Applications of carbonic acid solution for developing conversion coatings on Mg alloy

YU Bing-lung(余秉隆), LIN Jun-kai(林俊凱), UAN Jun-yen(汪俊延)

Department of Materials Science and Engineering, National Chung Hsing University,

250 kuo-kuang Rd., Taichung 402, Taiwan, China

Received 23 September 2009; accepted 30 January 2010

Abstract: Works on exploring an environmentally clean method for producing an Mg,Al-hydrotalcite (Mg₆Al₂(OH)₁₆CO₃·4H₂O) layer and/or calcium carbonate (CaCO₃) layer on Mg alloy in a carbonic acid solution system (aqueous HCO₃^-/CO₃^2- or Ca²⁺/HCO₃^-) at 50 ?C were reviewed. Conversion treatment for the Mg,Al-hydrotalcite conversion coating was as follows. Mg alloy was treated  first in acidic HCO₃^-/CO₃^2- aqueous for precursor layer formation on Mg alloy surface and then in alkaline HCO₃^-/CO₃^2- aqueous to form a crystallized Mg,Al-hydrotalcite coating. Duration of an Mg,Al-hydrotalcite coating on Mg alloy surface was reduced from 12 h to 4 h by the conversion treatment. On the other hand, for reducing the formation time of CaCO₃ coating on Mg alloy, the aqueous Ca²⁺/HCO₃^- with a saturated Ca²⁺ content was employed for developing a CaCO₃ coating on Mg alloy. A dense CaCO₃ coating could yield on Mg alloy surface in 2 h. Corrosion rate (corrosion current density, J_corr) of the Mg,Al-hydrotalcite-coated sample and CaCO₃-coated AZ91D sample was 7-10 μA/cm², roughly two orders less than the J_corr of the as-diecast sample (about 200 μA/cm²). No corrosion spot on the Mg,Al-hydrotalcite-coated sample and CaCO₃-coated sample was observed after 72 h and 192 h salt spray test, respectively.

Key words: Mg alloy; AZ91D alloy; corrosion; conversion coating; carbonic acid

1 Introduction

Magnesium alloys have a high specific strength with a density roughly two-third that of aluminum and one-quarter that of iron[1]. These characteristics make magnesium alloys extremely attractive in vehicle applications[2-3]. However, the alloys are susceptible to corrosion in practical environments due to their high electrochemical activity[4]. Surface treatment is a basic method to improve the corrosion resistance of magnesium alloys. Many surface treatments have been adopted to increase the corrosion resistance of magnesium alloys, such as chemical conversion coating[1], anodizing[5-6], electrochemical plating[1], electroless nickel plating[7], and coating with pure magnesium by physical vapor deposition (PVD)[8-9], plasma-assisted chemical vapor deposition (PACVD)[10] as well as selected etching surface treatment[11-12]. Chemical conversion treatment is the most common surface pretreatment with low cost for improving the corrosion resistance of Mg alloys[13]. The treatment is typically based on chromate solutions[13], although hexavalent chromium is a toxic substance that pollutes environment and is detrimental to health[1]. Many studies have investigated several chrome-free conversion coating of magnesium alloys, including phosphate[14-17], phosphate-permanganate[18-19], stannate[19-23], vanadate[24], cobalt(Ⅲ) hex coordinated complex[25], cerate[26-29] or lanthanite[27] or praseodymate[27] and others. Although the above studies contributed to the substitution of Cr⁶⁺-free conversion coating process, it may also have some potential risks to the environment, such as the environmental pollution of heavy metal ions and phosphorus. Moreover, the above mentioned chemical conversion coatings may make it difficult to recycle post-consumed Mg product scraps (such as from automotive components) into Mg ingots that fulfill ASTM specifications[1, 30-31]. One main reason is that impurities from the conversion coating contaminate the magnesium melt[30, 32].

In the previous study of the authors, an environmentally clean method for synthesizing a chemical conversion coating on Mg alloys in carbonic acid solution system (aqueous HCO₃^-/CO₃^2- or Ca²⁺/HCO₃^-)[33-34] was investigated. An Mg,Al- hydrotalcite (Mg₆Al₂(OH)₁₆CO₃·4H₂O) layer and aragonitic CaCO₃/Mg,Al-hydrotalcite two-layer coating was developed on AZ91D Mg alloy in an aqueous HCO₃^-/CO₃^2- solution and aqueous Ca²⁺/HCO₃^- solution at 50 ?C, respectively[33-34]. The results showed that the coatings could protect the metal substrate against corrosion. However, durations as long as at least 12 h were required to develop an aragonitic CaCO₃/Mg,Al-hydrotalcite or Mg,Al-hydrotalcite coating that enables to protect the Mg alloy from corrosion[33-34]. Therefore, how to shorten the treatment time to form an Mg, Al-hydrotalcite coating and/or calcium carbonate coating on AZ91D Mg alloy was also reported in our recent work. The microstructure, crystal structure and corrosion resistance of the coatings were reported.

