Ultrathin chemisorbed polymer coatings : corrosion ... · Ultrathin Chemisorbed Polymer Coatings:...

Ultrathin Chemisorbed PolymerCoatings: Corrosion Protection and

Nanostructuring of ZnO

Dissertationzur

Erlangung des Grades"Doktor der Naturwissenschaften"

an der Fakultät für Chemie und Biochemieder Ruhr-Universität Bochum

von

Danish Iqbalaus Muskat, Oman

This doctoral study was carried out under the supervision of Dr. Andreas Erbe andProf. Dr. Martin Stratmann in the group of Interface Spectroscopy in Department ofInterface Chemistry and Surface Engineering, at Max-Planck-Institut für EisenforschungGmbH (MPIE), Düsseldorf, Germany in the framework of the International Max PlanckResearch School for Surface and Interface Engineering in Advanced Materials (IMPRS-SurMat) during the period from July 2011 till November 2014.

Advisors: Dr. Andreas ErbeProf. Dr. Martin Muhler

Dedication

To my beloved parents,

Rashid Iqbal & Shahina Rashid

i

Acknowledgments

There are many people who helped me in several ways for successful completion of myPhD work. First of all, I would like to express my sincere gratitude to my adviser,Dr. Andreas Erbe for giving me an opportunity to work on a creative topic.Andreas, it was great to work with you. I learned a lot from you, specially how onecan do science by keeping things simple and straight. Moreover, I also acknowledgeyour scientific directions and helpful discussions during my work, which helped mein completion of this thesis and to enhance my scientific skills. I am also grateful toProf. Dr. Martin Muhler for kindly accepting to act as a second reviewer andthe time he invested on me. I furthermore thank Prof. Dr. Martin Stratmannfor his continuous support and scientific remarks that proved to be reliable guidanceat all times.I would like to take this opportunity to thank Julian Rechmann for his hard

work and support he put on in this project. I would also like to emphasize onthe importance of all my former and current colleagues at MPIE. In particular,Dr. Adnan Sarfraz’s helpful, critical but constructive discussions and inspiringsuggestions have played a key role. I also like to express my special thanks toDr. Asif Bashir, who continuously supported me throughout my PhD. I amthankful to my good friends at MPIE Abdulrahman Altin, Tai Hai Tran, AshokVimalanandan, and Dr. Georgi Genchev. Guys, I had a great with you,specially when we were on conferences.I am grateful to Dr. Simanthini Nayak, Dr. Paul Schneider, Petra Ebing-

haus, Bei Bei Pang, Dr. Ying Chen, Stefanie Pengel, and other membersof my group for providing a conducive atmosphere. I also owe thanks to my officemates, Dr. Tabrisur Khan, Dr. Rakesh Moirangthem, Waldemar Kriegerand Miriam Lange for providing a nice and friendly atmosphere. I gratefully ac-knowledge the organizers of SURMAT program, for providing a platform to studentswhere they can explore the field of surface science. I thank Dr. Rebekka Loschenand Elke Gatermann for all the administrative assistance and all SurMat studentsfor useful discussions. I also thank Cornelia Arckel and other colleagues at GOdepartment and MPIE for their continued support throughout this work.Finally, I am grateful to my family for supporting me through pains and pleasures

and motivating me through difficult times.

iii

Zusammenfassung

Das Ziel dieser Arbeit ist in erster Linie das grundlegende Verständnisvon Adhäsionseigenschaften von Beschichtungen an Polymer/Metall Grenz-flächen, insbesondere die durch kathodische Delamination initiierte Schich-tablösung von Beschichtungen von Metalloberflächen. Durch eine thermischinitiierte freie radikalische Polymerisation wurden ultradünne Polymethylme-thacrylat und Polysytrol schichten auf die Metalloberfläche aufgebracht undsystematisch untersucht.Die Ablösung der Polymere durch die kathodischeDelamination wurde mittels der Raster-Kelvin-Sonde analysiert. Desweiterenwurden durch den Einsatz von Vernetzermolekülen unterschiedliche Vernet-zungsgrade eingestellt und daren Delamination hin untersucht. Die Studienhaben gezeigt, dass durch die kovalente Bindung von Polymerbeschichtungenauf eine, Metalloberfläche die kathodische Delamination um den Faktor ∼10verlangsamt wird. Für PMMA zeigte darüber hinaus die Vernetzung der ange-wandten Polymere weitere Verlangsamung um den Faktor ∼10, Für PS hattedie Vernetzung nur geringen Einfluss auf die Delamina kinetik.Ein weiter Aspekt dieser Arbeit befasst sich mit der Untersuchung der resul-

tierenden Korrosionsprodukte nach der Delamination. Hierbei konnte die Aus-bildung von Zinkoxid Nanostäbchen beobachtet werden. Der, kathodische De-laminationsprozesses von Polymerbeschichtungen ermöglicht somit eine einfa-che und ökomische Herstellung von Zinkoxid-Nanostäbchen. Desweiteren wur-den die gezielt hergestellten Nanostäbchen auf die fotokatalytische Aktivitäthin untersucht. Diese zeigten eine deutliche aerobe Fotooxidation des Farb-stoffs Methylorange durch Bestrahlung mit rotem Licht, welches normaler-weise nur mit Halbleitermaterialien kleiner Bandlücken nutzbar ist. Zum Ver-ständnis des Wachstumsprozesses dieser Nanostäbchen wurde in situ Raman-Spektroskopie angewandt. Die Messungen zeigen eine anfängliche Phase mitdefektreichen Zinkoxid Strukturen, wohingegen in der fortgeschritten Wachs-tumsphase Zinkoxid mit Wurzit Struktur dominiert.In dieser Arbeit wurde gezeigt, dass die Grenzflächenstabilität zwischen Me-

tall und Polymer aber auch die Metalloxid Eigenschaften einen entschiedenenEinfluss auf die Delamination von Polymerbeschichtungen haben. Außerdemwurde dargelegt, dass die eigentlich unerwünschte Korrosion von Metallen zurgroßflächigen Herstellung von Zinkoxid Nanostäbchen genutzt werden kann.

v

Abstract

The primary aim of this work is to provide better understanding of adhe-sion properties at polymer/metal interface as well as the deadhesion processof polymer coatings from metal surfaces due to cathodic delamination. Ultra-thin Poly(methylmethacrylate) [PMMA] and poly(styrene) [PS] coatings thatwere covalently attached to the zinc substrate were synthesized. The stabilityof these coatings against cathodic delamination was characterized using scan-ning Kelvin probe (SKP). The results indicate that covalently bound polymercoatings to metal substrate shows reduction in rate of cathodic delaminationby a factor of ∼10. Moreover, adding a cross-linker to a polymer coatingfurther slows down delamination, with another factor of ∼10. Moreover, de-lamination rates are reaction-controlled for the system investigated here, asshown by linear time dependence.The other area of interest in this work was to study the corrosion products

formed at the metal surface as a result of deadhesion of polymer coatingdue to cathodic delamination. The analysis of corrosion products shows theformation of ZnO nanorods at the surface of the metal. Hence, this studyalso provides a very simplified and economically feasible route to fabricateZnO nanorods via a tailored corrosion process. These ZnO nanorods act asphotocatalysts for aerobic photooxidation of organic dye methyl orange underillumination with red light, which is normally accessible only to narrow bandsemiconductors. Furthermore, the growth of these ZnO nanords was studiedby in situ Raman experiments. The obtained in situ data reveals that atinitial stages of synthesis, preferentially defect rich ZnO grows while in thelater stages,wurtzite ZnO growth becomes more pronounced.In the present work, the important factors in adhesion and disbonding pro-

cesses of covalently bound polymer coatings has been evaluated. It is shownthat the interfacial chemistry at polymer metal interface and the properties ofoxide itself strongly influences the delamination properties of a polymer coat-ing. Moreover, it was demonstrated that usually unwanted corrosion processcan be utilized to produce large areas of ZnO nanorods.

vii

Contents

1 Introduction 11.1 Corrosion of metals . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Polymer coatings for corrosion protection . . . . . . . . . . . . . . . . 2

1.2.1 Adhesion at polymer/metal interface . . . . . . . . . . . . . . 31.3 Deadhesion of polymer coatings from metallic substrates . . . . . . . 6

1.3.1 Cathodic delamination . . . . . . . . . . . . . . . . . . . . . . 61.3.2 Anodic delamination . . . . . . . . . . . . . . . . . . . . . . . 71.3.3 Migration of species through polymer coatings . . . . . . . . . 8

1.4 Corrosion products . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101.4.1 Corrosion products of zinc in aqueous solutions . . . . . . . . 101.4.2 Corrosion process for fabrication of Zn(O) nanostructures . . . 12

2 Aims and Concepts 152.1 Motivation and aims of the work . . . . . . . . . . . . . . . . . . . . 152.2 Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

2.2.1 Surface of interest . . . . . . . . . . . . . . . . . . . . . . . . . 172.2.2 Synthesis of chemisorbed polymer coatings . . . . . . . . . . . 172.2.3 Model corrosion experiments . . . . . . . . . . . . . . . . . . . 192.2.4 Study of corrosion products . . . . . . . . . . . . . . . . . . . 20

2.3 Thesis outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

3 Experimental Techniques 233.1 Analytical techniques for characterization of surface modification . . . 23

3.1.1 Infrared (IR) spectroscopy . . . . . . . . . . . . . . . . . . . . 233.1.2 X-ray photoelectron spectroscopy (XPS) . . . . . . . . . . . . 233.1.3 Time-of-Flight secondary ion mass spectrometry (TOF-SIMS) 243.1.4 UV/Vis spectroscopy . . . . . . . . . . . . . . . . . . . . . . . 243.1.5 Ellipsometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

3.2 Analytical techniques for characterization of corrosion behavior . . . . 253.2.1 Scanning Kelvin probe (SKP) . . . . . . . . . . . . . . . . . . 253.2.2 Linear polarization . . . . . . . . . . . . . . . . . . . . . . . . 263.2.3 Electrochemical impedance spectroscopy (EIS) . . . . . . . . . 27

3.3 Analytical techniques for characterization of corrosion products . . . 283.3.1 Electron microscopy . . . . . . . . . . . . . . . . . . . . . . . 28

ix

Contents

3.3.2 Atomic force microscopy (AFM) . . . . . . . . . . . . . . . . . 283.3.3 X-ray diffraction (XRD) . . . . . . . . . . . . . . . . . . . . . 293.3.4 Raman spectroscopy . . . . . . . . . . . . . . . . . . . . . . . 29

3.4 Sample preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293.4.1 Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293.4.2 Surface modification . . . . . . . . . . . . . . . . . . . . . . . 303.4.3 Synthesis of ZnO nanostructures . . . . . . . . . . . . . . . . . 313.4.4 Electrochemical evaluation of stability of polymer coatings . . 323.4.5 Photocatalytic decomposition of methyl orange (MO) . . . . . 32

4 PMMA model coatings bound via organosilanes 354.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354.2 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . 36

4.2.1 Surface modification . . . . . . . . . . . . . . . . . . . . . . . 364.2.2 Model corrosion experiments . . . . . . . . . . . . . . . . . . . 404.2.3 Discussion of the chemistry of cathodic delamination . . . . . 45

4.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

5 PS model coatings bound via organosilanes 495.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495.2 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . 50

5.2.1 Synthesis of polymer coatings . . . . . . . . . . . . . . . . . . 505.2.2 Model corrosion experiments . . . . . . . . . . . . . . . . . . . 53

5.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

6 PMMA model coatings bound via ester linkage 576.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 576.2 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . 58

6.2.1 Surface modification . . . . . . . . . . . . . . . . . . . . . . . 586.2.2 Model corrosion experiments . . . . . . . . . . . . . . . . . . . 61

6.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

7 Comprehensive discussion on cathodic delamination 65

8 Delamination kinetics on ZnO 698.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 698.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

9 Growth of ZnO nanorods 779.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 779.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78

9.2.1 Structure and morphology . . . . . . . . . . . . . . . . . . . . 789.2.2 Light absorption and photocatalytic activity . . . . . . . . . . 81

x

Contents

9.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

10 In situ Raman studies on the growth of ZnO nanorods 8710.1 Experimental details . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

10.1.1 Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8810.1.2 Synthesis of ZnO nanostructures . . . . . . . . . . . . . . . . . 8910.1.3 Analytical techniques . . . . . . . . . . . . . . . . . . . . . . . 89

10.2 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . 8910.2.1 Change in surface morphology during growth . . . . . . . . . 9310.2.2 Effect of electrolyte concentration . . . . . . . . . . . . . . . . 9410.2.3 Optical properties of ZnO nanorods . . . . . . . . . . . . . . . 95

10.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

11 Summary and Outlook 9911.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9911.2 Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

Appendix 1251 List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1272 List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1293 Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1304 List of publications till November 2014 . . . . . . . . . . . . . . . . . 1315 Conference oral presentations & posters . . . . . . . . . . . . . . . . . 1316 CV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

xi

Chapter 1: Introduction

Corrosion is a common problem which we are facing for over a hundred years. Itis a naturally occurring phenomenon termed as deterioration of a material and itsproperties due to reaction with its surrounding environment. Like other naturalhazards, corrosion can lead to dangerous and expensive damages in almost everyareas from commodity products like pluming systems, sewer utility lines to high-tech products like airplanes, rockets etc. Several years ago, a study conducted byU.S. Federal Highway Administration (FHWA) showed that the total estimated costof damages and prevention of corrosion is approximately 3.1% of the nations grossdomestic product (GDP)1. This study points to the fact that corrosion not onlyleads to loss of desired properties but also has huge economic costs that are neededfor the repairs or complete replacement of a corroded material. However, the figuresmentioned above are only the direct economic costs of corrosion. The indirect costs(such as plant down time, loss of product, loss of efficiency etc) resulting from actualor possible corrosion can be even more but they are difficult to evaluate2[pp. 55].In particular, corrosion of metals is of significant research interest due to its eco-

nomic and commercial importance. Hence, several methods have been developed bymaterial scientist and engineers which are time proven to minimize/prevent corro-sion. These methods include corrosion protection by metallic coatings3, conversioncoatings4,5, with the aid of inhibitors6,7, self-healing coatings8–10, and by organiccoatings11,12. In the following chapters of this work, corrosion of metals and theirprotection by organic coatings will be discussed in detail. In particular, aliphatic(PMMA) and aromatic (PS) polymer coatings having silane and ester linkage withzinc substrate has been synthesized and their resistance against cathodic delami-nation has been investigated. Moreover, corrosion products formed on top of thesurface as a result of cathodic delamination is another area of interest that wouldbe discussed in detail.

1.1 Corrosion of metalsMetallic corrosion has been the subject of many research articles and text books13–15.Metallic corrosion is an electrochemical process which involves transfer of a charge.Hence, it is useful to first briefly discuss the electrochemistry of corrosion beforecontinuing to methods of corrosion protection. Metallic corrosion is an heteroge-

1

Chapter 1. Introduction

neous electrochemical reaction that usually occurs at metal-environment interface.If we consider a corrosion of metal in an aqueous solution (which usually containsdissolved ionic species), they result in oxidation of metal leading to the formationof ionic species that takes electrons which comes from the metal. As an example ageneric oxidation reaction of metal (M) is shown below:

M→ Mz+ + ze− (1.1)

However, the above reaction also called half reaction is an incomplete reaction as thefree electrons formed due to oxidation of metal must be consumed. These liberatedelectrons are consumed eventually by a reduction reaction which occurs on the samesurface14[pp.3-8]. A common example of reduction reactions are hydrogen evolutionand oxygen reduction, which are shown below.

2 H+ + 2 e− → H2 (1.2)O2 + 2 H2O + 4 e− → 4 OH− (1.3)O2 + 4 H+ + 4 e− → H2 + 2 OH− (1.4)

The location where oxidation of metal takes place is termed as anode, while the placewhere reduction reaction takes place is called cathode and these are the electrodes.

1.2 Polymer coatings for corrosion protectionPolymers are a class of synthetic materials that consist of small molecules (monomer)linked together through a covalent bonding forming long chain molecules calledmacromolecules16. Due to huge diversity and ease of modification in the propertiesof polymers, they are nearly used in all of our current commodity and industrialproducts. One of the important application area of polymers is there use as coatingsto protect metals against corrosion. Moreover, they are also responsible for improvedsurface properties like roughness and aesthetic appearance. Polymers prevent thecorrosion of metals by isolating them from the corrosive environment which leadsto suppression of both anodic dissolution of metal and the corresponding cathodicreaction.Polymers can be synthesized in a number of ways such as addition polymeriza-

tion and condensation polymerization. Free radical polymerization is one type ofaddition polymerization that is widely used to synthesize commercial polymers. Infree radical polymerization, monomers add one by one to an active side of growingchain. Free radical polymerization reaction mainly consist of three steps i.e. ini-tiation, propagation, and termination17. In the initiation step, an initiator (suchas AIBN, BPO) decomposes into free radicals either by heating or shining lighton them (depending on the type of initiator used) in presence of monomers. Themonomer consist of unsaturated carbon-carbon double bond that reacts with an un-paired electrons in a radical. In this reaction, the active center of the radical "grabs"

2

1.2. Polymer coatings for corrosion protection

one of the electrons from the unsaturated bond of the monomer, leading to a newactive center at the either end of the chain. The next step is a propagation step,where at the active end of the chain new monomers are connected and continues togrow. Finally the growth of polymer chains are stopped by a termination reactionwhere two active ends of growing chains are joined together or as a result of hydro-gen extraction. The steps of free radical polymerization are shown schematically inFigure 1.1 .

InitiatorDecomposition

Initiation

Propagation

Termination

N NC CH3C CH3

CH3

CN

CH3

CN

N NC CH3C CH3

CH3

CN

CH3

CN

H3C C

CH3

CN

C CH3

CH3

CN

N N

R

I

I

R

I

R

I

R R

I

R R

II

RR

I

R R

I

RR

II

RRRR

n

nnn

n

I

R R

n

II I

RR

n

R(

(

n

or

Figure 1.1: Schematic representation of steps in classical free radical polymeriza-tion18.

1.2.1 Adhesion at polymer/metal interfacePolymer coatings are applied to inorganic substrates as a liquid solution, primarilyby dip coating, spray coating, and brushing etc. After the coating process, liquidphase is evaporated at room temperature or by heating at higher temperatures,leaving a dry polymer coating covering the metal surface. However, the factor whichplays a more crucial role in corrosion protection of metals is the barrier propertiesof a polymer coating and adhesion at polymer/metal interface. Polymer coatingsmainly interact with metal/oxide surfaces in three ways15:a) mechanical interlocking, where two surfaces interact via interlocking due to

roughness,

3


b) by physisorption, where coating interacts with the inorganic substrate via phys-ical forces like van der Waals forces and dispersion forces (London),c) by chemisorption, where the polymer coating makes a covalent bond with the

substrate.Mechanical interlocking is usually observed for rough surfaces where polymer can

go in to the cavities and attach well to the surface. Recently, it has been shownthat increase in surface roughness leads to better protection against cathodic de-lamination of epoxy coated steel19. It has been well established in the literaturethat acid-base interactions between polymer/metal interface plays a major role inadhesion forces15. Moreover, it has been shown that when molecular interactiondominates, it results from the chemical bonding between reactive species on poly-mer and the substrate or due to molecular interactions between these two20.Wetting of the surface plays an important role because for any type of interaction,

the coating and substrate must be very close to each other. This can be achieved bywetting of the surface which is determined by the surface tension of substrate and acoating material21. To achieve complete wetting, the substrate should have highersurface energy than the coating material. This can be expressed in mathematicalterm as a spreading coefficient "S", which is given by the difference in the surfaceenergy of covered and uncovered surface22.

S = γSV − (γSl + γlV ) (1.5)

where γSV and γSl are the solid-liquid and solid-air surface energies, respectively,while γlV is the interfacial energy of the liquid. In order to have complete wettingof the surface, the value of spreading coefficient should be positive. Higher valuesof "S" indicates that the surface would wet more rapidly. The adhesion strengthbetween substrate and a coating material can be quantified by work of adhesion.Work of adhesion (Wa) is defined as the force applied to separate the interface perunit area.

Wa = γα + γβ − γαβ (1.6)

where, γα and γβ are the surface tension of substrate and coating material, respec-tively, while γαβ is the interfacial tension between substrate and coating material.Oxide surfaces can interact in various ways with different functional groups of thecoating23. Either the surface hydroxyl groups of oxide or the functional group oforganic molecule can act as the acid (as proton donor is Bronsted sense or the elec-tron acceptor in lewis sense) or as a base (proton aceptor or electron donor). Thebond between the functional groups of the oxide surface and a polymer coating canrange from fully covalent bonding to very weak ionic bonding. It has been shownthat various functional groups such as silanes24,25, phosphonates26, carboxylates27,catechols28, and amines29 are good candidates for covalent bonding with oxide cov-ered inorganic substrates. Some examples of covalent bonding between functionalgroups of substrate and coating is shown in Figure 1.2.

4

1.2. Polymer coatings for corrosion protection

OH OH OHOHOH OH

O

OSi

OSi

O

R R

O

SiO

O O

R

Silane

P

O O

O R

P

O O

O R

Phosphonate

P

HO OH

O R

O OH

R

O O

R

O O

R

Carboxylate

HO OH

R

O O

R

O O

R

Catechol

Figure 1.2: Schematic representation of covalent modification of oxide surfaces withspecific functional groups.

As known from classical chemistry, generally covalent bonding is stronger thanionic bonding, therefore for better corrosion protection properties, covalent bondingbetween substrate and a coating material is highly desired. In general, there aretwo approaches to achieve the covalent bonding at metal/oxide/polymer interface.One is the modification of the surface with self assembled monolayer (SAM) andthen application of the coating material30. The second approach is to coat the oxidesurface with end functionalized polymers which leads to covalent bonding with thesubstrate13. Both approaches have their own advantages and disadvantages, suchas monolayers provide more molecular control, however, mechanical robustness iscompromised. On the other hand, prefunctionalized polymer coatings has betterbarrier properties but the degree of covalent bonding between the substrate and acoating material is reduced due to steric hindrances.

5


1.3 Deadhesion of polymer coatings from metallicsubstrates

A complete polymer coating based protective system should provide an efficient bar-rier against water, oxygen and ionic species leading to life long protection of metallicsubstrate against corrosion. However, due to presence of defects, pores and cracksin an polymer coating system, water, oxygen and ions can approach metal/polymerinterface. These corrosive species leads to corrosion of metal underneath the organicfilm by several mechanisms. Two possibilities which are of high technical importanceare cathodic and anodic reactions at polymer/metal interface. In a corrosion reac-tion, both cathodic and anodic reactions occur at the same time, however, one ofthem will be the rate determining step. The metal ions produced due to anodic pro-cess, leads to increase in ionic force underneath the coating, giving rise to formationof blister due to osmotic phenomenon14[pp. 510-515]. These blisters can grow bydifferent mechanisms, known as, cathodic and anodic delamination.

1.3.1 Cathodic delaminationAs mentioned in section 1.3, one mechanism of polymer coating deadhesion fromthe metallic substrate is cathodic delamination. It has been shown that the majordriving force for cathodic delamination in presence of humid air is cathodic reaction(see equation 1.3)31. The hydroxyl and metal ions produced as a result of cathodicand anodic reactions, respectively, can form complexes (specially for zinc and iron)and precipitate as metal oxides and hydroxides. Studies has also indicated that thepH beneath the coating, where cathodic reaction takes place is relatively high dueto oxygen reduction reaction which produces hydroxyl ions32. These very high localpH value leads to deadhesion at polymer/metal interface possibly due to breakage ofcovalent bonding between the polymer and the metal. The complete mechanism ofcathodic delamination was proposed by Leidheiser et. al31 and can be summarizedas follows.

• Corrosive species such as water, oxygen and ions penetrates through the poly-mer coating.

• Low resistance pathways are created between bulk electrolyte and the sub-strate.

• Anodic reaction leads to oxidation of metal at the surface.

• To compensate anodic reaction, corresponding reduction of oxygen (cathodicreaction) takes place at the metal surface and leads to initiation of delamina-tion.

• Due to corrosion reactions in a confined area, corrosion products precipitateson the metal surface. This leads to drop in a pH at the anodic reaction site.

6

1.3. Deadhesion of polymer coatings from metallic substrates

• This causes the cathodic regions to be more basic (pH >10) and anodic regionsmore acidic (pH <10).

O2

Zn(OH)2

Zn(OH)2

Zn(OH)2

Zn(OH)2

Corrosionproducts

Zn2+

Zn2+

Zn2+

Zn2+

Zn Zn + 2e2+ -

Anode

Zn2+

Zn2+

Zn2+

Zn2+

Zn2+

Zn2+

Cathode

O2

Zn(OH)2

OH-

OH-

OH-

OH-

OH-OH

-

Increasing pHAcidic Basic

Polymercoating

O2

ELECTROLYTE

Figure 1.3: Schematic representation of cathodic delamination of polymer coatedzinc substrate.

The effect of higher pH on the degradation of interfacial bond has been supportedby the existence of carboxylated species at the interface33. On the other hand, atthe anodic side of corrosion reaction, local pH value is relatively low (i.e. <10). Therate of cathodic delamination is determined by rate of diffusion of oxygen throughthe polymer coating (in the case where oxygen reduction reaction is dominant) andby diffusion of alkali metal cations through the coating to cathodic region32. It wasalso shown that the rate of cathodic delamination strongly depends on the coatingthickness suggesting that the rate of diffusion of reactants through the coating wasan important variable.

1.3.2 Anodic delaminationCathodic delamination usually requires presence of ions, water molecules, and anelectron conducting interface that allows electrochemical reactions to occur. Hence,cathodic delamination usually dominates at relatively high humidity levels and onmaterials which form (semi)conducting oxides. Whereas, for wet conditions with lowhumidity levels and insulating interfaces, such as aluminum, anodic-driven deadhe-sion reaction will be dominant34. Anodic deadhesion processes can be classifiedas chemical reaction that takes place underneath an organic coating in which themajor deadhesion process is the anodic corrosion reaction15[pp. 131]. This typeof corrosion process is also termed as filiform corrosion (FFC) due to growth ofthread like filaments35. Although FFC is more common on organic-coated alu-minum36,37, it has been also observed on magnesium and steels36,38. It usuallyarises from surface defects present in the protective film in the presence of solubleionic species. The mechanism of FFC is still debated. Various processes includ-ing cathodic disbondment39,40, anodic undermining41, and physical disbonding29,42

7


though electro(osmotic) forces are proposed as the primary cause. As an example,consider FFC on organic coated iron, that consist of electrolyte filled filament heads,containing Fe2+ and Fe3+ cations and complexing anions. Generally, it is believedthat under theses conditions, the iron substrate is oxidized at the head front (an-ode) (see equation 1.9) and oxygen is reduced in the vicinity of the head-tail junction(cathode) (see equation 1.8)40. This leads to differential aeration because oxygenand water is supplied to the filament head by diffusion through the porous tail39.The potential gradient created between anode and cathode, forces anions to migrateto the front and cations to the back. Oxidation of iron is thought to result in theundercutting of the organic coating at the filament head leading edge, and has beenproposed as the principal cause of coating deadhesion43.

O2 + 2 H2O + 4 e− → 4 OH− (1.7)Fe→ Fe2+ + 2 e− (1.8)

1.3.3 Migration of species through polymer coatingsThe life time of organic-coated systems strongly relies on the protective properties ofa coating material. Polymer coatings not only reduce the rate of corrosion on metalbut they also act as a barrier to the penetration of external species. Corrosion reac-tions underneath the polymer coating requires an aqueous medium (that providesconductivity) and species for the cathodic reaction (usually oxygen)15[pp. 131-132].As the corrosion process cannot proceed more rapidly than the slowest mechanisticstep, limiting the rate of transport of one of the reactants will slow down the overallcorrosion process44. Hence, the transport of aggressive species (such as water andoxygen) through the polymer coating is a matter of concern and is discussed in thefollowing subsection.

Transport of oxygen

The main cathodic reaction in the corrosion process at neutral pH is oxygen reduc-tion (see figure 1.3). Hence, the diffusion of oxygen through the polymer coatingplays a significant role in initiation and progression of a corrosion reaction. However,the effect of oxygen permeation through an organic coating is still debatable14[pp.504-505]. For instance, Baumann45 has determined the oxygen permeability throughseveral paint films and concluded that oxygen diffusion through a coating plays arole in defining the final rate of corrosion. While on the other hand, Mayne hasstated that oxygen diffusion could not be a rate determining factor in the corrosionprocess46. Moreover, Stratmann and coworkers47 have studied the effect of oxygenpartial pressures on the delamination rate of epoxy based polymer coated iron sub-strate and estimated the diffusivity of molecular oxygen as 1.7× 10−8cm2s−1. Thediffusion of oxygen through polymer coating strongly depends on the compositionand morphology of the polymeric material. Specially, polymer films with less defects

8

1.3. Deadhesion of polymer coatings from metallic substrates

and pores in a molecular structure could significantly reduce the diffusion of oxygenthrough the polymer matrix.

Transport of water

Water is one of the important partner in initiating a corrosion process underneatha polymer coating48. The important question to know is how and at what ratethe water approaches a polymer/metal interface. Does water permeate through thepolymer matrix due to defects present in a coating or it just follow a random walkthrough the coating to approach an interface. Some of the previous studies haveshown that the polymer glass transition temperature (Tg) strongly influences thesolubility of water in a coating. In particular, low Tg leads to higher permeability ofwater through a coating which leads to higher corrosion rates49. Moreover, perme-ation of water leads to swelling of a polymer film that induces internal stresses andas a result of these internal stresses the Tg of a coating can be lowered50.It has been demonstrated that water uptake of a polymer coating increases on

applying cathodic potential51. The increased rate of water uptake was a directconsequence of the development of more effective diffusion pathways through thecoating. Hence, it is of technical interest to improve the barrier property of coatingsagainst water permeation through the polymer. This can be achieved by increas-ing the coating thickness, low porosity, or by use of several layers of coating withadditional additives (like pigments)14[pp. 506-507]. The water uptake by organicfilms (applied or free standing) can be measured by weight measurements52,53 eitheras a weight gain during absorption or weight loss during drying. More practicaland easier way to measure water uptake by organic films is by in situ capacitancedetermination using electrochemical impedance. Higher water content in a coatingleads to higher capacitance values. Equation that is widely used to determine thewater content in a organic film is Brasher-Kingsbury equation54.

φ = log[Cc(t)/Cc(0)]log(80) (1.9)

where, φ is the volume fraction of water in a coating. Cc(t) and Cc(0) are the coatingcapacitance at time t and of the dry coating, respectively.

Transport of ions

Organic coatings applied to metal surfaces in absence of inhibitive pigments retardsthe corrosion reaction by providing a barrier to the flow of ions46. Studies havebeen made to reveal the factors which control this resistance and the mechanismby which ions flow through the polymer film. For instance, Kittleberger and Elm55

measured diffusion rate of sodium chloride through several layers of organic coatingand found a linear relationship between the rate of diffusion and the reciprocal ofthe film resistance. This linear relationship suggests that sodium chloride diffused

9


through the coating as ions and not as ion pairs or undissociated molecules, sincethe diffusion through the unionized coating material would not effect the resistance.Furthermore, they also found that rate of diffusion of ions is much smaller than therate of diffusion of either water or oxygen. To further verify this behavior, Mayneand coworkers56,57 conducted a study over cast varnish coating to determine theresistance of the film that is in contact with potassium chloride. They observedtwo opposite behaviors, either the conductivity increased with increasing electrolyteconcentration or the conductivity decreased when the solution became more conduc-tive. The interpretation of decrease in conductivity with reduced solution resistancewas associated with water uptake by the film. Moreover, it was also concludedthat increase in conductivity with increased electrolyte concentration was a resultof differences in cross-linking density within the film structure.

1.4 Corrosion productsThe corrosion products discussed in this section are the solid products formed at thesurface of a corroding metal. Corrosion products formed are the first layers whichare exposed directly to the environment, hence they greatly influence the corrosionbehavior of the metal. Consequently, it is important to understand the process ofcorrosion product formation and what are the factors which effect them. In thissection, the focus would be on the corrosion products formed as a result of zinc cor-rosion in aqueous solutions and how they can be used to fabricate nanostructureswhich can play a vital role in various application areas such as nanoelectronics58,59,sensors60 and photocatalysis61,62. Generally speaking, corrosion products differ incomposition, structure, morphology, size, and properties depending on the condi-tions of the corrosion process63[pp. 157].

