Electrodeposition of chemically and mechanically protective Al-coatings on AZ91D Mg alloy
Haiyan Yang,Xingwu Guo,Guohua Wu,Wenjiang Ding,Nick Birbilis
DOI: https://doi.org/10.1016/j.corsci.2010.09.047
IF: 7.72
2011-01-01
Corrosion Science
Abstract:Electrodeposition of aluminium coatings upon AZ91D in aluminum chloride/1-ethyl-3-methylimidazolium chloride ionic liquid was achieved. Post-plating heat treatment processes compatible with AZ91D solution treatment (420 °C) and aging treatment (200 °C) were explored to improve coatings adhesion and hardness, and to maintain corrosion resistance. 420 °C treatment produced a β-phase (Mg 17 Al 12 ) enriched two-phase coating; whilst treatment at 200 °C leads to a tri-layer structure, rich in γ-phase (Mg 2 Al 3 ). The 200 °C treatment was shown to be most effective for corrosion resistance, eradicating water reduction as the principal cathodic reaction and increasing surface hardness. Keywords A. Magnesium A. Metal coatings A. Intermetallics B. Polarization C. Electrodeposited films 1 Introduction The surface properties of cast Mg alloys, including low corrosion and poor wear resistance, limits their wider application [1] . Surface modification is therefore necessary to improve the corrosion resistance of Mg alloys in essentially all cases. There are many possible coating technologies available for Mg and its alloys [2–18] . Conversion coatings are the simplest coatings, but such coatings do not readily permit for deposition of a range of other metals in a manner that allows metallurgical bonding. As a result, electroplating is the most effective processes for introducing metallic coatings upon an Mg substrate [19–22] . Metallic coatings upon Mg serve a large market and are of great interest owing to: (1) the low resistivity of the coated Mg alloy can be maintained, (2) electromagnetic shielding properties can be maintained, (3) metallic coatings have a good decorative appearance and (4) metallic coatings can resist light damage unlike most conversion coatings. Protection of Mg with metallic Al protection has received extensive attention to date, since Al coatings do not significantly increase the density of the material, corrosion resistance of Al is good (along with a mature anodizing industry) and recyclability of the overall material can be maintained since Al is a primary alloying element for the AZ series of Mg alloys. Recently, authors have even explored thermal/cold spray technology and laser cladding to achieve Al coatings upon Mg [23,3,24,25] . Cold spray has certain limitations of its own, namely cost, and hence parallel studies into Al-coatings continue. Of such studies, it has been noted that successful electrodeposition of Al and its alloys can be performed upon Mg from acidic aluminum chloride 1-ethyl-3-methylimidazolium chloride (AlCl 3 -EMIC) ionic liquids (ILs) with the proper molar ratios [21,26,27] . However, besides the corrosion resistance that may be imparted from such coatings, the adhesion between the substrate and the coating is another important characteristic to be considered. It is evident that as-plated Al has peeled off from Mg substrate in some of the cross-sectional micrographs in previous works [21,27] . In air, Mg is prone to form an oxide/hydroxide layer quickly, and this may prevent the adhesion of the electrodeposited layer if the oxide persists in the electroplating medium used. To address this in our previous work, various pretreatments for AZ91D substrates were conducted to improve adhesion. However, the quality of the coating was not significantly enhanced. Furthermore, the wear resistance of the Al coating was also limited. Consequently, in order to enhance coating adhesion strength and subsequent wear resistance, a process combining electrodeposition and the heat treatment is explored herein. As background to the response of Al coatings to thermal exposure some background is given here. Previous studies have confirmed that diffusion coatings can be obtained by heating Mg in contact with