning at least some release of anions took place also in the

delamination set-up, but this was not enough to stop corrosion in the defect. As the

reduction front proceeded from the defect further into the coating, the mechanism of

charge compensation quickly turned to more or less exclusive cation incorporation, in

accordance with fig. 6.13. Hence, no substantial molybdate release could occur [146].

6.2.1 Conclusion

The results presented above show that continuous coatings of conducting redox polymer

will fail to provide corrosion protection in presence of larger defects, which they cannot

passivate, and will show a fast break-down of the whole coating by fast reduction. This

fast reduction is caused by high cation mobility in the reduced polymer, which is due

to the gradual transformation of the polymer into an highway for fast cation transport

with increasing progress of the reduction front, as schematically explained in fig. 6.13.

It is proposed that this is true for all kinds of redox polymer, regardless of kind of

dopant, polymerization conditions, etc. It is assumed that this fast cation transport

will also occur in composite coatings containing pigments or filaments of conducting

polymer in a nonconductive matrix polymer where high conductivity is reached by

extended percolation networks of the conducting polymer. The cation transport would

then occur via these percolation networks.

Chapter 7

Non continuous ICP films: on the

role of the polarization of the

metal/non-conductive polymer

interface by ICP patterns.

7.1 Introduction

The possible mechanisms for corrosion protection proposed for conductive polymers

that were discussed in the former chapters are based on different properties of the con-

ductive polymer. Often some properties which are recommended from the one point

of view are highly undesired after consideration of other mechanisms. For example the

mechanism based on the defect polarization demands highly oxidized and conductive

polymer which could act as a charge storage and be able to provide the high cur-

rents necessary for the metal passivation. But on the other hand this would result

in fast coating break-down, if the defect is too large to be passivated. Furthermore,

the delamination experiment performed in the O18 containing atmosphere showed that

rather less conductive coatings would be required to ensure the shift of the oxygen

reduction site from the metal-polymer interface toward the bulk of the the polymer.

Finally the idea of the ”intelligent coating” which assumes the release of corrosion

inhibitors can be realized only when the conductive polymer assures good anion trans-

port through the polymer matrix. But above investigations showed that on the length

967. Non continuous ICP films: on the role of the polarization of the

metal/non-conductive polymer interface by ICP patterns.

scales discussed here any continuous ICP film would eventually turn to high cationic

conductivity which may accelerate the corrosion process. Properties which are prefer-

able according to one mechanism are unwanted considering other models. Therefore

the application of the continuous conductive coating as a protective coating seems not

to be easy. The complexity of this system requests careful consideration of all aspects

of the corrosion inhibition. The question which arises after above conclusions is: do the

models for corrosion inhibition presented in this thesis and that are broadly discussed

in the literature already cover the whole potential of conductive polymers? Are all

models and all possible ways, how conductive polymers could be used, known? The

large number of recent publications concerning conductive polymers shows that the

number of unsolved issues and problems still attracts scientists. Therefore the search

for models and possibilities for corrosion protection is still going on. In this chapter

new results of the investigation for new protection mechanisms will be presented.

Most of the proposed mechanisms for corrosion protection by the ICPs such as the “self

healing” mechanism or defect polarization are based on the assumption that there is

a good ionic and electronic contact between the polymer and the defect. The ionic

contact is realized by contact between ICP and the electrolyte which ensures an ion

exchange between the defect and the polymer. The electronic contact is provided by the

metallic substrate. Under these conditions the polypyrrole due to its oxidative proper-

ties may provide a protection of the corroding area (passivation of defect). However,

the experiments performed in the former chapter proved that the polypyrrole coating

can not be treated as a barrier for the ion transport. In contrary, it may even accel-

erate the cation migration and hence may increase the corrosion process at the defect

and the delamination of the coating, if the defect can not be passivated. The cation

incorporation may be reduced by the careful choice of the doping anion, however the

complete elimination of this effect might not be possible. This issue seems to be the

greatest danger for the polypyrrole application as a corrosion inhibitor. Hence a new

tactic aiming at the reduction of the influence of the ionic transport has to lay as a

basis for the development of the new corrosion protection model. In this chapter a

possibility for protection under limited cationic transport condition will be discussed

as a promising method for corrosion inhibition by conductive polymers: the use of ICP

additions to non-conductive polymer coatings.

7.1. Introduction 97

7.1.1 Interface polarization by conductive polymer islands at

the interface metal/non-conductive polymer.

In the following we will discuss the formation of the galvanic element during the de-

lamination experiment. The electrochemical model of this situation was proposed by

Stratmann [23] and is presented in fig. 7.1. The upper part of the figure shows a

schematic cross-section through a polymer coated sample with a defect down to the

metal. The defect in the coating which is situated in the middle of the sample will

initialize the corrosion process. Shortly after injection of the electrolyte to the defect

the corrosion starts at the defect. The activation and dissolution of the iron is accom-

panied by a cathodic shift of the electrode potential. Both reactions, the anodic iron

oxidation and the oxygen reduction as the main counter reaction, have high rates be-

cause the oxygen easily diffuses through the electrolyte and the iron corrodes actively

in the chloride containing electrolyte. Also under the now delaminating coating near

the defect reactions have to balance each other. Due to the passivated metal-polymer

interface which impedes the metal dissolution both reaction rates are much smaller

and the electrode potential reaches more positive value, because the anodic reaction is

practically fully inhibited. At this potential the oxygen reduction is strongly limited.

The lower part of fig. 7.1 shows the situation when the galvanic coupling between the

defect and the intact area is established. The potential difference between these two

areas becomes a driving force for the delamination process. Some electrons produced

in the defect will flow to the intact area through the bulk of the metal. The electronic

current has to be compensated by the ionic current flowing along the interface. This

establishes a coupling between defect site and the interface, resulting in pulling down

of the potential at the interface toward the one of the defect, thusly enabling a higher

oxygen reduction rate and by this accelerating the interface degradation. The idea now

is: if a ICP were now located nearby, it could prevent this pulling down by the defect

by a pulling-up by the ICP and thusly preventing a decrease of the electrode potential

and thusly prevent an increase in oxygen reduction. Figure 7.2 shows the scheme of

a delamination experiment performed on the sample consisting of the metal substrate

coated with a non-conductive lacquer containing conductive pigments. The beginning

of the experiment is similar to the typical delamination experiment. The potential

profile measured by the SKP (indicated by yellow curve) shows three regions according

to the Stratmann’s model. The first zone indicates the region of the completely delam-

inated coating. The thick electrolyte layer under the delaminated coating ensures good



Figure 7.1: The electrochemical scheme of the corrosion process. Top figure shows the

profile of sample. Blue and red dashed lines represent the rate of the oxygen reduction

and the iron dissolution respectively.

ion transport, hence the ionic resistance is very low. This region may be recognized on

the SKP profile as a plateau of the low potential which begins right at the defect. The

second region represents the area where the coating disbondment is not complete. The

limited ion transport in this region is reflected in the ohmic slope measured by SKP.

The end of this region may be defined by a kind of boundary potential above which

oxygen reduction practically does not take place anymore. The last region points out

the intact area where the oxygen reduction is not possible. The sharp transition of the

potential between regions two and three is called the delamination front. According to

figure 7.1 the rate of the oxygen reduction drops significantly at sufficiently high poten-

tial. However, the delamination front can not be considered as a sharp border which

divides the delamination region from the intact area. Rather it has to be recognized as

an area where the active oxygen reduction takes place at small rates, the smaller the

more into the intact area. Figure 7.2 shows the situation when the conductive particle

7.1. Introduction 99

(a) (b)

Figure 7.2: The model of the polarization of the interface by the conductive polymer

during the delamination process.

stays in the way of the moving front of the delamination. When the delamination

reaches the conductive polymer, galvanic coupling between the polymer and delami-

nating area will be established, i.e. the ICP will start to get slowly reduced. In order to

maintain the charge neutrality of the polymer, its reduction must be accompanied by

the transport of ions. As shown in former chapters, cations supplied by the defect may

be incorporated to the polymer matrix or anions which are already present inside the

polymer may be expelled. In both cases the ionic current at the interface that supplies

the oxygen reduction needs either to increase or -if the interface cannot provide this

additional transport- the potentials would go up or at least stop to decrease. The ionic

resistance of the delaminated region is low due to thick layer of the electrolyte present

between the de-adhered coating and the metal surface. The enlargement of the ionic

current will not result in significant changes of the potential at the interface at this

region. At the partially delaminated area, however, where the coating still presents

good adhesion to the substrate the ionic resistance is high and thusly limits the rate

of the ICP reduction. The expected shape of the SKP profile after the delamination

front contacted with the conductive polymer is marked by the red line in figure 7.2. A



part of the region where the oxygen reduction was possible gets polarized above the

boundary potential. In consequence of that, a small zone where the oxygen reduction

is inhibited will be created. This region will be called in the following “protection

zone”, because as long as the conductive polymer is able to keep the high potential

inside the zone, the further delamination will be stopped. Of course the potential of

the ICP will slowly get lower due to the slow polymer reduction. Hence, the size of

the protection zone will gradually decrease. Due to shrinking of the protection zone

its resistivity will decrease allowing the flow of higher ionic currents and acceleration

of the polymer reduction. Finally, the partially reduced polymer will not be capable

to provide any protection and the delamination front will reach directly the ICP. Than

the large amount of ions supplied through the fully delaminated interface directly to

the ICP will cause a fast reduction of the conductive pigment. The fast reduction of

the ICP after the immediate contact of the ICP with the fully delaminated interface

has to be considered as a weak point of presented mechanism. However, it is assumed

that by choice of an optimal distribution of the ICP (size of and distance between

ICP particles) the fast break down could be more than overcompensated by the initial

stopping of the delamination process through the establishment of the protection zone.

Hence, the overall delamination protection provided by the pigmented coating may be

significant.

7.2 Delamination test of an isolated PANI dot ap-

plied on a chromium substrate.

The theoretical considerations presented in the former paragraph do not give a clear

answer about the likely outcome of the combination of positive and negative effects.

If the undesired effect of the fast reduction of the polymer significantly exceeds the

positive outcome of the protection it may be difficult to prove the existence of the

“protection zone”. The success of the mechanism depends on the ability of the con-

ductive polymer to polarize the metal-polymer interface above the onset potential of

the oxygen reduction. Even slight polarization may result in the creation of the ex-

pected protection zone. Taking into account that the onset potential of the oxygen

reduction is an individual parameter for the given material [162], the careful choice of

the substrate make it easier to clearly observe the postulated effect. Accordingly, the

7.2. Delamination test of an isolated PANI dot applied on a chromium substrate. 101

first delamination experiment described was carried out on chromium as this should

show a higher polarization effect with increased influence to the overall delamination

behavior, compared to iron or gold. Due to the band structure of the n-semiconducting

passive layer on zinc, chromium and iron, for instance, the oxygen reduction is inhib-

ited at potentials of about 50-100 mV higher than the respective flat band potentials.

Hence, for chromium the oxygen reduction is inhibited already at lower potentials, than

on iron. Gold has no semiconducting surface layer, and hence oxygen reduction can

occur still at quite high potentials. Thanks to that the existence of the investigated

mechanism could be proved.

7.2.1 Investigation of the delamination performance of non-

conducting polymer on chromium.

The experimental setup is presented in figure 7.4. The sample consisted of a chromium

layer evaporated on a glass plate. Evaporation was performed using the PVD system

which was already described in the chapter 3. The obtained thickness of the chromium

layer was 100nm. The surface of the chromium was covered by the native passive layer

which has grown due to the contact to the atmospheric oxygen. According to Sunseri

[159] the native passive layer presents low or slightly p-type conductivity. However,

an investigation of photocurrents done by Kim [160] showed that due to a duplex

composition of the chromium passive layer its polarization over a certain potential

range results in a change of the type of the conductivity: polarization at cathodic

potentials induces a p-type conductivity while the polarization at anodic ones results

in n-type behavior. The oxygen reduction on chromium should be influenced by this

complex oxide structure. According to Delahay [163] oxegen reduction is practically

inhibited above potentials of about -200 mV. Therefore polarization of the interface

over this potential should significantly reduce the rate of the oxygen reduction.

The above assumptions were experimentally investigated by means of a delamination

test. At the beginning of the experiment the potential at the defect was set to a positive

value which will prohibit any delamination. Then, it was gradually decreased until the

delamination started. The onset and further progress of the delamination front can

be interpreted as the onset of the electron transfer through the oxide surface of the

chromium substrate.



d /

�m

0

500

1000

1500

2000

2500

3000

E / V

vs. SH

E

�0.5

�0.4

�0.3

�0.2

�0.1

0

0.1

0.2

0.3

0.4

t / min0 200 400 600 800 1000 1200 1400

Delamination distanceApplied potential

Figure 7.3: Progress of the delamination of the PVB coating under different potentials

applied to the defect.

The delamination test was performed using a glass substrate coated with 100µm

chromium layer. As a top coat poly-vinyl-butyral (PVB) was chosen. This was ap-

plied onto the substrate from ethanolic solution. During the spincoating the ethanol

evaporates and the PVB polymerizes and creates a continuous layer. It was found ex-

perimentally that spincoating of the 10% solution of the PVB in ethanol at 2000RPM

for 20s ensures a flat and a homogenous layer of the PVB. The advantage of PVB

over commercial lacquers is that its composition is well known and the viscosity of the

solution can be controlled directly by the concentration of the PVB in the solution.

