, , ( , , )

1 Introduction

Because of the rapid activation, high cell voltage, wide voltage range, high power density capability, relatively low density, low electrode potential and long unactivated storage life[1-4], magnesium alloys have been developed as anode materials used in seawater battery system and cathodic protection, such as sonobuoys, beacons, emergency equipment, balloon batteries and life jackets[5-8]. However, a critical limitation for the service of magnesium anode is their susceptibility to corrosion[9-11].

The corrosion of magnesium anode is mainly controlled by the composition of the α-Mg matrix, the volume fraction and the electrochemical properties of the second phase[12]. AP65 is one of these magnesium anodes with the nominal composition of 6% Al, 5% Pb (mass fraction) and balance Mg. It is reported that aluminum added in the magnesium matrix can enhance the corrosion resistance of the magnesium alloy in NaCl aqueous solution[12-15]. It has also been demonstrated that adding lead in the matrix improves the corrosion resistance of magnesium[16]. Udhayan et al[17]studied the corrosion behavior of AP65 in various concentrations of magnesium perchlorate solutions and found that its electrode/electrolyte interfacial process was determined by an activation-controlled reaction. Zinc is another alloying element added into the magnesium matrix to improve the electrochemical performance of the magnesium anodes. Cao et al[4]reported the electrochemical oxidation behavior of Mg-Li-Al-Ce-Zn and Mg-Li-Al-Ce-Zn-Mn anodes in NaCl aqueous solution and found that zinc and manganese could improve the electrochemical activity of the magnesium anode. Badawy et al[18]studied the electrochemical behavior of Mg-Al-Zn and Mg-Al-Zn-Mn alloys in different electrolytes and found that zinc decreased the corrosion rates of the alloys and made the open circuit potentials of the alloys more negative. But so far, there are few reports about the corrosion behavior of AP65 with zinc in the magnesium matrix in NaCl aqueous solution and its corrosion mechanism is not clearly understood. The aim of this work is to study the corrosion behaviors of AP65 with and without zinc in the magnesium matrix immersed in NaCl aqueous solution, and to summarize the corrosion mechanism of AP65 under different heat treatment conditions.

Article ID：1673-2812(2018)03-0408-11

Receiveddate：2016-06-01；Modifieddate：2016-12-26

Foundationitem：Funded by Natural Science Foundation of Hunan province (2016JJ2147) and Science and Technology Plan Projects of Hunan Province (2015JC3004)

Biography：YIN Liyong, Senior Engineer, E-mail: fengyanmse@csu.edu.cn.

2 Experimental Procedures

Mg-Al-Pb (denoted as A) and Mg-Al-Pb-Zn (denoted as B) alloys with nominal compositions of 6% Al, 5%Pb, 1% Zn (mass fraction) and balance Mg were prepared from ingots of pure magnesium (99.99%), pure aluminum (99.99%), pure lead (99.99%) and pure zinc (99.99%) by induction melting at 750℃ with argon protection. The chemical compositions determined by the atomic absorption spectrometry are listed in Table 1. Solution treatments (T4) were carried out at 400℃ for 24 h in argon atmosphere followed by water quench. The fraction of argon was 0.1 m3/h. The as-cast and T4 treated specimens were successively ground to 1200 grit by SiC paper, polished using 0.5 μm diamond paste, ultrasonically cleaned in acetone, and finally dried in cold air. The metallographic specimen was etched with 1% natal solution (HNO3in ethanol). The microstructure was examined by optical microscopy. The phase structure was measured by D/Max 2550 X-ray diffraction (XRD) with Cu Kαradiation. The scan rang of 2θ was from 10° to 80° at a scan rate of 2 (°)/min.