2 Experimental

2.1 Materials

The AZ91D die-cast magnesium alloy adopted herein was the same as that used in the earlier works[8-9, 11-12, 34-35]. It has the composition (mass fraction) of 8.8% Al, 0.69% Zn, 0.212% Mn, 0.02% Si, 0.002% Cu, 0.005% Fe, 0.001% Ni and balance Mg. The original die cast plate had an area of 300 mm×240 mm and a thickness of 1.4 mm. Square coupon specimens were cut from the plate with size of 1.4 mm×20 mm×20 mm.

2.2 Preparation of Mg,Al-hydrotalcite-coated sample

AZ91D sample was ground using SiC paper (1 500 grit) and then cleaned in ethyl alcohol in an ultrasonic cleaner. HCO₃^-/CO₃^2- aqueous solution was prepared at room temperature by bubbling CO₂ gas through 1 000 mL of deionized water. The CO₂ gas that did not immediately dissolve in water was recycled. The recycled CO₂ gas was then recharged into the water to generate the HCO₃^-/CO₃^2- solution. The flow rate of CO₂ gas was 1 dm³/min. Consequently, 20 min sufficed to minimize the pH of the solution (about 4.3). The CO₂ gas removed from industrial emissions is also believed to be able to fulfill the purpose of this work, as it is used to produce aqueous HCO₃^-/CO₃^2-. The carbonic acid solution was heated to 50 ?C in a water bath. Then, six square coupons were immersed in the solution at 50 ?C for a particular period. The immersion time changed from 1 h to 24 h. The aforementioned treatment in which samples were statically immersed in aqueous HCO₃^-/CO₃^2- to form a conversion layer on their surface was denoted as CO₂-A treatment, hereafter. For instant, CO₂-A-1h means that the treatment was performed for 1 h. Different treatment times are denoted similarly. AZ91D samples were immersed in HCO₃^-/CO₃^2- solution at 50 ?C for 1 h (CO₂-A-1h), 2 h (CO₂-A- 2h)…, 24 h (CO₂-A-24h), In another experiment, six square coupons were immersed in carbonic acid solution at 50 ?C for 2 h while CO₂ gas was continually bubbled through the solution. The pH of the solution was kept in the range between 4 and 6. This treatment is called CO₂-B treatment. The conditions of CO₂-A and CO₂-B treatment are listed in Table 1. The pH of an HCO₃^-/CO₃^2- solution was increased and kept at pH 11.5 by dropwise addition of 1.25 mol/L aqueous NaOH with vigorous stirring during the mixing. The samples after CO₂-B treatment were further dipped in the pH 11.5 HCO₃^-/CO₃^2- solution. This treatment was denoted as, for example, CO₂-B-2h/pH11.5-2h. The notation indicates that the AZ91D sample underwent CO₂-B-2h treatment first, and then was dipped into HCO₃^-/CO₃^2- solution at pH 11.5 at 50 ?C for 2 h.

Table 1 Conditions of CO₂-A[33] and CO₂-B treatment[36]

