- Email: [email protected]
Impedance monitoring of corrosion degradation of plasma electrolytic oxidation coatings (PEO) on magnesium alloy
Accepted Manuscript Impedance monitoring of corrosion degradation of plasma electrolytic oxidation coatings (PEO) on magnesium alloy L. Gawel, L. Nieuzyla, G. Nawrat, K. Darowicki, P. Slepski PII:
S0925-8388(17)32119-9
DOI:
10.1016/j.jallcom.2017.06.120
Reference:
JALCOM 42189
To appear in:
Journal of Alloys and Compounds
Received Date: 27 March 2017 Revised Date:
3 June 2017
Accepted Date: 11 June 2017
Please cite this article as: L. Gawel, L. Nieuzyla, G. Nawrat, K. Darowicki, P. Slepski, Impedance monitoring of corrosion degradation of plasma electrolytic oxidation coatings (PEO) on magnesium alloy, Journal of Alloys and Compounds (2017), doi: 10.1016/j.jallcom.2017.06.120. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Impedance monitoring of corrosion degradation of plasma electrolytic oxidation coatings (PEO) on magnesium alloy Gawel L.a,*, Nieuzyla L.b, Nawrat G.b, Darowicki K.a, Slepski P.a a
b
RI PT
Department of Electrochemistry, Corrosion and Material Engineering, Gdansk University of Technology, 11/12 Narutowicza Street, 80-233 Gdansk, Poland
Department of Inorganic Chemistry, Analytical Chemistry and Electrochemistry, Silesian University of Technology, 6 Boleslawa Krzywoustego Street, 44-100 Gliwice, Poland Corresponding author: [email protected]
M AN U
SC
*
Abstract
EP
Keywords
TE D
Impedance technique based on continuous multisinusoidal current perturbation was used to explain degradation process of plasma electrolytic oxidation coatings (PEO) on magnesium alloy. An obtained impedance spectrum was fitted by electrical equivalent circuit to describe properties of conversion layer on magnesium alloy. Additionally, galvanostatic mode maintained during experiment, with DC current equal to zero, allowed to achieve corrosion potential in degradation process. This methodology has advantage over more conventional methods. It allows observing parameters change in function of time and explanation of the degradation process of PEO on magnesium alloy.
AC C
plasma electrolytic oxidation, magnesium, corrosion, electrochemical impedance spectroscopy, galvanostatic mode, biodegradable materials
1
ACCEPTED MANUSCRIPT 1. Introduction
M AN U
SC
RI PT
Exceptional properties of magnesium alloys, such as high strength, low density, the possibility of various types of treatment and the possibility of recycling, result in growing interest in these materials, especially in the aircraft, automotive and electronic industry. Magnesium alloys are used in aviation for the production of engine housings and transmissions. It results from low density of this metal. The use of its alloys can reduce the weight of planes significantly without any substantial effect on their durability. However, magnesium also has many undesirable properties, including low corrosion resistance, low abrasion resistance, low creep resistance and high-chemical reactivity, including flammability. All of which limit the range of its applications. The main factor limiting a wide application of magnesium is its particularly high corrosivity, which results from an exceptionally low potential E°=-2.37 V. Magnesium and its alloys undergo galvanic corrosion as, in contact with other construction metals, it is always the anode, which is always subject to corrosive dissolution. In order to improve corrosion resistance, alloys of high metallurgical purity are used. These metals usually include iron, nickel and copper. In addition, to improve corrosion resistance of magnesium alloys, various surface treatment methods, such as electrochemical plating, conversion coatings, anodising, organic coatings and gas-phase deposition process, are used [1-3]. Magnesium alloys, as AZ91, with moderate corrosion resistance, strength, and cost is often used in automotive and aerospace field [4-7]. Currently, researchers are inclined to establish biodegradable magnesium alloys and thin protective surface layers on its surface [8-12].
AC C
EP
TE D
The formation of oxide coatings in the process of plasma electrolytic oxidation (PEO) is the method which is promising. This method involves the formation of a thin adherent hard ceramic-like coating as a result of complex electro, thermal, plasma and chemical reactions. The PEO layer formation process uses various forms of DC, AC unipolar and bipolar current modes [13,14]. It takes place in alkaline solutions containing various additions, mostly silicates which influence the properties of the coating [15-18]. High micro- hardness and low wearing rate improved the feasibility of alloys with PEO coating application in transportation and aviation industries [19]. Corrosion resistance enhancement is the main reason to employ PEO processes for manufacture of magnesium biodegradable implants. It is basically required that biodegradable implants maintain well enough corrosion resistance and mechanical properties[20-24]. Therefore, further research of materials based on oxide coating on magnesium alloys is an important development direction. The research on corrosion resistance of magnesium alloys covered with PEO coatings is rarely conducted using direct current methods, mostly due to the barrier characteristics of the protective layer. An alternative current method (Electrochemical Impedance Spectroscopy, EIS) is most often used for this purpose, which allows both detection of corrosion processes and analysis of PEO coating properties. EIS measurements are usually conducted on samples exposed in a corrosive environment of a sodium chloride solution. Frequency Response Analysis (FRA) is used to involve sequential sinusoidal excitation to provide impedance characteristic. This approach requires maintenance constant potential 2
ACCEPTED MANUSCRIPT regime over a significant period of time. It can directly affect on the corrosion process because the open circuit potential of corroding magnesium alloy with PEO layer exhibit significant changes over time of exposition as a result of the material structure changes [7,25,26].
M AN U
SC
RI PT
The presented impedance results, obtained in a traditional way, in the formation of spectra contained from one to three time constants [25,27,28]. Unfortunately, the interpretation of results presented by various authors is not consistent as various phenomena are attributed to the formation of individual time constants [26,29,30]. There are many reasons for this situation, e.g. varying degradation rate or varied times of impedance measurements. To explain the inaccuracies within the interpretation of impedance spectra representing the PEO coating degradation process, the authors suggest using the impedance method in the galvanostatic mode based on continuous multisinusoidal excitation. This method is also known as Dynamic Electrochemical impedance Spectroscopy (DEIS) [31,32]. Such a solution is used to obtain instantaneous impedance spectra and, as a result, to monitor the impedance of the degrading material under actual conditions, i.e. without keeping the artificially imposed corrosive potential, which occurs under the conditions of measurements in the potentiostatic mode. 2. Experimental
AC C
EP
TE D
Samples used in testing purposes (with dimensions of 90x53x10 mm) were made of the AZ91D alloy (composition [% by weight] Al: 8.77; Zn: 0.74; Mn: 0.18; Si<0.01; Fe<0.001; Cu<0.001; Ca<0.01; Ni: 0.001; Mg residue). Before the PEO process, the surface was smoothed using sandpaper of an increasing gradation up to 2500, ultrasonically cleaned in acetone and rinsed with demineralised water. The PEO process was conducted at 5°C in a solution containing hydrated sodium silicate, Na2SiO3·5H2O (7.5 g/dm3) and sodium hydroxide, NaOH (4.5 g/dm3) dissolved in deionised water. The micro plasma oxidation process was conducted for no longer than 20 minutes at the electrolysis voltage up to 500 V using basic parameters as follows: final voltage 420 V, averaged anodic current density 4.0 A/dm2, unit charge 10 As/cm2, current frequency 1,000 Hz and current filling of impulses 30%. The morphology of the coatings obtained was analysed using a scanning electron microscope (SEM) with energy-dispersive X-ray spectroscopy (EDX) analysis– Hitachi TM3000. The impedance measurements were performed over a disk-shaped surface of 1.3 cm2 determined by putting a vessel with a sealed opening in the bottom onto the sample. With such a sample arrangement it was possible to eliminate edge effects. An electrochemical measurement was obtained in 3.5% NaCl. The three electrode system consisted of measured material, SCE reference electrode and platinum mesh counter electrode. The excitation signal had the form of the sum of current sinusoids from the range 4.5 kHz to 300 mHz or 4.5 kHz to 3 Hz. Amplitudes and shifts of individual components of alternate current excitation was selected so that the voltage AC response signal could have an amplitude not exceeding 30 mV “peak to peak”, ac current perturbation did not influence on corrosion process. The DC constant component of current excitation was equal to 0, so the investigation was carried out at free corrosion potential. Signals were generated and recorded using the National 3
ACCEPTED MANUSCRIPT Instrument card PXI 4461, and Autolab 302N was used as a galvanostat. Controlling and result analysis were performed using software developed in the LabView environment. The ZsimpWin 3.21, electrochemical impedance data analysis software, was used to obtain electrical parameters by iterative process. 3. Results and discussion
RI PT
SEM images (Fig. 1) present the appearance and a cross -section of the tested samples with the visible base material and the PEO coating. The sample was given in the form of the AZ91D alloy covered with an oxide coating in the base solution containing sodium silicate and sodium hydroxide. The coating, approximately 5 µm thick, reveals various pores with varying depth, in addition to micro-cracks and other defects.