Table 1 Chemical compositions of Mg-Al-Pb (A) and Mg-Al-Pb-Zn (B) alloys /mass fraction %

Potentiodynamic polarization curves were measured by IM6ex potentiostat with a standard three-electrode glass cell. Each measurement started immediately after the specimen was immersed in 3.5% (mass fraction) NaCl aqueous solution at 25℃. The electrolyte solution was made with analytical grade reagents and deionized water. The potentiodynamic polarization curves were started by stepping the potential several hundred mV negative to the open circuit potential and then polarizing in an anodic direction at a scan rate of 1 mV/s. Each specimen was encapsulated in epoxy resin with an open surface of 10×10 mm exposed to NaCl aqueous solution. The specimen surface was ground successively to 1200 grit SiC paper. A platinum gauze was used as the counter electrode and a saturated calomel electrode (SCE) as the reference electrode. All potentials were referred to the SCE. Polarization curves were used to measure the corrosion current density,icorr, by Tafel extrapolation of both anodic and cathodic branch of the polarization curve. The corrosion current density (icorr(mA/cm2)) is related to the average penetration rate (Pi, mm/y) using [9,12]:

Pi=22.85icorr

(1)

The corrosion rate was also determined by the evolved hydrogen during immersion test. Each specimen was encapsulated in epoxy resin with a surface of 10×10 mm exposed to 3.5% NaCl solution at 25°C for 72 h. Before immersion test, the specimen was ground successively to 1200 grit SiC paper. The evolved hydrogen of the specimen in the course of immersion was collected by a funnel just above the specimen, and then went into a burette and gradually displaced the test solution in the burette. In this way, the kinetics of the evolved hydrogen can be easily determined by reading the height of the test solution level in the burette. The hydrogen evolution rate was evaluated as:

(2)

Where:v(H2)-hydrogen evolution rate, ml/(cm2·h);V(VH2)-volume of evolved hydrogen, ml; surface area of specimen, cm2; immersion time, h.

For each specimen, the hydrogen evolution rate was converted to an average penetration rate (PH2, mm/y) using[9,12]:

PH2=54.696v(H2)

(3)

In order to observe the corroded morphology and measure the surface roughness of the corroded area on the specimen surface, each specimen was encapsulated in epoxy resin with a surface of 10mm×10 mm exposed to 300 ml 3.5% NaCl solution at 25℃ for various periods. Before each test, the specimen was ground successively to 1200 grit with SiC paper. In order to remove the corrosion products with minimal dissolution of base alloy, chemical cleaning of corroded specimen was carried out in 200g/L CrO3+10g/L AgNO3solution. The corroded surface morphologies without corroded products were observed by Phillips XL30 scanning electron microscopy (SEM) using secondary electron (SE) imaging. The chemical compositions of the second phases and the magnesium matrix in the as-cast specimens were obtained by emission spectrum analysis (ESA). The surface roughness of the corroded area on the specimen surfaces immersed for various periods were measured by Wyko NT9100 surface profiler.

3 Results

3.1 Microstructure

Fig.1 shows the optical micrographs of A alloys under different heat treatment conditions. For as-cast specimens, the microstructures consist of primary α-Mg grains plus a second phase distributed non-continuously along and/or adjacent to the boundaries of the α-Mg grains. According to the XRD pattern shown in Fig.2, the second phases in as-cast specimens of both A and B alloys are mainly Mg17Al12(the β-phase). After T4 treatment, the β-phase dissolves in α-Mg phase which can be visualized in Fig.1(c). Zinc does not form any compound with other elements owing to its high solubility in magnesium.

3.2 Potentiodynamic polarization curves

Fig.3 shows the potentiodynamic polarization curves of A and B alloys in different heat treatment conditions immersed in 3.5% NaCl aqueous solution at 25℃. Each measurement started immediately after the specimen was immersed in NaCl aqueous solution. The polarization curves are not symmetrical in the anodic and cathodic branches. The current density increases in the anodic polarization branch is much greater than that in the cathodic branch. At more negative potentials than the corrosion potential (Ecorr), evolution of hydrogen dominates and results in cathodic current. At more positive potentials than Ecorr, magnesium oxidation dominates and the metal is continuously dissolved with the help of Cl-ions, which can make the oxide film break down. The corrosion current densities (icorr) and the corrosion potentials (Ecorr) of all these specimens are listed in Table 2. The corrosion current density can be ranked in a decreasing order: A-T4>B as-cast>B-T4>A as-cast. The corrosion current density (icorr(mA/cm2) is related to the average penetration rate (Pj(mm/y)) using Eq (1) and the results are listed in Table 3. For an alloy, the solution treatment results in the decrease of the corrosion rate, because the β-phase without zinc mainly serves as the corrosion barrier to inhibit the corrosion process.