Al powder or bulk materials [28–30] . Those diffusion coatings can improve both corrosion and wear resistance of AZ91D Mg alloy through the formation of an Mg/Al intermetallic compound. Shigematsu et al. [28] have created an Al diffusion coating mainly consists of α-phase (Mg) and β-phase (Al 12 Mg 17 ) on AZ91D Mg alloy by heating the sample imbedded in Al powder. Since the applied temperature was higher than the eutectic temperature of Mg/Al 12 Mg 17 (437 °C), surface cracks were observed within the layer. Through the addition of Zn into the diffusion process Zhang et al. [31] and Zhu et al. [32] have produced Mg 17 Al 12 -enriched layers on an AZ91 substrate at lower temperatures (430 °C and 420 °C), however, the layers formed were very heterogeneous. Zhang et al. [33,34] have used a technique known as surface mechanical attrition treatment (SMAT) to form a nanocrystalline surface layer on AZ91 alloy, combined with surface alloying using Al powder. Short-circuit grain boundary diffusion of the Al into the nanocrystalline layer lowers the diffusion coating temperature to 380 °C. Unfortunately, SMAT is not suited to components with complex shapes. Additionally, all of the above reported approaches only produce alloyed layers with a higher proportion of the Mg 17 Al 12 phase than exists in the as-cast Mg matrix; whilst none of them displayed continuous metallic coatings. In the present work, Al coatings were deposited onto AZ91D by electrodeposition from AlCl 3 -EMIC IL, with subsequent heat treatment carried out. The aim is to develop a cost effective and simple method to form protective Al coatings on Mg surfaces. Considering practical alloy processing heat treatments for AZ91D, i.e. solution treatment ∼420 °C and aging treatment ∼200 °C, these temperatures were studied. The effect of the heat treatment on the surface properties of the Al coatings on AZ91D Mg alloy was investigated, including the structure, hardness and corrosion resistance. 2 Experimental 2.1 Electrodeposition and coating heat treatment An acidic IL with a molar ratio of AlCl 3 -EMIC of 1.5:1 was prepared by slow addition of AlCl 3 (anhydrous powder, 99.99%, Aldrich) into EMIC (99%, Aldrich) at room temperature. All chemicals were handled under a purified nitrogen atmosphere in a glove box (H 2 O ⩽ 1 ppm, O 2 ⩽ 1 ppm). The mixture was continuously stirred (by magnetic bar) for 24 h to ensure uniformity. As-cast AZ91D Mg alloy (nominal composition Mg-9 wt%Al-1 wt% Zn) was used in this study. Specimens were cut into disks with the geometry of 11.3 mm and 5 mm thickness. A 1 mm hole was drilled and tapped on the side of the specimen so that a copper wire could be attached. The specimens were then fixed to the copper wire and mounted with an appropriate resin exposing a free surface of 1 cm 2 to be electroplated. Prior to electrodeposition, each sample was ground with SiC papers to 1000 grit in the glove box. The Al coatings were deposited onto AZ91D by electrodeposition from the AlCl 3 -EMIC IL at room temperature. Electrodeposition was carried using a conventional three-electrode cell arrangement and a Parstat 2273 Potentiostat/Galvanostat (Princeton Applied Research). The AZ91 was assembled as the working electrode, and a pure Al sheet and an Al wire were used for the counter and reference electrodes, respectively. Cathodic deposition was performed under a galvanostatically at 15 mA/cm 2 for 1 h. After deposition, the samples were thoroughly cleaned using distilled water and air dried. The Al-coated AZ91D samples were heat treated in a vacuum oven with a vacuum pressure <10 −3 Pa at 420 °C for 2 h or 200 °C for 12 h and air cooled. 2.2 Characterization The surface and cross-section morphology were observed using a JEOL JSM-6460 SEM, coupled with energy dispersive X-ray spectrometry (EDXS). Phase identification was determined by X-ray diffraction (D/MAX 2550V) with a Cu-Kα target. 