2.3 Preparation of CaCO₃-coated sample

AZ91D samples were degreased, and then were cleaned ultrasonically in distilled water. To prepare the Ca²⁺/HCO₃^- solution, 0.5 g and 0.7 g CaCO₃ powder was respectively mixed with 1 000 mL distilled water at room temperature. Experiments involved CO₂ gas bubbling through the CaCO₃/water slurry to dissolve the CaCO₃ compound. The flow rate of the CO₂ gas was the same as that used in the preparation of HCO₃^-/CO₃^2- aqueous solution. The CO₂ gas bubbled through the CaCO₃ water/slurry until the CaCO₃ had dissolved in water (typically taking 45 min) to yield 1 000 mL Ca²⁺/HCO₃^- solution. The CO₂ gas not immediately dissolved in water was recycled and then recharged into the water. The solution was filtered through a filter paper before it was used for the conversion hard coating experiment. The content (mass fraction) of Ca²⁺ in the aqueous Ca²⁺/HCO₃^- solution (0.5 g CaCO₃) was about 1.2×10^-4 g/mL while that in the aqueous Ca²⁺/HCO₃^- solution (0.7 g CaCO₃) was about 2.2×10^-4 g/mL (determined by an ion-specific meter model HI 93752, HANNA instruments). Five square coupon specimens with the surfaces facing upward were statically immersed in the Ca²⁺/HCO₃^- solution with Ca²⁺ content of 1.2×10^-4 and 2.2×10^-4 g/mL, respectively. The Ca²⁺/HCO₃^- solution was heated to 50 ?C in a water bath. The specimen immersion in the Ca²⁺/HCO₃^- solution with Ca²⁺ content of 1.2×10^-4 g/mL at 50 ?C was denoted as CaCO₃-treatment A. The immersion time changed from 1 h to 12 h. CaCO₃-A-1h means that the treatment was performed for 1 h. Similarly, AZ91D samples were immersed in Ca²⁺/HCO₃^- solution at 50 ?C for 1 h (CaCO₃-A-1h), 2 h (CaCO₃-A-2h), …, 12 h (CaCO₃-A-12h). In another experiment, five square coupon specimens with the surfaces facing upward were statically immersed in the Ca²⁺/HCO₃^- solution with Ca²⁺ content of 2.2×10^-4 at 50 ?C for 2 h. This treatment is called CaCO₃-B treatment. The conditions of CaCO₃-A and CaCO₃-B treatment are listed in Table 2.

Table 2  Experimental conditions for CaCO₃-A[34] and CaCO₃-B treatment

2.4 Microstructure observation

Backscattered electron imaging (BEI) system in a JEOL JSM-6700F field emission scanning electron microscope (FE-SEM) was adopted to study the microstructure. The crystallographic structure of the specimens was analyzed by glancing angle X-ray diffraction (GAXRD) using Cu K_α1 (1.540 5 ?) radiation.

2.5 Corrosion test

Electrochemical polarization tests and salt spray tests were employed to determine the corrosion resistance of samples. Electrochemical polarization tests were performed in a corrosion cell that contained 270 mL of 3.5% (mass fraction) NaCl solutions at room temperature at a scan rate of 0.5 mV/s. All electrochemical measurements were made using a Princeton Applied Research model 263A Potentiostat/ Galvanostat and M352 software. The area of the coating exposed to the NaCl solution was 1 cm². Platinum gauze was used as a counter electrode and silver/silver chloride (Ag/AgCl) electrode was used as the reference. At least four experiments were performed for each experimental case. Samples were subjected to a salt spray test (ASTM B117 standard[35]). Salt spray chamber was maintained at 35 ?C and the spray jet atomized continuously to convert salt solution into uniform small droplets.

3 Results and discussion

3.1 Reducing formation time of Mg,Al-hydrotalcite coating layer on AZ91D

In our previous study[33], CO₂-A treatment was employed to form Mg,Al-hydrotalcite layer on AZ91D sample surface. The GAXRD patterns of the samples CO₂-A-1h, CO₂-A-4h, CO₂-A-6h, CO₂-A-12h and CO₂-A-24h are presented in Fig.1. As shown in Fig.1, after treatment time of 1 h, 4 h and 6 h, weak X-ray peaks of Mg,Al-hydrotalcite were observed. Prolonging the treatment caused the GAXRD patterns to yield intense peaks of Mg,Al-hydrotalcite (JCPDS X-ray diffraction file No.22-0700). Fig.2 displays the backscattered electron images of cross-sectional microstructures of the CO₂-A sample. Fig.2(a) shows the cross-sectional microstructure of the CO₂-A-1h sample. When the immersion time was increased to 6 h (Fig.2(b)), a uniform precursor layer could be observed. Fig.2(c) displays the cross-sectional microstructure of the CO₂-A-24h sample. As indicated in Fig.2(c), the thickness of the Mg,Al-hydrotalcite layer was 5-8 μm. Fig.2(d) displays the plain-view microstructure of the CO₂-A-24h sample. Several network-like cracks at the coating layer were observed. Small cracks were distributed on the conversion coating layer during dehydration, which improved the adhesion of subsequent paint layers or organic coatings to the surface of the magnesium alloy substrate[13].