Fig.1 Optical micrographs of Mg-Al-Pb (A) alloys: (a) A-as cast, (b) closed-up view of (a) and (c) A-T4.

Fig.2 XRD pattern of Mg-Al-Pb (A) and Mg-Al-Pb-Zn (B) alloys: (a) A-T4, (b) B-T4, (c) A-as cast and (d) B-as cast.

Fig.3 Potentiodynamic polarization curves of Mg-Al-Pb (A) and Mg-Al-Pb-Zn (B) alloys in 3.5% NaCl solution at 25℃: (a) T4 treated specimens of A and B, (b) as-cast specimens of A and B, (c) T4 treated and as-cast specimens of A and (d) T4 treated and as-cast specimens of B

Table 2 Corrosion potentials (Ecorr) and corrosion current densities (icorr) of Mg-Al-Pb (A) and Mg-Al-Pb-Zn (B) alloys in different heat treatment conditions

3.3 Hydrogen evolution

Fig.4 shows the hydrogen evolution volume (ml/cm2), as a function of immersion time (h), of A and B alloys in different heat treatment conditions immersed in 3.5% NaCl aqueous solution for 72h. For each specimen, there is initially an incubation period during which there is a small rate of hydrogen evolution. Thereafter there is an increase in hydrogen evolution with increasing immersion time. For long time immersion, the rate of hydrogen evolution became linear and the corrosion comes into steady state. In T4 treated specimens, the hydrogen evolution volume of A alloy is higher than that of B alloy during short time immersion. After long time immersion, the hydrogen evolution volume of B alloy is higher than that of A alloy. In as-cast specimens, the hydrogen evolution volume of B alloy is higher than that of A alloy during the whole process of immersion test. To the same alloy under different heat treatment conditions, the hydrogen evolution volume of T4 treated specimen is higher than that of as-cast specimen after long time immersion. The hydrogen evolution volume can be ranked in a decreasing order after 72h immersion: B-T4>A-T4>B-as cast>A-as cast. The average hydrogen evolution rate ((ν(H2)(mg/cm2/h)) for 6 and 72h immersion, calculated using Eq(2), is related to the average penetration rate (PH2(mm/y)) using Eq (3) and the results are listed in Table 3.

Fig.4 Hydrogen evolution volume as a function of immersion time of Mg-Al-Pb (A) and Mg-Al-Pb-Zn (B) alloys in 3.5% NaCl solution at 25℃Table 3 Corrosion rate (mm·a-1) for A and B alloys estimated from the corrosion current density (Pi), and evolved hydrogen (PH2)

Corrosionrate/mm·a-1A?T4B?T4A?ascastB?ascastPi12 135 485 148 82PH2，6himmersion15 319 726 779 86PH2，72himmersion81 8593 1248 1472 34

3.4 Surface roughness

Fig.5 shows the surface roughness (μm) of the corroded area on the specimen surfaces, as a function of immersion time (h), of A and B alloys under different heat treatment conditions immersed in 3.5% NaCl aqueous solution for 72h. The surface roughness of the corroded area reflects the actual surface area during the corrosion process to some extent. For each specimen, the shape of the surface roughness/immersion time curve is approximately S shaped. There is initially an incubation period during which the surface roughness is small or even decreases with immersion time. Thereafter there is a rapid increase in surface roughness with increasing immersion time. After long time immersion, the surface roughness comes into steady state. The surface roughness can be ranked in a decreasing order after 72h immersion: B-T4>A-T4>B as-cast>A as-cast, which is consistent with that of hydrogen evolution volume.

