Effect of Al and Y addition on corrosion properties of magnesium alloy

1 Introduction

Magnesium alloys are widely used in aviation and automobile industries due to their excellent properties (Das et al. 2022; Gogheri et al. 2020; Zhao et al. 2021). However, low strength and poor corrosion resistance are two major factors affecting the application of magnesium alloys (Esmaily et al. 2017; Song et al. 1998). Currently, efforts have been made to significantly improve the strength of magnesium alloys (Wu et al. 2011; Zeng et al. 2021). However, the corrosion issues of magnesium alloys remain unresolved. The main reasons for the corrosion susceptibility of magnesium alloys are twofold: (i) The alloy matrix itself is prone to corrosion. The standard electrode potential of magnesium is −2.37 V_SCE (SCE, standard calomel electrode), indicating high chemical reactivity (Milazzo et al. 1978). Moreover, in most magnesium alloys, the potential of the second phase is more positive than that of the magnesium matrix, making the magnesium matrix susceptible to anodic corrosion dissolution in humid and solution environments. (ii) The generated oxide film is not dense enough (PB < 1), resulting in limited protection capability for the matrix (Bland et al. 2017).

Currently, researchers both domestically and internationally are working on improving the corrosion resistance of magnesium alloys through alloying (Chu et al. 2024; Li et al. 2021; Wang et al. 2010; Xiao et al. 2020), reducing impurity elements (Azzeddine et al. 2020; Gandel et al. 2011; Yang et al. 2018), micro-arc oxidation (Azarian and Khoei 2021; Chen et al. 2019; Farshid and Kharaziha 2020; He et al. 2021), surface paint spraying (Song 2009; Wu et al. 2016), and other methods. It is believed that it is difficult to suppress the corrosion of magnesium alloys by reducing the anodic dissolution of the magnesium matrix (Huang et al. 2020; Song et al. 2016). Therefore, improving the protective properties of the magnesium alloy surface film may be the only way to achieve acceptable corrosion resistance (Cao et al. 2016). Compared to complex and inefficient surface treatment processes, alloying and microalloying can improve the corrosion resistance of magnesium alloy films from the intrinsic perspective, while also considering mechanical properties (Li et al. 2021, 2020). There has been extensive research on this, but it mainly focuses on adding alloying elements to existing alloys and the corrosion resistance of Mg-X binary alloys (Cao et al. 2013; He et al. 2022; Liu et al. 2010; Shi et al. 2013; Xie et al. 2023; Xu et al. 2016).

Mg–Al based alloys are widely used in industrial applications, with AZ, AM, and AE series alloys having good castability and yield strength. In terms of corrosion characteristics, the addition of Al can facilitate the formation of a protective film, inhibiting corrosion of the alloy. However, the addition of Al can lead to the formation of the Mg₁₇Al₁₂ phase. Due to variations in the content and distribution of the Mg₁₇Al₁₂ phase, it can either accelerate or hinder the corrosion of α-Mg (Cui et al. 2018; Raman 2004). Rare earth Y has a standard electrode potential similar to that of Mg (−2.372 V_SCE) and is considered another element that could potentially improve the corrosion performance of Mg alloys (Liu et al. 2010). Velikokhatnyi and Kumta (2010) used density functional theory calculations and found that the addition of Y to Mg facilitates the formation of a more stable and less chemically reactive hydroxide protection film. This may result in these alloys having higher corrosion resistance. Xu et al. (2016) studied the effect of different Y contents on the corrosion performance of binary Mg–Y alloys and found that Y can significantly improve the corrosion resistance of Mg–Y alloys. However, excessive Y can deteriorate the corrosion performance of the alloy. Liu et al. (2009) investigated the corrosion behavior of Mg–Y alloys in solution and found that the addition of Y can form a protective film. It is noteworthy that the study pointed out that excessive Y elements can form second phases, which accelerate the corrosion of the alloy in 0.1 M NaCl solution. However, in 0.1 M Na₂SO₄ solution, although more second phases are present, the protective film on the alloy surface is stronger. Furthermore, many researchers have found that the addition of Y elements to existing alloys can either increase or decrease the corrosion rate of the alloy. Additionally, studies by Cao et al. (2013) and Shi et al. (2013) investigated and summarized the corrosion performance of Mg-X binary alloys, finding that the corrosion rates of binary alloys are generally higher than that of high-purity magnesium. From the aspect of alloy composition design, there is relatively less research on ternary alloys and alloys with more added elements. Most studies focus on common magnesium alloys such as AZ series, including AZ91 (Liu et al. 2018; Taltavull et al. 2014) and AZ31 (Feliu et al. 2019; Kim et al. 2013; Saikrishna et al. 2016; Wu et al. 2023). Zhu et al. (2022) designed and prepared Mg–11Y–1Al alloy, and found that a small amount of Al addition can assist Y in forming Y₂O₃/Y(OH)₃ film on the alloy surface, thereby enhancing the corrosion resistance of the alloy. Kim et al. (2019) investigated the effect of adding Y on the removal of iron from magnesium melt. They found that the corrosion rate of Mg–3Al-0.2Y was 1.4 mm/year. This reduction in corrosion rate is mainly due to the small amount of Y, which lowers the Fe content. From the above literature review, it can be concluded that the solid solution of Al and Y in Mg alloys helps the alloy form a protective film, thus reducing the corrosion rate of the alloy (Zhu et al. 2022). Additionally, the element Y forms a secondary phase with the impurity element Fe in the alloy and precipitates at the bottom of the molten crucible, thereby reducing the Fe concentration in the alloy (Kim et al. 2019). However, the influence of second-phase particles containing Al/Y on the corrosion performance of Mg alloys tends to accelerate corrosion, although the results remain uncertain.

