Nucleation and growth behavior of coating film on Mg–Al–Zn alloy with

different surface topographies via plasma electrolytic oxidation

Nisa Nashraha, Baek Sung Huna, Young Gun Koa,*

a
Materials Electrochemistry Group, School of Materials Science and Engineering,
Yeungnam University, Gyeongsan 38541, Republic of Korea

*Corresponding author. E-mail address: [email protected] (Y.G. Ko)

Abstract

This work was made to investigate how nucleation and growth behavior of the coating film

were affected by surface topographies of Mg–Al–Zn alloy substrate during the initial stage of

plasma electrolytic oxidation (PEO). To satisfy this end, a single substrate was prepared by

mechanical treatment exhibiting rough and smooth regions with an equal area on the surface.

The rough region with prominent hills and grooves induced the breakdown of passive film,

which was indicated by an early appearance of plasma discharge on the rough region since

nucleation of coating film occurred preferentially around the hills. However, the coating film

grown on the grooves was somewhat thicker and more porous than the film grown on the hills

and smooth regions. This was due to the fact that the growth of the coating film was found to

be localized in the presence of rough region, which was in line with the discharge activities.

Herein, the nucleation and growth behavior during the initial stage of PEO will be discussed

schematically on the basis of microstructural interpretation.

Keywords: Mg alloy, surface roughness, plasma electrolytic oxidation, coating film

1
1. Introduction

Mg and its alloys have been of great interest for miscellaneous applications ranging from

automobile manufacturing to bio-implant engineering for their high electromagnetic

shielding, good specific strength, and acceptable biocompatibility [1–3]. Nevertheless, poor

corrosion resistance remained a major shortcoming, which limited their practical

implementation [4,5]. Efforts to address such drawback have developed protective coating

via surface modification, including chemical conversion coating, anodizing, and plasma

electrolytic oxidation (PEO) [6–8]. PEO, where the metallic surface was reformed through

plasma-assisted electrochemical reaction in an ecofriendly electrolyte, was favorable for

growing a thick coating layer with strong adhesion to the metal [9,10].

Previous investigations of PEO on Mg alloys showed that the microstructure of the coating

film varies with processing variables, such as electrolyte composition, electrical regime, and

substrate condition [11–16]. Among them, substrate features, such as its grain size and

intermetallic phase have been reported to induce different growth patterns of the coating

film on the surface [17,18]. In addition, a modification in the surface structure prior to PEO

was also reported to influence the microstructure of the final coating film. For instance,

Huang et al. [19] demonstrated the effect of micro-grooving on a YL113 Al substrate prior

to 150 min of PEO treatment. The coating film formed on the grooved sample, whose

surface roughness was 0.6 µm, was thicker than that formed on the non-grooved sample.

This was attributed to an increase in the partial growth rate of the coating film from 0.66 to

1.0 µm·cm-1 by the presence of the grooved structure [19]. Another study reported by Zhu et

al. [20], showed that the variation in the surface roughness of Ti–6Al–4V alloy significantly

altered the structural features of the coating film formed on the substrate [20]. A decrease in

surface roughness may reduce Gibbs free energy, which facilitated the homogeneous growth

of the coating film and hence suppressed the formation of micropores. Consequently, the

2
sample with the lowest roughness produced a coating film with a relatively compact

structure. Most recently, PEO treatment of an Al alloy with a trapezoidal structure,

produced by high-precision machining, resulted in a coating film with large micropores and

visible microcracks. However, the hardness of the sample with trapezoidal structure was

~1.3 times higher than that of the sample with a nearly flat structure. This outcome was

assigned to the high content of crystalline α-Al2O3, ɣ-Al2O3, and 3Al2O3.2SiO2, in the final

coating induced by the high temperature and stress concentration at the sharp edge of the

trapezoidal interface [21]. Studies suggested that an increase in surface roughness results in

a thick yet porous coating film, whereas with decreasing surface roughness, a coating film

with a more compact structure is obtained.

While it is accepted that the surface roughness of the substrate may control the final

microstructure of the coating films, its role in determining the nucleation and growth

phenomena during the initial stage of PEO remains unclear. Moreover, previous studies

have only focused on the variation of surface roughness between different substrates, while

the use of a single substrate to maintain the consistency of the main variable should provide

a facile approach beneficial for investigating the nucleation and growth process in a direct

manner.

Thus, the present study aims to evaluate the nucleation and growth behavior of a coating

film formed on a single Mg–Al–Zn alloy substrate having two different surface structures

together during the initial stage of PEO. The nucleation and growth phenomena of the

coating film are discussed based on microstructural interpretation. The corrosion assessment

by means of potentiodynamic polarization test is used to describe the polarization behavior

of the coating film related to surface roughness.

