Journal Search Engine
Search Advanced Search Adode Reader(link)
Download PDF Export Citaion korean bibliography PMC previewer
ISSN : 1225-0562(Print)
ISSN : 2287-7258(Online)
Korean Journal of Materials Research Vol.28 No.10 pp.544-550

Effect of Power Mode of Plasma Anodization on the Properties of formed Oxide Films on AZ91D Magnesium Alloy

Sung-Hyung Lee1,2†, Hitoshi Yashiro2, Song-Zhu Kure-Chu3
1GEO Nation Research institute, Tokyo 102-0093, Japan
2Department of Chemistry and Biological Science, Iwate University, Morioka, Iwate 020-8550, Japan
3Materials Function and Design, Nagoya Institute of Technology, Nagoya, Aichi 466-855, Japan
Corresponding author
E-Mail (S.-H. Lee, Iwate Univ.)
June 12, 2018 July 27, 2018 September 17, 2018


The passivation of AZ91D Mg alloys by plasma anodization requires deliberate choice of process parameters due to the presence of large amounts of structural defects. We study the dependence of pore formation, surface roughness and corrosion resistance on voltage by comparing the direct current (DC) mode and the pulse wave (pulse) mode in which anodization is performed. In the DC plasma anodization mode, the thickness of the electrolytic oxide film of the AZ91D alloy is uneven. In the pulse mode, the thickness is relatively uniform and the formed thin film has a three-layer structure. The pulse mode creates less roughness, uniform thickness and improved corrosion resistance. Thus, the change of power mode from DC to pulse at 150 V decreases the surface roughness (Ra) from 0.9 μm to 0.1 μm and increases the corrosion resistance in rating number (RN) from 5 to 9.5. Our study shows that an optimal oxide film can be obtained with a pulse voltage of 150 V, which produces an excellent coating on the AZ91D casting alloy.


    © Materials Research Society of Korea. All rights reserved.

    This is an Open-Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License ( which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

    1. Introduction

    Lately, the use of Mg alloy is rapidly expanding for mobile electronic devices, as well as lightweight automobiles and industrial applications.

    Mg alloys are not only lightweight but have attractive properties in strength, thermal conductivity, electrical conductivity, and electromagnetic interference shielding ability. However, the use of Mg alloys is critically limited by their inherent chemical reactivity if their surfaces are not modified adequately.

    Thus, various surface treatment methods have been applied to address this problem of Mg alloys. Among those methods, anodizing protects the surface most effectively. Various types of existing anodizing methods can form porous oxide films on Mg surface, but the formed films do not possess sufficient thickness, hardness, or corrosion resistance.1-6) Plasma anodizing is a type of surface treatment that forms oxide films electrochemically with relatively better properties with help of plasma generation in an aqueous solution. This method uses a low-concentration neutral/alkaline electrolyte without harmful substances such as sulfuric acid or chromic acid.

    Plasma anodizing consists of complex processes including simultaneous formation of oxidation layer, dielectric breakdown, dissolution of the formed oxidation layer, and gas generation.7-8) When the voltage applied during the treatment becomes higher than the dielectric breakdown voltage, the oxide film layer that had formed breaks down and plasma is generated. In addition, the metal ions produced by the electrolysis of substrate react with hydroxide ions supplied to the surface to form a very strong oxide film.

    The goal of this study is to find an optimal condition that can form a better oxide film by the plasma anodizing process. It has been attempted to understand the influence of parameters on the properties of oxide films formed in DC mode and pulse mode among several types of power modes.

    2. Experimental Methods

    As described in our earlier paper,1) the plasma anodizing system was comprised of an electrolyte, a high-voltage, high-current pulse-reverse rectifier(NF, BP4650), a cooler (Lab Companion, RW-1025G), stainless steel(type 316) heat exchanger, and an air bubble system.

    It is estimated that the core temperature of the discharge created during plasma anodizing treatment is over 2200 °C. Because of heat generated from such dielectric breakdown, the temperature of the electrolyte solution increases as operating time increases. To control the temperature of the electrolyte solution, a stainless steel heat exchanger was selected.

    A circulation system enables proper control of the temperature of the electrolyte solution. Oxygen gas generated on the anode during the plasma anodizing was effectively eliminated using strong pump flow without any additional vortex equipment. The system was designed to align the direction of the cold electrolyte solution being supplied to the position of the specimen. By positioning the specimen parallel to the direction of the electrolyte supply, the oxygen gas adsorbed on the surface was eliminated by the force of the flow of the electrolyte solution.1)

    An insulated gate bipolar transistor(IGBT) type multimeter was used to control the pulse width and bipolar pulse. In the pulse mode, the cycle consists of 500 ms of positive (+) time(on-time) and 500 ms(off-time). AZ91D specimens with chemical composition shown in Table 1 were machined into 12 × 40 mm with 3 mm thickness and used as an anode. Meanwhile, the graphite plates were used as a cathode. Since the current density varies depending on the interelectrode distance and their size, graphite plates having a size of 10 × 30 mm were placed each side of specimen to form a uniform oxide film layer.