In summary, current research on improving the corrosion resistance of magnesium alloys mainly focuses on Mg-X binary alloys and micro-alloying of existing alloys, with relatively less research on the ternary Mg–Al–Y alloys with composite additions of Al and Y. Therefore, in this work, Mg–2Al–2Y alloy was designed and prepared by compound adding Al and Y elements into the alloy. To investigate the synergistic effects of Al and Y elements on the corrosion performance of magnesium alloys, Mg–2Al and Mg–2Y alloys were used as the baseline references. By comparing these with the Mg–2Al–2Y alloy, the individual effects of Al and Y, as well as their synergistic interactions, were systematically revealed, providing a theoretical basis for further optimization. By analyzing the corrosion morphology and composition of the materials, the mechanism of the combined addition of Al and Y on the corrosion resistance of the alloy was elucidated.

2 Materials and methods

3 Results

3.1 Microstructure

The backscattered images of the alloy, as shown in Figure 1. Energy-dispersive X-ray spectroscopy (EDS) analysis was conducted on particles from different regions, with the results summarized in Table 2. In the LP–Mg alloy and Mg–2Al alloy, circular particles (indicated by arrows A and D) were found to contain a significant amount of oxygen but no other elements besides Mg, suggesting these are oxides formed during sample polishing. Additionally, a few bright white particles were observed in LP–Mg (arrow B), which EDS identified as Fe-rich phases. In the Mg–2Al alloy, Al-rich phases were detected. The Mg–2Y alloy showed numerous bright white particles, identified as Y-rich phases by EDS. Unlike the Mg-X binary alloys, the second phase in the Mg–2Al–2Y ternary alloy was distributed along the dendritic grain boundaries, forming a network structure. It was identified as Al₂Y phases by EDS analysis.

Figure 1:

SEM images of studied alloys: (a) LP–Mg; (b) Mg–2Al; (c) Mg–2Y; (d) Mg–2Al–2Y.

Table 2:

Chemical compositions of the microconstituents in Mg alloy (at.%).

	Mg	Al	Y	Si	Fe	O
A	82.85	–	–	0.27	–	16.89
B	96.75	–	–	1.47	1.78	–
C	81.07	18.82	–	0.11	–	–
D	89.82	2.34	–	0.19	–	11.77
E	16.20	0.22	83.53	0.05	–	–
F	94.42	–	5.47	0.11	–	–
G	82.27	11.91	5.70	0.11	–	–
H	79.63	13.13	7.23	0.02	–	–

XRD analysis was conducted on the as-cast alloy, and the results were depicted in Figure 2. LP–Mg, Mg–2Al, and Mg–2Y alloys exhibit only peaks corresponding to the α-Mg matrix, with no detection of a second phase. This may be due to the lower content of the second phase. In the Mg–2Al–2Y alloy, a peak corresponding to Al₂Y is detected, confirming the presence of Al₂Y in the alloy. This finding is consistent with the results of EDS analysis.

Figure 2:

XRD patterns of LP–Mg and studied alloys.

3.2 Electrochemical tests

Figure 3 present the OCP and PDP curves. The OCP show that the OPC of LP–Mg is the more positive values (−1.67 V_SCE), and it changes little with increasing time, show as Figure 3(a). The OCP of the Mg–2Al–2Y alloy reaches a stable state first and fluctuates within a certain range with increasing time. The OCP of Mg–2Al and Mg–2Y alloys is the lowest and gradually increases with time. Figure 3(b) shows the Potentiodynamic polarization curves obtained after immersing the alloy in a 3.5 % NaCl solution for 30 min. Compared to LP–Mg, the alloys exhibit a significant breakdown potential (E_b) in the anodic region. At the same potential, the alloy has smaller anodic and cathodic current densities.