1.4.1 Corrosion products of zinc in aqueous solutionsThe most common corrosion product of zinc in diluted salt solutions such as KCl orNaCl at room temperature is ZnO, with zinc hydroxide present in various amountsas minor component63[pp. 171]. Generally, ZnO is expected to crystallize by thedehydration of zinc complexes (such as zinc hydroxide) that are formed during zinccorrosion in aqueous salt solution64. Moreover, in solutions of certain salts such asH3PO4, zinc can form compounds of low solubility (such as Zn2P2O7) and corrosionproducts can be concentrated with the corresponding zinc compounds63[pp. 171].

Effect of solution pH

The decisive factor in the formation of zinc corrosion product is the solution andsurface pH. This dependency is illustrated in a Pourbiax diagram (1.4). This dia-gram describes the relative formation of species in a certain pH-potential region. For

10

1.4. Corrosion products

instance, in acidic solutions, zinc has high solubility and dissolves with the forma-tion of Zn2+ ions. However, no corrosion product could be formed on corroding zincsurface as zinc complexes are soluble under acidic conditions (pH <5)65–67. Solubil-ity of zinc decreases as the solution pH increases, that results in the precipitationof Zn(OH)2 when certain pH value is approached. Roetheli et al.68 have studiedthe corrosion of zinc in HCl and NaOH aqueous solutions of different pH and hasconcluded that no corrosion product could be formed on the zinc surface when thesolution pH was either below 5 or more than 13.5. The corrosion products formedon galvanized steel in various concentrations of alkaline solution (KOH) was studiedby Macias and Andrade69. They found that initially the main corrosion productformed was ZnO, but it transforms to Zn(OH)2 for more than 30 days of immersionin a solution.

Zn2+

ZnO/Zn(OH)2

Zn

O2-

2

pH-2 0 4 8 12 16

Zn(s)

E(V

)

1.6

1.2

0.8

0.4

0

-0.4

-0.8

-1.2

(a)

(b)

Figure 1.4: Potential-pH equilibrium diagram (pourbaix) for zinc. Potential refersto standard hydrogen electrode70.

The morphology and structure of corrosion products formed in phosphate solu-tions also depends greatly on the solution pH71–73. The zinc phosphonate filmsformed are stable in the pH range of 3-1274,75. It was found that the film dissolvedcompletely at low pH, whereas at high pH they are converted to Zn(OH)2, whichthen decomposes to ZnO or dissolves as zincate ions.

Growth process

One of the fundamental studies on the growth and transformation of zinc corrosionproducts in aqueous solutions has been conducted by Feitknecht76. It was foundthat amorphous hydroxide is precipitated from dilute solution of zinc chloride, and

11


changes with time to ZnO or Zn(OH)2 via following reactions:

Zn2+ + 2 OH− → Zn(OH)2 (1.10)Zn(OH)2 → ZnO + H2O (pH = 7− 9) (1.11)Zn(OH)2 → β−Zn(OH)2 (pH = 11− 12) (1.12)

At higher Cl− concentration and more acidic pH values (pH<7), then two differenthydroxide chlorides (a and b) can be formed by the following reaction mechanism.

Zn(OH)2 + Zn2+ + 2 Cl− → 6 Zn(OH)2 · ZnCl2 (a) (1.13)7 Zn2+ + 12 OH− + 2 Cl− → 6 Zn(OH)2 · ZnCl2 (a) (1.14)4 ZnO + 4 H2O + Zn2+ + 2 Cl− → 4 Zn(OH)2 · ZnCl2 (b) (1.15)5 Zn2+ + 8 OH− + 2 Cl− → 4 Zn(OH)2 · ZnCl2 (b) (1.16)

Similarly, zinc carbonates such as smithsonite (ZnCO3) and hydroxycarbonate mixedsalt hydrozincite [Zn5(CO3)2(OH)6] can be obtained by precipitating a solution ofzinc salt with a mixture of sodium hydroxide and sodium carbonate77. The com-position of the corrosion product formed on zinc surfaces from different aqueoussolutions may not be uniform and it has been shown by several research groups thatformed corrosion layer varies depending on the corroding environment78–80. The im-portant properties of the corrosion product formed are compactness of the formedlayer, adhesion to the substrate, electrical conductivity, hardness and resistance todissolution. These properties determine the chemical stability and electrochemicalproperties of the formed zinc corrosion products14[pp. 176].

1.4.2 Corrosion process for fabrication of Zn(O) nanostructuresGenerally speaking, corrosion process is a highly undesired process which leads to aloss of material and product properties. However, in recent years it has been shownthat corrosion process can be utilized to synthesize various nanostructured materi-als. In particular, by customizing the corrosion process, one could synthesize variousnanostructured materials such as TiO2

81, CdO82, and CuO83–85. As mentioned insection 1.4.1, corrosion is strongly influenced by oxygen content, solution pH, ionconcentration, and surrounding environment, therefore to control and optimize thecorrosion process, it is important to control these parameters. By using controlledcorrosion process, several ZnO nanostructures has been synthesized. For instance,Yu et al.86 has fabricated ZnO nanorods on zinc foil by corrosion in aqueous for-mamide solution. The growth was initiated by immersing zinc foil in a 5 wt %formamide aqueous solution and heating at 65 oC for 24 h. The oxidation of zinc inwater is relatively low due to presence of surface passivating layer. Therefore, for-mamide and high temperature was utilized to increase the release of zinc ions due tocorrosion, which eventually leads to growth of uniform hexagonal ZnO nanorods thatwere highly oriented with the growth direction along the c-axis (perpendicular to the

12

1.4. Corrosion products

substrate surface) by dehydration of zinc hydroxide complex. Using similar growthstrategy, Zhang and coworkers87,88 have synthesized highly oriented and denselypacked ZnO nanorods on zinc substrate by corrosion of zinc in formamide solutionat 65 oC. The corrosion of zinc has been further utilized by Zhao and coworkers89to synthesize pine nano-tree like structures on zinc foil. The synthesis of tree likestructure was carried out by corrosion of zinc that was dipped in a aqueous solutionof ethylenediamine and heating at 140 oC for 10 h. The resulting corrosion productwas dense arrays of the complex nanostructures which resemble a "pine-tree for-est"89. More recently, the substrate free fabrication of ZnO nanorod film was grownby hydrothermal approach90, which again works on a principle of zinc corrosion andprecipitation to form corrosion products. The overall growth mechanism of ZnOnanorods via hydrothermal approach is well documented91, which is expressed by afollowing reaction.

Zn2+ + 2 OH− → Zn(OH)2 → ZnO + H2O (1.17)

In hydrothermal approach, the rate of release of OH− ions is controlled by additionof inhibitors such as hexamethylenetetramine (HMTA) which leads to the controlledcrystallization of ZnO92. The mechanism suggested by Wang et al.90 for substratefree growth of self supporting ZnO nanowire arrays can be expressed as follows.

Zn2+ + nNH3 → Zn(NH3)n2+ (n = 1, 2, 3 or 4) (1.18)

Zn(NH3)42+ + 2 OH− → ZnO + 4 NH3 + H2O (1.19)

Examples from the literature shown above demonstrates the potential of metalliccorrosion for the fabrication of ZnO nanostructrues. Besides the good scalability,metal corrosion offers a economical way to prepare nanomaterials with greatly re-duced energy consumption.

13

Chapter 2: Aims and Concepts

2.1 Motivation and aims of the workOrganic coatings are an important constituent of modern corrosion protection sys-tems93. In particular, synthetic polymer coatings are widely applied to increasethe durability of the protection system. For effective protection against corrosion,polymer coatings should have sufficient barrier properties and adhesion to the metalsurface in presence of water, ions, and other corrosive species. The barrier prop-erties of polymer coatings has been significantly modified by addition of additivesand use of multi-lamellar coatings, which leads to reduced defects and pores in thecoating14[pp. 510-515]. Moreover, adhesion between metal and polymer has beenenhanced by replacing physical adhesion forces to covalent bonding94,95. Currentindustrial protective polymer coatings are usually several micrometers up to tens ofmicrometers thick. From the application side, thinner coatings are desirable due toeconomic reasons, though it is hard to realize that the thinner coatings would havesame barrier properties.Deadhesion of polymer coatings from the metal or oxide substrate due to corro-

sive electrochemical processes is a common problem which could lead to completeloss of desired material properties. Particularly on steel, including galvanized steel,as well as on zinc, cathodic delamination starting on a electrolyte-filled defect isone of the main mechanisms of deadhesion of a coating96,97. However, the dead-hesion process at polymer/metal interface is a more difficult problem due to lackof detailed understanding of bonding chemistry at polymer/metal interface. Oneof the problem behind this issue is that direct analytic probing of such a buriedinterface is difficult, as most available non-destructive experimental techniques arenot interface-specific with the notable exception of sum-frequency generation spec-troscopy98,99. Consequently, typically the interface is destroyed, and subsequentlythe detailed analytical features of surface analysis are available13,94,100. Therefore,it is crucial to develop coating methods and strategies that can lead to the betterunderstanding of adhesion and deadhesion processes at metal/polymer interface.Ultrathin plasma polymer coatings have been considered as a possible alternate,however, poor understanding of interfacial bonding chemistry at polymer/metal in-terface still causes a problem101. As a pure analytic approach to an understandingof metal(oxide)/polymer interfaces can give only limited answers, here a syntheticapproach is employed, in which the metal(oxide)/polymer interface is defined by

15

Chapter 2. Aims and Concepts

chemical synthesis. This synthetic approach permits to carry out detailed electro-chemical experiments with the coating in place.Thus, the primary aim of this work is to synthesize ultrathin polymer coatings

that are covalently bound to a metallic surface and has a well defined interfacialchemistry. This approach allows the better understanding of interfacial chemistryat the metal/polymer interface and allows to systematically study the bonding anddeadhesion behavior between polymer and a metal surface. In addition to the adhe-

Zinc sheet

growth of chemisorbed polymer coating

Delamination

Top coat

Oxide surface

coatingElectrolyte

Zinc sheet

Zinc sheet

Zinc sheet

Zinc sheet

chemisorbed polymer coating



Top coat

Top coat

Zinc sheet

growth of chemisorbed polymer coating

Delamination

Top coat

Oxide surface

coatingElectrolyte

Zinc sheet

Zinc sheet

Zinc sheet

Zinc sheet




Top coat

Top coat

(a)(a)

Zinc sheet

growth of ZnO nanostructure

Delamination

Top coat

Oxide surface

coatingElectrolyte

Zinc sheet

Zinc sheet

Zinc sheet

Zinc sheet

Top coat

Top coat

ZnO nanostructure

ZnO nanostructure

ZnO nanostructure

(b)

Figure 2.1: Schematic representing the route for the delamination studies on mod-ified zinc substrate. (a) Chemisorbed polymer coatings, (b) on ZnOnanostructured surface.

sion strength, oxide chemistry at the polymer/metal interface also plays a significantrole in defining the final corrosion behavior. Most reactive metal substrates are cov-ered with the native oxide layer, while during the corrosion process an oxide layeris formed as a corrosion product. Corrosion products formed on the metal surfacecould strongly influence the further corrosion of a metal. In particular, formed corro-sion products can be passivating, that is blocking anodic and/or cathodic reactions,thus significantly reducing the rate of corrosion, or non passivating (having little or

16

2.2. Approach

no effect on the corrosion). The growth of these products depends on the corrosionenvironment and conditions such as pH, temperature, and humidity63. Moreover,tailoring the corrosion process can lead to the fabrication of oxide nanostructuresthat are potential candidates for various optoelectronic applications. Hence, theother area of interest in this work is to study the corrosion products formed on themetal surfaces and how these products could influence the cathodic delaminationprocess of organic coatings. Study of the corrosion products formed as a resultof corrosion reaction could provide better understanding of the corrosion reactionstaking place at the surface.

2.2 Approach

2.2.1 Surface of interestZinc is a widely used material for corrosion protection of steels. Moreover, it is oftencoated with polymers to enhance the corrosion protection properties. Therefore, inthis work zinc is used as model substrate to study the deadhesion process at thepolymer/metal interface. This allows to mimic the corrosion processes that takesplace on galvanized steels that are coated with polymers.

2.2.2 Synthesis of chemisorbed polymer coatingsAs mentioned earlier, the protective properties of organic coatings applied on metalsurfaces are determined substantially by barrier properties of a protective coatingand the adhesion strength between polymer and metal. Therefore, in order to en-hance the adhesion strength, covalently bonded polymer coatings were synthesizedin situ on top of metal substrates. Covalently bound polymer coatings on inorganicsubstrates are typically synthesized by two routes (a) a grafting to approach102–106,where pre-synthesized end functionalized polymers are covalently bonded to thesubstrate or (b) by grafting from/onto approach106–111 where polymer is grown insitu on the solid substrate. A special case are ultrathin, highly cross-linked plasmapolymer films101,112–114. In this work, grafting onto approach to synthesize ultra-thin covalently bound polymer coatings have been carried out. Initially, the zincsubstrate was functionalized with small molecules having specific functional groupsthat leads to formation of covalent bonding between metal and an organic molecule.The advantage which small organic molecules have over end functionalized bulkpolymers is that due to steric hindrance in end functionalized polymers, the numberof covalent bonds between polymer and metal is significantly reduced. Hence, thedeadhseion process can be accelerated due to less number of covalent bonds betweenmetal and polymer.The adhesion strength between metal and polymer is determined by the func-

tional groups that are present in polymer which makes covalent bonding with the

17


OH OH OH OH OH OH OHOH OH OH OH OH OH OH

ethanol+

water

O

SiO O

SiO

O O


Toluene+TEA

Cl

O

O O

Silane Carboxylate

Figure 2.2: Schematic representation of functionalization of zinc resulting in silaneand carboxylic bonds.

surface hydroxyl groups95,115. In particular, polymer coatings with silane and car-boxylic functionalities are often used in industrial polymer coatings because of theirdesirable structural and interfacial performances13,116. Hence, organic moleculeswith functionalities that leads to silane and ester bonding were employed. Thesmall chain length of the functionalizing molecule leads to a high degree of covalentbonding with well defined interfacial chemistry. Later on, aliphatic (polymethyl-methacrylate, PMMA) and aromatic (Polystyrene, PS) polymers were grown on topof the functionalized zinc substrate by performing thermally initiated free radicalpolymerization using the vinyl functionality as polymerizable group in the anchor-ing molecule. This synthesis strategy leads to ultrathin linear polymer coatingswith well defined interfacial chemistry that can be used to fundamentally study thedeadhesion behavior and bonding strength at polymer/metal interface.The other factor which also plays a vital role in protection of metal against cor-

rosion is permeability of aggressive species such as water and oxygen through thepolymer coating. Water and oxygen can access metal surface due to presence of poreand defects in a polymer bulk structure. The diffusion of these aggressive speciesthrough the polymer coating can be significantly reduced by increasing the degree ofcross-linking between the polymer chains117. Therefore, in addition to linear coat-ings, cross-linked polymer coatings with varying degree of cross-linking were alsosynthesized. The polymer chains were cross-linked by addition of bi-functional vinyl

18

2.2. Approach

molecule during the polymerization reaction. The different amount of cross-linkerleads to polymer coatings with varying degree of cross-linking that exhibits differentdegree of permeation level for water and oxygen.

n = 2 for EDA

n = 6 for HDA

+

toluene, AIBN, 70 Co

O

SiO O

n

O O

O

OnO

O

O

On

O

O

O

O

O

SiO O

Silane

oror

O

SiO O

O

O

n

Figure 2.3: Schematic representation of synthesis of PMMA and PS polymers (linearand cross-linked) on silane functionalized zinc.

2.2.3 Model corrosion experimentsScanning Kelvin probe (SKP) is a well established technique to study the deadhe-sion of polymer coatings from metal substrates due to cathodic delamination118,119.Hence, SKP was used as a primary analytical technique to study the deadhesion be-havior at polymer/metal interface. Moreover, linear polarization experiments wereperformed on these chemisorbed polymer coatings to study the effect on corrosionbehavior of zinc.

19


2.2.4 Study of corrosion productsCorrosion products that are formed during the corrosion reaction can substantiallyinfluence the further corrosion process taking place on top of coated metal sur-faces120,121. Hence, it is of interest to study the corrosion products that are formedduring the corrosion process. In this work, the corrosion process was monitoredby performing in situ Raman experiments during the cathodic delamination pro-cess. This allows to study the chemical and crystal structure of oxide formed as acorrosion product during the cathodic delamination process.

2.3 Thesis outlineFigure 2.4 schematically represents the outline of the thesis. Apart from the firstthree chapters, present work can be divided in to three parts A, B and C.First part (A) corresponds to Chapters 4,5,6 and 7 in which the delamination ki-

netics of ultrathin chemisorbed PMMA and PS coatings has been discussed. Chapter4 investigates the delamination kinetics of silane bonded PMMA coating to a zincsubstrate. The corrosion properties has been studied using SKP, linear polarizationexperiments, and EIS. Chapter 5 represents the studies on the cathodic delamina-tion kinetics of silane bonded PS coatings using SKP, while Chapter 6 investigatesthe delamination kinetics of ester bonded PMMA coating. The overall comparisonof differently bonded PMMA and PS coatings has been discussed in Chapter 7.Second part (B) of this thesis corresponds to Chapter 8, which investigates the

effect of different ZnO morphologies grown on zinc substrate on the cathodic de-lamination. Different ZnO structures namely twin-plate, nanorod and sphere weregrown on top of zinc via a hydrothermal approach and their delamination kineticshas been studied using SKP.Third part (C) of this thesis corresponds to Chapters 9 and 10. This part shows

the investigation of growth of ZnO nanorods on zinc substrate via a tailored corrosionprocess at room temperature. The growth mechanism of ZnO nanorods has beendiscussed in Chapter 9, while the growth kinetics of ZnO nanorods studied via insitu Raman spectroscopy is discussed in Chapter 10. Each chapter also consists ofthe introduction part, where the topic is specifically introduced.

20

2.3. Thesis outline

Introduction

Chapter 1

Chapter 2

Aims and concepts

Chapter 3

Experimental details

Delaminationof Chemisorbed

Polymers

Chapter 4

Delamination kinetics ofsilane bonded PMMA coatings

Chapter 5

Delamination kinetics ofsilane bonded PS coatings

Chapter 6

Delamination kinetics ofester bonded PMMA coatings

Chapter 7

Comprehensive discussion ofdelamination kinetics

Surface oxides

Delamination kinetics on ZnOwith different morphologies

Chapter 8

CorrosionProducts

Chapter 9

Corrosion for fabrication ofZnO nanorods

Chapter 10

In situ Raman studieson growth of ZnO nanorods

Chapter 11

Summary & outlook

Figure 2.4: Layout of flow chart depicting the schematic of thesis chapters.

21

Chapter 3: Experimental Techniques

The aim of this section is to give a brief introduction to the experimental techniquesemployed for the analysis of the samples as well as to provide the details about howthe samples were prepared and characterized.

3.1 Analytical techniques for characterization ofsurface modification

3.1.1 Infrared (IR) spectroscopyIR spectroscopy is a well established technique for identification and characteriza-tion of the molecular structure, orientation and intermolecular bonding of organicmolecules adsorbed on metallic and non metallic surfaces. Vibrational spectra areused as characteristic fingerprints for adsorbate molecule, adsorption configurationand structures. If a molecule has a dipole moment, then a molecule can absorb theIR light, however, only at certain fixed frequencies. Thus, an IR spectrum of lightreflected from surface will show absorption peaks that are characteristic of moleculeand its mode of bonding to the surface. These are the basis of IR spectroscopy.FTIR spectroscopy is nothing but just the Fourier transform of the reflected lightfrom the molecule that is under investigation122.In this work, modified samples were characterized by IR measurements performed

on a Bruker Vertex 70v Fourier transform IR spectrometer (Bruker, Karlsruhe, Ger-many). IR spectra were taken with a spectral resolution of 4 cm−1 at an angleof incidence of 80o using p-polarized light. A liquid nitrogen cooled, middle bandmercury cadmium telluride detector was used for detection. Prior to surface modi-fication, background spectra were obtained from freshly cleaned zinc samples. Thereflectance absorbance spectra shown in this work were recorded against these back-grounds.

3.1.2 X-ray photoelectron spectroscopy (XPS)XPS is a powerful surface sensitive technique that is widely used for the surfaceanalysis of materials. XPS works on the principle of photoelectric effect that isused for obtaining qualitative and quantitative information about the elemental

23

Chapter 3. Experimental Techniques

composition and of the surface123. When the sample surface is bombarded withX-rays having characteristic energy, the electrons from the core levels of the sampleare emitted. The energy of these emitted electrons is a "‘fingerprint"’ of the eachelement present in a material and their chemical environment. Consequently, XPSprovides elemental specificity and the chemical environment of the elements presentat the surface123.To analyze the surface composition, modified zinc samples were characterized

using XPS (Quantum 2000, Physical Electronics, Chanhassen, MN, USA) at a take-off angle of 45o, with a monochromatic Al Kα source (1486.6 eV) at a pass energy of23.5 and energy resolution of 0.2 eV. Survey scans were performed to scan the samplefor relevant elements (pass energy = 100 eV, energy step = 0.5 eV). Measurementswere formed by Adnan Sarfraz.

3.1.3 Time-of-Flight secondary ion mass spectrometry(TOF-SIMS)

TOF-SIMS is also a well developed surface sensitive analytical technique that uses apulsed ion beam to remove molecules from the very outermost surface of the sample.The molecules become dissociated ions (positive or negative) after the bombardmentwith the incident ion beam. These generated secondary particles are then acceleratedinto a flight path on their way towards a detector and their mass is determined bymeasuring the exact time at which they reach the detector (i.e. time-of-flight). The"Time-of-Flight" of an ion is proportional to the square root of its mass, thus all thedifferent masses are separated during the flight and can be detected individually124.In this work, TOF-SIMS was used as a tool to analyze the interfacial region at

metal/polymer interface. TOF-SIMS measurements have been executed with a PHITRIFT CE (Physical Electronics), applying a gallium ion gun on a spot size of 100µm × 100 µm. The primary gallium ion beam was used with an energy of 15 kVand the same ion beam was used for sputtering on a spot size of 200 µm × 200 µmto remove ambient impurities before measuring the mass spectrum. Experimentswere conducted by Adnan Sarfraz.

3.1.4 UV/Vis spectroscopyUV/Vis spectroscopy is a widely used technique to study the optical properties ofmaterials. It uses light in ultraviolet-visible range to induce electronic transitions ina molecule/material under investigation. It is well known that when a material isirradiated with a light, it absorbs some light having a wavelength that matches a pos-sible electronic transition within the molecule and excites an electron from groundstate (HUMO) to an excited state (LUMO). By studying the reflectance/absorptionspectra of the exposed sample, one could draw useful information about the concen-tration of the molecule or the optical properties of a material125.In this work, UV/Vis spectroscopy was employed for two primary experiments.

24

3.2. Analytical techniques for characterization of corrosion behavior

One was to study the degradation of a dye (MO) under the visible light illuminationand in presence of a catalyst. Second was to study the optical properties of assynthesized ZnO nanostructures. The UV/Vis spectrometer used in this study wasPerkin Elmer Lambda 900.

3.1.5 EllipsometryEllipsometry is an optical technique that is widely used to determine the thicknessand optical constants of thin films. It is based on the measurement of change in thepolarization of light after reflection at normal/non-normal incidence on the surfaceto study. Usually in ellipsometry linearly polarized light is used which becomeselliptically polarized after reflection from the surface of interest or vice-versa. El-lipsometric measurement gives two independent angles i.e. ∆ and Ψ, where Ψ isthe amplitude ratio and ∆ is the phase difference of p and s-polarized light. El-lipsometry is an absolute measurement method i.e. it does not need any reference.However, it does not give directly the physical parameters of the sample (thicknessand refractive index). In order to describe the sample it is always necessary to usethe model126.To determine the thickness of as synthesized polymer coatings, ellipsometric mea-

surements were performed using a UV/visible spectroscopic ellipsometer (SE 800,Sentech Instruments GmbH, Berlin/Krailling, Germany) in the wavelength range of400-800 nm. Measurements at three spots per sample were taken with an incidentangle of 70. Differences in ellipsometric parameters with respect to the bare zincsubstrate were analyzed. Film thickness of the polymer layer on samples, regardlessof modification, was obtained by fixing the refractive index of the polymer coatingsto 1.5, and considering it in first approximation as wavelength-independent. Thevalue of 1.5 is close to the refractive index of PMMA of ≈1.48127.

3.2 Analytical techniques for characterization ofcorrosion behavior

3.2.1 Scanning Kelvin probe (SKP)SKP is a well established technique to study the deadhesion behavior of polymercoatings from metal/semiconductor surfaces. In principle, SKP measures a Voltapotential (or work function) distribution over metal or semiconductor surfaces usinga vibrating capacitor method47,128,129. Kelvin probe usually consist of a conductingtip that is electrically in contact with the sample that is measured. When the workfunction of the tip and the sample is different, there will be a flow of current fromone materiel to another in order to align the Fermi levels. The Volta potentialdifference is measured by zeroing the flow of the above mentioned current/electronsby applying external voltage. This applied external voltage, which is identical to the

25


Volta potential difference, is given by the energy difference between the Fermi levelof the sample with respect to the Fermi level of the probe, which is known in manycases34. The work function of a bare metal surface under high vacuum conditionsis a well-defined quantity34. However, when the metal surface is covered with anelectrolyte or polymers, then electron has to pass through additional interfaces andthe Volta potential differences is strongly affected by potential differences at theseinterfaces. The same applies to the electrode potential and hence, there is a simplerelationship between Volta potential difference (∆ΨRef

Pol ) and the electrode potentialEcorr,

Ecorr =WRef

F− χPol − εRef1/2

+ ∆ΨRef

Pol (3.1)

WRef corresponds to the work function of the reference metal, χpol is the poten-tial of polymer surface and εRef1/2 is the half-cell potential of the reference metal.Therefore, SKP allows the measurement of corrosion potentials without touchingthe surface under investigation across a dielectric medium of infinite resistance. Forthe reactive metals such as zinc, the surface is usually covered with a thin nativeoxide layer that significantly influences the reactivity at the composite. In such acase, the Volta potential difference is given by the sum of the potential differencesover all interfaces34. In case of cathodic delamination, an additional liquid interfaceis formed between metal/oxide and polymer. In such a case, metal/liquid interfaceis treated as conventional electrochemical interface, however, an additional galvanicpotential difference ∆ΦD is taken into account for electrolyte/polymer interface.

Ecorr = WRef

F− χPol − εRef1/2 + ∆ΨRef

Pol + ∆ΦD (3.2)

Therefore, (∆ΨRefPol ) allows the measurement of corrosion potential at the me-

tal/polymer buried interface only if ∆ΦDis known34.Delamination experiments were performed on a commercial SKP system from

KM Soft Control (Wicinski - Wicinski GbR, Wuppertal, Germany) with a 100 µmNiCr tip in humid air and nitrogen atmospheres. Kelvin probe was calibrated tothe standard hydrogen electrode (SHE) against a Cu/CuSO4 reference electrode.

3.2.2 Linear polarizationLinear polarization is an electrochemical technique that estimates the rate of cor-rosion of metals by measuring the corrosion current. When a coated metal surfaceis exposed to a conducting electrolyte of sufficient oxidizing power, the metal willstart to corrode by oxidation of metal referred as anodic reaction. The dissolutionof metal leads to formation of free electrons which will flow to the cathodic side,where they will be consumed by a reduction reaction. However, direct measurementof corrosion current (icorr) is difficult as anodic and cathodic sites shifts during the

26

3.2. Analytical techniques for characterization of corrosion behavior

corrosion. This problem can be resolved by applying external potential shifts (E),which will produce a measurable current flow (I) at the corroding electrode. Thebehavior of this externally imposed potential is governed by degree of difficulty inanodic and cathodic reactions, which is termed as polarization resistance (Rp). Thesmaller the rate of anodic and cathodic reaction, the smaller the value of Rp for thegiven potential shift. This current can be related to icorr by the following relation,

Rp = βaβc2.3 icorr(βa + βc)

(3.3)

where, βa and βa are Tafel slopes of anodic and cathodic branches, respectively.In this work, linear polarization experiments were executed in 0.1 M KCl elec-

trolyte by using a Volta-lab Radiometer PST050 potentiostat at a scan rate of 5mV/s. All experiments were performed in a self-made 3 electrode setup. As counterelectrode, a graphite stick was used. A commercial Ag/AgCl/3 M KCl referenceelectrode (Metrohm, Filderstadt, Germany) served as reference electrode. All po-tentials are quoted here versus standard hydrogen electrode (SHE). The corrosioncurrent densities icorr were calculated from the polarization resistance Rp accordingto an established procedure130. Rp was determined as the slope of the linear polar-ization curve from -15 mV to +15 mV around Ecorr. The Tafel slopes of the anodicand cathodic branch, βa and βc, respectively, were obtained from linear regions ofthe Tafel plots. Corrosion current densities were then calculated as130.

3.2.3 Electrochemical impedance spectroscopy (EIS)EIS is a useful analytical tool to study the properties and performance of polymercoated metal systems. The AC impedance of an electrochemical cell is conventionallymeasured by applying an AC potential and then measuring the current throughthe cell. The ratio of potential and current is impedance of the cell, at a certainfrequency131.Using a Gamry PCI4/Series G Family potentiostat (Gamry Instruments Inc.)

with the same electrode setup as for linear polarization experiments, electrochemicalimpedance spectroscopy (EIS) was performed in the frequency range of 10−1 to105 Hz (10 points/decade of frequency f) with 15 mV amplitude of the sinusoidalvoltage modulation, at the corrosion potential. The experiments were conducted in aborate buffer solution, containing 0.2 M H3BO3, 0.05 M Na2B4O7 and 0.1 M Na2SO4.The pH of the solution was measured as 8.4 ± 0.4. The used electrolyte providesgood conductivity to perform the EIS experiments, and is expected to suppressundesired formation of (sometimes protective) corrosion products. Samples werePVB coated, and have been prepared as for the SKP measurements. No strongchanges in the spectra have been observed after even several hours of exposure timeto the solution, as water uptake starts to be decisive only after >24 h. EchemAnalyst software (Gamry Instruments) has been used for fitting of the obtained

27


data.

3.3 Analytical techniques for characterization ofcorrosion products

3.3.1 Electron microscopyGenerally speaking, electron microscopy is primarily used to study the structureand morphology of a material that is not visible to our naked eyes or to a lightmicroscope. It uses a beam of electrons to create an image of the specimen. Twomajor types of electron microscopes in use today are scanning electron (SEM) mi-croscopy and transmission electron microscopy (TEM). In SEM, the image is createdby secondary electrons that are emitted from the surface due to excitation by theprimary electron beam. While in case of TEM, the image is created by a partiallytransmitted electron beam through a very thin specimen that carries informationabout the structure of the sample132.In this work, the surface morphology of the samples were obtained by using SEM

from Zeiss (LEO 1550 VP) which was also equipped with energy dispersive X-rayspectrometer (Oxford Instruments). TEM measurements were performed by A.Kostka using JEOL JEM-2200FS operated at 200 kV.

3.3.2 Atomic force microscopy (AFM)AFM is a tool that measures a topography as well as physical properties of a surfaceby scanning over the sample via a sharpened tip. The measurement principle ofAFM is based upon the forces acting between the surface and a tip also knownas cantilever. In order to measure the surface topography, the probe is placedclose enough to the sample such that it can interact with the forces associatedto the surface. The AFM tip scans across the sample surface by keeping forcesbetween the sample and probe constant. The image of the surface is reconstructedby monitoring the precise motion of the tip which moves towards or away fromthe surface during the scan depending on the forces acting on it. In commercialAFM systems, multiple measurement modes are present such as contact mode andtapping mode. In contact mode, the tip is dragged across the sample surface andthe surface topography is measured by looking at the deflection of the sample. Intapping mode, the cantilever is driven to oscillate near its resonance frequency closeto the surface and the surface morphology is measured by detecting the changes inthe oscillation of the cantilever133.In this work, AFM images were recorded by a Digital Instruments Dimension

3100 AFM in tapping mode, employing Si micro-cantilever tips with radius <10 nmand a resonant frequency of ∼318 kHz. The measurements were performed by AsifBashir.

