Corrosion Science 112 (2016) 289–307

Contents lists available at ScienceDirect

Corrosion Science
journal homepage: www.elsevier.com/locate/corsci

A comparative study of corrosion inhibitors on hot-dip galvanized

steel
Ioannis A. Kartsonakis a , Stefan G. Stanciu b , Alecs A. Matei b , Radu Hristu b ,
Antonis Karantonis a , Costas A. Charitidis a,∗
a
National Technical University of Athens, School of Chemical Engineering, “Research Unit of Advanced, Composite, Nano Materials & Nanotechnology”, 9
Heroon Polytechneiou St., Zografos, GR-157 80, Greece
b
Center for Microscopy-Microanalysis and Information Processing, University Politehnica Bucharest, Splaiul Independentei 313, Sector 6, Bucharest,
Romania

a r t i c l e i n f o a b s t r a c t

Article history: This study investigates the morphology, composition and corrosion resistance of hot dip galvanized
Received 16 May 2016 (HDG) steel panels treated by immersion in NaCl solution in the presence of six protective compounds:
Received in revised form 29 July 2016 8-hydroxyquinoline, 2-mercaptobenzothiazole, 5-amino-1,3,4-thiadiazole-2-thiol, calcium phosphate,
Accepted 30 July 2016
cerium nitrate and sodium molybdate. The results reveal that upon corrosion new compounds are formed
Available online 1 August 2016
onto the exposed areas of the treated panels. The presence of these compounds on the surface of the HDG
steel panels hinders the corrosion process by reducing the rate of the anodic and cathodic reactions. The
Keywords:
conducted experiments provide new insights on the corrosion mechanisms of HDG steel.
A. Steel
A. Zinc © 2016 Elsevier Ltd. All rights reserved.
B. EIS
B. SEM
B. X-ray diffraction
C. Pitting corrosion

1. Introduction in galvanized coatings is based on sacrificially dissolving of zinc.

However, the corrosion rate of zinc is affected by the current pass-
The Hot-Dip Galvanized steel (HDG) represents one of the most ing per unit of area and by the relative zinc and steel exposed areas
extensively used types of steel taking part in almost every major [5], therefore there are still many gaps that need to be filled for
application of industries where iron or steel are used because it fully comprehending this topic. In this context, we estimate that
exhibits a great set of mechanical properties, being well resistive to more complex models are required for shedding more light on
abrasion or impact. Moreover, in the same time, HDG steel demon- what type of strategies should be followed for developing novel
strates good resistance to thermal gradients and to low and medium highly-efficient protective compounds compatible with HDG.
aggressive corrosive environments. Such anti-corrosive properties The exposure of bare steel in NaCl solution results in corrosion
are connected to the fact that galvanized coatings develop a met- initiation of iron. In the case of HDG steel, zinc, which constitutes
allurgical bond through the formation of a series of iron and zinc the outer layer, is also attacked by chlorides notwithstanding that
alloys layers with protective abilities. However, even in the pres- higher NaCl solutions are required. However, the overall corrosion
ence of these layers, HDG steel can still be influenced by aggressive process depends on the state of the galvanized surface [6]. Zinc
corrosive environments [1], thus there is a great need to iden- oxide is a common corrosion product in diluted solutions of salts
tify or develop new families of protective compounds capable to such as NaCl, with zinc hydroxide present in various amounts as
alleviate this problem. The studies reported so far on the topic a minor component. In case of concentrated solutions of NaCl, the
of galvanized steel corrosion have conventionally described the corresponding zinc salts may form zinc hydroxide, in addition to
corrosion behaviour of galvanized steel on the basis of the zinc zinc oxide. At a HDG coating the succession of layers, starting from
layer corrosion [2–4], due to the fact that the protection of steel the pure steel phases, is a Fe-Zn alloy with increasingly lower iron
content resulting in the outer phase being a zinc-rich solid solution
with approximately 0.003 wt.% Fe [2]. Results reported in the sur-
vey performed by Yadav et al. [7] demonstrate that the corrosion
∗ Corresponding author. performance of the Fe-Zn/Zn intermetallic phases that are formed
E-mail address: [email protected] (C.A. Charitidis).

http://dx.doi.org/10.1016/j.corsci.2016.07.030
0010-938X/© 2016 Elsevier Ltd. All rights reserved.
290 I.A. Kartsonakis et al. / Corrosion Science 112 (2016) 289–307

during the hot-dip galvanization process have an important con- the corrosion inhibiting action of cerium nitrate [Ce(NO3 )3 ] and
tribution towards the corrosion behaviour of HDG steel. However, sodium molybdate (Na2 MoO4 ) onto HDG steel, there are yet no
corrosion rate of zinc can be reduced via the addition of corrosion studies available in the literature that address the influence of 8-
inhibitors. These inhibitors are divided in two categories; those hydroxyquinoline (8HQ), 5-amino-1,3,4-thiadiazole-2-thiol (ATT),
that operate as a surface film agents due to the effect of precipi- calcium phosphate [Ca3 (PO4 )2 ] and 2-mercaptobenzothiazole
tation of insoluble compounds onto the metal surface (adsorption (MBT) to corrosion protection of HDG steel. Consequently, the
inhibitors) and those that passivate the metal surface of HDG steel evaluation of the inhibiting action of Ce(NO3 )3 and Na2 MoO4 was
(anodic inhibitors) [8]. Consequently, the corrosion protection of accomplished together with the four aforementioned inhibitors in
zinc contributes to the protection of the galvanized steel. order to obtain comparable results under the same experimental
With respect to protection of zinc to corrosive environments conditions using the same corrosive environment (NaCl solution).
using corrosion inhibitors, several studies have demonstrated It is mentioned that although Yang et al. [25] investigated MBT as
cerium ions as an effective solution in this regard. Fatima et al. [9] corrosion inhibitor for zinc surface, the corrosive environment that
investigated the electrochemical behaviour of galvanized steel pre- was used was NaClO4 , whereas in our current work the experiments
treated with bis-[triethoxysilylpropyl] tetrasulfide silane solutions were conducted in NaCl solution. The difference is that the Cl− are
modified with CeO2 nanoparticles and cerium ions. The results more aggressive and increase the dissolution of zinc compared to
revealed improved barrier properties of the silane films due to the ClO4 − [26].
presence of nanoparticles and the cerium ions. Studies with the The corrosion protection effectiveness of the complexes or
same silane precursor were also performed by Cabral et al. [10] indi- oxides or salts that are formed on the metal alloy surface, as well
cating that the presence of cerium ions into the coating decreases as the morphological modifications that occur on the surface of
the porosity and increases the thickness resulting in corrosion pro- the HDG steel panels following exposure, was evaluated by three
tection. Kong et al. [11] further proved that the incorporation of families of techniques. Electrochemical techniques [electrochem-
cerium nitrate into silane based coating improves the corrosion ical impedance spectroscopy (EIS), potentiodynamic polarization
protection of HDG steel by reducing the rate of both anodic and (PP) method and linear polarization resistance (LPR)] were used for
cathodic reaction in the corrosion process. Deflorian et al. [12] corrosion estimation of HDG steel. Microscopy techniques [scan-
demonstrated that cerium oxides operate as corrosion inhibitors ning electron microscopy (SEM), digital microscopy and reflectance
for HDG steel due to their potential to act as a complexing agent confocal scanning laser microscopy (RCSLM)] were used for collect-
reacting with the zinc ions in the anodic sites leading to the for- ing qualitative information on the surface damage and structural
mation of an insoluble film of corrosion products. Moreover, in changes that occurs on the metal surface as a result of the cor-
the works of Aramaki [13,14] it was demonstrated that a corro- rosion process. Optical cues on the effects of exposing the HDG
sion protective film of hydrated CeO2 and Ce2 O3 is formed on the steel panels to the considered environments were also collected at
zinc surface after its exposure to aqueous solution of NaCl includ- macroscopic scale using digital photography. Finally, spectroscopy
ing Ce(NO3 )3 or CeCl3 . Finally, Arenas et al. [15] investigated the methods [energy dispersive X-ray analysis spectroscopy (EDS) and
inhibitor behaviour of CeCl3 for HDG steel in aerated NaCl solu- X-ray diffraction (XRD)] were conducted to evaluate the forma-
tions demonstrating that Ce forms a continuous film that consists tion of potential complexes on the metal alloy surface following
of a mixture of oxides/hydroxides of cerium(III) and cerium(IV) on its interaction with the six considered corrosion inhibitors, in the
the surface of the metal substrate after prolonged immersion time. corrosive environment.
The inhibitive action of Na2 MoO4 on HDG steel using NaCl solu-
tion was studied by Volovitch et al. [16] indicating that protective
Mo-rich films are formed onto the metal surface in pH values up to 2. Experimental method
12; however, the oxidation state of Mo depends on the pH value.
Moreover, lots of studies have also been performed on the corrosion 2.1. Reagents and solutions
protection of HDG steel using molybdates and phosphates in silane
based coatings [17–20] indicating improved corrosion resistance Inhibitors: 8-hydroxyquinoline (8HQ, Acros-Organics), 5-
of the coatings due to the reaction of molybdate and phosphate amino-1,3,4-thiadiazole-2-thiol (ATT, Acros-Organics), calcium
ions with metal cations resulting in the formation of insoluble com- phosphate [Ca3 (PO4 )2 , Fluka AG], cerium nitrate [Ce(NO3 )3 , Sigma-
pounds on the anodic sites of the zinc surface. Moreover, enhanced Aldrich] 2-mercaptobenzothiazole (MBT, Acros-Organics) and
inhibition was observed in the presence of magnesium and calcium sodium molybdate (Na2 MoO4 , Acros-Organics) as well as other
ions [21,22]. chemical reagents used in our experiment were of analytical
Organic corrosion inhibitors have also been studied for the pro- reagent grade. The HDG substrates were received from Chemet-
tection of HDG steel. In the work of Ogle et al. [23] the effect all GmbH (Frankfurt). According the hot dip galvanization process
of l-cysteine on the Zn dissolution in 0.5 M NaCl was evaluated the steel was coated with a layer of zinc by immersing the metal
revealing that l-cysteine inhibited zinc dissolution for short expo- in a bath of molten zinc at a temperature of around 449 ◦ C. The
sures at slightly acidic or slightly basic pH due to the complex average thickness of the galvanization layer amounts 14 ␮m. Cor-
formation of l-cysteine with Zn-ions on the substrate. El-Shayed rosion studies were performed in 50 mM NaCl (Table 1) solutions
et al. [24] demonstrated improved corrosion protection after for- prepared with distilled water. The aforementioned low concentra-
mation of both phytic acid and diethylene triamine pentamethane tions of chloride ions were used in consideration of decreasing the
phosphonic acid self-assembled monolayers on HDG steel surface. rate of the corrosion processes in order to better evaluate the corro-
Furthermore, in the work of Yang et al. [25] it was demonstrated sion inhibition mechanism, since HDG steels were exposed to the
that MBT forms a film onto zinc surface that exhibits corrosion corrosive environment uncoated. The concentration of Ce(NO3 )3
protection properties when exposed to NaClO4 solution. and Na2 MoO4 inorganic corrosion inhibitors was 0.5 g/L. However,
The main goal of the presented experiments is to identify which the addition of 0.5 g of 8HQ, ATT, Ca3 (PO4 )2 and MBT corrosion
inhibitors perform best at restricting the aforementioned corro- inhibitors into 1 L water resulted in the preparation of saturated
sion process. In this regard the effect of six organic and inorganic aqueous solutions due to their low solubility in water. The HDG
inhibitors was investigated, by exposing HDG steel specimens steel panels were ground sequentially using 1200 and 2000 grade
to sodium chloride solution in their presence. Despite the fact SiC papers and then cleaned in agreement with ASTM D6386-99
that previous studies have been already performed related with (reapproved 2005) prior to being used in the conducted experi-
 I.A. Kartsonakis et al. / Corrosion Science 112 (2016) 289–307 291

Table 1
Composition of the solutions for testing the effectiveness of the inhibitors onto HDG steel.