Fig.5 Surface roughness of corroding specimens as a function of immersion time of Mg-Al-Pb (A) and Mg-Al-Pb-Zn (B) alloys in 3.5% NaCl solution at 25℃: (a) T4 treated specimen of A, (b) T4 treated specimen of B, (c) as-cast specimen of A, (d) as-cast specimen of B

3.6 Corrosion morphology

Fig.6 shows the corrosion morphologies under SEM (secondary electron) of A and B alloys in different heat treatment conditions immersed in 3.5% NaCl aqueous solution for 6 h. For T4 treated specimens, the alloy suffered pitting corrosion to

some extent from the low magnification morphology (Fig.6(a)). The corrosion pits mainly distributed along and/or adjacent to the grain boundaries from the high magnification morphology (Fig.6(b)). Less severely attack is observed for B alloy (Fig.6(c)). For as-cast specimens, both alloys suffered attack to some extent and the corrosion spots mainly present in the vicinity of second phases (Fig.6(d-e)). These results indicate that the corrosion takes place along the grain boundaries for T4 treated specimens while adjacent to the second phases for as-cast specimens.

Fig.7 shows the SEM (secondary electron) corrosion morphologies of A and B alloys in different heat treatment conditions immersed in 3.5% NaCl aqueous solution for 48 h. The corrosion morphology of B-T4 specimen, which is not shown in Fig.7, is similar to that of A-T4 specimen. The chemical compositions of the selected points in the as-cast specimens of A and B alloys obtained by emission

Fig.6 SE (secondary electron) corrosion morphologies of Mg-Al-Pb (A) and Mg-Al-Pb-Zn (B) alloys immersed in 3.5% NaCl solution at 25℃ for 6h: (a) A-T4, (b) closed-up view of (a), (c) B-T4, (d) A-as cast and (e) B-as cast

Fig.7 SE (secondary electron) corrosion morphologies of Mg-Al-Pb (A) and Mg-Al-Pb-Zn (B) alloys immersed in 3.5% NaCl solution at 25℃ for 48h: (a) A-T4, (b) closed-up view of (a), (c) A-as cast and (d) B-as cast

spectrum analysis (ESA) are listed in Table 4. The corrosion area is enlarged and spread to the whole surface of the specimen for 48 h immersion. For T4 treated specimens, many pits are still observed in the low magnification morphologies (Fig.7(a)) and these pits are mainly situated along and/or adjacent to the grain boundaries (Fig.7(b)). This result indicates that the corrosion develops downward, and still along the grain boundaries in the form of pitting corrosion, after it spreads to the whole surface of the alloy. For as-cast specimens, the second phases are still embedded in the α-Mg phase after 48h immersion, which can be observed in Fig.7(c) and (d). The corrosion pits are still surrounding the second phases, which is more obvious in B as-cast specimen. Besides, many pits are observed in the corroded surfaces. This result implies that the corrosion develops downward, and they are always surrounding the second phases, after it spreads to the whole surface of the alloy for a long time immersion.

Table 4 Chemical compositions of selected points in as-cast specimens obtained by emission spectrum analysis (ESA) /mass fraction,%

