Accepted Manuscript

Title: The combined use of scanning vibrating electrodetechnique and micro-potentiometry to assess the self-repairprocesses in defects on “smart” coatings applied to galvanizedsteel

Authors: M. Taryba, S.V. Lamaka, D. Snihirova, M.G.S.Ferreira, M.F. Montemor, W.K. Wijting, S. Toews, G.Grundmeier

PII: S0013-4686(11)00263-5DOI: doi:10.1016/j.electacta.2011.02.048Reference: EA 16789

To appear in: Electrochimica Acta

Received date: 25-8-2010Revised date: 9-2-2011Accepted date: 10-2-2011

Please cite this article as: M. Taryba, S.V. Lamaka, D. Snihirova, M.G.S. Ferreira,M.F. Montemor, W.K. Wijting, S. Toews, G. Grundmeier, The combined use ofscanning vibrating electrode technique and micro-potentiometry to assess the self-repairprocesses in defects on “smart” coatings applied to galvanized steel, ElectrochimicaActa (2010), doi:10.1016/j.electacta.2011.02.048

This is a PDF file of an unedited manuscript that has been accepted for publication.As a service to our customers we are providing this early version of the manuscript.The manuscript will undergo copyediting, typesetting, and review of the resulting proofbefore it is published in its final form. Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers thatapply to the journal pertain.

Page 1 of 36

Accep

ted

Man

uscr

ipt

Weldable primers were modified with submicron containers loaded with corrosion inhibitors

SVET and micro-potentiometry were used to study the corrosion inhibition ability

Submicron containers do not damage the barrier properties of model primers

Artificial defects of 50x50 μm in a coating can be easily analyzed by SVET and SIET

Inhibiting dissolution of sacrificial Zn may result in detrimental dissolution of Fe

*Research Highlights

Page 2 of 36

Accep

ted

Man

uscr

ipt

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

1

The combined use of scanning vibrating electrode technique and

micro-potentiometry to assess the self-repair processes in defects on “smart”

coatings applied to galvanized steel

M.Taryba1, S. V. Lamaka1*, D. Snihirova1, M.G.S. Ferreira1,2, M.F.Montemor1, W.K.

Wijting3, S. Toews3, G. Grundmeier3

1ICEMS, Instituto Superior Tecnico, UTL, Av. Rovisco Pais, 1049-001 Lisbon, Portugal

2CICECO, Dep.Ceramics and Glass Eng., University of Aveiro, 3810-193, Aveiro, Portugal

3Institute for Polymer Materials and Processes, University of Paderborn, 33098 Paderborn,

Germany

Abstract

Model weldable primer coatings for galvanized steel were modified with submicron

containers loaded with corrosion inhibitors. This procedure aims at introducing a new

functionality in the thin coatings: self-repair ability. The assessment of this property demands

new protocols and new approaches, combining conventional electrochemical methods with

electrochemical and analytical techniques of micrometer spatial resolution. Thus, in this work

model defects were created in the coatings by using a focused ion beam (FIB). The coated

samples, containing the model defects, were immersed in a NaCl 0.05 M solution and the

corrosion inhibition ability was studied using the scanning vibrating electrode technique

(SVET) and the scanning ion-selective electrode technique (SIET). SVET-SIET measurements

were performed quasi-simultaneously. Qualitative chemical analysis was performed by SEM

combined with EDS. Complementary studies were carried out by electrochemical impedance

spectroscopy (EIS) to assess the effect of the containers filled with corrosion inhibitors on the

barrier properties of the coatings. The electrochemical results highlight the importance of the

combined use of integral and localized electrochemical techniques to extract information for a

better understanding of the corrosion processes and corresponding repair of active microscopic

defects formed on thin coatings containing inhibitor filled containers.

Keywords: SVET; micro-potentiometry; SIET; galvanized steel; self-repair.

*Corresponding author: sviatlana.lamaka@ist.utl.pt , +351 218 417 996.

1. Introduction

Organic and inorganic coatings have been widely applied for the protection of metal

alloys against corrosion with significant progress in recent years. However, despite essential

improvements in anti-corrosion coating technologies, problems persist in the long-term

Page 3 of 36

Accep

ted

Man

uscr

ipt

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

2

protection of metals in aggressive environments at very high cost. An increased interest has

been observed in approaches that combine different classes of materials, achieving more

complex and feed-back active structures in order to obtain new synergistic effects with respect

to the functional performance. A key strategic topic is the combination of nanostructured

materials and eco-friendly hybrid and/or organic polymers to develop new coatings that

enhance the product performance and introduce new functionalities as for example self-healing

ability.

Nowadays, the study of different coating systems (pre-treatments, primers, inorganic and

hybrid sol-gel films, conductive polymers and water based coatings) with self-healing ability is

an important topic [1], with a significant increase in the number of publications in the last five

years.

Based on literature it is possible to identify two main types of electrochemical mechanisms

for the self-healing of corrosion processes:

a) Healing due to the formation of protective layers of corrosion products, which block the

access of aggressive species and/or oxygen towards the active sites [2-4].

b) Healing due to the formation of protective species, resulting from the presence of organic or

inorganic corrosion inhibitors, often incorporated into “smart” particles [5-13].

These mechanisms of self-healing have been studied in different aggressive environments,

mainly on aluminum and zinc alloys, including cut edges. It has been shown that the corrosion

mechanisms and, consequently, the mechanisms of self-healing are dependent upon the

substrate composition and nature of the corrosion products formed [2-3, 14]. According to [2],

corrosion of zinc coupons immersed in NaCl electrolytes reveals slight acidification of the

anodic zones and major alkalinization of the cathodic sites. Corrosion products precipitate

between anodic and cathodic areas, where pH is favorable for formation of insoluble species.

In coated metallic substrates, both corrosion and self-healing mechanisms can be localized

over very small areas and generally involve changes in the electrochemical potential, current

density and acid-base equilibrium reactions. Furthermore, the processes may become even

more complex due to cathodic delamination or anodic undermining phenomena under buried

interfaces. The effective characterization of the self-repair ability is an acknowledged need due

to the novelty and dynamic behavior of these systems. In this perspective, the combined use of

integral and spatially-resolved electrochemical techniques becomes fundamental. Some of the

most successful spatially resolved electrochemical tools that have been used to investigate

localized corrosion phenomena in uncoated and coated substrates are shortly addressed below.