2. Material and method

3
A commercial Mg–Al–Zn alloy purchased from POSCO (Republic of Korea) was used as

the substrate in this study without further processing prior to PEO. As we confirmed from

the manufacturer, the substrate consisted of 2.89 Al, 0.96 Zn, 0.31 Mn, 0.15 Fe, 0.12 Si, and

balanced Mg (in wt.%) which was listed in Table 1. The sample having a dimension of 15

(length) × 25 (width) × 4 (thickness) mm3 worked as an anode and a stainless steel wire was

employed as a cathode. The sample with a dimension of 15 (length) × 25 (width) × 4

(thickness) mm3 worked as an anode and a stainless steel wire was employed as a cathode.

The sample was polished mechanically with emery paper of grade 2400 to obtain a smooth

finish. The half surface of the substrate was grinded again using emery paper of grade 80 to

produce rough and smooth regions of equal area. The rough and smooth regions are denoted

as Rr and Sr, respectively. Surface topographies of the initial substrate was evaluated by

atomic force microscopy (AFM, Xe-100). An arithmetic mean roughness (Ra) used

commonly was chosen to quantify the surface roughness of the present samples by means of

non-contacting assessment ‘across’ the geographical structure. The Ra values was measured

by the Eq. (1) below,

l
1
Ra =
l ∫ Z ( x ) dx
0

(1)

where Z(x) is a function representing surface roughness and l is a measurable length. The Ra

values of rough and smooth region calculated based on AFM measurement were found to be

~1 and ~0.03 µm, respectively as shown in Figure 1. Prior to use, the sample was rinsed

with distilled water, cleaned ultrasonically with acetone, and dried in a stream of warm air.

Subsequent PEO treatments were conducted in an aqueous electrolyte containing KOH and

Na2SiO3 at a concentration of 0.1 M and 0.05 M, respectively (pH: 11.83, conductivity:

12.39 mS·cm-1) with current density of 50 mA·cm -2 and frequency of 60 Hz. Details of the

4
experimental design can be found elsewhere [22].

In this study, the nucleation and growth stages were distinguished based on the

characteristics of plasma discharges. The first appearance of plasma discharges was

considered the nucleation stage (N), which were identified at 20 and 25 s for Rr and Sr,

respectively. The proliferation of plasma discharges up to 40 s was considered the growth

stage (G). Accordingly, the Rr and Sr at the nucleation and growth stages are denoted as Rr-

N, Sr-N, Rr-G, and Sr-G, respectively. The visual appearance of the plasma discharges on

the sample surface was captured with a digital single-lens reflex camera (Canon EOS-700D)

over time, and the intensity of the plasma discharges was analyzed using image analysis

software. The morphological surface and cross section of the coating film were identified by

field-emission scanning electron microscopy (FE-SEM; S-4800, Hitachi) to study the

nucleation and growth phenomena of the coating film.

The polarization behavior of the coating film were evaluated through potentiodynamic

polarization test using a potentiostat (Interface 1000, Gamry Instruments) in a 3.5 wt.%

NaCl solution. To study the polarization behavior related to surface

roughness, four samples were used, namely Rr-N, Sr-N, Rr-G, and Sr-G. The assessment

was carried out in three-electrode system, consisting of each sample as the working

electrode, Pt as counter electrode, and Ag/AgCl as reference electrode. Prior to polarization

test, each sample was immersed in the solution for 1 h until reaching a steady potential at

open circuit. During potentiodynamic polarization test, the sample was scanned from -0.25

to +0.4 V with respect to the open circuit potential at a scan rate of 1 mV ·s-1. Each

assessment was repeated at least three times to ensure that the data obtained would be

feasible. Electrochemical parameters, such as the corrosion current density (icorr), corrosion

potential (Ecorr), and slopes of the anodic and cathodic reactions (βa and βc, respectively),

were iterated through Tafel extrapolation. The polarization resistance (Rp) was calculated

5
using the representative equation proposed by Stern-Geary [23] shown in Eq. (2) below

(2)

3. Results and discussion

3.1 Plasma discharges at the initial stage of PEO

The occurrence of plasma discharges during the initial stage of PEO was influenced by the

initial surface topographies of the Mg–Al–Zn alloy substrate. The effect of surface

roughness on the behavior of plasma discharges was evaluated at the onset of breakdown

and during proliferation by considering voltage-time curve shown in Figure 2.