M AN U
SC
This is a characteristic image of protective coatings on magnesium alloys, formed in the process of plasma electrolytic oxidation. The porous structure results from the sparking electrical discharge, which accompanies the electrolytic oxidation conducted at a high voltage. The indication of the inner structure of oxide coatings is given by the number, size and size distribution of pores visible on the surface structure with round pores being open channels after discharges and longitudinal pores as clearances between discharge products. The number and size of pores depends on factors such as alloy composition, bath composition, current characteristics used, application time, and so forth [33-35].
TE D
If a coating with appropriate tightness is absent, the electrolyte penetrates the outer coating during the exposure in 3.5% NaCl. This triggers a process which degrades the PEO coatings and the magnesium alloy. The changes which then occur cause a change in the corrosion potential. The following figure presents example records of changes in the corrosion potential, which was obtained during 19 hours of measurements (Fig. 2). The visible fluctuations with an amplitude of approximately 100 mV result from changes occurring locally on the surface of the tested sample.
AC C
EP
A conventional impedance measurement using the FRA method, i.e. the method of subsequent sinusoidal excitations under the conditions of potential fluctuations, which means non-stationary conditions, cannot be performed. The solution is to take an impedance measurement using a multisinusoidal signal. Such an approach limits the excitation time which is necessary to obtain an impedance spectrum by the time the lowest frequency in the package of summed sinusoids is detected. By comparing the lowest analysed frequency to the dynamics of changes in the system, the stationary condition can be achieved. Moreover, using continuous multisinusoidal current excitation, not containing a constant component, continuous impedance changes can be obtained when the measurement system is degrading freely [36-39]. This methodology was applied to the analysed case. In the first approach, a signal, which had sinusoidal components from the range 4.5 kHz to 300 mHz, was applied. For decomposition purposes, to obtain instantaneous impedance spectra, 10-second fragments of voltage and current signals (three periods of the lowest analysed frequency) were used. The obtained impedance spectra had various forms: from deformed semicircles during the
4
ACCEPTED MANUSCRIPT initial period of exposure, to forms with distinguished characteristic time constants towards the end of the measurement (Fig. 3).
SC
RI PT
Unfortunately, the majority of spectra were not suitable for any analysis due to high deformations. This was caused by having an excessively high dynamics of changes in the system and, as a result, the impedance within the 10-second period necessary to obtain a single spectrum. In order to the stationary condition within the time limit to obtain the instantaneous spectrum and, consequently, representative impedance characteristic curves, it was necessary to reduce the range of the analysed frequencies. The excitation signal within the range 4.5 kHz to 3 Hz, which required 1-second recordings for analysis purposes, made it possible to solve the non-stationary problem and obtain proper impedance spectra in the function of the exposure time (Fig. 4). Attention must be paid to the fact that the reduction of the lower frequency range of 0.3 Hz to 3 Hz, also brought negative effects. The third inductive time constant could not be determined.
TE D
M AN U
Bearing in mind the image of the obtained spectra (decreasing of two flattened semicircles) and the knowledge that, in the tested system, penetration by the electrolyte occurs together with the degradation of the PEO barrier coating and corrosion of the magnesium alloy, the analysis of individual impedance spectra was conducted based on the diagram of the equivalent circuit R0(Q1(R1(Q2R2))) (Fig. 5). Chi-squared value of the order of 3×10−5 or below was chosen to be acceptable for a given model. The results of the adjustment of the individual elements in the circuit in the time function were compared with the results of the corrosion potential, as presented in the diagram below (Fig. 6). The resistance R0, which represents the electrolyte resistance between working and reference electrodes, was not taken into account as it was an insignificant element.
AC C
EP
The elements Q1 and R1 represent the first time constant, which is detected in the upper range of frequencies. The values of the constant phase element continue to increase with a slight plateau during the 3 to 8 hours of exposure. They do not show any significant correlation with the corrosion potential of the sample. The resistance R1 shows a decreasing trend, which is particularly strong during the initial period of exposure. Very strong fluctuations of the resistance can be observed between 3 and 8 hours. The fluctuations are highly correlated with the fluctuations of the corrosion potential of the sample. After this period, the decreasing trend of the resistance persists but quick changes in the resistance become smaller and smaller. The presented changes in impedance parameters are no less than the result of reactions leading to the destruction of the PEO oxide coating and corrosion of the substrate. During the exposure in a water environment, conversion of the protective coating follows the equation (1): MgO + H2O → Mg(OH)2
(1)
This leads to a gradual thinning and partial degradation of the coating. In addition, penetration by the electrolyte occurs through pores and creating cracks between pores. The conversion of the coating is associated with a change in the dielectric constant which, when combined with the thinning of the coating, leads to a gradual increase in capacitance, is reflected by a change in the Q1 parameter. The electrolyte penetration through the PEO 5
ACCEPTED MANUSCRIPT oxide coating is represented by the decrease in the R1 resistance. The occurrence of fluctuations of this parameter should be attributed to local clogging of the pores by magnesium hydroxide which has a large volume, as compared to the initial MgO and Mg. The electrolyte penetrating the PEO oxide coating reaches the substrate, inducing the oxidation process (2): Mg → Mg2+ + 2e-
(2)
RI PT
At a further stage, cations which are formed with the participation of Cl- and OH- anions can be transformed as follows (3) [2,40]: xMg2+ + Cl- + yOH- + zH2O → Mgx(OH)yCl · zH2O
(3)
Mg ⇄ Mg+ + e-
SC
The magnesium oxidation process is also analysed as a two-stage hypothetical reaction (4-5) [14, 41,42]:
M AN U
Mg+ + H2O → Mg2+ + OH- + ½H2
(4) (5)
Needless to say, the oxidation reaction is accompanied by the reduction (6): 2H2O + 2e- → H2 + 2OH-
(6)
AC C
EP
TE D
The aforementioned phenomena generate a time constant based on charge transfer resistance and the capacity of the double electric layer. The degradation reactions, together with the exposure time, occur on a larger and larger surface causing a decrease in the measured resistance of the Rct charge transfer and an increase in the Cdl double electric layer. Moreover, attention should be paid to the fact that the aforementioned processes occur, especially during the initial period, at the bottom of narrow pores and cause the following phenomena: changes in the electrolyte concentrations and composition by the surface, blocking the pores with corrosion products, unblocking the pores by released hydrogen etc. They generate fluctuations of the corrosion potential, charge transfer resistance and the capacitance of the double electric layer. Such changes are described by R2 and Q2, i.e. the aforementioned phenomena should be attributed to these parameters. Q2 and R2 are elements constituting the second low-frequency time constant. The values of both elements fluctuate in accordance with the fluctuations of the corrosion potential. However, values of the fluctuation become smaller and smaller towards the end of the exposure. It should also be noticed that a very high increase in Q2 and a decrease in R2 can be observed over the first 2 hours. The general trend of both elements is consistent with the trend of the elements relating to the first time constant: an increase in the constant-phase element and a decrease in the resistance. Impedance changes for narrower period of time between 7.5 and 8.4 h of the exposition have been presented on Fig. 7 to highlight the scale of the non-stationary problem related to surface changes during time of exposition. Within less than one hour, the impedance of the material rapidly fluctuates. Impedance spectra fluctuation are correlated with corrosion potential presented in Fig. 8. Moreover on Fig. 8 changes of the equivalent 6
ACCEPTED MANUSCRIPT circuit elements during potential fluctuation have been presented. The behavior of the individual parameters is similar to previous conclusions. However, the particular attention should be paid to the fact, how much this parameters changes during short period of time. This results showed how important is time of exposition and time of the measurement of the studied PEO coating.
SC
RI PT
While explaining the origin of individual time constants, one should also mention the presence of the inductive constant, which is detected in the analysed system within the milihertz range. Unfortunately, the dynamics of the system degradation prevents the monitoring of this range by means of the impedance method. The most frequent explanation for the induction is the process of adsorption of intermediate products. This phenomenon is observed not only in the case of corrosion of magnesium alloys, but also in relation to aluminium and iron alloys [43-46].
4. Conclusions
TE D
M AN U
Additionally degradation of the coating surface before and after 19 h of exposition in 3.5% NaCl is presented on Fig. 9. Local pores have considerably expanded and micro cracks between pores have occurred during exposition. Clogging of the pores by corrosion products could lead to increasing of local stress and creating micro cracks. Confirmation of that is higher weight % of Mg around the pores (Fig. 9b), which could be directly related with emerging corrosion products as Mg(OH)2 and MgCl2 [47] described by first constant. Decrease of the total weight percentage of Mg on the surface suggests a gradual thinning of the coating during exposition. Furthermore, values of the weight % of aluminum and silicon on the coating surface after exposition increased significantly in compare to the coating without exposition. It is related to presence of the stable MgAl2O4 and Mg2SiO4/MgSiO3 [48,49] which has been formed during PEO process.
EP
The process of protective PEO coating degradation on magnesium alloys and corrosion of the substrate has a non-stationary nature. This prevents the use of a large number of research methods, including the conventional impedance method FRA.