The Scanning Vibrating Electrode Technique (SVET), that detects local changes of the

current density due to the potential differences originated in the surface as consequence of the

ionic fluxes generated between the anodic and cathodic sites, has become one of the most

effective and widely used techniques to detect localized corrosion phenomena in bare and

coated substrates [2-4, 13, 15-24].

Page 4 of 36

Accep

ted

Man

uscr

ipt

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

3

The micro-potentiometric measurements with ion-selective microelectrodes can be

performed by using Scanning Ion-selective Electrode Technique (SIET) and SECM in

potentiometric mode. This method provides information about local activity (concentration) of

specific ions in solution (e.g. H+) making possible to trace changes in the acid-base equilibriums

associated with the electrochemical processes [2, 3, 8, 16-20, 25-28]. Mg2+, Zn2+, Na+ and Cl-

selective microelectrodes have also been reported to be used for studying localized corrosion

[16-18, 29, 30].

The localized electrochemical techniques described above present advantages and

limitations, and the need of complementary application has been stressed in literature [2-4, 16-

20, 24, 28, 29]. In this work a complementary combination of different electrochemical

techniques was used to study the corrosion and self-repair processes in model defects. For this

purpose, model weldable primers were applied to galvanized steel substrates. Additionally the

primers were modified with submicron containers loaded with different corrosion inhibitors.

The barrier properties of five different coatings were first assessed by electrochemical

impedance spectroscopy. Based on EIS ranking, the corrosion mechanisms and the self-repair

ability of two best coatings and the reference sample were studied using quasi-simultaneous

SVET and SIET measurements, complemented with SEM/EDX analysis. The results highlight

the need of new experimental protocols and the synergistic combination of SVET and micro-

potentiometry with conventional electrochemistry and scanning electron microscopy, providing

new information both on the corrosion processes and the self-repair effects occurring on defects

formed in coated galvanized steel.

2. Experimental

2.1. Materials

The substrate used for coating application consisted of hot dip galvanized steel, thickness

0.8 mm, coated with 22 m zinc layers on both sides. This substrate was supplied by the

industrial company coated with model weldable primers without nanocontainers. (Reference

sample M0) or modified with different containers loaded with corrosion inhibitors as described

in Table 1. Layered double hydroxide (LDH) nanocontainers were prepared according to the

procedures described in [31]. Preparation of the titania containers and loading with organic

inhibitors is detailed in [32]. Application of natural halloysites for corrosion protection was first

reported in [6]. The halloysite nanocontainers used in this work were prepared and

characterized as described in [33]. Apart from nanocontainers, the model weldable primer

contains spherical zinc particles (50% w/w) with diameter ranging from 2 to 8 m embedded in

an organic epoxy matrix. Along with Zn particles, the primer also contains silica and organic

additives meant to inhibit the dissolution of the zinc particles and prolong primer’s service life.

This primer is a model system based on a formulation used in the automotive industry. The

Page 5 of 36

Accep

ted

Man

uscr

ipt

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

4

cross-section SEM image of the reference sample supplemented with Fe, Zn and Si EDS

elemental mapping is presented in Fig. 1.

Two sets of samples were prepared: one set was used for evaluating the barrier properties

of the modified coatings by EIS. Based on these results the best samples were chosen for the

formation of Focused Ion Beam (FIB) defects and further studied by SVET and SIET.

2.2. Preparation of FIB defects and their SEM/EDS characterization

Defect preparation in organic coatings by means of the FIB technique consists of cutting

the coating through Ga-ion milling while the prepared cut is controlled by means of Field-

Emission Electron Microscopy (FE-SEM) [34]. Based on the relative movement of coated sample

and ion beam, three dimensional defects can be prepared. To the best knowledge of the authors

FIB was not yet applied to prepare 3D defects in organic coatings for a subsequent

electrochemical analysis. The coating around the defects remains effectively unchanged. In

each coated sample a defect with a size of 50 x 50 μm was made by means of the FIB technique

(Zeiss Neon 40). The FIB beam was placed perpendicularly to the surface, milling away the

coating until a defect was created. FIB was used to produce microsized defects that intend to

replicate small defects formed in the coating. The technique is suitable to produce identical

defects because not only the size of the defect can be well defined, but also the depth is

controlled. Regularly, SEM images were taken and EDS analysis was performed to control

composition and depth.

A Scanning Electron Microscope, coupled with Energy Dispersive X-ray Spectroscopy

(SEM/EDS), was used for examining the microstructure and the chemical composition of the

samples after SVET-SIET measurements. A semi-in-lens Hitachi SU-70 UHR Schottky

(Analytical) FE-SEM microscope coupled with a Bruker EDS detector was used. An electron

beam energy of 15 keV or 25 keV was applied for SEM analysis and EDS mapping.

2.3. Electrochemical techniques

All the coated specimens (without defects) were studied by EIS in order to evaluate the

effect of the containers on the barrier properties of the coating. Measurements were taken in 0.5

M NaCl. The EIS were performed using a Gamry FAS2 Femtostat with a PCI4 Controller in a

frequency range from 100 kHz down to 0.01 or 0.003 Hz. All the spectra were recorded at open

circuit potential, applying a 10 mV sinusoidal perturbations (rms signal). A conventional three-

electrode cell was used, consisting of a saturated calomel reference electrode, a platinum wire as

counter electrode and the coated metallic coupon as working electrode. The area of the working

electrode was approximately 3.4 cm2. The cell was placed in a Faraday cage.

The evolution of the corrosion activity and the self-healing processes on the samples

showing the best barrier properties were studied by the SVET and SIET techniques. A

commercial device from Applicable Electronics, controlled by the ASET program

(Sciencewares), was used to perform the SVET and the SIET measurements.

Page 6 of 36

Accep

ted

Man

uscr

ipt

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

5

The FIB defect was in the centre of the scanned area. The area exposed to the electrolyte

was approximately 0.9 x 0.7 mm2 for the sample M0, 0.5 x 0.4 mm2 for the M1 sample and 0.5 x

0.5 mm2 for the M4 sample. The rest of the sample’s surface was isolated using bee wax.