3.1.1 Initiation of plasma discharges at the onset of breakdown

Fig.2 displayed the voltage-time curve of the sample, which consisted of rough and smooth

regions coated by PEO at 50 mA·cm-2 for 60 s. Several images were provided as insets in

Fig. 2, showing the real image of the present sample prior to PEO and the appearance of

plasma discharges at 20 and 40 s during PEO along with the line graphs demonstrating a

number of discharges as a function of intensity. From the voltage–time curve, two stages of

PEO were identified based on the behavior of the responding voltage. Stage I was

characterized by a linear increment in the responding voltage due to an increase in electrical

resistance of the passive film growing on the substrate under constant current condition,

consistent with the linear relationship of Ohm’s law. The linear increase in voltage

continued until the resistance of the passive film was unable to sustain the intensive electric

field, resulting in the breakdown of the passive film. The initial breakdown of the passive

film occurs at ~20 s, upon reaching ~227 V, i.e., breakdown voltage, which was

accompanied by the appearance of plasma discharges across the substrate surface. This

6
event indicated a transition from the conventional anodic oxidation in stage I to PEO in

stage II. As shown in the inset of Fig. 2; the plasma discharges observed in the Rr (left) at

~20 s preceded the appearance of plasma discharges in the Sr (right) at ~25 s . The

difference in the ignition times of plasma discharges could be attributed to the

nonhomogeneous distribution of the electrical field across the substrate-electrolyte

interface, which suggested that the electrical resistance of the preformed passive film

differed between regions.

According to the breakdown theory of Ikonopisov [24], initial breakdown took place

through the defects in the passive film due to their relatively low electrical resistance. In this

study, considering the high level of surface irregularities in the Rr, the electrical resistance

of the passive film formed in this region was possibly lower than that of the passive film

formed in the Sr. Consequently, the passive film in the Rr of the substrate would be inclined

to experience an early breakdown.

After the initial breakdown, the slope of the responding voltage in stage II is significantly

lower than that in stage I because the growth of the coating film relies on the activity of

plasma discharges.

3.1.2. Proliferation of plasma discharges

The activity of plasma discharges intensified after 40 s, the behavior differed between two

regions of the present substrate. It was apparent that the discharge activities in the Rr were

localized in some areas, while a uniform distribution of plasma discharges appeared across

the Sr of the substrate. To analyze such characteristic of the plasma discharges, the graph in

the inset of Fig. 2 presented the number of plasma discharges as a function of their

intensities. The associated number of plasma discharge and their intensities observed in Rr

of the sample were higher than those observed on the Sr at 40 s. Unlike in the Rr of the

substrate, the low incidence of surface irregularities in the Sr maintained a homogeneous

7
distribution of the electrical field, hence inhibited localization of breakdown spot.

Troughton et al. [25] postulated that plasma discharges might be sustained in some local

areas containing defects until the defects are sealed completely. It is, accordingly, presumed

that plasma discharges should be concentrated in certain locations of the rough region

owing to the presence of numerous hills and grooves (i.e., structural defects). This behavior

was demonstrated by the plasma discharges apparent in the left region of the substrate

surface, after 40 s of PEO as displayed in the inset of Fig. 2.

Our research suggested that the plasma discharges exhibited different characteristics not

only during ignition but also during their proliferation. Such a difference in the

manifestation of the plasma discharges was expected to influence the microstructure of the

coating film that would explain its nucleation and growth phenomena.

3.2 Nucleation and growth behaviors of coating film considering surface roughness

The nucleation and growth behaviors of the coating film were affected by the surface

roughness of the initial substrate due to the difference in micro-regional characteristics of

the passive film formed on the substrate, where the presence of grooves in the Rr triggered

the localized growth of the coating film. Both the nucleation and growth phenomena could

be explained by surface and cross-sectional morphologies of the coating film grown on both

rough and smooth regions.

3.2.1. Effect of surface roughness on coating film formation

Figure 3 shows the microstructures of the coating film formed on the rough and smooth

regions of the substrate at the onset of breakdown. The surface morphology of the coating

film in the Rr shown in Fig. 3a, was characterized by two areas i.e., hills and grooves. The

surface of the Rr was not covered evenly by the coating film; the hills appeared to be

coated, whereas the coating on the grooves was incomplete. In contrast, the surface of the Sr

8
was coated uniformly, as depicted in Fig. 3b. These observations were confirmed by the

cross-section images presented in Figs. 3c and d. As seen in Fig. 3c, the oxide debris

nucleateD on the hills, suggesting that the breakdown initiated on the hills earlier than in the

adjacent grooves, which resulted in an incomplete coating film. Meanwhile, as seen in Fig.