AC C
Corrosion potential changes clearly proves the non-stationarity of the measured PEO coating. Classic methods of the measurements often require a long period of time to study exposed materials and force constant potential different then open corrosion potential. As results, the informations may differ from the real state. As presented in this paper, measured coating system is rapidly changing its surface properties during time. Therefore, fast obtain the information about measured system and time of exposition before measurements are extremely important. As presented in this paper, it is possible to obtain instantaneous impedance characteristic curves and monitor impedance changes by using a multisinusoidal signal within an appropriate range of frequencies. The performance of such tests in the galvanostatic mode, while keeping the zero resultant current between the working and counter electrodes, made it possible to track changes both in impedance parameters and in the corrosion potential. The analysis showed that the high-frequency range of the spectrum is determined by the properties of the barrier layer. It is possible to track its penetration and 7
ACCEPTED MANUSCRIPT degradation by the electrolyte in detail. Within the low-frequency range, the detection of charge transfer resistance and the capacitance of the double electric layer occur.
References
RI PT
[1] J.E. Gray, B. Luan, Protective coatings on magnesium and its alloys — a critical review, J. Alloys Compd. 336 (2002) 88–113. doi: 10.1016/S0925-8388(01)01899-0. [2] Y. Gao, A. Yerokhin, A. Matthews, Deposition and evaluation of duplex hydroxyapatite and plasma electrolytic oxidation coatings on magnesium, Surf. Coatings Technol. 269 (2015) 170–182. doi: 10.1016/j.surfcoat.2015.01.030.
SC
[3] J. Liang, P.B. Srinivasan, C. Blawert, W. Dietzel, Comparison of electrochemical corrosion behaviour of MgO and ZrO2 coatings on AM50 magnesium alloy formed by plasma electrolytic oxidation, Corros. Sci. 51 (2009) 2483–2492. doi: 10.1016/j.corsci.2009.06.034.
M AN U
[4] S. Begum, D.L. Chen, S. Xu, A.A. Luo, Strain-controlled low-cycle fatigue properties of a newly developed extruded magnesium alloy, Metall. Mater. Trans. A Phys. Metall. Mater. Sci. 39 (2008) 3014–3026. doi:10.1007/s11661-008-9677-0. [5] C.R. Hutchinson, J.F. Nie, S. Gorsse, Modeling the precipitation processes and strengthening mechanisms in a Mg-Al-(Zn) AZ91 alloy, Metall. Mater. Trans. A. 36 (2005) 2093–2105. doi:10.1007/s11661-005-0330-x. [6] Z. Li, A.A. Luo, Q. Wang, H. Zou, J. Dai, L. Peng, Fatigue characteristics of sand-cast AZ91D magnesium alloy, J. Magnes. Alloy. 5 (2017) 1–12. doi:10.1016/j.jma.2017.03.001.
TE D
[7] L. Wang, B.-P. Zhang, T. Shinohara, Corrosion behavior of AZ91 magnesium alloy in dilute NaCl solutions, Mater. Des. 31 (2010) 857–863. doi:10.1016/j.matdes.2009.07.049.
EP
[8] T.M. Mukhametkaliyev, M.A. Surmeneva, A. Vladescu, C.M. Cotrut, M. Braic, M. Dinu, M.D. Vranceanu, I. Pana, M. Mueller, R.A. Surmenev, A biodegradable AZ91 magnesium alloy coated with a thin nanostructured hydroxyapatite for improving the corrosion resistance, Mater. Sci. Eng. C. 75 (2017) 95–103. doi:10.1016/j.msec.2017.02.033.
AC C
[9] C.L. Liu, J. Jiang, M. Wang, Y.J. Wang, P.K. Chu, W.J. Huang, Degradation and Biocompatibility of WE43, ZK60, and AZ91 Biodegradable Magnesium Alloys, Adv. Mater. Res. 287–290 (2011) 2008– 2014. doi:10.4028/www.scientific.net/AMR.287-290.2008. [10] Z. Wen, S. Duan, C. Dai, F. Yang, F. Zhang, Biodegradability and surface chemistry of AZ31D compared with AZ91 magnesium alloy in a modified simulated body fluid, Int. J. Electrochem. Sci. 9 (2014) 7846–7864. [11] R. Rojaee, M. Fathi, K. Raeissi, Controlling the degradation rate of AZ91 magnesium alloy via solgel derived nanostructured hydroxyapatite coating, Mater. Sci. Eng. C. 33 (2013) 3817–3825. doi:10.1016/j.msec.2013.05.014. [12] A.R. Bharat, B.S.S. Kumar, R.J.P. Sarath, P. Akhil, G.P.K. Reddy, B.R. Sunil, Influence of Al Content on the Corrosion Behavior of Biodegradable Magnesium Alloys in Simulated Physiological Solution, J. Adv. Biomed. Eng. Technol. 2 (2015) 17–23.
8
ACCEPTED MANUSCRIPT [13] S. V. Gnedenkov, O.A. Khrisanfova, A.G. Zavidnaya, S.L. Sinebryukhov, V.S. Egorkin, M. V. Nistratova, A. Yerokhin, A. Matthews, PEO coatings obtained on an Mg-Mn type alloy under unipolar and bipolar modes in silicate-containing electrolytes, Surf. Coatings Technol. 204 (2010) 2316–2322. doi: 10.1016/j.surfcoat.2009.12.024. [14] Y. Gao, A. Yerokhin, A. Matthews, Effect of current mode on PEO treatment of magnesium in Caand P-containing electrolyte and resulting coatings, Appl. Surf. Sci. 316 (2014) 558–567. doi: 10.1016/j.apsusc.2014.08.035.
RI PT
[15] A. Ghasemi, V.S. Raja, C. Blawert, W. Dietzel, K.U. Kainer, Study of the structure and corrosion behavior of PEO coatings on AM50 magnesium alloy by electrochemical impedance spectroscopy, Surf. Coatings Technol. 202 (2008) 3513–3518. doi: 10.1016/j.surfcoat.2007.12.033.
SC
[16] S.V. Gnedenkov, S.L. Sinebryukhov, D.V. Mashtalyar, I.M. Imshinetskiy, A.S. Gnedenkov, A.V. Samokhin, Y.V. Tsvetkov, Protective composite coatings obtained by plasma electrolytic oxidation on magnesium alloy MA8, Vacuum. 120 (2015) 107–114. doi: 10.1016/j.vacuum.2015.02.004.
M AN U
[17] H. Duan, C. Yan, F. Wang, Effect of electrolyte additives on performance of plasma electrolytic oxidation films formed on magnesium alloy AZ91D, Electrochim. Acta. 52 (2007) 3785–3793. doi: 10.1016/j.electacta.2006.10.066. [18] D. Sreekanth, N. Rameshbabu, K. Venkateswarlu, Effect of various additives on morphology and corrosion behavior of ceramic coatings developed on AZ31 magnesium alloy by plasma electrolytic oxidation, Ceram. Int. 38 (2012) 4607–4615. doi: 10.1016/j.ceramint.2012.02.040. [19] X. Li, X. Liu, B.L. Luan, Corrosion and wear properties of PEO coatings formed on AM60B alloy in NaAlO2 electrolytes, Appl. Surf. Sci. 257 (2011) 9135–9141. doi: 10.1016/j.apsusc.2011.05.115.
TE D
[20] D.R. Marsh, G. Li, The biology of fracture healing: optimising outcome, Br. Med. Bull 55 (1999) 856–869. doi: 10.1258/0007142991902673. [21] Y. Wang, F. Wang, M. Xu, B. Zhao, L. Guo, J. Ouyang, Microstructure and corrosion behavior of coated AZ91 alloy by microarc oxidation for biomedical application, Appl. Surf. Sci. 255 (2009) 9124– 9131. doi: 10.1016/j.apsusc.2009.06.116.
AC C
EP
[22] P. Wan, X. Lin, L. Tan, L. Li, W. Li, K. Yang, Influence of albumin and inorganic ions on electrochemical corrosion behavior of plasma electrolytic oxidation coated magnesium for surgical implants, Appl. Surf. Sci. 282 (2013) 186–194. doi: 10.1016/j.apsusc.2013.05.100. [23] Q. Zhao, X. Guo, X. Dang, J. Hao, J. Lai, K. Wang, Preparation and properties of composite MAO/ECD coatings on magnesium alloy, Colloids Surf. B Biointerfaces102 (2013) 321–326. doi: 10.1016/j.colsurfb.2012.07.040. [24] S. Chen, S. Guan, W. Li, H. Wang, J. Chen, Y. Wang, H. Wang, In vivo degradation and bone response of a composite coating on Mg–Zn–Ca alloy prepared by microarc oxidation and electrochemical deposition, J. Biomed. Mater. Res. B. Appl Biomater, 100 (2012) 533–543. doi: 10.1002/jbm.b.31982. [25] Y. Gu, S. Bandopadhyay, C. Chen, Y. Guo, C. Ning, Effect of oxidation time on the corrosion behavior of micro-arc oxidation produced AZ31 magnesium alloys in simulated body fluid, J. Alloys Compd. 543 (2012) 109–117. doi: 10.1016/j.jallcom.2012.07.130.