4 Discussion

4.1 Corrosion evaluation

The average corrosion penetration rate, Pi, and PH2, evaluated from the corrosion current density and evolved hydrogen, respectively, are summarized in Table 3. For each specimen, the corrosion penetration rate, Pi, is similar to PH2immersed for 6h which are much smaller than those immersed for 72h. This phenomenon is consistent with what Zhao reported about the corrosion behavior of AZ31, AZ91, AM30, AM60 and ZE41 magnesium alloys in NaCl solution[9, 12]. According to Zhao et al[9], the corrosion rate evaluated from the corrosion current density may relate to the onset of corrosion (6h immersion), whereas the corrosion rate from the hydrogen evolution measurement relates to corrosion averaged over a considerable period of time including a considerable long time corrosion (72h immersion) after corrosion onset. According to Fig.4, there is an incubation period at the corrosion onset during which the hydrogen evolution rate is low. The hydrogen evolution rate increases with increasing immersion time and became linear after long time immersion. This is attributed to the increasing actual surface area during the process of immersion[9]. The actual surface area of the corroding specimen is controlled by the corroded apparent area and the surface roughness of the corroded area. During short time immersion, only some sites on the surface are corroded, indicating that the corroded apparent area is small. The corroded apparent area is enlarged and expanded to the whole surface after long time immersion. The hydrogen mainly evolved from the corroded area during immersion. According to Fig.5, the surface roughness of the corroded area is small or even decreases with immersion time during short time immersion but the actual surface area is enlarged due to the expanded corroded apparent area. Thereafter the surface roughness increases rapidly with increasing immersion time and comes into steady state after long time immersion. Fig.8 shows the relationship between the surface roughness and the average hydrogen evolution rate of A and B alloys for various periods of immersion in 3.5% NaCl solution. For a long time immersion, the increasing rate of the surface roughness with increasing immersion time is similar to that of the average hydrogen evolution rate. This means the hydrogen evolution rate of A and B alloys increase linearly with the increasing surface roughness after long time immersion when steady state has been reached. The reason is that the whole surface is corroded after long time immersion and the actual surface area of the corroding specimen is mainly determined by the surface roughness. During short time immersion, the increasing rate of the surface roughness as the increasing immersion time is lower than that of the hydrogen evolution rate, indicating the linear relationship between the surface roughness and the average hydrogen evolution rate is not existed. The reason is that the surface roughness decreases while the hydrogen evolution rate increases with immersion time during short time immersion, which can be attributed to the enlarged corroded apparent area, leading to the enlarged actual surface area. The average hydrogen evolution rate after 72 h immersion is directly related to the corroded surface roughness as shown in Fig.9. It provides a cross plot of the independent measurements of hydrogen evolution rate and corroded surface roughness after 72 h immersion when the corrosion is well established. There is a linear relationship between the hydrogen evolution rate and the corroded surface roughness after long time immersion when the corrosion comes into steady state. However, this relationship is not probable for short time immersion.

Fig.8 Relationships between average hydrogen evolution rate (ml/cm2/h) and surface roughness (μm) of corroding surfaces of Mg-Al-Pb (A) and Mg-Al-Pb-Zn (B) immersed in 3.5% NaCl solution at 25℃ for various periods: (a) T4 treated specimen of A, (b) T4 treated specimen of B, (c) as-cast specimen of A and (d) as-cast specimen of B

Fig.9 Cross plot of independent measurements of hydrogen evolution rate and corroded surface roughness for 72h immersion

According to Table 3, for T4 treated specimens, the corrosion rate of B alloy is lower than that of A alloy during short time immersion (6 h). After long time immersion (72 h), the corrosion rate of B alloy is larger than that of A alloy. This means that zinc inhibits the corrosion of T4 treated Mg-Al-Pb anode during incubation period while it accelerates the corrosion of Mg-Al-Pb anode after long time immersion when steady state has been established. In T4 treated specimen, there is a meta-stable, partially protective film on the α-Mg matrix[19]. According to Fig.3(a), the cathodic current densities of both T4 treated specimens gradually increase with the enhancement of cathodic potential. At the same cathodic potential, the cathodic current density of B alloy is lower than that of A alloy, indicating that zinc with low exchange current density of hydrogen evolution inhibits the hydrogen evolved from the surface of alloy and makes the partially protective film more stable. The less severely attacked corrosion morphology of B alloy after 6h immersion shown in Fig.6(c) can also support this point. The corrosion rate of B-T4 treated specimen could be higher once the protective film is broken down due to long time immersion when steady state has been established. According to Table.1, the content of iron (impurity) in B alloy is larger than that of A alloy, which might be another reason for higher corrosion rate of B alloy after long time immersion. For as-cast specimens, only Mg17Al12(β-phase) exists as the second phase in the alloys according to XRD spectrum shown in Fig.2(c) and (d). The reason is that lead dissolves in the magnesium matrix with a high solubility during melting and casting. The content of zinc is low and it cannot be detected by XRD spectrum. According to ESA results shown in Table 4, zinc is mainly distributed in the second phases of B as-cast specimen. For as-cast specimens, the hydrogen evolution volume of B alloy is larger than that of A alloy during the whole process of immersion (Fig.4). The corrosion morphologies of the as-cast specimens after 48h immersion indicate that the corrosion pit surrounding the second phases for B alloy (Fig.7(d)) is deeper than that of A alloy (Fig.7(c)). This means that the corrosion of the as-cast specimen is more severely with zinc in the second phases. For A alloys, the corrosion rate of T4 treated specimen is higher than that of as-cast specimen during the whole process of immersion. For B alloys, the corrosion rate of T4 treated specimen is lower than that of as-cast specimen for short time immersion while it is higher than that of as-cast specimen for long time immersion. In magnesium alloys, the α-Mg matrix corrodes due to its very negative free corrosion potential and there is the tendency for the corrosion rate of the α-Mg matrix to be accelerated by micro-galvanic coupling between the α-Mg matrix and the β-phase[13]. However, this accelerated effect by micro-galvanic coupling is only found in B as-cast alloy during short time immersion. According to Song[20], the β-phase serves mainly as a galvanic cathode and accelerates the corrosion of the α-Mg matrix if it is present as a small fraction while it may act mainly as an anodic barrier against the overall corrosion of the alloy if its fraction is high. According to Fig.3(c), at the same cathodic potential, the cathodic current density of A as-cast specimen is lower than that of A-T4 specimen, indicating that the β-phase in A alloy inhibits the hydrogen evolved from the surface of the alloy and act as a corrosion barrier to hinder corrosion. In B as-cast specimen, with zinc in the β-phase, the corrosion of α-Mg matrix is accelerated in the beginning owing to micro-galvanic corrosion. When the corrosion attack reaches the β-phases, the corrosion is retarded to a certain extent.