28

3.4. Sample preparation

3.3.3 X-ray diffraction (XRD)XRD is a type of analytical technique that reveals the information about the crystalstructure, chemical composition and physical properties of a material. In XRD thesample is illuminated with an X-ray (while either moving the sample or a detector),that leads to a diffraction pattern. These diffraction patterns are then analyzed tounderstand the crystallographic properties of a material. In order to analyze thinfilms, XRD is performed in a grazing incidence (GI) geometry. GIXRD employeslow incident angles for the incoming X-ray beam, so that diffraction can be madesurface sensitive134.In this work, XRD (Bruker-AXS D8) measurements were performed in a grazing

incidence geometry with a Cu-Kα as a X-ray source.

3.3.4 Raman spectroscopyRaman spectroscopy is a useful non-destructive analytical tool that is widely usedto study the structure and properties of various types of organic and inorganicmaterials. Raman spectroscopy works on the principle of inelastic scattering of amonochromatic light that provides molecular fingerprint related to the vibrational,rotational, and other low frequency transitions of molecules135.In this work, Raman spectroscopy was employed to study the growth kinetics

and properties of ZnO nanostructures. In situ measurements on the growth of ZnOnanowires were conducted using a Horiba Jobin Yvon LabRAM confocal Ramanmicroscope with a Argon laser as a light source (emission wavelength; 514.342 nm)and a power of 12 mW. To carry out in situ Raman measurements, the assemblyof sample mounted on zinc substrate was introduced into a custom made chamberwith a glass window opening to allow spectral acquisition. To minimize electrolyteevaporation, the humidity was controlled ≥90% with the help of humid/dry air flow.Ex situ measurements were performed via alpha300M confocal Raman microscopesystem (Witec, Ulm, Germany), with an argon laser source (emission wavelength:532 nm). For photoluminescence experiments, the UV laser source was employedwith an emission wavelength of 355 nm. Measurements were performed throughmicroscope objective of 100x (numerical aperture = 0.75) in reflection geometry.

3.4 Sample preparation

3.4.1 MaterialsZinc sheets (purity, 99.95%) with a thickness of 1.5 mm were obtained from Good-fellow (Cambridge, UK). Vinyltrimethoxy silane (VTS), ethylene glycol diacrylate(EDA), hexanediol diacrylate (HDA), methyl methacrylate (MMA), styrene, azo-bisisobutyronitrile (AIBN), hexamethylenetetramine (HMT), toluene, and tetrahy-drofuran (THF) were supplied by Sigma-Aldrich (Steinheim, Germany). MMA was

29


purified by distillation from CaH2 under N2 atmosphere, while AIBN was recrys-tallized in ethanol prior to use. All other chemicals were used as received unlessotherwise noted. Zinc substrates (15 mm × 15 mm) were initially mechanicallyground with SiC paper up to 4000 grit, followed by polishing with silicon paste(1µm) to have a smooth surface. Polished samples were ultrasonically cleaned inacetone and dried under a nitrogen stream. Prior to functionalization, zinc sub-strates were immersed in 0.1 M NaOH for 1 min to increase the concentration ofhydroxyl groups12. The substrates were then thoroughly washed with deionizedwater. All synthesis steps were performed under air exclusion in standard Schlenktubes.

3.4.2 Surface modificationFunctionalization of zinc with silane

The overall scheme surface modification is illustrated in Figure 2.2 After the hydrox-ide treatment, zinc surfaces were modified with VTS by a silanization reaction136,to obtain a vinyl-terminated, silane modified substrate. Functionalization was car-ried out at a room temperature by immersion of a polished zinc substrate in anethanol/water mixture (90:10) containing 0.007 M of VTS. After 24 h, the sam-ple was removed, cleaned sequentially with excess acetone and ethanol to removephysisorbed molecules, and dried under a pressurized nitrogen stream.

Functionalization of zinc with ester

The overall synthesis strategy is illustrated schematically in Figure 2.2. Function-alization reaction was carried out by adding 0.0275 mol of triethylamine in a 40mL of anhydrous THF. The reaction mixture was then cooled down to 0 oC with aice bath. Afterward, 0.0075 mol of acryloyl chloride was added drop wise under acontinuous stirring over a period of 15 min. Zinc sample was then introduced in toa solution and reaction temperature was raised to room temperature by removingthe ice bath. After 24 h, the sample was removed, cleaned sequentially with excessacetone and ethanol to remove physisorbed molecules, and dried under a pressurizednitrogen stream.

Synthesis of polymer coatings

For polymerization reactions, initially toluene (40 mL) was degassed by performingthree freeze-thaw-pump cycles. Subsequently, MMA/styrene was added to reacha final concentration of 0.007 M, and AIBN was added to the solution in a con-centration of 1 wt% relative to MMA/styrene. The mixture was stirred until ahomogeneous solution was obtained. A functionalized zinc sample was then intro-duced into the reaction mixture and the temperature was raised to 70 C in order toinitiate a polymerization reaction. After 24 h the zinc sample was removed, washed

30


with excess acetone, followed by ultrasonic cleaning for 1 min in a toluene solutionin order to remove any remaining traces of polymer which is not covalently linkedto the zinc. Later, the samples were dried in a nitrogen stream.For the preparation of polymer coatings with different cross-linking densities on

zinc substrate, the same polymerization procedure was employed, with different mo-lar fractions (25% or 50%) of the respective cross-linking agent (EDA or HDA) addedto the solution. The total monomer concentration (EDA/HDA + MMA/styrene)was kept constant.

3.4.3 Synthesis of ZnO nanostructuresHydrothermal growth of ZnO nanostructures

A solution containing zinc nitrate hexahydrate Zn (NO3)2 .6H2O and hexamethylen-etetramine (HMT) was prepared in ultrapure water, containing each at a concen-tration of 50 mM. The solution was stirred at room temperature with a magneticstirrer at 1200 rpm until a clear solution was obtained (10 min). For the synthesis ofrod-like ZnO, the obtained solutions were directly heated (see below) without anycapping agent. For the synthesis of twin-plate-shaped ZnO, gelatin was dissolved inthe obtained solution with a concentration of 5 g.L−1. For the synthesis of sphericalZnO particles, a 40 mM solution of trisodium citrate (Na3C6H5O7) in ultra-purewater is mixed with the ZnNO3/HMT solution with a volume ratio of 1:2 in a glassbeaker.Next, a clean polished Zn sheet substrate was put into the respective solution in a

glass vial. The vial was tightly closed and then placed inside an oven for 90 min at atemperature of 90 oC. After letting the solution cool down to room temperature, themodified zinc substrates were removed from the solution and washed several timeswith water and then with ethanol to remove contaminants. Finally, the sampleswere dried in a nitrogen stream and kept in a desiccator until used.

Growth of ZnO nanorods via tailored corrosion route

Initially polished zinc sheets (1 cm × 1 cm × 1.5 mm) were spin coated with 5weight% solution of PVB in ethanol, followed by drying, to obtain a ≈2 µm thickpolymer layer. Afterwards, the coated sample was placed on a zinc base plate and areservoir was built at the edge of the sample for electrolyte insertion with the helpof fast drying two component adhesive. Subsequently, a small defect was createdat the edge of the coated sample with the help of a blade. The assembly wasthen placed into a self-made humidity chamber. In order to initiate the growth ofZnO nanorods via a tailored corrosion process, few drops of 0.1 M (unless notedotherwise) aqueous KCl solution was added. Finally, after 6-8 h the electrolyte wasremoved and substrates were washed with water and dried in an N2 flow at roomtemperature.

31


3.4.4 Electrochemical evaluation of stability of polymer coatingsThe electrochemical stability of coatings were carried out by using SKP, linear po-larization and EIS (unless otherwise noted).Prior to SKP experiments, zinc samples were spin coated with 5 wt% poly(vinyl

butyral) [PVB] in ethanol to yield a ≈1 µm thick polymer film. This PVB coatingprevents the spreading of the electrolyte above the modified zinc samples. Unmodi-fied zinc substrates were spin coated with PVB and delamination experiments wereperformed under similar conditions to evaluate the effect of PVB alone. In orderto initiate a cathodic delamination process, an artificial defect was created at theedge of the sample with a scalpel, and the defect was filled with 1 M KCl. Thesamples were subsequently introduced into a SKP chamber at 92% to 95% relativehumidity and at a temperature around 23 C. The progress of the delamination frontwas analyzed as described elsewhere12. As position of the delamination front, thefirst point which shows the potential of the intact interface was used. Typically, 3to 5 samples for each given preparation were analyzed. The results presented areresults of the median sample, i.e. after sorting the samples in the order of increasingstability, the sample with the middle performance was selected for presentation.

3.4.5 Photocatalytic decomposition of methyl orange (MO)Photocatalysis experiments were performed in a home made continuous flow setupwhich is depicted schematically in Figure 3.1. Zinc foil with ZnO nanorods grown ontop was placed on the bottom of a crystallization dish. Next to the foil, a magneticstirrer was placed on the bottom of the dish. Initially, a 10−5 M aqueous solution ofmethyl orange (20 mL) was added into the dish, and the solution was stirred withthe help of a magnetic stirrer throughout the experiment.

Light Source(633 nm)

PerstaticPump

UV/Vis Spectrometer

Photocatalyst

Water+

MethylorangeCuvette

Figure 3.1: Schematic representation of in situ photocatalysis experiment.

Before starting the photocatalytic decomposition of MO, ZnO nanorod arrayswere immersed in to the reaction chamber for ∼30 min without illumination to

32


establish adsorption/desorption equilibrium. Part of the MO solution was thenpumped into a quartz flow cell, placed inside a UV/visible spectrometer (PerkinElmer Lambda 900), using a peristaltic pump at a flow rate of 50 µL s−1. To ensurethe presence of only a specific visible wavelength of light, the sample in the glassreservoir was illuminated with a HeNe laser (633 nm, 1.96 eV) of 20 mW poweroutput with a beam expander, yielding a power density of ∼10 mW cm−2 at thelocation of the ZnO nanorods. The degradation of the dye was monitored using theUV/visible spectrometer by measuring the absorption spectra at different intervals,and evaluating the peak absorbance at 464 nm. The molar absorption coefficient ofMO at 464 nm was determined as 4 · 105 L·mol−1·cm−1 from a concentration series.

33

Chapter 4: Delamination kinetics of ultrathinPMMA model coatings on zincbound via organosilanes∗

4.1 IntroductionCathodic delamination studies have recently focused on the role of ion transport13,137,including quantification of transport rates138. Migration, rather than diffusion hasbeen shown to be an important process for transport along the interface139. Theeffect of surface morphology was also investigated19,140. Comprehensive parameterscans have been carried out141. For a better understanding of the chemical pro-cesses, delamination experiments have been combined with infrared spectroscopy142.Mathematical models of the electrochemical processes have been developed, whichpermit a detailed analysis of the kinetics143. Most studies leave the surface chem-istry ill-defined, or use complex industrial coatings, which permit only a globalunderstanding of the processes. Nevertheless, the detailed understanding of theelectrochemistry has lead to novel coatings with self-healing properties8,144.Investigations of the direct linkage of coatings to the surface is subject of current

interest94 and, e.g., bifunctional silane linkers have been found to decrease degra-dation rates95. To overcome problems with the definition of interfacial chemicalbonds, here an approach were an ultrathin polymer coating was synthesized is pre-sented. Metallic zinc substrates were modified with vinyltrimethoxysilane (VTS),defining the covalent linkage at the metal(oxide)/polymer interface. Subsequently,these surface-functionalized zinc substrates were subjected to a thermally-initiatedcopolymerization of methyl methacrylate (MMA) in the presence or absence of cross-linkers. Resulting poly(methyl methacrylate) [PMMA] coatings have been charac-terized and their delamination behavior has been studied using SKP and electro-chemical linear polarization experiments.

∗This chapter contains results published as a scientific article:Iqbal, D. et al. Synthesis of ultrathin poly(methyl methacrylate) model coatings bound viaorganosilanes to zinc and investigation of their delamination kinetics. ACS Appl. Mater.Interfaces (2014). doi:10.1021/am504992r

35

Chapter 4. PMMA model coatings bound via organosilanes

4.2 Results and discussion

4.2.1 Surface modificationZinc samples were first treated with a reactive silane coupling agent (VTS) thatcontains a polymerizable vinyl group. As shown in Figure 4.1, the zinc surface wasmodified with VTS via hydrolysis of the monomer in ethanol/water (90:10) mixtureat room temperature145–149. The addition of water to the reaction mixture promotesthe hydrolysis of methoxy groups that are present in VTS, resulting in the formationof Zn-O-Si bonds, as will be shown below.

n = 2 for EDA

n = 6 for HDA

+

ethanol

+

water


O

SiO O

O

SiO O

O

O

O O

O

On

O O

O

On

Si

O

O O

O

OO

O

n


Figure 4.1: Scheme of the surface modification steps carried out in this chapter.

To examine the composition of the functionalized zinc surface IR, XP and massspectra were recorded. Figure 4.2a shows an IR spectrum of a VTS-modified zincsurface. The bands at 3020, 2965, 2921, and 2850 cm−1 are assigned to C–H stretch-ing vibrations150. Furthermore, the peaks at 1265, 1110, 1018, and 824 cm−1 aremodes typically observed for organic siloxane-based structures151. The modificationof the zinc surface with VTS was further verified by conducting XPS measurements.Figure 4.2b shows a high-resolution XP spectrum of the Si 2p signal at the func-tionalized zinc sample. The presence of this Si 2p signal can be clearly identified,indicating the presence of the silicon on the zinc surface.

36

4.2. Results and discussion

3500 3000 2500 2000 1500 1000-0.001

0.000

0.001

0.002

0.003

82

4

96

010

3211

10

12

65

29

2129

65

Ab

so

rba

nc

e

Wavenumber / cm-1

(a)

110 108 106 104 102 100 98 96

800

1000

1200

1400

1600

Inte

ns

ity

/a

rb.

un

its

Binding energy / eV

Si 2p

(b)

Figure 4.2: (a) IR spectrum of VTS-modified zinc. (b) Si 2p region of the XPspectrum of VTS-modified zinc. The presence of the signal confirms thepresence of silane on the surface.

To further verify the presence of chemical bonding between zinc and the silanecoupling agent, ToF-SIMS analysis was performed. Initially, the top layer of VTSwas sputtered away by ion etching, to analyze the interfacial region between thenative oxide on zinc and the organosilane. In the region between organic film andnative oxide, the fragment ZnOSi+ was detected, as shown in Figure 4.3. The ionpattern of the fragment ZnOSi+ should reflect the isotopic abundance of the elementspresent; the observed distribution is in close agreement to the calculated distribu-tion152. Deviations at the peaks of species with lower abundance at ≈109 amuand ≈111 amu are attributed to contributions of organic fragments to the totallymeasured count rate. The presence of the ZnOSi+ fragment is an indication of theformation of a covalent bond between organosilane and the native oxide on zinc152.It is known that hydrolyzed alkoxysilanes may react with each other to form a densepolymeric film on the surface146, while care is needed to ensure the formation of amonolayer136,153. Here, ellipsometric analysis yields a thickness of 2-3 nm of thesilane layers, clearly above monolayer coverage, but still significantly below 10 nm.Formation of a covalent bond between PMMA and the VTS-modified surface was

achieved by performing a thermally initiated free radical polymerization of MMAin the presence of the modified zinc substrate with AIBN as initiator (see Figure4.1). Two fractions of PMMA after the copolymerization reaction were obtained,a covalently surface-bound fraction and a fraction without covalent bond to thesurface. The fraction that is physisorbed on the surface was washed with excesstoluene to remove the nonbounded polymer. Figure 4.5 Ia shows the measuredIR spectra after copolymerization. Appearance of characteristic C=O stretchingmode as a band at 1735-1740 cm−1 is an indication of the presence of an acrylatebound to the zinc sample150. In addition to copolymerization with the monofunc-tional monomer, polymerization was also carried out in the presence of bifunctionalacrylates with different chain length and different molar fractions (25% and 50%).

37


107 108 109 110 111 112

0

10

20

30

40

50

66ZnO

30Si

+

67ZnO

29Si

+

68ZnO

28Si

+

66ZnO

29Si

+

67ZnO

28Si

+

64ZnO

30Si

+

66ZnO

28Si

+

64ZnO

29Si

+

Co

un

ts

m z-1

/ amu

64ZnO

28Si

+

Figure 4.3: SIMS showing the isotope distribution of ZnOSi+ fragments from thesurface of VTS-modified zinc. Expected isotope distribution is shown asvertical lines (|) ca. 0.3 amu below the experimental peak.

1200 1000 800 600 400 2000

1800

3600

5400

7200

9000

10800

O KLL

O1s

Inte

nsity

/ ar

b. u

nits

Binding energy / eV

C1s

Figure 4.4: XPS survey spectrum on chemisorbed PMMA coating without cross-linker. Absence of substrate peak indicates the uniform coverage of thesurface.

38


The presence of such monomers in the copolymerization results in films consistingof cross-linked polymers. Figure 4.5 Ib-e shows the IR spectra of polymer filmsprepared in the presence of cross-linker. Furthermore, XP spectra (Figure 4.5 II)of polymer-modified samples show the expected C 1s and O 1s peaks. The highresolution C 1s spectrum can be decomposed in to four distinct peaks at 284.8 eV,285.4 eV, 286.4 eV and 288.8 eV that originate from aliphatic carbon, C–COO, C–O,and COO, respectivelyThe fraction of COO was quantified to be in ranges between 8 and 17% of the

total carbon, without a clear trend between the different polymers, but with highrepeatability between different samples of the same polymer. This fraction is slightlylower than expected (for HDA homopolymers 16.7%, for pure PMMA 20% and forEDA homopolymers 25%). The fraction detected by XPS is, however, also affectedby the distribution of the elements inside the thin films. The results here mayindicate that carboyxl groups may be enriched inside the film, and oxygen-poor parts

( )I ( )II

294 292 290 288 286 284 282

(a)

Binding energy / eV

(b)

(c)

Inte

ns

ity

/a

.u. (d)

(e)O C O/O C OC O

CC

O

OC C / C H

1x103

Inte

ns

ity

/a

.u.

/

Figure 4.5: (I) IR and (II) C 1s XP spectra of different covalently-bound polymercoatings on zinc, (a) PMMA without cross-linker, (b) PMMA + 25%EDA, (c) PMMA + 50% EDA, (d) PMMA + 25% HDA, (e) PMMA +50% HDA.

39


Table 4.1: Polymer film thickness obtained from ellipsometric measurements forPMMA samples with different cross-linker ratios

Sample Thickness / nmPMMA without cross-linker 25± 525% EDA 12± 250% EDA 9± 225% HDA 14± 350% HDA 12± 3

of the polymer chains may be present directly at the surface. Initiator fragmentsinside the film may also lower the COO fraction. In all samples after polymerization,the Si 2p peak remains centered at 102 eV, as shown for the VTS-modified surfacein Figure 4.2b. This peak position is typical for organosilanes indicating that thesestructures are still present in an unmodified form after polymerization. In the XPspectra, Zn peaks are absent expect for films with a thickness <10 nm, confirminga full coverage of the zinc with polymer on the level of XPS sensitivity (see Figure4.4).The thickness of the obtained polymer films was analyzed by ellipsometry. As

shown in Table 4.1, increase in the cross-linker fraction results in the reduction ofpolymer film thickness. This observation is tentatively attributed to a higher rateof quenching of the propagation reaction in case larger amounts of cross-linker areused154. Ellipsometric spectra could be well-described by a simple ambient-film-substrate model, which confirms that the films are not inhomogeneous on the lengthscale of several 10s of nm.

4.2.2 Model corrosion experimentsIn order to study the stability of synthesized silane-bonded PMMA coatings againstcathodic delamination, the propagation of a delamination front from a defect filledwith 1 M KCl was monitored by SKP measurements. Figure 4.6 shows the SKP po-tential profile recorded on spin coated PVB, as well as on covalently bound PMMA.Scans were obtained by scanning along the surface from the artificial defect, andshow the recorded potential as a function of distance from the defect for differenttimes. In the initial stages of the experiment, the SKP measures the potential of anintact (i.e. non-corroded) polymer/oxide/metal interface all over the sample. How-ever, with time, cathodic delamination is initiated in a humid air atmosphere onboth samples as shown in Figure 4.6. Consequently, during passage of a delamina-tion front, two distinct potential levels are observed12. The more negative potential(≈ -0.7 V) indicates actively corroding zinc in the region were the polymer is alreadydelaminated, while the more positive potentials (≈ -0.2 V) indicates an intact poly-mer/metal interface12. Between these two potentials, a sharp transition region canbe observed, which is the position of the delamination front12. The progress of the

40


0 2000 4000 6000 8000-0.8

-0.6

-0.4

-0.2

0.0

E/

Vv

s.

SH

E

Distance / µm

intiation of delamination

scans with interval of 120 min

after 18h

(a)

0 2000 4000 6000 8000 10000 12000

-0.90

-0.75

-0.60

-0.45

-0.30

E/

Vv

s.

SH

E

Distance / µm

initiation of delamination


after 48 min

(b)

Figure 4.6: SKP delamination profile of (a) covalently bound PMMA on zinc and(b) weakly bound PVB used for reference purposes, both in air.

delamination front with time t is plotted in Figure 4.7a for several coatings. Plottingthe same results on a double logarithmic scale (Figure 4.7b) enables to determinethe exponent α of the time dependence. While for diffusion controlled processes,the position d of the delamination front travels as d ∼ t1/2 (α = 1/2)100, a time de-pendence as d ∼ t1 (α = 1) is expected if a first-order reaction is rate-determining.Alternatively, d ∼ t1 may indicate migration in a constant static electric field. Here,exponents α ≈ 1 are found, with maximum 5% deviation for all except one portionof a curve, which will be discussed below.Hence, the delamination rate r was determined from the slope of a linear fit in

Figure 4.7a. The rates for different samples are shown in Table 4.2. Figure 4.7ashows that the progress of the delamination front is inhibited on zinc substrate mod-ified with covalently-bound PMMA without cross-linker (r = 0.27± 0.02 mm h−1),compared to the reference sample with spin-coated PVB (r = 11.0 ± 0.05 mm h−1)

41


and spin-coated PVB on VTS-modified zinc (r = 0.52 ± 0.02 mm h−1). Whenchanging from humid air atmosphere to humid nitrogen, no cathodic delaminationwas observed. In this case, the potential remained at the potential of the intactpolymer/metal interface. (Attempts to produce PMMA films by spin coating whichwere not covalently linked but have a thickness comparable to the thickness of thesynthesized layers failed.)

0 200 400 600 800 1000 1200 1400

0

1000

2000

3000

4000

5000no CL

VTS

25% HDA

50% HDA

25% EDA

50% EDA

PVB

De

lam

ina

tio

nfr

on

t/

µm

Time / min

(a)

10 100 100010

100

1000

10000

d ~ t1/2

De

lam

ina

tio

nfr

on

t/

µm

Time / min

no CL

VTS

25% HDA

50% HDA

25% EDA

50% EDA

PVB

d ~ t(b)

Figure 4.7: Comparison of the delamination rates for polymer coatings with differentcross-linker amounts (a) on a linear scale and (b) on a double logarithmicscale. In the plot (b), two dotted lines show the expected slopes for d ∼ tand d ∼

√t, as indicated in the graph.

In order to gain further insight into the effect of network chain length and degreeof cross-linking on the stability of these coatings against corrosion in general andagainst cathodic delamination in particular, additional coatings were synthesized

42


Table 4.2: Delamination rates and standard deviations for the different samples,based on linear fits of plots as in Figure 4.7. All samples were coveredwith PVB, as explained in the experimental section.

Sample Delamination rate / mm h−1

pure PVB 11.2± 0.5VTS 0.52± 0.03no cross-linker 0.27± 0.0225% EDA 0.22± 0.0250% EDA 0.042± 0.00425% HDA 0.039± 0.00350% HDA 0.028± 0.004

on VTS-modified zinc substrates with two different cross-linker chain length. Thedifferent molar ratios employed should result in a different degree of cross-linking.Delamination was monitored by SKP, and the delamination rates obtained are in-cluded in Table 4.2. In one case of the cross-linked samples, namely in the case ofcoatings containing 50 % HDA cross-linker, a t1 behavior is found initially, whichat larger times changes to t1/2. Such a transition is observed only for this partic-ular type of samples. Overall, the results show a higher rate of delamination forEDA-cross-linked samples compared to HDA-cross-linked samples. That is, poly-mer coatings with a shorter cross-linker chain length delaminate faster than coatingswith a longer chain between the two polymerizable functional groups of the cross-linker. Moreover, an increase in the amount of cross-linker leads to a decrease inthe delamination rate. EDA-cross-linked samples delaminate, however, at a similarrate than non-cross-linked samples.Furthermore, investigation of the samples after a delamination experiment with

scanning electron microscopy reveals the growth of crystalline ZnO corrosion prod-ucts. The growth of zinc oxide/hydroxide structures after cathodic delaminationhas been shown in the past12,140. The corrosion resistance of different coatings wasalso evaluated by performing linear polarization experiments. Typical Tafel plotsobtained for bare zinc and coated zinc are shown in Figure 4.8 (It must be noted thatas opposed to the SKP experiments, no PVB coating was present for these experi-ments.). Corresponding corrosion current rates were calculated from the polarizationresistance130, and are tabulated in Table 4.3. The highest value of corrosion cur-rent were naturally recorded for bare, uncoated zinc. Consequently, zinc oxidationrates (from oxide formation and dissolution) as well as the accompanying cathodicreaction, most likely the oxygen reduction reaction in this system are highest forpure zinc. Overall corrosion currents are lower for coated zinc samples, showingthat indeed the coatings have the expected corrosion-inhibiting effect. Comparingthe corrosion currents of different samples (Table 4.3) show that all cross-linkedcoatings exhibits lower corrosion currents than coatings without cross-linking. The

43


Table 4.3: Corrosion current densities icorr for differently coated samples in 0.1 MKCl. dec= decade of current. “no cross-linker” refers to non-cross-linkedPMMA.

Sample βa / mV dec−1 βc / mV dec−1 Rp / Ω cm2 icorr / µA cm−2

zinc uncoated 104± 4 176± 3 317± 15 89± 5no cross-linker 105± 4 196± 8 780± 60 38± 425% EDA 43± 3 161± 2 860± 30 17± 150% EDA 54± 2 168± 3 1120± 30 15.8± 0.725% HDA 62± 2 150± 3 1560± 50 12.2± 0.650% HDA 58± 2 137± 3 4200± 200 4.2± 0.3

sample with the lowest overall corrosion current is the same that shows the lowestdelamination rate. However, the difference between EDA and HDA is not clearlyvisible in the corrosion currents. Both coatings also inhibit the dissolution as ex-pected, which shows in the lower currents in the anodic branch of the plots in Figure4.8.

-0.95 -0.90 -0.85 -0.80 -0.75 -0.70

1E-3

0.01

0.1

Zinc (without coating)

PMMA without crosslinker

25 % EDA

50 % EDA

log

10

[i

/m

Ac

m-2

]

E / V vs. SHE

(a)

-0.95 -0.90 -0.85 -0.80 -0.75 -0.70

1E-3

0.01

0.1

Zinc (without coating)

PMMA without crosslinker

25 % HDA

50 % HDA

log

10

[i

/m

Ac

m-2

]

E / V vs. SHE

(b)

Figure 4.8: Polarization curves in the form of Tafel plots in 0.1 M KCl for zinc withdifferent coatings, (a) with EDA cross-linker, (b) with HDA cross-linker.

The typical Bode plots of the investigated materials after 2 h of immersion aregiven in Figure 4.9. The 50% HDA cross-linked polymer indicates a prevailing ca-pacitive behavior. The corresponding EIS spectrum shows only one time constantafter 2 h of immersion. However, in case of the other samples the phase angle isstarting to decrease at lower frequencies, a consequence of the electrolyte ingressinto the polymer. As a consequence of the penetrating solution, electrochemical re-actions can take place at the metal surface, resulting in the presence of a second timeconstant in the Bode plots155. The first process at high frequencies can be attributedto the coating capacitance and resistance. The second process at low frequenciescan be related to the charge-transfer resistance and the electrical double-layer ca-

44


10-1

100

101

102

103

104

105

-80

-60

-40

-20

0

50% HDA

25% HDA

50% EDA

25% EDA

0% CL

Zn+PVB

Zn

Am

pli

tud

e/

°

f / Hz

10-1

100

101

102

103

104

105

101

102

103

104

105

106

107

|Z|

/O

hm

cm

-2

f / Hz

50% HDA

25% HDA

50% EDA

25% EDA

0% CL

Zn+PVB

Zn

(a) (b)

Figure 4.9: Bode plots (top: modulus of impedance |Z|, bottom: phase) for theselected polymer systems as indicated in the graph after 2 h of immersionin borate buffer. The straight lines represent the fitted curves of thecorresponding sample.

pacitance at the metal/solution interface.155 Moreover, based on the decrease of theimpedance modulus at lowest frequencies (10−1 Hz), a comparison of the protectiveproperties of the selected systems can be made, via the charge transfer resistance.As a tendency, the more hydrophobic HDA cross-linker is granting better protectionfrom water ingress than EDA cross-linker. Higher cross-linker concentrations alsoslow down the penetration of the electrolyte to the metal surface.Plasma polymers are ultrathin polymer films which are typically highly cross-

linked. Charge transfer/polarisation resistance found here for coatings by free radi-cal polymerization are in similar order of magnitude comparing with literature dataon plasma polymers156. Slowest delamination rates found here are of same order aspublished results for plasma polymers on steel with even lower electrolyte concen-tration.101

4.2.3 Discussion of the chemistry of cathodic delaminationFirst, it is obvious that the presence of a coating on metallic zinc reduces the cor-rosion rates, as it inhibits interfacial electron transfer reactions through blockingtransport paths and active sites. Second, a covalently bound coating delaminatesconsiderably slower than a coating which is attached to the surface by van der Waalsforces only. It is known that modification of oxide-covered metal surfaces with cova-lently bound polymers inhibits the oxygen reduction reaction at the polymer/metalinterface as a result of a strong chemical bond157–159.The linear dependence of the delamination front location from the experiment

time shows that diffusion (e.g. of oxygen or water through the coating) is notrate limiting100. Indeed, for low oxygen partial pressure, the contribution of iondiffusion to overall transport was reported to be overestimated13. Consequently, thedifferences observed for delamination rates and corrosion currents between the cross-

45


linked polymer coatings cannot be explained by differences in transport through thethin polymer films. Likewise, the relatively small differences between the polymercoating thicknesses cannot explain the observed differences, as the thinner coatingswith high molar ratios of cross-linkers show lower corrosion currents than the slightlythicker coatings.The overall amount of hydrophobic groups is one factor contributing to the ob-

served differences between the coatings. Large fractions of HDA in the monomersolution lead to the presence of a large fraction of hydrophobic alkyl chains in thefinal polymer. EDA as cross-linker on the other hand does not significantly alter hy-drophilicity of the polymer layers compared to PMMA. Hydrophobic polymer layerscannot take up as much water as less hydrophobic coatings. Lower availability ofthe solvent water during the corrosion process is decreasing the corrosion rate andthe delamination rate. The coating which is expected to be the most hydrophobicchanges to a transport-controlled delamination mechanism at longer exposure.The behavior of the coatings in linear polarization experiments and in delam-

ination experiments is not the same. Coatings without cross-linker show similardelamination rates as EDA-cross-linked coatings, while their corrosion currents aretwice as high. As cross-linked coatings block transport routes in between differentpolymer chains, they decrease the corrosion current, but not the delamination rate.Delamination rates do overall roughly follow the same trends as observed for thelow-frequency impedance modulus, which reflects the charge transfer resistance.The observed t1-delamination kinetics may be determined by migration of ions in

a constant electrostatic potential gradient, or by a first-order chemical reaction. Formigration in the electric field, ion flux is proportional to the electric field gradient(e.g.139). When comparing the delamination curves (e.g. shown in 4.7), it becomesobvious that the potential difference between active and non-active region is verysimilar for all polymers. With the lateral resolution used here, the distance abovewhich the potential drop is observed is also quite similar for the different samples.Hence, the differences in delamination rate between the different samples caused bymigration in the electric field alone should be rather small. Therefore focus is onthe discussion of the chemical reactions as responsible for the observed differencesbetween the coatings.Considering cathodic delamination, two types of reactions could be rate deter-

mining, either the oxygen reduction reaction (ORR), which drives the delaminationfront, or the deadhesion process, which involves breaking of covalent bonds in thereactions here. It is hard to see why the rate of the ORR should depend on theamount of cross-linker present in the preparation. Instead, the ORR rate, as anyrate of an electron transfer reaction, is a sensitive probe of the state of the interface,especially of defects in the layers, i.e. free metal surface which is available to the elec-tron transfer reaction. While it is possible that the presence of the cross-linker in thecopolymerization reaction modifies the interface, it is less likely. All surfaces weresubjected to identical treatments before polymerization, so the metal(oxide)/silaneinterface is supposed to be identical. In the actual deadhesion process, several pro-

46

4.3. Summary

cesses could potentially occur. It is known that the hydoxide as reaction productof the ORR leads to strongly alkaline conditions in the vicinity of the delaminationfront12. Hydroxide could involve in a number of follow up reactions: (a) in thealkaline dissolution of zinc oxide under hydroxozincate formation according to

ZnO + 2OH− + H2O −−→ Zn(OH)42−, (4.1)

or (b) in the alkaline hydrolysis of Si−O bonds under silicate formation accordingto

R(1)−Si(−OR(2))3 + 3H2O OH−−−−→ R(1)Si(OH)3 + 3R(2)OH. (4.2)

Both reactions are possible at pH > 11.160 Both dissolution of ZnO and hydrolysisof Si-O bonds are not expected to be affected by the presence of cross-linked struc-tures in the polymer film. In both cases, the inorganic part of the metal/polymerinterface will dissolve. A further degradation reaction is an alkaline ester hydrolysisof the PMMA under formation of carboxylate and alcohol according to

R(1)−COO−R(2) + OH− −−→ R(1)−COO− + R(2)OH. (4.3)

This reaction does, however, not lead to a breakage of the polymer main chain,and hence not to de-adhesion of the complete coating. Alkaline ester hydrolysiswould, however, affect the cross-linking, as both cross-linkers link polymer chainsvia ester bonds. As an alternative to processes involving OH−, radical reactions areother candidates which could lead to the deadhesion of the polymer films. Radi-cal intermediates have been shown to occur in the ORR161,162. Here, only radicalreaction can explain all the results without involvement of a complex coupling ofpolymer degradation and e.g. zinc oxide dissolution. Previous reports have shown adecrease in delamination rate in the presence of radical scavengers163.