9
ACCEPTED MANUSCRIPT [26] A. Castellanos, A. Altube, J.M. Vega, E. García-Lecina, J.A. Díez, H.J. Grande, Effect of different post-treatments on the corrosion resistance and tribological properties of AZ91D magnesium alloy coated PEO, Surf. Coatings Technol. 278 (2015) 99–107. doi: 10.1016/j.surfcoat.2015.07.017. [27] Y. Song, K. Dong, D. Shan, E.H. Han, Investigation of a novel self-sealing pore micro-arc oxidation film on AM60 magnesium alloy, J. Magnes. Alloy. 1 (2013) 82–87. doi:10.1016/j.jma.2013.02.009.
RI PT
[28] L. Pezzato, K. Brunelli, S. Gross, M. Magrini, M. Dabalà, Effect of process parameters of plasma electrolytic oxidation on microstructure and corrosion properties of magnesium alloys, J. Appl. Electrochem. 44 (2014) 867–879. doi: 10.1007/s10800-014-0695-x. [29] Y. Wang, J. Wang, J. Zhang, Z. Zhang, Characteristics of anodic coatings oxidized to different voltage on AZ91D Mg alloy by micro-arc oxidization technique, Mater. Corros. 56 (2005) 88–92. doi: 10.1002/maco.200403822.
SC
[30] T. Chen, W. Xue, Y. Li, X. Liu, J. Du, Corrosion behavior of friction stir welded AZ31B magnesium alloy with plasma electrolytic oxidation coating formed in silicate electrolyte, Mater. Chem. Phys. 144 (2014) 462–469. doi: 10.1016/j.matchemphys.2014.01.020.
M AN U
[31] K.Darowicki, Theoretical description of the measuring method of instantaneous impedance spectra, J. Electroanal. Chem. 486 (2000) 101-105. doi: 10.1016/S0022-0728(00)00110-8 [32] K. Darowicki, P. Slepski, Dynamic electrochemical impedance spectroscopy of the first order electrode reaction, J. Electroanal. Chem. 547 (2003) 1-8. doi: 10.1016/S0022-0728(03)00154-2 [33] A.M. Kumar, S.H. Kwon, H.C. Jung, Y.H. Park, H.J. Kim, K.S. Shin, Fabrication and Electrochemical Corrosion Behavior of PEO Coatings on Strip-Cast AZ31Mg Alloy in 3 . 5 % NaCl Solution, Ind. Eng. Chem. Res. 53 (2014) 9703−9713. doi: 10.1021/ie404329s
TE D
[34] R.O. Hussein, D.O. Northwood, X. Nie, The effect of processing parameters and substrate composition on the corrosion resistance of plasma electrolytic oxidation (PEO) coated magnesium alloys, Surf. Coatings Technol. 237 (2013) 357–368. doi: 10.1016/j.surfcoat.2013.09.021.
EP
[35] P. Su, X. Wu, Y. Guo, Z. Jiang, Effects of cathode current density on structure and corrosion resistance of plasma electrolytic oxidation coatings formed on ZK60 Mg alloy, J. Alloys Compd. 475 (2009) 773–777. doi: 10.1016/j.jallcom.2008.08.030.
AC C
[36] P. Slepski, K. Darowicki, K. Andrearczyk, On-line measurement of cell impedance during charging and discharging process, J. Electroanal. Chem. 633 (2009) 121–126. doi: 10.1016/j.jelechem.2009.05.002. [37] J. Ryl, J. Wysocka, P. Slepski, K. Darowicki, Instantaneous impedance monitoring of synergistic effect between cavitation erosion and corrosion processes, Electrochim. Acta. 203 (2015) 388–395. doi: 10.1016/j.electacta.2016.01.216. [38] H. Gerengi, P. Slepski, E. Ozgan, M. Kurtay, Investigation of corrosion behavior of 6060 and 6082 aluminum alloys under simulated acid rain conditions, Mater. Corros. 66 (2015) 233–240. doi: 10.1002/maco.201307287. [39] J. Ryl, K. Darowicki, P. Slepski, Evaluation of cavitation erosion-corrosion degradation of mild steel by means of dynamic impedance spectroscopy in galvanostatic mode, Corros. Sci. 53 (2011) 1873–1879. doi: 10.1016/j.corsci.2011.02.004.
10
ACCEPTED MANUSCRIPT [40] X. Zhou, L. Jiang, P. Wu, Y. Sun, Y. Yu, G. Wei, H. Ge, Effect of aggressive ions on degradation of WE43 magnesium alloy in physiological environment, Int. J. Electrochem. Sci. 9 (2014) 304–314. [41] G. Song, D.S.T. John, J. Nairn, The anodic dissolution of magnesium and sulphate solutions in chloride, Corros. Sci. 39 (1997) 1981–2004. doi: 10.1016/S0010-938X(97)00090-5. [42] F.H. Cao, V.H. Len, Z. Zhang, J.Q. Zhang, Corrosion behavior of magnesium and its alloy in NaCl solution, Russ. J. Electrochem. 43 (2007) 837–843. doi: 10.1134/S1023193507070142.
RI PT
[43] X. Li, S. Deng, X. Xie, Experimental and theoretical study on corrosion inhibition ofoxime compounds for aluminium in HCl solution, Corros. Sci. 81 (2014) 162–175. doi: 10.1016/j.corsci.2013.12.021.
SC
[44] N. Chaubey, V.K. Singh, Savita, M.A. Quraishi, E.E. Ebenso, Corrosion inhibition of aluminium alloy in alkaline media by Neolamarkia Cadamba bark extract as a green inhibitor, Int. J. Electrochem. Sci. 10 (2015) 504–518.
M AN U
[45] M. Kissi, M. Bouklah, B. Hammouti, M. Benkaddour, Establishment of equivalent circuits from electrochemical impedance spectroscopy study of corrosion inhibition of steel by pyrazine in sulphuric acidic solution, Appl. Surf. Sci. 252 (2006) 4190–4197. doi: 10.1016/j.apsusc.2005.06.035. [46] G.A. Zhang, Y.F. Cheng, On the fundamentals of electrochemical corrosion of X65 steel in CO2containing formation water in the presence of acetic acid in petroleum production, Corros. Sci. 51 (2009) 87–94. doi: 10.1016/j.corsci.2008.10.013. [47] H.S. Ryu, D.-S. Park, S.-H. Hong, Improved corrosion protection of AZ31 magnesium alloy through plasma electrolytic oxidation and aerosol deposition duplex treatment, Surf. Coatings Technol. 219 (2013) 82–87. doi: 10.1016/j.surfcoat.2013.01.008.
TE D
[48] M. Daroonparvar, M.A.M. Yajid, N.M. Yusof, H.R. Bakhsheshi-Rad, Preparation and corrosion resistance of a nanocomposite plasma electrolytic oxidation coating on Mg-1%Ca alloy formed in aluminate electrolyte containing titania nano-additives, J. Alloys Compd. 688 (2016) 841–857. doi: 10.1016/j.jallcom.2016.07.081.
AC C
EP
[49] G. Barati Darband, M. Aliofkhazraei, P. Hamghalam, N. Valizade, Plasma electrolytic oxidation of magnesium and its alloys: Mechanism, properties and applications, J. Magnes. Alloy. 5 (2017) 74– 132. doi: 10.1016/j.jma.2017.02.004.
11
ACCEPTED MANUSCRIPT Figure captions Fig. 1 SEM images of the oxide coating formed at 5 °C using 30% impulse filling with current and the impulse current frequency of 1,000 Hz on the AZ91D magnesium alloy: a) Coating surface 0.002 mm2; b-c) Cross section of the coating with visible canals through during plasma discharges; d) Distribution of the coating thickness.
RI PT
Fig. 2 Example records of changes in the corrosion potential during a 20-hour measurement. Fig. 3 Example of instantaneous impedance spectra obtained after 4 min. (+), 9 h (◊) and 19 h (). Frequency excitation within the range 4.5 kHz to 0.3 Hz
SC
Fig. 4 Changes in the impedance characteristic curve of the material during the experiment within the excitation frequency range 4.5 kHz to 3 Hz.
M AN U
Fig. 5 Equivalent circuit model used to describe corrosion degradation process of PEO on magnesium alloy. Fig. 6 Change in the value of individual elements in the equivalent circuit (R1, Q1, R2, Q2) as regards changes in the corrosion potential in the time function (grey doped lines - corrosion potential. Fig. 7 Changes in the impedance characteristic curve of the material during the experiment for short period of time within the excitation frequency range 4.5 kHz to 3 Hz.
TE D
Fig. 8 Change in the value of individual elements in the equivalent circuit (R1, Q1, R2, Q2) as regards changes in the corrosion potential for short period of time (grey doped lines corrosion potential.
AC C
EP
Fig. 9 EDX analysis of coating surface before and after 19 h exposition.
12
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Highlights
EP
TE D
M AN U
SC
RI PT
Impedance technique with continuous multisinusoidal current perturbation was used. An explanation of corrosion process of PEO coated magnesium alloy is proposed. Galvanostatic mode allowed to achieve corrosion potential in time of experiment. Changes of equivalent circuit parameters allowed to explain corrosion process.