2 θ / θ scanning mode was obtained in the 2 θ = 20–80° range and the scanning speed was 5°/min with a step of 0.01°. Nanoindentation test was performed on the sample surfaces using a CSM nanohardness tester. The indentation load was fixed at 2000 μN and the loading/unloading rate remained constant at 4000 μN per minute. A stabilization time of 5 s between the loading and unloading stages was used. The elastic modulus (EIT) and hardness (nHV) were calculated from the load–displacement curves as per [35] . The elastic modulus (EIT) was determined upon the slope ( S ) of the unloading curve using the following equation, which is in fact a measure of the elastic property of the surface [36] . (1) E r = π S 2 A where A is the projected area after the indentation. To evaluate the effect of heat treatment on corrosion resistance, two methods were used. The measurement of hydrogen evolution was employed in order to measure the cumulative corrosion evolved during an extended period of immersion in 3.5 wt.% NaCl solution at 25 ± 2 °C for 72 h. Additionally, electrochemical behavior was studied by potentiodynamic polarization. A high density graphite electrode and a saturated calomel electrode (SCE) were used as the counter electrode and the reference electrode, respectively. Prior to polarization, the specimens remained at open circuit for 30 min to reach a steady value. Scans were obtained from 100 mV below OCP to and scanned upwards at a rate of 0.5 mV/s. 3 Results and discussion 3.1 Microstructure and morphology of Al coatings 3.1.1 As electrodeposited The as electrodeposited Al-coated sample was silver-gray in appearance and the surface coverage was complete as shown in Fig. 1 a. Fig. 1 b shows the corresponding electron micrograph of the as electrodeposited coating, revealing that a dense faceted microstructure with the crystal size in the range of 2–5 μm was clearly observed. A typical cross-section micrograph of the Al-coated Mg alloy is presented in Fig. 1 c. What can be seen, and unlike pack diffusion coatings, is that the Al coating is continuous and the thickness of the deposited layer was uniform (at approximately 18 μm, which is close to the theoretical value according to Faraday’s law). This suggests that the deposition current efficiency was high and adjusting the total charge applied could control the thickness of the deposit. Fig. 1 d is a typical EDXS line scan across the layer and into the substrate, as depicted by the line on the micrograph of the sectioned specimens. Whilst EDXS is semi-quantitative, a uniform distribution of Al through the deposited layer is readily observed, along with a detectible amount of O detected at the interface between Al layer and the AZ91D substrate. In addition to the EDXS, phase analysis was carried out using XRD, and the acquired diffraction patterns are shown in Fig. 2 . Fig. 2 includes a control curve for the AZ91D substrate as “Curve a”, revealing the dual-phase nature as expected previous (α-Mg solid solution phase and Mg 17 Al 12 β-phase). “Curve b” represents the Al-coated Mg alloy sample. A diffraction pattern ascribed to pure Al, with a face-centered-cubic (FCC) structure, is clearly identified. In Curve b, the signals from the Mg alloy substrate were highly suppressed as the surface was covered by the deposited Al layer. Routine handling of the as electrodeposited specimen, along with simple adhesion peel-off tests (using the tape test) could be used to assess the adhesion of the deposited layer. Unfortunately, the adhesion strength was not high, and following tape testing, bubbles were observed, indicating that the adhesion between the Al coating and the Mg substrate is poor. A possible reason for this may be the high content of O at the interface. As a result, this adhesion issue, along with the softness of pure Al, gives rise to the notion of post electrodeposition heat treatment. 