Our recent study[36] demonstrated that the formation of an Mg,Al-hydrotalcite structure on the AZ91D sample in HCO₃^-/CO₃^2- solution at 50 ?C was strongly related to the pH of solution. A precursor layer of Mg,Al-hydrotalcite first covered on sample surface in acid HCO₃^-/CO₃^2- solution. As the CO₂-A treatment time was increased to at least 12 h, the solution turned from acidic to alkaline and the precursor layer transformed into the layer of crystalline Mg,Al- hydrotalcite[36]. The above results were exploited to shorten the time required to prepare an crystalline Mg,Al-hydrotalcite layer on AZ91D. AZ91D sample was first immersed in acidic HCO₃^-/CO₃^2- solution (pH 4-6) for precursor layer formation (CO₂-B) and then was

Fig.1 GAXRD patterns for sample with different CO₂-A treatment time[33]

Fig.2 Cross-sectional microstructures of sample after CO₂-A treatment for different time: (a) 1 h, (b) 6 h, (c) 24 h[33]; (d) Surface morphology of sample after CO₂-A treatment for     2 h[36]

immersed in alkaline HCO₃^-/CO₃^2- to form crystallized Mg,Al-hydrotalcite coating. Fig.3(a) shows the GAXRD patterns of the sample after CO₂-B-2h treatment. As shown in Fig.3(a), weak and broad peaks of Mg,Al- hydrotalcite from CO₂-B-2h sample were detected, suggesting the formation of the precursor layer of Mg,Al-hydrotalcite. The GAXRD pattern of the CO₂-B-2h sample (Fig.3(a)) was similar with that of CO₂-A-1h, CO₂-A-4h and CO₂-A-6h sample (see the patterns in Fig.1). Fig.3(b) shows the cross-sectional microstructure of the CO₂-B-2h sample. As presented in Fig.3(b), a precursor layer exists on the surface of the CO₂-B-2h sample. The thickness of the Mg,Al- hydrotalcite layer was similar with that of the CO₂-A-24h sample. The samples after CO₂-B-2h treatment were subsequently immersed in an alkaline environment (pH 11.5). Fig.4 shows that the GAXRD patterns of the sample CO₂-B-2h/pH11.5-1h. The X-ray peaks of Mg,Al-hydrotalcite on the sample CO₂-B-2h/pH11.5-1h were strong (as indicated in Fig.4). Thus, the presented treatment method could be utilized to shorten the treatment time from at least 12 h to 3 h to form an crystalline Mg,Al-hydrotalcite layer on AZ91D magnesium alloy. Moreover, to form a compact Mg,Al-hydrotalcite layer, the CO₂-B-2h sample was immersed in pH 11.5 carbonic acid solution for 2 h. Fig.5(a) presents the surface microstructure of the CO₂-B-2h/pH11.5-2h sample. As indicated in Fig.5(a), this sample had crystalline Mg, Al-hydrotalcite coating layer, which, however, still exhibited several network-

Fig.3 GAXRD pattern (a) and cross-sectional microstructure (b) of CO₂-2h sample[36]

Fig.4 GAXRD pattern of CO₂-B-2h/pH11.5-1h sample[36]

Fig.5 FE-SEM BEI images of CO₂-B-2h/pH11.5-2h sample: (a) Surface observation with arrows denoting coating materials in crevice; (b) Cross-sectional microstructure; (c) Coating material in crevice (as denoted by arrow)[36]

like cracks. The arrows in Fig.5(a) denoted that there were coating materials within the cracks. The cross-sectional microstructure of the CO₂-B-2h/ pH11.5-2h was shown in Figs.5(b) and (c). The thickness of the Mg,Al-hydrotalcite layer was 5-8 μm (see Fig.5(c)). As shown in Fig.5(c), there was coating material in the crevice, avoiding the exposure of substrate metal to the environment. Hence, although the network-like cracks were observed on the coated sample surface, the crack did not penetrate the layer directly to the substrate metal.

3.2 Corrosion properties of Mg,Al-hydrotalcite conversion coating on AZ91D

The polarization curves of the samples are plotted in Fig.6. The as-cast AZ91D sample and various CO₂-2h/ pH11.5-treated samples were measured in 3.5% NaCl

Fig.6 Polarization curves of as-cast AZ91D sample and CO₂-B-2h samples being treated in carbonic acid solution of pH 11.5 for different periods[36]

solution. The corrosion potential (φ_corr) of the CO₂-2h/ pH11.5-2h sample was about -1.39 V (vs Ag/AgCl), while that of the as-cast AZ91D sample was around -1.45 V (vs Ag/AgCl). The AZ91D substrate had corrosion current density (J_corr) of about 250 μA/cm² and the CO₂-2h sample had J_corr of about 100 μA/cm². As shown in Fig.6, the J_corr of the CO₂-2h/pH11.5 samples was lower than that of the CO₂-2h sample. J_corr would decrease as the immersion time in pH 11.5 HCO₃^-/CO₃^2- solution at 50 ?C increased. The J_corr of CO₂-2h/pH 11.5-2h sample could be down to about 10 μA/cm².