The viscosity of the coating is an important parameter which influences the thickness

and the homogeneity of the spincoated layer. The obtained results are presented in

fig. 7.3. At the beginning of the experiment the potential of 250 mV was applied. This

potential excludes the reduction of the oxygen so no delamination could be observed.

Then the potential was decreased. The development of a delamination front was ob-

served, when the potential applied to the defect was set to -360 mV. Following changes

of the potential did not have a significant effect.

7.2.2 Polarization of the chromium-polymer interface.

As polyaniline has a higher redox poetntial tan polypyrrole, the effect should be more

pronounced for polyaniline. Furthermore, polyaniline is easier to be dissolved in organic

7.2. Delamination test of an isolated PANI dot applied on a chromium substrate. 103

Figure 7.4: The experimental setup for the investigation of the polarization of the

PVB-chromium interface by the conductive polymer.

solvents, making it easier to apply single dots at the surface. Therefore for further in-

vestigations a polyaniline dispersion in xylene was chosen. The dispersion supplied by

Merck can be easily applied on the surface. According to the producer the polyaniline

is doped with organic sulfonic acid. As a top coat again poly-vinyl-butyral (PVB) was

chosen. The sample for the experiment was prepared following the sequence: first the

chromium surface was masked with adhesive tape having a rectangular hole. Then the

polyaniline was applied using the spincoater at 2000RPM for 20s. After 10 minutes the

adhesive tape was removed and an additional layer of the PVB solution in ethanol was

spincoated. Figure 7.5 shows potential profiles obtained during the delamination exper-

iment. Measurements were taken every 15min. The first profile (marked by the yellow

line) shows two plateaus of different potentials. The plateau at the potential of 200 mV

indicates the area coated only by the PVB coating. The other plateau where the poten-

tial reaches the value of 600 mV refers to the area coated at first by the polyaniline and

then by the PVB coating. The black vertical line which in visible in the figure, shows

the border which separates both regions. The defect was filled with 1M KCL solution

and polarized using the potentiostat. In order to initialize the delamination, the poten-

tial applied to the defect was set to -450 mV vs. SHE. At this potential the chromium

oxide presents relatively good conductivity so the oxygen reduction could take place.

For the first six scans a very fast progress of the delamination front can be observed.

The only slight ohmic drop following the delamination front indicates that the coating

easily de-adhered from the metal substrate ensuring a very good ion transport along

the metal-polymer interface. However, when the delamination front approached the

area coated with the polyaniline its progress was completely stopped and the polyani-



line gets slowly reduced, but only partly as the reduction goes only down to a potential

of about 0 VSHE. Obviously the potential cannot be fully pulled down, because of the

slope between the PANI and the potential in the delaminated area. This is the IR drop

caused by the flow of ions along the high resistance intact interface in the protection

area. An equilibrium has established between ion flow along the interface and reduction

rate of the polyaniline and this resulted in a slope that allows reduction only down to

0 VSHE. The magnitude of the potential slope between the delaminated area and the

intact area is given by the product of I · R, where I is the ionic current and R is the

ionic resistance of the metal-polymer interface. The high potentials in the protection

zone effectively hinder oxygen reduction and stop the subsequent delamination of the

coating. Next potential profiles shows at the PANI site the development of a plateau

at about 0V vs. SHE and creation of the PANI reduction front. The positive potential

of the plateau excludes the possibility of the oxygen reduction. Therefore the region

marked by this plateau can not to be considered as delaminated but it indicates just

partially reduced polyaniline. The reduction of the polyaniline advanced as a second

front which separated partially reduced polyaniline from the intact oxidized one. It can

be observed that the reduction progresses much slower than the delamination, which

is completely halted. The thick layer of the conductive polymer stores a large amount

of charge. Hence its reduction demands much more charge comparing to the charge

which is necessary for the delamination of the interface. But much more important: the

polarized area where the coating still adheres to the substrate works as a bottle-neck

for the ion transport to the polymer and limits the rate of the polymer reduction. The

model experiment described above confirmed the existence of the protection zone in

front of the polyaniline pigment. The interface polarization provided by the polymer

managed to stop the oxygen reduction and the progress of the delamination. Even

partial reduction of the conductive polymer did not reduce the protection provided by

the polyaniline, because the potential the PANI is reduced to is still higher than the

potential where oxygen reduction can set in at the chromium surface! The obtained

results show that the proposed mechanism of the interface polarization by ICP can

be successfully utilized for corrosion protection. However, the investigated system was

designed especially to enhance the positive effect of the interface polarization. In the

next section the same experiment will be carried out on iron, where oxygen reduction

can occur up to higher potentials than on chromiun, thusly lowering the protection

potential of the ICP (as the difference between ICP potential and onset potential for

oxygen reduction is smaller).

7.3. Interface polarization by a PANI dot on the iron. 105

500 1000 1500 2000 2500 3000 3500 4000 4500

−400

−200

0

200

400

600

d / µm

E /

mV

Polyaniline + PVB coating

PVB coating

I · R ↑XX

XXy

second frontPPi

(a)

0 50 100 150 200 250 3000

500

1000

1500

2000

2500

3000

3500

t / minE

/ m

V(b)

Figure 7.5: Results of the delamination test performed on the chromium sample, (a)

potential profiles, (b) the progress of the delamination.

Figure 7.6: The experimental setup for the delamination of the iron sample.

7.3 Interface polarization by a PANI dot on the

iron.

The idea of the following experiment is to check whether the investigated mechanism

can work also on electrochemically more active substrates such as iron. The schematic

drawing presenting the experimental setup is showed in figure 7.6. The substrate con-

sisted of an iron rectangular plate with dimensions of 20x10mm which was ground with

grinding paper up to a grade of 1000. Then it was cleaned in ethanol in an ultrasound

bath for 15 minutes. Finally the substrate was dried in a nitrogen flow. Then the

polyaniline and the PVB top coat were applied following the same procedure described

in the former paragraph. Using a conductive paste the prepared sample was glued to

the metal plate which acted as a defect. The defect was placed on the left side of the



sample and filled with a 1M KCL solution employed as electrolyte. The scanning area

was chosen in such a way that allowed to monitor changes of the potential over the

surface covered as well by the PVB coating as by the PANI and the PVB layer. The

scanning begun at the starting point located 500µm from the edge of the sample and

proceeded along the X axis. When the last point of the line was measured, the SKP

tip was moved along the Y axis and started to measure the next line. The distance

between measurement points in the X direction was chosen 50µm and in the Y direc-

tion 100µm. All measurements were performed every 1.5 h. Figure 7.7a shows maps

of the potential obtained after the measurement. Color scales visible in the figure refer

to the measured potential. Three zones of different potential may be easily pointed

out. The first zone, where the potential reached a value of -450 mV indicates the

delaminated area. The second region marked by the plateau of the potential of 100

mV refers to the area where the metal/PVB interface was still intact. The last and

third zone is indicated by the high potential of the polyaniline. As is clearly visible

in figure 7.7a the delamination front which is indicated by the low potential started

to proceed from the defect toward the middle of the sample. The presented figures

give the misleading impression that the delamination front was not homogeneous, i.e.

that the delamination front advanced slower at the region 0-750µm and faster at the

region up to 3000µm. This illusion is caused by the slow scanning rate of the SKP

apparatus. During the time when the SKP managed to finish the scanning of a single

line the delamination front already advanced so much that during scanning of the next

line its position was already significantly shifted. This shows that the delamination

of PVB from iron proceeds at quite high speed, most likley also due to a relatively

low adhesion. The delamination speed decreases with increasing distance from the de-

fect, therefore in the next fig. 7.7b the effect of the slanting delamination front is only

slightly visible and in figure 7.7c disappears completely. The next figures show the

situation when the delamination front reached the PANI. Hence, the protective zone

should be formed and further delamination process inhibited. However, the following

measurements show no effect of a protective zone, although the delamination seems to

come to stop right at the PANI. A smooth gradation of the potential between the de-

fect and the polyaniline border suggest that the whole metal-polymer interface between

the defect and the polyaniline got polarized, however due to the low resistance of the

delaminated interface only slightly. The good galvanic coupling obviously established

between the defect and the polymer results in the fast reduction of the conductive

polymer. The fast reduction of the polyaniline stripe indicates that there is no real

7.3. Interface polarization by a PANI dot on the iron. 107

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 4500 5500 6500 75000

500

1000

1500

2000

2500

3000

d / µm

d / µ

m

(a)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 4500 5500 6500 75000

500

1000

1500

2000

2500

3000

d / µm

d / µ

m

(b)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 4500 5500 6500 75000

500

1000

1500

2000

2500

3000

d / µm

d / µ

m

(c)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 4500 5500 6500 75000

500

1000

1500

2000

2500

3000

d / µm

d / µ

m

(d)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 4500 5500 6500 75000

500

1000

1500

2000

2500

3000

d / µm

d / µ

m

(e)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 4500 5500 6500 75000

500

1000

1500

2000

2500

3000

d / µmd

/ µm

(f)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 4500 5500 6500 75000

500

1000

1500

2000

2500

3000

d / µm

d / µ

m

(g)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 4500 5500 6500 75000

500

1000

1500

2000

2500

3000

d / µm

d / µ

m

(h)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 4500 5500 6500 75000

500

1000

1500

2000

2500

3000

d / µm

d / µ

m

(i)

Figure 7.7: Maps of the potential measured during the delamination of the model clear

coat applied on the iron substrate with a dot of PANI (right hand side). Measurements

were taken every 1.5h.

protection zone formed. On the other hand a certain effect is clearly visible. The

potential profiles presented in fig. 7.8 show the evolution of the potential during the

experiment. The following profiles represent the data measured in the middle of the

sample. Profiles measured in the beginning of the experiment shows that the delami-

nation front progressed very fast. The lack of ohmic drop between the defect and the

delamination front suggest that there is a very weak bonding between the iron and the

PVB coating. Due to the delamination process the PVB coating seems to de-adhere

completely allowing the massive ingress of the electrolyte which entailed high rates of



1000 2000 3000 4000 5000 6000 7000 8000−500

−400

−300

−200

−100

0

100

200

300

400

500

d / µm

E /

mV

Figure 7.8: Profiles of the potential obtained from maps presented on the fig. 7.7 at the

position 1800µm on the Y axis. Note the formation of the IR drop at the delaminated

interface caused by massive discharging of the PANI.

ion transport. Why is a protection zone not observed in this case?. As described above

a good adhesion of the coating to the metal surface is a key condition which has to be

satisfied so that the interface polarization can work. In the case of this experiment the

interface adhesion obviously is too weak as to allow the formation of a protective zone.

Therefore when the delamination front reached the polyaniline the fast reduction of

conductive coating took place. The discharge of the polymer coating is accompanied

by high ionic currents indicated by the significant increase of the ohmic drop across

all the delaminated interface between defect and PANI, despite the low resistance at

this delaminated interface. The good ion transport between the defect and PANI is

responsible for the massive reduction of the polymer. Therefore the weak metal/PVB

interface or rather the good ion mobility along this interface are what is responsible

for the failure of the polarization mechanism.

7.4 Delamination test of the PANI dot on gold.

In order to check that the interface polarization mechanism is able also to inhibit

delamination from metals that show oxygen reduction also at higher potentials than

7.4. Delamination test of the PANI dot on gold. 109

chromium, experiments were also carried out on gold. However, other than the iron

/PVB interface the gold/PVB interface is quite stable. The lack of the OH groups

at the gold surface ensures relatively slow delamination cased by poor incorporation

of water and ions. The gold layers were evaporated on glass substrates according

to the procedure described in the chapter 4. The polyaniline drop was placed on the

substrate using a glass pipe. The border of the drop was situated 2,5 mm from the edge

of substrate. Then the sample was dried in air at the room temperature for 30 minutes.

When the xylene was evaporated, the sample was coated with the 10% ethanol solution

of the PVB. The thus prepared sample was used for the delamination experiment. The

experimental setup and the calibration procedure of the SKP apparatus was already

described in chapter 5. The defect was polarized by potentiostat at a potential of -

440mV vs SHE. In order to mark the area where the delamination occurred the sample

was kept in a humidifier and the electrolyte remained in the defect after the end of the

experiment, but without polarization of the defect. Ions present in the defect diffused

toward the delamination front due to the gradient of the concentration between the

defect and the delaminated area. After a few hours the sample was slowly dried in the

room temperature and KCL crystals appeared under the PVB coating. These crystals

indicate the presence of the electrolyte under the PVB coating. Therefore they act as

a marker of the delaminated area. Figure 7.9 shows the picture of the sample after the

experiment. Two general regions are easily recognized. The first area extends from

the left edge of the sample and surrounds the black spot of polyaniline. The presence

of the KCL crystals confirms that in this region the coating deadhesion caused by

the delamination process took place. The most interesting observation is that the

shape of the boundary of the delaminated area exactly corresponds to the shape of the

polyaniline drop. This indicates that the advancing delamination front was stopped in

front of the polyaniline drop and could not proceed further. The black arrow visible

in fig. 7.9 shows the path which the delamination front had to follow. It is important

to note that the delamination front did not reach the polyaniline but stopped roughly

100µm in front of it. This thin stripe of the not delaminated coating which separates

the polyaniline from the delaminated area seems to act as a protective shield, i.e. this

obviously is the protection zone. The delamination front managed to pass the PANI

dot around and progressed behind it, but the protection provided by the polyaniline

still worked. However, the protection of the zone is not homogeneous around the drop.