3.1.2 After heat treatment at 420 °C/2 h Fig. 3 a shows the surface morphology of the Al coating on the AZ91D Mg alloy after heat treatment at 420 °C/2 h. It is seen that the surface morphology of the deposited Al coating varied significantly after heat treatment. This could be due to large scale diffusion and atomic rearrangement, possibly aided by some local melting. The cross-section morphology in Fig. 3 b shows that a two-phase structure coating with the thickness of 100–140 μm evolved on the surface of the specimen, and integrated with the substrate by metallurgical bonding. This, as mentioned by Spencer et al. [23] should greatly improve the adhesion strength of the coating. In the coating itself, a light phase formed a continuous network, and a dark phase was co-mingled in the light phase. EDXS results of the two phases labeled in Fig. 3 a are shown in Table 1 . As expected from the backscattered image contrast, the Al content of the light phase was much higher than that of the dark phase. The composition of the light phase was similar to β-phase (Al 12 Mg 17 ), and indeed this phase was confirmed by XRD, as shown in Fig. 3 c. However, instead of peaks of Al phase, only peaks of α-phase (Mg) and β-phase (Mg 17 Al 12 ) were detected with no remaining FCC (Al) phase remaining. As a result, it is seen that the 420 °C/2 h treatment results in a coating composed of α-Mg and β-phase (Mg 17 Al 12 ). Furthermore, the two-phase surface is rich in β-phase. 3.1.3 After heat treatment at 200 °C/12 h The top down surface morphology of the electrodeposited Al coating before and after the heat treatment at 200 °C was similar (dense faceted structure). However, in response to the 200 °C/12 h heat treatment, the changes that occurred in the coating morphology and structure are observed via the cross-section morphology, as shown in Fig. 4 a. The coating is composed of three distinct layers after the heat treatment at 200 °C. In addition to this, it is obvious that a boundary and some voids have formed in the mid layer. This is a phenomenon that may have been expected based on recent work of Spencer and co-workers on the heat treatment of various cold spray coatings. The authors attribute this to the Kirkendall effect that arises from to asymmetrical diffusion [37] . We note however, that the extent of voids observed is actually less than that which is observed following cold spray coating, presumably because cold spray coatings have significantly different strain characteristics and also the grain sizes (coating and substrate) are likely much smaller and diffusion more rapid. Based on the location of the voids, the boundary in the mid layer is probably the original interface of the Al coating and Mg substrate since the thickness of the layer above the boundary is ∼18 μm in total, as shown in Fig. 4 a. The tri-layer structure is further studied by EDXS line profile analysis, as shown in Fig. 4 b. Combination the EDXS analysis of the three layer summarized in Table 1 , the most probable three phases were estimated as: the first layer is Al(Mg) solid solution, the second layer is γ-phase (Mg 2 Al 3 ), and the third layer is β-phase (Mg 17 Al 12 ). To quantify this, the structure was further studied by XRD, and a solid solution Al (Mg) phase, γ-phase (Mg 2 Al 3 ) and β-phase (Mg 17 Al 12 ), were identified in the spectrum, as shown in Fig. 4 c. Consequently we propose that the three-layer coating labeled in Fig. 4 a is comprised of: I: top layer, Al (Mg) solid solution II: mid layer, γ-phase (Mg 2 Al 3 ) III: bottom layer, β-phase (Mg 17 Al 12 ) We note however, that the top layer was found to be removed/detached somewhat easily, owing to poor adhesion. This is somewhat of a concern and an area that is very worthy of future work (discussed further below). In instances where the top solid solution layer is removed, the cross-section morphology is seen in Fig. 4 d. Hence, a two-layer structure of γ-phase and β-phase remains very robust (due to metallurgical bonding) after heat treatment at 200 °C/12 h. 3.2 Nanomechanical coating properties The nanomechanical behavior of the coating was specifically evaluated by nanoindentation testing. Fig. 5 shows the typical nanoindentation load–displacement curves for the Al coating for all conditions studied