Samples M2 and M3 were not tested, since they demonstrated the weakest barrier properties

among the samples tested by EIS. The local ionic current density and pH were mapped at 35 x

35 grid generating 1225 data points. Measurements were taken in 0.05 M NaCl solution.

An insulated Pt-Ir probe (Microprobes Inc.) was used as vibrating electrode for SVET

measurements. A spherical layer (d=12 ± 2µm) of platinum black was deposited on the exposed

tip of the probe. The probe was placed 100±3 µm above the surface, vibrating in the planes

perpendicular (Z) and parallel (X) to the sample’s surface. The vibration frequencies of the

probe were 124 Hz (Z) and 325 Hz (X).

The localized pH measurements were carried out using pH-selective glass-capillary

microelectrodes. Silanized glass micropipettes were back-filled with the inner filling solution

and tip-filled with the H+-selective ionophore-based membrane. The ion-selective membrane

was prepared in-house and was composed of 6 wt% 4-nonadecylpyridine, 12 mol% potassium

tetrakis(4-chlorophenyl)borate and membrane solvent 2-nitrophenyloctyl ether. All reagents for

the membrane of pH-SME were Selectophore grade products from Fluka. The diameter of the

tip of the glass-capillary microelectrodes was 2 ± 0.3 μm. The pH-selective microelectrodes were

calibrated using commercially available pH buffers in a range from pH = 2 to 10 and

demonstrated linear Nernstian response -54.4 ± 0.9 mV/pH. The local activity of H+ was

detected 30±3 µm above the surface.

SVET and SIET scans were taken sequentially to trace the evolution of corrosion process in

the case of the M0 sample. Quasi-simultaneous SVET-SIET measurements were made on the M1

and M4 samples. To this end the vibrating probe of SVET and the pH-selective microelectrode

of SIET were brought up together to map the ionic current density and pH distribution

simultaneously. The 1.5 second time lag between reading SVET and SIET signals required due

to the response time of the pH-selective probe. The details of methodology of quasi-

simultaneous measurements are presented elsewhere [35].

3. Results and discussion

3.1. Barrier properties of the model primers

The addition of submicron particles loaded with corrosion inhibitors in the model primer

aims at introducing self-healing ability in the coating; however it is of outmost importance that

these additives do not damage the protective barrier properties. Otherwise, the protective

performance of the coating will be reduced and the addition of containers is not recommended,

despite the fact that they could impart corrosion inhibiting ability.

Page 7 of 36

Accep

ted

Man

uscr

ipt

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

6

EIS measurements were performed to evaluate the barrier properties and, therefore, the

anti-corrosion performance of the coated model systems. Fig. 2 depicts the EIS results obtained

for the samples M0, M1, M2, M3 and M4 (as listed in Table 1) after 2 hours and after 48 hours of

immersion in NaCl 0.5 M. At the early stages of immersion all the spectra revealed a capacitive

response in the high frequency domain, which can be related to the barrier properties of the

coating. In the low frequency range there is a resistive response, which is a consequence of the

presence of zinc microparticles or electrolyte uptake and formation of conductive pathways

through the coating. The M4 and M1 samples showed the highest coating resistance at the early

immersion stages, which was approximately 1 MΩ. At this early stage, all the samples showed

total impedance values identical or above those of the blank sample, showing that the additives

filled with corrosion inhibitors do not damage the barrier properties of the coating.

As the immersion time elapsed, the total impedance decreased for all the samples. After 2

days of immersion all the spectra presented similar behavior, showing impedance values in the

range 30 to 130 kΩ cm2. At this stage, the sample M1 revealed the highest overall impedance

values.

Since the additives must not damage the barrier properties of the coating, its resistance

was determined by fitting procedures. The equivalent circuit described elsewhere [7, 15]

consisted of an association of two time constants connected in parallel. Each time constant

included a resistive element and a constant phase element. This association, also including a

series resistive element to account for the electrolyte resistance, was used to fit the experimental

impedance spectra and to calculate the evolution of the coating resistance during immersion in

0.5M NaCl, Fig. 3. The fast decrease of the coating resistance at the beginning of immersion is

related with penetration of the electrolyte in the pores of the coating. Sample M2 showed

barrier properties that were very close to the blank coating, whereas the M4 and M1 systems

provided the highest coating resistances. This resistance showed a sudden drop during the first

day of immersion and, after 2 days, the coating resistance became stable and similar to that of

the blank sample. Concerning long term barrier properties, the M1 sample revealed the highest

coating resistances.

The impedance results clearly show that the submicron containers added to the model

weldable primer do not damage the barrier properties of the coating. Moreover, in some cases

the additive could improve the protective performance at least at the early stages of immersion.

Since the M1 and M4 samples revealed the best performance regarding the barrier properties,

these samples were chosen to create the FIB defects and to study the self-healing mechanisms as

presented in the next sections.

3.2. FIB defects analysis

To assess the self-healing ability, a model defect was prepared by means of FIB, at 30 kV

and 200 pA. The samples were milled for 3 µm. SEM images of the defect in the M0 coating are

Page 8 of 36

Accep

ted

Man

uscr

ipt

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

7

shown in Fig. 4. Since the coating is a weldable primer it contains zinc particles, which are

evident in the SEM/EDS images.

EDS maps of the defect are shown in Fig. 5. In the defect the content of C, O, and Si were

significantly lowered after the defect formation as consequence of the coating removal. The

substrate consists of a galvanized layer, so the signal of zinc in the defect increased,

comparatively to that of the coating.

Fig. 6 depicts the cross-section SEM image and the EDS spectrum of the defect formed in

the M1 coating. The largest and round shaped particles are the zinc powder pigments and the

longest ones, as highlighted in the square, are the added halloysites. The analysis reveals the

presence of silicon from the coating and zinc from the substrate. Rather high content of

aluminum as well as traces of magnesium and phosphorus accounts for the halloysites, which

are aluminosilicate clay mineral [6, 33].

In Fig. 7a and b a cross-section and a top view of the defect formed in the M4 coating are

presented. The EDS spectra obtained over the highlighted areas (continuous and dashed) are

presented in Fig. 7c and d, respectively. Fig. 7c reveals the presence of titanium, which comes

from the titanium dioxide particles used as inhibitor reservoir.