In the lower part of the figure there seems to be a weak point at the interface. At the

point marked with the red circle the zone seems to be broken and the delamination



Figure 7.9: The photography of the gold sample coated with the polyaniline dot (black

spot) and PVB coating after the delamination experiment.

front contacts directly the polyaniline. The change of the color of the PANI from black

to brown, visible at the place where the zone was broken, suggests that polyaniline got

reduced at this site. This hypothesis can be confirmed by the potential maps obtained

using SKP for a similar experiment. Fig. 7.10a shows the situation when the progress

of the delamination is stopped by the PANI. The black circle in the middle of the

sample shows the position of the PANI dot. The delaminated area indicated by the

low potential measured by the SKP assumes the shape of the polyaniline drop and

further progress of the delamination front is restrained by the protective zone created

around the polyaniline. Of course the delamination may still advance at areas that

lie out of the range of the polyaniline influence. That may be noticed at the position

corresponding to Y axis coordinates below 500 and above 6000µm. The delamination

front changes its position on every following scan, while at the position of 3500µm

it does not advance. Next figures show the break down of the protective zone. At

the beginning, the weak crack in the protection zone at y=2000 micrometers allows

a direct contact of the delaminated area to the polyaniline. The good ion transport

provided through this crack permits a very fast reduction of the conductive coating.

7.4. Delamination test of the PANI dot on gold. 111

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 4500 55000

1000

2000

3000

4000

5000

6000

d / µm

d / µ

m

(a)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 4500 55000

1000

2000

3000

4000

5000

6000

d / µm

d / µ

m(b)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 4500 55000

1000

2000

3000

4000

5000

6000

d / µm

d / µ

m

(c)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 4500 55000

1000

2000

3000

4000

5000

6000

d / µm

d / µ

m

(d)

Figure 7.10: Maps of the potential obtained during the delamination experiment per-

formed on the gold substrate coated with the polyaniline pigment and the PVB coating.

The position of the polyaniline dot is marked by the black dashed line.

The protective zone which can be maintained only by the high potential of the polymer,

decreases it size so that it finally disappears. This way the small crack in the protection

zone expands and leads to the massive polymer reduction, clearly visible on fig. 7.10d.

However the zone stays still intact at places where the high potential of the polymer

is still present. The reason which leads to the brake down of the protection zone lies

most likely in a local weakness of the bonding between polymer and metal, i.e. the

stability of the protection zone depends on the quality of the metal/polymer interface.

Its local weakness will enhance ion transport and decrease the polarization effect or in

the worst case leads to the creation of a direct connection between the delaminated



area and the PANI. The random occurrence of such a break down areas suggests that

it may be caused by some impurities present on the sample prior the top coat was

applied. The experiment described made the existence of the protection zone ahead of

the conductive coating clearly visible.

7.5 Patterns of the conductive coating.

The results presented above show that the conductive polymer can polarize the metal-

polymer interface and create a protection zone where the oxygen reduction is prohib-

ited. Hence, the delamination process can be slowed down or even completely stopped.

However, if the metal/polymer interface is not good enough, the protective zone can

be broken. Then due to good ionic conductivity at the interface a fast polymer reduc-

tion occurs and leads to the fast progress of the delamination. This high sensitivity

of the protection zone to the quality of the interface is a weak point of the polariza-

tion mechanism. However, this problem can be minimized. This can be achieved by

decreasing the size of the ICP dots and using them in high density, i.e. by using the

conductive coating as a pigment inside the insulating matrix. Separated particles will

provide the protective shields but the insulating coating which separates particles of

the ICP will work as a barrier for the ion transport between particles. A localized

break down of the protection zone will lead to the reduction of just one small parti-

cle of the conductive polymer instead of the whole coating as it is in the case of the

continuous conductive dot. If the size of conductive particles is relatively small com-

pared to the size of the zone created by them, then the effect of the protection would

overcome the fast progress of the delamination after the break down of the protection

zone. The following investigation will focus on the delamination behavior of according

model samples. The scheme of the samples is presented in the figure 7.11 which shows

a pattern of polyaniline dots applied on a gold substrate. Each of the polyaniline dots

should create its own protection zone. If the distance separating individual dots is

smaller than the size of the shielding zone, then the overlapping of protection zones

should provide a continuous protection of the sample and detain the progress of the de-

lamination front. Therefore, the proper spacing of the dots is an important issue which

may decide about failure or success of the protection. The following experiments will

aim on delamination tests performed on samples with different spacing between PANI

dots.

7.5. Patterns of the conductive coating. 113

(a) (b)

Figure 7.11: The idea to use a pattern of conductive polymer, (a) the 3d arrangement

of the conductive particles, (b) the idea of overlapping protection zones.

7.5.1 Preparation of the polyaniline pattern.

The patterned samples were prepared using the setup shown in figure 7.12. The setup

consists of a sample stage driven by Physical Instruments motors which allow the

motion of the sample in all directions. High accuracy of the motors ensured the precise

positioning of the sample during the preparation of the pattern. As substrate gold

layers evaporated on the glass plate were used. The substrates had a rectangular

shape with dimension of 12 x 10 mm. The process of the pattern preparation was

controlled by a specially designed software which allowed an automatic preparation of

the pattern with desired spacing between polyaniline dots. Polyaniline dots were made

using a glass capillary which was prepared by the glass Narishige Model PC-10.After

puling, the tip of the capillary was ground in order to obtain a flat ending. The final

outer diameter of the capillary was around 100µm. Then the capillary was fixed over

the sample stage and filled with xylene dispersion of the polyaniline. A peristaltic

pump was applied in order to provide a constant flow of the polyaniline dispersion

which is necessary to avoid clogging of the capillary by drying polyaniline particles and

ensure a homogeneous size of PANI dots. The pumping speed was chosen to 1l/h. Dots

were prepared according the following procedure: at first, motors adjusted the sample

position in the X and Y axes. Then the sample was moved up until the capillary just

touches the sample surface so the polyaniline could stick to the gold surface. Next the



Figure 7.12: The scheme of the setup for the preparation of the PANI pattern.

sample stage was lowered and moved to the another point. All parameters such as

motors speed, pumping speed and the time when the capillary contacted the sample

surface were chosen and adjusted experimentally. Figure 7.13 shows photographs of

obtained patterns. The shown samples differ in the spacing between the polyaniline

dots. The photographs presented in the figures 7.13a, 7.13b, 7.13c, 7.13d show PANI

patterns with spacings between the centers of the dots of 300µm, 500µm, 750µm and

1000µm respectively. The obtained dots can be classified as reproducible with only

small variation of size and shape. All samples consisted of three areas: a pattern

free area, extending from the edge of the substrate up to 1500µm inside the sample.

A second region was the patterned area, containing a few rows of PANI dots and

extending for the next 1000µm. The third zone was again without the polyaniline

pattern. In the first part of the experiment the delamination will progress on the

pattern free area. Than, after 1500µm it will meet the ICP pattern. At the end of

the experiment the delamination front will enter again the area free of the pattern.

Hence the delamination can be observed on both areas: containing PANI dots and not

containing them. The comparison of the measured speeds of the delamination will give

an answer to the question, if the patterned samples can provide a protection and slow

down the progress of the delamination front.


(a) (b)

(c) (d)

Figure 7.13: Pictures of patterns of the polyaniline on gold substrates, spacing between

centers of dots: (a)-300µm, (b)-500µm,(c)750µm, (d)-1000µm.

7.5.2 Investigation of the delamination properties of the polyani-

line patterns with different spacing between conductive

pigments.

The following investigations will focus on the influence of the spacing between dots of

ICP within a pattern on the overall delamination behavior. Results of the delamination

test performed on the sample with the pattern spacing of 300µm are shown in fig. 7.14.

The time interval between the potential maps was 1h. Black vertical lines indicate

the region where the PANI pattern was applied. At the beginning of the experi-

ment (fig. 7.14a, 7.14b, 7.14c) a homogeneously progressing delamination front can be



−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

200

400

600

800

1000

d / µm

d / µ

m

Pattern

A

B

(a)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

200

400

600

800

1000

d / µm

d / µ

m

Pattern

(b)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

200

400

600

800

1000

d / µm

d / µ

m

Pattern

(c)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

200

400

600

800

1000

d / µm

d / µ

m

Pattern

(d)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

200

400

600

800

1000

d / µm

d / µ

m

Pattern

(e)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

200

400

600

800

1000

d / µmd

/ µm

Pattern

(f)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

200

400

600

800

1000

d / µm

d / µ

m

Pattern

(g)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

200

400

600

800

1000

d / µm

d / µ

m

Pattern

(h)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

200

400

600

800

1000

d / µm

d / µ

m

Pattern

(i)

Figure 7.14: Maps of the potential obtained during the delamination test of the polyani-

line pattern with the spacing between dots of 300µm. Measurements taken every 1h.

Black vertical lines marks the patterned region. The pattern is not resolved by SKP

due to the similarity between the potentials of the area PVB7gold and PVB/PANI.

observed. When the the delamination front reached the patterned region the delam-

ination seemed to stop. In the upper part of fig. 7.14e the beginning of the down of

the protection is visible. The following figures show the progress of the delamination

front indicating ceasing of the protection. However, the delamination front did not

advanced homogeneously. At the lower part of fig. 7.14g the protection seemed to stay

intact and still hinder the delamination progress. The position of the front shifted only

slightly in comparison to fig. 7.14e. Finally in fig. 7.14i one can see that the delami-

nation front advanced quickly on the non patterned area after surrounding the rest of


the intact pattern. The fig. 7.15a presents the potential profiles measured along the

path “A” marked on the fig. 7.14i. The presented curves were measured every 15min.

A significant slow down of the delamination progress which is visible at the position

of 1500µm indicates that the front reached the protection zone. When the protection

was broken fast reduction of the polyaniline dot took place. That resulted in the fast

advance of the the delamination front. However, the delamination front was stopped

again in the front of the next row of dots. The following profiles shows the develop-

ment of the “volley” of the low potential at the position of 3000µm. The drop of the

potential in this region was caused by the delamination front which managed to pass

the patterned zone in the upper part of the sample where the protection was not so

effective. The progressing front started to surround the remaining part of the pattern

where the protection still worked. It has to be noted that there is no indication that

the delamination front was able to rush between the polyaniline dots. This indicates

that the applied spacing of the dots of 300µm is small enough to ensure an overlapping

of the protection zones created by separate dots. Figure 7.15b shows the potential

profiles measured along the path “B”. The delamination front slowed down in front of

the first row of dots but the protection provided by the polyaniline was too weak to

effectively restrain the delamination front. The next rows of dots also did not provide

efficient protection as in the case of the profile presented on the fig. 7.15a. Obviously

one part of the measured sample presented significantly lower resistance against the

delamination, maybe because the dots have been thinner in the upper part than in

the lower part of the sample. There were other indications that the applied amount of

PANI per dot was not very reproducible with the chosen method.

Above measurements confirm the presumption that the method applied for the pro-

duction of the polyaniline pattern does not ensure the necessary homogeneity of the

obtained patterns. The forgoing results confirm that the patterned sample can provide

effective protection against the delamination. However, the quality of the pattern is

an essential factor which determines the durability of the provided protection.

The following figures show maps of the potential obtained for samples with a spacing

between dots of 750µm. Now the positions of the dots are clearly indicated by the

slightly higher potential of the polyaniline. As in the case of the former experiments

the delamination started from the left side of the sample. The homogeneous delam-

ination front progressed with almost constant speed. There was no indication for a

significant slow down of the delamination speed in the patterned area. The presented



500 1000 1500 2000 2500 3000 3500−500

−400

−300

−200

−100

0

100

200

300

400

500

d / µm

E /

mV

(a)

500 1000 1500 2000 2500 3000 3500−500

−400

−300

−200

−100

0

100

200

300

400

500

d / µmE

/ m

V

(b)

Figure 7.15: The potential profiles measured over regions marked on the fig. 7.14i.

Profiles (a),(b) were measured along paths “A”, “B” respectively.

results show that a separation distance of 750µm is definitely too large and cannot

provide any protection. The protection areas created by the individual dots did not

overlap and the progress of the delamination front could not be restrained. Figure

7.17a shows potential profiles measured during the experiment. The presented scans

were measured along the path which crossed the dots from the first and the last row of

the polyaniline pattern. The tracing of this path is marked in fig. 7.16a. All measure-

ments were taken every 30 minutes with the exception of profiles marked with dashed

line. The time interval between them was 1h. As presented in the figure below, the

PANI dots did not slow down the delamination at all. In addition the fast reduction

of the polyaniline enhanced the progress of the delamination. The progress of the de-

lamination presented in fig. 7.17b shows the square root behavior which is typical for

the delamination process. Hence the presence of polyaniline dots did not influence the

overall delamination behavior.

7.5.3 Examination of the delamination properties of the polyani-

line patterns with increased volume of the conductive

polymer.