    As shown in Fig. 1, distilled water and ethanol were used to remove impurities adhering to the surface of the raw material.

    A mixture of 2 g/L of sodium aluminate, 5 g/L of sodium hydroxide, and 10 g/L of sodium silicate was used as an electrolyte solution.

    In order to observe the plasma anodized specimens were cut, cold mounted, and polished with silicon carbide(SiC) paper. cross-section and composition of plasma-anodized samples were observed by field emission scanning electron microscopy(FESEM). To observe the crystalline phase appearing in the oxide film, X-ray diffraction(XRD) patterns were recorded over a range of 10-90° at 1.2°/min.

    In order to evaluate morphological characteristics, surface roughness was measured using surface roughness tester(MITUTOYO, SJ-400) and the roughness average (Ra) was used. The measurement was made over the surface of 0.8 mm for 3 times for each specimen. The Ra represents an average surface roughness value from center line to the surface contour.

    Electrochemical impedance spectroscopy(EIS) was measured using a potentiostat/galvanostat(10 V / 2 A, ZIVE SP2, WonATech, Co., Ltd) to evaluate the corrosion resistance of the treated specimens. A frequency response analyzer(FRA) was used to evaluate the EIS. In addition, a saltwater spray tester was used to perform a salt spray test according to the American Society for Testing and Materials(ASTM) - B117 standard test using a 5 wt% NaCl solution maintained at 35 °C for 72 h.

    3. Results and Discussion

    The effect of applied voltage on the formation of oxide film was evaluated in both DC and pulse modes at 50 °C. The treatment time was set to 15 min while the voltage applied to the specimen was changed from 50 to 200 V.

    Fig. 2 shows the cross- sectional structure and composition of the oxide films treated with DC and pulse modes at 150 V.

    In both EDS element mapping results showing the distribution of Mg, O and Si in the oxide film, Mg, O and Si are the main components of the oxide film and Mg and O are uniformly distributed.

    In the pulse mode, the thickness of the cross section of the oxide film was relatively uniform, and a three-layer structure of the oxide film was recognized. The porous structure of the outermost layer may be advantageous for further coating with good paint adhesion. The highdensity interlayer is the leading part for excellent mechanical properties(abrasion resistance) and corrosion resistance, and the interconnection layer is a very thin layer through which electricity passes from the material to the oxide film.9)

    Fig. 3 shows the XRD patterns of raw AZ91D and that treated with pulse plasma anodizing at 150 V. The results show clear differences in the peak intensities and crystal phases. After the plasma anodizing treatment, three ceramic phases, MgO, Mg2SiO4, and MgAl2O4, are identified on the AZ91D.

    The chemical species appearing during the anodization process are produced by various phenomena, such as the sample dissolution and electrolysis of the electrolyte solution, after which a ceramic phase is formed through a series of reactions. The MgO formation mechanism is similar to that observed in conventional Mg anodization techniques. As can be seen from equation,1) Mg2+ is transferred from the specimen to the oxidation layer/ electrolyte solution interface, and OH ions are transferred from the electrolyte solution to the sample/oxidation layer interface. Therefore, MgO simultaneously forms at the oxidation layer/electrolyte solution interface and at the sample/oxidation layer interface,

    Mg 2+ + 2Oc H- =MgO+ H 2 O.

    Mg2SiO4 can be generated by the reaction of SiO32− anions with Mg2+. Na2SiO3 in the electrolyte solution dissociates to produce SiO32− ions, which react with Mg2+ to form Mg2SiO4,

    Na 2 SiO 3 = 2Na +  + SiO 3 2-
    2Mg 2+ + SiO 3 2- + 2O H- =Mg 2 SiO 4 + H 2 O.

    MgAl2O4 can also form via the reaction

    Mg 2+ + 2AlO 2 -=MgAl 2 O 4 .

    Since AZ91D contains Al, the MgAl2O4 phase formed during the plasma anodization treatment has a stable spinel structure, which improves the corrosion resistance of the material by passivation.10)

    Fig. 4 shows FESEM images of surface oxides treated in DC and pulse modes. In the DC mode surface of the oxide layer showed a non-uniform, porous morphology. Applying a high DC voltage of 200 V results in higher non-uniformity and larger pore size. The pores that appeared on the surface were formed by oxygen bubbles generated when a micro-arc formed on the surface during the plasma anodizing treatment. The size of the micro-arc is proportional to the magnitude of current,9) and thus, a higher applied voltage resulted in larger pores.