3d, a thin coating film with a thickness of ~1 µm was detected on the Sr, which was

possibly the result of homogeneous nucleation induced by a uniform passive film preformed

on the smooth surface.

In general thin film theory, the variation in the surface structure of the substrate led to

inconsistent morphological features, including microstructure and thickness of the film

formed on the substrate. If the substrate was rough, the surface diffusion of ionic molecules

adsorbed to the interfaces of the solid material could exhibit different behaviors as a result

of structural gradient [26]. The correlation between the surface geometry of the metallic

substrate and its affinity towards ionic molecules in the wet electrolyte was reported by Zuo

et al. [27]. They exposed 316L stainless steel, with different surface finishes, to a 0.01 M

NaCl solution to evaluate pitting corrosion. Depending on the surface roughness, the aspect

ratio of the surface grooves, which was defined as the width (w) to depth (d) ratio,

influenced the chemical reaction between the substrate and ionic species in the electrolyte.

They suggested that the lower the aspect ratio (w/d), the greater the localized decay of the

metallic substrate owing to the higher surface area [27]. The decay of metallic materials

during pitting corrosion and the formation of a passive film during PEO were expected to

occur in a similar manner, in terms of the reaction between the metallic substrate and the

electrolyte in solution [28,29]. Considering the fact that the grooves possessed a higher

surface area than the hills, the chemical reaction was expected to be concentrated in the

former, resulting in a passive film that was presumably thicker than that formed on the hills

or on the Sr. As a result, the distribution of the electric field was likely non-uniform and

9
possibly concentrated around the hills coated with the thinner passive film. This triggered

avalanches of electrons to move across the hills with ease, igniting plasma discharges.

Thus, it was suggested that the high surface roughness induced an early breakdown due to

the nonuniform thickness of the passive film exhibiting nonhomogeneous electrical

resistance. Since the nucleation phenomena in both regions were found to be directly

affected by surface roughness, the difference in growth behavior of the coating film was

expected.

Figure 4 presents the morphological structure of the coating film after ~40 s of PEO, where

the plasma discharges participated in the growth stage. Figs. 4a and b showed the surface

morphologies of the coating film grown in the Sr and Sr, respectively. As seen in Fig. 4a,

the topographical features that defined the hill and groove areas were still evident after ~40

s of PEO treatment. It was apparent that the distribution of micropores in the coating film in

the Sr was not uniform, i.e., numerous micropores of different sizes were detected in the

film grown on the groove, while the film grown on the hill was relatively compact. In

contrast, a uniform film covered the Sr, as depicted in Fig. 4b. These results were evident in

the cross-sectional morphologies shown in Figs. 4c and d. Although nucleation was first

observed on the hills, the coating film grown on the grooves appeared thicker and more

porous than that grown on the hills after ~40 s of PEO (Fig. 4c). However, across the Sr,

continuous homogeneous nucleation produced a coating film with a relatively compact

structure and thickness of ~2.5 µm (Fig. 4d). The morphologies of coating films produced

through PEO were typically microporous or relatively compact structures [30,31]. These

microstructures of the coating films were considered the fingerprints of the discharge

activities, where each pore was associated with a discharge phenomenon that occurred

during PEO [32,33]. In the case of the Rr, the activities of the plasma discharges were

expected to be more pronounced in the grooves considering the numerous micropores

10
identified in the film grown on the groove. This result was consistent with the nonuniform

distribution of plasma discharges in the Sr at ~40 s, as shown in the inset of Fig. 1 . Although

the nucleation of the coating film in the Sr occur on the hills was earlier than that on the

grooves, the thickness of the coating film formed in the Sr was relatively comparable to that

formed in the Sr after 40 s of PEO. This was attributed to the fact that the coating film on the

groove was moderately thicker than that on the hills because of the higher growth rate of the

coating film formed in the Rr than that in the Sr. Such different growth rates corresponded to

the intense plasma discharges observed concentrated in the grooves with a thinner coating

film since nucleation first occurred on the hills, as evidenced in Fig. 3c.