The above presented results show that the proper spacing between the polyaniline

pigments decides about the failure or the success of the protection provided by the


−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

500

1000

1500

d / µm

d / µ

m

A

(a)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

500

1000

1500

d / µmd

/ µm

(b)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

500

1000

1500

d / µm

d / µ

m

(c)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

500

1000

1500

d / µm

d / µ

m

(d)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

500

1000

1500

d / µm

d / µ

m

(e)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

500

1000

1500

d / µm

d / µ

m(f)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

500

1000

1500

d / µm

d / µ

m

(g)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

500

1000

1500

d / µm

d / µ

m

(h)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

500

1000

1500

d / µm

d / µ

m

(i)

Figure 7.16: Potential maps obtained during the delamination test of the polyaniline

pattern with dot spacing 750µm. Measurements taken every 1h.

conductive polymer. Separated, distant dots were not able to stop the progress of

delamination. Only the collective cooperation of single polyaniline pigments can assure

the inhibition of the delamination. However, the duration of the protection provided

by the polyaniline pattern with dot spacing of 300µm may still be improved. The

endurance of the protection zone should be limited by the charge stored inside the

polyaniline which is necessary to maintain the positive potential in front of the dot.

Therefore an increase of the volume of the single dot should result in an increase of its

durability and as a result the resistance of the whole pattern should be improved. As

already mentioned above lateral dimensions of the single dot should be small so the

reduction of polyaniline dot, caused by the break down of the protection, would not



0 500 1000 1500 2000 2500 3000−500

−400

−300

−200

−100

0

100

200

300

400

500

d / µm

E /

mV

(a)

0 200 400 600 800 1000500

1000

1500

2000

2500

3000

3500

t / mind

/ µm

(b)

Figure 7.17: (a) The potential profiles measured along path “A” marked in fig. 7.16a,

(b) progress of the delamination front.

result in a fast progress of the delamination in the extended PANI dots. Hence, the

increase of the volume should be achieved only by enlarging the height of PANI dots.

At the same time the spacing between dots should be kept as small as possible in order

to assure the overlapping of protection zones. Enlarging of the volume of the PANI dot

was achieved by the repeated application of single polyaniline dots at the same place.

Despite of the precise positioning of the sample during the multiple placement of dots,

their lateral size always slightly increased a little by this procedure. This caused the

joining of single dots during the preparation. Therefore all attempts to produce the

pattern with a spacing of the dots of 300µm failed. The smallest dot spacing allowed by

this preparation method was 500µm. According to the above, two polyaniline patterns

were prepared. First, with a dot spacing of 500µm and the second with a spacing

of 1000µm. The height of the obtained dots was measured using SKP. As shown in

fig. 7.18 the topography of the obtained pattern was not homogeneous. Some of the

dots reached the height of 10µm while others scarcely crossed the height of 1µm, which

confirms that the volume per application was not well controllable and changed steadily

over time, but not erratically. The prepared patterns were coated and dried according

to the procedure described in the former paragraph. Then the protective properties

of both samples were examined by means of the delamination test. Fig. 7.19 shows

the potential maps obtained during the delamination of the sample with a dot spacing

of 1000µm. The positions of polyaniline dots can be easy recognized as circular areas

indicated by the slightly lower potential than for rest of the PVB coating. The fact


0

1

2

3

4

5

6

7

8

9

10

01000

20003000

40005000

0

1000

2000

30000

5

10

d / µmd / µm

Figure 7.18: The topography of the polyaniline pattern with dot spacing of 1000µm.

that the potential of the PANI dots was slightly lower than that of the intact interface

was at first surprising, but as the intact interface has already a very high potential

in all the experiments the potential of PANI and the intact interface were practically

similar. Slight variations in both potentials could lead to the PANI being slightly lower

or at nearly the same value. The delamination front progressed from the left side. The

polyaniline dots seemed to be an obstacle for the advancing front which clearly slowed

down in front of each of them. However the progress of the delamination between PANI

dots did not suffer any obstruction due to the limited range of the influence of the

polyaniline. Therefore the advancing front manged to surround dots which lead to the

formation of “islands” of the high potential shown in fig. 7.19e, confirming the existence

of a protection zone surrounding the particles. The presented results show that dots

with larger volume of the polyaniline can provide better protection as compared to the

patterns examined in the paragraph 7.5.2, but still the range of the protection provided

by the dots is too small to prevent the overall delamination of the coating (no overlap

of protection zones). Figure 7.20 shows maps of the potential obtained during the

delamination experiment performed on the sample with the pattern spacing of 500µm.

Like in the case of the above experiments the defect was placed on the left side of the

sample and the delamination process was initialized by cathodic polarization of the

defect to -440 mV vs. SHE. The direction of the progress of the delamination front is

marked by small arrows visible in figures. Already the first two measurements show

that the delamination front did not progress homogeneously and seemed to slow down

in front of the polyaniline. However, in the area between the dots the delamination



−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 45000

200

400

600

800

1000

1200

d / µm

d / µ

m

(a)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 45000

200

400

600

800

1000

1200

d / µm

d / µ

m

(b)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 45000

200

400

600

800

1000

1200

d / µm

d / µ

m

(c)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 45000

200

400

600

800

1000

1200

d / µm

d / µ

m

(d)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 45000

200

400

600

800

1000

1200

d / µm

d / µ

m

(e)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 45000

200

400

600

800

1000

1200

d / µmd

/ µm

(f)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 45000

200

400

600

800

1000

1200

d / µm

d / µ

m

(g)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 45000

200

400

600

800

1000

1200

d / µm

d / µ

m

(h)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 3500 45000

200

400

600

800

1000

1200

d / µm

d / µ

m

(i)

Figure 7.19: Maps of the potential measured during the delamination experiment mea-

sured over the sample with 1000µm dots spacing. Measurements taken every 1h.

is not slowed down. It is surprising that the observed reduction of the delamination

speed appeared already c.a 500µm from the edge of the polyaniline dot. However, such

a long range effect is also observable in fig. 7.19! Additionally the potential of the gold

surface in the area between the delamination front and the polyaniline dots remained

unchanged and no slope due to polarization can be observed. So there is no evidence

for the galvanic coupling between the delaminated area and the polyaniline. This long

distance effect cannot be explained by the simple interface polarization mechanism

proposed above and will be discussed in detail in the next paragraph. Fig. 7.20d shows

the situation when the delamination front hit the first row of the polyaniline pattern.

The shape of the front clearly indicates that the delamination was not that much slowed


down in the area between dots but was significantly slowed down in front of them. The

delamination front which already passed the area between the dots hit on directly on

the protection zones created by the second row of dots. Then it changed direction in

order to follow the area which was not protected. There is a narrow trail of the high

potential behind dots marked by the red ellipse on the fig. 7.20f. This indicates that

the delamination front started to surround the polyaniline dot. Finally the polyaniline

dots from the second row were completely surrounded by the delaminated area. This

measurements proves that polyaniline dots with increased thickness are a more effective

obstacle than the thinner ones. Each of them managed to slow down the progress of

the delamination front. The spacing between dots seemed to be still too large as the

results presented in fig. 7.20 show that the delamination front may pass aside the PANI

dots.

However, it is not able to pass the next row of the pattern until all dots from the

previous row are completely reduced and do not any more provide protection (see

fig. 7.20g). This conclusion indicates that the overall progress of the delamination

front is restricted by the presence of the pattern of dots. The situation presented in

the fig. 7.20h where the delamination front advances at full speed after leaving the

pattern despite of still active dots behind shows that the spacing was still too wide.

A significant problem which appears during the analysis of the results is the evalua-

tion of the delamination speed. In the case of a typical delamination experiment the

position of the front which separates the intact and the delaminated area is relatively

homogeneous and the evaluation of its progress does not procure any difficulties. Here

the delamination front does not advance homogeneously so its position will be cal-

culated as an average of the whole sample along the Y axis. Figure 7.21 shows the

progress of the delamination measured for three different samples. The reference sam-

ple consisted of a PVB coating. The delamination progress evaluated for this sample

is marked by the black line. The blue one shows the results obtained for the patterned

sample. As shown in the figure the progress of the delamination calculated for the

continuous PANI coating is even higher than for the coating which does not contain

the conductive polymer. The reason for that is the high cation transport through the

coating which was already discussed in the chapter 5. The progress of the delamination

for the patterned sample shows that the delamination front slowed down by a factor

of (3?) in front of the first row. When the protection by the first row has broken

down, the delamination sped up to stop again in front of the second row. After the



−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

200

400

600

800

1000

1200

d / µm

d / µ

m

(a)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

200

400

600

800

1000

1200

d / µm

d / µ

m

(b)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

200

400

600

800

1000

1200

d / µm

d / µ

m

(c)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

200

400

600

800

1000

1200

d / µm

d / µ

m

(d)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

200

400

600

800

1000

1200

d / µm

d / µ

m

(e)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

200

400

600

800

1000

1200

d / µmd

/ µm

(f)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

200

400

600

800

1000

1200

d / µm

d / µ

m

(g)

−500

−400

−300

−200

−100

0

100

200

300

400

500

500 1500 2500 35000

200

400

600

800

1000

1200

d / µm

d / µ

m

(h)

Figure 7.20: Maps of the potential measured during the delamination of the sample

with the spacing of polyaniline dots of 500µm. Measurements taken every 1h.

front passed second row it accelerated again. The third row did not have a measurable

effect, most likely because it contained too little PANI.

The presented measurements show the positive effect of the polyaniline pattern. The

progress of the delamination slowed down significantly in front of the conductive par-

ticles. The size of the observed protection zone was surprisingly large. The slow down

of the delamination can be observed even a few hundreds micrometers in front of the

polyaniline dots. Additionally, the potential profiles measured by the SKP do not show

any potential changes in the protection zones surrounding the polyaniline. This does

not look like an active polarization of the interface, or at least there is no indication

for this. How can this be explained?


Figure 7.21: The progress of the delamination for measured for: reference sample

coated with the PVB, continuous layer of polyaniline plus a layer of the PVB, and the

sample with the polyaniline pattern plus a coating of PVB. The green rectangles mark

the position of the rows of the PANI dots.

The model of the interface polarization is based on the assumption that the galvanic

coupling between the defect and the conductive particle is established by the migration

of ions. The driving force for this process is constituted by the potential difference

between the defect and the intact area. However, the migration is not the only method

by which ions may be transported along the polymer-metal interface. According to the

Nernst-Einstein equation which is shown below:

ji = kici

[

ziF∂φ

∂x+ RT

(

∂ln ci

∂x

)]

(7.1)

ji - flux, ci - concentration of species i, T - absolute temperature, R - gas constant, zi - charge on ion

i, F - Faraday constant, x - distance in direction x, φ - potential in volts.

The rate of the ion transport depends of two terms. The first one represents the

migration of ions induced by the potential gradient. During the delamination the

migration of ions helps to keep the charge neutrality of the interface at sites where the

oxygen reduction takes place. The other term refers to the diffusion of ions caused by

the concentration gradient. As shown Leng [24] diffusion of ions takes place during



the delamination experiment performed in the nitrogen atmosphere. Diffusion of ions

establishes the galvanic coupling between the defect and the oxide layer which is present

on the metal surface. A diffusion front moves ahead of the delamination and creates

an ion-enriched zone ahead of the delamination front. The selective ingress of cations

into the interface leads to a decrease of potential and hence is an important step for the

further advance of the delamination. This ion diffusion does not influence the adhesion

of the coating [24]. As during its slow reduction the conducting polymer acts as a drain

for cations and a source of released anions, its presence ensures a high concentration

gradient in the protection zone. This gradient could be enough to provide the ion flow

required to prevent the pulling down of the interracial potential by ingress of cations

from the defect; or at least just a very small IR drop may be needed. That could

help to explain observed effect of the long range influence of the conductive pigments

without notable potential slopes.

In the case of the experiment described in the paragraph 7.5.3 the potential of the

intact gold surface was around 500mV vs. SHE. At this potential the reduction of the

oxygen is inhibited. Therefore the delamination of the coating has to be initialized by

the previous reduction of this surface potential. Hence the sufficient protection may

be achieved just by maintaining the initial potential of the gold surface. There is no

need to polarize it more positively. Very low ionic current flowing to the conductive

particle may result in a very small IR drop which might be undetected by the Kelvin

Probe. Thus the conductive particle may act as a drain for cations which consumes

all available cations for its own reduction and prevents the decrease of the potential in

the protection zone around it. Figure 7.22 illustrates the idea of the ion draining by

the conductive pigment.

7.6 On the diffusion at polymer—metal interfaces:

an investigation on polymer—iron.

One puzzling observation made in the experiments described above is that no IR drop

was observable in the protection zone surrounding the PANI dots on gold. Although the

potential difference between PANI and gold is small, a was to be expected. This is not

the case, meaning that the required ion flow along the interface, either the scavenging

7.6. On the diffusion at polymer—metal interfaces: an investigation onpolymer—iron. 127

Figure 7.22: Draining of cations by the conductive polymer prevents pull down of the

potential due to ingress of cations. Also the release of anions from the ICP (source)

would ensure an inhibition of a decrease of potential due to cation ingress from the

delaminated area into the protection zone by counterbalancing the charge by anions.

of cations by the ICP or the neutralization of the them by released anions, is much

smaller than it could be.

As already pointed out above, maybe diffusion in the concentration gradient between

delaminated area and the PANI site, possibly in combination with some migration in

a very small potential gradient, ensures the necessary coupling to prevent the pulling

down of the potential in the protection zone. However, this hypothesis accepts one

puzzling fact: the ion flow could be higher as the driving force of a higher potential

gradient is not made use of. On the other hand an equivalent observation was made

by Leng et al. [24] for the diffusion/migration of cations from a defect into the poly-

mer/metal interface in nitrogen atmosphere. Since under this condition the interface

is not delaminated the resistance against ion diffusion/migration remains high at the

interface and should be about the same at all sites where ions have progressed into.