    In the pulse mode, the oxide surface showed a relatively constant pore size up to a 150 V voltage condition and a small amount of released material and microcracks or craters(sinkhole / pipe) were detected at 200 V.

    In both conditions, as the voltage increased, the pores became larger and non uniform, but the pores of the oxidized surface in the pulse mode were relatively constant in size and less in defect. The non-uniform growth pattern of the oxide film and oxygen bubble trapping during the growth process resulted in the formation of a wider pore layer in the ceramic layer. Fig. 5 shows crosssectional images of the specimen to observe the thickness of the oxide layer grown as a function of applied voltage.

    The thickness of the oxide layer treated with DC and pulse modes increased from 8 to 36 μm depending on the anodization voltage. As the applied voltage increased in the plasma anodizing process, the oxide layer became thick. In addition, the oxide layer of the specimen treated at 200 V was 2 or 3 times thicker than that treated at 50 V.

    In the pulse mode, the growth rate of the thickness was relatively low. It is attributed to 500 ms(off-time) of the pulse mode.

    As the treatment voltage increases or the treatment time becomes longer, the oxide film grows thicker. In turn, as the oxide film becomes thicker, the electrical resistance increases, thus affecting generation of plasma.11) Therefore, the current flowing on the specimen surface during the plasma anodizing treatment does not increase proportionately to the voltage level. Moreover, the higher applied voltage increases the strength of the plasma, thereby dissolving a greater amount of the material from the surface. Thus, a strong micro-arc is sustained for a long time with a voltage level that has exceeded the dielectric breakdown voltage.

    Fig. 6 shows optical microphotographs of the oxide films formed on the AZ91D material after plasma anodization at fixed voltages of 50-200 V in DC and pulse modes.

    The surface of the oxide film formed in all samples contains many products that generate crater traces, pores generated by secretions, and craters. This phenomenon is caused by the plasma explosion of the surface of the sample at a high temperature at which molten Mg reacts rapidly with the electrolyte and is cooled.

    We found that the size of the pores produced by the plasma and the amount of product generated by the emission increased with increasing voltage.

    Fig. 7 shows the surface roughness of the AZ91D material as a function of plasma anodization voltage in DC and pulse modes, as indicated by Ra.

    As a result of controlling the voltage at 50-200 V in the pulse mode, the pore size in the oxide film was relatively constant and the surface roughness was in the range of Ra 0.1 to 0.4 μm. DC mode voltage control produced craters and microcracks and the surface roughness Ra ranged from 0.6 to 1.4 μm Therefore, it is more preferable to perform voltage control in pulse mode rather than DC mode in order to obtain an oxide film having a uniform surface.

    These results are consistent with the trends observed in surface morphology(see Figs. 6).

    Fig. 8 shows the corrosion resistance of the samples measured in a 3 mol dm−3 NaCl solution using EIS as a function of anodizing voltage. In this figure, corrosion resistance value is related to the magnitude of the AC impedance at 10 Hz, because the polarization resistance is usually dominant at this frequency. As expected, as the voltage of the plasma-anodization increased, corrosion resistance also increased.

    The corrosion resistance was higher when pulse voltage was applied, which means that the oxide layer formed under pulse voltage condition is more protective.

    Fig. 9 shows the results of a salt spray test conducted for 72 hours in accordance with the ASTM standard to evaluate the corrosion properties of the plasma anodized Mg alloy specimens.

    The experiment was carried out using a 5 wt% NaCl solution and the temperature was maintained at 35 °C. The corrosion properties were quantified by evaluating the average grade number(RN).

    The untreated AZ91D material and the plasma anodized material with DC mode at 150 and 200 V exhibited RN values of 3.5, 5 and 7, respectively, while the plasma anodized ones in pulse mode at 150-200 V showed a RN value of 9.5.

    These results mirror the trend observed in the surface roughness measurements and in the EIS data, shown in Figs. 7 and 8.

    Fig. 10 shows the cross-sectional structure of the pulse anodized AZ91D after the salt spray test. The aggressive electrolyte, i.e. chloride, penetrates the oxide film cracks and promotes material corrosion. The final corrosion products, MgO and Mg(OH)2, accumulate in the cracks of the oxide film and can interfere with the further progress of the corrosion process.12) However, higher Cl− ion concentrations at the interface accelerate corrosion. Figs. 10 and 11 shows that the entire oxide film is destroyed by Cl− ion and the corrosion product spreads on the top surface of the oxide film and the protection of the oxide film is lost. Fig. 11 shows that the corrosion products (a) consist of Mg, O and Cl− ions and that Mg, O and Si are homogeneously distributed in the corrosion free region (b).