Reagent Concentration (mM) Reagent Solubility in water (mg/100 g) Sample

1. NaCl 50 mM – – HDG-blank
2. NaCl 50 mM + 0.5 g/L 8HQ 3.447 (saturated) 55.6 HDG-8HQ
3. NaCl 50 mM + 0.5 g/L ATT 0.0037 (saturated) 0.5 HDG-ATT
4. NaCl 50 mM + 0.5 g/L Ca3 (PO4 )2 0.064 (saturated) 2.0 HDG-Ca3 (PO4 )2
5. NaCl 50 mM + 0.5 g/L Ce(NO3 )3 1.151 very soluble HDG-Ce(NO3 )3
6. NaCl 50 mM + 0.5 g/L MBT 1.136 (saturated) 19.0 HDG-MBT
7. NaCl 50 mM + 0.5 g/L Na2 MoO4 2.428 very soluble HDG-Na2 MoO4

ments. This procedure is necessary due to the fact that the exposed a saturated silver chloride electrode [Ag/AgCl,KCl(sat) ] as reference
zinc coating interacts with the environment to form zinc oxides at and a platinum foil as counter electrode was used for all the PP
first, then zinc hydroxides, and finally zinc carbonates [27]. Accord- measurements. Polarization curves were assessed on the bare alloy
ing to the aforementioned ASTM, the cleaning procedure includes after different exposure intervals. Anodic and cathodic branches of
the degreasing of HDG steel panels in acetone, then their immer- polarization curves were recorded separately at room temperature
sion in a NaOH solution of pH 11 for 5 min at 60 ◦ C and finally their with the potential scan rate of 1 mVs−1 in the anodic and in the
rinsing with distilled water and drying in air. cathodic directions, starting from the open circuit potential (OCP,
EOC ). Concerning the LPR measurements, the potential range was
2.2. Characterization ±2.5 mV versus OCP and the scan rate was 0.1 mVs−1 . The ratio of
the applied potential to the applied current (E/I) is defined as
The surface morphology of the HDG steel panels was examined the Rp (polarization resistance) and calculated from the slope of the
using multiple imaging techniques providing information at micro- potential versus current plot [28].
and macroscopic scales. For collecting micro-scale information SEM The EIS experiments were performed at room temperature,
imaging was performed using a PHILIPS Quanta Inspect (FEI Com- at the OCP, using the same potentiostat instrument connected
pany) microscope with W (tungsten) filament 25 KV equipped with with a SI 1260 Impedance/Phase-Gain analyser and the afore-
EDAX GENESIS (AMETEX PROCESS & ANALYTICAL INSTRUMENTS). mentioned three-electrode electrochemical cell. The measuring
Microscopic assessment was further performed by using at first frequency ranged from 100 kHz down to 5 mHz, and the rms voltage
a Leica TCS SP Laser Scanning Microscope working in reflectance was 10 mV. Spectra were treated using the Z-view Software (Scrib-
workmode. For the conducted RCSLM investigations the samples ner Associates Incorporated, SAI) utilizing the adequate equivalent
were illuminated with a 633 nm HeNe laser line and light reflected electric circuits. The corrosion inhibition efficiency, (%), was esti-
from the sample was collected with a 10 × 0.3 NA HCX PL Fluorite mated by using the equation:
objective, which offers a 1 mm x 1 mm field of view. Further on, Rct(inhibitor) − Rct(bareMetal)
microscopic investigations at larger scales were conducted with an (%) = x100 (1)
Rct(inhibitor)
AIGO GE5 Digital Microscope using a pseudo-coloration strategy.
Finally, for placing micro-scale information into a broader con- where Rct(inhibitor) and Rct(bareMetal) represent the charge transfer
text, optical details at macroscopic scale were collected using a resistances in the presence and in the absence of inhibitor, respec-
HTC Dual 4 MP, f/2.0, 27 mm autofocus, dual-LED (dual tone) flash tively [29,30].
Camera. Using this camera, all samples were imaged from a fixed
distance, using a preset macro mode for achieving sharp focus at 3. Results and discussion
short distance. The chemical analysis was conducted via Fourier
Transform Infrared Spectroscopy (FT-IR) using a FT-IR attenu- 3.1. Electrochemical studies
ated total reflectance Agilent Cary 630 instrument in the range of
400–4000 cm1 . Furthermore, chemical analysis was accomplished All the electrochemical characterizations, PP, LPR and EIS were
via Micro-Raman measurements using a Renishaw inVia spectrom- conducted in a pH range from 5.8 to 6.3, depending on the type of
eter working in backscattering configuration and equipped with each potential corrosion inhibitor. It is mentioned that all the elec-
a near-infrared diode laser emitting at 785 nm. The spectra were trochemical measurements were performed at least in triplicate.
recorded by focusing the laser beam on the sample surface and A representative curve is depicted for each sample. The presence
adjusting the light power so that 1 mW is provided for a spot of of an inhibitor can shift the OCP to more positive values indicat-
about 1 ␮m diameter. ing a decrease to the oxidation rate of the system. On the other
For identifying the compounds that are formed on the exposed hand, a negative potential shift could possible denote inhibition of
HDG panels, XRD measurements were conducted via an APD the cathodic reactions. The measurement of the OCP with immer-
2000 diffractometer [GNR, Agrate Conturbia (Novara), Italy] work- sion time is very useful to determine the time where the system is
ing in a Bragg-Brentano geometry and using CuK␣ radiation stabilized in order for PP characterization to be performed. Fig. 1a
(␭ = 0.1540598 nm). The performed XRD investigations took place depicts the evolution of OCP versus the exposure duration of HDG
under the following instrumental settings: 40 kV tube voltage, steel samples to 50 mM NaCl at room temperature. It can be seen
30 mA tube current, 0.02◦ 2␪ step-size and 1 s counting time per that after 1 h of immersion in the corrosive environment, the pres-
step. All XRD data was collected in the range of 5◦ –45◦ (2␪). ence of the inhibitors (apart from 8HQ) results in an increase of
The conductivity measurements of the HDG steel samples the OCP compared to the one of bare HDG steel denoting improved
exposed to corrosive environment in the presence or absence anodic inhibition effect that can be ascribed to the formation of a
of corrosion inhibitors were performed using a thermo scientific protective layer. Among all the corrosion inhibitors, the addition of
ORION 3 STAR Conductivity Benchtop instrument equipped with Na2 MoO4 exhibits the highest OCP. On the other hand, the pres-
an ORION 013005MD 4-Electrode Conductivity Cell. A Princeton ence of 8HQ induces neither an increase nor a decrease of the OCP;
Applied Research EG & G 263A potentiostat instrument was used for a behaviour that can be attributed to the simultaneous inhibition of
the PP and LPR measurements. A three-electrode electrochemical the cathodic reactions and also to the formation of a protective layer
cell, consisting of a working electrode (≈1.0 cm2 of exposed area), that inhibits the anodic reactions. The OCP measurements were
292 I.A. Kartsonakis et al. / Corrosion Science 112 (2016) 289–307

Table 2
Tabulated values of EOC obtained from PP technique and of Rp estimated via LPR
method acquired for all the samples after exposure to 50 mM NaCl.

Samples EOC [V vs EOC [V vs Rp (kohm cm2 )

Ag/AgCl,KCl(sat) ] Ag/AgCl,KCl(sat) ] (168 h exposure)
(1 h exposure) (168 h exposure)

HDG-blank −1.063 −0.967 2.049

HDG-8HQ −1.051 −0.953 53.839
HDG-ATT −1.023 −0.866 4.841
HDG-Ca3 (PO4 )2 −1.030 −0.956 3.489
HDG-Ce(NO3 )3 −1.005 −0.835 82.803
HDG-MBT −1.040 −0.957 10.814
HDG-Na2 MoO4 −0.984 −0.909 11.314

values compared to HDG-blank denoting that they are less prone

to corrosion after 1 h of exposure to the corrosive environment.
Considering the cathodic branches of the polarization curves
after 168 h of immersion in the corrosive environment (Fig. 2c) it
can be seen that all the inhibitors, apart from Ca3 (PO4 )2 , could pos-
sible inhibit the cathodic reactions as they present lower cathodic
current densities compared to the HDG-blank sample. Further-
more, the estimation of the corresponding anodic branches of the
polarization curves clearly denotes that the presence of all the
inhibitors into the corrosive environment results in lower anodic
current densities in comparison with the HDG-blank sample that
can be assigned to the formation of corresponding protective layers
(Fig. 2d). This outcome is further confirmed by the tabulated values
of EOC obtained from PP technique after the exposure of the samples
to 50 mM NaCl solution for 168 h (Table 2), where all the samples
demonstrate more positive EOC values compared to the one of the
HDG-blank sample.
It is important to mention that the PP measurements were car-
ried out in a pH range from 5.8 to 6.3 and that the anodic and
cathodic branches of polarization curves were recorded in a poten-
tial range between −1.28 V and 0 V vs. Ag/AgCl,KCl(sat) electrode.
Fig. 1. a) The evolution of OCP versus time in case of the HDG steel samples exposed Therefore, it may be remarked, with respect to the Pourbaix dia-
to 50 mM NaCl at room temperature, b) the reproducibility of the OCP measurements gram of zinc [27,31,32] that in all the PP measurements a reduction
in case of the HDG-ATT, HDG-MBT and HDG-Na2 MoO4 samples.
of water and oxygen takes place. Moreover, in the aforementioned
range of potential, dissolution of zinc exists but, be that as it may,
performed in triplicate, in identical conditions in order to confirm the oxidation of zinc does not involve oxidation of water. Finally,
the reliability of the collected data sets. Fig. 1b demonstrates the it should be noted that the PP method was used only for quali-
reproducibility of the data sets collected for the HDG-ATT, HDG- tative estimations because the determination of Rp and corrosion
MBT and HDG-Na2 MoO4 samples. It should be taken into account current (icorr ) values via the linear fitting of the polarization curves
that the three aforementioned samples were selected indicatively according to the Tafel method was impossible, due to the fact that
and that for each system with inhibitor, three data sets are depicted there were two reduction processes in the system; the reduction
using three lines, under the same identifying data point. Therefore, of oxygen and the reduction of water [33]. Furthermore, it should
it may be remarked that as the immersion time elapses, all three be taken into account that there is no clear linear region in the
lines of each sample demonstrate similar OCP values. These results anodic branches in order to evaluate the corrosion current density.
clearly demonstrate the repeatability, and hence reliability, of the Moreover, despite the fact that zinc is usually corroded via general
performed measurements. corrosion, galvanic corrosion and wet storage stain corrosion, verti-
The cathodic and anodic PP curves obtained after immersion cally placed zinc panels can also be corroded with the formation of
of HDG steel samples in corrosive environment for 1 h and 168 h pits (pitting corrosion) in distilled water, at room temperature and
(7 days) in the presence or absence of corrosion inhibitors are in an open air system [27,34]. Consequently, the Tafel method can-
demonstrated in Fig. 2a, b and c, d, respectively. Taking into consid- not be used as it only yields an average uniform corrosion rate and
eration the cathodic branches of the PP curves after 1 h of exposure is not appropriate for localized corrosion, such as pitting corrosion
(Fig. 2a), it may be remarked that the samples HDG-Ca3 (PO4 )2 , [27,35].
HDG-Ce(NO3 )3 and HDG-MBT demonstrate lower cathodic current The obtained Rp values after the performance of LPR character-
densities compared to the HDG-blank, indicating that the addi- ization technique are tabulated in Table 2. The LPR method was
tion of Ca3 (PO4 )2 , Ce(NO3 )3 and MBT inhibitors into the corrosive performed by immersing the samples to the corrosive environ-
environment could possible reduce the cathodic reactions. Further- ment for 168 h. Considering the LPR results, it is clearly denoted
more, studying the anodic branches of the polarization curves, it that Ce(NO3 )3 and 8HQ inhibitors demonstrate enhanced corro-
is mentioned that all the samples exhibit lower anodic current sion protection to HDG steel as both HDG-Ce(NO3 )3 and HDG-8HQ
densities compared to the HDG-blank sample, meaning an effec- samples have Rp values one order of magnitude higher compared
tive inhibition of the anodic reactions (Fig. 2b). Regarding the EOC to those of HDG-blank and the other samples.
values according to Table 2, it is observed that all the HDG steel The EIS plots of HDG steel samples after exposure to corrosive
samples treated with corrosion inhibitors demonstrate nobler EOC environment in the presence or absence of corrosion inhibitors are
 I.A. Kartsonakis et al. / Corrosion Science 112 (2016) 289–307 293

Fig. 2. The polarization curves recorded for the HDG steel samples in the presence or absence of inhibitors after exposure in 50 mM NaCl at room temperature: (a) cathodic
after 1 h (b) anodic after 1 h, (c) cathodic after 168 h and (d) anodic after 168 h.