As the potential at the intact interface is higher than at the defect and the ion con-

centration is zero, there are two driving gradients for the cation movement: an electric

potential and a concentration gradient.

At the front the inwards moving cations lead to a partial reduction of highly oxidized

states in the passive layer, which results in lowering of the potential. As the cation

migration occurs along an interface of same resistance one would assume that the

potential profiles should develop as shown in fig. 7.23a. However, what is observed is

shown in fig. 7.23b, i.e. the potential profile evolution looks just like the one during

delamination. For the case of delamination this makes sense: where the interface

is delaminated high ion mobility exists and the delaminated area is in direct Ohmic



(a) (b)

(c) (d)

Figure 7.23: Evolution of the potential profile due to the diffusion of ions along the

interface. Diffusion of ions along the metal/insulating coating interface: (a) expected

potential profile, linearly increasing potential, (b) the observed shape of the potential

profile. Evolution of the potential profile after an establishment of the galvanic coupling

between the conductive particle and the defect: (c) linearly increasing potential in the

protection zone (d) the observed potential profiles after coupling to the conductive

pigment.

contact with the electrolyte in the defect. Hence, the potential is pulled down to

the defect potential and the ion migration along this interface does not cause a high

potential drop (IR drop). However, in nitrogen atmosphere this is not the case and

even a very small cation migration would cause a potential drop along the interface.

The fact that the potential is pulled down in the area where cations have progressed

into and is flat suggests that the ion flow is much lower than it could be. The reason

is unknown yet and remains subject for further extensive research beyond the scope of

this work.


It is just emphasized that the observation made in the protection zone is in agreement

with this phenomenon already observed by Leng et al.: in the protection zone the

interface is not yet delaminated and the resistance is still high. Along the interface a

small ion current has to flow in order to keep up the potential. Just as discussed above,

no potential drop is caused by this, indicating that this flow is supported either by just

diffusion and or migration in a much smaller gradient than it could be. Hence, not

profiles like shown in fig. 7.23c are observed, but rather ones like shown in fig. 7.23d.

In order to get more information about this phenomenon (and to check the observation

inside the protection zones around PANI on gold also for iron, in order to be directly

able to compare with the Leng results) a model experiment was carried out for a

polymer coated iron oxide sample partly covered by PANI. The sample was placed in

nitrogen atmosphere and brought into contact with a negatively polarized defect.

Bulk iron is not suitable for extended experiments as it might react with the ICP (i.e.

the iron might start to corrode while the ICP is reduced). This was observed several

times and can be coped only by use of inhibitor anions as dopants to the ICP. Thus

for this basic experiment in order to avoid the interaction between the polyaniline and

the bulk iron a special sample was prepared. The substrate consisted of a platinum

plate coated with a 50µm thick iron oxide layer. The procedure of the preparation of

the oxide was described in detail by Asteman [161]. The polyaniline drop was placed

on the oxide layer in a distance of 2000µm from the edge of the sample. Then the

whole sample was coated with the 10% methanol solution of the PVB. The PVB layer

was prepared using the spincoater at the rotation of 2000RMP for 20 seconds. The

potential of the defect was adjusted using a potentiostat like in the case of former

measurements. A value of 0V vs. SHE which was chosen as a polarization potential,

which should exclude the possiblity of oxygen reduction (by traces of oxygen in the

nitrogen atmosphere; as these experiments required several days this precaution was

necessary in order to exclude any delamination of the interface). A constant nitrogen

flow through the SKP chamber was kept during the experiment. The fig. 7.24 shows

the obtained profiles of the potential. The black vertical line indicates the position of

the edge of the polyaniline dot.

The presented potential profiles were measured every 12 hours. As in the case described

by Leng [24] the diffusion/migration of ions from the defect along the interface resulted

in potential profiles similar to the ones observed during delamination, although no



0 500 1000 1500 2000 2500−100

0

100

200

300

400

500

600

d / µm

E /

mV

Figure 7.24: Potential profiles obtained during the diffusion experiment performed on

the iron oxide in nitrogen atmosphere. The red line indicates the potential profile

which was measured just after the introduction of oxygen. The black line indicates the

position of the edge of the polyaniline dot.

oxygen is available and the adhesion stays intact. Incorporated ions couple the oxide

with the defect and allow the partial reduction of the oxide (i.e. they caused an increase

in the Fe2+/Fe3+ ratio). This causes a cathodic shift of the Fermi level of the oxide.

This causes the plateau around 0 mV, where the potential has not yet reached 0mV not

enough cations have ingressed yet. Where the potential is still unaltered, no cations

have reached yet. Hence, the moving front indicates the progress of the cations. As

shown on the fig. 7.24 the diffusion of ions progressed as a moving front with a constant

speed. However in the vicinity of the polyaniline the progress slowed down and finally

it stopped completely at 250µm in front of the polyaniline. Also the slope of the front

increased slightly. The gradual decrease of the potential of the polyaniline, which is a

result of the reduction of the polymer, confirms that the coupling between the polymer

and the defect was established. However, a polarization of the interface could not be

observed. This means that the ionic current flowing to the polymer has to be very low

but large enough to restrain the reduction of the iron oxide and the further progress

of the diffusion front. The red line shows the situation after the introduction of the

oxygen to the chamber. The presence of the oxygen in the atmosphere is assumed

to have caused the reoxidation of the iron oxide and subsequent change of the ratioFe2+

Fe3+ . The shape of the red curve gives the impression that the potential increased

with the distance from the defect. This effect is caused by the low scanning speed of


the SKP. The scan started at the defect area and continued toward the polyaniline. In

the meantime the iron oxide was oxidizing and its potential was increasing during the

scanning.

Again the observation is not as shown in fig. 7.23a, confirming the above discussed.

7.6.1 Diffusion of anions and cations.

The bending of energy levels which arise on the metal surface influences the electron

transfer from the bulk of the metal to the surface. As discussed above diffusing cations

and anions may allow changes of th oxidation states at the metal surface and modu-

late the band bending. Hence, the electron transfer through the oxide layer may be

promoted or blocked. In consequence, the delamination process may stay inhibited

or enhanced. Therefore further investigations of the ion diffusion may help to find

additional ways for further corrosion protection. Especially in the case of conductive

polymers which are doped with anions. These anions may be released during the reduc-

tion of the polymer and their influence on the delamination process could be significant.

The concept of intelligent “self healing” mechanism for corrosion protection assumes,

that anions stored in the polymer matrix can be released and stop the corrosion process

at the defect. In order to study the individual aspects of anions and cations, anions

and cations should remain separated during investigations. The process of the diffusion

is difficult to control, however the isolation of anions and cations may be achieved by

the choice of ions with different mobility. For example as electrolyte, which will act as

a source of ions, a solution containing small (mobile) cations and large (not mobile)

anions can be applied. This will ensure the fast diffusion of cations and very slow

diffusion of anions. On the other hand the solution which is consisted of small anions

and large cations will allow to investigate the effect of diffusing anions. Figure 7.25

shows the idea of the diffusion experiment and expected results. The sample consisted

of an iron plate coated with insulating lacquer. On the left side of the sample the elec-

trolyte reservoir has to be placed. The electrolyte has to have a direct contact to the

metal-polymer interface so ions can diffuse through it. The potential of the interface is

monitored by the SKP. The oxygen free atmosphere will exclude the possibility of the

reoxidation of the passive layer, so its state could be observed by SKP. The left part of

the figure refers to the situation when the defect is filled with the solution containing

the small cations and large and not mobile anions. Quickly diffusing cations couple the



Figure 7.25: The influence of the diffusing ions to the surface potential, (a) diffusion

primarily of anions, (b) diffusion primarily of cations.

oxide to the defect and initialize its reduction. As a result of that the Fermi level of the

oxide should shift and lower the potential of the oxide. The change of the potential can

be recorded by SKP. The expected potential profile is presented in the lower part of

the figure. The front which separates the low and the high potential indicates position

which has been reached by the moving cations, as already discussed above. Figure

7.25b shows the situation when the applied electrolyte consists of small anions and

large cations. In contrast to the first case diffusing anions should increase the surface

dipole and therefore the work function should slightly increase. So the potential profile

measured by the SKP should have a small peak which will indicate the position of

diffusing anions. Cations which slowly follow anions will enable the reduction of the

oxide and the subsequent cathodic shift of the potential. However, the potential front

should be significantly smeared out comparing to the case of the diffusion of small

cations. The potential profile should take a shape which is presented in the lower part

of the fig. 7.25b

The theoretical model described above was proved experimentally. The sample used in

the experiment consisted of the iron plate which was ground with the grinding paper

up to grade of 1000 and cleaned with ethanol. The PVB coating used in former exper-

iments presents very weak adhesion to the iron substrate. A very fast progress of the

delamination measured in the case of this coating (paragraph 7.3) suggests that the

diffusion is very quick and its observation may meet difficulties. Therefore the MS-Top

coat lacquer was chosen as a top coat. Its adhesion is much better and the diffusion


progresses with low enough speed which allows the measurement. The lacquer was

spincoated according to parameters used in the chapter 5. The potential of the defect

was adjusted by the potentiostat according to the procedure described above. Experi-

ments were performed in nitrogen atmosphere in order to avoid the delamination of the

coating. In the first experiment the diffusion of sodium cations was investigated. As

an electrolyte the 0.1M tosylate sodium salt was used. Large tosylic anions which has

a low mobility will diffuse much slower than small sodium cations so their influence can

be neglected. The fig. 7.26a presents potential profiles obtained during the experiment.

All measurements were taken every 12 hours. The first three scans show the change of

the potential over the whole sample. This indicates the reduction of the oxide due to

incorporation of cations (and the absence of oxygen which would keep up the poten-

tial). A potential front is clearly seen close to the defect. The following measurements

show the progress of the diffusion front which advances with almost constant speed.

In the meantime the potential over the rest of the sample slightly decreased due to

slow reduction of the oxide. Three regions can be distinguished in the measured pro-

files of the potential. The first plateau indicates completely reduced oxide. Then the

potential front can be recognized. It separates the area which is reduced down to the

defect potential from the intact interface. The last region is the intact passive layer.

The shape of the obtained profile corresponds to the presented model (fig. 7.25) and

confirms the proposed assumptions. Figure 7.26b shows the potential profiles obtained

in the experiment dedicated to the investigation of the diffusion of anions. In order to

reduce the influence of diffusing cations tetrabutylammonium chloride salt was used as

electrolyte. A large tetrabutylammonium cations diffuses much slower than the small

chloride anions. Therefore, like in the case of the former experiment, the influence of

less mobile ion was reduced. The first two measurements show the reduction of the

oxide over the whole sample like in the case of the former experiment. The difference

appears from the second scan. The potential profile seems to consist of many steps

and small plateaus which constitute a kind of the potential front. However, it is very

shallow comparing to the front measured in the case of the diffusion of sodium. This

shows that the decrease of the potential which is a result of the migration of cations is

partially compensated by the fast diffusion of anions which raised the measured poten-

tial, leading to the smeared out potential front. This shows that indeed anion diffusion

can buffer the reducing effect of cation ingress. Doping the conductive polymer with

mobile anions that can be easily expelled from the polymer matrix during the reduction

of the polymer would have exactly this effect.



0 1000 2000 3000 4000−500

−400

−300

−200

−100

0

d / µm

E /

mV

(a)

0 1000 2000 3000 4000−500

−400

−300

−200

−100

0

d / µmE

/ m

V

(b)

Figure 7.26: Diffusion of anions and cations in the nitrogen atmosphere. (a) 0.1M

tosylate sodium salt solution in the defect, (b) 0.1M tetrabutylammonium chloride salt

solution in the defect. SKP profiles were measured every 12 hours.

The measured profiles do not show the expected increase of the potential at the front of

the diffusion. This might be caused by the small increase of the surface dipole induced

by anions or by the weak lateral resolution of the Kelvin Probe. Despite of this the

shape of measured profiles in general resembles and confirms the theoretical model.

7.7 Conclusion

The results presented above show that a suitable application of the conductive polymer

may result in a significant increase of the corrosion protection. It was proved that in

the surrounding of the polyaniline pigment a pulling down at the potential of the

interface can be prevented for a long time. Despite of the approaching delamination

the protective zone created by the conductive particle is able to hinder the decrease

of the surface potential and the subsequent oxygen reduction. It was found that the

protective zone is created as a consequence of the galvanic coupling between the defect

and the polymer. The galvanic coupling is realized by ions which diffuse along the intact

polymer-metal interface. The reduction of the polymer has to be accompanied by the

cation incorporation or anion expulsion. Both of which prevents the pulling down of the

electrode potential in the protection zone. The preserved high potential of the surface

inhibits the oxygen process and further progress of the delamination of the coating.

7.7. Conclusion 135

The durability of the protection depends on the ability of the polymer to consume

cations or to release anions. This ability decreases with the gradual reduction of the

polymer. Finally, the partially reduced polymer can not longer keep up the potential

and ingressing cations initialize a decrease in potential. Oxygen reduction can take

place and delamination can proceed. Another very important factor which influences

the durability of the protection zone is the quality of the metal-polymer interface. Weak

interfaces which allow high rates of the ion transport decrease significantly the time

of the provided protection or even exclude the creation of the zone. The durability

of the protection can be increased by enlarging the volume and the oxidation state

of the conductive particle. However, the lateral dimensions of the pigment should be

minimized in order to reduce the negative result of the break down of the protection

once the ICP is spent. It was proved that the collective cooperation of protective

zones created by neighboring pigments rises the resistivity of the whole pattern to

the delamination. For PVB/gold this effect can be observed, if the distance between

edges of particles does not exceed 300µm. In the case of the larger spacing between

pigments the delamination front can manage to circumvent the conductive pigment and

the particle gets quickly reduced as a consequence of the global interface break-down.