AC C
• • • •
S0925-8388(17)32119-9
DOI:
10.1016/j.jallcom.2017.06.120
Reference:
JALCOM 42189
To appear in:
Journal of Alloys and Compounds
Received Date: 27 March 2017 Revised Date:
3 June 2017
Accepted Date: 11 June 2017
Please cite this article as: L. Gawel, L. Nieuzyla, G. Nawrat, K. Darowicki, P. Slepski, Impedance monitoring of corrosion degradation of plasma electrolytic oxidation coatings (PEO) on magnesium alloy, Journal of Alloys and Compounds (2017), doi: 10.1016/j.jallcom.2017.06.120. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Impedance monitoring of corrosion degradation of plasma electrolytic oxidation coatings (PEO) on magnesium alloy Gawel L.a,*, Nieuzyla L.b, Nawrat G.b, Darowicki K.a, Slepski P.a a
b
RI PT
Department of Electrochemistry, Corrosion and Material Engineering, Gdansk University of Technology, 11/12 Narutowicza Street, 80-233 Gdansk, Poland
Department of Inorganic Chemistry, Analytical Chemistry and Electrochemistry, Silesian University of Technology, 6 Boleslawa Krzywoustego Street, 44-100 Gliwice, Poland Corresponding author: [email protected]
M AN U
SC
*
Abstract
EP
Keywords
TE D
Impedance technique based on continuous multisinusoidal current perturbation was used to explain degradation process of plasma electrolytic oxidation coatings (PEO) on magnesium alloy. An obtained impedance spectrum was fitted by electrical equivalent circuit to describe properties of conversion layer on magnesium alloy. Additionally, galvanostatic mode maintained during experiment, with DC current equal to zero, allowed to achieve corrosion potential in degradation process. This methodology has advantage over more conventional methods. It allows observing parameters change in function of time and explanation of the degradation process of PEO on magnesium alloy.
AC C
plasma electrolytic oxidation, magnesium, corrosion, electrochemical impedance spectroscopy, galvanostatic mode, biodegradable materials
1
ACCEPTED MANUSCRIPT 1. Introduction
M AN U
SC
RI PT
Exceptional properties of magnesium alloys, such as high strength, low density, the possibility of various types of treatment and the possibility of recycling, result in growing interest in these materials, especially in the aircraft, automotive and electronic industry. Magnesium alloys are used in aviation for the production of engine housings and transmissions. It results from low density of this metal. The use of its alloys can reduce the weight of planes significantly without any substantial effect on their durability. However, magnesium also has many undesirable properties, including low corrosion resistance, low abrasion resistance, low creep resistance and high-chemical reactivity, including flammability. All of which limit the range of its applications. The main factor limiting a wide application of magnesium is its particularly high corrosivity, which results from an exceptionally low potential E°=-2.37 V. Magnesium and its alloys undergo galvanic corrosion as, in contact with other construction metals, it is always the anode, which is always subject to corrosive dissolution. In order to improve corrosion resistance, alloys of high metallurgical purity are used. These metals usually include iron, nickel and copper. In addition, to improve corrosion resistance of magnesium alloys, various surface treatment methods, such as electrochemical plating, conversion coatings, anodising, organic coatings and gas-phase deposition process, are used [1-3]. Magnesium alloys, as AZ91, with moderate corrosion resistance, strength, and cost is often used in automotive and aerospace field [4-7]. Currently, researchers are inclined to establish biodegradable magnesium alloys and thin protective surface layers on its surface [8-12].
AC C
EP
TE D
The formation of oxide coatings in the process of plasma electrolytic oxidation (PEO) is the method which is promising. This method involves the formation of a thin adherent hard ceramic-like coating as a result of complex electro, thermal, plasma and chemical reactions. The PEO layer formation process uses various forms of DC, AC unipolar and bipolar current modes [13,14]. It takes place in alkaline solutions containing various additions, mostly silicates which influence the properties of the coating [15-18]. High micro- hardness and low wearing rate improved the feasibility of alloys with PEO coating application in transportation and aviation industries [19]. Corrosion resistance enhancement is the main reason to employ PEO processes for manufacture of magnesium biodegradable implants. It is basically required that biodegradable implants maintain well enough corrosion resistance and mechanical properties[20-24]. Therefore, further research of materials based on oxide coating on magnesium alloys is an important development direction. The research on corrosion resistance of magnesium alloys covered with PEO coatings is rarely conducted using direct current methods, mostly due to the barrier characteristics of the protective layer. An alternative current method (Electrochemical Impedance Spectroscopy, EIS) is most often used for this purpose, which allows both detection of corrosion processes and analysis of PEO coating properties. EIS measurements are usually conducted on samples exposed in a corrosive environment of a sodium chloride solution. Frequency Response Analysis (FRA) is used to involve sequential sinusoidal excitation to provide impedance characteristic. This approach requires maintenance constant potential 2
ACCEPTED MANUSCRIPT regime over a significant period of time. It can directly affect on the corrosion process because the open circuit potential of corroding magnesium alloy with PEO layer exhibit significant changes over time of exposition as a result of the material structure changes [7,25,26].
M AN U
SC
RI PT
The presented impedance results, obtained in a traditional way, in the formation of spectra contained from one to three time constants [25,27,28]. Unfortunately, the interpretation of results presented by various authors is not consistent as various phenomena are attributed to the formation of individual time constants [26,29,30]. There are many reasons for this situation, e.g. varying degradation rate or varied times of impedance measurements. To explain the inaccuracies within the interpretation of impedance spectra representing the PEO coating degradation process, the authors suggest using the impedance method in the galvanostatic mode based on continuous multisinusoidal excitation. This method is also known as Dynamic Electrochemical impedance Spectroscopy (DEIS) [31,32]. Such a solution is used to obtain instantaneous impedance spectra and, as a result, to monitor the impedance of the degrading material under actual conditions, i.e. without keeping the artificially imposed corrosive potential, which occurs under the conditions of measurements in the potentiostatic mode. 2. Experimental
AC C
EP
TE D
Samples used in testing purposes (with dimensions of 90x53x10 mm) were made of the AZ91D alloy (composition [% by weight] Al: 8.77; Zn: 0.74; Mn: 0.18; Si<0.01; Fe<0.001; Cu<0.001; Ca<0.01; Ni: 0.001; Mg residue). Before the PEO process, the surface was smoothed using sandpaper of an increasing gradation up to 2500, ultrasonically cleaned in acetone and rinsed with demineralised water. The PEO process was conducted at 5°C in a solution containing hydrated sodium silicate, Na2SiO3·5H2O (7.5 g/dm3) and sodium hydroxide, NaOH (4.5 g/dm3) dissolved in deionised water. The micro plasma oxidation process was conducted for no longer than 20 minutes at the electrolysis voltage up to 500 V using basic parameters as follows: final voltage 420 V, averaged anodic current density 4.0 A/dm2, unit charge 10 As/cm2, current frequency 1,000 Hz and current filling of impulses 30%. The morphology of the coatings obtained was analysed using a scanning electron microscope (SEM) with energy-dispersive X-ray spectroscopy (EDX) analysis– Hitachi TM3000. The impedance measurements were performed over a disk-shaped surface of 1.3 cm2 determined by putting a vessel with a sealed opening in the bottom onto the sample. With such a sample arrangement it was possible to eliminate edge effects. An electrochemical measurement was obtained in 3.5% NaCl. The three electrode system consisted of measured material, SCE reference electrode and platinum mesh counter electrode. The excitation signal had the form of the sum of current sinusoids from the range 4.5 kHz to 300 mHz or 4.5 kHz to 3 Hz. Amplitudes and shifts of individual components of alternate current excitation was selected so that the voltage AC response signal could have an amplitude not exceeding 30 mV “peak to peak”, ac current perturbation did not influence on corrosion process. The DC constant component of current excitation was equal to 0, so the investigation was carried out at free corrosion potential. Signals were generated and recorded using the National 3
ACCEPTED MANUSCRIPT Instrument card PXI 4461, and Autolab 302N was used as a galvanostat. Controlling and result analysis were performed using software developed in the LabView environment. The ZsimpWin 3.21, electrochemical impedance data analysis software, was used to obtain electrical parameters by iterative process. 3. Results and discussion
RI PT
SEM images (Fig. 1) present the appearance and a cross -section of the tested samples with the visible base material and the PEO coating. The sample was given in the form of the AZ91D alloy covered with an oxide coating in the base solution containing sodium silicate and sodium hydroxide. The coating, approximately 5 µm thick, reveals various pores with varying depth, in addition to micro-cracks and other defects.
M AN U
SC
This is a characteristic image of protective coatings on magnesium alloys, formed in the process of plasma electrolytic oxidation. The porous structure results from the sparking electrical discharge, which accompanies the electrolytic oxidation conducted at a high voltage. The indication of the inner structure of oxide coatings is given by the number, size and size distribution of pores visible on the surface structure with round pores being open channels after discharges and longitudinal pores as clearances between discharge products. The number and size of pores depends on factors such as alloy composition, bath composition, current characteristics used, application time, and so forth [33-35].