4.3 SummaryFree radical copolymerization of dissolved and surface-bound monomers has beenused to synthesize PMMA coatings with thicknesses in the order of 10-20 nm onthe industrially relevant metal zinc. Adding certain amounts of cross-linkers enablesa systematic variation of the degree of cross-linking in the resulting coatings. Thecoatings prepared here are covalently linked to the base metal via organosilanes; suchcoatings show a larger resistance towards cathodic delamination than coatings boundexclusively via van der Waals interactions. The delamination rate decreases to<1/10only by introducing covalent bonds to the base metal. Introduction of cross-linkersfurther slows down delamination, with another factor of ∼10 between non-cross-linked PMMA and samples with 50% cross-linker. Comparing the two cross-linkersused in this study, the more hydrophobic HDA yields the lowest delamination ratesof all samples investigated in this study, of 0.028± 0.004 mm h−1.

47

Chapter 5: Delamination kinetics of model PScoatings on zinc bound via organosi-lanes

5.1 IntroductionThe barrier properties of polymer coating against aggressive species like oxygen andwater plays a vital role in defining the final corrosion protection properties of acoating22. The permeability of these species through the coating is defined by thestructure and morphology of the protective organic film117. The diffusion of gaseslike oxygen, nitrogen e.t.c is governed by a diffusion process present in the poly-mer, as these species hardly interact with the polymer. Therefore, the structuralattributes such as holes/defects in a polymer structure could significantly increasethe diffusion of oxygen through the coating. On the other hand, permeability ofwater is strongly governed by the solubility due to strong interaction with the poly-mer117. It has been shown that although water and oxygen may not be the ratedetermining step in corrosion, however they are essential for initiation of a corrosionreaction. Water is essential to establish a electrochemical corrosion cell beneaththe coating164 and oxygen is to complete the corrosion reaction117. In particular,close packing of polymer chains and higher fractions of hydrophobic groups leads toenhanced resistance against permeation of water and oxygen165. Moreover, it hasbeen shown that polymers containing stiff chain molecules (such as aromatic rings)tends to significantly reduce the diffusion of water molecules through the polymerchains166.In the previous chapter, the effect of chemisorbed linear and cross-linked PMMA

coatings on the cathodic delamination rate has been studied. PMMA is a aliphaticpolymer that has higher chain flexibility when compared to aromatic polymers suchas PS. Since aromatic polymers compared to aliphatic polymers exhibits enhancedbarrier properties against water and oxygen117, it is expected that they also showenhanced resistance against cathodic delamination process. Hence, the aim of thiswork is to study the effect of PS polymer, linked via silane bond to the zinc surface,on the cathodic delamination kinetics of a coating. Moreover, cross-linked PS coat-ings were also synthesized by addition of aliphatic bifunctional molecule during thepolymerization reaction.

49

Chapter 5. PS model coatings bound via organosilanes

n = 2 for EDA

n = 6 for HDA

+


O

SiO O

O

SiO O

n

O O

O

On

O O

O

On

Figure 5.1: Schematic representation of the synthetic route for the synthesis of silanebonded linear and cross-linked PS coatings.


5.2.1 Synthesis of polymer coatingsThe modification of zinc substrate with VTS and characterization with IR, XPS andTOF-SIMS were carried out using the synthetic route explained in chapter 4.After successful functionalization of zinc surface with VTS, PS was grown on top

of functionalized zinc surface via thermally initiated free radical polymerization ofstyrene monomer in presence of AIBN as a thermal initiator (see Figure 5.1). As inthe case of free radical polymerization of MMA explained in chapter 4, two fractionsof PS polymer were formed. One fraction was covalently bonded to the substrate,due to surface initiated polymerization reaction, while the other fraction was eitherphysisorbed on the surface or present in the solvent phase. The physisorbed polymerwas removed by access cleaning with acetone and toluene.In addition to linear PS polymer, cross-linked PS coatings were synthesized on top

of functionalized zinc samples by adding a bifunctional cross-linking molecule at thestart of polymerization reaction that leads to cross-linked polymer coatings. Thebifunctional molecules used were of different chain lengths (i.e. EDA and HDA)

50


3000 2500 2000 1500 1000

(e)

(d)

(c)

(b)

C-H

ben

d

C-O

str

etc

h

C-C

/C=

C

CH

str

etc

haro

mati

c

C-H

ben

d

C=

O

Wavenumber / cm-1

str

etc

h

aro

mti

cstr

etc

h

str

etc

h

CH

(a)

Ab

so

rpti

on

Figure 5.2: IR spectra of different covalently-bound polymer coatings on zinc, (a)PS without cross-linker, (b) PS + 25% EDA, (c) PS + 50% EDA, (d)PS + 25% HDA, (e) PS + 50% HDA.

Table 5.1: FTIR stretching vibration position (cm−1) of linear and cross-linked PScoatings synthesized on zinc surface based on refrence150.

Peak Assignment PSPS+

25% EDA

PS+

50% EDA

PS+

25% HDA

PS+

50% HDA

C=C-C [s] 15891494

15921490

15791492

15981490

15941444

C-H [s, aromatic] 3060 3060 3058 3062 3056CH2 [as] 2917 2922 2923 2922 2921CH2 [ss] 2848 2848 2854 2851 2850CH2 [bend] 1446 1445 1446 1450 1444C=O [s] - 1737 1737 1731 1731*s = stretching, ss = symmetric stretching, as = antisymmetric stretching

51


294 292 290 288 286 284 282 280

(e)

(d)

(c)

(b)

Inte

ns

ity

(a.u

.)

Binding energy / eV

C-C/C-H

O-C=O

C-O

(a)

C-COO

Figure 5.3: C 1s region of the XP spectrum of zinc with different polymer coatings,(a) PS without cross-linker, (b) PS + 25% EDA, (c) PS + 50% EDA, (d)PS + 25% HDA, (e) PS + 50% HDA. The line represents the respectivepeak position for the atom shown in bold on its label.

and were added in different molar ratios (i.e. 25 mol% and 50 mol%) to yieldpolymer coatings with different morphology and cross-linking densities. Linear andcross-linked polymer coated zinc samples was then characterized by IR and XPSto determine the chemical composition of the surface. Figure 5.2a-e shows themeasured IR spectrum on chemisorbed PS coating grafted to zinc substrate viasilane bond. The appearance of bands at 1589 cm−1 and 1494 cm−1 is assignedto the aromatic ring stretching arising from the ring structure present in the PSpolymer150. Moreover, there is an additional band present at around 3060 cm−1 thatis assigned to the C-H stretching mode of aromatic ring150. Hence, the presence ofthe aromatic ring stretching modes confirms the successful growth of PS polymercoating on top of the silane modified zinc substrate. In case of cross-linked coatings,the presence of C=O stretching mode at around 1725 cm−1 confirms the presence

52


Table 5.2: Thickness of as synthesized linear and cross-linked PS coatings obtainedfrom ellipsometric measurements.

Sample Thickness / nmPS without cross-linker 15± 325% EDA 10± 250% EDA 8± 225% HDA 8± 150% HDA 10± 3

of ester functionality that comes from the cross-linking molecule. The detailedassignments of IR spectra is summarized in Table 5.1.Figure 5.3 shows the high resolution C 1s XPS spectra of chemisorbed linear and

cross-linked PS coating grown on zinc substrate. The peak is deconvulated in tofour sub peaks as show in Figure 5.3a-e. C-H peak is assigned at 284.8 eV whicharises from the polymer backbone and partially from the impurities present on thesurface due to exposure to open air. Carbonate peaks are assigned at 285.4 eV,286.4 eV and 288.8 eV arises from the presence of acrylate based cross-linker in thepolymer coating.The thickness of the as synthesized polymer coatings was determined by ellipsom-

etry and obtained values are summarized in Table 5.2.

5.2.2 Model corrosion experimentsThe stability of these as synthesized chemisorbed linear and cross-linked coatingsagainst cathodic delamination was characterized by SKP. Prior to the SKP experi-ments, polymerized zinc samples were spin coated with PVB to yield ≈1 µm thickphysisorbed polymer film to prevent the immediate spreading of an electrolyte. SKPexperiments were performed under similar experimental conditions as carried out forchemisorbed PMMA coatings explained in chapter 4. Figure 5.4 shows the SKP po-tential profile recorded on spin coated PVB, as well as on covalently bound linearPS coating. The delamination rates were calculated from the plot of progress indelamination front with time and is summarized in Table 5.3. Similarly, the sameresults were plotted on a double logarithmic scale (see Figure 5.5b) which enables todetermine the exponent α of the time dependence. Delamination rates are initiallyreaction controlled, while on larger time scales, cathodic delamination process be-comes diffusion limited. This phenomenon is attributed to the higher hydrophobicityof PS polymer due to presence of aliphatic chains that exhibits lower water solu-bility. Hence, on larger times scales, the cathodic delamination process transformsfrom reaction controlled to diffusion limited process.Figure 5.4 shows the progress of delamination front for physisorbed PVB coat-

ing on zinc as well as for chemisorbed linear PS coating. It is quite evident fromthe Figure 5.5 that physisorbed PVB coating has very high delamination veloc-

53


0 2000 4000 6000 8000 10000 12000

-0.90

-0.75

-0.60

-0.45

-0.30

E/

Vv

s.

SH

E

Distance / µm


scans with interval of 8 min.

after 48 min.

0 500 1000 1500 2000

-0.60

-0.45

-0.30

-0.15

0.00

E/

Vv

s.

SH

E

Distance / µm


scans with interval of 1.5 h

after 24 h

(a) (b)

Figure 5.4: SKP delamination profile of (a) covalently bound PS on zinc and (b)weakly bound PVB used for reference purposes, both in air.

ity (r = 11.0 ± 0.05 mm h−1) as compared to chemisorbed linear PS coating(r = 0.05±0.005 mm h−1) due to weak interaction forces with the substrate mate-rial. However, there is not much difference in the delamination rates for cross-linkedPS coatings. The possible explanation for this could be attributed to the additionof aliphatic cross-linking molecules. Addition of aliphatic cross-linking molecules ef-fects the close packing of aromatic rings and increases the hydrophilicity of polymerchains due to presence of hydrophillic groups in a cross-linking molecule. Hence,there is not a significant decrease in the delamination rates for cross-linked PS coat-ings. This fact could be further verified by comparing the delamination rates ofPS coatings with different molar ratios of EDA and HDA. It could be seen in Ta-ble 5.3 that the addition of higher amounts of cross-linker leads to an increase inthe delamination rates of PS coatings. Moreover, cross-linker with longer aliphaticchain also increases the delamination rate. This affect can be again explained bythe morphology and structure of the polymer chains. Higher amounts of aliphaticcross-linker disturbs the close packing of the aromatic chains and hence, more freevolume is present in the polymer chain structure. Similarly, the cross-linker withlonger aliphatic chain (i.e. HDA) destroys the close polymer chain packing due tohigher chain mobility. The delamination rates for linear and cross-linked PS coatingsare summarized in Table 5.3.

54

5.3. Summary

0 220 440 660 880 1100 1320

0

400

800

1200

1600

2000

2400

De

lam

ina

tio

nfr

on

td

/µ

m

Time t / min

VTS

no CL

25% EDA

50% EDA

25% HDA

50% HDA

PVB

10 100 1000

10

100

1000

10000

d ~ t1/2

d ~ t

VTS

no CL

25% EDA

50% EDA

25% HDA

50% HDA

PVB

De

lam

ina

tio

nfr

on

td

/µ

m

Time t / min

(a) (b)

Figure 5.5: Comparison of the delamination rates for polymer coatings with differentcross-linker amounts (a) on a linear scale and (b) on a double logarithmicscale. In the plot (b), two lines show the expected slopes for d ∼ t andd ∼√t, as indicated in the graph.

Table 5.3: Comparison of delamination rates of linear and cross-linked PS coatings


PSno cross-linker 0.05± 0.00525% EDA 0.024± 0.0250% EDA 0.028± 0.00525% HDA 0.036± 0.00550% HDA 0.039± 0.003

5.3 SummaryThis chapter shows the investigation on the cathodic delamination kinetics of lin-ear and cross-linked PS coatings that are covalently linked to the base metal viaorganosilanes. These ultrathin coatings were synthesized on top of modified zincsubstrate via thermally initiated free radical polymerization of styrene. The resultsshows that the chemisorbed PS coatings exhibits slower delamination rates thanphysisorbed PVB coating. Moreover, addition of aliphatic cross-linking agents doesnot leads to a significant reduction in the cathodic delamination rates. In particular,PS coatings cross-linked with EDA exhibits slower delamination rates than the coat-ings cross-linked with HDA. The addition of the cross-linker with longer aliphaticchain (HDA) effects the close packing of aromatic groups that are present in the PSpolymer and hence, leads to lower resistance against cathodic delamination.

55

Chapter 6: Delamination kinetics of modelPMMA coatings bound to zincvia ester linkage

6.1 IntroductionPolymer coatings with specific functionalities are often used in industrial polymercoatings because of their desirable structural and interfacial performances. In par-ticular, carboxylic functionalities has gained much attention due to their impor-tance in understanding the fundamental adhesion mechanisms at polymer/metalinterfaces94,167. The bonding properties at polymer/metal interface is affected byseveral factors. The most crucial factor is the end functional group that leads toa covalent bonding between polymer and metal. Moreover, the metal substrate isoften pretreated before application of a polymer coating in order to enhance theadhesion strength and improve the corrosion performance168. It has been shownthat carboxylic functionality can attach to the zinc oxide surface via coordinativebonding167,169.In the chapter 4 and 5, it has been demonstrated that covalently adhered PMMA

and PS coatings linked to zinc substrate via silane bond leads to enhanced protectionof polymer coatings against cathodic delamination. Moreover, it was also shownthat the morphology and hydrophilicity of polymer coating also plays a role indefining the deadhesion properties of a coatings due to different barrier propertiesagainst aggressive species like oxygen and water. Previous studies have shown thatthe adhesion strength between metal and polymer is determined by the functionalgroups that are present in polymer which makes covalent bonding with the surfacehydroxyl groups of zinc95,115. Therefore in this chapter, the cathodic delaminationkinetics of linear and cross-linked PMMA coatings that are attached to the metallicsubstrate via ester linkage is investigated. In order to obtain polymer coating boundto zinc via ester linkage, the zinc substrate was functionalized with acryloyl chlorideinstead of VTS.

57

Chapter 6. PMMA model coatings bound via ester linkage


6.2.1 Surface modificationInitially, zinc substrate was treated with acryloyl chloride that leads to ester linkageat metal/orgnic interface. As shown in Figure 6.1, the functionalization reactionwas carried out at room temperature in THF and in the presence of triethylamine.Triethylamine acts as a buffer and reacts with the hydrochloric acid (HCl) producedduring the reaction forming triethylammoniumchloride which precipitates and is notactive in the reaction anymore. As this is an equilibrium reaction, removal of HClis necessary for the completion of the reaction otherwise back reaction will occur,which results in the cleavage of ester bond by the HCl produced during the re-action. Successful functionalization could be proved by IR, XPS and TOF-SIMSexperiments. Figure 6.2a shows the IR spectrum of acrlyol chloride modified zincsubstrate. The bands at 2924 cm−1 and 2865 cm−1 are assigned to anti symetricand symmetric stretching modes of CH2, respectively150. Moreover, the presence

n = 2 for EDA

n = 6 for HDA

+

TEA

+

THF


O

O

Cl

O O

O

On

O O

O

On

O

O

O

O

O

O O

O

O

n


Figure 6.1: Scheme of the surface modification steps carried out in this chapter.

58


of band centered at 1725 cm−1 confirms the presence of C=O group. The forma-tion of covalent bonding between zinc and acryloyl chloride can be confirmed bythe appearance of the band at arround 1600 cm−1 and 1440 cm−1 that is assignedto the antisymmetric and symmetric stretching vibrational modes of CO−

2167,170.

Presence of both symmetric and antisymmetric stretching modes indicates that car-boxylate bonds are inclined at a certain angle with respect to the surface. Figure

3000 2500 2000 1500 1000

-0.001

0.000

0.001

0.002

1600

995-985

(C-H)

1230-1050

(C-O)

1450

(C-H)

(COO-)

1725

(C=O)

2925-2850

(C-H)

Ab

so

rba

nc

e

Wavenumber / cm-1

(a)

292 290 288 286 284 282

1000

1500

2000

2500

3000

3500

4000

C-O

COO-

Binding energy / eV

C-C/C-H(b)

Inte

ns

ity

a.u

./

Figure 6.2: (a) IR spectrum of acryloyl chloride modified zinc. (b) C 1s region ofthe XP spectrum of acryloyl chloride modified zinc.

6.2b shows the high resolution C 1s XP spectra. The peak is decomposed intoC-C/C-H (284.8eV), C-O (286.4 eV) and CO−

2 (289.1 eV) sub peaks. To furtherverify the successful modification of zinc surface with acryloyl chloride, TOF-SIMSmeasurements were performed. Figure 6.3 depicts the positive TOF-SIMS spectrameasured at the interfacial region of modified zinc substrate with assignment of main

59


93 94 95 96 97 98 99

0

500

1000

1500

2000

70ZnOC

+

70ZnO

13C

+

70ZnOCH

+

68ZnO

13CH

+

68ZnO

13C

+

68ZnOCH

+

67ZnO

13C

+

68ZnOC

+

67ZnOC

+

66ZnOCH

+

ZnO13

C+ 66

ZnO13

C+

66ZnO

13CH

+

67ZnOCH

+

66ZnOCH

+

ZnO13

CH+

ZnOCH+

Co

un

ts

m z-1

/ amu

Figure 6.3: SIMS showing the isotope distribution of ZnOCH+ fragments from thesurface of acryloyl chloride modified zinc. Expected isotope distributionis shown as vertical lines (|) ca. 0.3 amu below the experimental peak.

ions. The presence of the fragment ZnOCH+ is an indicative of presence of esterlinkage between polymer and a metal substrate. Moreover, the ion pattern of thefragment ZnOCH+ should reflect the isotopic abundance of the elements present;the observed distribution is in close agreement to the calculated distribution.The growth of linear and cross-linked polymer coatings on functionalized zinc

surface that is covalently bonded to the anchoring molecule was achieved by usingsimilar synthetic route explained in chapter 4. To verify the successful growth ofpolymer on functionalized zinc substrate, the polymerized samples were character-ized with IR and XPS. Figure 6.4 I demonstrates the IR spectra measured on assynthesized PMMA coatings. The presence of peak at around 1725 cm−1 is assignedto C=O150 and that confirms the presence of PMMA. Moreover, the increase inthe intensity of C-H stretching bands at around 2960-2850 cm−1 also depicts thepresence of long aliphatic polymer chains. The detailed assignment of IR peaks aresummarized in Table 6.1.Moreover, XP spectra (Figure 6.4 II) of polymer-modified samples show the ex-

pected C 1s and O 1s peaks. The high resolution C 1s spectrum can be decomposedin to four distinct peaks at 284.8 eV, 285.4 eV, 286.4 eV and 288.8 eV that originatefrom aliphatic carbon, C–COO, C–O, and COO, respectively.

60


3000 2500 2000 1500 1000

Wavenumber /cm-1

Ab

so

rpti

on

1270-1000

(C-O)

1760-1735

(C-H)

2990-2850

(C=O)

(a)

(b)

(c)

(d)

(e)

I

294 292 290 288 286 284 282

Binding energy / eV

C-C/C-HC-COO

C-O

O=C-O

(a)

(b)

(c)

(d)

(e)

Inte

ns

ity

/ a

.u.

II

Figure 6.4: IR and C 1s XP spectra of different covalently-bound polymer coatingson zinc, (a) PMMA without cross-linker, (b) PMMA + 25% EDA, (c)PMMA + 50% EDA, (d) PMMA + 25% HDA, (e) PMMA + 50% HDA.Dotted red lines are indicating the area of interest in both IR and XPspectra.

Ellipsometry was used as an analytical tool to determine the thickness of as syn-thesized linear and cross-linked PMMA coatings. The obtained polymer thicknessvalues are summarized in Table 6.2.

6.2.2 Model corrosion experimentsModel corrosion experiments were performed as explained in chapter 4. Figure 6.5demonstrates the SKP potential profiles measured over zinc sample that is modifiedwith PMMA (having a ester linkage with the substrate) and coated with PVB alongwith the physisorbed PVB coated zinc.It is shown in Figure 6.6a that the progress of the delamination front is fastest in

the case of spin coated PVB, while the progress of the delamination front is inhibitedin case of chemically modified zinc samples. The increased delamination velocity of

61


Table 6.1: FTIR stretching vibration position (cm−1) of AC and different PMMAcoatings with and without cross-linker synthesized on zinc surface basedon refrence150

PeakAssignment AC PMMA

PMMA+

25% EDA

PMMA+

50% EDA

PMMA+

25% HDA

PMMA+

50% HDACH3 [as] - 2993 2992 2987 2984 2993CH3 [ss] - 2956 2953 2948 2954 2948CH2 [as] 2924 2917 2922 2923 2922 2921CH2 [ss] 2865 2848 2848 2854 2851 2850CH2[bend] 1455 1440 1445 1460 1446 1453C=O [s] 1730 1733 1738 1737 1730 1736C=C [s] 1635 - - - - -COO− [as] 1590 1587 - 1578 1600 1581C-O [stretching] 1150 1149 1153 1153 1160 1166*s = stretching, ss = symmetric stretching, as = antisymmetric stretching

Table 6.2: Polymer film thickness obtained from ellipsometric measurements for lin-ear PMMA samples along with coatings having different cross-linker ra-tios bounded to zinc via ester linkage.

Sample Thickness / nmPMMA without cross-linker 25± 625% EDA 14± 350% EDA 13± 325% HDA 22± 350% HDA 18± 6

PVB coated zinc could be explained by the weak van der Waals interaction forcesat polymer/metal interface.In order to gain further insight into the effect of network chain length and degree

of cross-linking on the stability of these coatings against cathodic delamination,additional coatings were synthesized on zinc substrates with different cross-linkerchain length and degree of cross-linking. Following a similar experimental proce-dure, these cross-linked polymer coatings were exposed to delamination experimentsand progress of cathodic delamination was tracked by SKP. Figure 6.6 shows theposition of delamination front with respect to time for coatings cross-linked withEDA and HDA together with different degree of cross-linking. When comparing thedelamination rates of EDA and HDA cross-linked PMMA, it can be clearly seen thatrate of cathodic delamination is fast on a coatings cross-linked with EDA comparedto HDA cross-linked polymer coatings. This could be explained by the fact that

62


0 1500 3000 4500 6000 7500-0.8

-0.6

-0.4

-0.2

0.0

E/

Vv

s.

SH

E

Distance / µm



after 3.5 hr

(a)

0 2000 4000 6000 8000 10000

-0.90

-0.75

-0.60

-0.45

-0.30

E/

Vv

s.

SH

E

Distance / µm



after 48 min.

(b)

Figure 6.5: SKP delamination profile of (a) covalently bound PMMA on zinc and(b) weakly bound PVB used for reference purposes, both in air

Table 6.3: Delamination rates and standard deviations for the different samples,based on linear fits of plots as in Figure 6.6. All samples were coveredwith PVB, as explained in the experimental section.


pure PVB 11.2± 0.5no cross-linker 1.32± 0.125% EDA 0.67± 0.0450% EDA 0.55± 0.0425% HDA 0.44± 0.0650% HDA 0.25± 0.03

EDA has a shorter chain length, which leads to a polymer network structure wherechains mobility is hindered and has increased voids/pore in a coating. Moreover,due to shorter aliphatic chain length, it leads to coatings with lower hydrophobicity.Hence, the EDA cross-linked coatings exhibits lower resistance against diffusion ofwater/oxygen and this eventually increases the oxidative degradation of the poly-mer/metal interface. Moreover, increasing the cross-linking density by increasing theamount of EDA from 25 mol% to 50 mol% results in deceleration in the degradationrate of the interface as a result of polymer network with high degree of cross-linkingbetween the polymer chains. Thus, there is less free volume available for oxygenand water molecules to diffuse through the network structure, which further reducesthe rate of oxygen/reduction reaction in the humid air atmosphere.On the other hand, HDA has a longer aliphatic chain length which makes the re-

sulting polymer structure more hydrophobic, dense and flexible. Consequently diffu-sion of water through the polymer network structure is inhibited and thus, progressof delamination front is slowed down. Similarly, increase in the concentration of

63


0 100 200 300 400 500 600

0

2000

4000

6000

8000

De

lam

ina

tio

nfr

on

td

/µ

m

Time t / min

PVB

no CL

25% EDA

50% EDA

25% HDA

50% HDA

10 100

100

1000

10000

d ~ t

d ~ t1/2

De

lam

ina

tio

nfr

on

td

/µ

m

Time t / min

PVB

no CL

25% EDA

50% EDA

25% HDA

50% HDA

(a) (b)

Figure 6.6: Comparison of the delamination rates for polymer coatings with differentcross-linker amounts (a) on a linear scale and (b) on a double logarithmicscale. In the plot (b), two lines show the expected slopes for d ∼ t andd ∼√t, as indicated in the graph.

HDA from 25 mol% to 50 mol% results in further deceleration in the degradationrate of the interface as a result of increased hydrophobicity of the polymer.

6.3 SummaryLinear ultrathin PMMA coatings that are covalently bonded to zinc/oxide surfacevia ester linkage were synthesized by thermally initiated copolymerization reac-tion. The results yields that covalently grafted PMMA coating lead to significantinhibition in cathodic delamination rate due to enhanced adhesion forces at poly-mer/metal interface. Furthermore, addition of the cross-linker leads to the decel-eration in the delamination rate as a result of denser polymer network structureand reduced number of defects/pores in the coating. However, delamination ratesof EDA cross-linked polymer coatings are higher in comparison to HDA cross-linkedsamples. Higher concentration of EDA results in the network structure with higherhydrophilicity (compared to HDA), which leads to higher water uptake through thepolymer coating. On contrary, cross-linker with longer chain length (HDA) showsenhanced resistance against cathodic delamination process due to higher hydropho-bicity of polymer coating.

64

Chapter 7: Comprehensive discussion on ca-thodic delamination of PMMA &PS with different surface linkage

The presence of a polymer coating on metallic zinc leads to reduction in the corrosionrates, by inhibiting interfacial electron transfer reactions through blocking diffusionpaths and active sites. The nature of bonding between polymer and metal alsoplays a role in defining the final rates of cathodic delamination. In particular, poly-mer coatings bound to metallic surface via covalent bonding leads to considerablereduction in the corrosion rates in comparison to coatings that are physisorbed tothe substrate material. It has been shown that modification of oxide-covered metalsurfaces with covalently bound polymers inhibits the oxygen reduction reaction atthe polymer/metal interface as a result the strong chemical bond157–159.In this work, PMMA and PS coatings with different cross-linking densities that

are bound to zinc substrate via silane and ester linkage has been synthesized via freeradical polymerization and their cathodic delamination behavior is monitored withSKP. The obtained delamination rates are summarized in Table 7.1. It can be seenthat the delamination rate is highest for the reference sample that was only coatedwith PVB. This fast delamination rate of PVB is a result of weak van der Waalsinteraction with the substrate material. Now, comparing the delamination rates ofchemisorbed polymers, PS coatings that are bound to zinc via silane linkage has thelowest, while the ester bonded PMMA coatings has the highest delamination rates.In order to explain the differences in the delamination rates between these chemi-

sorbed coatings, the factors which could play a role in the cathodic delaminationprocess have to be considered. These factors are the nature of chemical bonding atpolymer/metal interface, overall amount of hydrophobic groups in a coating, andmorphology of the polymer itself. Considering the delamination rates of PMMApolymer bonded to zinc via silane and ester linkage, there is a quite significantdifference in the delamination rates for both the coating systems. In particular,the linear PMMA coating bonded to zinc via ester linkage delaminate faster by thefactor of almost 10 then PMMA bound to zinc via silane linkage. The possibleexplanation for the lower resistance against cathodic delamination for ester boundPMMA coating is the nature of ester bond itself. Ester linkage is quite susceptibleto alkaline hydrolysis and as cathodic delamination process leads to high alkalinepH at the metal/polymer interface, this leads to rapid breakage of metal/polymer

65

Chapter 7. Comprehensive discussion on cathodic delamination

chemical bond. The alkaline hydrolysis of ester linkage is shown in the equationbelow,

R(1)−COO−R(2) + OH− −−→ R(1)−COO− + R(2)OH. (7.1)

0 300 600 900 1200 1500

0

1000

2000

3000

4000

Time t / min

PMMA with ester linkage

PMMA with silane linkage

PS with silane linkage

De

lam

ina

tio

n f

ron

t/

µm

d

Figure 7.1: Comparison of delamination rates of chemisorbed linear PS and PMMApolymer coatings.