4.2 Corrosion mechanism

The corrosion mechanism of A and B alloys in different heat treatment conditions is schematically shown in Fig.10. For T4 treated specimens, the corrosion happens along the grain boundaries in the form of pitting corrosion. The grain boundary is not stable thermodynamically thus it is more easily to be attacked during immersion test. According to Zhang et al[21], the pit corrosion is dominated by the synergistic effect of the pit initiation and pit growth. The pit initiation rate increases with immersion time after long time immersion and the corrosion spreads to the whole surface and develops downward, which still takes place along the grain boundaries in the form of pitting corrosion. For as-cast specimens, the corrosion is initiated surrounding the second phases due to the effect of micro-galvanic coupling. After long time immersion, the corrosion spread to the whole surface and develops downward, which is still initiated near the second phases.

Fig.10 Sketch map for the corrosion mechanism of Mg-Al-Pb (A) and Mg-Al-Pb-Zn (B) alloys in different heat treatment conditions

5 Conclusion

1. There is an incubation period at the corrosion onset of Mg-Al-Pb and Mg-Al-Pb-Zn anodes during which the corrosion rate is low. The hydrogen evolution rate of Mg-Al-Pb anode increases with increasing surface roughness of corroding alloys after long time immersion when steady state has been established.

2. Zinc inhibits the corrosion of T4 treated Mg-Al-Pb anode during incubation period while it accelerates the corrosion when the corrosion comes into steady state. The second phases in as-cast Mg-Al-Pb anode act as a corrosion barrier to hinder thecorrosion especially after long time immersion.

3. For T4 treated Mg-Al-Pb anode, the corrosion takes place along the grain boundaries. It spreads to the whole surface after long time immersion and develops downward, but still along the grain boundaries. For as-cast specimens, the corrosion occurs around the second phases. For long time immersion, it spread to the whole surface and develops downward, which still initiates surrounding the second phases.

Reference

[1] Renuka R. Influence of Allotropic Modifications of Sulphur on the Cell Voltage in Mg-CuI(S) Seawater Activated Battery[J]. Materials Chemistry and Physics, 1999, 59(1): 42～48.

[2] Feng Y, Wang R C, Yu K, Peng C Q, Li W X. Influence of Ga and Hg on Microstructure and Electrochemical Corrosion Behavior of Mg Alloy Anode Materials[J]. Transactions of Nonferrous Metals Society of China, 2007, 17(6): 1363～1366.

[3] Renuk R. AgCl and Ag2S as Additives to CuI in Mg-CuI Seawater Activated Batteries[J]. Journal of Applied Electrochemistry, 1997, 27(12): 1394～1397.