demonstrated in Fig. 3. Considering the EIS Bode curves after 3 h both HDG-8HQ and HDG-Ce(NO3 )3 samples increase as the immer-
of exposure (Fig. 3a), it may be remarked that almost all the sys- sion time elapses from 3 h to 72 h. Regarding the EIS Bode curves
tems including HDG steel and inhibitors illustrate higher EIS moduli obtained after 168 h of exposure to corrosive environment (Fig. 3e),
compared to the one of the HDG-blank sample indicating that the it is clearly denoted that all inhibitors, except for Ca3 (PO4 )2 , protect
presence of inhibitors into the corrosive environment enhances HDG steel against corrosion as all the corresponding impedance
the HDG steel corrosion protection. Moreover, it should be men- moduli are higher compared to the one of HDG-blank. The cor-
tioned that the HDG-Ce(NO3 )3 sample demonstrates the highest EIS responding Nyquist plots (Fig. 3f) also illustrate that the systems
modulus meaning that the addition of Ce(NO3 )3 inhibitor into the containing all the inhibitors, apart from Ca3 (PO4 )2 , present higher
corrosive environment presents the best protection of HDG steel impedance values at low frequencies of the corresponding semicir-
against corrosion. On the other hand, the addition of Ca3 (PO4 )2 cles compared to the one of the system without inhibitor indicating
compound into the NaCl solution does not seem to improve the that the aforementioned inhibitors protect the metal alloy from
anticorrosive properties of the HDG steel. These results are fur- corrosion.
ther confirmed by the EIS Nyquist plots (Fig. 3b), where it can be The EIS Bode and Nyquist plots for the untreated HDG steel
observed that the samples HDG-8HQ, HDG-ATT, HDG-Ce(NO3 )3 , after 3 h of immersion in corrosive environment are characterized
HDG-MBT and HDG-Na2 MoO4 depict higher EIS values at the low by two time constants, one in the middle frequency range that is
frequency range in comparison with the HDG-blank sample. Esti- assigned to the corrosion process [12,36] and a second one in the
mating the EIS Bode curves after 72 h of exposure to corrosive low frequencies that is attributed to the formation onto the gal-
environment (Fig. 3c), it can be seen that apart from HDG-Ca3 (PO4 )2 vanized surface of a combined film due to an air-formed layer of
sample, all the other samples demonstrate higher impedance oxides [37] and to an oxide/hydroxide layer from the immersion
values in the low frequency range compared to the HDG-blank in the corrosive environment [38] (Fig. 3a, b). As the immersion
indicating that they protect HDG steel against corrosion. The best time elapses from 3 h to 168 h, only one time constant in the
corrosion protection is presented by the HDG-Ce(NO3 )3 sample middle-low frequency range is depicted for the untreated HDG
as it has the highest impedance modulus. The corresponding EIS steel, which is attributed to the corrosion process [12,36] (Fig. 3c–f).
Nyquist plots (Fig. 3d) further confirm the aforementioned results. This result denotes that the film assigned to air-formed oxide layer
Moreover, according to EIS Bode and Nyquist plots (Fig. 3a-d), it is or to an oxide/hydroxide layer from the immersion in the corrosive
mentioned that the impedance values in the low frequency range of environment was dissolved in the NaCl solution. As corrosion pro-
294 I.A. Kartsonakis et al. / Corrosion Science 112 (2016) 289–307

Fig. 3. EIS plots obtained for the HDG steel samples after immersion in 50 mM NaCl solution in the presence or absence of inhibitors at room temperature: Bode plots: a) 3 h,
c) 72 h and e) 168 h. Nyquist plots: b) 3 h, d) 72 h and f) 168 h.
 I.A. Kartsonakis et al. / Corrosion Science 112 (2016) 289–307 295

Fig. 3. (Continued)

gresses from 3 h to 168 h, the relative relaxation process is shifted [37]. Estimating the EIS Bode and Nyquist plots for the HDG-8HQ
to lower frequencies due to the increase of the corrosion active sample (Fig. 3a–d), it can be seen that three time constants are illus-
area and the corresponding increase of the double layer capacitance trated, one in the high frequency range that is assigned to a film
296 I.A. Kartsonakis et al. / Corrosion Science 112 (2016) 289–307

formation between the galvanized surface and the 8HQ compound

[39], a second one in the middle frequency range that is ascribed
to corrosion process [12,36] and a third one in the low frequencies
that is due to the air-formed film of metal oxides [37] and/or to an
oxide/hydroxide layer from the immersion in the corrosive envi-
ronment [38]. As the immersion time elapses from 72 h to 168 h,
the time constant in the low frequency range is ceased indicating
that the aforementioned metal oxides film was dissolved to the
corrosive environment (Fig. 3e, f).
Taking into consideration the EIS Bode and Nyquist plots for
the HDG-ATT sample (Fig. 3a–f), it may be remarked that two time
constants are demonstrated, one in the middle frequencies that is
attributed to corrosion process (capacitive semicircle) [12,36] and
another one in the low frequency range that is ascribed to surface
roughness and inhomogeneity due to adsorption of [H+ ] (induc-
tive semicircle) [30]. Regarding the EIS Bode and Nyquist plots for
the HDG-Ce(NO3 )3 sample (Fig. 3a–f), it is important to mention
that two time constants appear; one in the middle-high frequency
range that is assigned to the formation of a cerium oxide film on
the galvanized surface [40,41] and a second broad one that covers
the middle and low frequency range and ascribed to both corrosion
process [12,36] as well as the formation of a metal oxide/hydroxide
film [37].
Estimating the EIS Bode and Nyquist plots for the HDG-MBT
sample after 3 h of immersion in corrosive environment (Fig. 3a,
b), it is observed that two time constants appear; one broad
time constant from the middle-high to the middle-low frequency
range (capacitive semicircle) that includes the corrosion process
as well as the formation of a complex between the metal and the
MBT [25,42,43] and a second time constant in the low frequency
range (inductive semicircle) due to the surface inhomogeneity
and the adsorption of [H+ ] [30]. As the immersion time elapses
from 3 h to 168 h, the inductive semicircle in the low frequency
range disappears; however, the aforementioned broad time con-
stant is separated into two time constants. A first one appears in
the middle-high frequency range that is ascribed to the forma-
tion of a complex between the metal and the MBT [42,43] and a
second one is demonstrated in the middle-low frequencies due to Fig. 4. Equivalent circuit used for numerical simulation of the EIS data obtained for
corrosion process [44] (Fig. 3c–f). Considering the curves of both the HDG steel panels after immersion in 50 mM NaCl solution in the presence or
absence of inhibitors.
EIS Bode and Nyquist for the HDG-Na2 MoO4 (Fig. 3a–f), it may be
remarked that for the first 72 h two time constants appear; one
broad in the middle-high frequency range that is ascribed to cor- [46]. Using the Cole-Cole approach together with CPE [47,48], the
rosion process as well as a Mo-rich film between the galvanized capacitance can be calculated from the fittings by:

surface and MoO4 −2 ions and a second one in the low frequencies
n RY0
that is attributed to the formation of both a metal oxide/hydroxide C= (3)
Rn
layer [12,32,36]. As the immersion time elapses, one broad time
constant is observed in the low-middle frequency range that is The equivalent circuit for the bare HDG steel for the first
assigned to corrosion process as well as a Mo-rich film between 3 h of immersion in corrosive environment includes five compo-
the galvanized surface and MoO4 −2 ions. nents; the solution resistance (Rsol ), the charge transfer resistance
The interpretation of the EIS results for the bare HDG steel panels together with a double layer capacitance due to the time constant
in the absence and presence of corrosion inhibitors in 50 mM NaCl in the middle frequency range (Rct -CPEdl ) and the resistance and
solution, was performed by numerical fitting, using the equivalent the capacitance of the air-formed layer of zinc oxides and/or the
circuit illustrated in Fig. 4. In this equivalent circuit, constant phase oxide/hydroxide layer from the immersion in the corrosive envi-
elements (CPE) were used instead of pure capacitors. This modifi- ronment onto the galvanized surface due to the relaxation process
cation is obligatory when the phase shift of a capacitor is different at the low frequencies (Rox -CPEox ) (Fig. 4b) [12,36,37]. However,
from −90◦ [45]. The impedance of an R-CPE parallel association is the meaning of the time constant in the low frequency range is not
given by: clear and it can be related with diffusion processes either by oxy-
gen diffusion or by the evolution of the Zn(I) and Zn(II) equilibrium
R at the layer/substrate interface [30,36,38,46,49]. Nevertheless, the
ZR−CPE = n (2)
1 + RY0 (jω) introduction of Warburg impedance did not improve the numer-
ical fitting of the EIS results. As the immersion time elapses the
where Y0 is the admittance of the CPE and n is the CPE expo- zinc oxides film disappears and the equivalent circuit for the bare
nent. Constant phase elements correspond to a capacitor when the HDG steel includes only three components (Fig. 4a). It is impor-
CPE exponent (n) is one. Additionally, when 0.5 < n < 1, CPE repre- tant to mention that despite the fact that in the work of Pedraza
sents a non-ideal capacitor. Finally, when (n) equals to 0.5, the CPE et al. [38] similar capacity values with ours are demonstrated for
describes a diffusion process and represents a Warburg impedance the time constant at low frequencies that is assigned to an oxide
 I.A. Kartsonakis et al. / Corrosion Science 112 (2016) 289–307 297

Table 3
Fitting parameters according to equivalent circuits of Fig. 4 after samples exposure to 50 mM NaCl solution.

3 h into NaCl HDG-blank HDG-8HQ HDG-ATT HDG-Ce(NO3 )3 HDG-MBT HDG-Na2 MoO4

2
Rs (kohm cm ) 0.234 0.228 0.218 0.217 0.239 0.233
CPEinh -T (␮F cm−2 s−n ) – 0.127 – 6.94 12.2 –
CPEinh -P – 0.92005 – 0.75811 0.73032 –
Rinh (kohm cm2 ) – 0.187 – 0.585 1.20 –
Cinh (␮F cm−2 ) – 0.0503 – 1.20 2.56 –
CPEdl -T (␮F cm−2 s−n ) 42.3 20.5 17.1 17.1 6.92 25.7
CPEdl -P 0.84751 0.69458 0.9205 0.80532 0.71649 0.75329
Rct (kohm cm2 ) 0.612 2.56 1.55 40.8 6.17 1.96
Cdl (␮F cm−2 ) 21.9 5.61 12.5 15.7 1.99 9.66
CPEox -T (mF cm−2 s−n ) 27.0 18.0 – – – 2.24
CPEox -P 0.84393 0.99998 – – – 0.7556
Rox (kohm cm2 ) 0.159 3.13 – – – 0.510
Cox (mF cm−2 ) 35.3 18.0 – – – 2.34
RL (kohm cm2 ) – – 4.89 – 17.4 –
L (kH cm2 ) – – 27.9 – 277.0 –
(%) – 76.1 60.5 98.5 90.1 68.8

72 h into NaCl HDG-blank HDG-8HQ HDG-ATT HDG-Ce(NO3 )3 HDG-MBT HDG-Na2 MoO4

2
Rs (kohm cm ) 0.236 0.225 0.227 0.218 0.234 0.268
CPEinh -T (␮F cm−2 s−n ) – 0.0834 – 8.09 45.6 –
CPEinh -P – 0.84453 – 0.74648 0.67477 –
Rinh (kohm cm2 ) – 3.82 – 0.501 0.580 –
Cinh (␮F cm−2 ) – 0.0189 – 1.25 7.92 –
CPEdl -T (␮F cm−2 s−n ) 1410 90.2 67.5 22.3 61.1 81.4
CPEdl -P 0.55758 0.50549 0.95362 0.8088 0.71432 0.7486
Rct (kohm cm2 ) 0.625 18.6 1.77 94.1 3.87 3.22
Cdl (␮F cm−2 ) 1280 150 60.9 26.6 34.3 51.9
CPEox -T (mF cm−2 s−n ) – 19.5 – – – 6.08
CPEox -P – 0.99999 – – – 0.99998
Rox (kohm cm2 ) – 1.05 – – – 0.982
Cox (mF cm−2 ) – 19.5 – – – 6.08
RL (kohm cm2 ) – – 3.15 – – –
L (kH cm2 ) – – 96.8 – – –
(%) – 96.6 64.7 99.3 83.9 80.6