Figure 3:

Electrochemical test results of LP–Mg and Mg alloy in 3.5 wt%NaCl solution: (a) OCP curves; (b) PDP curves.

The corrosion current density (i_corr), corrosion potential (E_corr, V_SCE), cathodic Tafel slope (β _c ), and anodic Tafel slope (β_a) of the sample extracted from the polarization curve are shown in Table 3. Additionally, the polarization resistance (R_p) and corrosion rate (P_i) of the material can be calculated using the i_corr, β_a, and β_c. R_p can be calculated using formula (5) (Bakhsheshi-Rad et al. 2012, 2016).

(5) R p = β a β c 2.303 β a + β c i corr

Table 3:

Polarization curve fitting parameters of LP–Mg and studied alloys in 3.5 wt%NaCl solution.

Alloy	Ecorr (VSCE)	icorr (mA/cm2)	Eb (VSCE)	βa (mV/decade)	−βc (mV/decade)	Rp (Ωcm2)	Pi (mm/year)
LP–Mg	−1.63	1.93E-2		9	12	80	0.412
Mg–2Al	−1.64	5.87E-4	−1.58	17	24	4,678	0.013
Mg–2Y	−1.66	1.12E-3	−1.58	10	12	1,459	0.025
Mg–2Al–2Y	−1.65	1.39E-3	−1.55	14	13	1,316	0.029

P_i can be calculated using formula (6) (Bakhsheshi-Rad et al. 2012, 2016, Shi et al. 2010, Zhao et al. 2009).

(6) P i = 22.85 i corr

It is worth noting that formula (6) is an empirical formula specifically for calculating the corrosion rate of magnesium alloys and is only applicable to magnesium and its alloys.

From Table 3, it can be seen that the E_corr difference among the four materials is not significant. The alloy exhibits a lower i_corr (one or two orders of magnitude lower than LP–Mg), a substantial increase in R _p , and one order of magnitude decrease in P_i. The i_corr and P _i of Mg–2Al are the lowest. The E_b of the Mg–2Al–2Y alloy is the most positive, indicating that the film formed on the surface of the Mg–2Al–2Y alloy is more resistant to breakdown.

Figure 4 shows the impedance spectra obtained from the specimens. Figure 4(a) shows the Nyquist plot. The impedance of LP–Mg consists of two parts: high-frequency capacitive reactance and low-frequency inductive reactance. The Nyquist plot of the Mg alloy mainly consists of high-frequency capacitive arc and low-frequency capacitive arc. Figure 4(b) is the Bode plot of impedance versus frequency. As the frequency changes from high to mid-range, the impedance of the material increases. At low frequencies, the impedance value decreases for LP–Mg, while it increases for the alloy. Figure 4(c) is the Bode plot of phase angle versus frequency, which includes two peaks and one valley. The peaks represent the capacitive circuit, while the valley represents the low-frequency inductive circuit. Combining the corrosion behavior of the four materials and previous research results (Li et al. 2016; Baril et al. 2007), the impedance data of LP–Mg were fitted with the equivalent circuit model shown in Figure 5(a), whereas the magnesium alloy data were analyzed using the model in Figure 5(b). The fitting results are presented in Table 4. Here, R _s represents the solution resistance, R_t represents charge transfer resistance, R_f represents corrosion film resistance, and R _l represents low-frequency inductive resistance. CPE _t represents double-layer capacitance, CPE _f represents corrosion film capacitance. The electric double layer at the interface between the electrode and the solution is generally equivalent to a capacitor, known as the electric double layer capacitance. However, experiments have shown that the frequency response characteristics of the solid electrode’s electric double layer capacitance do not fully match the behavior of a “pure capacitor”, exhibiting deviations. This phenomenon is commonly referred to as the “dispersive effect”. Considering the “dispersive effect”, a constant phase angle element (CPE) is used instead of pure capacitance (Baril et al. 2007; Deng et al. 2021; Feng et al. 2022; Huang et al. 2020; Song and Unocic 2015; Zhang et al. 2011). L represents low-frequency inductance. The impedance of the CPE is given by the equation (7) (Huang et al. 2016):

(7) Z CPE = 1 Q j ω n

where Q is the CPE coefficient with units of Ω⁻¹·cm⁻²·sⁿ, j is the imaginary unit (j =  − 1 , ω is the angular frequency (ω = 2πf), n is the exponent that describes the deviation from ideal capacitive behavior (0 ≤ n ≤ 1). In Table 4, CPE-T and CPE-P represent the parameters Q and n, respectively.