3.3. Combined SVET and SIET study

3.3.1. Reference sample (M0) - blank primer

Figs. 8a,b show the optical micrographs of the area tested by the SVET and SIET. The dark

square dot in the middle is the FIB defect. Fig. 8a was taken after one day of immersion in

0.05M NaCl. A whitish border of corrosion products of roughly circular shape is visible in the

central and lower left part of the sample. Fig. 8b indicates the position of the SVET and SIET

scans. Figs. 8c,d show the SIET and SVET maps, respectively.

The SVET map depicted in Fig. 8d show the anodic activity (positive ionic current fluxes)

over the defect and corresponding cathodic activity (negative current fluxes) in the right part of

the scanned area. Corrosion activity was detected at the very beginning of the immersion and

there was a gradual increase of the anodic and cathodic current density with time. After one

day of immersion the anodic current density attained values above 80 μA∙cm-2 and the cathodic

current density remained around -40 μA∙cm-2.

The pH map (Fig. 8c) reveals significant acidification over the defect and alkalinization in

the upper and right parts of the scanned area. The pH recorded in the alkaline zone attained

values 9.1 while the lowest measured pH in acidic area reached 3.5. Such profound acidification

over the defect is worthy of special attention. Assuming that the anodic dissolution would

generate only Zn2+ due to dissolution of the galvanized layer and zinc particles from the primer,

the pH in the anodic zone would not be lower than 5.0. The change of pH is predetermined by

the hydrolysis of Zn2+ cations:

Page 9 of 36

Accep

ted

Man

uscr

ipt

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

8

Zn Zn2+ +2e- (1)

nH[Zn(OH)n]OnHZn n)-(2

2

2

pK 1

hyd = 7.96 K 1

st = 1.10·106 (2)

According to the equation (2), the expected pH values are: pH = 4.0 if [Zn2+] = 1M, pH= 5.0

if [Zn2+] = 0.01M and pH= 6.0 if [Zn2+] = 0.0001M. The Kst1 value was taken from [36]. pK1

hyd was

calculated from K 1

st . For the studied sample, with a defect size of 50x50 µm, the local

concentration of Zn2+ is unlikely to be higher than 0.01M. For comparison, the maximum Zn2+

concentration measured using Zn2+-selective microelectrode over an anodically polarized 500

µm diameter Zn wire was found to be 0.01M [18]. Accessed by a disc electrode, maximum

concentration of Zn2+ during galvanic corrosion of a model Zn/steel couple was found to be

6.5∙10-3M [37].

However, the pH values detected by SIET in the middle of anodic zone are lower than 4. A

plausible explanation for such acidic pH can be the presence of Fe3+ ions, which cause a

considerable decrease of pH, due to hydrolysis of Fe3+ ions:

Fe Fe2+ +2e- (3)

Fe2+ Fe3+ +e- (4)

mH][Fe(OH)OmHFe m)(3

m2

3 pK 1

hyd = 2.20 K 1

st = 6.31·1011 (5)

The equilibrium reactions for the hydrolysis of Fe3+ (eq. 5) show that the expected pH values

are: pH = 2.3 if [Fe3+] = 0.01M and pH= 4.0 if [Fe3+] = 0.0001M. Thus, dissolution and hydrolysis

of Fe3+ are likely to contribute to the highly acidic pH values detected by micro-potentiometry,

Fig. 8c. The presence of the red rust zone around the defect in the optical images (Fig. 8a)

corroborates this explanation.

The reduction of dissolved oxygen is the main cathodic reaction, producing hydroxyl ions,

which are responsible for the increased pH.

O2 + 2H2O + 4e- 4OH- (6)

Comparing the pH map, Fig.8c, with the optical micrograph, Fig. 8a, it is evident that the

distribution of corrosion products follows the pH map. The boarder of the acidified zone, the

distorted circle around the anodic spot characterized by pH values in the range 6-7, Fig. 8c,

match the whitish deposits of corrosion products that are visible in Fig. 8a. The light blue curve

Page 10 of 36

Accep

ted

Man

uscr

ipt

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

9

in the lower middle part of the SVET map, Fig. 8d, also indicates that the current densities are

decreased over the area where the precipitation of corrosion products occurred. According to

Fig. 9a, this roughly circular area of corrosion products is likely to be composed of zinc oxide

ZnO or hydroxide Zn(OH)2 or hydroxylchloride Zn5(OH)8Cl2, which precipitate from solution

as pH reaches 6.3 and higher. Investigations performed by Ogle et al. on cut edges also revealed

the presence of zinc-based corrosion products precipitated in zones of intermediate pH, which

were in that case identified as ZnO and 3Zn(OH)2·2ZnCO3 [4] or Zn5(OH)8Cl2 [2] working as

cathodic inhibitors.

The pH values observed on the anodic sites of the M0 sample are more acidic, while the

alkalinization in cathodic areas is milder than reported in previous publications on cut edge

galvanized steel. The two most likely explanations for this fact are: i) A coating with a micro-

size defect (50x50 m) obviously generates less activity than the cut edge with an area of 10000

x 810 m, as, for example, tested by Ogle et al [4]. ii) In their localized pH measurements Ogle’s

group, as well as Ding and Hihara [28], use Hydrogen I cocktail B (Fluka, Ref.95293) for

producing glass-capillary microelectrodes. Remarkably, its functional pH range is 5.5 to 12 [39].

Thus, even if present, lower pH values simply could not be quantified with the tool used. In this

work we used another ionophore to produce pH-selective microelectrodes: 4-

nonadecylpyridine-based cocktail demonstrated a linear Nernstian response in pH ranging

from 2 to 10 making the detection of acidified sites possible. For dynamic and concentration

calibration curves of the microelectrodes based on tridodecylamine (Hydrogen I, Fluka) and 4-

nonadecylpyridine, interested readers can refer to our review [29].

Fig. 10 shows the results of the SEM/EDS measurements performed on the same area as

depicted in Fig. 8a. Depletion of zinc (relatively to initial even distribution of Zn-rich primer)

and accumulation of chlorine around the defect is shown in Fig.10 c and d. Silicon, which is part

of the formulation of the weldable primer, was also detected, being depleted in the same areas

as zinc, Fig.10b. This result reveals decomposition of the primer in the acidified area. SEM

observation of the defect after immersion indicates that the defect is partly blocked with the

corrosion products, Fig. 10e. The EDS spectrum over the defect, Fig. 10f, reveals the presence of

O, Cl, Zn and Fe. Modeling of ionic equilibrium in the solution containing Fe3+, Zn2+, Cl- and

CO32- suggests that the species blocking the defect are likely to be nonstoichiometric iron

hydroxylchloride Fe(OH)2.7Cl0.3, which starts precipitating at pH as low as 1, Fig. 9b. This

supports the statement that co-dissolution of zinc and iron occurred at the corroding defect,

taking active part in the ionic flux detected in the anodic site of the SVET maps.