herein. As a baseline, the mechanical properties of the AZ91D Mg substrate were also measured. The elastic modulus (EIT) and the nano-hardness (nHV) determined from the load–displacement are listed in Table 2 . The hardness of the AZ91D decreased from 1.1 GPa to 0.7 GPa with the as-electrodeposited layer was tested. After heat treatment at 420 °C/2 h, the hardness significantly increased. This was attributed to the strengthening effect of β-phase (Mg 17 Al 12 ) [33] . Given the size of the nano-indent being necessarily small, it was possible, in fact necessary, to determine the nanomechanical properties of the 420 °C/2 h coating according to the properties of the two distinct (light and dark) phases. It is interesting to find that the hardness of α-Mg in the coating is much higher than that of AZ91D Mg substrate. This can be attributed to the solute content in α-Mg that comprises the second phase of the coating (with Al and Mg known to very effectively solid solution strengthen the other). In combination, assuming that α and β phase play an equal role in hardness, according the volume percentage of these two phases by metallography, an average hardness can be calculated as 3.07 GPa, which is nearly three times higher than that of AZ91D substrate. Following heat treatment at 200 °C/12 h surface hardness is further increased to 3.5 GPa which is greater than three times higher than that of AZ91D substrate. This indicates that the intermetallic compounds γ and β phase have significantly higher hardness than the Mg substrate as expected. We concede that hardness is not a perfect measure of wear resistance, nor in fact is any single wear test since wear of a given material may vary substantially depending on the mode of wear. For this study, in the absence of a requirement for a specific mode of wear resistance, hardness will be taken as a generic measure of abrasion, scratch and wear resistance. Hence, the increase of hardness implies that the mechanical and wear resistance of the Al coating is improved after heat treatment. Table 2 also shows the variation of elastic modulus (EIT) before and after heat treatment. The elastic modulus (EIT) of the AZ91D increased when it was coated Al layer, and further increased after heat treatment, suggesting that the intermetallic compounds have higher elastic modulus, which is generally true of intermetallics, however more importantly for this work, confirming that a metallurgical bond has occurred. 3.3 Corrosion performance of Al-coated AZ91D To determine both the longer-term performance and also the mechanistic aspects relating to the Al-coating, the hydrogen evolution test and potentiodymanic polarization were both conducted. 3.3.1 Hydrogen evolution tests The typical representation of the hydrogen evolved over 72 h immersion in 3.5 wt.% NaCl solution is seen for all specimen tested in Fig. 6 . As is well documented, measuring the volume of hydrogen evolution is equivalent to measuring the weight-loss of Mg dissolved, and the measured hydrogen evolution rate is equal to the weight-loss rate if both of them have been converted into the same unit since the reduction of water is the sole cathodic reaction at the potentials where Mg corrosion occurs [38] . The results showed that the rate of the hydrogen evolution for the AZ91D substrate increased with immersion time, signaling that the corrosion rate of the Mg substrate not only occurred, but proceeded with immersion time. For the specimen that was Al-coated and heat-treated at 420 °C/2 h, some hydrogen evolution was detected and increased with time indicating that corrosion was proceeding. Clearly, the slope of the line was more subtle than the uncoated AZ91, hence the rate of hydrogen evolution was reduced suggesting that the 420 °C/2 h Al coating provide some protection for AZ91D. What is interesting to note, is that for the Al coating, no hydrogen evolved during immersion and similarly no hydrogen evolved for the 200 °C/12 h Al coated specimen. This observation reveals several important