168 h into NaCl HDG-blank HDG-8HQ HDG-ATT HDG-Ce(NO3 )3 HDG-MBT HDG-Na2 MoO4

Rs (kohm cm2 ) 0.239 0.237 0.218 0.216 0.264 0.250

CPEinh -T (␮F cm−2 s−n ) – 0.0852 – 7.54 57.9 –
CPEHF -P – 0.83348 – 0.72994 0.64754 –
Rinh (kohm cm2 ) – 5.91 – 0.550 0.863 –
Cinh (␮F cm−2 ) – 0.0187 – 0.991 11.3 –
CPEdl -T (␮F cm−2 s−n ) 2540 68.9 37.1 27.3 121 149
CPEdl -P 0.63402 0.5873 0.91025 0.8486 0.69103 0.76979
Rct (kohm cm2 ) 0.856 38.4 2.37 88.2 5.01 4.84
Cdl (␮F cm−2 ) 3980 137 29.2 31.9 96.7 135
CPEox -T (mF cm−2 s−n ) – – – – – 50.8
CPEox -P – – – – – 0.99995
Rox (kohm cm2 ) – – – – – 0.184
Cox (mF cm−2 ) – – – – – 50.8
RL (kohm cm2 ) – – 0.837 – – –
L (kH cm2 ) – – 238 – – –
(%) – 97.8 63.9 99.0 82.9 82.3

layer, in our current study the calculated capacity value for the HDG-Ce(NO3 )3 sample for all time periods of exposure to corrosive
time constant in the low frequencies that is due to an air-formed environment is schematic represented in Fig. 4c and contains the
film of metal oxides and/or to an oxide/hydroxide layer from the solution resistance (Rsol ), the capacitance of the cerium oxide film
immersion in the corrosive environment should be considered with in the middle-high frequency range (Rinh -CPEinh ), and the compo-
caution (35.3 mF cm−2 , Table 3). nents Rct -CPEdl in the middle-low frequencies due to the corrosion
The equivalent circuit for the HDG-8HQ sample for the first 3 h process.
of immersion is represented by Fig. 4d and contains seven compo- The EIS results of the HDG-ATT sample can be fitted by the equiv-
nents; the solution resistance (Rsol ), the capacitance of the complex alent circuits of Fig. 4e that includes the solution resistance (Rsol ),
film between the metal and the 8HQ in the high frequency range the components Rct -CPEdl in the middle frequencies and two addi-
(Rinh -CPEinh ), the components Rct -CPEdl in the middle frequencies tional components in the very low frequency range that are due to
due to corrosion process, and the components (Rox -CPEox ) in the the inductive semicircle; the resistance of an inductor (RL ) and the
low frequency range that are attributed to the air-formed zinc inductance of an inductor element (L). Estimating the EIS results of
oxides film and/or to an oxide/hydroxide layer from the immer- the HDG-MBT sample after 3 h of immersion, it may be remarked
sion in the corrosive environment [12,36–38]. As the immersion that they can be fitted by the equivalent circuit of Fig. 4e. However,
time increases from 72 h to 168 h, the EIS results are fitted by the as the immersion time passes from 3 h to 168 h, the appropriate
equivalent circuit of Fig. 4c due to the disappearance of the time circuit is the one of Fig. 4c. Finally, the fitting of the HDG-Na2 MoO4
constant in the low frequency range. The equivalent circuit for the EIS results can be performed by the circuit of Fig. 4d for the first
298 I.A. Kartsonakis et al. / Corrosion Science 112 (2016) 289–307

72 h and as the immersion time increases from 72 h to 168 h, the

EIS results are fitted by the equivalent circuit of Fig. 4c due to the
disappearance of the time constant in the low frequency range. The
EIS fitting parameters according to equivalent circuits of Fig. 4 after
samples exposure to 50 mM NaCl solution are tabulated in Table 3.
It should be mentioned that according to the presented Rp
and Rct values of Tables 2 and 3, both LPR and EIS methods con-
firm that HDG-Ce(NO3 )3 and HDG-8HQ samples demonstrate the
best anticorrosive behaviour after 168 h immersion in corrosive
environment. However, regarding the CPEdl -P values (CPEdl -P rep-
resents the n which is the CPE exponent in equations 2 and 3)
calculated for the sample HDG-8HQ after exposure to 50 mM NaCl
solution for 72 h (0.50549) and 168 h (0.58730), as well as for the
sample HDG-blank after exposure for 72 h (0.55758) it may be
remarked that these elements cannot be associated to a non-ideal
capacitor (Table 3) [45]. However, it should be taken into account
that Tables 2 and 3 do not present the mean values of the LPR
measurements and EIS fitting parameters.
The evolution of the mean values together with the correspond-
ing error bars of the EIS fitting parameters Rct , Cdl , Rox , Cox , Rinh
and Cinh of the HDG steel panels after their exposure to 50 mM
NaCl solution for 168 h in the presence or absence of corrosion
inhibitors are illustrated in Fig. 5 (and Supplementary Fig. S1).
Regarding the estimating values of the double layer capacitance
together with the charge transfer resistance of the aforementioned
systems (Fig. 5a,b), it may be remarked that the system including
the inhibitor Ce(NO3 )3 depicts the highest Rct values together with
low Cdl values for all the immersion time intervals. Furthermore,
very small variations are observed in the double layer capacitance
of this system denoting that the corrosion reaction area does not
undergo changes. It is important to mention that high Rct values
mean high resistance to the corrosion process and that low Cdl val-
ues correspond to low corrosion reaction area [50]. Consequently,
the inhibitor Ce(NO3 )3 provides the best protection to HDG steel
against corrosion. Moreover, it is observed that after 72 h of immer-
sion time, 8HQ exhibits enhanced Rct values and low Cdl values
indicating that it is an effective inhibitor against corrosion of HDG
steel (Table 3 and Fig. 5).
Considering the evolution of capacitance and resistance of the
air-formed layer of oxides and/or the oxide/hydroxide layer onto
the galvanized surface during immersion in 50 mM NaCl solution,
it is demonstrated that in the system without inhibitors the zinc
oxide layer disappears as the immersion time elapses due to the
corrosion process (Supplementary Fig. S1c). Regarding the Rinh and
Cinh EIS parameters, it is important to mention that the complex
film between the metal and 8HQ of the system HDG-8HQ exhibits
enhanced resistance values that continuously increase denoting Fig. 5. The evolution of the mean values of the EIS fitting parameters as a function
of time: a) Rct , b) Cdl .
improved corrosion protection (Supplementary Fig. S1e). On the
other hand, the cerium oxide film in the system HDG-Ce(NO3 )3
presents small variations of Rinh and Cinh values denoting that the values of (%), 98.5 and 90.1, respectively. After 72 h of immersion,
corrosion reaction area does not undergo changes (Supplementary the (%) value of MBT slightly decreases, whereas the samples HDG-
Fig. S1e,d). 8HQ and HDG- Ce(NO3 )3 demonstrate the highest values, 96.6 and
Taking into account the numerical fitting of the EIS results 99.3, respectively. Finally, as the immersion time elapses to 168 h,
(Table 3), it may be remarked that the resistance solution of all both samples HDG-8HQ and HDG- Ce(NO3 )3 illustrate higher values
the systems was ranged from 216 to 268 ohm cm2 . Moreover, in of (%) compared to the other systems, 97.8 and 99.0, respectively.
order to be clarified what the relevance of the conductivity to the To sum up, taking into account the (%) values that are tabulated in
obtained results is, conductivity measurements were performed Table 3, it is clearly denoted that the presence of the inhibitors 8HQ,
to all the corrosive solutions in the absence or present of the ATT, Ce(NO3 )3 , MBT and Na2 MoO4 into the NaCl solution provide
inhibitors. The acquired conductivity values were between 4.93 and enhanced corrosion inhibition to HDG steel but, be that as it may,
5.27 mS cm−1 . Therefore, it can be safely assumed that the conduc- 8HQ and Ce(NO3 )3 , depict the highest (%) values.
tivity of the solutions does not influence the comparison between
different systems. 3.2. Morphology
Table 3 demonstrates the tabulated values of (%) derived from
equation (1) after 3, 72 and 168 h of samples exposure to 50 mM The surface SEM images as well as the corresponding tabu-
NaCl, respectively. The results indicate that for the first 3 h the solu- lated values of the HDG steel panels EDS analyses prior and after
tions containing Ce(NO3 )3 and MBT inhibitors present the highest their exposure to 50 mM NaCl solution for 168 h in the presence
 I.A. Kartsonakis et al. / Corrosion Science 112 (2016) 289–307 299

Fig. 6. SEM surface images of the HDG steel panels after exposure for 168 h: a) NaCl, b) NaCl + 8HQ, c) NaCl + MBT.

Table 4
Tabulated values of wt.% element concentration of bare HDG steel panels unexposed to corrosive environment; and after exposure to 50 mM NaCl solution for 168 h.

C O Fe Zn Ce P N Mo S Cl

HDG-unexposed – 2.21 8.53 89.26 – – – – – –

HDG-blank 7.29 23.09 2.73 65.72 – – – – – 1.17
HDG-8HQ 49.93 12.81 1.52 28.46 – – 7.27 – – –
HDG-ATT 7.41 11.82 1.57 69.33 – – 4.62 – 3.47 1.78
HDG-Ca3 (PO4 )2 6.13 2.74 2.23 86.54 – 1.02 – – – 1.34
HDG-Ce(NO3 )3 5.13 16.36 1.31 55.02 20.10 – – – – 2.09
HDG-MBT 27.41 8.44 1.25 42.06 – – 4.92 – 13.32 1.65
HDG-Na2 MoO4 5.85 21.81 1.29 67.11 – – – 3.77 – 0.17

or absence of corrosion inhibitors, are demonstrated in Fig. 6 (and assigned to the formation of zinc oxide on the surface of the panel
Supplementary Fig. S2) and Table 4, respectively. The surface of the due to its interaction with the atmosphere [51].
HDG steel panel prior to its immersion in the corrosive environment The surface of the untreated HDG steel panel that was immersed
does not depict any macroscopic defects but, be that as it may, the in the corrosive environment exhibits delaminated areas of 100 ␮m
SEM image with the highest magnification illustrates stripes and as well as cracks and needled-shaped substances that occur due
pores (Supplementary Fig. S2a). The EDS analysis reveals that the to corrosion related morphological modifications (Fig. 6a). More-
surface of the panel mainly consists of the zinc element (Table 4); over, an increase of oxygen values is observed due to the formation
however, the presence of the iron element in the outer layer of of corrosion products (Table 4). According to Fig. 6b, the addition
the panel is attributed to the HDG process where annealing of the of 8HQ into the corrosive environment results in a unique mor-
hot dipped panels results in the existence of 6 wt.% of iron in the phology onto the surface of HDG steel that consists of carbon and
outer layer [27]. Finally, the presence of small amounts of oxygen is nitrogen, as revealed by the EDS analysis (Table 4). The presence of
300 I.A. Kartsonakis et al. / Corrosion Science 112 (2016) 289–307