Indeed, the model weldable primer contains additives which aim to hinder the dissolution

of the zinc particles and prolong primer’s service life. However, if sacrificial zinc is being

dissolved too slowly, the exposed steel substrate can undergo anodic dissolution resulting in

the major acidification of the defect area (eqs. 3-5). SVET alone could not reveal the chemistry

of the anodic reaction, while SIET suggested that both Zn and Fe take part in anodic dissolution.

This essential conclusion about the detrimental effect of the inhibiting additives in a particular

Page 11 of 36

Accep

ted

Man

uscr

ipt

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

10

primer formulation could not be attained by only using localized electrochemical technique that

measures net positive and negative currents such as SVET. This example shows the importance

of complementary use of SVET and SIET for elucidating the chemical mechanisms of corrosion

reactions.

3.3.2. Primers modified with nanocontainers loaded with inhibitors (M1 and M4)

Quasi-simultaneous SVET-SIET measurements were performed on the M1 and M4

samples. Figs. 11-a show the optical micrographs of the M1 sample and corresponding SIET

(Figs. 11-b) and SVET (Figs. 11-c) maps after different immersion times.

According to the SVET scan Fig. 11-1c, high corrosion activity could be detected after one

hour of immersion, attaining values around 80 μA∙cm-2, which are very close to those observed

for the blank sample, M0. The SIET map also shows acidification in the defect, characterized by

a decrease of the pH to values around 5, Fig.11-1b.

As the immersion time elapsed, both pH and current density show important changes.

After 8 hours of immersion (Fig. 11-2 a, b, c) the pH measured in the anodic zone ranges

between 4.6 and 5.0, while pH in cathodic zone is around 6. As compared to the reference

sample, the pH changes are more attenuated, both in the anodic and cathodic zones. Again, as it

was mentioned in part 3.3.1, if the anodic dissolution would generate only zinc cations Zn2+,

then the pH in the anodic zones could not decrease to values lower than 5.0. However, the pH

values detected by SIET in the anodic zone reached 4.6, suggesting that Fe3+ ions were also

generated simultaneously with Zn2+ during the corrosion reactions. After approximately one

day of immersion the current is lower than that observed for the reference sample M0 and the

pH values in the anodic zone range between 4.5 and 4.1. Anodic and cathodic current densities

are in the range 40 to -30 μA∙cm-2. Comparatively to the reference sample, there is a slight

decrease of the anodic current density and a narrower pH window. The cathodic areas showed

pH values around 7, much lower than these observed for the reference sample (around 9).

SEM/EDS study was performed right after the SVET/SIET experiments, in the upper part

of the scanned area as limited by the rectangle in Fig. 12c. The results clearly demonstrate the

presence of traces of iron precipitated around the defect. This precipitates may account for

mixed zinc-iron oxides as shown in Fig. 9b. Furthermore, a clear absence of Si was detected over

the defect, Fig. 13 c, but its distribution is nearly uniform around the defect, suggesting that the

primer organic matrix was not delaminated.

Fig. 13 shows the results of SEM/EDS observations of a bigger area around the defect. The

EDS maps were taken for several elements in order to assess the composition of the precipitates

covering the defect.

It can be seen that the defect is mainly covered by Zn-containing corrosion products. Also

significant amounts of P and O are detected, while Cl does not seem to contribute to formation

Page 12 of 36

Accep

ted

Man

uscr

ipt

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

11

of corrosion products in the defect. The phosphorous comes from the halloysites, as phosphate

salts are used to separate the halloysite nanotubes from the bulk clay. Accumulation of O, P and

Zn over the defect suggests formation of insoluble corrosion products, like Zn3(PO4)2 which can

be formed at pH as low as 4 [36]. The Si distribution around the defect is nearly uniform. In

contrast to the results obtained for the blank sample, it is evident that the coating is still intact

around the defect and that the corrosion process did not spread laterally as observed for the

blank sample. Although the acidification in the anodic zone does not attain the values observed

for the blank sample, it is still present for the M1 sample. In this case acidification front,

extending from the centre of the defect, is not as strong as for example in M0, delaying

dissolution of the zinc and therefore coating degradation. On the other hand, the halloysites

introduced in M1 are loaded with 2-mercaptobenzothiazole which may contribute for the

narrower pH variations, suggesting a decrease of the corrosion activity. MBTA has been

reported as an effective corrosion inhibitor for copper, aluminum, zinc and iron [40, 41]. The

inhibition efficiency is likely to be the consequence of inhibitor’s adsorption on the metallic

surface and formation of complexes with the metal cations.

The M4 sample was also studied using the same approach: quasi-simultaneous SVET-SIET

measurements followed by SEM/EDS examination of the sample after immersion test. The

visual appearance of the defect and area around it at the beginning of immersion tests is shown

in Fig. 14 1a. Optical micrograph of the same sample after 46 hours of immersion is shown in

Fig.14 1b.

Figs. 14 2a-5a show the SIET maps and Figs.14 2b-5b depict the SVET maps. The pH maps

are in a good agreement with the distribution of ionic currents. In general, the corrosion activity

in the M4 sample remained relatively weak throughout 46 hours of immersion. No corrosion

activity was registered by SVET in the defect of the M4 sample during the first 9 hours of

immersion, Fig. 14-2b and 3b. However, SIET detected the first signs of corrosion attack at the

very early stage of immersion, after 1 hour: slight alkalinization around the defect with the sign

of anodic dissolution in the defect, Fig. 14 2a and 3a. SVET registered corrosion onset after 1 day

of immersion, when a slight increase of the anodic activity was noticed, Fig. 14 4b. Further

progress of corrosion activity results in the appearance of anodic and cathodic activity as

detected by SVET and SIET (Fig. 14 5a and 5b).