Moreover, cross-linked PMMA coatings bound to zinc via ester and silane linkageleads to further reduction in the cathodic delamination rates. It is obvious thatthe this larger resistance against the delamination rate is not due to the nature ofchemical bond, as the bond with the metallic substrate remains the same for bothlinear and cross-linked polymers. As can be seen from Table 7.1, the cross-linkingof PMMA coatings leads to reduction in the rate of cathodic delamination for bothsilane and ester bonding. This could be explained by the fact that cross-linkingof the coating leads to denser polymer structures, with increased number of esterlinkage between the polymer chains. Due to significantly increased number of cross-linking between the polymer chains, the degradation of polymer via ester hydrolysistakes more time. Consequently, the overall degradation process of a polymer coatingis slowed down.Now, if we look at the delamination rates of linear PMMA and PS coatings bound

to zinc via silane linkage, PS coatings show enhanced resistance against cathodicdelamination. This inhibition in the delamination velocity could be attributed tohigher hydrophobicity and dense packing of polymer chains due to presence of aro-matic rings in the PS polymer. It has been shown that dense packing of polymerchains exhibits superior barrier properties against aggressive species such as oxy-

66

Time t / min

0 250 500 750 1000 1250

0

450

900

1350

1800

2250

Silane bound PS

Silane bound PMMA

Ester bound PMMA

De

lam

ina

tio

n f

ron

t/

µm

d

Figure 7.2: Comparison of delamination rates of chemisorbed 50% EDA cross-linkedPS and PMMA polymer coatings.

gen and water which are the primary cause of initiation of corrosion reaction117.While on the other hand, PMMA polymer is less hydrophobic due to presence ofhydrophillic groups. Hydrophobic polymer layers cannot take up as much wateras less hydrophobic coatings. This lower availability of the solvent water duringthe corrosion process leads to reduction in the corrosion rate and the delaminationrate. Moreover, when comparing the delamination rates of cross-linked PMMA andPS coatings, there is not a significant difference. This can be again explained bythe morphology and hydrophobicity of polymer chains. Although, the addition ofaliphatic cross-linking agents to the PS chains leads to increased cross-linking be-tween polymer chains, however, they effect the close packing of aromatic rings thatare present in the polymer chain and increases the hydrophilicity of the coating dueto presence of hydrophillic groups in the cross-linking molecule. Hence, there isno significant reduction in the cathodic delamination rate of cross-linked PS coat-ings. On the other hand, cross-linking of PMMA coatings leads to denser polymernetwork that has enhanced barrier properties against diffusion of aggressive species(independent of surface linkage) like oxygen and water leading to higher resistanceagainst cathodic delamination.The coatings prepared here via a wet chemical process lead to similar delam-

ination rates, corrosion current densities and charge transfer resistance values asplasma polymers prepared by a gas phase process. Plasma polymers, on the otherhand, are ultrathin polymer films which are typically highly cross-linked. Chargetransfer/polarization resistance are in similar order of magnitude comparing with

67

Chapter 7. Comprehensive discussion on cathodic delamination

Table 7.1: Delamination rates and standard deviations for the different samples. Allsamples were covered with PVB, as explained in the experimental section.Sumarizes Tables 5.3, 4.2, 6.3


PS(silane linkage)

PMMA(silane linkage)

PMMA(ester linkage)

no cross-linker 0.05± 0.005 0.27± 0.02 1.32± 0.1225% EDA 0.048± 0.02 0.22± 0.02 0.67± 0.0450% EDA 0.023± 0.005 0.042± 0.004 0.55± 0.0425% HDA 0.036± 0.005 0.039± 0.003 0.44± 0.0650% HDA 0.039± 0.003 0.028± 0.004 0.25± 0.03

literature data. Also, slowest delamination rates found here are of same order com-pared to published results for plasma polymers on steel101. Hence, the coatingsprepared here via a wet chemical process are comparable to plasma polymers pre-pared by a gas phase process.Delamination rates are reaction-controlled for the PMMA coating systems, inves-

tigated here, as shown by the t1 dependence. In particular, diffusion limitation canbe excluded, as this would lead to a t1/2 dependence. Because delamination ratesdepend on the presence of cross-linkers and the degree of cross-linking. Therefore,the rate-limiting reaction must be related to the polymer of the coating. This obser-vation rules out alkaline zinc oxide dissolution and siloxane/ester bond hydrolysisas sole reactions leading to de-adhesion. Radical reactions as side reactions of theoxygen reduction can explain the observed trends in delamination rate with varia-tion of interface composition. On the other hand, delamination rates are partiallydiffusion limited on larger time scales for silane bound PS coatings. This diffu-sion limited dependence is attributed to the higher hydrophobicity of PS coatingthat shows enhanced resistance against permeation of water and oxygen. Hence,over time the cathodic delamination process for silane bound PS coatings becomesdiffusion limited. In summary, delamination mechanisms discussed above is not cen-tered around radical reactions, needs to involve an unlikely coupling by an unknownmechanism between alkaline dissolution of inorganic surface linkages and polymerdegradation. The situation encountered in this work is significantly different fromthe situation in a previous work, were a larger cross-linker amount was found to leadto higher delamination rates in latex polymers because of the increase in interfacialfree volume50. Interfacial free volume can form in significant amount in coatingswhich form by coalescence of latex particles, but is less likely to occur in significantamount in coatings prepared directly by free radical polymerization, as used here.Especially, if interfacial free volume forms, its relation to the cross-linking is notstraightforward.

68

Chapter 8: Study of polymer coating delam-ination kinetics on zinc modifiedwith zinc oxide of different mor-phologies∗

8.1 IntroductionIndustrially prepared material surfaces often possess complex morphologies and to-pographies. Moreover, Zn is always covered with a thin oxide layer, which deter-mines its adhesion to the polymer167,171,172, which is related to its resistance againstcorrosion. Previous studies have shown that interfacial oxide chemistry plays an im-portant role in adhesion between polymer/metal interfaces. In particular, tailoringthe hydroxyl/oxide ratio could lead to the enhancement in the adhesion at poly-mer/oxide/metal interface169. The native oxide differs in its electronic structuresubstantially from the crystalline oxide, especially due to the presence of intra-gap states, which may affect adhesion properties173,174. Naturally, the thickness ofthe oxide films depends on the electrochemical polarization and atmospheric con-ditions175,176. The dissolution kinetics of ZnO under the conditions of cathodicdelamination may, however, also be a crucial factor when determining the rate ofdelamination.ZnO itself has become an interesting functional materials in a number of areas,

e.g. for application in LEDs177, solar cells178, catalysis179, and nano-sensing180.For most of these applications, micro- or nanostructured surfaces, or small particlesare used, where "small" refers to size parameters below 10 µm. The properties ofmicro- and nanostructured ZnO can be affected via control of size and shape181,182.Therefore, a variety of different crystal shapes has been synthesized, ranging fromrods/tubes86, belts183, spheres184, to more complex ones like urchins185, flowers186,twin-plates187 and pine trees89.In corrosion protection, the formation of a layer of protective corrosion products

on a surface is crucial120,121,188. By deposition of artificial corrosion products on the

∗This chapter contains results published as a scientific article:Iqbal, D. et al. Study of polymer coating delamination kinetics on zinc modified with zinc oxideof different morphologies. Mater. Corros.2014, 65, 370-375.

69

Chapter 8. Delamination kinetics on ZnO

surface of steel and galvanised steel, the effect of oxides on the oxygen reductionwas recently investigated188. This chapter follows a similar idea: ZnO particles withdifferent shapes are deposited on Zn surfaces, and delamination of a model polymercoating is investigated. ZnO crystals of different sizes and shapes (determiningoverall surface morphology) exhibit different stability towards wet etching in a basicmedium, as occurs during cathodic delamination. ZnO dissolution in an alkalinemedium has shown to take place at crystal edges189. Consequently, the rate ofdelamination of a polymer coating should be affected by the oxide surface on topof the zinc coating. A test of this hypothesis is subject of this chapter. Therefore,spherical, rod-shaped and twin-plate-shaped ZnO particles have been synthesized viaa sol gel route, and deposited on a metallic Zn substrate. Subsequently, poly(vinylbutyral) (PVB) coatings, as a model for weakly bound polymers on the zinc, havebeen deposited, and the delamination kinetics was studied in a humid air atmosphereby scanning Kelvin probe (SKP) measurements.

8.2 ResultsDifferent microstructures with a certain distribution in size and diameter of theZnO particles were grown on polished Zn sheet using the hydrothermal techniquedescribed above. Figure 8.1 shows the SEM images of twin-plate-shaped (a), rod-shaped (b) and spherical (c) ZnO microstructures precipitated on the Zn sheetsubstrate. Rod-like structures were synthesized without any surface capping agent,due to the tendency of c-axis growth of ZnO91. To obtain other morphologies of ZnO,gelatin and sodium citrate were used for selective protection of one surface. Thedetailed growth mechanism of twin-plate-shaped ZnO has been discussed in187,190,while the respective mechanism for growth of spherical particles is described in184,191.By varying the growth time and concentration of the surface protecting agents, thesize and morphology of the ZnO microstructure can easily be varied.XRD data, also shown in Figure 8.1, confirms the crystalline nature of the pro-

duced ZnO particle-covered surfaces. Comparing the relative intensity ratios of thedifferent ZnO peaks to powder diffraction file No. 00-005-0664 for ZnO shows slightdeviations. The most prominent peak is in all cases the 101 peak, though its in-tensity compared to other peaks in all patterns in this work is higher than in thereference data. On the other hand, the second most intense 100 peak from the refer-ence data is less prominent in the data obtained here. While the patterns for spheresand twin-plates are quite similar, the rod’s pattern shows only a small intensity ofthe 100 peak, but a substantial 002 peak. There is clearly some preferential growthdirection, the overall deviations from the powder data does not indicate a strongdominance of one direction, as would be expected for c-axis growth on surfaces91.While there is some evidence for c-axis growth in the rod-shaped particles, in allcases the patterns are dominated by the 101 peak, belonging to one of the stablenon-polar surfaces of ZnO91,192. Figure 8.2 shows averaged SKP potential scans ofthe delamination experiments of PVB coatings on Zn and Zn modified with ZnO

70

8.2. Results

Figure 8.1: SEM micrographs and XRD patterns of grown ZnO structures on metal-lic zinc. Left panel: SEM images of (a) ZnO twin plates, (b) ZnO rods,

of different morphologies. Initially, SKP measures a potential of the intact poly-mer/oxide/metal interface. However, with time, cathodic delamination is initiatedin a humid air atmosphere on all the samples, as can be seen in Figure 8.2. Theprogress of the delamination front with time is plotted in 8.3 for three represen-tative examples from each surface morphology, and the slope of a linear fit wasdetermined. This slope is discussed as delamination velocity below. It is evidentfrom Figure 8.3 that progress of the delamination front is by far fastest for PVB onZn without particles, where a delamination rate of ≈ 11 mm h−1 is observed. Thestatistical error for all delamination rates reported in this work, including those onthe surface with spherical, twin-plate-shaped and rod-like particle, is 0.05 mm/h,which was obtained as the standard deviation from three samples treated accordingto the same scheme. The actual uncertainty is likely to be higher, due to systematicerrors and uncertainties in readout and sample preparation. While on the modifiedsurfaces, thick layers of ZnO microstructures are present, the untreated Zn possessesonly its native oxide layer. The observed strong reduction in delamination rate may

71


0 1000 2000 3000 4000 5000-0.8

-0.6

-0.4

-0.2

0.0

0.2

Distance / µm

E/

Vv

s.

SH

E

0 min

scans with 12min interval

after 147 min

0 2000 4000 6000 8000 10000 12000

-0.90

-0.75

-0.60

-0.45

-0.30

E/

Vv

s.

SH

E

Distance / µm



after 48 min.

(a)

0 1000 2000 3000 4000 5000

-0.8

-0.6

-0.4

-0.2

0.0

0.2

E/

Vv

s.

SH

E

Distance / µm

0 min


after 65 min

0 1000 2000 3000 4000 5000

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

Distance / µm

E/

Vv

s.

SH

E

0 min


after 560 min

(b)

(d)(c)

Figure 8.2: SKP delamination profiles in humid air on PVB coated zinc (a) withoutany artificial oxide, (b) with ZnO twin plates, (c) with ZnO rods, and(d) with ZnO spheres.

therefore have two origins. First, it may be related to an inhibition of the oxygenreduction reaction (ORR), as was observed after deposition of simonkolleite corro-sion products on Zn188. Second, the deposited thick layer may protect longer Znagainst dissolution, as the ZnO layer needs more time to dissolve under the alkalineconditions present after the delamination front compared to the thin native oxide.It is also important to note that during delamination process, reprecipitation, i.e.formation of zinc hydroxide from solution, takes place, which does not passivate thesurface due to porous nature of the film formed193,194.Now, focusing on the delamination rates of the different ZnO-modified samples,

namely containing ZnO spheres, rods, or twin-plates (Figure 8.3), a clear differencebetween the different surface morphologies is observed. Delamination velocity ofcathodic delamination is slowest in the case of deposited ZnO spheres, where it is≈ 0.2 mm h−1, while it is fastest on ZnO twin-plate samples with ≈ 2 mm h−1.Rod-shaped structures show a rate of ≈ 1 mm h−1, i.e. in between spheres and twinplates.For all three different morphologies, a reduction in the ORR activity is expected.

72

8.2. Results

0 50 100 150 2000

1100

2200

3300

4400

5500

Del

amin

atio

n fr

ont

/ µm

Time / min

PVB coated zinc twin-plate nanorod sphere

Figure 8.3: Progress of the delamination front with time on differently structuredsurfaces. The slope of the linear fit to the points is the delaminationrate.

Previous studies have shown that ZnO dissolution in an alkaline medium (as presentunder the conditions here in the passing of the delamination front) takes place atcrystal step edges189,195. Macroscopically, a sphere has a smooth surface, withoutsharp edges and corners, which may explain the particularly low delamination rateson surfaces covered with spherical particles. Sharp corners are present in the case ofrod-like and twin-plate structures. From the crystal structure of ZnO, one expects,after typical c-axis growth, the top and bottom of the cylinder to be terminatedwith a polar (0001) surface, while the sides should be terminated with the lowerenergy non-polar [e.g. (1010) or (1011)] surface91,192. Consequently, twin plates areexpected to have the highest fraction of polar surfaces, while this fraction is reducedin rod-shaped crystals. This difference is reflected in the faster dissolution of thetwin-plate covered Zn, compared to rod-covered Zn. The actual surface termina-tion of the spherical particles is not intuitively obvious. Attempts to conclude theorientation of the surfaces from the XRD data shown in 8.1 may be misleading, asthis shows the predominant crystal growth direction perpendicular to the surface,which is not necessarily the decisive growth direction. As the surface of a sphericalparticle is, however, homogenous, one should expect a non-polar surface with defectsto adjust in the curvature. These defects are, however, obviously not those needed

73


to facilitate dissolution. Kinks and edges on the other hand must be present on thetwin plates and rod, and are obvious sites for start of oxide dissolution under basicpH. Therefore, the cathodic delamination process is accelerated on ZnO rods andtwin-plates in comparison to spheres.

3 µm

(b)

(a)

3 µm

Figure 8.4: SEM micrographs after the delamination of PVB from (a) ZnO twin-plate-covered surface and (b) ZnO rod-covered surface.

A second factor which needs to be considered is the availability of surface regionswhich are not covered by µm thick oxides. In these regions, oxygen has easy accessto the metal, and the rate of the ORR is therefore faster. For a detailed analysisof the importance of this effect, however, a detailed study of the interface betweenoxide and metal is needed, which is rather difficult to perform. An inspection of thesurfaces after delamination also points to the importance of the oxide dissolution.SEM images after the cathodic delamination experiment are shown in Figure 8.4.All the twin-plate structures have been dissolved. Rod-like structures have alsobeen destroyed, however, the final morphologies resemble rods. Comparison to theinitial morphologies (see Figure 8.1) points to the importance of oxide dissolutionin determining the rate of cathodic delamination.Two more potential factors of influence on the delamination rate, though through

evidence here considered as not decisive, shall be discussed, (a) the adsorption ofthe capping agent and (b) the effect of the overall amount of ZnO in the interface,i.e. the layer thickness. Organic capping agents have been used for the synthesis of

74

8.3. Summary

spherical and twin-plate-shaped structures, while rod-shaped structures have beenprepared without capping agents. As the rate increases in the sequence sphere <rod < twin-plate, i.e. the delamination rate for the capping agent free system is inbetween those with capping agents, there is indirect evidence that the adsorption ofresidual capping agent in the delamination experiments is not decisively determiningthe delamination rate. If, on the other hand, the dissolution of ZnO is decisive,the total amount of ZnO in the interfacial region may contribute to the observeddifferences in delamination kinetics. On the basis of XRD data shown in Figure8.1, it can be concluded that the overall layer thickness of ZnO is lowest in case ofthe rod-shaped structures. In the case of these structures, the ration between thestrongest diffraction peaks from the substrate zinc to the strongest diffraction peakof ZnO is highest, indicating a lower layer thickness than obtained for twin-plate-shaped and spherical particles. On the basis of the sequence of delamination ratesobserved here it can be again inferred that the total layer thickness cannot be theonly determining factor, but that indeed the morphology plays a central role.

8.3 SummaryThe deposition of ZnO on Zn substantially decreases the delamination rate of poly-mer coatings on top. The morphology of ZnO plays a crucial role in the determina-tion of the delamination kinetics, with a factor of 10 between the lowest (sphericalparticle) and highest (twin-plate-shaped particle) delamination rates. A compari-son of the surface morphologies before and after the delamination experiments showsthat dissolution of the oxide in the interfacial region occurs. The kinetics of delam-ination is likely determined by the nature of the dominating surface terminationof ZnO. These results suggest that a careful control over surface morphologies ina pretreatment-like process offers the chance to significantly increase resistance ofpolymer-coated galvanized steel towards cathodic delamination.

75

Chapter 9: Sequential growth of zinc oxidenanorod arrays via a corrosion pro-cess∗

9.1 IntroductionAmongst the big challenges in semiconductor photocatalysis is to tune the bandgap,defect levels and morphology of the catalyst for efficient use of a large part of thevisible spectrum, in particular the red fraction of light, to catalyze chemical reac-tions196–198. TiO2 has been so far the most popular material of study as photocat-alysts, e.g. in different nanostructures199–201. A particular important application ofsemiconductor photocatalysis is the decomposition of organic substances in water viaaerobic photooxidation202,203, a potential route to remove persistent pollutants204.Using UV light, large conversions in decomposition of an organic dye model pol-lutant have been achieved recently202. For decomposition of organic substances, anumber of different materials are currently investigated, including low-cost, non-toxic ZnO61,205. Its band gap at UV photon energies canbe modified by crystalstrain62. While application with UV light is straight forward, more current focusis towards extending the operational range towards visible light. In this regard,most approaches extend the ZnO light absorption from the UV into the blue part ofthe visible61,206,207. Alternatively, metal nanoparticles can be conjugated to ZnO208.Supported catalysts with large surface areas can conveniently be realized as ZnOnanorod arrays209. Consequently, developed synthesis methods include chemicalvapour deposition210, laser decomposition of a suitable precursor211, vapour-liquid-solid growth212, pulsed laser deposition213,214, a flame transport method215, templatebased methods216, solvothermal techniques217, and solution phase approaches218,219.Synthesis can be interfaces with lithographic techniques for regular patterning220.All aforementioned preparation techniques either require the use of sophisticatedinstrumentation, or elevated temperature, with nanorods growing in parallel overthe sample area. Besides photocatalysis, other applications of ZnO nanorod arrays

∗This chapter contains results published as a scientific article:1.Iqbal, D. et al. Sequential growth of zinc oxide nanorod arrays at room temperature via acorrosion process: application in visible light photocatalysis. ACS Appl. Mater. Interfaces(2014). doi:10.1021/am504299v

77

Chapter 9. Growth of ZnO nanorods

include gas sensing217,221, refractive index sensing60, water splitting222, antibacterialcoatings223, and solar cells58,59. Corrosion processes may also be used to synthesizenanostructured materials224. Here, a strategy for sequentially fabricating highlyoriented, 1D ZnO nanorod arrays over large areas on metallic zinc via cathodic de-lamination of a polymer coating using aqueous KCl at room temperature has beenreported. Moreover, it have been demonstrated that the ZnO nanorods mediate thephoto-degradation of the organic dye methyl orange under illumination with redlight.

9.2 Results

9.2.1 Structure and morphologyZnO nanorods were grown for typically 6-8 h (maximum 15-18 h) by coating azinc substrate with poly(vinylbutyral) (PVB), subsequently preparing a defect inthe polymer coating, and exposing the defect to a drop of 0.1 M aqueous KCl(unless noted otherwise) at room temperature. The structure and morphology ofZnO nanorods grown over a zinc substrate were examined by scanning electronmicroscopy (SEM) and transmission electron microscopy (TEM) (Figure 9.1). TheSE micrographs demonstrate that well-ordered ZnO nanorods grow densely anduniformly over large areas (Figure 9.1a and 9.1b). Nanorod growth was observedover the full sample, typically 1 cm2, up to sizes of 4 cm2, in the samples usedhere. Nanorod diameters range between 20-70 nm (Figure 9.1a). Diameters werealso verified by tapping mode atomic force microscopy (AFM, Figure 9.1e). Thediameter of the ZnO nanorods observed by AFM is slightly higher than observed inSEM images. This difference results from the tip convolution due to the high aspectratio of ZnO nanorods225. The relatively flat top surfaces show that the surface israther smooth and grown nanorods have similar lengths. The length was found tobe 800-1000 nm from scanning TEM (STEM) images of single nanorods detachedfrom the ZnO nanorods array (Figure 9.1c). The lattice planes of the hexagonalnanorods can be clearly seen in the high resolution TEM image (HRTEM) froma single nanorod (Figure 9.1d), from which the selected-area diffraction (SAED)pattern (inset in Figure 9.1d) proves the crystalline nature with a (0001) growthdirection.The crystalline nature of the formed ZnO over the complete sample area is verified

by X-ray diffraction (XRD) (Figure 9.1f). In addition to metallic zinc peaks arisingfrom the substrate, the sharp (002) peak of ZnO indicates the highly preferentialgrowth of ZnO nanorods along the c-axis91.X-ray photoelectron spectroscopy (XPS; Figure 9.2a), energy dispersive X-ray

microanalysis (EDX) measurements (Figure 9.2b) and Raman spectroscopy (Figure9.2c) confirm the purity of the resulting phase. Figure 9.2a shows the XPS surveyspectrum obtained from surface of the nanorod arrays. Prominent peaks originatingfrom zinc and oxygen are obvious. A small amount of carbon (≈ 1%) is detected as

78

9.2. Results

contaminant. Except for zinc, oxygen and carbon, no other elements were detectedwithin the sensitivity of XPS. These results are consistent with EDX, where onlyzinc and oxygen can be detected (Figure 9.2b). The semiquantitative EDX analysisyields ≈ 55% Zn and ≈ 45% O.Figure 9.2c shows the Raman spectrum of a ZnO nanorod array. The three main

peaks are marked as A, B and C. The Raman peak at 334 cm−1 (A) is originatingfrom a multiphonon process226,227, while the sharp peak at 437 cm−1 (B) correspondsto the characteristic E2 mode of ZnO with an hexagonal Wurtzite lattice228,229. Ad-ditionally, a broad band detected around 578 cm−1 (C), corresponds to LO phonons(E1 and A1 modes)230. The observation of the LO phonon in the backscattering ge-ometry used here is an indication of the presence of structural defects, see discussionelsewhere174,230,231.The effect of different electrolyte concentrations on the morphology of ZnO nanos-

tructures was also investigated. In particular by exposing the zinc substrate tohigher concentration of KCl (1 M and 3 M), the diameter of 1D ZnO nanostructurereduced from 70 nm to 10 nm. The resulting wire structures are not isolated, as inthe case of lower electrolyte concentration. Instead of rods, which are observed forelectrolyte concentrations below 1 M, pyramidal structures are formed, as shown inFigure 9.3a and b. An extreme case is observed when 3 M KCl is used, where thinZnO nanorods are obtained (Figure 9.3c and 9.3d). The different diameters observedin electrolytes with different Cl− concentration are attributed to the modification ofgrowth rates on different crystal faces by Cl− adsorption. As faces phases posses dif-ferent polarities192,232, different susceptibility towards ion adsorption is reasonable.The observations made in this work indicate that the ZnO nanostructures grow af-ter the passage of a delamination front under the polymer coating, as illustratedin Figure 9.412. The polymer itself has the function to contain a confined reactionenvironment, and is not actively participating in the reaction. In the cathodic initialregion of the delamination front, oxygen reduction leads to the formation of alkalineconditions. After this part of the delamination front progressed through a certainregion, the region becomes the local anode, where zinc dissolution to Zn2+ occurs12.Above a critical Zn2+ concentration, when the pH is >9233–235, zinc oxide nucleates.Due to the instability of one of the fundamental surfaces of ZnO, the oxide grows inrod-shape, as observed here91,236. It is generally believed that low supersaturationlevels favour 1D ZnO nanostructure growth236. A schematic view of the process ofcathodic delamination and consequent corrosion product growth is shown in Figure9.4. In the absence of the polymer, i.e. when homogeneous corrosion of zinc occurs,no formation of rod-like nanostructures was observed. This cathodic delaminationbased nanorod synthesis approach is a sequential process, as opposed to the com-monly practised synthesis approaches, in which nanorods grow in parallel over thefull sample area210,211.

79


Figure 9.1: (a) SEM image of ZnO nanorod array grown over a zinc substrate viatailored corrosion process carried out at room temperature, (b) magnifiedcrossectional SEM image of ZnO nanorod arrays, (c) STEM micrographof single nanorod, (d) HRTEM image with SAED pattern in the insetof single nanorods, (e) AFM topography image and (f) X-ray diffractionpattern.

80

9.2. Results

1000 800 600 400 200 0

Zn 3p

Zn 3s

Zn LMMO 1s

O KLL

Zn 2p

Binding energy / eV

1 2 3 4 5 6 7 8 9 10

Zn

O

Zn

Energy / keV

300 400 500 600 700 800 900

C

B

A

Wavenumber / cm-1

(a) (b)

(c)

Figure 9.2: (a) XP survey spectrum, (b) EDX spectrum and (c) Raman spectrumof ZnO nanorod array.

9.2.2 Light absorption and photocatalytic activityIn general, electrochemically produced ZnO is known to have an electronic structurethat is rather different from the electronic structure of bulk ZnO173,174. In thin ZnOnanorods, quantum size effects may become important and defect levels lead to anabsorption of light in the visible213,231,237.Figure 9.5a shows the photoluminescence spectrum of the as-grown ZnO nanorod

arrays measured at room temperature with excitation at 325 nm (3.81 eV). Thenanorods exhibit a UV emission at ∼368 nm (∼3.4 eV), which corresponds to freeexcitonic emission209, while wide visible emission band in the range of 500-700 nm(2.5-2 eV) is attributed to different structural defects in ZnO209,231,238–240. Emissioncentred at 700 nm is generally attributed to oxygen interstitials213,241. Emission

81


Figure 9.3: SEM micrographs showing the morphology of ZnO nanostructures afterdelamination with different electrolyte concentrations: (a) and (b) 1 MKCl, (c) and (d) 3 M KCl.

at lower wavelengths is attributed to oxygen vacancies230,238. Because of the largepeak width, a unique assignment of the spectrum to certain dominant defects isdifficult, especially as other factors such as lattice strain cause energy shifts242. Arecent thermodynamic analysis of energy levels of point defects shows that ZnO isunstable towards Zn vacancy formation243. Here, it is concluded that very likelya combination of several types of defects contribute to the observed luminescence.A UV/visible reflection absorbance spectrum of ZnO nanorods on the Zn substrate(Figure 9.5b) shows a strong increase in absorbance below 390 nm (3.2 eV), becauseof the main electronic transition in ZnO. However, in addition, a strong, structurelessabsorption of light is present throughout the visible spectral range, which is likelythe result of a combination of multiple scattering of light in the nanorod arrays, andlight absorption by the nanorods174. The absorption is caused by defects in the ZnOcrystal structure, while the rod-like morphology efficiently traps light via a multiplescattering process.The absorption of visible light inspired the use of the ZnO structures in photo-

catalysis experiments using visible light. Previously, ZnO nanorods have been used

82

9.2. Results

Figure 9.4: Schematic representation of corrosion-driven growth of ZnO nanorodsarrays.

as photocatalysts with excitation energies above the band gap e.g. for the degrada-tion of pollutants or bacteria244,245. Here, the photocatalytic degradation of methylorange (MO) under illumination with visible light was investigated. After start-ing red illumination, UV/visible spectra of the MO solution were measured (Figure9.5c) and used to calculate the concentration of MO (Figure 9.5d). After 6 h ofillumination at 633 nm, more than 90% of MO was decomposed. Two types of con-trol experiments were conducted (Figure 9.5d). In the first control experiment, thesolution was irradiated with the laser light in the absence of nanorod arrays and nodecrease in MO concentration was observed (Figure 9.5d, green symbols). In thesecond control experiment, ZnO nanorod arrays were immersed in an MO aqueoussolution, and the solution’s absorption spectra were recorded in the absence of il-lumination (Figure 9.5d, red symbols). In this case, a certain low decompositionwas observed, but at significantly lower rate compared to the presence of illumina-tion. If the light source at 633 nm is replaced by a light source at 532 nm, i.e. at awavelength where the absorbance of MO is already significant, almost no decreasein MO concentration with time is observed. This observation is explained by thetotal absorption of light in the MO solution resulting in rather low intensity at thelocation of the ZnO nanorods. A presumable mechanism of the observed photocat-alytic activity is the binding of diradical dioxygen to the exciton generated by lightabsorption, and a subsequent reaction of the generated oxygen species with the or-ganic molecules. ZnO has been shown to facilitate the generation of reactive oxygenspecies207,237. “Free electron” defects inside the ZnO are also important for the cat-alytic activity in methanol synthesis246. While the TEM images and XRD patterns

83


Inte

nsity /

a.u

.

Wavelength / nm Time / min

Wavelength / nmPhoton energy / eV

Energy / eV

Figure 9.5: (a) Room-temperature photoluminescence spectrum of ZnO nanorod ar-rays, (b) reflection absorption spectrum of ZnO nanorod arrays (the pho-ton energies on the top axis are rounded equivalents to the wavelengthsat the corresponding label on the bottom), (c) UV/visible spectrum ofmethyl orange solution after different irradiation times, (d) molar con-centration c normalized to initial molar concentration c0 of methyl orangeas function of time under illumination in the presence (––) or absence(–N–) of ZnO nanorod arrays, as well as in the presence of ZnO nanorodarrays but in without illumination (–•–).

confirm that crystalline ZnO is dominating in the produced nanorods, the presenceof the visible photoluminescence,238 and the LO phonon in the Raman spectrum174,point to a prominent presence of defects in the produced structures. Several types ofdefects are likely to contribute. Defect-related electronic states inside the band gapare responsible for the absorption of visible light. As the rod-like morphology leadsto strong in-plane scattering of light, even low absorption coefficients will eventuallylead to photon absorption after multiple scattering events. The observed photocat-alytic activity is therefore a consequence of a combination of morphology and defect

84

9.3. Summary

structure in the ZnO.

9.3 SummaryIn conclusion, a tailored corrosion process based on cathodic delamination of apolymer film on metallic zinc yields dense arrays of ZnO nanorods. This room-temperature process is extraordinarily simple. Because it relies on a sequentialprocess, in this case the passage of a delamination front, the process can easily beupscaled to large areas. The resulting nanorod arrays have been shown to be ableto use the red part of the visible spectrum to photocatalytically decompose methylorange. The nanostructures produced by this method may potentially be used todecompose persistent pollutants. The wide availability of galvanised steel and theeasy scalability of the process will enable a fabrication of the nanostructures on am2 scale. The nanorod synthesis scheme may also enable other (e.g. optoelectronic)applications of ZnO nanorods on large areas.