things. One must recall that the measurement of hydrogen is specific to the dissolution of Mg and not Al, since the cathodic reaction for Mg dissolution is responsible for hydrogen evolution. As a result, from Fig 6 , it reveals that if any corrosion did occur on such specimens (the as deposited and 200 °C/12 h) that the corresponding cathodic reaction was not evolving hydrogen. Also, as importantly, it also indicated that after 72 h of testing, there was no breach in the protectiveness of the coating, since no hydrogen was detected, suggesting that the dissolution of Mg was suppressed. This is a significant finding, since it indicates rather starkly, that the 200 °C/12 h treated coating was generating a surface that is rich enough in Al to create a surface potential that does not support water reduction as the cathodic reaction, and hence dramatically improves the corrosion resistance of AZ91D in aqueous environments. This is not the case for the 420 °C/2 h coating, and hence the 420 °C/2 h treatment is not of significant utility for corrosion protection. The above results can also be contrast with reports in the literature, in which have reported that it may be possible to protect corrosion of Mg alloys if a continuous network of β phase is formed on the surface [39] . This is not necessarily true, since at best, the surface that would contain only β phase would still be subject to water reduction as the dominant cathodic reaction, and as reported in instances where authors have presented the electrochemical behavior of β phase alone, the electrochemical properties of β phase are not significantly greater (in an engineering context with respect to corrosion resistance) than a range of Mg alloys [40] . Any protective coating should offer, at a minimum, protection against water reduction. 3.3.2 Potentiodynamic testing Fig. 7 gives a typical potentiodynamic polarization curve for each of the specimens studied herein; with the results of potentiodynamic polarization summarized in Table 3 . As expected, the AZ91D reveals the highest corrosion current density and shows high rates of reaction at highly negative potentials characteristic of Mg alloys. The as-electrodeposited Al coating shows a response that is typical of pure Al, with a noticeable passivation range and a lower corrosion current density than AZ91D. After heat treatment at 420 °C/2 h, the resultant coating reveals a polarization response that is typical of the reported response of β phase. Whilst the corrosion current density is reduced to that of AZ91D, we see that the alteration is not dramatic. Also, whilst evidenced by the hydrogen evolution data, the polarization data also reveals that the kinetics of the cathodic branch of the polarization curves for the AZ91D and the 420 °C/2 h specimens are somewhat similar, consistent with water reduction being the dominant and shared cathodic reaction. As may have been predicted from the hydrogen evolution data, the 200 °C/12 h displays electrochemical characteristics more similar to that of the as-plated Al coating. The cathodic kinetics is governed by oxygen reduction, since the potential ( E corr ) at the surface of the 200 °C/12 h specimen is ennobled by several hundreds of mV from AZ91D. This therefore suggests that the Al (Mg) solid solution remaining on the surface and the γ-phase (Mg 2 Al 3 ) layer favor oxygen reduction. This is consistent with the results of Searles [41] and Birbilis [42] who have independently measured the electrochemical properties of γ-phase (Mg 2 Al 3 ), however in the context of Al alloys where it receives the notation of β phase in the said references. It was reported that the Mg 2 Al 3 is spontaneously passive in NaCl solution and that oxygen reduction remains the dominant cathodic reaction, however it was also noted that the Mg 2 Al 3 is not inherently a very corrosion resistant phase and hence would only offer limited protection in the long term if it were the continuous surface phase. This suggests that future work is needed to retain a wholly Al(Mg) solid solution phase on the AZ91D surface, and hence ensure high quality corrosion protection is imparted for the longer term. However, the results herein are self consistent with what is expected from the electrochemical behavior of intermetallic phases, and that heat treatment of Al coatings can produce a favorable coating with good mechanical behavior and not susceptible to water reduction and hence large scale attack in aqueous environments. 