Therefore, the CO₂-2h/pH11.5-2h sample exhibited greater corrosion resistance than the as-cast AZ91D sample. Figs.7(a) and (b) display the surface morphologies of the samples after the salt spray test. A total 25 pieces of CO₂-2h/pH11.5-2h samples were placed in the salt spray chamber. Only two samples had small corrosion spots after 72 h of the salt spray test.  Nevertheless, the surface area fraction of these corrosion spots on each of the two samples was less than 4%. Fig.7(a) shows an example of the CO₂-2h/pH11.5-2h sample surface after 72 h of the salt spray test. For comparison, as shown in Fig.7(b), the as-cast AZ91D sample was severely corroded after 12 h salt spray test.

3.3 Reducing formation time of CaCO₃ coating on AZ91D

Fig.8 shows the GAXRD patterns of sample after CaCO₃-A treatment for 10 min, 30 min, 2 h, 8 h and  12 h. The content (mass fraction) of Ca²⁺ in the aqueous Ca²⁺/HCO₃^- solution was about 1.2×10^-4 g/mL. The GAXRD patterns of the CaCO₃-A-10 min sample were composed mainly of Mg. For the sample immersed for 30 min in the Ca²⁺/HCO₃^- solution (see Fig.8), the X-ray diffraction pattern had intensity peaks for aragonitic calcium carbonate (JCPDS cards No. 01-0628). For the

Fig.7 Surface morphologies after salt spray tests: (a) CO₂-B-  2h/pH11.5-2h sample after salt spray test for 72 h; (b) As-cast sample after salt spray test for 12 h[36]

Fig.8 GAXRD patterns for sample with different CaCO₃-A treatment time[34]

sample immersed in the solution for 2 h, the peak of aragonite structure at a 2θ angle of 29.1? had a strong preferred orientation. The intensities of peaks at a 2θ angle of 29.1? increased as immersion duration increased (see Fig.8). Fig.9 presents the microstructures of the sample after CaCO₃-A treatment for 2 h and 12 h. Fig.9(a) displays the surface microstructure of the CaCO₃-A-2h sample. As displayed in Fig.9(a), some of the sample surface was not covered by aragonitic CaCO₃

Fig.9 Microstructures of sample after CaCO₃-A treatment for 2 h and 12 h: (a) Surface morphology of CaCO₃-A-2h sample; (b) Surface morphology of CaCO₃-A-12h sample; (c) Cross-sectional microstructure of CaCO₃-A-2h sample; (d) Cross-sectional microstructure of CaCO₃-A-12h sample[34]
crystals. Fig.9(b) shows the surface microstructure of the CaCO₃-A-12h sample, indicating that the aragonitic CaCO₃ crystals covered the CaCO₃-A-12h sample surface. Figs.9(c) and (d) present the backscattered electron images of the cross-sectional microstructure of the CaCO₃-A-2h and CaCO₃-A-12h samples, respectively. As shown in Figs.9(c) and (d), the aragonite layer on CaCO₃-A-12h sample surface was much thicker than that on CaCO₃-A-2h sample surface. The thickness of aragonite layer on CaCO₃-A-12h sample was (3.8±0.5) μm. An interlayer was observed between aragonitic CaCO₃ layer and the AZ91D substrate (see Figs.9(c) and (d)). The interlayer was composed of Mg,Al-hydrotalcite structure[33]. The Mg,Al- hydrotalcite was corrosion product due to the corrosion of AZ91D substrate surface in the Ca²⁺/HCO₃^- solution at 50 ?C. Fig.10 presents the GAXRD patterns of the as-cast AZ91D sample and the sample after CaCO₃-B treatment for 2 h. The aqueous Ca²⁺/HCO₃^- solution contained Ca²⁺ content up to about 2.2×10^-4 g/mL. The diffraction patterns of the CaCO₃-B-2h sample showed the peaks of CaCO₃ (JCPDS X-ray diffraction file No. 01-0837).