3.2.2. Proposed mechanism

All microstructural evidence suggested that the variation in the surface roughness of the Mg–

Al–Zn alloy substrate determined the nucleation and growth behaviors of the coating film

during the initial stage of PEO. The schematic illustration in Figure 5 shows the formation of

the coating film on a single substrate with rough and smooth regions chronologically. Before

the formation of the coating film, the first insulating layer, called passive film, was formed

through a chemical reaction between the metal in the substrate and the ionic species in the

electrolyte [24,34]. The reaction forming the passive film was affected locally by the surface

topography of the substrate. The Rr offered more sites for the accumulation of anions, such

as OH- and SiO32-, originating from the electrolyte, owing to the greater surface area of the

grooves compared to those of the hills and the Sr. This condition was shown in Fig. 5a. The

accumulated anions promoted the chemical reactions in the grooves, resulting in the

formation of a thicker passive film on the grooves compared to those formed on the hills or in

the Sr. Therefore, the resistance of the passive film against the electric field was not

uniform, and the thinner passive film possessed lower resistance, as depicted in Fig. 5b. The

passive film on the hills was probably thinner than that in the Sr, therefore the electric

11
charges required to induce breakdown on the hills should be lower than that required to

induce breakdown in the Sr. Accordingly, plasma discharges first appeared across the Rr,

particularly on the hills. After the ignition of plasma discharges on the hills, it was plausible

that the discharge events occurred across the Sr before occurring in the grooves coated by the

thickest passive film [35]. These phenomena were confirmed by the microstructures of the

coating film in both regions after the ignition of the plasma discharges. As shown in Fig. 4,

the coating film appeared only on the hills of the Rr and across the entire Sr. This suggested

that the nucleation of the coating film on the grooves occurred later since the thickest

preformed passive film had the highest electrical resistance.

The presence of exposed surfaces in the grooves which was not completely covered by the

coating film, created a pathway with a relatively low electrical resistance. Thus, it was

believed that the discharge activity concentrates in the grooves, producing numerous

micropores in the film coating the grooves, as seen from the cross-section image in Fig. 4.

This local growth behavior facilitated plasma-assisted electrochemical reactions to thicken the

coating film by increasing the surrounding temperature [36,37]. This increased the growth

rate of the coating film on the grooves and caused it to exceed the growth rate of the coating

film on hills, thereby resulting in a thicker coating film, as shown in Fig. 4c and schematically

shown in Fig. 5c.

The present work revealed that surface topographies affected directly the ignition and

proliferation behaviors of plasma discharges, which, consequently, controlled the nucleation

and growth of the coating film, thus determining the overall thickness and porosity of the

coating film, as confirmed by the SEM images in Figs. 3 and 4.

3.3 Polarization behavior

The polarization behavior of the coating films was examined during a potentiodynamic

12
polarization test in a 3.5 wt.% NaCl solution to evaluate the electrochemical properties of

the coating film, formed in each region, in a corrosive environment. The polarization curves

of the coating film formed on rough and smooth regions during nucleation and growth,

namely Rr-N, Rr-G, Sr-N, and Sr-G, are displayed in Figure 6, along with a bar diagram in

the inset showing the polarization resistance of each sample. The electrochemical

parameters, such as icorr and Ecorr, iterated through Tafel extrapolation, and the Rp values are

tabulated in Table 2.

Based on the polarization curves shown in Fig. 6, the values of icorr decreased in the order of

Rr-N, Sr-N, Rr-G, and Sr-G, indicating a decrease in the corrosion kinetics , which could be

ascribed to the microstructure of the coating film in different regions. Meanwhile, the values

of Ecorr increased in the same order, indicating the decreasing tendency of the tested samples

to be oxidized. The significant difference in the Ecorr value of each sample was believed to be

related to the structure of the coating film. The Rr-N sample achieved the lowest value of Ecorr

due to the unevenly coated metallic surface resulting from preferential, localized nucleation.

The ability of the coating film to isolate the underlying metal from corrosive ions in an

acidic environment was closely related to the average thickness of the coating film. In the

present work, the average thicknesses of the coating film in the Rr-G and Sr-G are likely

comparable, although the coating film on the grooves in the Rr is somewhat thicker than that

in the Sr. The Sr-G, however, demonstrated better protection performance than the Rr-G. This

result is attributed to the compactness of the coating film in the Sr-G, which impedes the

diffusion of corrosive ions into the substrate, thereby improving the electrochemical

resistance. In contrast, the existence of numerous micro-pores in the coating film in the Rr-G,

enabled corrosive ions to penetrate to the metallic substrate. These results are also supported

by the polarization resistance values provided in the inset of Fig. 5. The error bars

corresponded to the respective standard deviations, which represented the homogeneity and

13
uniformity of the coating structure in each region. The Sr-G achieved the highest Rp and

lowest deviation due to the homogeneous structure of the coating film. The evaluation of the

polarization behavior showed that the coating with a low incidence of surface irregularities

offered greater metallic protection in a corrosive environment because the coating grown on

the smooth surface was more homogenous and uniform even in the nucleation stage compared

to that grown in the rough surface. Although the coating on Rr-G is thicker than that on Sr-G,

the numerous micropores concentrated in the valley allowed the corrosive ions to attack the

coating easily compared to Sr with compact structure. In this study, the homogeneity of the

coating film controlled by surface roughness would take the main control over the thickness,

in controlling the polarization behavior of the coating film.