    The corrosion of magnesium is the following oxidationreduction reaction:

    Mg+ 2H 2 O Mg(OH) 2 + H 2 .

    Mg in a neutral or alkaline solution is oxidized to Mg2+ and reacts with OH ions formed by the simultaneous reduction of water to form Mg(OH)2 on the magnesium surface.

    Further study is in progress to get rid of the defects in the passive film closely.

    4. Conclusion

    We have developed a plasma anodizing device suitable for the anodic oxidation of Mg alloys and have proven applicability for AZ91D alloys.

    Anodization is controlled by voltage, which can be optimized to obtain a uniform oxide layer on the Mg alloy surface. When the voltage was in the pulse mode of 150 V, the concentration of defects in the oxide layer was minimized. The surface roughness also decreased and the corrosion resistance increased from RN = 3.5 to RN = 9.5. Mg, Al, O and Si were uniformly distributed on the surface of the oxide film under optimized conditions. Cl− ions, known to promote corrosion, were detected in the corroded oxide layer.

    The roughness of the surface which is related to the size and shape of the pores and the resulting defects in the oxide film may have a greater effect on the durability of the material than the thickness of the oxide film.



    pretreatment of AZ91D magnesium alloy for plasma anodization.


    Cross-sectional view and EDS mapping of AZ91D after plasma anodization using (a) DC and (b) pulse at 150 V.


    XRD patterns of AZ91D (a) before and (b) after plasma anodizing at (a) AZ91D (b) 150 V.


    Surface morphologies of AZ91D after plasma anodizing using DC (a) and pulse (b) at (a, b) 50 V, (a-1, b-1) 100 V, (a-2, b-2) 150 V and (a-3, b-3) 200 V.


    Cross sections of AZ91D after plasma anodizing using DC (a) and pulse (b) at (a, b) 50 V, (a-1, b-1) 100 V, (a-2, b-2) 150 V and (a-3, b-3) 200 V.


    Digital microscope surface photographs(1000 × magnification) at DC (a) and pulse (b) at (a, b) 50 V, (a-1, b-1) 100 V, (a-2, b-2) 150 V and (a-3, b-3) 200 V.


    Effect of voltage on the surface roughness of the oxide film on the AZ91D surface after plasma anodization. DC (a) and pulse (b).


    Effect of applied DC and pulse voltage for plasma anodizing on the corrosion resistance of AZ91D as evaluated by EIS in a 3 mol dm−3 NaCl solution at 25 °C.


    Effect of DC and pulse voltage for plasma anodizing of AZ91D on the corrosion behavior by 72 hours salt spray test.


    Cross-sectional micrograph of the corrosion site of pulse plasma anodized AZ91D at 150 V after the salt spray test.


    Surface micrograph and EDS analysis data for the pulse plasma anodized material at 150 V after the salt spray test.


    Chemical composition of the AZ91D Mg alloy(mass %).


    1. S. H. Lee, H. Yashiro and S. Z. Kure-Chu, J. KoreanInst. Surf. Eng., 50, 432 (2017).
    2. B. L Mordike and T Ebert, Mater. Sci. Eng.: A, 302, 37 (2001).
    3. G. Song, A. Atrens, D. Stjohn, J. Nairn and Y. Li, Corros. Sci., 39, 855 (1997).
    4. Y. Ma, X. Nie, D. O. Northwood and H. Hu, Thin S olidFilms, 494, 296 (2006).
    5. J. Gray and B. Luan, J. Alloys Compd., 336, 88 (2002).
    6. P. B. Srinivasan, C. Blawert and W. Dietzel, Mater. Sci. Eng.: A, 494, 401 (2008).
    7. Y. Zhang, C. Yan, F. Wang, H. Lou and C. Cao, Surf. Coat. Technol., 161, 36 (2002).
    8. H. Duan, K. Du, C. Yan and F. Wang, Electrochim. Acta, 51, 2898 (2006).
    9. T. S. Narayanan, I. S. Park and M. H. Lee, Prog. Mater. Sci., 60, 1 (2014).
    10. L. Wang, J. Zhou, J. Liang and J. Chen, Surf. Coat. Technol., 206, 3109 (2012).
    11. R. O. Hussein, D. O. Northwood and X. Nie, Surf. Coat. Technol., 237, 357 (2013).
    12. S. J. Xia, R. Yue, R. G. RateickJr. and V. I. Birss., J. Electrochem. Soc., 151, B179 (2004).