carbon and nitrogen may indicate adsorption of organic molecules resentative RCSLM images for the bare HDG steel panel, and for
which could be attributed to the formation of a complex film onto the HDG steel panels exposed to the corrosive environment in the
the metal surface. The presence of ATT in the corrosive environ- presence of the considered inhibitors.
ment results in the formation of a complex onto the metal surface In Fig. 7a, it can be observed that the bare, unexposed, HDG
but, be that as it may, it does not inhibit the corrosion process and steel panel has a patterned texture, which can be associated either
macroscopic defects can be observed on the surface of the metal to topographic aspects or to oxidation related ones (or a mix-
alloy (Supplementary Fig. S2d). The EDS analysis reveals enhanced ture). Although covered in some parts by islet-type circular regions,
value of oxygen due to the corrosion attack (Table 4). Estimating this patterned texture can be also clearly observed also in the
the addition of Ca3 (PO4 )2 into the NaCl solution, the surface of the case Fig. 7b which means that exposing the HDG steel panel to
HDG steel panel depicts as well signs of the corrosion process in the NaCl + Ca3 (PO4 )2 does not results in the formation of new com-
form of cracks, denoting that (PO4 )2 −3 cannot effectively inhibit the pound with homogeneous repartition across the substrate surface.
corrosion process (Supplementary Fig. S2e). On the other hand, the The same situation was observed for NaCl + Na2 MoO4 (Supplemen-
oxygen value is low revealing that the formation of zinc oxides is tary Fig. S3). A case in which the opposite situation occurs, and the
limited in this case (Table 4). patterned texture is almost imperceptible as a result of a homo-
In the case of Ce(NO3 )3 addition in the NaCl solution it may geneous film deposited on top of the HDG steel surface, is the one
be remarked that the corrosion process has been inhibited, as when the substrate was exposed to NaCl + 8HQ (Fig. 7c). It may be
the panel surface exhibits less signs of corrosion compared to the remarked that the same situation was observed for NaCl + ATT (Sup-
HDG-blank (Supplementary Fig. S2f). Furthermore, the EDS analysis plementary Fig. S3). Intermediary situations were observed for the
reveals some amounts of cerium and enhanced amounts of oxygen cases when the samples were exposed to NaCl + Ce(NO3 )3 and to
that can be assigned to the formation of a cerium hydroxide com- NaCl + MBT (Supplementary Fig. S3).
plex onto the zinc surface that protects the metal against corrosion An important aspect to notice is that the compounds formed on
(Table 4). Regarding the presence of MBT into the corrosive envi- top of the HDG steel substrate following its exposure, represented
ronment, it can be mentioned that it improves the protection of in the RCSLM images either as separated islets or as homogeneous
HDG steel against corrosion by the formation of an insoluble film films, exhibit different optical properties accounting for different
(Fig. 6c, Table 4). Finally, the weathered surface of the metal alloy RCSLM signal intensities in the respective regions. Further work on
denote that the corrosion procedure has not been effectively inhib- correlation on RCSLM and XRD data could thus lead to the develop-
ited in the presence of Na2 MoO4 into the corrosive environment; ment of computer vision methods for the automated identification
however, the EDS analysis demonstrates amounts of molybdenum of compounds in RCSLM images. Such methods could be merged
element that is ascribed to a possible formation of a complex on with complementary methods based on textural features extracted
the surface between zinc and molybdate ions (Supplementary Fig. using fractal analysis or wavelet packet transform [52,53] for devel-
S2h, Table 4). oping novel automated corrosion monitoring strategies based on
The detection of small carbon amounts in the EDS analyses of multiple-expert machine learning strategies [54]. It is also impor-
samples that do not include organic protective compounds into the tant to mention, that no damage in the form of microcracks can
NaCl solution, can be attributed either to the sensitivity of the EDAX be observed in all of the evaluated cases; however, low reflecting
GENESIS instrument or to contamination of the SEM chamber or to islets might also be attributed to pitting defects, and not only to
the formation of zinc carbonate due to the presence of carbonates a potential un-impactful presence of light absorbing compounds
in the water solution [34]. The preparation of the NaCl solutions which result from the chemical reactions that take place during
was accomplished by using distilled water that contains carbonates HDG panel exposure.
due to the dissolving and hydrolysis of atmospheric carbon dioxide. Digital microscopy was used for collecting additional opti-
However, it should be mentioned that the XRD measurements did cal information on the evaluated HDG Panels. The sample area
not demonstrate the formation of zinc carbonate compounds on imaged with digital microscopy was 9 times larger than the area
the HDG steel surface. imaged with RCSLM, the dimension of the observed field of view
Taking into account the aforementioned results, it may be being of 3 × 3 mm. The optical information contained in the digital
remarked that SEM represents a superb tool for characterizing microscopy images correlates well to the optical information of the
the morphology of the investigated HDG panels at micro-scale. RCSLM images, as can be observed in Fig. 8 (and Supplementary Fig.
However, the field of view available with SEM systems is limited, S4).
and thus gaining a broader perspective on the effects of exposing A last step for morphology assessment was imaging the whole
the HDG steel panels to the corrosive environment based on SEM exposed areas of the HDG steel panels by digital photography (Fig. 9
alone is difficult and prone to sampling errors. In this regard, three and Supplementary Fig. S5). Supplementary Fig. S5a demonstrates
additional optical tools have been employed to support this task: the surface of the metal unexposed to NaCl solution where no signs
RCSLM, digital microscopy, and digital photography. of corrosion appear. The area inside the circle of the plastic o-ring
In RCSLM a laser beam is used to illuminate the sample, and the is the exposed HDG steel panel area to the corrosive environment.
image is formed by collecting light that is reflected by the sam- It can be observed that all the metals depict craters or delami-
ple. Bright regions of RCSLM images correspond to highly reflective nated areas due to corrosion onset, apart from the metals that were
sample regions, while dark regions correspond to either imaging exposed to NaCl with the presence of 8HQ (Fig. 9b) and Ce(NO3 )3
areas with inclined geometries or sample regions where the num- (Fig. 9c) that illustrate a slightly green film due to the formation of a
ber of reflected photons is low due to absorption issues. In the complex between the 8HQ and the zinc of the panel outer layer and
case of the investigated HDG steel panels image regions with dif- a yellow film that is assigned to cerium hydroxide, respectively.
ferent signal intensities in the collected RCSLM micrographs most
likely correspond to sample areas with different chemical config- 3.3. X-Ray diffraction analysis
urations; these translate to different optical properties in terms of
reflectance/absorption. Such different chemical configurations cor- Panels exposed to 50 mM NaCl solution for 168 h were inves-
respond to the compounds resulted after exposing the HDG steel tigated by XRD at room temperature. Fig. 10a illustrates the
panels to the corrosive environment in the presence of the eval- diffraction pattern of the HDG steel panel before exposure in the
uated inhibitors. These compounds are discussed in section 3.3 NaCl corrosive environment. The collected XRD data reveal that the
which presents the results obtained with XRD. Fig. 7 depicts rep- d-spacing of the investigated HDG steel are in close agreement with
 I.A. Kartsonakis et al. / Corrosion Science 112 (2016) 289–307 301

Fig. 7. RCSLM images of the HDG panels a) unexposed to corrosive environment; and after exposure for 168 h to: b) NaCl + Ca3 (PO4 )2 , c) NaCl + 8HQ.

Fig. 8. Digital microscopy images of the HDG panels a) unexposed to corrosive environment; and after exposure for 168 h to: b) NaCl + Ca3 (PO4 )2 , c) NaCl + 8HQ.

Fig. 9. Digital photography images of the HDG panels after exposure for 168 h: a) NaCl, b) NaCl + 8HQ, c) NaCl + Ce(NO3 )3 .

the corresponding values given for zinc standard in International 50 mM + ATT comes along to the detection of XRD diffraction peaks
Centre for Diffraction Data (ICDD) card no. 04-0831. that are ascribed to the Zn substrate, simonkolleite and zincite
The XRD patterns collected for the HDG steel panels after their (Fig. 10b). Furthermore, the exposure of the HDG steel panels to
exposure to 50 mM NaCl solution for 168 h in the presence of the NaCl 50 mM + Ca3 (PO4 )2 leads to the verification of XRD diffrac-
evaluated corrosion inhibitors are also depicted in Fig. 10. The col- tion peaks that are attributed to the Zn substrate, zinc hydroxide,
lected XRD spectra reveal that the exposure of the HDG steel panels simonkolleite and zincite, Fig. 10c. In addition, XRD diffraction
to NaCl 50 mM + 8HQ leads to the detection of XRD diffraction peaks corresponding to the Zn substrate, simonkolleite and zincite
peaks corresponding to the Zn substrate, zinc hydroxide [Zn(OH)2 ], are ascertained after immersion of the HDG steel panels to NaCl
simonkolleite [Zn5 (OH)8 Cl2 ·H2 O], zincite (ZnO) and Zn-8HQ•2H2 O 50 mM + Ce(NO3 )3 (Fig. 10d). Likewise, the exposure of the HDG
(Fig. 10a). Moreover, the immersion of the HDG steel panels to NaCl steel panels to NaCl 50 mM + MBT leads to the detection of XRD
302 I.A. Kartsonakis et al. / Corrosion Science 112 (2016) 289–307

Fig. 11. Raman spectra of the HDG steel panels after their exposure to 50 mM NaCl
solution for 168 h in the presence or absence of corrosion inhibitors.

stretching vibration, a peak at 1177 cm−1 that is attributed to PO2

stretching vibration, two bands at 1248 and 1352 cm−1 that are
assigned to P = O stretching vibration and a peak at 1465 cm−1 that
is ascribed to P-OH vibration [64–66].
The spectrum of HDG-Ce(NO3 )3 illustrates a strong peak at
454 cm−1 that is ascribed to the Ce O symmetric vibration of CeO2
[67,68]. The spectrum of HDG-MBT exhibits bands at: 505 cm−1
attributed to the vibration of C C, 1020 cm−1 due to symmetri-
cal stretching of C-S, 1246 and 1373 cm−1 assigned to the rocking
vibrations of C H. The rocking vibrations at 611 and 868 cm−1 can
be attributed to the interaction between Zn surface and S atom of
MBT. The weak peaks at 730, 870 and 1459 cm−1 are assigned to the
Fig. 10. XRD patterns of the HDG steel panels a) unexposed to corrosive envi- N C C wagging, N C S scissoring and N C symmetrical vibra-
ronment; and after exposure for 168 h to: b) NaCl + 8HQ, c) NaCl + ATT, d) NaCl tions. The peak at 1373 cm−1 is ascribed to N H rocking vibration
+Ca3 (PO4 )2 , e) NaCl + Ce(NO3 )3 , f) NaCl + MBT, g) NaCl + Na2 MoO4 . [25]. In the spectrum of the sample HDG-Na2 MoO4 the charac-
teristic peaks of Mo = O symmetric vibration at 914 cm−1 due to
diffraction peaks corresponding to the Zn substrate and zincite the presence of Mo2 O7 −2 and Mo = O bending mode vibration at
(Fig. 10e). Finally, the immersion of the HDG steel panels to NaCl 319 cm−1 due to the MoO4 −2 are clearly observed [69]. Further-
50 mM + Na2 MoO4 comes along to the detection of XRD diffraction more, the band at 862 cm−1 is assigned to the vibration of MoO4 −2
peaks that are ascribed to the Zn substrate, simonkolleite, zincite tetrahedra [70].
and zinc hydroxide (Fig. 10f). According to the work of Autengruber
et al. [55] simonkolleite, which could be detected for all compounds 3.5. FT-IR spectroscopy analysis
except MBT, provides a high corrosion resistance of HDG steel due
to its barrier properties. Fig. 12 illustrates the FT-IR spectra of the HDG steel panels after
their exposure to 50 mM NaCl solution for 168 h in the presence
3.4. Raman spectroscopy analysis or absence of corrosion inhibitors. The spectrum of the HDG-blank
sample exhibits the characteristic peaks of zinc oxide and Zn(OH)2
The Raman spectra of the HDG steel panels after their exposure due to the exposure of HDG steel into the corrosive environment.
to 50 mM NaCl solution for 168 h in the presence or absence of cor- The peaks at 429 and 580 cm−1 are assigned to the stretching vibra-
rosion inhibitors are demonstrated in Fig. 11. It can be observed that tion of Zn-O [71,72]. The broad absorption band between 3300 and
all the spectra exhibit additional peaks compared to the spectrum of 3500 cm−1 as well as the peaks at 520, 530, 734, 831, 1035, 1080
the HDG-blank. The spectrum of the sample HDG-blank illustrates and 1380 cm−1 can be attributed to Zn(OH)2 as well as to the O H
a peak at 430 cm−1 that is assigned to the ZnO [56]. The spectrum of stretching and wagging vibrations of H2 O present in ZnO, indicat-
the sample HDG-8HQ reveals a band at 1385 cm−1 that corresponds ing the existence of water absorbed on the surface of HDG steel
to the C H bending and stretching of the quinoline ring, a peak at after its exposure to the corrosive environment [72,73].
157 cm−1 that is ascribed to the torsion of quinoline ring and peaks The spectrum of the HDG-8HQ sample reveals well defined
at 507, 740, 1070, 1115, 1143 and 1278 cm−1 due to the in plane bands of the ring stretching modes of 8HQ (491, 785, 823, 1059,
bending of the quinoline ring [57,58] [59] [60]. The spectrum of 1154, 1512, 1581, 1626 cm−1 ). The peaks at 1286 and 1273 cm−1
the sample HDG-ATT depicts a peak at 375 cm−1 due to the N C S correspond to the C N stretching vibrations. The bands at 1231
vibration of the ring and a peak at 248 cm−1 that is assigned to the and 1222 cm−1 refer to C O stretching vibrations, and at 575 and
C-S, N C C and C S N vibrations [61–63]. The HDG-Ca3 (PO4 )2 466 cm−1 correspond to in-plane and out of plane bending of
spectrum exhibits a band at 1046 cm−1 due to PO4 −3 asymmetric C O, respectively. The O H stretching vibrations are depicted in
 I.A. Kartsonakis et al. / Corrosion Science 112 (2016) 289–307 303