TE D
If a coating with appropriate tightness is absent, the electrolyte penetrates the outer coating during the exposure in 3.5% NaCl. This triggers a process which degrades the PEO coatings and the magnesium alloy. The changes which then occur cause a change in the corrosion potential. The following figure presents example records of changes in the corrosion potential, which was obtained during 19 hours of measurements (Fig. 2). The visible fluctuations with an amplitude of approximately 100 mV result from changes occurring locally on the surface of the tested sample.
AC C
EP
A conventional impedance measurement using the FRA method, i.e. the method of subsequent sinusoidal excitations under the conditions of potential fluctuations, which means non-stationary conditions, cannot be performed. The solution is to take an impedance measurement using a multisinusoidal signal. Such an approach limits the excitation time which is necessary to obtain an impedance spectrum by the time the lowest frequency in the package of summed sinusoids is detected. By comparing the lowest analysed frequency to the dynamics of changes in the system, the stationary condition can be achieved. Moreover, using continuous multisinusoidal current excitation, not containing a constant component, continuous impedance changes can be obtained when the measurement system is degrading freely [36-39]. This methodology was applied to the analysed case. In the first approach, a signal, which had sinusoidal components from the range 4.5 kHz to 300 mHz, was applied. For decomposition purposes, to obtain instantaneous impedance spectra, 10-second fragments of voltage and current signals (three periods of the lowest analysed frequency) were used. The obtained impedance spectra had various forms: from deformed semicircles during the
4
ACCEPTED MANUSCRIPT initial period of exposure, to forms with distinguished characteristic time constants towards the end of the measurement (Fig. 3).
SC
RI PT
Unfortunately, the majority of spectra were not suitable for any analysis due to high deformations. This was caused by having an excessively high dynamics of changes in the system and, as a result, the impedance within the 10-second period necessary to obtain a single spectrum. In order to the stationary condition within the time limit to obtain the instantaneous spectrum and, consequently, representative impedance characteristic curves, it was necessary to reduce the range of the analysed frequencies. The excitation signal within the range 4.5 kHz to 3 Hz, which required 1-second recordings for analysis purposes, made it possible to solve the non-stationary problem and obtain proper impedance spectra in the function of the exposure time (Fig. 4). Attention must be paid to the fact that the reduction of the lower frequency range of 0.3 Hz to 3 Hz, also brought negative effects. The third inductive time constant could not be determined.
TE D
M AN U
Bearing in mind the image of the obtained spectra (decreasing of two flattened semicircles) and the knowledge that, in the tested system, penetration by the electrolyte occurs together with the degradation of the PEO barrier coating and corrosion of the magnesium alloy, the analysis of individual impedance spectra was conducted based on the diagram of the equivalent circuit R0(Q1(R1(Q2R2))) (Fig. 5). Chi-squared value of the order of 3×10−5 or below was chosen to be acceptable for a given model. The results of the adjustment of the individual elements in the circuit in the time function were compared with the results of the corrosion potential, as presented in the diagram below (Fig. 6). The resistance R0, which represents the electrolyte resistance between working and reference electrodes, was not taken into account as it was an insignificant element.
AC C
EP
The elements Q1 and R1 represent the first time constant, which is detected in the upper range of frequencies. The values of the constant phase element continue to increase with a slight plateau during the 3 to 8 hours of exposure. They do not show any significant correlation with the corrosion potential of the sample. The resistance R1 shows a decreasing trend, which is particularly strong during the initial period of exposure. Very strong fluctuations of the resistance can be observed between 3 and 8 hours. The fluctuations are highly correlated with the fluctuations of the corrosion potential of the sample. After this period, the decreasing trend of the resistance persists but quick changes in the resistance become smaller and smaller. The presented changes in impedance parameters are no less than the result of reactions leading to the destruction of the PEO oxide coating and corrosion of the substrate. During the exposure in a water environment, conversion of the protective coating follows the equation (1): MgO + H2O → Mg(OH)2
(1)
This leads to a gradual thinning and partial degradation of the coating. In addition, penetration by the electrolyte occurs through pores and creating cracks between pores. The conversion of the coating is associated with a change in the dielectric constant which, when combined with the thinning of the coating, leads to a gradual increase in capacitance, is reflected by a change in the Q1 parameter. The electrolyte penetration through the PEO 5
ACCEPTED MANUSCRIPT oxide coating is represented by the decrease in the R1 resistance. The occurrence of fluctuations of this parameter should be attributed to local clogging of the pores by magnesium hydroxide which has a large volume, as compared to the initial MgO and Mg. The electrolyte penetrating the PEO oxide coating reaches the substrate, inducing the oxidation process (2): Mg → Mg2+ + 2e-
(2)
RI PT
At a further stage, cations which are formed with the participation of Cl- and OH- anions can be transformed as follows (3) [2,40]: xMg2+ + Cl- + yOH- + zH2O → Mgx(OH)yCl · zH2O
(3)
Mg ⇄ Mg+ + e-
SC
The magnesium oxidation process is also analysed as a two-stage hypothetical reaction (4-5) [14, 41,42]:
M AN U
Mg+ + H2O → Mg2+ + OH- + ½H2
(4) (5)
Needless to say, the oxidation reaction is accompanied by the reduction (6): 2H2O + 2e- → H2 + 2OH-
(6)
AC C
EP
TE D
The aforementioned phenomena generate a time constant based on charge transfer resistance and the capacity of the double electric layer. The degradation reactions, together with the exposure time, occur on a larger and larger surface causing a decrease in the measured resistance of the Rct charge transfer and an increase in the Cdl double electric layer. Moreover, attention should be paid to the fact that the aforementioned processes occur, especially during the initial period, at the bottom of narrow pores and cause the following phenomena: changes in the electrolyte concentrations and composition by the surface, blocking the pores with corrosion products, unblocking the pores by released hydrogen etc. They generate fluctuations of the corrosion potential, charge transfer resistance and the capacitance of the double electric layer. Such changes are described by R2 and Q2, i.e. the aforementioned phenomena should be attributed to these parameters. Q2 and R2 are elements constituting the second low-frequency time constant. The values of both elements fluctuate in accordance with the fluctuations of the corrosion potential. However, values of the fluctuation become smaller and smaller towards the end of the exposure. It should also be noticed that a very high increase in Q2 and a decrease in R2 can be observed over the first 2 hours. The general trend of both elements is consistent with the trend of the elements relating to the first time constant: an increase in the constant-phase element and a decrease in the resistance. Impedance changes for narrower period of time between 7.5 and 8.4 h of the exposition have been presented on Fig. 7 to highlight the scale of the non-stationary problem related to surface changes during time of exposition. Within less than one hour, the impedance of the material rapidly fluctuates. Impedance spectra fluctuation are correlated with corrosion potential presented in Fig. 8. Moreover on Fig. 8 changes of the equivalent 6
ACCEPTED MANUSCRIPT circuit elements during potential fluctuation have been presented. The behavior of the individual parameters is similar to previous conclusions. However, the particular attention should be paid to the fact, how much this parameters changes during short period of time. This results showed how important is time of exposition and time of the measurement of the studied PEO coating.
SC
RI PT
While explaining the origin of individual time constants, one should also mention the presence of the inductive constant, which is detected in the analysed system within the milihertz range. Unfortunately, the dynamics of the system degradation prevents the monitoring of this range by means of the impedance method. The most frequent explanation for the induction is the process of adsorption of intermediate products. This phenomenon is observed not only in the case of corrosion of magnesium alloys, but also in relation to aluminium and iron alloys [43-46].
4. Conclusions
TE D
M AN U
Additionally degradation of the coating surface before and after 19 h of exposition in 3.5% NaCl is presented on Fig. 9. Local pores have considerably expanded and micro cracks between pores have occurred during exposition. Clogging of the pores by corrosion products could lead to increasing of local stress and creating micro cracks. Confirmation of that is higher weight % of Mg around the pores (Fig. 9b), which could be directly related with emerging corrosion products as Mg(OH)2 and MgCl2 [47] described by first constant. Decrease of the total weight percentage of Mg on the surface suggests a gradual thinning of the coating during exposition. Furthermore, values of the weight % of aluminum and silicon on the coating surface after exposition increased significantly in compare to the coating without exposition. It is related to presence of the stable MgAl2O4 and Mg2SiO4/MgSiO3 [48,49] which has been formed during PEO process.
EP
The process of protective PEO coating degradation on magnesium alloys and corrosion of the substrate has a non-stationary nature. This prevents the use of a large number of research methods, including the conventional impedance method FRA.