The pH measured in the anodic areas ranges between 5.1 and 6.1 during the first day of

immersion, while the pH in the cathodic area remains within 6.2-7.1 during the same period of

time. These pH values can be caused by anodic dissolution and hydrolysis of zinc cations solely,

without involving iron as it was the case for M0 and M1 samples. The pH over the anodic and

cathodic areas change as the corrosion process proceeds and the products of anodic and

cathodic processes accumulate. The pH measured in the anodic zone decreases to 5.1-5.7 and

the pH of the cathodic area slightly increases to 6.5-7.1. After 2 days of immersion, the pH and

the current density remain similar to the values obtained after one day, and one order of

Page 13 of 36

Accep

ted

Man

uscr

ipt

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

12

magnitude below the values measured for the blank sample, revealing an important inhibiting

effect in this system.

Corrosion activity measured by SVET remained relatively low (within ±8 μA∙cm-2) for the

M4 sample during the entire immersion period.

The SEM/EDS maps obtained after the immersion test do not reveal significant changes

on the coating and over the defect, which is outlined by the dotted square in the SEM image,

Fig.15. The SEM/EDS study reveals uniform distribution of Zn, O and Si. It also reveals the

presence of titanium from the added nanocontainers modified with 8-hydroxiquinoline. Thus,

the inhibiting effect detected for this sample is a consequence of the presence of this inhibitor.

Inhibition by 8-hydroxyquinoline on zinc substrates was found to be predominantly anodic

resulting in a change of the mechanism of zinc dissolution [42]. Identical behavior has been

reported for mild steel [43]. Literature reports that the main process of inhibition in the presence

of 8-HQ seems to be the adsorption of the 8-HQ on the anodic sites, which slows down the

corrosion rate and prevents adsorption of chloride ions. Formation of complexes between 8-

hydroxyquinoline and dissolved cations is also likely to occur, which account for some mixed

inhibition effect [44, 45].

The results obtained in this work show that the containers added to the weldable model

primer, do not damage the barrier properties. The M1 and M4 systems presented improved

barrier properties comparatively to the reference system. The self-healing ability of these

nanocontainers was demonstrated by the quasi-simultaneous SVET and SIET measurements.

Correlation of local current density and pH is performed by matching SVET and SIET maps,

Figs 8, 11 and 14. A very good interdependence is observed: acidified zones correspond to

anodic currents in SVET maps and alkaline sites match cathodic currents.

Part of the corrosion activity might be hidden below the planes of SVET and SIET scans

due to formation of galvanic couples between the steel substrate and zinc coating in the defect.

Even though, it should be proportional for defects in all samples. One of the advantages of the

FIB technique is that it allows for the formation of reproducible defects. As the same primer was

used and similar defects were created, the comparative analysis of several different samples

tested under the same conditions produces reliable data.

Complementary SVET-SIET experiments allow assessing relevant information regarding

corrosion activity and self-healing potential. Thus, corrosion activity is relatively high for M0

sample (reference sample) coated with the model primer. The values of current density and pH

remain within the range -40 to +80 μA∙cm-2 and 3.5 to 9.1 respectively for 24 hours of immersion

in 0.05M NaCl solution. Low pH values are related with important iron dissolution. Presence of

corrosion products containing Fe was confirmed by the SEM/EDS analysis. Unexpected

dissolution of iron in the microdefect is an important warning that inhibiting of the zinc anode

may impair its protective properties. Thus, addition of corrosion inhibitor into the weldable

primer needs to be fully tested to optimize inhibition and dissolution of the sacrificial zinc.

Page 14 of 36

Accep

ted

Man

uscr

ipt

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

13

The sample M1 coated with the model primer modified by MBTA loaded halloysites

demonstrated corrosion activity with peak current densities reaching -40 to +80 μA∙cm-2.

Decrease of corrosion activity to -7 to + 32 μA∙cm-2 after 8 hours suggests a weak self-healing

effect which did not last long as corrosion intensified again to -20 to +60 μA∙cm-2 after 26 hours

of immersion.

The sample M4 coated with model primer modified by 8-hydroxyquinoline loaded in TiO2

nanoparticles exhibits the lowest corrosion activity and resists the corrosion attack for several

hours of immersion. Ionic current density did not exceed ±8 μA∙cm-2 and pH values remain in

the range 5.1 – 7.1 suggesting very weak anodic activity, which did not intensify throughout

two days of immersion. This reveals that these containers have an inhibiting effect on the

corrosion activity, and therefore induce self-healing ability in the FIB defects.

4. Conclusions

Electrochemical impedance spectroscopy revealed that the addition of containers filled

with organic inhibitors do not damage the barrier properties of weldable model primers applied

to galvanized steel substrates.

It was found that an artificial defect of 50x50 μm in a coating on HDG-steel size can be

easily studied by localized electrochemical techniques as SVET and SIET. The SVET-SIET

measurements clearly identify zones of cathodic and anodic activity, both in the maps of pH

and ionic current distributions for all three samples. The advantages of co-operative application

of SVET and micro-potentiometry measurements are emphasized. The combination of these

measurements with localized chemical analysis highlights the changes in the composition and

supports the statements made from the pH changes. On the reference sample the intense

acidification over the defect shows that Fe3+ species are formed together with Zn2+ species,

which was confirmed by the chemical EDS analysis. Thus, iron was clearly identified in the

anodic zones, when strong acidification was measured. The addition of inhibiting species

extending the service life of weldable primer led to the passivation of sacrificial Zn, with the

detrimental dissolution of Fe as a result.

The study reveals that the coatings modified with the containers filled with corrosion

inhibitors are able to induce an important corrosion inhibitive effect and that, in the case of the

M4 sample the corrosion activity was very low, indicating the self-healing ability of the

containers loaded with the corrosion inhibitor.

Summarizing, this work demonstrates the complementary use of the localized

electrochemical techniques to detect local changes both on the ionic activity and pH. This

information provides important insights into mechanism of corrosion processes and

demonstrates how to assess the self-healing ability of modified coatings over localized defects.

Page 15 of 36

Accep

ted

Man

uscr

ipt

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

14

Acknowledgments: The authors acknowledge financial support from projects

REDE/1509/RME/2005, PTDC/CTM/108446/2008 (FCT, Portugal), and European FP7

“MUST” NMP3-LA-2008-214261 and Marie Curie IRSES project “SISET” GA-2010-269282.