Fig.10 GAXRD patterns for as-cast AZ91D and sample after CaCO₃-B treatment for 2 h

Fig.11 presents the microstructure of the sample after CaCO₃-B treatment for 2 h. As shown in Fig.11(a), rhombohedra-shaped calcite crystals were covered on the sample surface. Fig.11(b) presents the backscattered electron images of the cross-sectional microstructure of the CaCO₃-B-2h sample. An Mg,Al-hydrotalcite layer was also observed between calcite CaCO₃ coating and the AZ91D substrate (see Fig.11(b)). By comparing the surface microstructures between the samples after CaCO₃-A treatment and CaCO₃-B treatment, it was found that the content of Ca²⁺ can remarkably affect the polymorph form of CaCO₃. Moreover, only 2 h for CaCO₃-B treatment was needed to have a CaCO₃ film covering on sample surface. For comparison, much more

Fig.11 SEM surface morphology (a) and cross-sectional microstructure (b) of CaCO₃-B-2 h sample

than 2 h was needed for CaCO₃-A treatment to have a continuous CaCO₃ coating on the Mg sample. The mechanism for the formation of aragonitic CaCO₃ coating or calcite CaCO₃ coating was proposed in our previous work[34]. However, the reason that Ca²⁺ content effectively changes the polymorph of CaCO₃ on Mg alloy remains unclear.

3.4 Corrosion properties of CaCO₃ coating on AZ91D

The electrochemical test results of samples are compared in Fig.12. Corrosion potential (φ_corr) of the AZ91D substrate was about -1.45 V (vs Ag/AgCl). The CaCO₃-A-12h and CaCO₃-B-2h sample remained at the same level as the AZ91D substrate. The AZ91D substrate had corrosion current density (J_corr) of about

Fig.12 Polarization curves for as-cast AZ91D, CaCO₃-A-12 h and CaCO₃-B-2 h samples

250 μA/cm² while the J_corr of the CaCO₃-B-2 h sample was reduced to about 7 μA/cm². The J_corr of the CaCO₃-A-12h sample was higher than that of the CaCO₃-B-2h sample, suggesting that the CaCO₃ coating on CaCO₃-B-2h sample was denser than CaCO₃ coating on CaCO₃-A-12h sample surface. Fig.13 displays the surface morphologies for the samples after the salt spray test. Fig.13(a) shows that no corrosion spot was observed on CaCO₃-A-12h sample after 43 h salt spray test. As indicated in Fig.13(b), corrosion spot was absent on the CaCO₃-B-2h sample surface after 192 h of the salt spray test. Therefore, calcite CaCO₃ coating (CaCO₃-B-2h) protected the Mg alloy from corrosion in a relatively longer time than the aragonitic CaCO₃ coating did after CaCO₃-A treatment.

Fig.13 Surface morphologies on sample surface after salt spray test: (a) CaCO₃-A-12 h sample after 48 h test; (b) CaCO₃-B-2h sample after 192 h test

4 Conclusions

1) An environmentally clean method was explored to develop an Mg,Al-hydrotalcite (Mg₆Al₂(OH)₁₆- CO₃·4H₂O) layer and/or calcium carbonate (CaCO₃) layer on Mg alloy for improving corrosion resistance of the alloy in NaCl environment.

2) To reduce the process time of developing Mg,Al-hydrotalcite conversion coating, an AZ91D sample must first be treated in an acidic HCO₃^-/CO₃^2- bath for precursor layer formation, and then in an alkaline HCO₃^-/CO₃^2- bath to form crystallized Mg,Al- hydrotalcite coating. The treatment time could be reduced to 4 h. The AZ91D sample with the crystalline Mg,Al-hydrotalcite conversion coating had nobler φ_corr (-1.39 V (vs Ag/AgCl)) than that of substrate AZ91D (-1.45 V (vs Ag/AgCl)). The J_corr of the AZ91D sample with the crystalline Mg,Al-hydrotalcite conversion coating (about 10 μA/cm²) was evidently lower than that of the substrate metal (about 250 μA/cm²). No corrosion spot on the crystalline Mg,Al-hydrotalcite-coated sample was observed after a 72 h salt spray test.

3) The aqueous Ca²⁺/HCO₃^2- with a saturated Ca²⁺ content was employed for rapidly developing a CaCO₃ coating on Mg alloy. A calcite CaCO₃/ Mg,Al-hydrotalcite coating could yield on Mg alloy surface in 2 h. The J_corr of the sample with CaCO₃/ Mg,Al-hydrotalcite coating could be down to about 7 μA/cm². Corrosion spot on the calcite CaCO₃/ Mg,Al-hydrotalcite coating was absent after a 192 h salt spray test.