AC C
Corrosion potential changes clearly proves the non-stationarity of the measured PEO coating. Classic methods of the measurements often require a long period of time to study exposed materials and force constant potential different then open corrosion potential. As results, the informations may differ from the real state. As presented in this paper, measured coating system is rapidly changing its surface properties during time. Therefore, fast obtain the information about measured system and time of exposition before measurements are extremely important. As presented in this paper, it is possible to obtain instantaneous impedance characteristic curves and monitor impedance changes by using a multisinusoidal signal within an appropriate range of frequencies. The performance of such tests in the galvanostatic mode, while keeping the zero resultant current between the working and counter electrodes, made it possible to track changes both in impedance parameters and in the corrosion potential. The analysis showed that the high-frequency range of the spectrum is determined by the properties of the barrier layer. It is possible to track its penetration and 7
ACCEPTED MANUSCRIPT degradation by the electrolyte in detail. Within the low-frequency range, the detection of charge transfer resistance and the capacitance of the double electric layer occur.
References
RI PT
[1] J.E. Gray, B. Luan, Protective coatings on magnesium and its alloys — a critical review, J. Alloys Compd. 336 (2002) 88–113. doi: 10.1016/S0925-8388(01)01899-0. [2] Y. Gao, A. Yerokhin, A. Matthews, Deposition and evaluation of duplex hydroxyapatite and plasma electrolytic oxidation coatings on magnesium, Surf. Coatings Technol. 269 (2015) 170–182. doi: 10.1016/j.surfcoat.2015.01.030.
SC
[3] J. Liang, P.B. Srinivasan, C. Blawert, W. Dietzel, Comparison of electrochemical corrosion behaviour of MgO and ZrO2 coatings on AM50 magnesium alloy formed by plasma electrolytic oxidation, Corros. Sci. 51 (2009) 2483–2492. doi: 10.1016/j.corsci.2009.06.034.
M AN U
[4] S. Begum, D.L. Chen, S. Xu, A.A. Luo, Strain-controlled low-cycle fatigue properties of a newly developed extruded magnesium alloy, Metall. Mater. Trans. A Phys. Metall. Mater. Sci. 39 (2008) 3014–3026. doi:10.1007/s11661-008-9677-0. [5] C.R. Hutchinson, J.F. Nie, S. Gorsse, Modeling the precipitation processes and strengthening mechanisms in a Mg-Al-(Zn) AZ91 alloy, Metall. Mater. Trans. A. 36 (2005) 2093–2105. doi:10.1007/s11661-005-0330-x. [6] Z. Li, A.A. Luo, Q. Wang, H. Zou, J. Dai, L. Peng, Fatigue characteristics of sand-cast AZ91D magnesium alloy, J. Magnes. Alloy. 5 (2017) 1–12. doi:10.1016/j.jma.2017.03.001.
TE D
[7] L. Wang, B.-P. Zhang, T. Shinohara, Corrosion behavior of AZ91 magnesium alloy in dilute NaCl solutions, Mater. Des. 31 (2010) 857–863. doi:10.1016/j.matdes.2009.07.049.
EP
[8] T.M. Mukhametkaliyev, M.A. Surmeneva, A. Vladescu, C.M. Cotrut, M. Braic, M. Dinu, M.D. Vranceanu, I. Pana, M. Mueller, R.A. Surmenev, A biodegradable AZ91 magnesium alloy coated with a thin nanostructured hydroxyapatite for improving the corrosion resistance, Mater. Sci. Eng. C. 75 (2017) 95–103. doi:10.1016/j.msec.2017.02.033.
AC C
[9] C.L. Liu, J. Jiang, M. Wang, Y.J. Wang, P.K. Chu, W.J. Huang, Degradation and Biocompatibility of WE43, ZK60, and AZ91 Biodegradable Magnesium Alloys, Adv. Mater. Res. 287–290 (2011) 2008– 2014. doi:10.4028/www.scientific.net/AMR.287-290.2008. [10] Z. Wen, S. Duan, C. Dai, F. Yang, F. Zhang, Biodegradability and surface chemistry of AZ31D compared with AZ91 magnesium alloy in a modified simulated body fluid, Int. J. Electrochem. Sci. 9 (2014) 7846–7864. [11] R. Rojaee, M. Fathi, K. Raeissi, Controlling the degradation rate of AZ91 magnesium alloy via solgel derived nanostructured hydroxyapatite coating, Mater. Sci. Eng. C. 33 (2013) 3817–3825. doi:10.1016/j.msec.2013.05.014. [12] A.R. Bharat, B.S.S. Kumar, R.J.P. Sarath, P. Akhil, G.P.K. Reddy, B.R. Sunil, Influence of Al Content on the Corrosion Behavior of Biodegradable Magnesium Alloys in Simulated Physiological Solution, J. Adv. Biomed. Eng. Technol. 2 (2015) 17–23.
8
ACCEPTED MANUSCRIPT [13] S. V. Gnedenkov, O.A. Khrisanfova, A.G. Zavidnaya, S.L. Sinebryukhov, V.S. Egorkin, M. V. Nistratova, A. Yerokhin, A. Matthews, PEO coatings obtained on an Mg-Mn type alloy under unipolar and bipolar modes in silicate-containing electrolytes, Surf. Coatings Technol. 204 (2010) 2316–2322. doi: 10.1016/j.surfcoat.2009.12.024. [14] Y. Gao, A. Yerokhin, A. Matthews, Effect of current mode on PEO treatment of magnesium in Caand P-containing electrolyte and resulting coatings, Appl. Surf. Sci. 316 (2014) 558–567. doi: 10.1016/j.apsusc.2014.08.035.
RI PT
[15] A. Ghasemi, V.S. Raja, C. Blawert, W. Dietzel, K.U. Kainer, Study of the structure and corrosion behavior of PEO coatings on AM50 magnesium alloy by electrochemical impedance spectroscopy, Surf. Coatings Technol. 202 (2008) 3513–3518. doi: 10.1016/j.surfcoat.2007.12.033.
SC
[16] S.V. Gnedenkov, S.L. Sinebryukhov, D.V. Mashtalyar, I.M. Imshinetskiy, A.S. Gnedenkov, A.V. Samokhin, Y.V. Tsvetkov, Protective composite coatings obtained by plasma electrolytic oxidation on magnesium alloy MA8, Vacuum. 120 (2015) 107–114. doi: 10.1016/j.vacuum.2015.02.004.
M AN U
[17] H. Duan, C. Yan, F. Wang, Effect of electrolyte additives on performance of plasma electrolytic oxidation films formed on magnesium alloy AZ91D, Electrochim. Acta. 52 (2007) 3785–3793. doi: 10.1016/j.electacta.2006.10.066. [18] D. Sreekanth, N. Rameshbabu, K. Venkateswarlu, Effect of various additives on morphology and corrosion behavior of ceramic coatings developed on AZ31 magnesium alloy by plasma electrolytic oxidation, Ceram. Int. 38 (2012) 4607–4615. doi: 10.1016/j.ceramint.2012.02.040. [19] X. Li, X. Liu, B.L. Luan, Corrosion and wear properties of PEO coatings formed on AM60B alloy in NaAlO2 electrolytes, Appl. Surf. Sci. 257 (2011) 9135–9141. doi: 10.1016/j.apsusc.2011.05.115.
TE D
[20] D.R. Marsh, G. Li, The biology of fracture healing: optimising outcome, Br. Med. Bull 55 (1999) 856–869. doi: 10.1258/0007142991902673. [21] Y. Wang, F. Wang, M. Xu, B. Zhao, L. Guo, J. Ouyang, Microstructure and corrosion behavior of coated AZ91 alloy by microarc oxidation for biomedical application, Appl. Surf. Sci. 255 (2009) 9124– 9131. doi: 10.1016/j.apsusc.2009.06.116.
AC C
EP
[22] P. Wan, X. Lin, L. Tan, L. Li, W. Li, K. Yang, Influence of albumin and inorganic ions on electrochemical corrosion behavior of plasma electrolytic oxidation coated magnesium for surgical implants, Appl. Surf. Sci. 282 (2013) 186–194. doi: 10.1016/j.apsusc.2013.05.100. [23] Q. Zhao, X. Guo, X. Dang, J. Hao, J. Lai, K. Wang, Preparation and properties of composite MAO/ECD coatings on magnesium alloy, Colloids Surf. B Biointerfaces102 (2013) 321–326. doi: 10.1016/j.colsurfb.2012.07.040. [24] S. Chen, S. Guan, W. Li, H. Wang, J. Chen, Y. Wang, H. Wang, In vivo degradation and bone response of a composite coating on Mg–Zn–Ca alloy prepared by microarc oxidation and electrochemical deposition, J. Biomed. Mater. Res. B. Appl Biomater, 100 (2012) 533–543. doi: 10.1002/jbm.b.31982. [25] Y. Gu, S. Bandopadhyay, C. Chen, Y. Guo, C. Ning, Effect of oxidation time on the corrosion behavior of micro-arc oxidation produced AZ31 magnesium alloys in simulated body fluid, J. Alloys Compd. 543 (2012) 109–117. doi: 10.1016/j.jallcom.2012.07.130.