Diffusion experiments showed that the type of the doping anion, which is present in the

polymer matrix, may enhance the polarization effect. Small and mobile anions, which

could be released from the polymer, can increase or at least maintain the surface dipole

and in the consequence balance the influence of diffusing cations. Thus the reduction

of the potential of the interface and subsequent delamination can be hindered.

Chapter 8

Main conclusions and outlook

The investigations presented in the former chapters showed that the application of

conductive polymers as corrosion inhibiting paint meets significant difficulties. The

examined mechanisms for corrosion protection by ICPs appeared to be valid only un-

der very special experimental conditions which are much different from real corrosion

conditions.

The immersion test which is described in detail in chapter 4 showed that PPy has

limited ability to passivate defects in the coating. This is due to the fact that only the

small volume of the polymer which is galvanically coupled with the defect is involved

in the passivation process. Only the charge which is stored in the active part of the

polymer can be released for the passivation. Hence, in the case of large defects which

demand large currents for the passivation, the currents delivered by the ICP may

not be sufficiently large. This is especially true for the case of atmospheric corrosion

conditions. Therefore only very small pinholes can be passivated by conductive coating.

During repetitive discharging of PPy, the polymer showed an ability to reoxidize when

the galvanic coupling between the defect and the polymer was broken. Hence, there

is chance that this feature of ICPs can help of provide the protection to pinholes

during wet-dry cycling. In the wet period of the cycle the ICP supplies the passivation

currents but in the dry period it can recover its charge and provide a protection during

subsequent wet cycle.

Investigations of the ion transport through the conductive coating presented in chap-

ter 7 showed that when the ionic transport occurs over large distances ICPs generally

become cation permselective. Cations migrate through the polymer by hopping be-

138 8. Main conclusions and outlook

tween the fixed anions remaining in the polymer matrix. Good ionic conductivity of

ICPs results in acceleration of the kinetics of the polymer reduction. This property of

conductive polymers is responsible for the massive reduction of the conductive coating

during delamination test. Moreover, this cation permselectivity suppresses the wanted

release of anions from the polymer required for the “self healing” of the coating.

Despite of difficulties related to the application of ICPs as a corrosion protective paint

the utilization of ICP for corrosion protection is still possible. However the above

issues concerning the ion transport have to be solved. This problems can be overcome

if the polymer is applied in the form of conductive pattern or dispersion of conductive

particles within an insulating matrix, preventing too macroscopic percolation networks

of the conducting polymer. As it was shown in the chapter 7 separated “islands”

of conductive polymer are able to polarize the metal-polymer interface and create

the “protection zone” which significantly inhibits the progress of the delamination

of the coating. At the same time separation of conductive particles by insulating

matrix hinders the fast cation transport through the coating and excludes the rapid

discharge of the conductive coating as it is in the case of the continuous conductive

coatings. Measurements presented in the chapter 7 showed that the application of

conductive pattern can significantly increase the corrosion protection provided by ICP.

Therefore further research should aim on the optimization of the filling of coatings by

the conductive pigments with respect to the pigment lateral size, their volume and

spacing between pigments.

Bibliography

[1] U.R. Evans, Nature, 206 (1965) 980.

[2] U.R. Evans, Proc. Cambridge Phil. Soc., 22(1924) 54.

[3] N.-G. Vannerberg, T. Syderberger, Corr.Sci. 10 (1970) 43.

[4] G. Lorin, La phospahtation des Metaux Edi-tions Eyrolles, Paris, 1973

[5] G.W. Jernstedt, Trans. Electrochem. Soc.83 (1943) 361.

[6] H. Kaesche, Metallic Corrosion, NationalAssociation of Corrosion Engineers, 1440South Creek Drive, Houston, Texas 77084.

[7] T. Biestek, J. Weber, Electrolytic andChemical Converision Coatings, PressLimited-Redhill, Surrey, UK, 1976.

[8] S. Wernick, R. Pinner, P.G Sheasby, Chem-ical converion coatings, The Surface Treat-ment and Finishning of Aluminium and itsAlloys, 5th ed.,ASM Internation, Metals,Park, Ohio, 1987, pp. 220-284, Vol. 1.

[9] W. Marchand, Electroplat. Met. Finish. 14(1961) 43

[10] S. Spring, K. Woods, Met. Finish., 79 198149

[11] J. Zaho, G.S. Frankel, R.L. McCreery, J.Electrochem. Soc. 145 (1998) 228

[12] N. Sato, Corrosion 45 1989 354

[13] N. sato, Corros. Sci. 31 1990 1.

[14] L. Xia, E. Akiyama, G. Frankel et. al., J.Electrochem Soc. 147 (2000) 2556.

[15] P.O. O’Brien, A. Kortenkamp, Trans. Met.Chem. 20 (1995) 636.

[16] IARC, Monograph on the Evaluation of theCrcinogenic Risk of Chemicals to Humans,Chromium, Nickel and Welding, Interna-tional Agency for the Research on Cancer,Lyon, 1990, Vol. 49

[17] Y. Suzuki, K. Fukuda, Arch. Toxicol. 64(1990) 169.

[18] K. Ogle, R.G. Buchheit, Encyclopedia ofElectrochemistry, Volume 4, Corrosion andOxides Films, 460

[19] J. Boxall, J.A. Fanhofer, S.C. Warren, J. OilCol Chem. Assoc. 55 (1972) 24

[20] J.O. Mayne, The Mechanism of the pro-tective Action of Paints in Corrosion (Ed.:L.L.,Dhreir) 2nd ed., Newnes-Butterworths,London, 1979 Vol.II

[21] C.A. Kumins, Off. Digest 34 (1962) 843.

[22] G. Grundmeier, A. Simmoes, Encyclopediaof Electrochemistry, Volume 4, 500

[23] A. Leng, H. Streckel, M. Stratmann, Corr.Sci. 41 (1998) p. 547-578

[24] A. Leng, H. Streckel, M. Stratmann, Corr.Sci. 41 (1998) p. 579-597

[25] A. Leng, H. Streckel, K. Hofmann, M. Strat-mann, Corr. Sci. 41 (1998) p. 599-620

[26] A.J. Heeger, Rev. Mod. Phys. 73 (2001) 681.

[27] A.G. MacDiarmid, Rev. Mod. Phys. 73(2001) 701.

[28] H. shirakawa, Rev. Mod. Phys. 73 (2001)713.

140 BIBLIOGRAPHY

[29] S. Geetha, Chepuri R.K. Rao, M. Vijayan,D.C. Trivedi, Anal. Chim. Acta 568 (2006)119-125.

[30] A. Malinauskas, R. Garjonyte, R.Mazeikiene, I. Jureviciute, Talanta 64(2004) 121-129.

[31] J.H. Burroughes et al., Nature (1990) 347,p. 539-541.

[32] D.D.C. Bradley et al., Synth. Met. (1991)43, p. 3135-3141.

[33] D. Braun, Mater. Today 5 (2000) 33-39

[34] D. Nicolas-Debarnot, F. Poncin-Epaillard,Anal. Chimi. Acta 475 (2003) 1-15.

[35] J.-M. Breguet, S. Johansson, W. Driesen1,U. Simu, 10th International Conference onNew Actuators (Actuator 2006), pages 374-381, 2006, Bremen, Germany, 14-16 June2006.

[36] J.D.W. Madden, B. Schmid, M. Hechinger,S.R. Lafontaine, P.G.A. Madden, F.S.Hover, R. Kimball and I.W. Hunter, IEEEJournal of Oceanic Engineering 2 (2004) 1-10

[37] J. Pettersson, B. Ramsey, D. Harrison, J.Pow. Sour. 157 (2006) 28-34.

[38] M. Nechtstein, Phys. Rev. Lett., 44 (1980)356.

[39] H. Shirakawa, L. MacDiarmid, C.K.Chiang,A.J. Heeger, J. Chem. Soc., Chem. Com-mun. 16 (1977) 578-579;

[40] H. Naarmann, Electronic Properties ofConjugated Polymers, (H. Kuzmany, M.Mehring and S. Roth (Eds.), Springer-Verlag, Sol. State Sci., 76 (1987) 12.

[41] A. F. Diaz, J. Electrochem. Soc., 123 (1989)115.

[42] G. Tourillon and F. Garnier, J. Electroanal.Chem., 135 (1982) 173.

[43] R. J. Waltman, J. Bargon and A. F. Diaz,J. Phys. Chem., 87 (1983) 1459.

[44] T. Yamamoto, K. Sanechika and A. Ya-mamoto, J. Polym. Sci., Polym. Lett. Ed.,18 (1980) 9.

[45] G. Tourillon and F. Garnier, J. Electrochem.Soc., 130 (1983) 2042.

[46] A. G. Mc Diarmid, Polym. Prep. Am. Chem.Soc. Div. Polym. Chem., 25 (1984) 248.

[47] E. Genies and C. Tsintavis, J. Electroanal.Chem., 195 (1985) 109.

[48] A. F. Diaz, J. Castillo, K. K. Kanazawa, J.A. Logan, M. Salmon and O. Fajardo, J.Electroanal. Chem., 133 (1982) 233.

[49] J. Bargon, S. Mohmand and R. Waltman,Mol. Cryst. Liq. Cryst., 93 (1983) 279.

[50] D. M. Ivory, J. Chem. Phys., 71 (1979) 1506.

[51] M. Satoh, K. Kaneto and K. Yoshino, J.Chem. Soc., Chem. Commun., (1985) 1629.

[52] J. L. Bredas, R. R. Chance and R. Silbey,Phys. Rev., Part B., B26(10) (1982) 5843.

[53] E. M. Genies, G. Bidan and A. F. Diaz, J.Electroanal. Chem., 149 (1983) 101-113.

[54] R.J. Waltman, J. Bargon, A.F. Diaz, J.Phys. Chem. 87 (1983) 1459-1463.

[55] Y. Wei, G-W Jang, C-C Chan, J. Polym.Sci. Part C: Polym Lett 28 (1990) 219-225.

[56] Y. Wei, C-C Chan, J. Tian, GW Jang, KFHsueh, Chem. Mater. 3 (1991) 888-897.

[57] J. Ouyang, Y. Li, Polymer 38 (1997) 1971-1976.

[58] N.V. Blinowa, J. Stejskal, M. Trochova, J.Prokes, M. Omastova, Eur. Pol. J. 43 (2007)2331-2341

[59] F.R.Diaz, C.O. Sanchez, M.A. del valle,L.H. Tagle, J.C. Bernede and Y. Tregouet,Synth. Met. 92 (1998) 99-106

[60] Yangyong Wang, Xinli Jing, Junhua Kong,Synt. Met. 157 (2007) 269-275

[61] C.A. Ferreira, S. Aeiyach, J.J. Aron andP.C. Lacaze, Electrochimica Acta 41 (1996)1801-1809.

[62] Ramakrishnan Rajagopalan, Jude O. Iroh,Electrochim. Acta 46 (2001) 2443-2455.

BIBLIOGRAPHY 141

[63] J.L. Camalet, J.C Lacroix, S. Aeiyach, K.Chane-Ching, P.C. Lacaze, J. Electroanal.Chem 416 (1996) 179-182

[64] J.L. Camalet, J.C Lacroix, S. Aeiyach, K.Chane-Ching, P.C Lacaze, Synth. Met. 93(1998) 133-142.

[65] A. De Bruyne, J.L. Delplancke, R. Winand,Surf. Coat. Tech. 99 (1998) 118-124.

[66] G. Zotti, G. Schiavon, Electrochim. Acta 34(1989) 881-884.

[67] S. Kuwabata, J.Nakamura, H. Yoneyama, J.Chem. Soc. Chem. Commun., (1998) 799.

[68] W.K. Lu, R.L. Elsenbaumer, B. WesslingSynth Met (1995) 71:2163-2166

[69] P.J. Kinlen, D.C. Silverman, C.R. Jeffereys,Synthetic Metals 85 (1997) 1327-1332

[70] J.O. Iroh, W. Su, electrochimica Acta 46(2000) 15-45 Electrochemistry Communica-tions 1 (1999) 180-183

[71] T.D. Nguyen, M. Keddam, H. Takenouti,Electrchem. and Solid-State Letters 6 (2003)B25-B28

[72] P.T. Nguyen, U. Rammelt, and W. Plieth,Electrochim. Acta 50 (2005) 1757.

[73] B. Wessling, Mater. Corr. 47 (1996) 439.

[74] M. Rohwerder and M. Stratmann, MRSBull. 24 (1999) 43.

[75] M. Kendig, M. Hon, L. Warren, Progres inOrganic Coatings 47 (2003) 183-189

[76] J. Reut, A. Opik, and K. Idla, Synth. Metals102 (1999) 1392.

[77] P. T. Nguyen, U. Rammelt, and W. Plieth,Electrochim. Acta 50 (2005) 1757.

[78] R. J. Holness, G. Williams, D. A. Worsley,and H. N. McMurray, J. Electrochem. Soc.,152(2) B73-B81 (2005)

[79] M. Stratmann, H. Streckel, Corr. Sci. 30(1990) 681.

[80] M. Stratmann, H. Streckel, Corr. Sci. 30(1990) 697.