4. Conclusion

The nucleation and growth behaviors of the coating film grown on an Al-Mg-Zn alloy during

the initial stage of PEO were investigated by considering two different surface topographies

of a single substrate. The breakdown phenomenon resulting in the appearance of plasma

discharges occurred ~5 s earlier in the rough region than that in the smooth region because

nucleation of the coating film favored the areas having the lowest electrical resistance, i.e., the

hills of the rough region. This result suggested that a high incidence of surface irregularities

of the substrate triggered the ignition of plasma discharges, promoting early nucleation. Once

plasma discharges proliferated, the growth of the coating film in the Rr was concentrated in

the grooves. Although this local growth behavior resulted in a coating film that was thicker on

the grooves than in the hills and in the Sr of the substrate, a greater number of micropores

formed in the coating film grown on the grooves due to the localized plasma discharges.

Determined by the surface roughness, the homogeneity of the coating film was found to have

the main role in affecting the electrochemical resistance of the coating film against corrosive

14
environment.

Declaration of interest

The authors declare that they have no known competing financial interests or personal

relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This work was supported by the Mid-Level Researcher National Project of the National

Research Foundation (NRF) funded by the Ministry of Science and ICT, Republic of Korea

(NRF-2020R1A2C2004192), and supported partly by the Competency Development Program

for Industry Specialist of the Korea Institute for Advancement of Technology (KIAT) funded

by the Ministry of Trade, Industry, and Energy, Republic of Korea (P0002019).

References

[1] Z.Z. Yin, W.C. Qi, R.C. Zeng, X.B. Chen, C.D. Gu, S.K. Guan, Y.F. Zheng, Advances in
coatings on biodegradable magnesium alloys, J. Magnes. Alloy. 8 (2020) 42–65.
[2] Y. Galindez, E. Correa, A.A. Zuleta, A. Valencia-Escobar, D. Calderon, L. Toro, P. Chacón, F.
Echeverría E, Improved Mg–Al–Zn Magnesium Alloys Produced by High Energy Milling and
Hot Sintering, Met. Mater. Int. (2019).
[3] S. Feliu, A. Samaniego, V. Barranco, A.A. El-Hadad, I. Llorente, C. Serra, J.C. Galván, A study
on the relationships between corrosion properties and chemistry of thermally oxidised surface
films formed on polished commercial magnesium alloys AZ31 and AZ61, Appl. Surf. Sci. 295
(2014) 219–230.
[4] Esmaily, J.E. Svensson, S. Fajardo, N. Birbilis, G.S. Frankel, S. Virtanen, R. Arrabal, S. Thomas,
L.G. Johansson, Fundamentals and advances in magnesium alloy corrosion, Prog. Mater. Sci. 89
(2017) 92–193.
[5] J. Song, J. She, D. Chen, F. Pan, Latest research advances on magnesium and magnesium alloys
worldwide, J. Magnes. Alloy. 8 (2020) 1–41.Francis, Y. Yang, A.R. Boccaccini, A new strategy
for developing chitosan conversion coating on magnesium substrates for orthopedic implants,
Appl. Surf. Sci. 466 (2019) 854-862.
15
[6] A. Francis, Y. Yang, A.R. Boccaccini, A new strategy for developing chitosan conversion coating
on magnesium substrates for orthopedic implants, Appl. Surf. Sci. 466 (2019) 854-862.
[7] W. Ximei, Z. Liqun, L. Huicong, L. Weiping, Influence of surface pretreatment on the anodizing
film of Mg alloy and the mechanism of the ultrasound during the pretreatment, Surf. Coatings
Technol. 202 (2008) 4210-4217.
[8] W. Al Zoubi, Y.G. Ko, Chemical stability of synergistic inorganic materials for enhancing
electrochemical performance, Compos. Sci. Technol. 199 (2020) 108383.
[9] W. Al Zoubi, Y.G. Ko, Flowerlike Organic-Inorganic Coating Responsible for Extraordinary
Corrosion Resistance via Self-assembly of an Organic Compound, ACS Sustain. Chem. Eng. 6
(2018) 3546–3555.
[10] W. Al Zoubi, M.P. Kamil, H. Yang, Y.G. Ko, Electron donor and acceptor agents responsible for
surface modification optimizing electrochemical performance, ACS Appl. Mater. Interfaces 9
(2017) 28967-28979.
[11] M. Kaseem, Y.G. Ko, Effect of starch on the corrosion behavior of Al-Mg-Si alloy processed by
micro arc oxidation from an ecofriendly electrolyte system, Bioelectrochemistry. 128 (2019)
133-139.
[12] J. Martin, A. V. Nominé, J. Stef, A. Nominé, J.X. Zou, G. Henrion, T. Grosdidier, The influence
of metallurgical state of substrate on the efficiency of plasma electrolytic oxidation (PEO)
process on magnesium alloy, Mater. Des. 178 (2019) 10785.
[13] S.S. Farhadi, M. Aliofkhazraei, G. Barati Darband, A. Abolhasani, A. Sabour Rouhaghdam,
Corrosion and wettability of PEO coatings on magnesium by addition of potassium stearate, J.
Magnes. Alloy. 5 (2017) 210–216.
[14] L. An, Y. Ma, L. Sun, Z. Wang, S. Wang, Investigation of mutual effects among additives in
electrolyte for plasma electrolytic oxidation on magnesium alloys, J. Magnes. Alloy. 8 (2020)
523-536.
[15] B. Salami, A. Afshar, A. Mazaheri, The effect of sodium silicate concentration on microstructure
and corrosion properties of MAO-coated magnesium alloy AZ31 in simulated body fluid, J.
Magnes. Alloy. 2 (2014) 72–77.
[16] M. Kaseem, J.H. Min, Y.G. Ko, Corrosion behavior of Al-1wt% Mg-0.85wt%Si alloy coated by
micro-arc-oxidation using TiO2 and Na2MoO4 additives: Role of current density, J. Alloys
Compd. 723 (2017) 448–455.
[17] S. Fatimah, H.W. Yang, M.P. Kamil, Y.G. Ko, Control of surface plasma discharge considering
the crystalline size of Al substrate, Appl. Surf. Sci. 477 (2018) 60–70.
[18] K.M. Lee, Y.G. Ko, D.H. Shin, Effect of icosahedral phase on growth behavior of thin oxide film
on MgZn12Y1.7 alloy via micro arc oxidation, Thin Solid Films. 531 (2013).
[19] H. Huang, X. Wei, J. Yang, J. Wang, Influence of surface micro grooving pretreatment on MAO