85

Chapter 10: Non destructive in situ Ramanstudies on the growth of ZnOnanorods

IntroductionIn recent years, there has been a significant interest in synthesis of nanostructuredmaterials owing to their unique and novel applications in catalysis247, sensors248,electronics249, photonics250, and mechanics251. In particular, one dimensional (1D)ZnO nanostructures such as wires and rods have been extensively researched in thepast few years due to their excellent optical, mechanical, and electrical properties252.As 1D ZnO nanomaterials offers large surface area availability and quantum con-finement effect, they exhibit a very fast response, high carrier mobility and novelproperties which would improve the performance of future nanodevices. In order toproduce high performance 1D ZnO based nanodevices, the synthesis of high qual-ity ZnO nanorods is desired. Thus significant efforts have been devoted to thedevelopment of methods to synthesize high quality 1D ZnO nanostructures, suchas chemical vapor deposition (CVD), vapor-liquid-solid growth (VLS)253,254, pulsedlaser deposition214, and hydrothermal approach86,90. In recent times, hydrothermalapproach has gained interest for the synthesis of 1D ZnO nanostructures due totheir simplicity, commercial feasibility, and scalability255. The pioneer work in thisarea has been carried out by Vayssieres et al.91, where they have synthesized ZnOnanorods on various heterogeneous substrates via solution phase approach. Morerecently, our group has further simplified this growth method by synthesizing ZnOnanorods via tailored corrosion process at room temperature.Despite of significant progress in the synthesis of ZnO nanostructures, there are

still some unanswered questions concerning the performance and reliability of ZnObased devices256. For instance, for all synthesized ZnO nanostructures, there isa trade-off between conductivity and near band edge emission (NBE) efficiency.High conductivity in ZnO is achieved by increasing the number of defects in aZnO crystal structure that provides sufficient free carriers. However, this comeswith a compromise in NBE emission efficiency due to direct interband radiativerecombination257. Moreover, application of ZnO nanomaterials in a visible lightphotocatalysis is strongly dependent on the concentration of oxygen deficiencies

87

Chapter 10. In situ Raman studies on the growth of ZnO nanorods

or other native defects in a crystal structure, which is responsible for visible lightabsorption in ZnO. Hence, one of the factors which play a vital role in achieving thedesired property of ZnO based nanodevice is the optimal concentration of defects in ananomaterial. A key towards resolving these issues lies in the detailed understandingof ZnO nanostructure growth process. Most of the previous studies performed toprobe the growth kinetics of ZnO nanostructures are achieved by using the UV-Vis spectroscopy258,259. More recently, laser assisted hydrothermal growth of ZnOnanowires was investigated based on optical in situ growth monitoring260. However,both of the above mentioned strategies only gives insight into the growth of crystalsand no information on the evolution of crystal/electronic structure of ZnO couldbe obtained. Indeed, properties of ZnO nanostructures depend not only on theirsize but also on the crystal structure. Hence, it also vital to study the evolutionof crystal structure during the growth of ZnO nanostructure. X-ray diffraction is awidely used technique to probe the crystal structure of nanomaterials; however it isevidently quite difficult to use it for in situ measurement during the hydrothermalgrowth. Hence, another fast and non-destructive technique is required for in situmonitoring of crystal structure and growth kinetics of ZnO.Raman spectroscopy is a well-established nondestructive technique to study the

chemical composition, crystal structure and defects in a nanomaterial. It has beenwidely used to probe the physical state, phonon confinement, and crystal structureof ZnO135. ZnO belongs to the P63mc space group, where all atoms occupy C3υsites. According to group theory, four Raman active modes of A1, E1, and E2(high and low) are expected for wurtzite ZnO. The phonons belonging to A1 andE1 symmetry are polar and therefore exhibit different transverse-optical (TO) andlongitudinal-optical (LO) frequencies261. On the other hand, E2 modes are due tothe vibration of Zn sub-lattice (E2-low) or O sublattice (E2-high). With the Ramanactive modes of ZnO, recently Raman spectroscopy is utilized to determine thecrystal orientation of anisotropic ZnO nanorods252. Moreover, it has been employedfor in situ analysis of electronic and structural aspects of passive ZnO film grownunder alkaline conditions262. This provides us the motivational grounds to utilizeRaman spectroscopy for in situ studies on growth kinetics of ZnO nanorods.In this chapter, kinetic studies on the growth of ZnO nanorods synthesized via

a tailored corrosion process by performing in situ Raman experiments has beencarried out. Furthermore, the effect of salt concentration on the growth and crystalstructure of ZnO nanorods was also studied.

10.1 Experimental details

10.1.1 MaterialsPolyvinyl butyral (PVB), absolute ethanol and KCl were obtained from SigmaAldrich and used as received. Zinc sheets (purity, 99.95%) having a thickness of1.5 mm were obtained from Goodfellow (Cambridge, UK). Zinc sheets were initially

88


mechanically polished using 4000 grit polishing paper and then cleaned ultrasoni-cally in ethanol.

10.1.2 Synthesis of ZnO nanostructuresInitially polished zinc sheets (1 cm × 1 cm × 1.5 mm) were spin coated with 5 wt%solution of PVB in ethanol followed by drying in an oven at 70 oC, to obtain 2 µmthick polymer layer. Afterwards, small defect in the polymer coating was created atthe edge of the sample and the sample was placed on zinc base plate. Then smallreservoir was built around the edge of the sample with the help of fast drying twocomponents adhesive to place an electrolyte. To start the growth of ZnO nanowiresvia tailored corrosion process, reservoir was filled with 1 M aqueous KCl solution(unless noted otherwise). After completion ZnO nanowires growth, PVB coatingwas removed by washing with ethanol, followed by rinsing with distilled water toremove any traces of remaining KCl. The washed samples were dried in nitrogenstream and stored in vacuum for further analysis.

10.1.3 Analytical techniquesThe surface morphology ZnO nanowires were obtained by Scanning Electron Mi-croscopy (SEM, Zeiss/LEO 1550 VP). Before introducing the samples into SEMchamber, PVB coating was removed by excess cleaning with ethanol in order toavoid sample charging. Crystallographic information about the nanorods was ob-tained by X-ray diffraction (XRD, Bruker-AXS D8 with a Cu-Kα source) in grazingincidence geometry.To carryout in situ Raman measurements, the assembly of sample mounted on zinc

substrate was introduced into a custom made chamber with a glass window openingto allow spectral acquisition. To minimize electrolyte evaporation, the humidity wascontrolled ≥90% with the help of humid/dry air flow. The in situ Raman spectra’swere recorded using a Horiba Jobin Yvon LabRAM confocal Raman microscopewith a Argon laser source (emission wavelength is 514.342 nm) and a power of12 mW. The typical spectral acquisition time was in the range of 20 to 30 s. Themeasurements were performed in reflection geometry with a 50x-objective for sampleillumination and scattered light harvesting. The laser spot size on the sample surfacewas approximately 1 µm.

10.2 Results and discussionPrior to the growth of nanorods, zinc substrate was spin coated with 5 wt% PVBin order to have a thin layer of polymer covering the whole surface. The presence ofthin polymer layer hinders the immediate spreading of an electrolyte over the wholesample surface and only allows an ultra-thin layer of an electrolyte to go throughpolymer/metal interface. The sample was then introduced into a custom made

89


chamber, where the humidity was controlled above 90% via flow of humid/dry air.Later on, the coated zinc was exposed to 1 M KCl in order to grow ZnO nanorods viacathodic delamination process. The detailed mechanism of ZnO nanorods growth viacathodic delamination is explained in chapter 8. The setup for in situ measurementsis shown schematically in Figure 10.1.

O2

Zn(OH)2

Zn(OH)2

Zn(OH)2

Zn(OH)2

Fully grownZnO NR Intermediate

stage

Zn2+

Zn2+

Zn2+

Zn2+

Zn Zn + 2e2+ -

Anode

Zn2+

Zn2+

Zn2+

Zn2+

Zn2+

Zn2+

Zn2+

Zn2+

Zn2+

Cathode

Zn

O n

an

oro

d

Zn

O n

an

oro

d

Zn

O n

an

oro

d

Zn

O n

an

oro

d

Zn

O n

an

oro

d

Zn

O n

an

oro

d

O2

Zn(OH)2

OH-

OH-

OH-

OH-

OH-

OH-

Zinc

Electrolyte Polymer Coating

Increasing pHAcidic Basic

Humid Atmosphere

RamanSpectrometer

0 5 10 15 20 2580

85

90

95

100

Humidity (%)

Temperature

Time (min.)

Hu

mid

ity

(%)

20

22

24

26

28

30

Te

mp

era

ture

(oC

)

250 500 750 1000

Tim

e

565

Inte

ns

ity

[a.u

.]

Wavenumber [cm-1]

332

440

Air InletAir Outlet

Humidity/Temp. Sensor

Laser

Figure 10.1: Schematic representation of experimental setup for in situ Raman mea-surements of ZnO nanorods growth via cathodic delamination.

The growth kinetics was monitored at a measurement spot via in situ Raman ex-periment. Figure 10.2 shows the measured Raman spectra over a corroding sample asa function of time. At the initial stages of experiment, hardly any discernible peakscould be observed for freshly polished and polymer coated zinc substrate. However,as the electrolyte/delamination front approaches the measurement spot, the bandat 565 cm−1 begins to evolve (see Figure 10.2a). This peak is attributed to A1-LOphonon mode of ZnO, indicating the growth of amorphous ZnO228. This growth iscaused by degradation of bonds at metal/polymer interface which induces local pH

90


changes in the vicinity of delamination front by oxygen reduction (pH increase) orby formation of zinc-hydroxide complexes (pH decrease). Zn(OH)2 is believed to bea precursor that forms during corrosion of zinc and which eventually precipitatesas the solution becomes supersaturated to form ZnO. However, no evidence for thepresence of Zn(OH)2 within the sensitivity level of these in situ Raman measure-ments could be found. This is in agreement with the previous studies performedon zinc under alkaline conditions262. As the time progresses, one additional bandevolve at 438 cm−1 and starts to grow. This peak is assigned to E2-high phononmode of ZnO, indicating the growth of wurtzite ZnO crystals228. After 12 h, nofurther increase in the peak intensity has been observed, which indicates that thegrowth of crystalline ZnO is no more pronounced. The increase in a peak height of

300 400 500 600 700 800 900 1000

330 cm-1

565 cm-1

Wavenumber

0 min

intermediate scans(interval:1 hr)

after 12 hr

438 cm-1

0 150 300 450 600 7500.0

0.5

1.0

1.5

2.0

Growth Time / min

wurtzite & amorphous ZnO ratio

0 150 300 450 600 750

0

1500

3000

4500

6000

Growth Time / min

wurtzite ZnO bandamourphous ZnO band

Inte

ns

ity

Ra

tio

Inte

ns

ity

Ra

tio

cm-1

(a) (b)

(c)

/

Inte

nsit

y / a

.u.

Figure 10.2: In situ monitoring of ZnO nanorods growth via tailored corrosion pro-cess with 1 M KCl. (a) In situ Raman spectra acquired during growthof ZnO nanorods as a function of time, (b) intensity ratios of wurtziteZnO and defect rich ZnO peaks with respect to the baseline, (c) inten-sity ratio of wurtzite ZnO peak vs amorphous ZnO peak.

wurtzite ZnO (438 cm−1) and amorphous ZnO (565 cm−1) as a function of time isdepicted in Figure 10.2b. The plot clearly demonstrates the monotonic increase inthe peak intensities ratio over time. The possible explanation of this phenomenon

91


is the depletion in the concentration of Zn(OH)2 or neutralization of the interfacialpH, which is responsible for the growth of ZnO by dehydration. As ZnO grows onzinc, the rate of zinc oxidation goes down due to passivation of the surface. Thisalso leads to decrease in the rate of O2 reduction (as it has to diffuse through theformed passive film) which is responsible for the formation of OH− ions. Hence, theinterfacial pH value gets lower, which prevents any further precipitation of ZnO.To gain insight into the competing behavior of two growing peaks, the height

ratio of wurtzite ZnO peak (440 cm−1) versus amorphous/defect rich ZnO peak(565 cm−1) as a function of time (see Figure 10.2c) is plotted. It can be seen that atinitial stages of growth, peak height ratio results in a value lower than 1, indicatinggrowth of defect rich ZnO. However, after the initial period of the experiment (afterca. 2 h), growth of wurtzite ZnO becomes dominant as can be seen in Figure 10.2c.Finally after approximately 12 h the ratio becomes stable, indicating that no furtherchanges in the crystal structure of growing ZnO takes place. This leads to concludethat initially, growth of amorphous ZnO is favored, while in the later stages morecrystalline ZnO nanorods grow.Figure 10.3 shows the Raman and PL spectra performed along the length of the

nanorod cluster using 514 nm and 355 nm excitation wavelenght, respectively. Forbetter visualization of differences in the peak intensities along the length of the rod,different colors are used in a spectra. Figure 10.3I shows the room temperaturePL mapping performed (using 355 nm excitation wavelength) along the length ofthe nanorod cluster. The spectrum is marked with A, B and C corresponding tothe lower region, middle region and upper region of the nanorod, respectively. It isquite evident that the intensity of green emission which is related to the presenceof oxygen vacancies (or other point defects)238 is different along the length of thenanorod cluster. On one hand, the PL intensity of green emission is lower for upperregion (C) of ZnO nanorods, indicating that the ZnO nanorod cluster has a lowdefect density263 and good crystallinity264. While on the other hand, higher greenemission intensity at the bottom of the rod represents the higher density of oxygen-related defects. On the same nanorod cluster, Raman mapping was performedalong the length using 532 nm excitation source. The change in the intensity couldbe identified for two main bands centered at around 438 cm−1 (E2-high) and 565cm−1 (E1-LO) (see Figure 10.3II). Initially, at the lower end of the nanorod cluster(marked with A), the intensity of E1 band is higher indicating the presence greateramount of defects, while in the upper region (marked with C) the intensity of E1band gets lower indicating presence of more crystalline ZnO. The region markedwith B indicates the presence of both crystalline and defect rich ZnO. Hence, ex situRaman analysis performed along the length of nanorod cluster, further confirms thestatement that at the initial stages, ZnO nanorod has a defect rich structure whicheventually gets more crystalline as the growth proceeds.

92


II

Inte

ns

ity

/a

.u.

Wavenumber / cm-1

350 400 450 500 550 600 650

C

B

A

I

Inte

nsit

y/a.u

.

Wavelength / nm

450 465 480 495 510

C

B

A

Substrate

ZnO

nanoro

d

B

A

C

400 nm

III

Figure 10.3: Raman mapping of ZnO nanorod arrays. (I) PL spectra along thelength of the nanorod array with 355 nm excitation wavelength, (II)Raman spectra along the length of the nanorod array with excitationwavelength of 514.32 nm, (III) SEM image of a nanorod cluster used inex-situ experiments.

10.2.1 Change in surface morphology during growthTo gain further insight in to the changes in the morphology of growing ZnO nanorods,SEM micrographs with their corresponding Raman spectra were obtained after spe-cific time intervals (see Figure 10.4). As shown in Figure 10.4-Ia, the zinc surfaceis covered with the thin oxide layer during the initial stages of growth. This grownoxide is amorphous which is evident from the attributed Raman spectra (Figure10.4-Ic) where only the peak at 568 cm−1 is present that corresponds to the defectrich/amorphous ZnO. After approximately an hour, nanorod type structure of ZnOstarts to grow as can be seen in Figure 10.4II, which can be confirmed by appear-ance of 440 cm−1 peak in the acquired Raman spectra. Finally after 12 h, the ZnOnanorods are grown upto 1 µm (see 10.4-III) and the wurtzite ZnO peak is dominant

93


400 nm

500 nm 400 nm

600 nm

600 nm

600 nm

300 400 500 600 700

326 cm-1

567 cm-1

440 cm-1

300 400 500 600 700

332 cm-1

567 cm-1

440 cm-1

300 400 500 600 700

568 cm-1

324 cm-1

I II III

a

b

c

Wavenumber cm-1

/ Wavenumber cm-1

/ Wavenumber cm-1

/

Inte

ns

ity

/ a

.u.

Inte

ns

ity

/ a

.u.

Inte

ns

ity

/ a

.u.

Figure 10.4: SEM micrographs with their corresponding Raman spectra. (I) after15 mins, (II) after 1 h, (III) after 12 h. (a) top view, (b) cross-sectionalview (c) ex situ Raman spectra.

in the corresponding Raman spectra. The comparison of the relative height of twobands at 440 cm−1 and 568 cm−1 suggests that the band at 440 cm−1 increases moreintensively with time then the band at 568 cm−1. This observation also suggeststhat the concentration of defects in a growing ZnO crystal reduces as the growthprogresses.

10.2.2 Effect of electrolyte concentrationPrevious studies have shown that corrosion of zinc is accelerated by increase inthe concentration of Cl− ions in aqueous solutions265,266. Hence, it is important tostudy the effect of chloride concentration on the growth kinetics of ZnO nanorods.Consequently, PVB coated zinc surface was exposed to 3 M KCl in a humid airatmosphere and the growth kinetics was monitored via in situ Raman experiments.Figure 10.5a depicts the measured Raman spectra over a corroding zinc sample asa function of time. As in the case of 1 M KCl, initially the peak at 566 cm−1

appears which corresponds to growth of defect rich ZnO. Later on, wurtzite ZnO

94


peak evolves at 440 cm−1 and grows over increasing period of time. However, increasein the intensity of wurtzite ZnO band stops approximately after approximately 4 h.This suggests that no further growth or modification of ZnO takes place. To trackthe changes in the crystal structure, the height ratio of wurtzite ZnO peak anddefect rich ZnO peak is plotted as performed in the case of 1 M KCl. As shownin 10.5c, for approximately initial 30 minutes the ratio stays below 1, which isan indication that growth of defect rich ZnO is more pronounced then growth ofwurtzite ZnO. It is also interesting to see a sharp dip in a curve after nearly 15 min,which shows an increased rate in the growth of defect rich ZnO. Afterwards, theratio goes over 1, indicating that now wurtzite ZnO crystal structure is growingmore intensively. Finally, the ratio becomes stable after approximately 4 h whichindicates that there are no more changes in the crystal structure taking place. Theresult depicts a similar trend for growth of ZnO nanorods in both 1 M KCl and 3 MKCl except the growth rate is increased in the latter case. This increased rate ingrowth of ZnO nanorods is caused by fast zinc corrosion due to higher electrolyteconcentration. Moreover, slightly higher ratio of amorphous/defect rich ZnO peaksuggest the presence of higher number of oxygen deficiencies in ZnO nanorods grownwith 3 M KCl in comparison with rods grown with 1 M KCl.

10.2.3 Optical properties of ZnO nanorodsAs known from the literature, the optical properties of ZnO nanostructures such asphotoluminescence and visible light absorption strongly depends on the presence ofdefects that are present in the crystal of an oxide. Indeed, in situ Raman measure-ments showed above reveals the presence of higher number of oxygen deficiencies(or point defects) evident from the higher oxygen deficient peak intensity (see Fig-ure 10.5) in a nanostructure grown with 3 M KCl. Hence, to verify this statement,UV/Vis absorption spectra (in reflection geometry) of as grown ZnO nanostructureswith 1 M and 3 M KCl were measured and are shown in 10.7. The spectra showa strong absorption at around 372 nm, which corresponds to the charge-transferprocess from valance band to conduction band. However, the light absorption in thevisible range is higher for wires synthesized with 3 M KCl then for 1 M KCl. Thiscould be explained by the fact that the higher concentration of oxygen defects (orpoint defects) in nanorods grown with 3 M KCl leads to increased absorption in thevisible light region.

95


300 450 600 750 900

331 cm-1

566 cm-1

Wavenumber / cm-1

0 min

intermediate scans (interval: 30min)

270440 cm

-1

0 50 100 150 200 250 300

0.0

0.3

0.6

0.9

1.2

1.5

1.8

Growth Time / min

wurtzite & amorphous ZnO ratio

0 50 100 150 200 250 300

0

1500

3000

4500

6000wurtzite ZnO bandamorphous ZnO band

Growth Time / min

Inte

nsit

y R

ati

oIn

ten

sit

y R

ati

omin

(a) (b)

(c)

Inte

nsit

y / a

.u.

Figure 10.5: In situ monitoring of ZnO nanorods growth via tailored corrosion pro-cess with 3 M KCl. (a) In situ Raman spectra acquired during growthof ZnO nanorods as a function of time, (b) Intensity ratios of wurtziteZnO and defect rich ZnO peaks with respect to the baseline, (c) inten-sity ratio of wurtzite ZnO peak vs amorphous ZnO peak.

96


400 nm

400 nm

600 nm

600 nm

300 400 500 600 700

Wavenumber / cm-1

300 400 500 600 700

Wavenumber / cm-1

Inte

ns

ity

Inte

ns

ity

327 cm-1

440 cm-1

567 cm-1

440 cm-1

567 cm-1

I II

a

b

c

Figure 10.6: SEM micrographs and their corresponding Raman spectra measured exsitu after different delamination times. (I) after 15 min, (II) after 4 h,(a) top view, (b) cross-sectional view (c) corresponding ex situ Ramanspectra.

97


300 400 500 600 700 8000.9

1.2

1.5

1.8

2.1

2.4

2.7

Ab

so

rpti

on

Wavelenght / nm

1M KCl

3M KCl

Figure 10.7: UV/VIS reflection absorption spectrum of ZnO nanorod arrays grownwith 1 M KCl and 3 M KCl respectively.

10.3 SummaryIn situ Raman studies of ZnO nanorod growth reveals that at the initial stages ofgrowth, nucleation of defect rich ZnO takes place, while in the later stages growth ofwurtzite ZnO becomes pronounced. The growth of nanorods stops due to depletionin the concentration of zinc hydroxide complexes which is responsible for the growthof ZnO via dehydration. Moreover, it was also shown that the nanorod growthkinetics could be accelerated by increasing the concentration of an electrolyte. Theseresults provide a key insight into the growth kinetics and evolution of the crystalstructure of ZnO nanorods which could play a vital role in their future optoelectronicapplications.

98

Chapter 11: Summary and Outlook

11.1 SummaryUltrathin polymer coatings were synthesized by thermally initiated free radical poly-merization on top of zinc substrates. These coatings were covalently bound to thesurface with well defined interfacial chemistry. In order to alter the adhesion prop-erties, polymer coatings were grown on top of silane and ester functionalized zincsamples. Moreover, PMMA and PS were employed as a model polymers to study thecathodic delamination behavior as they mimic the properties of both aliphatic andaromatic based polymers used as polymer coatings in corrosion protection. In orderto alter the polymer structure, PMMA and PS coatings were crosslinked by addingbifunctional molecule having different aliphatic chain length (ethylene diacrylate,EDA and hexanediol diacrylate, HDA) and were added in different molar ratios (i.e.25 mol% and 50 mol%). The thickness of the model polymer coatings were in therange of 10-25 nm.Scanning Kelvin probe measurements shows that chemisorbed polymer coatings

show larger resistance against cathodic delamination than physisorbed PVB coating.The delamination rate decreases to <1/10 only by introducing covalent bonds tothe base metal. Moreover, it was also shown that silane bonded PMMA coatingsexhibit larger resistance against cathodic delamination than coatings attached tothe substrate via ester bond. This enhanced resistance is a result of strong silanebond at the polymer/metal interface. While for ester bound PMMA coating, thebond at polymer/metal interface is relatively weak and susceptible to alkaline esterhydrolysis. Introduction of a cross-linker to a silane and ester bonded PMMA coatingfurther slows down the delamination rate, with another factor of ∼10 between non-cross-linked PMMA and samples with 50% cross-linker. Comparing the two cross-linkers used in this study, the more hydrophobic HDA yields the lowest delaminationrates of all samples.Similarly, PS coatings bound to zinc via organosilanes show enhanced resistance

against cathodic delamination when compared with coating bound exclusively viavan der Waals interactions. Furthermore, PS coating without cross-linker showslarger resistance toward cathodic delamination when compared to linear PMMAcoatings. This enhanced resistance is attributed to higher hydrophobicity of PS.However, adding a cross-linker to PS coatings does not significantly reduces thedelamination rate.

99

Chapter 11. Summary and Outlook

The cathodic delamination rates of PMMA coatings synthesized in this work onmetal substrates are not diffusion limited but are reaction controlled. On the otherhand, for PS coatings, initially the cathodic delamination is reaction controlled butbecomes diffusion controlled after long times.For a study of the dependence of cathodic delamination on interface morphol-

ogy, different ZnO structures (i.e. twin-plate, nanorod, and sphere) were synthe-sized. PVB was used as a model coating to exclude the covalent interaction atpolymer/oxide/metal interface. The delamination was fastest for twin-plate-shapedparticles, and lowest for spherical ZnO particles. Moreover, the comparison of thesurface morphologies before and after the delamination experiments shows that dis-solution of the oxide in the interfacial region occurs. The kinetics of delamination islikely determined by the nature of the dominating surface termination of ZnO. Theseresults suggest that controlling the oxide surface morphology offers the chance tosignificantly increase resistance of polymer-coated galvanized steel towards cathodicdelamination.After evaluating the effect of different ZnO morphologies on the cathodic delam-

ination kinetics, it was of interest to study the corrosion products formed at thesurface as a result of zinc corrosion. Hence, corroded zinc substrates were character-ized by SEM to analyze the corrosion products formed at the surface. The resultsrevealed the formation of ZnO nanorod structures at the surface. This provides aunique and tailored corrosion process based on cathodic delamination of a polymerfilm to yield dense arrays of ZnO nanorods. This synthesis route is extraordinarilysimple and carried out at room temperature, hence can easily be up-scaled to largescales. Moreover, the resulting nanorod arrays have been shown to be able to usethe red part of the visible spectrum to photocatalytically decompose methyl orange.The investigation of growth of ZnO nanorods with in situ Raman experiments

were performed with the aim to investigate and get further insights into the growthmechanism via the tailored corrosion process. The results show that the defectconcentration was different along the length of the nanorods. In particular, at initialstages of growth, nucleation of defect rich ZnO takes place, while in the later stagesgrowth of wurtzite ZnO becomes pronounced. Moreover, it was also shown thatdefect concentration in ZnO nanorods could be altered by varying the concentrationof electrolyte used for the cathodic delamination process.The results presented in this work may contribute to the fundamental understand-

ing of interactions at polymer/metal interface. In particular the disbonding kineticsof polymer coatings from metal substrates due to cathodic delamination. It furtherallows to utilize the mostly undesired cathodic delamination process to synthesizeZnO nanorods at large scales via a very simple and economically feasible process.Moreover, the application of these as synthesized ZnO nanorods in visible light pho-tocatalytic decomposition of methyl orange provides a potential way to be used todecompose persistent pollutants.

100

11.2. Outlook

11.2 OutlookBased on the fundamental and technical understanding achieved in this work, asolid basis for further investigations and step out to industrial applications can beproposed. Consequently, it would be valuable to study the delamination kineticsof polymer coatings bonded to metal substrate via different covalent bonds suchas ether and amide linkages. Both amide and ether linkages have already beenshown to bound to inorganic substrates via covalent bonds30. Furthermore, in or-der to achieve polymer coatings with well defined chain lengths, it is also desirableto utilize controlled radical polymerization approaches such as atom transfer freeradical polymerization. Comparing the delamination kinetics of these differentlybonded polymer coatings to the metal substrates could provide the basis for the im-provements in the commercially applied polymer coatings for corrosion protection.Furthermore, the tailored corrosion route developed in this work for the fabricationof ZnO nanorods on metallic zinc at room temperature could be used to grow ZnOnanostructures on other inorganic substrates. In particular, for optoelectronic appli-cations such as solar cells, ZnO nanorods grown on surfaces like silicon and/or poly-mers are of high technological interest. However, due to complexities in the growthprocess of ZnO nanostructures, their fabrication on top of these inorganic substratesis limited. The proposed route in this work for fabrication of ZnO nanorods pro-vides mild growth conditions that could be used grow ZnO nanostructures on topof substrates interesting for optoelectronic applications.

101

Bibliography

[1] Koch, G. H., Brongers, M. P. H., Thompson, N. G., Virmani, Y. P., Payer,J. H., Corrosion cost and preventive strategies in the united states, Technicalreport, Transportation Research Board, 2002.

[2] Uhlig, H. H., Revie, R. W., Corrosion and corrosion control: an introductionto corrosion science and engineering, Wiley, 1985.

[3] Gavrila, M., Millet, J. P., Mazille, H., Marchandise, D., Cuntz, J. M., Corro-sion behaviour of zinc–nickel coatings, electrodeposited on steel, Surf. Coat.Technol., 123(2–3), 164, 2000.

[4] Taheri, P., Lill, K., de Wit, J., Mol, J., Terryn, H., Effects of zinc surface acid-based properties on formation mechanisms and interfacial bonding propertiesof zirconium-based conversion layers, J. Phys. Chem. C , 116(15), 8426, 2012.

[5] Sarfraz, A., Posner, R., Lange, M. M., Lill, K., Erbe, A., Role of intermetallicsand copper in the deposition of ZrO2 conversion coatings on AA6014, J. Elec-trochem. Soc., 161(12), C509, 2014.

[6] El Azhar, M., Mernari, B., Traisnel, M., Bentiss, F., Lagrenée, M., Corrosioninhibition of mild steel by the new class of inhibitors [2,5-bis(n-pyridyl)-1,3,4-thiadiazoles] in acidic media, Corr. Sci., 43(12), 2229, 2001.

[7] Bentiss, F., Lagrenee, M., Traisnel, M., Hornez, J. C., The corrosion inhibitionof mild steel in acidic media by a new triazole derivative, Corr. Sci., 41(4),789, 1999.

[8] Vimalanandan, A., Lv, L.-P., Tran, T. H., Landfester, K., Crespy, D., Rohw-erder, M., Redox-responsive self-healing for corrosion protection, Adv. Mater.,25(48), 6980, 2013.

[9] Zheludkevich, M. L., Shchukin, D. G., Yasakau, K. A., Möhwald, H., Ferreira,M. G. S., Anticorrosion coatings with self-healing effect based on nanocon-tainers impregnated with corrosion inhibitor, Chem. Mater., 19(3), 402, 2007.

[10] Shchukin, D. G., Zheludkevich, M., Yasakau, K., Lamaka, S., Ferreira, M.G. S., Möhwald, H., Layer-by-layer assembled nanocontainers for self-healingcorrosion protection, Adv. Mater., 18(13), 1672, 2006.

103

Bibliography

[11] Iqbal, D., Rechmann, J., Sarfraz, A., Altin, A., Genchev, G., Erbe, A., Synthe-sis of ultrathin poly(methyl methacrylate) model coatings bound via organosi-lanes to zinc and investigation of their delamination kinetics, ACS Appl. Mater.Interfaces, 2014, DOI: 10.1021/am504992r.

[12] Fürbeth, W., Stratmann, M., The delamination of polymeric coatings fromelectrogalvanised steel – a mechanistic approach.: Part 1: delamination froma defect with intact zinc layer, Corr. Sci., 43(2), 207, 2001.

[13] Posner, R., Santa, M., Grundmeier, G., Wet- and corrosive de-adhesion pro-cesses of water-borne epoxy film coated steel i. interface potentials and char-acteristics of ion transport processes, J. Electrochem. Soc., 158(3), C29, 2011.

[14] Bard, A. J., Stratmann, M., Frankel, G. S., Encyclopedia of Electrochemistry.Volume 4: Corrosion and Oxide Films, Weinheim: Wiley-VCH, 2003.

[15] Dickie, R. A., Floyd, F. L., editors, Polymeric Materials for Corrosion Control,volume 322 of ACS Symposium Series, American Chemical Society, Washing-ton, DC, 1986.

[16] Harris, F., Introduction to polymer chemistry, J. Chem. Educ., 58(11), 837,1981.

[17] Zhao, B., Brittain, W. J., Polymer brushes: surface-immobilized macro-molecules, Prog. Polym. Sci., 25(5), 677, 2000.

[18] Moad, G., Solomon, D. H., 1 - introduction, in G. M. H. Solomon, editor, TheChemistry of Radical Polymerization (Second Edition), pages 1–9, ElsevierScience Ltd, Amsterdam, 2005.

[19] Khun, N. W., Frankel, G. S., Effects of surface roughness, texture and polymerdegradation on cathodic delamination of epoxy coated steel samples, Corr.Sci., 67, 152, 2013.

[20] Hare, C. H., Paint film degradation mechanisms and control, Technical report,Transportation Research Board, 2002.

[21] De Wit, F. M., Polymer bonding and adhesion on aluminium and magnesiumalloys, Interface, 42, 316, 2010.

[22] Stratmann, M., Frankel, G. S., Encyclopedia of electrochemistry, Vol. 4 Cor-rosion and oxide films. sl Wiley, VCH, Weinheim, 2003.

[23] Cheever, G. D., Weiss, P., Interface Conversion for polymer coatings., Elsevier,New York, 1968.

104

Bibliography

[24] Herzer, N., Stephanie, S., Ulrich, N., Fabrication of patterned silane based self-assembled monolayers by photolithography and surface reactions on silicon-oxide substrates, Chem. Commun., 46(31), 5634, 2010.

[25] Fadeev, A. Y., McCarthy, T. J., , A., A new route to covalently attachedmonolayers:â reaction of hydridosilanes with titanium and other metal sur-faces, J. Am. Chem. Soc., 121(51), 12184, 1999.