3.4 General discussion The ability to electrodeposit Al coatings on AZ91D is a demonstration of the ability to deposit pure metals upon other metals via the wide potential windows presented by IL based electrolytes. There are many advantages to being able to perform such electrodeposition in the field of coating Mg alloys. First of all, one key advantage is that when compared to conversion coatings, minimum surface preparation is needed as conversion coatings have many pretreatment steps for surface chemistry homogenization. Since the whole surface of the specimen to be coated is polarized during electrodeposition, surface heterogeneity, owing to different electrochemical properties of second phases which are in effect at open circuit potential, are overwhelmed by the application of current allowing a uniform coating to evolve. Such a uniform coating is seen clearly in Fig. 1 c. In addition, electrodeposition is applicable to all Mg alloys, not just certain classes, which is a limitation for many conversion or electroless coating technologies. Secondly, the benefits of electrodeposition when compared to other methods such as cold spray include: potentially reduced costs, ability to coat complex shapes, and minimization of Kirkendall effect related voids on subsequent heat treatment. This work has revealed that there is some future work required in order to maintain an Al (Mg) solid solution layer on the surface as the ultimate goal of post electrodeposition treatment. However, in spite of this, good quality coatings of decreased corrosion rate and good mechanical properties were obtained herein, which allowed the corresponding cathodic reaction to shift from water to oxygen reduction. We have shown that a continuous net of β phase on the surface is not satisfactory for corrosion protection, and that a continuous γ is better, but not as good as a continuous α would be. With the above in mind, the field of electrodeposition is set to be transformed via IL based media, since it becomes possible to potentially coat Mg alloys with most metals (not reported here, but we are in parallel working on other metal coatings), which would make it possible to coat Mg alloys with alloys that can suppress certain intermetallic formation reaction by making the deposited metal a solid solution itself. 4 Conclusions A continuous and dense Al coating with a FCC structure can be electrodeposited upon AZ91D surface via an IL electrolyte. Whilst the Al coating imparts significant improvements in corrosion resistance to the substrate, the nanomechanical properties and bonding strength of the coating are not sufficient in the as-plated condition – which would ultimately have serious limitations on practical utility. Adhesion promotion and strength increased markedly from post-coating heat-treatment. Treatment at 420 °C/12 h resulted in a β-phase enriched two-phase coating (α-Mg + β). The hardness increases nearly three times, however the corrosion reduction afforded by such a β-phase network is not significant, with the surface capable of supporting water reduction as the dominant cathodic reaction and hence dissolution still occurred at relatively high rates, albeit lower than the base AZ91D. Post-coating heat-treatment at 200 °C/2 h generated a tri-layer structure (α-Al(Mg)/β/γ). This surface showed no appreciable hydrogen evolution during immersion in 3.5 wt.