[81] M. Stratmann, H. Streckel, K.T. Kim, S.Crockett, Corr. Sci. 30 (1990) 715.

[82] I.D. Baikie, S. Mackenzie, P.J.Z. Estrup,J.A. Meyer, Rev. Sci. Instrum. 62 (1991)1326.

[83] G.S. Frankel, M. Stratmann, M. Rohwerder,A. Michalik, B. Maier, J. Dora, M. Wicinski,Corr. Sci. 49 (2007) 2021-2036

[84] J.Reut, A.Opic, K. Idla, Synthetic Metals102 (1999) 1392-1393

[85] Yongfang Li, Bohua Deng, Gufeng He, J.App. Pol. Sci. 79 (2001) 350-355.

[86] F.Beck, M.Dahlhaus, J. Electroanal. Chem.,357 (1993) 289-300.

[87] Ming Zhou, Jurgen Heinze, ElectrochimicaActa, 44 (1999) 1733-1748.

[88] Yongfang Li, Electrochimica Acta, 42 (1997)203-210.

[89] T. Raudsepp, M. Marandi, T. Tamm, Elec-trochimica Acta (in press).

[90] J. Tamm, A. Hallik, A. Alumaa, V. Sam-melseg, Electrochimica Acta 42 (1997) 2929-2934.

[91] K. Naoi, M. Lien, W.H. Smyrl, J. Elec-trochem Soc. 1991, 138, 440.

[92] Yongfang Li, Jianyong Ouyang, Jing Yang,Synth. Met. 74 (1995) 49-53

[93] J. Heinze, M. Dietrich and J. Mortensen,Makromol. Chem., Macromoi. Symp. 8, 73(1987).

[94] W. Wemet and G. Wegner, Makromol.Chem. 188 (1987) 1465 .

[95] F. Beck, P. Braun and F. Schloten, J. Elec-troanal. Chem. 267, (1989) 141 .

[96] Y. Li and R. Qian, Synth. Met. 28, (1989)Cl27 .

[97] F. Beck and M. Oberst, Makromol. Chem.,Macromol. Symp., 8 (1987) 97.

[98] P. Pfluger, M. Krounbi, G.B. Street, G.Weiser, J. Chem. Phys. 78 (1983) 3212.

[99] J.G. Eaves, H.S. Munro, D. Parker, Polym.Commun. 28 (1987) 39.

142 BIBLIOGRAPHY

[100] P. Pfluger, G.B. Street, J. Chem. Phys. 80(1984) 544.

[101] J. Joo, J.K. Lee, J.S. Baeck, K.H. Kim, E.J.Oh, J. Epstein, Synth. Met. 117 (2001) 45.

[102] D. Schmeisser, H. Naarmann, W. Gopel,Synth. Met. 59 (1993) 211.

[103] L. Ruangchuay, J. Schwank, Anavat Sirivat,Appl. Surf. Sci. 199 (2002) 128-137

[104] PHi MultiPak Software Manual, version8.1, ULVAC-PHI, Inc. 370 Enzo, Chi-gasaki City, Hanagawa Prefecture, 253-0084, Japan, 2006.

[105] C.D. Wagner, W.M. Riggs, L.E. Davis, J.F.Moulder, G.E. Mullenberg, Handbook of X-Ray Photoelectron Spectroscopy, PhysicalElectronics Division, Perkin-Elmer Corpo-ration, USA, p. 18.

[106] G. Beamson and D. Briggs, High Resolu-tion XPS of Organic Polymers. The Sci-enta ESCA300 Database. Wiley, Chichester(1992).

[107] Y. Niwa, H. Kobayashi, T. Tsuchiya, J.Chem. Phys. 60 (1974) 799.

[108] S.N. Kumar, G. Bouyssoux, F. Gaillard,Surf. Interf. Anal. 15 (1990) 531.

[109] B. Zaid, S. Aeiyach and P. C. Lacaze, Synth. Met . 65, 27 (1994).

[110] K. Idla, A.Talo, H.E.-M. Niemi, O.Forsenand S.Yiasaari Surf. and Interf. Anal. 25(1997) 837-854.

[111] K. Chiba, R. Ohmori, H. Tanigawa, T.Yoneoka, S. Tanaka, Fusion Eng. Des. 49/50(2000) 791.

[112] E.T. Kang, K.G. Neoh, K.L. Tan, Adv.Polym. Sci. 106 (1993) 135.

[113] E.T. Kang, K.G. Neoh, Y.K. Org, K.L. Tan,B.T.G. Tan, Synth. Met. 39 (1990) 69.

[114] J. Lei, Z. Cai, C.R. Martin, Synth. Met. 46(1992) 53.

[115] T. A. Skotheim (ed.), Handbook of Con-ducting Polymers, Vols. I and II. MarcelDekker, New York (1986).

[116] C. Malitesta, I. Losito, L. Sabbatini, P.G.Zambonin, J. Electron Spectrosc. Relat.Phenomena 76 (1995) 629.

[117] Ch. Froeck, A. Bartl and L. Dunsch, Elecr-rochimca Acta 40 (1995) 1421-1425.

[118] S. Pandey, Wataru Takashima, KeiichiKaneto, Thin Sol. Films 438-439 (2003) 206-211

[119] P. Herrasti, P. Ocon, Appl. Surf. Sci. 172(2001) 276-284

[120] P. Herrasti, L. Diaz, P. Ocon, A. Ibnez, E.Fatas Electrochimica Acta 49 (2004) 3693-3699

[121] R. J. Holness, G. Williams, D. A. Worsley,and H. N. McMurray, J. Electrochem. Soc.,152 (2) B73-B81 (2005)

[122] Kh. Ghanbari, M.F. Mousavi, M. Sham-sipur Electrochimica Acta (in press)

[123] D. Nicolas-Debarnot, F. Poncin-Epaillard,Anal. Chim. Acta 475 (2003) 1-15

[124] Gaoyi Han, Gaoguan Shi, Sensors and Acu-tators B 99 (2004) 525-531

[125] Y. Wu, G. Alici, G.M. Spinks, G.G. WallaceSynth. Met. (article in press)

[126] E.W.H. Jager, E. Smela, O. Ingans, I.Lundstrom, Synth. Met. 102 (1999) 1309-1310

[127] D. Braun, Mat. Today, 5 (2002) 32-39

[128] Y. Liu, T. Cui, K. Ybarahramayan SolidState Electronics 47 (2003) 811-814

[129] V. Saxena, B.D. Malhotra, Curr. App.Phys. 3 (2003) 293-305

[130] S. Asavapiriyanont, G.K. Chandler, G.A.Gunawardena and D. Pletcher, J. Elec-troanaL Chem., 177 (1984) 245.

[131] F. Beck, U. Barsch and R. Michaelis J. Elec-troanal. Chem., 351 (1993) 169-184

[132] F.Beck, P.Braun and M.Oberst, Bet. Bun-senges. Phys. Chem., 91 (1987) 967.

[133] P.A. Christensen and A. Hamnett, Elec-trochim. Acta, 36(1991) 1263.

BIBLIOGRAPHY 143

[134] G. Inzelt, V. Kertesz, A.-S. Nyback, J. SolidState Electrochem. 3 (1999) 251.

[135] J. N. Barisci, T.W. Lewis, G. M. Spinks,C. O. Too, and G. G.Wallace, J. Intel. Mat.Syst. Str. 9 (1998) 723.

[136] S. Kuwabata, C.R. Martin 91 (1994) 1-12

[137] U. Rammelt, L.M. Duc, W. Plieth, J. Appl.Electrochem. 35 (2005) 1225.

[138] M. Rohwerder, F. Turcu, Electrochim. Acta53 (2007) 290.

[139] J.R. Reynolds, M. Pyo, Y.J. Qiu, Synth.Met. 55 (2/3) (1993) 1388.

[140] U. Rammelt, S. Bischoff, M. El-Dessoki, R.Schulze,W. Plieth, L. Dunsch, J. Solid StateElectrochem. 3 (1999) 406.

[141] M.A. Vorotyntsev, E. Vieil, J. Heinze, J.Electroanal. Chem. 450 (1) (1998) 121.

[142] M. Zhou, M. Pagels, B. Geschke, J. Heinze,J. Phys. Chem. B 106 (39) (2002) 10065.

[143] Q.X. Zhou, C.J. Kolaskie, L.L. Miller, J.Electroanal. Chem. 223 (1/2) (1987) 283.

[144] U. Johanson, A. Marandi, T. Tamm, J.Tamm, Electrochim. Acta 50 (2005) 1523.

[145] G. Paliwoda-Porebska, M. Stratmann, M.Rohwerder,U. Rammelt, L. Minh Duc, W.Plieth, J. Solid State Electrochem. 10 (2006)730.

[146] G. Paliwoda, M. Stratmann, M. Rohwerder,K. Potje-Kamloth, Y. Lu, A.Z. Pich, H.-J.Adler, Corros. Sci. 47 (2005) 3216.

[147] T. Schauer, A.Joos, L.Dulog, C.D. Eisen-bach, Progres in Organic Coatings 33 (1998)20-27

[148] A. Mirmohseni, W.E. Price, G.G. Wallace,Journal of Membrane Science 100 (1995)239-248

[149] Y. Velmurugu, S. Skaarup, Ionics 11 (2005)370-374

[150] Hochun Lee, Haesik Yang, Juhyoun KwakJournal of Membrane Science 152 (1999) 61-70

[151] C. Weidlich, K.-M. Mangold, K. JuttnerElectrochimica Acta 50 (2005) 1547-1552

[152] C. Gabrielli, M. Keddam, N. Nadi, H. Per-rot J. of Electroanal. Chem. 485 (2000) 101-113

[153] Yongfang Li, electrochimica Acta 42 (1997),203-210

[154] K.M. Cheung, D. Bloor, G.C. Stevens, J. ofMat. Sci. 25 (1990) 3814-3838

[155] D.E. Tallman, G. Spinks, A. Dominis, G.G.Wallace, J. Solid State Electrochem (2002)6: 73-84

[156] A. W. Hassel, Koji FushimiHassel, M. Seo,Electrochem. Commun. 1 (5) (1999) 180,

[157] M.A de la Plaza, M.C. Izquierdo, Eur. Pol.Journal 42 (2006) 1446-1454

[158] C. Ehrenbeck, K. Juttner, Electrochim.Acta 41 (4) (1996) 511.

[159] C. Sunseri, S. Piazza, F. Di Quatro, J. Elec-trochem. Soc. 137 (1990) 2411.

[160] J. Kim, E. Cho, H. Kwon, ElectrochimicaAcgta 47 (2001) 415-421

[161] H.Asteman, K.Lill, A.W.Hassel,manuscript

[162] R. Hausbrand, M. Stratmann, M. Rohw-erder, submitted to JES

[163] Delahay and L. J. Stagg, J. Electrochem.Soc. 99(12) 546-548

144 BIBLIOGRAPHY

List of Figures

1.1 Scheme of the mechanism of the cathodic delamination . . . . . . . . . . . . . . . . . . 7

1.2 Different forms of polyacetylene, a) cis-transoid, b) trans-transoid. . . . . . . . . . . . 9

1.3 Idealized structrure of ICPs: a)polyacetylene, b)polythiophene, c)polypyrrole, d)polyaniline 10

1.4 Degenerated states in polyacetylene with reversed order of alternating bonds. . . . . . 11

1.5 Solitons in the polyacetyelene chain, a)neutral b)positive c)negative . . . . . . . . . . . 12

1.6 Band structure of the polymer with the soliton states . . . . . . . . . . . . . . . . . . 13

1.7 The band structure of polymer with bipolaron states . . . . . . . . . . . . . . . . . . . 13

1.8 Forms of the polyprrole chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

1.9 The mechanism of the polypyrrole polymerization. . . . . . . . . . . . . . . . . . . . . 15

2.1 Position of Fermi levels for a) not connected metals, b) connected metals. . . . . . . . 22

2.2 Scheme of the Kelvin Probe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

2.3 Compensation of the Volta potential difference by applied external potential . . . . . . 24

2.4 Scheme of the electronic circuit used for the nulling procedure. . . . . . . . . . . . . . 25

2.5 Scanning Kelvin Probe at Max-Planck-Institute in Duesseldorf . . . . . . . . . . . . . 26

2.6 The main panel of the Control Software for SKP at Max-Planck-Institute in Duesseldorf. 28

2.7 Front panel of the control program for the Scanning Kelvin Probe. . . . . . . . . . . . 29

2.8 Motor section in the SKP control program, panel “Scan”. . . . . . . . . . . . . . . . . 30

2.9 SKP control software, “Experiment” section, panel “Scan” . . . . . . . . . . . . . . . . 32

2.10 Data presentation, 3D graph window. . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

3.1 The setup for the electrochemical experiments. . . . . . . . . . . . . . . . . . . . . . . 39

146 LIST OF FIGURES

3.2 Potential during the deposition of the polypyrrole. . . . . . . . . . . . . . . . . . . . . 40

3.3 Structural formula of: a) tetrabutylammonium cation, b) tosylate anion. . . . . . . . . 42

3.4 Cyclic voltammogram of the PPy layer performed in 0.1M tetrabutylammonium tosy-

late solution at scanning speed of 10mV/s . . . . . . . . . . . . . . . . . . . . . . . . . 43