16
 process of aluminum alloy, Appl. Surf. Sci. 389 (2016) 1175-1181.
[20] L. Zhu, R.S. Petrova, J.P. Gashinski, Z. Yang, The effect of surface roughness on PEO-treated
Ti-6Al-4V alloy and corrosion resistance, Surf. Coatings Technol. 325 (2017) 22-29.
[21] J. Wang, S. Huang, H. Huang, M. He, P. Wangyang, L. Gu, Effect of micro-groove on
microstructure and performance of MAO ceramic coating fabricated on the surface of aluminum
alloy, J. Alloys Compd. 777 (2019) 94-101.
[22] S. Fatimah, F. Khoerunnisa, Y.G. Ko, Decoration of an inorganic layer with nickel (hydr)oxide
via green plasma electrolysis, RSC Adv. 8 (2018) 26804-26816.
[23] M. Stern, A.L. Geary, A theoretical analysis of the shape of polarization curves, J. Electrochem.
Soc. 104 (1957) 56-63.
[24] S. Ikonopisov, Theory of electrical breakdown during formation of barrier anodic films,
Electrochim. Acta. 22 (1977) 1077-1082.
[25] S.C. Troughton, A. Nominé, J. Dean, T.W. Clyne, Effect of individual discharge cascades on the
microstructure of plasma electrolytic oxidation coatings, Appl. Surf. Sci. 389 (2016) 260-269.
[26] L.B. Freund, S. Suresh, Thin Film Materials: stress, defect formation, and surface evolution,
Cambridge University Press, 2004.
[27] Y. Zuo, H. Wang, J. Xiong, The aspect ratio of surface grooves and metastable pitting of
stainless steel, Corros. Sci. 44 (2002) 25-35.
[28] J. Ninlachart, K.S. Raja, Passivation kinetics of Mg-Nd-Gd-Zn-Zr (EV31A) and Mg-Y-Nd-Gd-
Zr (WE43C) in NaOH solutions, J. Magnes. Alloy. 5 (2017) 254-270.
[29] B. Zhang, J. Wang, B. Wu, X.W. Guo, Y.J. Wang, D. Chen, Y.C. Zhang, K. Du, E.E. Oguzie,
X.L. Ma, Unmasking chloride attack on the passive film of metals, Nat. Commun. 9 (2018)
2559.
[30] M. Kaseem, T. Hussain, S.H. Baek, Y.G. Ko, Formation of stable coral reef-like structures via
self-assembly of functionalized polyvinyl alcohol for superior corrosion performance of AZ31
Mg alloy, Mater. Des. 193 (2020) 108823
[31] Y. kang Wu, Z. Yang, R. qiang Wang, G. rui Wu, D. Chen, D. dong Wang, X. tong Liu, D. long
Li, C. hong Guo, S. xue Yu, D. jiu Shen, P. Nash, An investigation of microstructure evolution
for plasma electrolytic oxidation (PEO) coated Al in an alkaline silicate electrolyte, Surf.
Coatings Technol. 351 (2018) 136–152.
[32] X. Zhang, S. Aliasghari, A. Němcová, T.L. Burnett, I. Kuběna, M. Šmíd, G.E. Thompson, P.
Skeldon, P.J. Withers, X-ray Computed Tomographic Investigation of the Porosity and
Morphology of Plasma Electrolytic Oxidation Coatings, ACS Appl. Mater. Interfaces. 8 (2016)
8801-8810.
[33] M.P. Kamil, W. Al Zoubi, D.K. Yoon, H.W. Yang, Y.G. Ko, Surface modulation of inorganic
layer via soft plasma electrolysis for optimizing chemical stability and catalytic activity, Chem.