[26] Chen, J., Ruther, R. E., Tan, Y., Bishop, L. M., Hamers, R. J., , J., Molec-ular adsorption on ZnO(1010) single-crystal surfaces: Morphology and chargetransfer, Langmuir , 28(28), 10437, 2012.

[27] Pujari, S. P., Li, Y., Regeling, R., Zuilhof, H., Tribology and stability oforganic monolayers on CrN: A comparison among silane, phosphonate, alkene,and alkyne chemistries, Langmuir , 29(33), 10405, 2013.

[28] Faure, E., Halusiak, E., Farina, F., Giamblanco, N., Motte, C., Poelman, M.,Archambeau, C., Van De Weerdt, C., Martial, J., Jérôme, C., Duwez, A.-S.,Detrembleur, C., Clay and DOPA containing polyelectrolyte multilayer filmfor imparting anticorrosion properties to galvanized steel, Langmuir , 28(5),2971, 2012.

[29] Ruan, C.-M., Bayer, T., Meth, S., Sukenik, C. N., Creation and character-ization of n-alkylthiol and n-alkylamine self-assembled monolayers on 316lstainless steel, Thin Solid Films, 419(1–2), 95, 2002.

[30] Pujari, S. P., Scheres, L., Marcelis, A. T. M., Zuilhof, H., Covalent surfacemodification of oxide surfaces, Angew. Chem. Int. Ed., 53(25), 6322, 2014.

[31] Leidheiser, H., Kendig, M., Mechanism of corrosion of polybutadiene-coatedsteel in aerated sodium-chloride, Corrosion, 32(2), 69, 1976.

[32] Leidheiser Jr., H., Wang, W., Igetoft, L., The mechanism for the cathodicdelamination of organic coatings from a metal surface, Prog. Org. Coat., 11(1),19, 1983.

[33] Hammond, J., Holubka, J., Dickie, R., Surface-analysis of interfacial chemistryin corrosion-induced paint adhesion loss, J. Coat. Technol., 51(655), 45, 1979.

[34] Grundmeier, G., Stratmann, M., Adhesion and de-adhesion mechanisms atpolymer/metal interfaces: mechanistic understanding based on in situ studiesof buried interfaces, Ann. Rev. Mater. Res., 35(1), 571, 2005.

[35] Van Loo, M., Laiderman, D. D., Bruhn, R. R., Filiform corrosion, Corrosion,9(8), 277, 1953.

105

Bibliography

[36] Sorensen, P. A., Kiil, S., Dam-Johansen, K., Weinell, C., Sørensen, P. A., Kiil,S., Dam-Johansen, K., Weinell, C., Anticorrosive coatings: a review, JCTResearch, 6(2), 135, 2009.

[37] Steele, G. D., Filiform corrosion on architectural aluminium: A review, Corros.Method M., 41(1), 8, 1994.

[38] Bautista, A., Filiform corrosion in polymer-coated metals, Prog. Org. Coat.,28(1), 49, 1996.

[39] Funke, W., Blistering of paint films and filiform corrosion, Prog. Org. Coat.,9(1), 29, 1981.

[40] Leblanc, P. P., Frankel, G. S., Investigation of filiform corrosion of epoxy-coated 1045 carbon steel by scanning kelvin probe force microscopy, J. Elec-trochem. Soc., 151(3), B105, 2004.

[41] Ruggeri, R. T., Beck, T. R., An analysis of mass transfer in filiform corrosion,Corrosion, 39(11), 452, 1983.

[42] Williams, G., McMurray, H. N., The mechanism of group (i) chloride initiatedfiliform corrosion on iron, Electrochem. Commun., 5(10), 871, 2003.

[43] Watson, T. M., Coleman, A. J., Williams, G., McMurray, H. N., The effect ofoxygen partial pressure on the filiform corrosion of organic coated iron, Corr.Sci., 2014.

[44] Walter, G., A critical-review of the protection of metals by paints, Corr. Sci.,26(1), 27, 1986.

[45] Baumann, K., Effective use of protective coatings for corrosion, Chem. Tech.,24(12), 780, 1972.

[46] Mayne, J., Mechanism of protection of iron and steel by paint, Anti-Corros.Methods Mater., 20(10), 3, 1973.

[47] Leng, A., Streckel, H., Hofmann, K., Stratmann, M., The delamination ofpolymeric coatings from steel part 3: Effect of the oxygen partial pressureon the delamination reaction and current distribution at the metal/polymerinterface, Corr. Sci., 41(3), 599, 1998.

[48] Nguyen, T., Byrd, E., Bentz, D., Lint, C., In situ measurement of water atthe organic coating/substrate interface, Prog. Org. Coat., 27(1–4), 181, 1996.

[49] Ji, W.-G., Hu, J.-M., Liu, L., Zhang, J.-Q., Cao, C.-N., Water uptake of epoxycoatings modified with γ-APS silane monomer, Prog. Org. Coat., 57(4), 439,2006.

106

Bibliography

[50] Posner, R., Marazita, M., Amthor, S., Roschmann, K. J., Grundmeier, G., In-fluence of interface chemistry and network density on interfacial ion transportkinetics for styrene/acrylate copolymer coated zinc and iron substrates, Corr.Sci., 52(3), 754, 2010.

[51] Parks, J., Leidheiser, H., Ionic migration through organic coatings and itsconsequences to corrosion, Ind. Eng. Chem. Prod. Rd., 25(1), 1, 1986.

[52] Walter, G., The application of impedance methods to study the effects ofwater-uptake and chloride-ion concentration on the degradation of paint films.1. attached films, Corr. Sci., 32(10), 1059, 1991.

[53] Lindqvist, S., Theory of dielectric-properties of heterogeneous substances ap-plied to water in a paint film, Corrosion, 41(2), 69, 1985.

[54] Brasher, D., Kingsbury, A., Electrical measurements in the study of immersedpaint coatings on metal .1. comparison between capacitance and gravimetricmethods of estimating water-uptake, J. Appl. Chem., 4(2), 62, 1954.

[55] Kittelberger, W., Elm, A., Diffusion of sodium chloride through various paintsystems, Ind. Eng. Chem. Prod. Rd., 44(2), 326, 1952.

[56] Kinsella, E. M., Mayne, J. E. O., Ionic conduction in polymer films i. influenceof electrolyte on resistance, Brit. Poly. J., 1(4), 173, 1969.

[57] Mayne, J. E. O., Scantlebury, J. D., Ionic conduction in polymer films: II.inhomogeneous structure of varnish films, Brit. Poly. J., 2(4), 240, 1970.

[58] Kim, H., Jeong, H., An, T. K., Park, C. E., Yong, K., Hybrid-type quantum-dot cosensitized ZnO nanowire solar cell with enhanced visible-light harvest-ing, ACS Appl. Mater. Interfaces, 5(2), 268, 2012.

[59] McCune, M., Zhang, W., Deng, Y., High efficiency dye-sensitized solarcells based on three-dimensional multilayered ZnO nanowire arrays with“caterpillar-like” structure, Nano Lett., 12(7), 3656, 2012.

[60] Moirangthem, R. S., Erbe, A., Interfacial refractive index sensing using visible-excited intrinsic zinc oxide photoluminescence coupled to whispering gallerymodes, Appl. Phys. Lett., 103(5), 051108, 2013.

[61] Wang, J., Wang, Z., Huang, B., Ma, Y., Liu, Y., Qin, X., Zhang, X., Dai,Y., Oxygen vacancy induced band-gap narrowing and enhanced visible lightphotocatalytic activity of ZnO, ACS Appl. Mater. Interfaces, 4(8), 4024, 2012.

[62] Wei, B., Zheng, K., Ji, Y., Zhang, Y., Zhang, Z., Han, X., Size-dependentbandgap modulation of ZnO nanowires by tensile strain, Nano Lett., 12(9),4595, 2012.

107

Bibliography

[63] Zhang, X. G., Corrosion and electrochemistry of zinc, Springer, 1996.

[64] Xu, S., Wang, Z. L., One-dimensional ZnO nanostructures: Solution growthand functional properties, Nano Res., 4(11), 1013, 2011.

[65] Laudise, R. A., Ballman, A. A., Hydrothermal synthesis of zinc oxide and zincsulfide, J. Phys. Chem., 64(5), 688, 1960.

[66] Li, W.-J., Shi, E.-W., Zhong, W.-Z., Yin, Z.-W., Growth mechanism andgrowth habit of oxide crystals, J. Cryst. Growth, 203(1–2), 186, 1999.

[67] Demianets, L. N., Kostomarov, D. V., Kuzmina, I. P., Pushko, S. V., Mech-anism of growth of ZnO single crystals from hydrothermal alkali solutions,Crystallogr. Rep., 47(1), S86, 2002.

[68] Roetheli, B. E., Cox, G. L., Littreal, W. B., Effect of pH on the corrosionproducts and corrosion rate of zinc in oxygenated aqueous solutions, Metalsand Alloys, 3, 73, 1932.

[69] Macias, A., Andrade, C., Corrosion of galvanized steel reinforcements in al-kaline solutions: Part 2: SEM study and identification of corrosion products,Brit. Corros. J., 22(2), 119, 1987.

[70] Klemm, S. O., Microelectrochemical characterization of Zn, ZnO and Zn-Mg alloys with online dissolution monitoring, Ph.D. thesis, Ruhr-UniversitätBochum, Fakultät für Chemie und Biochemie, Bochum, Germany, 2011.

[71] Jardy, A., Rosset, R., Wiart, R., Diphosphate coatings for protection of gal-vanized steel - quality-control by impedance measurements, J. Appl. Elec-trochem., 14(4), 537, 1984.

[72] Tegehall, Vannerberg, Tegehall, P. E., Vannerberg, N. G., Nucleation andformation of zinc phosphate conversion coating on cold-rolled steel, Corr. Sci.,32(5-6), 635, 1991.

[73] Jiang, L., Volovitch, P., Wolpers, M., Ogle, K., Activation and inhibition ofzn-al and zn-al-mg coatings on steel by nitrate in phosphoric acid solution,Corr. Sci, 60, 256, 2012.

[74] Vanooij, W., Sabata, A., Chemical and thermal-stability of phosphate layerson cold-rolled and electrogalvanized steels, J. Coat. Technol., 61(778), 51,1989.

[75] Vanooij, W., Sabata, A., Chemical-stability of phosphate conversion coatingson cold-rolled and electrogalvanized steels, Surf. Coat. Technol., 39(1-3), 667,1989.

108

Bibliography

[76] Feitknecht, W., Studies on the influence of chemical factors on the corrosionof metals, Chem. Ind., 36, 1102, 1959.

[77] Graedel, T. E., Corrosion mechanisms for zinc exposed to the atmosphere, J.Electrochem. Soc., 136(4), 193C, 1989.

[78] Sergi, G., Short, N., Page, C., Corrosion of galvanized and galvannealed steelin solutions of ph 9.0 to 14.0, Corrosion, 41(11), 618, 1985.

[79] Macías, A., Andrade, C., Corrosion of galvanized steel in dilute ca(OH)2 so-lutions (pH 11·1–12·6), Brit. Corros. J., 22(3), 162, 1987.

[80] Munier, G. B., Psota, L. A., Reagor, B. T., Russiello, B., Sinclair, J. D.,MUNIER, PSOTA, REAGOR, RUSSIELLO, SINCLAIR, Contamination ofelectronic equipment after an extended urban exposure, J. Electrochem. Soc.,127(2), 265, 1980.

[81] Zhang, H., Liu, P., Li, F., Liu, H., Wang, Y., Zhang, S., Guo, M., Cheng, H.,Zhao, H., Facile fabrication of anatase TiO2 microspheres on solid substratesand surface crystal facet transformation from 001 to 101, Chem. Eur. J.,17(21), 5949, 2011.

[82] Yu, H., Wang, D., Han, M. Y., Perpendicular branching in crystal growth of3d architecture-tuned cadmium hydroxide arrays: From oriented tripods tofaceted crystals, Adv. Mater., 20(12), 2276, 2008.

[83] Zhang, Z., Shao, X., Yu, H., Wang, Y., Han, M., Morphosynthesis and orna-mentation of 3d dendritic nanoarchitectures, Chem. Mater., 17(2), 332, 2005.

[84] Zhang, W., Wen, X., Yang, S., Berta, Y., Wang, Z., Single-crystalline scroll-type nanotube arrays of copper hydroxide synthesized at room temperature,Adv. Mater., 15(10), 822, 2003.

[85] Wen, X., Zhang, W., Yang, S., Synthesis of cu(OH)2 and CuO nanoribbonarrays on a copper surface, Langmuir , 19(14), 5898, 2003.

[86] Yu, H., Zhang, Z., Han, M., Hao, X., Zhu, F., A general low-temperature routefor large-scale fabrication of highly oriented ZnO nanorod/nanotube arrays,J. Am. Chem. Soc., 127(8), 2378, 2005.

[87] Zhang, Z., Yu, H., Shao, X., Han, M., Near-room-temperature production ofdiameter-tunable ZnO nanorod arrays through natural oxidation of zinc metal,Chem. Eur. J., 11(10), 3149, 2005.

[88] Zhang, Z., Yu, H., Wang, Y., Han, M.-Y., Aggregation-driven growth of well-oriented ZnO nanorod arrays, Nanotechnology, 17(12), 2994, 2006.

109

Bibliography

[89] Zhao, F., Li, X., Zheng, J.-G., Yang, X., Zhao, F., Wong, K. S., Wang, J.,Lin, W., Wu, M., Su, Q., ZnO pine-nanotree arrays grown from facile metalchemical corrosion and oxidation, Chem. Mat., 20(4), 1197, 2008.

[90] Wang, L., Tsan, D., Stoeber, B., Walus, K., Substrate-free fabrication of self-supporting ZnO nanowire arrays, Adv. Mater., 24(29), 3999, 2012.

[91] Vayssieres, L., Growth of arrayed nanorods and nanowires of ZnO from aque-ous solutions, Adv. Mater., 15(5), 464, 2003.

[92] Greene, L. E., Yuhas, B. D., Law, M., Zitoun, D., Yang, P., Solution-grownzinc oxide nanowires, Inorg. Chem., 45(19), 7535, 2006.

[93] de la Fuente, D., Bohm, M., Houyoux, C., Rohwerder, M., Morcillo, M., Thesettling of critical levels of soluble salts for painting, Prog. Org. Coat., 58(1),23, 2007.

[94] Taheri, P., Ghaffari, M., Flores, J. R., Hannour, F., de Wit, J. H. W., Mol,J. M. C., Terryn, H., Bonding mechanisms at buried interfaces between car-boxylic polymers and treated zinc surfaces, J. Phys. Chem. C , 117(6), 2780,2013.

[95] Santa, M., Posner, R., Grundmeier, G., Wet- and corrosive de-adhesion pro-cesses of water-borne epoxy film coated steel, J. Electrochem. Soc., 158(3),C36, 2011.

[96] McMurray, H. N., Williams, G., 2.14 - under film/coating corrosion, in B. Cot-tis, M. Graham, R. Lindsay, S. Lyon, T. Richardson, D. Scantlebury, H. Stott,editors, Shreir’s Corrosion, pages 988–1004, Elsevier, Oxford, 2010.

[97] Amirudin, A., Thierry, D., Corrosion mechanisms of phosphated zinc layerson steel as substrates for automotive coatings, Prog. Org. Coat., 28(1), 59,1996.

[98] Rangwalla, H., Dhinojwala, A., Probing hidden polymeric interfaces using IR-visible sum-frequency generation spectroscopy, J. Adhes., 80(1-2), 37, 2004.

[99] Zhang, C., Myers, J. N., Chen, Z., Elucidation of molecular structures atburied polymer interfaces and biological interfaces using sum frequency gen-eration vibrational spectroscopy, Soft Matter , 9(19), 4738, 2013.

[100] Leng, A., Streckel, H., Stratmann, M., The delamination of polymeric coatingsfrom steel. part 2: First stage of delamination, effect of type and concentra-tion of cations on delamination, chemical analysis of the interface, Corr. Sci.,41(3), 579, 1998.

[101] Grundmeier, G., Stratmann, M., Plasma polymerization – a new and promis-ing way for the corrosion protection of steel, Mater. Corros., 49(3), 150, 1998.

110

Bibliography

[102] Tran, Y., Auroy, P., Synthesis of poly(styrene sulfonate) brushes, J. Am.Chem. Soc., 123(16), 3644, 2001.

[103] Sofia, S. J., Premnath, V., Merrill, E. W., Poly(ethylene oxide) grafted tosilicon surfaces: Grafting density and protein adsorption, Macromolecules,31(15), 5059, 1998.

[104] Luzinov, I., Julthongpiput, D., Malz, H., Pionteck, J., Tsukruk, V. V.,Polystyrene layers grafted to epoxy-modified silicon surfaces, Macromolecules,33(3), 1043, 2000.

[105] Barbey, R., Lavanant, L., Paripovic, D., Schuwer, N., Sugnaux, C., Tugulu,S., Klok, H.-A., Polymer brushes via surface-initiated controlled radical poly-merization: Synthesis, characterization, properties, and applications, Chem.Rev., 109(11), 5437, 2009.

[106] Meyer, T., Spange, S., Hesse, S., Jager, C., Bellmann, C., Radical graftingpolymerization of vinylformamide with functionalized silica particles, Macro-mol. Chem. Phys., 204(4), 725, 2003.

[107] Laible, R., Hamann, K., Formation of chemically bound polymer layers onoxide surfaces and their role in colloidal stability, Adv. Colloid Interface Sci.,13(1–2), 65, 1980.

[108] Zhang, B., Hu, N., Wang, Y., Wang, Z., Wang, Y., Kong, E. S., Zhang, Y.,Poly(glycidyl methacrylates)-grafted zinc oxide nanowire by surface-initiatedatom transfer radical polymerization, Nano-Micro Lett., 2(4), 285, 2010.

[109] Prucker, O., Rühe, J., Mechanism of radical chain polymerizations initiated byazo compounds covalently bound to the surface of spherical particles, Macro-molecules, 31(3), 602, 1998.

[110] Prucker, O., Rühe, J., Synthesis of poly(styrene) monolayers attached to highsurface area silica gels through self-assembled monolayers of azo initiators,Macromolecules, 31(3), 592, 1998.

[111] Liu, H., O’Mahony, C. T., Audouin, F., Ventura, C., Morris, M., Heise, A.,Random poly(methyl methacrylate-co-styrene) brushes by ATRP to createneutral surfaces for block copolymer self-assembly, Macromol. Chem. Phys.,213(1), 108, 2012.

[112] Delattre, J. L., d, R., Agostino, Fracassi, F., Plasma-polymerized thiophenefilms for enhanced rubber–steel bonding, Appl. Surf. Sci., 252(10), 3912, 2006.

[113] Vasudev, M. C., Anderson, K. D., Bunning, T. J., Tsukruk, V. V., Naik, R. R.,Exploration of plasma-enhanced chemical vapor deposition as a method forthin-film fabrication with biological applications, ACS Appl. Mater. Interfaces,5(10), 3983, 2013.

111

Bibliography

[114] Li, P., Li, L., Wang, W., Jin, W., Liu, X., Yeung, K. W. K., Chu, P. K.,Enhanced corrosion resistance and hemocompatibility of biomedical NiTi alloyby atmospheric-pressure plasma polymerized fluorine-rich coating, Appl. Surf.Sci., 297, 109, 2014.

[115] Nazarov, A., Thierry, D., Volovitch, P., Ogle, K., An SKP and EIS investiga-tion of amine adsorption on zinc oxide surfaces, Surf. Interface Anal., 43(10),1286, 2011.

[116] Lee, I., Polymer adhesion vs substrate receptor group density, Macromolecules,33(7), 2680, 2000.

[117] Sangaj, N. S., Malshe, V. C., Permeability of polymers in protective organiccoatings, Prog. Org. Coat., 50(1), 28, 2004.

[118] Furbeth, W., Stratmann, M., Scanning kelvin probe investigations on thedelamination of polymeric coatings from metallic surfaces, Prog. Org. Coat.,39(1), 23, 2000.

[119] Nazarov, A., Prosek, T., Thierry, D., Application of EIS and SKP methodsfor the study of the zinc/polymer interface, Electrochim. Acta, 53(25), 7531,2008.

[120] Krieg, R., Rohwerder, M., Evers, S., Schuhmacher, B., Schauer-Pass, J., Ca-thodic self-healing at cut-edges: The effect of zn2+ and mg2+ ions, Corr. Sci.,65, 119, 2012.

[121] Volovitch, P., Vu, T. N., Allély, C., Abdel Aal, A., Ogle, K., Understandingcorrosion via corrosion product characterization: II. role of alloying elementsin improving the corrosion resistance of zn–al–mg coatings on steel, Corr. Sci.,53(8), 2437, 2011.

[122] Stuart, B., Infrared spectroscopy, Wiley Online Library, 2005.

[123] Sherwood, P. M., Auger and X-ray photoelectron spectroscopy, volume 1, J.Wiley and Sons Ltd, Newyork, NY, 1990.

[124] Grams, J., New Trends and Potentialities of ToF-SIMS in Surface Studies,Nova Publishers, 2007.

[125] Perkampus, H.-H., UV-VIS spectroscopy and its applications, Springer-Verlag,1992.

[126] Fujiwara, H., Spectroscopic Ellipsometry: Principles and Applications, JohnWiley & Sons, 2007.

[127] Brandrup, J., Immergut, E. H., Grulke, E. A., Abe, A., Bloch, D. R., Polymerhandbook, volume 89, Wiley New York, 1999.

112

Bibliography

[128] Stratmann, M., Streckel, H., On the atmospheric corrosion of metals whichare covered with thin electrolyte layers-i. verification of the experimental tech-nique, Corr. Sci., 30(6–7), 681, 1990.

[129] Leng, A., Streckel, H., Stratmann, M., The delamination of polymeric coatingsfrom steel. part 1: Calibration of the kelvinprobe and basic delaminationmechanism, Corr. Sci., 41(3), 547, 1998.

[130] Stern, M., Geary, A. L., Electrochemical polarization i . a theoretical analysisof the shape of polarization curves, J. Electrochem. Soc., 104(1), 56, 1957.

[131] Loveday, D., Peterson, P., Rodgers, B., Evaluation of organic coatings withelectrochemical impedance spectroscopy, JCT Coatings Tech., 8, 46, 2004.

[132] Egerton, R. F., Physical principles of electron microscopy, Springer, 2005.

[133] Meyer, E., Atomic force microscopy, Prog. Surf. Sci., 41(1), 3, 1992.

[134] Moore, D. M., Reynolds, R. C., X-ray Diffraction and the Identification andAnalysis of Clay Minerals, volume 378, Oxford university press Oxford, 1989.

[135] Kattumenu, R., Lee, C. H., Bliznyuk, V. N., Singamaneni, S., Micro-ramanspectroscopy of nanostructures, in C. S. S. R. Kumar, editor, Raman Spec-troscopy for Nanomaterials Characterization, pages 417–444, Springer BerlinHeidelberg, 2012.

[136] Niehoff, P., Ebbinghaus, P., Keil, P., Erbe, A., Monolayer formation ofoctyltrimethoxysilane and 7-octenyltrimethoxysilane on silicon (1 0 0) coveredwith native oxide, Applied Surface Science, 258(7), 3191, 2012.

[137] Posner, R., Wapner, K., Stratmann, M., Grundmeier, G., Transport processesof hydrated ions at polymer/oxide/metal interfaces: Part 1. transport at in-terfaces of polymer coated oxide covered iron and zinc substrates, Electrochim.Acta, 54(3), 891, 2009.

[138] Sorensen, P. A., Dam-Johansen, K., Weinell, C. E., Kiil, S., Cathodic delam-ination: Quantification of ionic transport rates along coating–steel interfaces,Prog. Org. Coat., 68(1–2), 70, 2010.

[139] Montoya, R., García-Galván, F. R., Jiménez-Morales, A., Galván, J. C., Acathodic delamination study of coatings with and without mechanical defects,Corr. Sci., 82, 432, 2014.

[140] Iqbal, D., Moirangthem, R. S., Bashir, A., Erbe, A., Study of polymer coatingdelamination kinetics on zinc modified with zinc oxide of different morpholo-gies, Mater. Corros., 65(4), 370, 2014.

113

Bibliography

[141] Sorensen, P. A., Dam-Johansen, K., Weinell, C. E., Kiil, S., Cathodic delam-ination of seawater-immersed anticorrosive coatings: Mapping of parametersaffecting the rate, Prog. Org. Coat., 68(4), 283, 2010.

[142] Nazarov, A., Olivier, M. G., Thierry, D., SKP and FT-IR microscopy studyof the paint corrosion de-adhesion from the surface of galvanized steel, Prog.Org. Coat., 74(2), 356, 2012.

[143] Huang, M.-W., Allely, C., Ogle, K., Orazem, M. E., A mathematical modelfor cathodic delamination of coated metal including a kinetic pH–porosityrelationship, J. Electrochem. Soc., 155(5), C279, 2008.

[144] Paliwoda-Porebska, G., Stratmann, M., Rohwerder, M., Potje-Kamloth, K.,Lu, Y., Pich, A. Z., Adler, H. J., On the development of polypyrrole coatingswith self-healing properties for iron corrosion protection, Corr. Sci., 47(12),3216, 2005.

[145] Mittal, K. L., Silanes and other coupling agents, volume 4, CRC Press, 2007.

[146] Bressy, C., Ngo, V. G., Ziarelli, F., Margaillan, A., New insights into theadsorption of 3-(trimethoxysilyl)propylmethacrylate on hydroxylated ZnOnanopowders, Langmuir , 28(6), 3290, 2012.

[147] Posthumus, W., Magusin, P., Brokken-Zijp, J. C. M., Tinnemans, A. H. A.,van der Linde, R., Surface modification of oxidic nanoparticles using 3-methacryloxypropyltrimethoxysilane, J. Colloid Interface Sci., 269(1), 109,2004.

[148] Rong, Y., Chen, H.-Z., Li, H.-Y., Wang, M., Encapsulation of titanium dioxideparticles by polystyrene via radical polymerization, Coloid. Surface. A, 253(1–3), 193, 2005.

[149] Rong, Y., Chen, H.-Z., Wu, G., Wang, M., Preparation and characterizationof titanium dioxide nanoparticle/polystyrene composites via radical polymer-ization, Mater. Chem. Phys., 91(2–3), 370, 2005.

[150] Nyquist, R. A., Interpreting infrared, Raman, and nuclear magnetic resonancespectra, Academic Press, 2001.

[151] Trasferetti, B. C., Davanzo, C. U., Bica de Moraes, M. A., LO-TO splittingsin plasma-deposited siloxane films, J. Phys. Chem. B, 107(39), 10699, 2003.

[152] Bexell, U., Olsson, M., Characterization of a non-organofunctional silane filmdeposited on al, zn and al-43.4zn-1.6si alloy-coated steel, Surf. Interface Anal.,31(3), 212, 2001.

[153] Zou, H., Wu, S., Shen, J., Polymer/silica nanocomposites: Preparation, char-acterization, properties, and applications, Chem. Rev., 108(9), 3893, 2008.

114

Bibliography

[154] Bhattacharya, A., Rawlins, J. W., Ray, P., Polymer grafting and crosslinking,Wiley Online Library, 2009.

[155] Grundmeier, G., Schmidt, W., Stratmann, M., Corrosion protection by organiccoatings: electrochemical mechanism and novel methods of investigation, Elec-trochim. Acta, 45(15–16), 2515, 2000.

[156] Yasuda, H., Yu, Q. S., Chen, M., Interfacial factors in corrosion protection:an EIS study of model systems, Prog. Org. Coat., 41(4), 273, 2001.

[157] Plueddemann, E., Pape, P., Bank, H., New coupling agents for improvedcorrosion-resistant composites, Polym.-Plast. Technol. Eng., 25(3-4), 223,1986.

[158] Watts, J. F., Rattana, A., Abel, M.-L., Interfacial chemistry of adhesives onhydrated aluminium and hydrated aluminium treated with an organosilane,Surf. Interface Anal., 36(11), 1449, 2004.

[159] Abel, M.-L., Watts, J. F., Examination of the interface of a model adhesivejoint by surface analysis: a study by XPS and ToF-SIMS, Surf. Interface Anal.,41(6), 508, 2009.

[160] Pourbaix, M., Atlas of electrochemical equilibria in aqueous solutions, 1974.

[161] Nayak, S., Biedermann, P. U., Stratmann, M., Erbe, A., A mechanistic studyof the electrochemical oxygen reduction on the model semiconductor n-ge(100)by ATR-IR and DFT, Phys. Chem. Chem. Phys., 15(16), 5771, 2013.

[162] Nayak, S., Biedermann, P. U., Stratmann, M., Erbe, A., In situ infrared spec-troscopic investigation of intermediates in the electrochemical oxygen reduc-tion on n-ge(100) in alkaline perchlorate and chloride electrolyte, Electrochim.Acta, 106, 472, 2013.

[163] Sorensen, P. A., Weinell, C. E., Dam-Johansen, K., Kiil, S., Reduction ofcathodic delamination rates of anticorrosive coatings using free radical scav-engers, J Coat Technol Res, 7(6), 773, 2010.

[164] Haagen, H., Funke, W., Prediction of corrosion protective properties of paintfilms by permeability data, J. Oil Colour Chem. As., 58(10), 359, 1975.

[165] Comyn, J., Polymer Permeability, Springer Science & Business Media, 1985.

[166] Ashley, R. J., Permeability and plastics packaging, in Polymer permeability,pages 269–308, Springer, 1985.

[167] Taheri, P., Wielant, T., Hauffman, T., Flores, J. R., Hannour, F., de Wit, J.H. W., Mol, J. M. C., Terryn, H., A comparison of the interfacial bondingproperties of carboxylic acid functional groups on zinc and iron substrates,Electrochim. Acta, 56(4), 1904, 2011.

115

Bibliography

[168] van den Brand, J., Blajiev, O., Beentjes, P. C. J., Terryn, H., de Wit, J. H. W.,Interaction of anhydride and carboxylic acid compounds with aluminum oxidesurfaces studied using infrared reflection absorption spectroscopy, Langmuir ,20(15), 6308, 2004.

[169] Taheri, P., Hauffman, T., Mol, J. M. C., Flores, J. R., Hannour, F., de Wit,J. H. W., Terryn, H., Molecular interactions of electroadsorbed carboxylicacid and succinic anhydride monomers on zinc surfaces, J. Phys. Chem. C ,115(34), 17054, 2011.

[170] Kataby, G., Cojocaru, M., Prozorov, R., Gedanken, A., Coating carboxylicacids on amorphous iron nanoparticles, Langmuir , 15(5), 1703, 1999.

[171] Thissen, P., Wielant, J., Koeyer, M., Toews, S., Grundmeier, G., Formationand stability of organophosphonic acid monolayers on ZnAl alloy coatings,Surf. Coat. Technol., 204(21-22), 3578, 2010.

[172] Noguera, C., Theoretical investigation of acid-base properties of oxide surfaces,J. Adhes., 57(1-4), 91, 1996.

[173] Zuo, J., Erbe, A., Optical and electronic properties of native zinc oxide filmson polycrystalline zn, Phys. Chem. Chem. Phys., 12(37), 11467, 2010.

[174] Chen, Y., Schneider, P., Liu, B.-J., Borodin, S., Ren, B., Erbe, A., Electronicstructure and morphology of dark oxides on zinc generated by electrochemicaltreatment, Phys. Chem. Chem. Phys., 15(24), 9812, 2013.

[175] Chen, Y., Schneider, P., Erbe, A., Investigation of native oxide growth on zincin different atmospheres by spectroscopic ellipsometry, Phys. Status Solidi A-Appl. Mat., 209(5), 846, 2012.

[176] Chen, Y., Erbe, A., In situ spectroscopic ellipsometry during electrochemicaltreatment of zinc in alkaline carbonate electrolyte, Surf. Sci., 607, 39, 2013.

[177] Park, W. I., Yi, G.-C., Electroluminescence in n-ZnO nanorod arrays verticallygrown on p-GaN, Adv. Mater., 16(1), 87, 2004.

[178] Law, M., Greene, L. E., Johnson, J. C., Saykally, R., Yang, P. D., Nanowiredye-sensitized solar cells, Nat. Mater., 4(6), 455, 2005.

[179] Baruah, S., Rafique, R. F., Dutta, J., Visible light photocatalysis by tailoringcrystal defects in zinc oxide nanostructures, Nano, 3(5), 399, 2008.