3.5 Cyclic voltammogram of the PPy layer performed in 0.1M NaCl solution at speed of

10mV/s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

3.6 A typical XPS survey spectrum of the polypyrrole layer after cycling in 0.1 tetrabuty-

lammonium tosylate solution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

3.7 Deconvolution of selected peaks obtained from detailed XPS measurement, a) C 1s

peak, b) N 1s peak, c) O 1s peak, d) S 2p peak. Details on the assignment of the

different species: see text. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

3.8 Pictures of the PPy layers obtained using Scanning Electron Microscope, a) thin layer

b) thick layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

3.9 The AFM image of the PPy sample, thickncess of the PPy layer of 1µm. . . . . . . . . 51

4.1 The idea of the defect protection by the conductive polymer . . . . . . . . . . . . . . . 54

4.2 Electrochemical behavior of the iron electrode. . . . . . . . . . . . . . . . . . . . . . . 55

4.3 Protection of pin holes by ICP induced passivation. . . . . . . . . . . . . . . . . . . . . 56

4.4 The experimental setup for the investiagtion of the “ennobling effect“. . . . . . . . . . 57

4.5 Immersion test in: a) K2SO4 at pH4 b) chloride containing electrolyte at pH6 . . . . 59

4.6 Cyclic discharging and recharging of the polypyrrole by connecting and disconnecting

it with the iron, lower graph: enlarged snapshot showing details of the iron protection

and break-down. Red curve shows the galvanic current, blue one the potential of the

polypyrrole, green one of the iron. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

4.7 Galvanostatic discharge of the polypyrrole layer in 1M KCl solution at the current

density of 1 µA/cm. The polymer thickness was 1 µm. In the case of contact with iron,

the current does not remain constant, but increases drastically at the potential when

the iron becomes active, i.e. the break down point will be at a different potential. . . . 61

5.1 The idea of the dislocation of the site of the oxygen reduction, a) delamiantion of

the insulating or slightly conductive coating. Reduction of the oxygen at the metal-

polymer interface leads to the delaminaiton of the coating, b) smearing out of the

oxygen reduction. The large volume of the polymer allow to neutralize large amount of

radicals and OH−, c) highly conductive coating. The site of the oxygne reduction was

shifted to the ICP/Top-Coat interface and results with the delamination of the Top-Coat. 64

LIST OF FIGURES 147

5.2 Experimental set-up for investigating a possible dislocation of the site of the oxygen

reduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

5.3 Results of the ToF SIMS measurements of the polypyrrole layer which was delaminated

in the 18O2 containing atmosphere, a) distribution of 18O2, b) distribution of 18OH−. 67

5.4 ToF-SIMS depth profiles, a) not delaminated area, b) delaminated area. . . . . . . . . 68

5.5 Mechanism of the overoxidation of polypyrrole. . . . . . . . . . . . . . . . . . . . . . . 69

5.6 Degradation of the polypyrole chain due to the overoxidation. . . . . . . . . . . . . . . 70

5.7 The scheme of the polypyrrole reduction in the alkaline enviroment without polarization. 70

5.8 Delamination of the polypyrrole coating in oxygen containing atmosphere. The defect

was place at the left site of the sample. Irreversible change of the PPy color is visible

in on the left side of the sample (PPy turned transparent). Intact PPy is marked by

the dark area on the right side of the sample. . . . . . . . . . . . . . . . . . . . . . . . 71

6.1 Schematic of the “intelligent self-healing coating”. The doping anions stored in the

polymer matrix are released from the coating due to the reduction of the polymer

caused by electrons produced at the corroding defect. Released anions passivate the

surface of the corroding metal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

6.2 Setup for the investigation of the contribution of oxygen and polymer reduction to the

delamination of the conductive coating. . . . . . . . . . . . . . . . . . . . . . . . . . . 76

6.3 Successive SKP profiles (time interval 30 min between profiles) measured by SKP,

first during reduction in oxygen-free nitrogen atmosphere (solid lines), then during

delamination in air (dotted lines). For sample (d) air was exchanged back to nitrogen

at the end of the experiment. The samples differ in the thickness of the tosylate doped

polypyrrole film: (a) 0.271 C/cm2 deposition charge (about 1.1µm), (b) 0.103 C/cm2

deposition charge (about 0.4µm), (c) 0.0501 C/cm2 deposition charge (about 0.2µm),

and (d) 0.022 C/cm2 deposition charge (about 0.1µm). . . . . . . . . . . . . . . . . . . 78

6.4 (a)Position of the reduction/delamination front vs. time for the four samples. (b)

Galvanic current between artificial defect and sample vs. time. The change to oxygen

containing atmosphere results in a sudden and steep increase in the current (see b) and

for the three thinner films in a significant slow down of the progress of the front; in

fact, for these three films the progress comes even basically to a stop for a period of

about 3 h (between 200 and 380 min). . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

6.5 (a) A 1µm polypyrrole film is first reduced in nitrogen atmosphere and then the po-

tential in the defect is switched to a higher value (e.g. +300 mV). The potentials in

the reduced area measured directly after the switching remain low. (b) If the same

experiment is performed in air the potentials go up immediately, indicating delamination. 80

148 LIST OF FIGURES

6.6 Cyclic voltammograms of polypyrrole doped with tosylate (0.1 C/cm2 deposition charge)

in 0.1M tetrabutylammonium chloride. Scan speed: 10 mV/s. . . . . . . . . . . . . . . 81

6.7 XPS spectra in the range between 150 and 350 eV binding energy of (a) polypyrrole

doped with tosylate and (b) with chloride (after anion exchange by cycling in tetra-

butylammonium chloride). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

6.8 Potential profiles during reduction in nitrogen. (a) For a polypyrrole film doped with

tosylate and with the big organic tetrabutylammonium cation in the defect (solid lines).

The progress is very slow, the polypyrrole does not get fully reduced even directly at

the border to the defect (potentials stay above -100 mV). After injection of KCl into

the defect a fast reduction front progresses into the polypyrrol (dotted lines). Time

intervals for the case of terabutylammonium in the defect: 3 h, after injection of KCl:

15 min! (b) For a polypyrrole film doped with chloride (exchanged) and again with

0.1M tetrabutytlammonium chloride in the defect (solid lines). Time interval between

successive profiles: 3 h. The progress is slow, but steady. Full reduction is observed.

After injection of KCl into the defect no increase in velocity is observed (dotted lines,

still 3 h time between scans). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

6.9 Comparison of the progress of the reduction front vs. time for chloride doped polypyr-

role in the presence of a small cation in the defect (black circles) and for the case of the

big tetrabutylammonium cation in the electrolyte (red triangles), data from fig. 6.8b).

The injection of KCl has no effect on the slow curve started with only the big cation in

the defect. (b) The effect of the cation size on reduction velocity for polypyrrole doped

with tosylate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

6.10 XPS spectra between 160 and 350 eV binding energy after the reduction (in the de-

lamination set-up, with KCl in the defect) of (a) tosylate doped and (b) chloride doped

polypyrrole. The charge neutralization mainly occurs via incorporation of the small

cation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

6.11 Current vs. progress of the reduction/delamination front for the about 1.1µm and for

the about 0.4µm thick polypyrrole film. In the nitrogen atmosphere only polypyrrole

reduction takes place, resulting in about three times higher currents for the thicker

sample (not at the beginning, but at later stages). After change to air oxygen reduction

at the interface and delamination of the films occur. While the thick coating shows

immediately after the gas exchange a nearly linear dependence on delaminated area,

for the thinner one this takes a while. . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

6.12 Schematic explanation of the role of cation transport in thick coatings and interfacial

transport at the delaminated interface for thin coatings. (a) In nitrogen (no delamina-

tion), (b) in air (delamination). For the thick coating delamination does not increase

ion transport significantly. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

LIST OF FIGURES 149

6.13 Schematic explanation of (a) the high mobility for cations in the presence of anions

remaining in the polymer. (b) If anions are quantitatively expelled from the conducting

polymer during reduction then the fast cation hopping is inhibited in the ion-free zone

(c). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

7.1 The electrochemical scheme of the corrosion process. Top figure shows the profile of

sample. Blue and red dashed lines represent the rate of the oxygen reduction and the

iron dissolution respectively. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

7.2 The model of the polarization of the interface by the conductive polymer during the

delamination process. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

7.3 Progress of the delamination of the PVB coating under different potentials applied to

the defect. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

7.4 The experimental setup for the investigation of the polarization of the PVB-chromium

interface by the conductive polymer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

7.5 Results of the delamination test performed on the chromium sample, (a) potential

profiles, (b) the progress of the delamination. . . . . . . . . . . . . . . . . . . . . . . . 105

7.6 The experimental setup for the delamination of the iron sample. . . . . . . . . . . . . 105

7.7 Maps of the potential measured during the delamination of the model clear coat applied

on the iron substrate with a dot of PANI (right hand side). Measurements were taken

every 1.5h. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

7.8 Profiles of the potential obtained from maps presented on the fig. 7.7 at the position

1800µm on the Y axis. Note the formation of the IR drop at the delaminated interface

caused by massive discharging of the PANI. . . . . . . . . . . . . . . . . . . . . . . . . 108

7.9 The photography of the gold sample coated with the polyaniline dot (black spot) and

PVB coating after the delamination experiment. . . . . . . . . . . . . . . . . . . . . . 110

7.10 Maps of the potential obtained during the delamination experiment performed on the

gold substrate coated with the polyaniline pigment and the PVB coating. The position

of the polyaniline dot is marked by the black dashed line. . . . . . . . . . . . . . . . . 111

7.11 The idea to use a pattern of conductive polymer, (a) the 3d arrangement of the con-

ductive particles, (b) the idea of overlapping protection zones. . . . . . . . . . . . . . . 113

7.12 The scheme of the setup for the preparation of the PANI pattern. . . . . . . . . . . . . 114

7.13 Pictures of patterns of the polyaniline on gold substrates, spacing between centers of

dots: (a)-300µm, (b)-500µm,(c)750µm, (d)-1000µm. . . . . . . . . . . . . . . . . . . . 115

150 LIST OF FIGURES

7.14 Maps of the potential obtained during the delamination test of the polyaniline pattern

with the spacing between dots of 300µm. Measurements taken every 1h. Black vertical

lines marks the patterned region. The pattern is not resolved by SKP due to the

similarity between the potentials of the area PVB7gold and PVB/PANI. . . . . . . . . 116

7.15 The potential profiles measured over regions marked on the fig. 7.14i. Profiles (a),(b)

were measured along paths “A”, “B” respectively. . . . . . . . . . . . . . . . . . . . . 118

7.16 Potential maps obtained during the delamination test of the polyaniline pattern with

dot spacing 750µm. Measurements taken every 1h. . . . . . . . . . . . . . . . . . . . . 119

7.17 (a) The potential profiles measured along path “A” marked in fig. 7.16a, (b) progress

of the delamination front. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120

7.18 The topography of the polyaniline pattern with dot spacing of 1000µm. . . . . . . . . 121

7.19 Maps of the potential measured during the delamination experiment measured over the

sample with 1000µm dots spacing. Measurements taken every 1h. . . . . . . . . . . . . 122

7.20 Maps of the potential measured during the delamination of the sample with the spacing

of polyaniline dots of 500µm. Measurements taken every 1h. . . . . . . . . . . . . . . . 124

7.21 The progress of the delamination for measured for: reference sample coated with the

PVB, continuous layer of polyaniline plus a layer of the PVB, and the sample with the

polyaniline pattern plus a coating of PVB. The green rectangles mark the position of

the rows of the PANI dots. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

7.22 Draining of cations by the conductive polymer prevents pull down of the potential due

to ingress of cations. Also the release of anions from the ICP (source) would ensure an

inhibition of a decrease of potential due to cation ingress from the delaminated area

into the protection zone by counterbalancing the charge by anions. . . . . . . . . . . . 127

7.23 Evolution of the potential profile due to the diffusion of ions along the interface. Diffu-

sion of ions along the metal/insulating coating interface: (a) expected potential profile,

linearly increasing potential, (b) the observed shape of the potential profile. Evolution

of the potential profile after an establishment of the galvanic coupling between the con-

ductive particle and the defect: (c) linearly increasing potential in the protection zone

(d) the observed potential profiles after coupling to the conductive pigment. . . . . . . 128

7.24 Potential profiles obtained during the diffusion experiment performed on the iron oxide

in nitrogen atmosphere. The red line indicates the potential profile which was measured

just after the introduction of oxygen. The black line indicates the position of the edge

of the polyaniline dot. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

7.25 The influence of the diffusing ions to the surface potential, (a) diffusion primarily of

anions, (b) diffusion primarily of cations. . . . . . . . . . . . . . . . . . . . . . . . . . . 132

LIST OF FIGURES 151

7.26 Diffusion of anions and cations in the nitrogen atmosphere. (a) 0.1M tosylate sodium

salt solution in the defect, (b) 0.1M tetrabutylammonium chloride salt solution in the

defect. SKP profiles were measured every 12 hours. . . . . . . . . . . . . . . . . . . . . 134

152 LIST OF FIGURES

List of Tables

1.1 Conductivity of the pylypyrrole prepared in various solvents and electrolytes [57]. . . . 17

3.1 Relative intensities of components of each envelope expressed as a percentage of the

total signal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

3.2 Ratios of measured elements. Theoretical values were calculated from the formula

C4H3N(C7H7O3S)n assuming that doping level equals to Stotal/Ntotal. . . . . . . . . . 50

conductive polymers for corrosion protection : a critical ...€¦ · conductive polymers for...

Documents