3.6. Corrosion inhibition mechanism

The inhibition mechanism against corrosion depends on the pH

of the system and also on the formation of complexes onto the
metal surface. As it was mentioned above, all the electrochemical
characterizations were performed in a pH range from 5.8 to 6.3.
Despite the fact that in acid aqueous solutions the attack of Zn on
the HDG steel surface is an acidic corrosion with hydrogen evolu-
tion, in aqueous solutions with pH higher than 4.5, the corrosion
rate is controlled by the quantity of diluted oxygen available and
its diffusion on the metal substrate [8]. The exposure of HDG steel
to sodium chloride solution results in the dissolution of zinc due
to the corrosion of the galvanized layer creating local acidification,
according to the following reactions [31,32,34,83]:

Zn → Zn+2 + 2e− (4)

Zn+2 + nH2 O → [Zn(OH)n ](2−n) + nH+ (5)

The creation of local acidification (chemical reaction 5) results

Fig. 12. FT-IR spectra of the HDG steel panels after their exposure to 50 mM NaCl
in the initiation of pitting corrosion [34]. Moreover, taking into con-
solution for 168 h in the presence or absence of corrosion inhibitors. sideration that ZnO is thermodynamically stable in pH range from
6 to 12 [84], it may be remarked that the ZnO, which had been
created on the HDG steel surface as a result of its interaction with
atmosphere, could be dissolved as stated in the following reactions:
region 2800–3300 cm−1 [74]. The spectrum of the sample HDG-
ZnO + 2H+ → Zn+2 + H2 O (6)
ATT depicts peaks at 1612, 3268 and 3330 cm−1 due to the NH2
stretching. The band at 3168 cm−1 is ascribed to the ring N H On the other hand, the reduction reactions that occur in acidic
stretching of the thione tautomer. The peaks at 1602 and 752 cm−1 aqueous solution are the reduction of water and the reduction of
are assigned to the in-plane and out-of-plane bending of the amine oxygen [32,33]:
group, respectively. The band at 1502 cm−1 is ascribed to the C N
vibration. The broad band at 700–770 cm−1 is attributed to the C = S 2H3 O+ + 2e− → H2 + 2H2 O (7)
stretching mode. The bands at 428 and 618 cm−1 are assigned to the + −
O2 + 2H + 2e → H2 O2 (8)
C-S and N C S vibrations [61,62].
+ −
In the spectra of HDG-Ca3 (PO4 )2 sample the bands in the H2 O2 + 2H + 2e → 2H2 O (9)
550–700 cm−1 region as well as the peaks at 430 and 492 cm−1
could be assigned to symmetric vibration of PO4 −3 units [75,76]. However, the presence of chloride anions into the solution
Furthermore, the peaks from 1000 to 1100 cm−1 are attributed to increases the solubility of Zn and results in the formation on the
the P-O stretching vibration. The asymmetric vibration of P-O-P metal surface of the insoluble corrosion product simonkolleite
groups are depicted at 935 cm−1 . Moreover, the bands from 1200 [Zn5 C12 (OH)8 ], according to the following reaction [32,85]:
to 1320 cm−1 are ascribed to the P = O stretching vibrations [76,77]. Zn(OH)2 + 4Zn+2 + 6OH− + 2C1− → Zn5 C12 (OH)8 (10)
The spectrum of the HDG-Ce(NO3 )3 sample illustrates an absorp-
tion band at 574 cm−1 that is attributed to Zn O Ce species [71]. Furthermore, due to the fact that the chloride concentration
The peak at 451 cm−1 is assigned to Ce-O stretching vibration [78]. is 50 mM > 10 mM as well as the pH is below 7, two additional
The bands at 650, 725 and 833 cm−1 as well as the peak at 897 cm−1 hydroxide chlorides (6Zn(OH)2 ·ZnC12 and 4Zn(OH)2 ·ZnC12 ) might
are attributed to Ce(OH)2+ [79]. be formed as a result of the following reactions [86]:
The spectrum of the HDG-MBT sample demonstrates the char-
Zn(OH)2 + Zn+2 + 2C1− → 6Zn(OH)2 ·ZnC12 (11)
acteristic peaks of MBT. In details, the peak at 428 cm−1 is ascribed
to C C torsion vibration, at 499 cm−1 is due to C C S bending 7Zn +2 − −
+ 12OH + 2C1 → 6Zn(OH)2 ·ZnC12 (12)
vibration, at 611 cm−1 is assigned to C C C bending vibration,
+2 −
at 723 cm−1 is attributed to C C torsion vibration, at 749 cm−1 4ZnO + 4H2 O + Zn + 2C1 → 4Zn(OH)2 ·ZnC12 (13)
is ascribed to C H and N H wagging vibrations, at 849 cm−1 is 5Zn +2 − −
+ 12OH + 2C1 → 4Zn(OH)2 ·ZnC12 (14)
due to C H wagging vibration, at 1015 cm−1 is assigned to out
of phase stretching vibration of C S in S C S and at 1084 cm−1 In general, the addition of Ca3 (PO4 )2 into the solution results in
is attributed to in phase stretching vibration of C S in S C S as the decreament of Zn solubility and the formation of zinc phosphate
well as to C C C bending and C = S stretching vibrations. More- film, which can be protective against corrosion. This film is created
over, the band at 1136 cm−1 is assigned to C H bending as well due to the precipitation of phophates from the solution onto the
as C-S stretching vibrations, at 1159 cm−1 is due to C H bending metal surface and the simultaneous diffusion of Zn+2 through the
vibration, at 1241 cm−1 is ascribed to C H bending as well as C N precipitated film [87] (Fig. 13), according to the reactions:
stretching vibrations, at 1312 cm−1 is attributed to C N stretching
Ca3 (PO4 )2 → 3Ca+2 + 2PO4 −3 (15)
as well as C H and N H bending vibrations and at 1371 cm−1 is
ascribed to C C and C N stretching vibrations. Finally, the peak at + −3 −2
H + PO4 ↔ HPO4 (16)
1453 cm−1 is due to C C and C N stretching as well as C H bending
+ −2 −
vibrations and at 1588 cm−1 is assigned to C C stretching vibration H + HPO4 ↔ H2 PO4 (17)
[80,81]. The spectrum of the HDG-Na2 MoO4 sample depicts peaks
Zn+2 + 2H2 PO4 − ↔ Zn(H2 PO4 )2 (18)
at 410, 727, 792, 845 and 943 cm−1 due to the MoO4 −2 . The peak at
1047 is assigned to M O stretching vibration [75,79,82]. 3Zn(H2 PO4 )2 ↔ Zn3 (PO4 )2 ·4H2 O + 4H3 PO4 (19)
304 I.A. Kartsonakis et al. / Corrosion Science 112 (2016) 289–307

Fig. 13. Schematic representation of formation of complex films onto the HDG steel between the zinc and the inhibitors.

However, taking into account that the experiments were per- is precipitated onto the zinc surface, acting as a protective coat-
formed in a pH range from 5.8 to 6.3, it may be remarked that ing against zinc corrosion. Regarding the reaction (8) the oxygen
these conditions do not induce the comprehensive and complete reduction produces H2 O2 (reaction 8) that can be further reduced
deposition of the aforementioned film [88] resulting in inadequate resulting in the local formation of OH− as well as to the oxidation of
protection as the electrochemical tests reveal in section 3.1. More- Ce+3 to Ce+4 . Consequently, Ce(OH)4 is formed (Fig. 13) according
over, it can be remarked that the calcium cations do not participate to the following reactions:
in the formation of a protective coating against corrosion onto the (21) H2 O2 + 2e− → 2OH−
zinc surface because the reaction of Ca(Zn(OH)3 )2 ·2H2 O formation:
Ce+3 + OH− + ½H2 O2 → Ce(OH)2 +2 (22)
2Zn(OH)4 −2 + Ca+2 +2H2 O → Ca(Zn(OH)3 )2 ·2H2 O + 2OH− (20) Ce(OH)2 +2 −
+ 2OH → Ce(OH)4 (23)

takes place only in higly alkaline environment (pH ≥ 12.5) where As reported by the literature the precipitation of Ce(OH)4 can
the Zn(OH)4 −2 can be formed [89]. Therefore, it can be noticed that result even at pH = 2.9 [90,91]. Thus, it may be mentioned that
the presence of Ca3 (PO4 )2 into the NaCl solution does not protect the addition of cerium nitrate into the NaCl solution increases the
effectively the HDG steel against corrosion; this can be also visually protection of HDG steel against corrosion.
observed in the RCSLM micrograph in Fig. 7b. The inhibition of HDG steel corrosion as a result of the presence
By estimating the electrochemical characterizations, it can be of 8HQ, ATT and MBT into the corrosive environment is ascribed
clearly observed that he addition of cerium nitrate into the NaCl to the adsorption of the organic molecules onto the anodic sites
solution increases the protection of HDG steel against corrosion of the metal surface resulting in the formation of a film complex
by the formation of a cerium hydroxide film onto the zinc surface. between the organic compounds and the zinc cations [92] (Fig. 13)
Nonetheless, it should be mentioned that the pH of the used cor- that inhibits the anodic reactions [83]. The film complex formation
rosive solutions is not alkaline [90], consequently, the formation is based on the electron donation of nitrogen and sulfur atoms from
of Ce(OH)3 is not stable. Nevertheless, a film based on Ce(OH)4 8HQ, ATT and MBT to the divalent zinc cation. It may be remarked
 I.A. Kartsonakis et al. / Corrosion Science 112 (2016) 289–307 305

that despite the fact that the divalent zinc cation is not strictly a − European Social Fund (ESF) and National Funds through the
transition element as it has filled its 3d10 orbitals, its atomic struc- Operational Program “Education and Lifelong Learning” (ELL). Con-
ture is similar with those of transition metals and therefore, creates tribution of S.G. Stanciu, A.A. Matei and R. Hristu was supported
complexes with electron donor atoms such as sulfur and nitrogen by the PN-II-RU-TE-2014-4-1803 (MICRONANO) Research Grant,
[93]. Consequently, it can be noticed that the presence of 8HQ, funded by the Romanian Executive Agency for Higher Education,
ATT and MBT into the NaCl solution enhances the protection of Research, Development and Innovation Funding (UEFISCDI).
HDG steel against corrosion. Visual cues that reflect this situation
can be observed in the RCSLM and digital microscopy images in Appendix A. Supplementary data
Figs. 7 and 8 (and Supplementary Figs. S3 and S4).
The influence of sodium molybdate to the corrosion protection Supplementary data associated with this article can be found,
of HDG steel is assigned to the formation of a complex between the in the online version, at http://dx.doi.org/10.1016/j.corsci.2016.07.
molybdate ions and the zinc. When the pH of an aqueous molybdate 030.
solution is roughly 6, the MoO4 −2 ions condense into heptamolyb-
date (paramolybdate) polyanion [94] (Fig. 13), according to the
References
following reaction:

7[MoO4 ]−2 + 8H+ ↔ [Mo7 O24 ]−6 + 4H2 O (24) [1] F.C. Porter, Corrosion Resistance of Zinc and Zinc Alloys, First edition, Marcel
Dekker, Inc, New York, 1994.
[2] A.P. Yadav, A. Nishikata, T. Tsuru, Electrochemical impedance study on
Furthermore, polyanions such as [Mo6 O21 ]−6 , [Mo8 O26 ]−4 , galvanized steel corrosion under cyclic wet-dry conditions-influence of time
[MoO3 (OH)]− , [MoO(OH)5 ]− and [(OH)4 OMo-O-MoO(OH)4 ]−2 can of wetness, Corros. Sci. 46 (2004) 169–181.
also exist [16,95]. Subsequently, taking into account that the afore- [3] A.P. Yadav, A. Nishikata, T. Tsuru, Degradation mechanism of galvanized steel
in wet-dry cyclic environment containing chloride ions, Corros. Sci. 46 (2004)
mentioned experiments were carried out using corrosive solutions 361–376.
with pH ranging from 5.8 to 6.3, the polymolybdate ions could [4] T.E. Graedel, Corrosion mechanisms for zinc exposed to the atmosphere, J.
be adsorbed onto the positive charged metal surface of HDG steel Electrochem. Soc. 136 (1989) 193C.
[5] F.C. Porter, General considerations: IV. How zinc can protect steel, in:
resulting in its protection of chloride penetration [16]. Thus, it may Corrosion Resistance of Zinc and Zinc Alloys, First edition, Marcel Dekker, Inc,
be remarked that the addition of sodium molybdate into the NaCl New York, 1994, pp. 83–87.
solution enhances the protection of HDG steel against corrosion. [6] C. Andrade, C. Alonso, Corrosion of galvanized steel in the presence of
chlorieds, in: S.R. Yeomans (Ed.), Galvanized Steel Reinforcement in Concrete,
Elsevier Ltd., Amsterdam, 2004, pp. 132–141.
4. Conclusion [7] A.P. Yadav, H. Katayama, K. Noda, H. Masuda, A. Nishikata, T. Tsuru, Effect of
Fe-Zn alloy layer on the corrosion resistance of galvanized steel in chloride
containing environments, Corros. Sci. 49 (2007) 3716–3731.
In this study, the protective capabilities of six potential [8] F.C. Porter, Resistance to aqueous corrosion: II. Influence of types of water, in:
inhibitors against HDG steel corrosion was evaluated. The HDG Corrosion Resistance of Zinc and Zinc Alloys, First edition, Marcel Dekker Inc,
New York, 1994, pp. 264–281.
steel panels were treated with 8HQ, ATT, Ca3 (PO4 )2 , Ce(NO3 )3 , MBT
[9] M.F. Montemor, M.G.S. Ferreira, Cerium salt activated nanoparticles as fillers
and Na2 MoO4 in 50 mM NaCl, at room temperature, in a pH range for silane films: evaluation of the corrosion inhibition performance on
from 5.8 to 6.3. The electrochemical characterizations reveal that galvanised steel substrates, Electrochim. Acta 52 (2007) 6976–6987.
[10] A.M. Cabral, W. Trabelsi, R. Serra, M.F. Montemor, M.L. Zheludkevich, M.G.S.
8HQ, ATT, Ce(NO3 )3 , MBT and Na2 MoO4 can be considered as cor-
Ferreira, The corrosion resistance of hot dip galvanised steel and AA2024-T3
rosion inhibitors of HDG steel against corrosion as they reduce the pre-treated with bis-[triethoxysilylpropyl] tetrasulfide solutions doped with
anodic and/or cathodic reactions of zinc. Moreover, the presence Ce(NO3 )3 , Corros. Sci. 48 (2006) 3740–3758.
of Ce(NO3 )3 into the corrosive environment exhibits the best cor- [11] G. Kong, J. Lu, H. Wu, Post treatment of silane and cerium salt as chromate
replacers on galvanized steel, J. Rare Earths 27 (2009) 164–168.
rosion protection as it illustrates the highest impedance modulus [12] F. Deflorian, M. Fedel, S. Rossi, P. Kamarchik, Evaluation of mechanically
and Rp value as the immersion time elapses. On the other hand, the treated cerium (IV) oxides as corrosion inhibitors for galvanized steel,
addition of Ca3 (PO4 )2 does not inhibit the corrosion process of HDG Electrochim. Acta 56 (2011) 7833–7844.
[13] K. Aramaki, Treatment of zinc surface with cerium(III) nitrate to prevent zinc
steel. XRD measurements and SEM characterization depict that the corrosion in aerated 0.5 M NaCl, Corros. Sci. 43 (2001) 2201–2215.
exposure of HDG steel to corrosive environment in the presence [14] K. Aramaki, The inhibition effects of cation inhibitors on corrosion of zinc in
or absence of inhibitors, results in the formation of various oxide, aerated 0.5 M NaCl, Corros. Sci. 43 (2001) 1573–1588.
[15] M.A. Arenas, J.J. de Damborenea, Growth mechanisms of cerium layers on
hydroxide and hydroxide-chloride compounds on the zinc surface galvanised steel, Electrochim. Acta 48 (2003) 3693–3698.
such as zinc hydroxide, simonkolleite and zincite. Considering the [16] V. Shkirskiy, P. Keil, H. Hintze-Bruening, F. Leroux, T. Stimpfling, D. Dragoe, K.
electrochemical, spectroscopy and morphology characterizations, Ogle, P. Volovitch, MoO4 2− as a soluble inhibitor for Zn in neutral and alkaline
solutions, Corros. Sci. 99 (2015) 31–41.
the corrosion protection mechanisms of HDG steel can be attributed
[17] G. Kong, J. Lu, S. Zhang, C. Che, H. Wu, A comparative study of
to the layers that are formed onto the metal surface due to the pres- molybdate/silane composite films on galvanized steel with different
ence of the inhibitors preventing chlorides penetration. In the case treatment processes, Surf. Coat. Technol. 205 (2010) 545–550.
[18] Y. Hamlaoui, L. Tifouti, F. Pedraza, Corrosion behaviour of
of 8HQ, ATT, and MBT, complex films are formed onto the HDG
molybdate-phosphate-silicate coatings on galvanized steel, Corros. Sci. 51
steel between the zinc and the nitrogen or sulfur atoms of the (2009) 2455–2462.
organic compounds; XRD measurements reveal the formation of [19] A.M. Simões, J. Torres, R. Picciochi, J.C.S. Fernandes, Corrosion inhibition at
Zn-8HQ·2H2 O in the presence of 8HQ. Furthermore, the addition of galvanized steel cut edges by phosphate pigments, Electrochim. Acta 54
(2009) 3857–3865.
Ce(NO3 )3 , Ca3 (PO4 )2 and Na2 MoO4 into the corrosive environment [20] C.-Y. Tsai, J.-S. Liu, P.-L. Chen, C.-S. Lin, Effect of Mg2+ on the microstructure
results in the formation of cerium hydroxide, zinc phosphate and and corrosion resistance of the phosphate conversion coating on hot-dip
complex film between zinc and molybdate ions, respectively. galvanized sheet steel, Corros. Sci. 52 (2010) 3907–3916.
[21] K.-h. Wu, L.-q. Zhu, W.-p. Li, H.-c. Liu, Effect of Ca2+ and Mg2+ on corrosion and
scaling of galvanized steel pipe in simulated geothermal water, Corros. Sci. 52
Acknowledgments (2010) 2244–2249.
[22] I.M. Zin, S.B. Lyon, V.I. Pokhmurskii, Corrosion control of galvanized steel using
a phosphate/calcium ion inhibitor mixture, Corros. Sci. 45 (2003) 777–788.
C.A. Charitidis and I.A. Kartsonakis acknowledge the financial [23] V. Shkirskiy, P. Keil, H. Hintze-Bruening, F. Leroux, F. Brisset, K. Ogle, P.
support of the research project “SELF-HEALING NANOMATERIALS Volovitch, The effects of l-cysteine on the inhibition and accelerated
dissolution processes of zinc metal, Corros. Sci. 100 (2015) 101–112.
FOR PROTECTION OF METAL ALLOYS”, CODE 3456& ACRONYM: [24] A.-R. El-Sayed, U. Harm, K.-M. Mangold, W. Fürbeth, Protection of galvanized
“SHELL” for the National Action “ARISTEIA II” of the Operational steel from corrosion in NaCl solution by coverage with phytic acid SAM
Program “Education and Lifelong Learning”, by the European Union modified with some cations and thiols, Corros. Sci. 55 (2012) 339–350.
306 I.A. Kartsonakis et al. / Corrosion Science 112 (2016) 289–307