[180] Wei, T.-Y., Yeh, P.-H., Lu, S.-Y., Lin-Wang, Z., Gigantic enhancement insensitivity using schottky contacted nanowire nanosensor, J. Am. Chem. Soc.,131(48), 17690, 2009.

116

Bibliography

[181] Grassian, V. H., When size really matters: Size-dependent properties andsurface chemistry of metal and metal oxide nanoparticles in gas and liquidphase environments, J. Phys. Chem. C , 112(47), 18303, 2008.

[182] Mudunkotuwa, I. A., Rupasinghe, T., Wu, C.-M., Grassian, V. H., Dissolu-tion of ZnO nanoparticles at circumneutral pH: A study of size effects in thepresence and absence of citric acid, Langmuir , 28(1), 396, 2012.

[183] Pan, Z. W., Dai, Z. R., Wang, Z. L., Nanobelts of semiconducting oxides,Science, 291(5510), 1947, 2001.

[184] Moirangthem, R. S., Cheng, P.-J., Chien, P. C.-H., Ngo, B. T. H., Chang,S.-W., Tien, C.-H., Chang, Y.-C., Optical cavity modes of a single crystallinezinc oxide microsphere, Opt. Express, 21(3), 3010, 2013.

[185] Wang, D., Song, C., Hue, Z., Shape-controlled synthesis of ZnO architectures,Cryst. Res. Technol., 43(1), 55, 2008.

[186] Shang, T.-M., Sun, J.-H., Zhou, Q.-F., Guan, M.-Y., Controlled synthesisof various morphologies of nanostructured zinc oxide: flower, nanoplate, andurchin, Cryst. Res. Technol., 42(10), 1002, 2007.

[187] Tseng, Y.-H., Liu, M.-H., Kuo, Y.-W., Chen, P., Chen, C.-T., Chen, Y.-F., Mou, C.-Y., Biomimetic ZnO plate twin-crystals periodical arrays, Chem.Commun., 48(26), 3215, 2012.

[188] Yoo, J. D., Volovitch, P., Abdel Aal, A., Allely, C., Ogle, K., The effect of anartificially synthesized simonkolleite layer on the corrosion of electrogalvanizedsteel, Corr. Sci., 70, 1, 2013.

[189] Valtiner, M., Torrelles, X., Pareek, A., Borodin, S., Gies, H., Grundmeier,G., In situ study of the polar ZnO(0001)-zn surface in alkaline electrolytes, J.Phys. Chem. C , 114(36), 15440, 2010.

[190] Tseng, Y.-H., Lin, H.-Y., Liu, M.-H., Chen, Y.-F., Mou, C.-Y., Biomimeticsynthesis of nacrelike faceted mesocrystals of ZnO-gelatin composite, J. Phys.Chem. C , 113(42), 18053, 2009.

[191] Kuo, C.-L., Kuo, T.-J., Huang, M. H., Hydrothermal synthesis of ZnO micro-spheres and hexagonal microrods with sheetlike and platelike nanostructures,J. Phys. Chem. B, 109(43), 20115, 2005.

[192] Woell, C., The chemistry and physics of zinc oxide surfaces, Prog. Surf. Sci.,82(2–3), 55, 2007.

[193] Conway, B., Kannangara, D., Zinc oxidation and redeposition processes inaqueous alkali and carbonate solutions .2. distinction between dissolution andoxide film formation processes, J. Electrochem. Soc., 134(4), 906, 1987.

117

Bibliography

[194] Mokaddem, M., Volovitch, P., Ogle, K., The anodic dissolution of zinc andzinc alloys in alkaline solution i oxide formation on electrogalvanized steel,Electrochim. Acta, 55(27), 7867, 2010.

[195] Nicholas, N. J., Ducker, W., Franks, G. V., Differential etching of ZnO nativeplanes under basic conditions, Langmuir , 28(13), 5633, 2012.

[196] Tong, H., Ouyang, S., Bi, Y., Umezawa, N., Oshikiri, M., Ye, J., Nano-photocatalytic materials: Possibilities and challenges, Adv. Mater., 24(2), 229,2012.

[197] Kisch, H., Semiconductor photocatalysis-mechanistic and synthetic aspects,Angew. Chem. Int. Ed., 52(3), 812, 2013.

[198] Habisreutinger, S. N., Schmidt-Mende, L., Stolarczyk, J. K., Photocatalyticreduction of CO2 on TiO2 and other semiconductors, Angew. Chem. Int. Ed.,52(29), 7372, 2013.

[199] Paramasivam, I., Jha, H., Liu, N., Schmuki, P., A review of photocatalysisusing self-organized TiO2 nanotubes and other ordered oxide nanostructures,Small, 8(20), 3073, 2012.

[200] Liu, N., Chen, X., Zhang, J., Schwank, J. W., A review on TiO2-based nan-otubes synthesized via hydrothermal method: Formation mechanism, struc-ture modification, and photocatalytic applications, Catal. Today, 225, 34,2014.

[201] Ong, W.-J., Tan, L.-L., Chai, S.-P., Yong, S.-T., Mohamed, A. R., Facet- de-pendent photocatalytic properties of TiO2- based composites for energy con-version and environmental remediation, ChemSusChem, 7(3), 690, 2014.

[202] Guo, W., Zhang, F., Lin, C., Wang, Z. L., Direct growth of TiO2 nanosheet ar-rays on carbon fibers for highly efficient photocatalytic degradation of methylorange, Adv. Mater., 24(35), 4761, 2012.

[203] Yang, T.-H., Huang, L.-D., Harn, Y.-W., Lin, C.-C., Chang, J.-K., Wu, C.-I., Wu, J.-M., High density unaggregated au nanoparticles on ZnO nanorodarrays function as efficient and recyclable photocatalysts for environmentalpurification, Small, 9(18), 3169, 2013.

[204] Lyu, J., Zhu, L., Burda, C., Considerations to improve adsorption and photo-catalysis of low concentration air pollutants on TiO2, Catal. Today, 225, 24,2014.

[205] Zhang, P., Li, B., Zhao, Z., Yu, C., Hu, C., Wu, S., Qiu, J., Furfural-induced hydrothermal synthesis of ZnO@c gemel hexagonal microrods withenhanced photocatalytic activity and stability, ACS Appl. Mater. Interfaces,6(11), 8560, 2014.

118

Bibliography

[206] Etacheri, V., Roshan, R., Kumar, V., Mg-doped ZnO nanoparticles for efficientsunlight-driven photocatalysis, ACS Appl. Mater. Interfaces, 4(5), 2717, 2012.

[207] Lipovsky, A., Tzitrinovich, Z., Friedmann, H., Applerot, G., Gedanken, A.,Lubart, R., EPR study of visible light-induced ROS generation by nanoparti-cles of ZnO, J. Phys. Chem. C , 113(36), 15997, 2009.

[208] Deng, Q., Duan, X., Ng, D. H. L., Tang, H., Yang, Y., Kong, M., Wu, Z., Cai,W., Wang, G., Ag nanoparticle decorated nanoporous ZnO microrods andtheir enhanced photocatalytic activities, ACS Appl. Mater. Interfaces, 4(11),6030, 2012.

[209] Djurisic, A. B., Leung, Y. H., Optical properties of ZnO nanostructures, Small,2(8–9), 944, 2006.

[210] Wu, J.-J., Liu, S.-C., Catalyst-free growth and characterization of ZnOnanorods, J. Phys. Chem. B, 106(37), 9546, 2002.

[211] Hong, S., Yeo, J., Manorotkul, W., Kang, H. W., Lee, J., Han, S., Rho, Y.,Suh, Y. D., Sung, H. J., Ko, S. H., Digital selective growth of a ZnO nanowirearray by large scale laser decomposition of zinc acetate, Nanoscale, 5(9), 3698,2013.

[212] Yang, F., Liu, W.-H., Wang, X.-W., Zheng, J., Shi, R.-Y., Zhao, H., Yang,H.-Q., Controllable low temperature vapor-solid growth and hexagonal diskenhanced field emission property of ZnO nanorod arrays and hexagonal nan-odisk networks, ACS Appl. Mater. Interfaces, 4(8), 3852, 2012.

[213] Willander, M., Nur, O., Bano, N., Sultana, K., Zinc oxide nanorod-basedheterostructures on solid and soft substrates for white-light-emitting diodeapplications, New J. Phys., 11(12), 125020, 2009.

[214] Choi, J. H., Tabata, H., Kawai, T., Initial preferred growth in zinc oxide thinfilms on si and amorphous substrates by a pulsed laser deposition, J. Cryst.Growth, 226(4), 493, 2001.

[215] Reimer, T., Paulowicz, I., Röder, R., Kaps, S., Lupan, O., Chemnitz, S., Be-necke, W., Ronning, C., Adelung, R., Mishra, Y. K., Single step integration ofZnO nano- and microneedles in si trenches by novel flame transport approach:Whispering gallery modes and photocatalytic properties, ACS Appl. Mater.Interfaces, 6(10), 7806, 2014.

[216] Li, Y., Meng, G. W., Zhang, L. D., Phillipp, F., Ordered semiconductor ZnOnanowire arrays and their photoluminescence properties, Applied Physics Let-ters, 76(15), 2011, 2000.

119

Bibliography

[217] Rai, P., Kwak, W.-K., Yu, Y.-T., Solvothermal synthesis of ZnO nanos-tructures and their morphology-dependent gas-sensing properties, ACS Appl.Mater. Interfaces, 5(8), 3026, 2013.

[218] Taghavi, M., Mattoli, V., Mazzolai, B., Filippeschi, C., Beccai, L., Synthe-sizing tubular and trapezoidal shaped ZnO nanowires by an aqueous solutionmethod, Nanoscale, 5(8), 3505, 2013.

[219] Pradel, K. C., Wu, W., Zhou, Y., Wen, X., Ding, Y., Wang, Z. L., Piezotroniceffect in solution-grown p-type ZnO nanowires and films, Nano Lett., 13(6),2647, 2013.

[220] Kim, S. B., Lee, W. W., Yi, J., Park, W. I., Kim, J.-S., Nichols, W. T., Simple,large-scale patterning of hydrophobic ZnO nanorod arrays, ACS Appl. Mater.Interfaces, 4(8), 3910, 2012.

[221] Park, S., An, S., Ko, H., Jin, C., Lee, C., Synthesis of nanograined ZnOnanowires and their enhanced gas sensing properties, ACS Appl. Mater. In-terfaces, 4(7), 3650, 2012.

[222] Pan, K., Dong, Y., Zhou, W., Pan, Q., Xie, Y., Xie, T., Tian, G., Wang, G.,Facile fabrication of hierarchical TiO2 nanobelt/ZnO nanorod heterogeneousnanostructure: An efficient photoanode for water splitting, ACS Appl. Mater.Interfaces, 5(17), 8314, 2013.

[223] Manna, J., Begum, G., Kumar, K. P., Misra, S., Rana, R. K., Enabling an-tibacterial coating via bioinspired mineralization of nanostructured ZnO onfabrics under mild conditions, ACS Appl. Mater. Interfaces, 5(10), 4457, 2013.

[224] Yu, H.-D., Zhang, Z., Han, M.-Y., Metal corrosion for nanofabrication, Small,8(17), 2621, 2012.

[225] Park, W. I., Yi, G.-C., Kim, J.-W., Park, S.-M., Schottky nanocontacts onZnO nanorod arrays, Applied Physics Letters, 82(24), 4358, 2003.

[226] Calleja, J., Cardona, M., Resonant raman-scattering in zno, Phys. Rev. B,16(8), 3753, 1977.

[227] Ozgur, U., Alivov, Y. I., Liu, C., Teke, A., Reshchikov, M. A., Dogan, S.,Avrutin, V., Cho, S.-J., Morkoc, H., A comprehensive review of ZnO materialsand devices, J. Appl. Phys., 98(4), 041301, 2005.

[228] Damen, T. C., Porto, S. P. S., Tell, B., Raman effect in zinc oxide, Phys. Rev.,142(2), 570, 1966.

[229] Arguello, C. A., Rousseau, D. L., Porto, S. P. S., First-order raman effect inwurtzite-type crystals, Phys. Rev., 181(3), 1351, 1969.

120

Bibliography

[230] Tay, Y. Y., Tan, T. T., Liang, M. H., Boey, F., Li, S., Nature of quasi-LOphonon in ZnO, Applied Physics Letters, 93(11), 111903, 2008.

[231] Wang, M., Zhou, Y., Zhang, Y., Kim, E. J., Hahn, S. H., Seong, S. G.,Near-infrared photoluminescence from ZnO, Applied Physics Letters, 100(10),101906, 2012.

[232] Dag, S., Wang, S., Wang, L.-W., Large surface dipole moments in ZnOnanorods, Nano Lett., 11(6), 2348, 2011.

[233] Yamabi, S., Imai, H., Growth conditions for wurtzite zinc oxide films in aque-ous solutions, J. Mater. Chem., 12(12), 3773, 2002.

[234] Peulon, S., Lincot, D., Mechanistic study of cathodic electrodeposition of zincoxide and zinc hydroxychloride films from oxygenated aqueous zinc chloridesolutions, J. Electrochem. Soc., 145(3), 864, 1998.

[235] Xu, C., Shin, P., Cao, L., Gao, D., Preferential growth of long ZnO nanowirearray and its application in dye-sensitized solar cells, J. Phys. Chem. C ,114(1), 125, 2009.

[236] Weintraub, B., Zhou, Z., Li, Y., Deng, Y., Solution synthesis of one-dimensional ZnO nanomaterials and their applications, Nanoscale, 2(9), 1573,2010.

[237] Guo, M. Y., Ng, A. M. C., Liu, F., Djurisic, A. B., Chan, W. K., Su, H.,Wong, K. S., Effect of native defects on photocatalytic properties of ZnO, J.Phys. Chem. C , 115(22), 11095, 2011.

[238] Vanheusden, K., Warren, W. L., Seager, C. H., Tallant, D. R., Voigt, J. A.,Gnade, B. E., Mechanisms behind green photoluminescence in ZnO phosphorpowders, J. Appl. Phys., 79(10), 7983, 1996.

[239] Wang, J., Liu, P., Fu, X., Li, Z., Han, W., Wang, X., Relationship betweenoxygen defects and the photocatalytic property of ZnO nanocrystals in nafionmembranes, Langmuir , 25(2), 1218, 2008.

[240] Zhao, D., Zhang, C., Zhang, X., Cai, L., Zhang, X., Luan, P., Zhang, Q.,Tu, M., Wang, Y., Zhou, W., Li, Z., Xie, S., Substrate-induced effects onthe optical properties of individual ZnO nanorods with different diameters,Nanoscale, 6(1), 483, 2014.

[241] Pierce, B. J., Hengehold, R., Depth-resolved cathodoluminescence of ion-implanted layers in zinc oxide, J. Appl. Phys., 47(2), 644, 1976.

[242] Xu, S., Guo, W., Du, S., Loy, M. M. T., Wang, N., Piezotronic effects on theoptical properties of ZnO nanowires, Nano Lett., 12(11), 5802, 2012.

121

Bibliography

[243] Todorova, M., Neugebauer, J., Connecting semiconductor defect chemistrywith electrochemistry: Impact of the electrolyte on the formation and concen-tration of point defects in ZnO, Surf. Sci., 2014.

[244] Lu, F., Cai, W., Zhang, Y., ZnO hierarchical micro/nanoarchitectures:Solvothermal synthesis and structurally enhanced photocatalytic performance,Adv. Funct. Mater., 18(7), 1047, 2008.

[245] Sapkota, A., Anceno, A. J., Baruah, S., Shipin, O. V., Dutta, J., Zinc oxidenanorod mediated visible light photoinactivation of model microbes in water,Nanotechnology, 22(21), 215703, 2011.

[246] Polarz, S., Strunk, J., Ischenko, V., van den Berg, M. W. E., Hinrichsen,O., Muhler, M., Driess, M., On the role of oxygen defects in the catalyticperformance of zinc oxide, Angew. Chem. Int. Ed., 45(18), 2965, 2006.

[247] Hu, Y. H., A highly efficient PhotocatalystHydrogenated black TiO2 for thephotocatalytic splitting of water, Angew. Chem. Int. Ed., 51(50), 12410, 2012.

[248] Lao, C. S., Park, M.-C., Kuang, Q., Deng, Y., Sood, A. K., Polla, D. L.,Wang, Z. L., Giant enhancement in UV response of ZnO nanobelts by polymersurface-functionalization, J. Am. Chem. Soc., 129(40), 12096, 2007.

[249] Haick, H., Hurley, P. T., Hochbaum, A. I., Yang, P., Lewis, N. S., Electri-cal characteristics and chemical stability of non-oxidized, methyl-terminatedsilicon nanowires, J. Am. Chem. Soc., 128(28), 8990, 2006.

[250] Yan, R., Gargas, D., Yang, P., Nanowire photonics, Nat. Photonics, 3(10),569, 2009.

[251] Dasgupta, N. P., Sun, J., Liu, C., Brittman, S., Andrews, S. C., Lim, J., Gao,H., Yan, R., Yang, P., 25th anniversary article: Semiconductor nanowiressynthesis, characterization, and applications, Adv. Mater., 26(14), 2137, 2014.

[252] Singamaneni, S., Gupta, M., Yang, R., Tomczak, M. M., Naik, R. R., Wang,Z. L., Tsukruk, V. V., Nondestructive in situ identification of crystal orienta-tion of anisotropic ZnO nanostructures, ACS Nano, 3(9), 2593, 2009.

[253] Morales, A. M., Lieber, C. M., A laser ablation method for the synthesis ofcrystalline semiconductor nanowires, Science, 279(5348), 208, 1998.

[254] Huang, M. H., Wu, Y. Y., Feick, H., Tran, N., Weber, E., Yang, P. D., Cat-alytic growth of zinc oxide nanowires by vapor transport, Adv. Mater., 13(2),113, 2001.

[255] Yeo, J., Hong, S., Wanit, M., Kang, H. W., Lee, D., Grigoropoulos, C. P.,Sung, H. J., Ko, S. H., Rapid, one-step, digital selective growth of ZnO

122

Bibliography

nanowires on 3d structures using laser induced hydrothermal growth, Adv.Funct. Mater., 23(26), 3316, 2013.

[256] Djurisic, A. B., Chen, X., Leung, Y. H., Ng, A. M. C., ZnO nanostructures:growth, properties and applications, J. Mater. Chem., 22(14), 6526, 2012.

[257] Ding, H., Pan, N., Ma, C., Wu, Y., Li, J., Cai, H., Zhang, K., Zhang, G.,Ren, W., Li, J., Luo, Y., Wang, X., Hou, J. G., Maximizing integrated opticaland electrical properties of a single ZnO nanowire through native interfacialdoping, Adv. Mater., 26(19), 3035, 2014.

[258] Viswanatha, R., Santra, P. K., Dasgupta, C., Sarma, D. D., Growth mecha-nism of nanocrystals in solution: ZnO, a case study, Phys. Rev. Lett., 98(25),255501, 2007.

[259] Wong, E. M., Bonevich, J. E., Searson, P. C., Growth kinetics of nanocrys-talline ZnO particles from colloidal suspensions, J. Phys. Chem. B, 102(40),7770, 1998.

[260] In, J. B., Ryu, S.-G., Lee, D., Ahn, S., Zheng, A. C., Hwang, D. J. S., Grig-oropoulos, C. P., Laser crystallization and localized growth of nanomaterialsfor solar applications, Proc. of SPIE , 8826, 2013.

[261] Gandhi, A. C., Hung, H.-J., Shih, P.-H., Cheng, C.-L., Ma, Y.-R., Wu, S. Y.,In situ confocal raman mapping study of a single ti-assisted ZnO nanowire,Nanoscale Res. Lett., 5(3), 581, 2010.

[262] Cai, W. B., Scherson, D. A., In situ raman spectroscopy of zinc electrodes inalkaline solutions, J. Electrochem. Soc., 150(5), B217, 2003.

[263] Zhang, X. Y., Dai, J. Y., Ong, H. C., Wang, N., Chan, H. L. W., Choy, C. L.,Hydrothermal synthesis of oriented ZnO nanobelts and their temperature de-pendent photoluminescence, Chem. Phys. Lett., 393(1-3), 17, 2004.

[264] Tang, Q., Zhou, W. J., Shen, J. M., Zhang, W., Kong, L. F., Qian, Y. T., Atemplate-free aqueous route to ZnO nanorod arrays with high optical property,Chem. Commun., 6, 712, 2004.

[265] Abd El Aal, E. E., Effect of cl− anions on zinc passivity in berate solution,Corr. Sci., 42(1), 1, 2000.

[266] Hurlen, T., Corrosion of zinc - effect of ph, Acta Chem. Scand., 16(6), 1346,1962.

123

Appendix

125

List of Figures

1.1 Schematic representation of steps in classical free radical polymerization 31.2 Schematic representation of covalent modification of oxide surfaces . . 51.3 Schematic representation of cathodic delamination of polymer coated

zinc substrate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71.4 Potential-pH equilibrium diagram (pourbaix) for zinc . . . . . . . . . 11

2.1 Schematic representing the route to delamination studies . . . . . . . 162.2 Schematic representation of functionalization of zinc . . . . . . . . . . 182.3 Schematic representation of synthesis of PMMA and PS polymers . . 192.4 Layout of flow chart depicting the schematic of thesis chapters. . . . . 21

3.1 Schematic representation of in situ photocatalysis experiment. . . . . 32

4.1 Scheme of the surface modification steps carried out in this chapter. . 364.2 IR and XP spectra of VTS modified zinc . . . . . . . . . . . . . . . . 374.3 TOF-SIMS spectra of VTS modified zinc . . . . . . . . . . . . . . . . 384.4 XPS survey spectrum on chemisorbed PMMA coating . . . . . . . . . 384.5 IR and C 1s XP spectra of silane bound PMMA coating . . . . . . . 394.6 SKP delamination profile of silane bound PMMA and physisorbed

PVB coated zinc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414.7 Comparison of the delamination rates for silane bound PMMA coatings 424.8 Tafel plots for silane bound PMMA coatings . . . . . . . . . . . . . . 444.9 Bode plots of silane bound PMMA coatings . . . . . . . . . . . . . . 45

5.1 Synthetic route for the synthesis of silane bonded PS coatings . . . . 505.2 IR spectra silane bound PS coatings . . . . . . . . . . . . . . . . . . . 515.3 C 1s region of the XP spectra of PS modified zinc . . . . . . . . . . . 525.4 SKP delamination profile of silane bound PS and physisorbed PVB

coating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545.5 Comparison of the delamination rate of PS modified zinc . . . . . . . 55

6.1 Scheme of the surface modification steps carried out in this chapter. . 586.2 (I) IR and (II) C 1s XP spectra of ester modified zinc . . . . . . . . . 596.3 TOF-SIMS spectrum of acryloyl chloride modified zinc . . . . . . . . 606.4 (I)IR and (II) C 1s XP spectra of ester bound PMMA coatings . . . . 61

127

List of Figures

6.5 SKP delamination profile of ester bound PMMA and physisorbedPVB coating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

6.6 Comparison of the delamination rate for ester bound PMMA coatings 64

7.1 Comparison of delamination rates of chemisorbed linear PS and PMMApolymer coatings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

7.2 Comparison of delamination rates of chemisorbed 50% EDA cross-linked PS and PMMA polymer coatings. . . . . . . . . . . . . . . . . 67

8.1 SEM micrograph and XRD pattern of grown ZnO structures on zinc . 718.2 SKP delamination profiles of zinc with and without ZnO nanostructures 728.3 Comparison of delamiantion rates . . . . . . . . . . . . . . . . . . . . 738.4 SEM micrographs after the delamination . . . . . . . . . . . . . . . . 74

9.1 SEM image of ZnO nanorod arrays . . . . . . . . . . . . . . . . . . . 809.2 XPS, EDX and Raman spectra of ZnO nanorod arrays . . . . . . . . 819.3 SEM micrograph of ZnO nanord grown with different electrolyte con-

centration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 829.4 Schematic representation of corrosion-driven growth of ZnO nanorods

arrays. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 839.5 (a) PL spectrum of ZnO nanorod arrays, (b) reflection absorption

spectrum of ZnO nanorod arrays, (c) UV/visible spectrum of MOsolution after different irradiation times, (d) molar concentration ofMO as function of time under illumination . . . . . . . . . . . . . . . 84

10.1 Experimental setup for in situ Raman measurement of ZnO nanorods 9010.2 In situ monitoring of ZnO nanorods growth with 1 M KCl . . . . . . 9110.3 Raman mapping of ZnO nanorod cluster . . . . . . . . . . . . . . . . 9310.4 SEM micrographs with their corresponding Raman spectra . . . . . . 9410.5 In situ monitoring of ZnO nanorods growth with 3 M KCl . . . . . . 9610.6 SEM micrographs and their corresponding Raman spectra measured

ex situ after different delamination times . . . . . . . . . . . . . . . . 9710.7 UV/VIS reflection absorption spectrum of ZnO nanorod arrays . . . . 98

128

List of Tables

4.1 Thickness of silane bound PMMA coatings . . . . . . . . . . . . . . . 404.2 Delamination rates and standard deviations of silane bound PMMA

coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434.3 Corrosion current densities of silane bound PMMA coatings . . . . . 44

5.1 IR stretching vibration position of PS coatings . . . . . . . . . . . . . 515.2 Thickness of PS coatings . . . . . . . . . . . . . . . . . . . . . . . . . 535.3 Comparison of delamination rate of PS coatings . . . . . . . . . . . . 55

6.1 IR stretching vibration position of PMMA coatings . . . . . . . . . . 626.2 Thickness of ester bound PMMA coatings . . . . . . . . . . . . . . . 626.3 Delamination rate and standard deviation of ester bound PMMA

coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

7.1 Delamination rates and standard deviations for the different samples 68

129

List of Tables

3 AbbreviationsAbbreviation Explanation

AC Alternating currentAFM Atomic force microscopyAIBN AzobisisobutyronitrilePVB poly(vinyl butyral)EDA Ethylene glycol diacrylateEDX Energy dispersive X-ray spectroscopyEIS Electrochemical impedance spectroscopyFTIR Fourier transform infrared spectroscopyGIXRD Grazing incidence X-ray diffractionHDA Hexanediol diacylateHMTA hexamethylenetetramineIR InfraredM MetalMMA Methyl methacrylatePMMA Poly(methyl methacrylate)MO Methyl orangePS Poly(styrene)SEM Scanning electron microscopeSKP Scanning Kelvin probeSTM Scanning Tunneling MicroscopyTHF TetrahydrofuranTEM Transmission electron microscopeTOF-SIMS Time-of-Flight secondary ion mass spectroscopyVTS Vinyltrimethoxy silaneUV-Vis Ultraviolet-VisibleXPS X-ray photoelectron spectroscopyXRD X-ray diffraction

130

4. List of publications till November 2014

4 List of publications till November 20141. D. Iqbal, J. Rechmann, A. Sarfraz, A. Altin, G. Genchev, A. Erbe, Synthesis

of ultrathin poly(methyl methacrylate) model coatings bound via organosi-lanes to zinc and investigation of their delamination kinetics, ACS AppliedMaterials & Interfaces, 2014, doi:10.1021/am504992r.

2. D. Iqbal, A. Kostka, A. Bashir, A. Sarfraz, Y. Chen, A.D. Wieck, A. Erbe,Sequential growth of zinc oxide nanorod arrays at room temperature via acorrosion process: application in visible light photocatalysis, ACS AppliedMaterials & Interfaces, 2014, doi:10.1021/am504299v.

3. D. Iqbal, R.S. Moirangthem, A. Bashir, A. Erbe, Study of polymer coatingdelamination kinetics on zinc modified with zinc oxide of different morpholo-gies, Materials & Corrosion, 65(4):370-375, 2014.

4. C.A. Laska, M. Auinger, P.U. Biedermann, D. Iqbal, N. Laska, J.D. Strycker,K.J.J. Mayrhofer, Effect of hydrogen carbonate and chloride on zinc corrosioninvestigated by a scanning flow cell system, Corrosion Science, submitted,2014.

5. W. Azzam, D. Iqbal, M. Ebqai, A. Bashir, An unexpectedly formation ofdense phases along with temperature-induced: self-assembled terphenylthio-late monolayers on Au(111), Langmiur, Submitted, 2014.

6. D. Iqbal, M.H Samiullah, Photo-Responsive Shape-Memory and Shape-Chan-ging Liquid-Crystal Polymer Networks, Materials. 2013; 6(1):116-142.

7. D. Iqbal, C. Melchert, M. Behl, A. Lendlein, S. Beuermann, Modeling of FreeRadical Polymerization of Azobenzene-based Linear Polymers, Mater. Res.Soc. Symp. Proc, 1403, 131-136, 2012.

5 Conference oral presentations & posters1. D. Iqbal, C. Melchert, M. Behl, A. Lendlein, S. Beuermann, Modeling of

Free Radical Polymerization of Azobenzene-based Linear Polymers, Mater.Res. Soc. Symp. Proc, 27-02 November 2012, Boston, USA.

2. D. Iqbal, A. Erbe, Simplified approach to synthesis of ZnO nanorods, 10thMaterials day 8-9 November 2012, Ruhr university Bochum, Bochum, Ger-many.

3. D. Iqbal, A. Sarfraz, A. Erbe, Ultrathin polymer coatings for delaminationstudies, Eurocorr, 1-5 September 2013, Estoril, Portugal.

131

List of Tables

4. D. Iqbal, A. Erbe, Facile route to grow crystalline ZnO nanorod arrays viacorrosion and their photocatalytic activity under visible Light, 224th ECSMeeting, 27-1 November 2013, San francisco, USA.

5. D. Iqbal, J. Rechmann, A. Sarfraz, A. Erbe, Ultrathin polymer coatings forcorrosion protection, Gordon Aqueous corrosion, 12-18 July 2014, New london,USA.

6. D. Iqbal, A. Sarfraz, A. Erbe, In-situ Raman studies on growth of ZnOnanowires, 24th International conference on Raman spectroscopy, 10-15 Au-gust 2014, Jena, Germany.

7. D. Iqbal, R.S. Moirangthem, A. Bashir, A. Erbe, Study of polymer coatingdelmination kinetics on zinc modified with zinc oxide of different morphologies,Eurocorr, 8-12 September 2014, Pisa, Italy.

8. A. Bashir, D. Iqbal, M. Rohwerder, A. Erbe, Delamination kinetics of chemi-sorbed polymer coatings on nanostructured metal surfaces, Eurocorr, 8-12September 2014, Pisa, Italy.

132

CVName: Danish IqbalDate of Birth: 21.07.1984Place of Birth: Muscat, Oman

University

2011–till now PhD student, Max-Planck Institut für Eisenforschung,GmbH, Faculty of Chemistry and Biochemistry, Ruhr Uni-versität Bochum, Germany

2009–2011 Master in Polymer Science, TU, HU, FU Berlin & Universityof Potsdam, Germany

2002–2002 Bachelors of Engineering in Polymer Technology, HamdardUniversity, Karachi, Pakistan (First Position)

2003–2004 Diploma In Polymer Technology, London Metropolitan Uni-versity, UK

2002–2003 One Year Certificate in Polymer Technology, LondonMetropolitan University, UK

School

2000–2002 Government College for Men (North Karachi), Karachi, Pak-istan

1993–2000 Ali Ali High School, Karachi, Pakistan

Work & Research Experience

Sep. 2010–May. 2011 Master thesis with focus on "Smart Polymers" at Helmholtzcenter for Biomaterial development, Berlin, Germany

2006–2009 Process/Production engineer at Dadex Eternit Ltd, Karachi,Pakistan

Aug. 2005–Sep. 2005 Internee at Tri-Pak Films Ltd, Karachi, Pakistan

Professional Training

Workshop Two days workshop on "Extrusion and Injection MouldingProcess of Plastics" by Wavin (Leading pipe manufacturingcompany in Europe)

Workshop One day workshop on industrial "Hydraulics and its Optimiza-tion"

Training Production module of SAP (Enterprise Software)

133

Ultrathin chemisorbed polymer coatings : corrosion ... · Ultrathin Chemisorbed Polymer Coatings:...

Documents

Transcript of Ultrathin chemisorbed polymer coatings : corrosion ... · Ultrathin Chemisorbed Polymer Coatings:...