% NaCl solution for 72 h, and providing high levels of protection for AZ91D substrate. It was noted however, that the outermost surface layer of α-Al (Mg) was weakly bonded to the surface (i.e. neighboring γ layer) and to remedy this, future work is needed. In the context of proving the optimum corrosion protection, the outermost surface should retain as high a concentration of Al as possible, and this could potentially be achieved by electrodeposition of alloyed Al to modify the formation of Mg-rich intermetallics (since such intermetallics are comparatively active and prone to supporting water reduction) whilst retaining a solid solution outermost layer. Acknowledgments The authors sincerely acknowledge the financial support from the National Basic Research Program of China (No. 2007CB613701), Program of Shanghai Subject Chief Scientist (No. 08XD14020), and Aerospace Science and Technology Innovation Fund of China Aerospace Science and Technology Corporation (No. 0502). References [1] R. Ambat W. Zhou Surf. Coat. Technol. 179 2004 124 134 [2] J.E. Gray B. Luan J. Alloys Comp. 336 2002 88 113 [3] M. Carboneras M.D. López P. Rodrigo M. Campo B. Torres E. Otero J. Rams Corros. Sci. 52 2010 761 768 [4] H. Ardelean I. Frateur P. Marcus Corros. Sci. 50 2008 1907 1918 [5] R. Arrabal E. Matykina F. Viejo P. Skeldon G.E. Thompson Corros. Sci. 50 2008 1744 1752 [6] T. Lei C. Ouyang W. Tang L.F. Li L.S. Zhou Corros. Sci. 52 2010 3504 3508 [7] B.L. Yu X.L. Pan J.Y. Uan Corros. Sci. 52 2010 1874 1878 [8] E. Rocca C. Juers J. Steinmetz Corros. Sci. 52 2010 2172 2178 [9] Y.W. Shao H. Huang T. Zhang G.Z. Meng F.H. Wang Corros. Sci. 51 2009 2906 2915 [10] W.J. Cheong B.L. Luan D.W. Shoesmith Corros. Sci. 49 2007 1777 1798 [11] L.C. Zhao C.X. Cui Q.Z. Wang S.J. Bu Corros. Sci. 52 2010 2228 2234 [12] H. Ardelean I. Frateur S. Zanna A. Atrens P. Marcus Corros. Sci. 51 2009 3030 3038 [13] T.F. Conceicao N. Scharnagl W. Dietzel K.U. Kainer Corros. Sci. 52 2010 3155 3157 [14] H.M. Wang R. Akid M. Gobara Corros. Sci. 52 2010 2565 2570 [15] T.F. Conceicao N. Scharnagl C. Blawert W. Dietzel K.U. Kainer Corros. Sci. 52 2010 2066 2079 [16] R.F. Zhang Corros. Sci. 52 2010 1285 1290 [17] J. Liang P.B. Srinivasan C. Blawert W. Dietzel Corros. Sci. 52 2010 540 547 [18] X.L. Liu T. Zhang Y.W. Shao G.Z. Meng F.H. Wang Corros. Sci. 52 2010 892 900 [19] Y.F. Jiang L.F. Liu C.Q. Zhai Y.P. Zhu W.J. Ding Thin Solid Films 484 2005 232 237 [20] C.A. Huang T.H. Wang T. Weirich V. Neubert Corros. Sci. 50 2008 1385 1390 [21] J.K. Chang S.Y. Chen W.T. Tsai M.J. Deng I.W. Sun Electrochem. Commun. 9 2007 1602 1607 [22] C.A. Huang C.K. Lin Y.H. Yeh Corros. Sci. 52 2010 1326 1332 [23] K. Spencer M.X. Zhang Scripta Mater. 61 2009 44 47 [24] Y.S. Tao T.Y. Xiong C. Sun L.Y. Kong X.Y. Cui T.F. Li G.L. Song Corros. Sci. 52 2010 3191 3197 [25] M. Qian D. Li S.B. Liu S.L. Gong Corros. Sci. 52 2010 3554 3560 [26] J.K. Chang S.Y. Chen W.T. Tsai M.J. Deng I.W. Sun J. Electrochem. Soc. 155 2008 C112 C116 [27] S.J. Pan W.T. Tsai J.K. Chang I.W. Sun Electrochim. Acta 55 2010 2158 2162 [28] I. Shigematsu M. Nakamura N. Siatou K. Shimojima J. Mater. Sci. Lett. 19 2000 473 475 [29] H.W. Huo Y. Li F.H. Wang J. Mater. Sci. Technol. 23 2007 379 382 [30] X.R. Li W. Liang X.G. Zhao Y. Zhang X.P. Fu F.C. Liu J. Alloys Comp. 471 2009 408 411 [31] M.X. Zhang P.M. Kelly J. Mater. Res. 17 2002 2477 2479 [32] L.Q. Zhu G.L. Song Surf. Coat. Technol. 200 2006 2834 2840 [33] H.Q. Sun Y.N. Shi M.X. Zhang K. Lu Surf. Coat. Technol. 202 2008 3947 3953 [34] M.X. Zhang Y.N. Shi H. Sun P.M. Kelly J. Nanosci. Nanotechnol. 8 2008 2724 2728 [35] W.C. Oliver G.M. Pharr J. Mater. Res. 19 2004 3 20 [36] H. Huang K. Winchester Y. Liu X.Z. Hu C.A. Musca J.M. Dell L. Faraone J. Micromech. Microeng. 15 2005 608 614 [37] K. Spencer M.X. Zhang Surf. Coat. Technol. 203 2009 3019 3025 [38] H.W. Huo Y. Li F.H. Wang Corros. Sci. 46 2004 1467 1477 [39] G.L. Song A. Atrens X. Wu B. Zhang Corros. Sci. 40 1998 1769 1791 [40] O. Lunder J.E. Lein T.Kr. Aune K. Nisancioglu Corrosion 45 1989 741 748 [41] J.L. Searles P.I. Gouma R.G. Buchheit Met. Mat. Trans. A 32A 2001 2859 2867 [42] N. Birbilis R.G. Buchheit J. Electrochem. Soc. 152 2005 B140 B151