[25] H. Yang, Y. Sun, J. Ji, W. Song, X. Zhu, Y. Yao, Z. Zhang, [52] R.M. Pidaparti, B.S. Aghazadeh, A. Whitfield, A.S. Rao, G.P. Mercier,
2-Mercaptobenzothiazole monolayers on zinc and silver surfaces for Classification of corrosion defects in NiAl bronze through image analysis,
anticorrosion, Corros. Sci. 50 (2008) 3160–3167. Corros. Sci. 52 (2010) 3661–3666.
[26] L.M. Baugh, Corrosion and polarization characteristics of zinc in neutral—acid [53] R.M. Pidaparti, B. Hinderliter, D. Maskey, Evaluation of corrosion growth on
media—I. Pure zinc in solutions of various sodium salts, Electrochim. Acta 24 SS304 based on textural and color features from image analysis, ISRN Corros.
(1979) 657–667. 2013 (2013) 1–7.
[27] X.G. Zhang, Zinc, in: R.W. Renie (Ed.), Uhlig’s Corrosion Handbook, John Wiley [54] V.C. Raykar, S. Yu, L.H. Zhao, A. Jerebko, C. Florin, G.H. Valadez, L. Bogoni, L.
& Sons Inc, New York, 2000, pp. 887–904. Moy, Supervised Learning from Multiple Experts: Whom to Trust When
[28] S.J. Kelly RG, D.W. Shoesmith, R.G. Buchheit, The polarization resistance Everyone Lies a Bit, Proceedings of the 26th Annual International Conference
method for determination of instantaneous corrosion rates, in: P.A. on Machine Learning, 2009, pp. 889–896.
Schweitzer (Ed.), Electrochemical Techniques in Corrosion Science and [55] R. Autengruber, G. Luckeneder, A.W. Hassel, Corrosion of press-hardened
Engineering, Marcel Dekker Inc, New York, 2002, pp. 125–150. galvanized steel, Corros. Sci. 63 (2012) 12–19.
[29] M.A. Migahed, A.M. Al-Sabagh, E.A. Khamis, E.G. Zaki, Quantum chemical [56] A. Souissi, M. Amlouk, H. Khemakhem, S. Guermazi, Deep analysis of Raman
calculations, synthesis and corrosion inhibition efficiency of spectra of ZnO:Mo and ZnO:In sprayed thin films along with LO and TA + LO
ethoxylated-[2-(2-{2-[2-(2-benzenesulfonylamino-ethylamino)- bands investigation, Superlattices Microstruct. 92 (2016) 294–302.
ethylamino]-ethylamino}-ethylamino)-ethyl]-4-alkyl-benzenesulfonamide [57] G.C. Kordas, A.C. Balaskas, I.A. Kartsonakis, E.K. Efthimiadou, A Raman study of
on API X65 steel surface under H2 S environment, J. Mol. Liq. 212 (2015) 8-hydroxyquinoline release from loaded TiO2 nanocontainer, Int. J. Struct.
360–371. Integr. 4 (2013) 121–126.
[30] Y. Zhu, M.L. Free, G. Yi, Electrochemical measurement, modeling, and [58] N. Muraki, M. Yoshikawa, Characterization of the interface between metal
prediction of corrosion inhibition efficiency of ternary mixtures of and tris (8-hydroxyquinoline) aluminum using surface-enhanced Raman
homologous surfactants in salt solution, Corros. Sci. 98 (2015) 417–429. scattering with glass cap encapsulation, Chem. Phys. Lett. 496 (2010) 91–94.
[31] P. Delahay, M. Pourbaix, P. Van Rysselberghe, Potential-pH diagram of zinc [59] S. Yurdakul, Fourier transform infrared and Raman spectroscopic studies on
and its applications to the study of zinc corrosion, J. Electrochem. Soc. 98 8-hydroxyquinoline metal(II) tetracyanonickelate complexes, J. Mol. Struct.
(1951) 101. 412 (1997) 231–237.
[32] X.G. Zhang, Electrochemical thermodynamics and kinetics, in: X.G. Zhang [60] N. Prabavathi, A. Nilufer, V. Krishnakumar, Spectroscopic (FT-IR, FT-Raman UV
(Ed.), Corrosion and Electrochemistry of Zinc, Springer Science+ Business and NMR) investigation, conformational stability, NLO, properties,
Media, LLC, New York, 1996, pp. 19–64. HOMO-LUMO and NBO analysis of hydroxyquinoline derivatives by density
[33] S.J. Kelly R.G, D.W. Shoesmith, R.G. Buchheit, Electrochemical functional theory calculations, Spectrochim. Acta Part A Mol. Biomol.
thermodynamics and kinetics of relevance to corrosion, in: P.A. Schweitzer Spectrosc. 114 (2013) 449–474.
(Ed.), Electrochemical Techniques in Corrosion Science and Engineering, [61] A.G.S. Prado, J.A.A. Sales, R.M. Carvalho, J.C. Rubim, C. Airoldi, Immobilization
Marcel Dekker, Inc, New York, 2002, pp. 9–54. of 5-amino-1,3,4-thiadiazole-thiol onto silica gel surface by heterogeneous
[34] X.G. Zhang, Corrosion forms, in: X.G. Zhang (Ed.), Corrosion and and homogeneous routes, J. Non-Cryst. Solids 333 (2004) 61–67.
Electrochemistry of Zinc, New York, Springer Science+ Business Media, LLC, [62] E.E. Chufán, J.C. Pedregosa, J. n. Borrás, Spectroscopic behaviour of metal–drug
1996, pp. 183–240. complexes. Infrared spectra of Cu(II) complexes with
[35] D.A. Shifler, D.M. Aylor, Testing in environments-seawater, in: R. Baboian 5-amino-1,3,4-thiadiazole-2-thiol (Hatm), Vib. Spectrosc. 15 (1997) 191–199.
(Ed.), Corrosion Tests and Standards Application and Interpretation, ASTM [63] A.E. Ozel, S.K. Gunduz, S. Celik, S. Akyuz, Structural and vibrational study on
International, Baltimore, 2005, pp. 362–379. monomer and dimer forms and water clusters of acetazolamide, J. Spectrosc.
[36] M. Mouanga, P. Berçot, Comparison of corrosion behaviour of zinc in NaCl and 2013 (2013) 1–13.
in NaOH solutions; Part II: Electrochemical analyses, Corros. Sci. 52 (2010) [64] R.L. Frost, An infrared and Raman spectroscopic study of natural zinc
3993–4000. phosphates, Spectrochim. Acta Part A Mol. Biomol. Spectrosc. 60 (2004)
[37] A.C. Bastos, M.G.S. Ferreira, A.M. Simões, Comparative electrochemical studies 1439–1445.
of zinc chromate and zinc phosphate as corrosion inhibitors for zinc, Prog. [65] D.A. Magdas, N.S. Vedeanu, D. Toloman, Study on the effect of vanadium oxide
Org. Coat. 52 (2005) 339–350. in calcium phosphate glasses by Raman, IR and UV–vis spectroscopy, J.
[38] Y. Hamlaoui, F. Pedraza, L. Tifouti, Corrosion monitoring of galvanised Non-Cryst. Solids 428 (2015) 151–155.
coatings through electrochemical impedance spectroscopy, Corros. Sci. 50 [66] S. Kaabi Falahieh Asl, S. Nemeth, M.J. Tan, Mechanism of calcium phosphate
(2008) 1558–1566. deposition in a hydrothermal coating process, Surf. Coat. Technol. 270 (2015)
[39] S.V. Lamaka, M.L. Zheludkevich, K.A. Yasakau, M.F. Montemor, M.G.S. Ferreira, 197–205.
High effective organic corrosion inhibitors for 2024 aluminium alloy, [67] Y. Hamlaoui, L. Tifouti, C. Remazeilles, F. Pedraza, Cathodic electrodeposition
Electrochim. Acta 52 (2007) 7231–7247. of cerium based oxides on carbon steel from concentrated cerium nitrate. Part
[40] M.F. Montemor, A.M. Simões, M.G.S. Ferreira, Composition and corrosion II: Influence of electrodeposition parameters and of the addition of PEG,
behaviour of galvanised steel treated with rare-earth salts: the effect of the Mater. Chem. Phys. 120 (2010) 172–180.
cation, Prog. Org. Coat. 44 (2002) 111–120. [68] M. Faisal, A.A. Ismail, A.A. Ibrahim, H. Bouzid, S.A. Al-Sayari, Highly efficient
[41] M. Olivier, A. Lanzutti, C. Motte, L. Fedrizzi, Influence of oxidizing ability of the photocatalyst based on Ce doped ZnO nanorods: controllable synthesis and
medium on the growth of lanthanide layers on galvanized steel, Corros. Sci. enhanced photocatalytic activity, Chem. Eng. J. 229 (2013) 225–233.
52 (2010) 1428–1439. [69] I.D. Dobrea, C.E. Ciocan, E. Dumitriu, M.I. Popa, E. Petit, V. Hulea, Raman
[42] A. Lutz, O. van den Berg, J. Wielant, I. De Graeve, H. Terryn, A multiple-action spectroscopy—Useful tool for studying the catalysts derived from Mo and
self-healing coating, Front. Mater. 2 (2016). V-oxyanion-intercalated layered double hydroxides, Appl. Clay Sci. 104
[43] M.F. Montemor, A.M. Cabral, M.L. Zheludkevich, M.G.S. Ferreira, The corrosion (2015) 205–210.
resistance of hot dip galvanized steel pretreated with bis-functional silanes [70] V.M. Anandakumar, M.A. Khadar, Synthesis, characterization and optical
modified with microsilica, Surf. Coat. Technol. 200 (2006) 2875–2885. properties of nanocrystalline lead molybdate, Phys. Status Solidi (a) 205
[44] M.F. Montemor, W. Trabelsi, M. Zheludevich, M.G.S. Ferreira, Modification of (2008) 2666–2672.
bis-silane solutions with rare-earth cations for improved corrosion protection [71] O. Bechambi, A. Touati, S. Sayadi, W. Najjar, Effect of cerium doping on the
of galvanized steel substrates, Prog. Org. Coat. 57 (2006) 67–77. textural, structural and optical properties of zinc oxide: role of cerium and
[45] C.H. Hsu, F. Mansfeld, Technical note: concerning the conversion of the hydrogen peroxide to enhance the photocatalytic degradation of endocrine
constant phase element parameter Y0into a capacitance, Corrosion 57 (2001) disrupting compounds, Mater. Sci. Semicond. Process. 39 (2015) 807–816.
747–748. [72] J. Kasperek, D. Verchere, D. Jacquet, N. Phillips, Analysis of the corrosion
[46] G. Kong, L. Lingyan, J. Lu, C. Che, Z. Zhong, Corrosion behavior of products on galvanized steels by FTIR spectroscopy, Mater. Chem. Phys. 56
lanthanum-based conversion coating modified with citric acid on hot dip (1998) 205–213.
galvanized steel in aerated 1 M NaCl solution, Corros. Sci. 53 (2011) [73] R.S. Jayasree, V.P. Mahadevan Pillai, V.U. Nayar, I. Odnevall, G. Keresztury,
1621–1626. Raman and infrared spectral analysis of corrosion products on zinc
[47] J.R. Macdonald, E. Barsoukov, Fundamentals of impedance spectroscopy, in: E. NaZn4 Cl(OH)6 SO4 ·6H2 O and Zn4 Cl2 (OH)4 SO4 ·5H2 O, Mater. Chem. Phys. 99
Barsoukov, J.R. Macdonald (Eds.), Impedance Spectroscopy Theory, (2006) 474–478.
Experiment, and Applications, A John Wiley & Sons, Inc., Publication, USA, [74] P. Pimchan, N. Khaorapapong, M. Sohmiya, M. Ogawa, In situ complexation of
2005, pp. 13–20. 8-hydroxyquinoline and 4,4’-bipyridine with zinc(II) in the interlayer space of
[48] K.S. Cole, Dispersion and absorption in dielectrics II. Direct current montmorillonite, Appl. Clay Sci. 95 (2014) 310–316.
characteristics, J. Chem. Phys. 10 (1942) 98. [75] M. Szumera, Structural investigations of silicate-phosphate glasses containing
[49] Y. Zhu, M.L. Free, G. Yi, The effects of surfactant concentration, adsorption, MoO3 by FTIR, raman and 31 P MAS NMR spectroscopies, Spectrochim. Acta
aggregation, and solution conditions on steel corrosion inhibition and Part A Mol. Biomol. Spectrosc. 130 (2014) 1–6.
associated modeling in aqueous media, Corros. Sci. 102 (2016) 233–250. [76] D. Rair, A. Rochdi, A. Majjane, T. Jermoumi, A. Chahine, M.E. Touhami,
[50] S. Shen, Y. Zuo, X. Zhao, The effects of 8-hydroxyquinoline on corrosion Synthesis and study by FTIR, 31P NMR and electrochemical impedance
performance of a Mg-rich coating on AZ91D magnesium alloy, Corros. Sci. 76 spectroscopy of vanadium zinc phosphate glasses prepared by sol-gel route, J.
(2013) 275–283. Non-Cryst. Solids 432 (2016) 459–465.
[51] D.A. Dragatogiannis, E.P. Koumoulos, I.A. Kartsonakis, C.A. Charitidis, [77] C. Renuka, A.B. Shinde, P.S.R. Krishna, C.N. Reddy, Structural analysis of
Deformation mechanism during nanoindentation creep and corrosion molybdo-zinc-phosphate glasses: neutron scattering FTIR, Raman scattering,
resistance of Zn, Int. J. Struct. Integr. 7 (2016) 47–69. MAS NMR studies, J. Mol. Struct. 1118 (2016) 83–90.
 I.A. Kartsonakis et al. / Corrosion Science 112 (2016) 289–307 307

[78] A. Arumugam, C. Karthikeyan, A.S. Haja Hameed, K. Gopinath, S. Gowri, V. [86] X.G. Zhang, Corrosion products, in: X.G. Zhang (Ed.), Corrosion and
Karthika, Synthesis of cerium oxide nanoparticles using Gloriosa superba L. Electrochemistry of Zinc, Springer Science+ Business Media, LLC, New York,
leaf extract and their structural, optical and antibacterial properties, Mater. 1996, pp. 157–182.
Sci. Eng. C Mater. Biol. Appl. 49 (2015) 408–415. [87] X.G. Zhang, Properties, products and processes, in: X.G. Zhang (Ed.), Corrosion
[79] M.A. Patel, B.A. Bhanvase, S.H. Sonawane, Production of cerium zinc and Electrochemistry of Zinc, Springer Science+ Business Media, LLC, New
molybdate nano pigment by innovative ultrasound assisted approach, York, 1996, pp. 1–18.
Ultrason. Sonochem. 20 (2013) 906–913. [88] X.G. Zhang, Corrosion in waters and aqueous solutions, in: X.G. Zhang (Ed.),
[80] I.A. Kartsonakis, A.C. Balaskas, G.C. Kordas, Influence of cerium molybdate Corrosion and Electrochemistry of Zinc, Springer Science+ Business Media,
containers on the corrosion performance of epoxy coated aluminium alloys LLC, New York, 1996, pp. 283–304.
2024-T3, Corros. Sci. 53 (2011) 3771–3779. [89] Z.Q. Tan, C.M. Hansson, Effect of surface condition on the initial corrosion of
[81] A.K. Rai, R. Singh, K.N. Singh, V.B. Singh, FTIR, Raman spectra and ab initio galvanized reinforcing steel embedded in concrete, Corros. Sci. 50 (2008)
calculations of 2-mercaptobenzothiazole, Spectrochim. Acta Part A Mol. 2512–2522.
Biomol. Spectrosc. 63 (2006) 483–490. [90] S.A. Hayes, P. Yu, T.J. O’Keefe, M.J. O’Keefe, J.O. Stoffer, The phase stability of
[82] B.A. Bhanvase, M.A. Patel, S.H. Sonawane, Kinetic properties of layer-by-layer cerium species in aqueous systems, J. Electrochem. Soc. 149 (2002) C623.
assembled cerium zinc molybdate nanocontainers during corrosion [91] B.R.W. Hinton, L. Wilson, The corrosion inhibition of zinc with cerous
inhibition, Corros. Sci. 88 (2014) 170–177. chloride, Corros. Sci. 29 (1989) 967–985.
[83] M. Taryba, S.V. Lamaka, D. Snihirova, M.G.S. Ferreira, M.F. Montemor, W.K. [92] A.C. Bastos, M.L. Zheludkevich, M.G.S. Ferreira, A SVET investigation on the
Wijting, S. Toews, G. Grundmeier, The combined use of scanning vibrating modification of zinc dust reactivity, Prog. Org. Coat. 63 (2008) 282–290.
electrode technique and micro-potentiometry to assess the self-repair [93] M. Gerloch, E.G. Constable, Complex stability and energetics, in: M. Gerloch,
processes in defects on smart coatings applied to galvanized steel, E.G. Constable (Eds.), Transition Metal Chemistry, Verlagsgesellschaft mbH,
Electrochim. Acta 56 (2011) 4475–4488. Weinheim VCH, New York, 1994, pp. 145–166.
[84] X.G. Zhang, Electrochemistry of zinc oxide, in: X.G. Zhang (Ed.), Corrosion and [94] Molybdenum and tungsten, in: N.N. Greenwood, A. Earnshaw (Eds.),
Electrochemistry of Zinc, Springer Science+ Business Media, LLC, New York, Chemistry of the Elements, Butterworth-Heinemann, Oxford, 1998, pp.
1996, pp. 93–124. 1022–1039.
[85] W. Miao, I.S. Cole, A.K. Neufeld, S. Furman, Pitting Corrosion of Zn and Zn-Al [95] G. Mu, X. Li, Q. Qu, J. Zhou, Molybdate and tungstate as corrosion inhibitors for
Coated Steels in pH 2–12 NaCl Solutions, J. Electrochem. Soc. 154 (2007) C7. cold rolling steel in hydrochloric acid solution, Corros. Sci. 48 (2006) 445–459.