References

1. W. Feng, S.H. Patel, M-Y. Young, J.L. Zunino, M. Xanthos, Adv. Polym. Technol. 26 (2007) 1.

2. K. Ogle, S. Morel, D. Jacquet, J. Electrochem. Soc. 153 (2006) B1.

3. F. Thébault, B. Vuillemin, R. Oltra, K. Ogle, C. Allely, Electrochim. Acta 53 (2008) 5226.

4. K. Ogle, V. Baudu, L. Garrigues, , X. Philippe, J. Electrochem. Soc. 147 (2000) 3654.

5. S.V. Lamaka, M.L. Zheludkevich, K.A. Yasakau, R. Serra, S.K. Poznyak, M.G.S. Ferreira, Prog. Org. Coat. 58 (2007) 127.

6. D.G. Shchukin, S.V. Lamaka, K.A. Yasakau, M.L. Zheludkevich, H. Möhwald, M.G.S. Ferreira, J. Phys. Chem. C, 112 (2008) 958.

7. M.F. Montemor, R. Pinto, M.G.S. Ferreira, Electrochim. Acta 54 (2009) 5179.

8. D. Snihirova, S.V. Lamaka, M. Taryba, A.M. Salak, S. Kallip, M.L. Zheludkevich, M.G.S. Ferreira, M.F. Montemor, ACS Appl. Mater. Interfaces 2 (2010) 3011.

9. G.Williams, H.N. McMurray, M.J. Loveridge, Electrochim. Acta 55 (2010) 1740.

10. P.J. Kinlen, Y. Ding, D.C. Silverman, Corrosion 58 (2002) 490.

11. G. Paliwoda-Porebska, M. Stratmann, M. Rohwerder, K. Potje-Kamloth, Y. Lu, A.Z. Pich, H.J.Adler, Corros. Sci. 47 (2005) 3216.

12. M. Kendig, M. Hon, L. Warren, Prog. Org. Coat. 47 (2003) 183.

13. S.V. Lamaka, D. G. Shchukin, D.V. Andreeva, M. L. Zheludkevich, H. Möhwald, M. G.S. Ferreira, Adv. Funct. Mat. 18 (2008) 3137.

14. M.W Kendig, R.G. Buchheit, Corrosion (Houston, TX, U. S.) 59 (2003) 379.

15. W. Trabelsi, P. Cecilio, M.G.S. Ferreira, M.F. Montemor, Prog. Org. Coat. 54 (2005) 276.

Page 16 of 36

Accep

ted

Man

uscr

ipt

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

15

16. S.V. Lamaka, O.V. Karavai, A.C. Bastos, M.L. Zheludkevich, M.G.S. Ferreira, Electrochem. Commun. 10 (2008) 259.

17. O.V. Karavai, A.C. Bastos, M.L. Zheludkevich, M.G. Taryba, S.V. Lamaka, M.G.S.Ferreira, Electrochim. Acta 55 (2010) 5401.

18. A.C. Bastos, M.G. Taryba, O.V. Karavai, M.L. Zheludkevich, S.V. Lamaka, M.G.S.Ferreira, Electrochem. Commun. 12 (2010) 394.

19. F. Thébault, B. Vuillemin, R. Oltra, C. Allely, K. Ogle, E. Tada, ECS Transactions 11 (2008) 91.

20. S.V. Lamaka, G. Knörnschild, D.V. Snihirova, M.G. Taryba, M.L. Zheludkevich, M.G.S. Ferreira, Electrochim. Acta 55 (2009) 131.

21. C. Scheffey, Rev. Sci. Instrum. 59 (1988) 787.

22. L.F. Jaffe, R. Nuccitelli, J. Cell Biol. 63 (1974) 614.

23. H.S. Isaacs, A.J. Aldykiewicz Jr., D. Thierry, T.C. Simpson, Corrosion (Houston, TX, U.S.) 52 (1996) 163.

24. A.M. Simões, A.C. Bastos, M.G.S. Ferreira, Y. González-García, S. González, R.M. Souto, Corros. Sci. 49 (2007) 726.

25. C. Wei, A.J. Bard, G. Nagy, K. Toth, Anal. Chem. 67 (1995) 1346.

26. B.R. Horrocks, M.V. Mirkin, D.T. Pierce, A.J. Bard, G. Nagy, K. Toth, Anal. Chem. 65 (1993) 1213.

27. M. Marenzana, A.M. Shipley, P. Squitiero, J.G. Kunkel, A. Rubinacci, Bone 37 (2005) 545.

28. H.B. Ding, L.H. Hihara, J. Electrochem. Soc. 155 (2008) 226.

29. S.V. Lamaka, R. M. Souto, M.G.S. Ferreira, in: A.M. Vilas (Ed.), Microscopy: Science,

Technology, Applications and Education, Edition 2010, Formatex Research Center, Badajoz,

Spain, 2010, vol. 3, pp. 2162-2173.

30. C.J. Lin, R.G. Du, T. Nguyen, Corrosion 56 (2000) 41.

31. M.L. Zheludkevich, S.K. Poznyak, L.M. Rodrigues, D. Raps, T. Hack, L.F. Dick, T. Nunes, M.G.S. Ferreira, Corros. Sci. 52 (2010) 602.

32. I.A. Kartsonakis, I. L. Danilidis, G. S. Pappas, G. C. Kordas, J.Nanosci. Nanotech. 10 (2010) 5912.

Page 17 of 36

Accep

ted

Man

uscr

ipt

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

16

33. D. Fix, D.V. Andreeva, Y.M. Lvov, D.G Shchukin, H. Möhwald, Adv. Func. Mater. 19 (2009) 1720.

34. R .Wirth, Eur. J. Mineral 16 (2004) 863.

35. S.V. Lamaka, M.G. Taryba, M.F. Montemor, H.S. Isaacs, M.G.S. Ferreira, Electrochem.

Commun. 13 (2011) 20.

36. Y.Y. Lur’e, Handbook of Analytical Chemistry, 6th ed., Himiya, M., 1989.

37. E. Tada, S. Satoh, H. Kaneko, Electrochim. Acta 49 (2004) 2279.

38. I. Puigdomenech, Program MEDUSA (Make equilibrium diagrams using sophisticated algorithms), Royal Institute of Technology, Stockholm, 1999.