4) The content of Ca²⁺ can affect the polymorph form of CaCO₃. The aragonitic CaCO₃ layer formed on sample surface in aqueous Ca²⁺/HCO₃^- solution with Ca²⁺ content of about 1.2×10^-4 g/mL while the calcite CaCO₃ layer formed on sample surface in aqueous Ca²⁺/HCO₃^- solution when the aqueous Ca²⁺/HCO₃^- solution with Ca²⁺ content up to about 2.2×10^-4 g/mL.

References

[1] GRAY J E, LUAN B. Protective coatings on magnesium and its alloys—A critical review [J]. J Alloy Compd, 2002, 336(1/2): 88-113.

[2] BALLERINI G, BARDI U, LAVACCHI A, MIGLIORINI D, DANIELE B, BIGNUCOLO R, CERAOLO G. Magnesium alloys for structural automotive applications [J]. Magnesium Industry, 2002, 3: 16-20.

[3] SONG G, ATRENS A, JOHN D S, WU X, NAIRN J. The anodic dissolution of magnesium in chloride and sulphate solution [J]. Corro Sci, 1997, 39 (10/11): 1981-2004.

[4] JONES D A. Principles and prevention of corrosion [M]. Second Edition. Prentice Hall, Englewood Cliffs, NJ, 1996: 40-74.

[5] SHI Z, SONG G, ATRENS A. Influence of the β phase on the corrosion performance of anodized coating on magnesium- aluminium alloys [J]. Corro Sci, 2005, 47(11): 2760-2777.

[6] SHI Z, SONG G, ATRENS A. Influence of anodising current on the corrosion resistance of anodised AZ91D magnesium alloy [J]. Corro Sci, 2006, 48(8): 1939-1959.

[7] CHEONG W J, LUAN B L, SHOESMITH D W. Protective coating on Mg AZ91D alloy—The effect of electroless nickel (EN) bath stabilizers on corrosion behavior of Ni-P deposit [J]. Corro Sci, 2007, 49(4): 1777-1798.

[8] YU B L, UAN J Y. Sacrificial Mg film anode for cathodic protection of die cast Mg-9%Al-1%Zn alloy in NaCl aqueous solution [J]. Scr Mater, 2006, 54(7): 1253-1257.

[9] YU B L, UAN J Y. Sacrificial magnesium film anode for cathodic protection of die casting AZ91D alloy[C]//Magnesium Technology 2006. San Antonio, 2006: 299-304.

[10] RIE K T, WOHLE J. Plasma-CVD of TiCN and ZrCN films on light metals [J]. Surf Coat Technol, 1999, 112(1/3): 226-229.

[11] LI C F, UAN J Y. Selected etching surface treatment for improving the corrosion resistance of die cast AZ91D thin plate[C]// Magnesium Technology 2006. San Antonio, 2006: 305-310.

[12] UAN J Y, LI C F, YU B L. Characterization and improvement in the corrosion performance of a hot-chamber diecast Mg alloy thin plate by the removal of interdendritic phases at the die chill layer [J]. Metall Mater Trans A, 2008, 39(3): 703-715.

[13] AVEDESIAN M M, BAKER H. Magnesium and magnesium alloys [M]. ASM Specialty Handbook, 1999: 1-71, 143.

[14] LIN C S, LEE C Y, LI W C, CHEN Y S, FANG G N. Formation of phosphate/permanganate conversion coating on AZ31 magnesium alloy [J]. J Electrochem Soc, 2006, 153(3): B90-B96.

[15] KOUISNI L, AZZI M, ZERTOUBI M, DALARD F, MAXIMOVITCH S. Phosphate coatings on magnesium alloy AM60. Part 1: Study of the formation and the growth of zinc phosphate films [J]. Surf Coat Technol, 2004, 185(1): 58-67.

[16] ZHOU W, SHAN D, HAN E H, KE W. Structure and formation mechanism of phosphate conversion coating on die-cast AZ91D. magnesium alloy [J]. Corro Sci, 2008, 50(2): 329-337.

[17] SONG Y, SHAN D, CHEN R, ZHANG F, HAN E H. Formation mechanism of phosphate conversion film on Mg-8.8Li alloy [J]. Corro Sci, 2009, 51(5): 62-69.

[18] HAWKE D, ALBRIGHT D L. A phosphate-permanganate conversion coating for magnesium [J]. Metal Finishing, 1995, 93(10): 34-38.