9
ACCEPTED MANUSCRIPT [26] A. Castellanos, A. Altube, J.M. Vega, E. García-Lecina, J.A. Díez, H.J. Grande, Effect of different post-treatments on the corrosion resistance and tribological properties of AZ91D magnesium alloy coated PEO, Surf. Coatings Technol. 278 (2015) 99–107. doi: 10.1016/j.surfcoat.2015.07.017. [27] Y. Song, K. Dong, D. Shan, E.H. Han, Investigation of a novel self-sealing pore micro-arc oxidation film on AM60 magnesium alloy, J. Magnes. Alloy. 1 (2013) 82–87. doi:10.1016/j.jma.2013.02.009.
RI PT
[28] L. Pezzato, K. Brunelli, S. Gross, M. Magrini, M. Dabalà, Effect of process parameters of plasma electrolytic oxidation on microstructure and corrosion properties of magnesium alloys, J. Appl. Electrochem. 44 (2014) 867–879. doi: 10.1007/s10800-014-0695-x. [29] Y. Wang, J. Wang, J. Zhang, Z. Zhang, Characteristics of anodic coatings oxidized to different voltage on AZ91D Mg alloy by micro-arc oxidization technique, Mater. Corros. 56 (2005) 88–92. doi: 10.1002/maco.200403822.
SC
[30] T. Chen, W. Xue, Y. Li, X. Liu, J. Du, Corrosion behavior of friction stir welded AZ31B magnesium alloy with plasma electrolytic oxidation coating formed in silicate electrolyte, Mater. Chem. Phys. 144 (2014) 462–469. doi: 10.1016/j.matchemphys.2014.01.020.
M AN U
[31] K.Darowicki, Theoretical description of the measuring method of instantaneous impedance spectra, J. Electroanal. Chem. 486 (2000) 101-105. doi: 10.1016/S0022-0728(00)00110-8 [32] K. Darowicki, P. Slepski, Dynamic electrochemical impedance spectroscopy of the first order electrode reaction, J. Electroanal. Chem. 547 (2003) 1-8. doi: 10.1016/S0022-0728(03)00154-2 [33] A.M. Kumar, S.H. Kwon, H.C. Jung, Y.H. Park, H.J. Kim, K.S. Shin, Fabrication and Electrochemical Corrosion Behavior of PEO Coatings on Strip-Cast AZ31Mg Alloy in 3 . 5 % NaCl Solution, Ind. Eng. Chem. Res. 53 (2014) 9703−9713. doi: 10.1021/ie404329s
TE D
[34] R.O. Hussein, D.O. Northwood, X. Nie, The effect of processing parameters and substrate composition on the corrosion resistance of plasma electrolytic oxidation (PEO) coated magnesium alloys, Surf. Coatings Technol. 237 (2013) 357–368. doi: 10.1016/j.surfcoat.2013.09.021.
EP
[35] P. Su, X. Wu, Y. Guo, Z. Jiang, Effects of cathode current density on structure and corrosion resistance of plasma electrolytic oxidation coatings formed on ZK60 Mg alloy, J. Alloys Compd. 475 (2009) 773–777. doi: 10.1016/j.jallcom.2008.08.030.
AC C
[36] P. Slepski, K. Darowicki, K. Andrearczyk, On-line measurement of cell impedance during charging and discharging process, J. Electroanal. Chem. 633 (2009) 121–126. doi: 10.1016/j.jelechem.2009.05.002. [37] J. Ryl, J. Wysocka, P. Slepski, K. Darowicki, Instantaneous impedance monitoring of synergistic effect between cavitation erosion and corrosion processes, Electrochim. Acta. 203 (2015) 388–395. doi: 10.1016/j.electacta.2016.01.216. [38] H. Gerengi, P. Slepski, E. Ozgan, M. Kurtay, Investigation of corrosion behavior of 6060 and 6082 aluminum alloys under simulated acid rain conditions, Mater. Corros. 66 (2015) 233–240. doi: 10.1002/maco.201307287. [39] J. Ryl, K. Darowicki, P. Slepski, Evaluation of cavitation erosion-corrosion degradation of mild steel by means of dynamic impedance spectroscopy in galvanostatic mode, Corros. Sci. 53 (2011) 1873–1879. doi: 10.1016/j.corsci.2011.02.004.
10
ACCEPTED MANUSCRIPT [40] X. Zhou, L. Jiang, P. Wu, Y. Sun, Y. Yu, G. Wei, H. Ge, Effect of aggressive ions on degradation of WE43 magnesium alloy in physiological environment, Int. J. Electrochem. Sci. 9 (2014) 304–314. [41] G. Song, D.S.T. John, J. Nairn, The anodic dissolution of magnesium and sulphate solutions in chloride, Corros. Sci. 39 (1997) 1981–2004. doi: 10.1016/S0010-938X(97)00090-5. [42] F.H. Cao, V.H. Len, Z. Zhang, J.Q. Zhang, Corrosion behavior of magnesium and its alloy in NaCl solution, Russ. J. Electrochem. 43 (2007) 837–843. doi: 10.1134/S1023193507070142.
RI PT
[43] X. Li, S. Deng, X. Xie, Experimental and theoretical study on corrosion inhibition ofoxime compounds for aluminium in HCl solution, Corros. Sci. 81 (2014) 162–175. doi: 10.1016/j.corsci.2013.12.021.
SC
[44] N. Chaubey, V.K. Singh, Savita, M.A. Quraishi, E.E. Ebenso, Corrosion inhibition of aluminium alloy in alkaline media by Neolamarkia Cadamba bark extract as a green inhibitor, Int. J. Electrochem. Sci. 10 (2015) 504–518.
M AN U
[45] M. Kissi, M. Bouklah, B. Hammouti, M. Benkaddour, Establishment of equivalent circuits from electrochemical impedance spectroscopy study of corrosion inhibition of steel by pyrazine in sulphuric acidic solution, Appl. Surf. Sci. 252 (2006) 4190–4197. doi: 10.1016/j.apsusc.2005.06.035. [46] G.A. Zhang, Y.F. Cheng, On the fundamentals of electrochemical corrosion of X65 steel in CO2containing formation water in the presence of acetic acid in petroleum production, Corros. Sci. 51 (2009) 87–94. doi: 10.1016/j.corsci.2008.10.013. [47] H.S. Ryu, D.-S. Park, S.-H. Hong, Improved corrosion protection of AZ31 magnesium alloy through plasma electrolytic oxidation and aerosol deposition duplex treatment, Surf. Coatings Technol. 219 (2013) 82–87. doi: 10.1016/j.surfcoat.2013.01.008.
TE D
[48] M. Daroonparvar, M.A.M. Yajid, N.M. Yusof, H.R. Bakhsheshi-Rad, Preparation and corrosion resistance of a nanocomposite plasma electrolytic oxidation coating on Mg-1%Ca alloy formed in aluminate electrolyte containing titania nano-additives, J. Alloys Compd. 688 (2016) 841–857. doi: 10.1016/j.jallcom.2016.07.081.
AC C
EP
[49] G. Barati Darband, M. Aliofkhazraei, P. Hamghalam, N. Valizade, Plasma electrolytic oxidation of magnesium and its alloys: Mechanism, properties and applications, J. Magnes. Alloy. 5 (2017) 74– 132. doi: 10.1016/j.jma.2017.02.004.
11
ACCEPTED MANUSCRIPT Figure captions Fig. 1 SEM images of the oxide coating formed at 5 °C using 30% impulse filling with current and the impulse current frequency of 1,000 Hz on the AZ91D magnesium alloy: a) Coating surface 0.002 mm2; b-c) Cross section of the coating with visible canals through during plasma discharges; d) Distribution of the coating thickness.
RI PT
Fig. 2 Example records of changes in the corrosion potential during a 20-hour measurement. Fig. 3 Example of instantaneous impedance spectra obtained after 4 min. (+), 9 h (◊) and 19 h (). Frequency excitation within the range 4.5 kHz to 0.3 Hz
SC
Fig. 4 Changes in the impedance characteristic curve of the material during the experiment within the excitation frequency range 4.5 kHz to 3 Hz.
M AN U
Fig. 5 Equivalent circuit model used to describe corrosion degradation process of PEO on magnesium alloy. Fig. 6 Change in the value of individual elements in the equivalent circuit (R1, Q1, R2, Q2) as regards changes in the corrosion potential in the time function (grey doped lines - corrosion potential. Fig. 7 Changes in the impedance characteristic curve of the material during the experiment for short period of time within the excitation frequency range 4.5 kHz to 3 Hz.
TE D
Fig. 8 Change in the value of individual elements in the equivalent circuit (R1, Q1, R2, Q2) as regards changes in the corrosion potential for short period of time (grey doped lines corrosion potential.
AC C
EP
Fig. 9 EDX analysis of coating surface before and after 19 h exposition.
12
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Highlights
EP
TE D
M AN U
SC
RI PT
Impedance technique with continuous multisinusoidal current perturbation was used. An explanation of corrosion process of PEO coated magnesium alloy is proposed. Galvanostatic mode allowed to achieve corrosion potential in time of experiment. Changes of equivalent circuit parameters allowed to explain corrosion process.
AC C
• • • •