17
 Eng. J. 391 (2020) 123614.
[34] S. Ikonopisov, A. Girginov, M. Machkova, Electrical breaking down of barrier anodic films
during their formation, Electrochim. Acta. 24 (1979) 451-456.
[35] Z. Ur Rehman, D. Choi, Investigation of ZrO2 nanoparticles concentration and processing time
effect on the localized PEO coatings formed on AZ91 alloy, J. Magnes. Alloy. 7 (2019) 555-565.
[36] A.L. Yerokhin, L.O. Snizhko, N.L. Gurevina, A. Leyland, A. Pilkington, A. Matthews,
Discharge characterization in plasma, J. Phys. D Appl. Phys. 36 (2003) 2110-2120.
[37] V. Dehnavi, W.J. Binns, J.J. Noël, D.W. Shoesmith, B.L. Luan, Growth behaviour of low-energy
plasma electrolytic oxidation coatings on a magnesium alloy, J. Magnes. Alloy. 6 (2018) 229-
237.

18
Table 1 Chemical constituent of the initial substrate used in PEO treatment.
Composition (wt.%)
Substrate
Mg Al Zn Mn Fe Si

Mg–Al–Zn Alloy Bal. 2.89 0.96 0.31 0.15 0.12

Table 2 Potentiodynamic polarization results of Sr-G, Rr-G, Sr-N, and Rr-N samples measured in 3.5
wt.% NaCl solution. The scan was from -0.25 to 0.4 V vs. OCP at a scan rate of 1 mV·s-1
Sample Ecorr (V) icorr (A·cm-2) βa (V·decade-1) βc (V·decade-1) Rp (104×Ω·cm2)

Sr-G -0.20 7.38×10-9 0.45 0.63 6.81 ±0.2

Rr-G -0.35 2.25×10-9 0.74 0.63 5.69 ±0.4

Sr-N -0.65 1.58×10-9 0.76 0.54 2.74 ±0.3

Rr-N -1.10 8.95×10-7 0.98 0.51 0.02 ±0.005

19
Figure caption

Fig. 1. Representative AFM images of an initial sample taken from two areas showing (a)

rough and (d) smooth region, respectively.

Fig. 2. Plot of the responding voltage vs. coating time of a Mg–Al–Zn substrate with rough

and smooth surface regions during PEO at 50 mA·cm-2 for 60 s. The insets show an

illustration of the substrate, the appearance of plasma discharges after different coating times,

and their intensity profiles.

Fig. 3. Representative SEM images showing the (a), (b) surface and (c), (d) cross-section

morphologies of the nucleation stage at 20 and 25 s in the rough and smooth regions,

respectively.

Fig. 4. Representative SEM images showing the (a), (b) surface and (c), (d) cross-section

morphologies of the growth stage at 40 s in the rough and smooth regions, respectively.

Fig. 5. Schematic illustration describing the formation of the coating film on a Mg–Al–Zn

alloy in chronological order: (a) the movement of anions during the formation of the passive

film, (b) the breakdown of passive film during nucleation of the coating film, and (c) the

growth of the coating film during the initial stage of PEO.

Fig. 6. Potentiodynamic polarization curve of Rr and Sr samples during a test performed in

3.5 wt. % NaCl solution. The inset displays a bar diagram showing the magnitudes of

polarization resistance for each sample.

20
Figure 1

Figure 2
21
22
Figure 3

23
Figure 4

24
Figure 5

25
Figure 6

26