[4] Cao D X, Wu L, Wang G L, Lu Y Z. Electrochemical Oxidation Behaviour of Mg-Li-Al-Ce-Zn and Mg-Li-Al-Ce-Zn-Mn in Sodium Chloride Solution[J]. Journal of Power Sources, 2008, 183(2): 799～804.

[5] Feng Y, Wang R C, Peng C Q, Wang N G. Influence of Mg21Ga5Hg3Compound on Electrochemical Properties of Mg-5%Hg-5%Ga Alloy[J]. Transactions of Nonferrous Metals Society of China, 2009, 19(1): 154～159.

[6] Fira S, Kibl L, Lin L W. Water-activated Disposable and Long Shelf Life Microbatteries[J]. Sensors and Actuators A: Physical, 2004, 111(1): 79～86.

[7] Rao K V. Performance Evaluation of Mg-AgCl Batteries for Underwater Propulsion[J]. Denfence Science Journal, 2001, 51(2): 161～170.

[8] Fidel G M, Juan M F, Ruben D R, Genesca J. Electrochemical Study on Magnesium Anodes in NaCl and CaSO4-Mg(OH)2Aqueous Solutions[J]. Electrochimica Acta, 2006, 51(8): 1820～1830.

[9] Zhao M C, Liu M, Song G L, Atrens A. Influence of pH and Chloride ion Concentration on the Corrosion of Mg Alloy ZE41[J]. Corrosion Science, 2008, 50(11): 3168～3178.

[10] Hiromoto S, Yamamoto A, Maruyama N, Somekawa H, Mukai T. Influence of pH and Flow on the Polarisation Behaviour of Pure Magnesium in Borate Buffer Solutions[J]. Corrosion Science, 2008, 50(12): 3561～3568.

[11] Shi Z M, Song G L, Atrens A. Influence of the β Phase on the Corrosion Performance of Anodised Coatings on Magnesium-aluminium Alloys[J]. Corrosion Science, 2005, 47(11): 27620～2777.

[12] Zhao M C, Schmutz P, Brunner S, et al. An Exploratory Study of the Corrosion of Mg Alloys Suring Interrupted Spray Testing[J]. Corrosion Science, 2009, 51(6): 1277～1292.

[13] Zhao M C, Liu M, Song G L, Atrens A. Influence of the β-phase Morphology on the Corrosion of the Mg Alloy AZ91[J]. Corrosion Science, 2008, 50(7): 1939～1953.

[14] Gu X N, Zheng Y F, Cheng Y, et al. In Vitro Corrosion and Biocompatibility of Binary Magnesium Alloys[J]. Biomaterials, 2009, 30(4): 484～498.

[15] Pardo A, Merion M C, Coy A E, et al. Influence of Microstructure and Composition on the Corrosion Behaviour of Mg/Al Alloys in Chloride Media[J]. Electrochimica Acta, 2008, 53(27): 7890～7902.

[16] Candan S, Unal M, Turkmen M, et al. Improvement of Mechanical and Corrosion Properties of Magnesium Alloy by Lead Addition[J]. Materials Science and Engineering: A., 2009, 501(1): 115～118.

[17] Udhayan R, Bhatt D P. On the Corrosion Behaviour of Magnesium and Its Alloys Using Electrochemical Techniques[J]. Journal of Power Sources,1996, 63(1): 103～107.

[18] Badawy W A, Hilal N H, et al. Electrochemical Behaviour of Mg and Some Mg Alloys in Aqueous Solutions of Different pH[J]. Electrochimical Acta, 2010, 55(6): 1880～1887.

[19] Zhou W, Shen T, Aung N N. Effect of Heat Treatment on Corrosion Behaviour of Magnesium Alloy AZ91D in Simulated Body Fluid[J]. Corrosion Science, 2010, 52(3): 1035～1041.

[20] Song G L, Atrens A, Dargusch M. Influence of Microstructure on the Corrosion of Diecast AZ91D[J]. Corrosion Science, 1999, 41(2): 2490～273.

[21] Zhang T, Chen C M, Shao Y W, et al Corrosion of Pure Magnesium under Thin Electrolyte Layers[J]. Electrochimica Acta, 2008, 53(27): 7921～7931.