39. Fluka ionophore catalogue. Available at

http://www.sigmaaldrich.com/etc/medialib/docs/Sigma/Instructions/h_hydrogen.Par.0001.

File.tmp/h_hydrogen.pdf. Accessed: February 04, 2011.

40. M. Ohsawa, W. Suetaka, Corros. Sci. 19 (1979) 709.

41. M.L. Zheludkevich, K.A. Yasakau, S.K. Poznyak, M.G.S. Ferreira, Corros. Sci. 47 (2005) 3368.

42. M.S. Abdel-Aal, Z.A. Ahmed, M.S Hassan, J. Appl. Electrochem. 22 (1992) 1104.

43. M.S. Abdel-Aal, M.S. Morad, Br. Corros. J. 36 (2001) 253.

44. S. Szunerits, D.R. Walt, Anal. Chem. 74 (2002) 886.

45. S.V Lamaka, M.L. Zheludkevich, K.A. Yasakau, M.F. Montemor, M.G.S. Ferreira, Electrochim. Acta 52 (2007) 7231.

Figure Captions

Fig. 1. SEM micrograph of cross-section of reference coating M0 (a), and corresponding EDS

elemental mapping of Fe (b), Zn (c) and Si (d).

Fig. 2. EIS Bode plots for the galvanized steel coated with the weldable primer modified with

different additives. Data is presented for 2 and 48 h of immersion in 0.5M NaCl.

Fig. 3. Evolution of the weldable coating resistance (determined by fitting) during immersion in

0.5M NaCl.

Page 18 of 36

Accep

ted

Man

uscr

ipt

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

17

Fig. 4. 50 x 50 µm FIB defect formed in the M1 primer coating applied to galvanized steel. Top

image: top view. Bottom image: Taken under an angle of 54 °.

Fig. 5. Sample M0. EDS mapping of the defect, showing the Carbon and Oxygen (left column

images) and Zinc and Silicon (right column images).

Fig. 6. FIB defect formed in the M1 primer coating applied to galvanized steel. a) Cross-section

view; b) EDS spectrum according to the white rectangle.

Fig. 7. 50 x 50 µm FIB defect formed in the M4 primer coating applied to galvanized steel. a)

Cross-section view; b) Top view; c) EDS spectrum taken at the white circle; d) EDS spectrum

taken at the dashed circle.

Fig. 8. Sample M0. Optical micrograph of the sample after one day of immersion in 0.05 M

NaCl, showing the whitish contour of corrosion products around the defect (a); optical

micrograph indicating the data points of the SVET and SIET scans (b); pH distribution (c), map

of ionic current density (d).

Fig. 9. Modelling of ionic equilibria as a function of pH in solution containing the following

ions: a) [Zn2+]= 0.01M, [Cl- ]= 0.05M, [CO32+]= 7·10-11M; b) [Zn2+]= 0.01M, [Fe3+]= 0.0005M, [Cl-

]= 0.05M, [CO32+]= 7·10-11M. The diagrams were made using Hydra/Medusa software [38].

Fig. 10. Sample M0. a) SEM images and corresponding EDS elemental distribution of b) Si; c)

Zn; d) Cl; e) SEM image of the defect; f) EDS spectrum of the area shown in e).

Fig. 11. Sample M1. Optical micrographs of the scanned sample (1a, 2a, 3a) with the white

rectangle limiting the area of the SVET and SIET scans; pH distributions (1b, 2b, 3b); and

current density distribution (1c, 2c, 3c). The measurements were taken after 1, 8 and 26 hours of

immersion in 0.05M NaCl.

Fig. 12. Sample M1. EDS elemental distribution of a) Fe and b) Si; c) pH map with white

rectangle showing the area of the EDS mapping.

Fig. 13. Sample M1. SEM image and EDS elemental mapping of the defect and area around it.

Fig. 14. Sample M4. Optical micrograph of the sample with the FIB defect after 1 hour (1a) and

46 hours (1b) of immersion in 0.05M NaCl; pH distribution (2a, 3a, 4a, 5a) and corresponding

Page 19 of 36

Accep

ted

Man

uscr

ipt

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

18

current density distribution (2b, 3b, 4b, 5b). The measurements were made after 1, 9, 21 and 46

hours of immersion. The white rectangle limits the area of the SVET and SIET scans.

Fig. 15. Sample M4. SEM image and EDS elemental mapping taken after 46 hours of immersion.

Page 20 of 36

Accep

ted

Man

uscr

ipt

Table 1. Composition of the model weldable primers containing different nanocontainers.

Nanocontainer / Inhibitor

M0 Primer (reference, without nanocontainers)

M1 Primer modified with halloysites/2-mercaptobenzothiazole (MBTA)

M2 Primer modified with LDH/vanadium oxide (VOx)

M3 Primer modified with TiO2/2-mercaptobenzothiazole (MBTA)

M4 Primer modified with TiO2/8-hydroxyquinoline (8-HQ)

Table 1

Page 21 of 36

Accep

ted

Man

uscr

ipt

Figure 1

Page 22 of 36

Accep

ted

Man

uscr

ipt

Figure 2

Page 23 of 36

Accep

ted

Man

uscr

ipt

Figure 3

Page 24 of 36

Accep

ted

Man

uscr

ipt

Figure 4

Page 25 of 36

Accep

ted

Man

uscr

ipt

Figure 5

Page 26 of 36

Accep

ted

Man

uscr

ipt

Figure 6

Page 27 of 36

Accep

ted

Man

uscr

ipt

Figure 7

Page 28 of 36

Accep

ted

Man

uscr

ipt

Figure 8

Page 29 of 36

Accep

ted

Man

uscr

ipt

Figure 9a

Page 30 of 36

Accep

ted

Man

uscr

ipt

Figure 9b

Page 31 of 36

Accep

ted

Man

uscr

ipt

Figure 10

Page 32 of 36

Accep

ted

Man

uscr

ipt

Figure 11

Page 33 of 36

Accep

ted

Man

uscr

ipt

Figure 12

Page 34 of 36

Accep

ted

Man

uscr

ipt

Figure 13

Page 35 of 36

Accep

ted

Man

uscr

ipt

Figure 14

Page 36 of 36

Accep

ted

Man

uscr

ipt

Figure 15