[19] ZUCCHI F, FRIGNANI A, GRASSI V, TRABANELLI G, MONTICELLI C. Stannate and permanganate conversion coatings on AZ31 magnesium alloy [J]. Corro Sci, 2007, 49(12): 4542-4552.

[20] HUO H, LI Y, WANG F. Corrosion of AZ91D magnesium alloy with a chemical conversion coating and electroless nickel layer [J]. Corro Sci, 2004, 46(6): 1467-1477.

[21] GONZALEZ-NUNEZ M A, NUNEZ-LOPEZ C A, SKELDON P, THOMPSON G E, KARIMZADEH H, P LYON, WILKS T E. A non-chromate conversion coating for magnesium alloys and magnesium-based metal matrix composite [J]. Corro Sci, 1995, 37(11): 1763-1772.

[22] ANICAI L, MASI R, SANTAMARIA M, QUARTO F D. A photoelectrochemical investigation of conversion coatings on Mg substrates [J]. Corro Sci, 2005, 47(12): 2883-2900.

[23] LIN C S, LIN H C, LIN K M, LAI W C. Formation and properties of stannate conversion coatings on AZ61 magnesium alloys [J]. Corro Sci, 2006, 48(1): 93-109.

[24] YANG K H, GER M D, HWU W H, SUNG Y, LIU Y C. Study of vanadium-based chemical conversion coating on the corrosion resistance of magnesium alloy [J]. Mater Chem Phys, 2007, 101(2/3): 480-485.

[25] KIM S J, ZHOU Y, ICHINO R, RYOICHI O, MASAZUMI T S. Characterization of the chemical conversion films that form on Mg-Al alloy in colloidal silica solution [J]. Met Mater Int, 2003, 9(2): 207-213.

[26] ARDELEAN H, FRATEUR I, MARCUS P. Corrosion protection of magnesium alloys by cerium, zirconium and niobium-bath conversion coatings [J]. Corro Sci, 2008, 50(7): 1907-1918.

[27] RUDD A L, BRESLIN C B, MANSFELD F. The corrosion protection afforded by rare earth conversion coatings applied to magnesium [J]. Corro Sci, 2000, 42(2): 275-288.

[28] BRUNELLI K, DABALA M, CALLIARI I, MAGRINI M. Effect of HCl pre-treatment on corrosion resistance of cerium-based conversion coatings on magnesium and magnesium alloys [J]. Corro Sci, 2005, 47(4): 989-1000.

[29] ZHONG X, LI Q, HU J, LU Y. Characterization and corrosion studies of ceria thin film based on fluorinated AZ91D magnesium alloy [J]. Corro Sci, 2008, 50(8): 2304-2309.

[30] HANKO G, ANTREKOWITSCH H, EBNER P. Recycling automotive magnesium scrap [J]. JOM, 2002, 54(2): 51-54.

[31] JAVAID A, ESSADIQI E, BELL S, DAVIS B. Literature review on magnesium recycling[C]//Magnesium Technology 2006. San Antonio, 2006: 7-12.

[32] SKAR J I, SIVERTSEN L K, OSTER J M[C]// ICEPAM-international Conference on Environmental Friendly Pre-treatment for Aluminum and Other Metals. Oslo, 2004: 1-4.

[33] LIN J K, HSIA C L, UAN J Y. Characterization of Mg,Al-hydrotalcite conversion film on Mg alloy and Cl^- and CO₃^2- anion-exchangeability of the film in a corrosive environment [J]. Scr Mater, 2007, 56(11): 927-930.

[34] UAN J Y, YU B L, PAN X L. Morphological and microstructural characterization of the aragonitic CaCO₃/Mg,Al-hydrotalcite coating on Mg-9 wt pct Al-1 wt pctZn alloyto protect against corrosion [J]. Metall Mater Trans A, 2008, 39(13): 3233-3245.

[35] ASTM Method B117. Standard test method of salt spray testing [S]. ASTM, Philadelphia, PA, 1990.

[36] LIN J K, UAN J Y. Formation of Mg,Al-hydrotalcite conversion coating on Mg alloy in aqueous HCO₃^-/CO₃^2- and corresponding protection against corrosion by the coating [J]. Corros Sci, 2009 51(5): 1181-1188.

(Edited by YANG Bing)

Foundation item: Project supported by the Ministry of Education Under the ATU Plan

Corresponding author: UAN Jun-yen; Tel: +886-4-22854913; +886-4-22857017; E-mail: jyuan@dragon.nchu.edu.tw

DOI: 10.1016/S1003-6326(09)60300-9