Figure 4:

EIS tests results of LP–Mg and studied alloys in 3.5 wt%NaCl solution: (a) Nyquist diagram; (b) and (c) bode diagram.

Figure 5:

Equivalent circuits for fitting the nyquist diagrams given in Figure 4: (a) LP–Mg; (b) Mg–2Al, Mg–2Y and Mg–2Al–2Y.

Table 4:

Fitting values for the equivalent circuits of LP–Mg and studied alloys immersed in 3.5 wt%NaCl solution.

	LP–Mg	Mg–2Al	Mg–2Y	Mg–2Al–2Y
Rs (Ω cm2)	30	10	5	40
CPEt-T(Ω−1·cm−2·sn)	30	800	897	2,163
CEPt-P	1.02	0.80	0.75	0.94
Rl (Ω cm2)	40	–	–	–
Rt (Ω cm2)	145	1,600	1,580	890
Rf (Ω cm2)	–	2,180	2,500	940
CPEf -P(Ω−1·cm−2·sn)	–	17	18	16
CEPf-P		0.94	0.91	0.95
L (H/cm2)	512	–	–	–
Rp (Ω cm2)	31	3,780	4,080	1,830

R _p represents polarization resistance, used to measure the total resistance of the electrode reaction. The reciprocal of polarization resistance, 1/R_p, is directly proportional to the corrosion rate of the alloy (Song et al. 2004). Different types of polarization resistance R_p can be calculated using the following equation (8) and (0). The R_p of LP–Mg is calculated using equation (8), while the R_p of the magnesium alloy is calculated using equation (9).

(8) R p = 1 1 R t + 1 R l

(9) R p = R t + R f

where R_t is charge transfer resistance, R_f is corrosion film resistance, and R _l represents low-frequency inductive resistance. The unit is Ω cm².

Observations of the surface corrosion morphology of the alloy after electrochemical testing were made using laser confocal microscopy, as shown in Figure 6. The surface of LP–Mg exhibits numerous deep corrosion groove, while the surfaces of other alloys only show pitting corrosion.

Figure 6:

The surface corrosion morphology of the alloy after electrochemical testing: (a) LP–Mg; (b) Mg–2Al; (c) Mg–2Y; (d) Mg–2Al–2Y.

3.3 Immersion test

Immersion experiments and hydrogen evolution experiments provide the most direct observation of the corrosion properties of materials. Figure 7 presents the hydrogen evolution curves and weight loss rates of the alloy after immersion in a 3.5 wt%NaCl solution for 7 days. Figure 7(a) shows the variation of hydrogen evolution per unit area over time for the alloy. It can be observed that the hydrogen evolution of Mg alloys is significantly reduced compared to LP–Mg. Mg–2Al–2Y exhibits the lowest hydrogen evolution after immersion in 3.5 wt%NaCl solution for 168 h. Figure 7(b) presents the average hydrogen evolution rate and weight loss rate of the materials after immersion for 168 h. The corrosion rate measured by weight loss method is slightly higher than that by hydrogen evolution test. The hydrogen evolution rate of Mg–2Al–2Y alloy after 168 h of immersion is 0.57 mm/y, while the weight loss rate is 1.01 mm/y, much lower than that of LP–Mg.

Figure 7:

The data obtained from immersion experiments. (a) Volume of evolved hydrogen and (b) average hydrogen evolution rate variation over time of LP–Mg and Mg alloy immersed in 3.5 wt%NaCl solution; (c) hydrogen evolution rate and weight loss rate in 3.5 wt%NaCl solution for 7 days.

XRD tests were conducted on the samples after immersion experiments, and the results shown in Figure 8. The results indicate that the main component in the corrosion film on the sample after immersion is Mg(OH)₂.

Figure 8:

XRD patterns of corrosion film of LP–Mg and Mg alloy immersed in 3.5 wt%NaCl solution for 3 days.

Figure 9 illustrates the corrosion morphology of the material surface after immersion in 3.5 wt%NaCl solution for 30 min and 7 days. This result suggests that after 30 min of immersion, the surface of LP–Mg exhibits severe corrosion, while the surfaces of Mg alloys show numerous small and shallow corrosion pits. After 7 days of immersion, the surface of LP–Mg exhibits extensive matrix detachment, forming deeper corrosion grooves. There is little difference in the corrosion morphology of magnesium alloys, with numerous corrosion pits observed on all surfaces. In comparison, severe corrosion occurs at the edge of the Mg–2Al alloy. It is noteworthy that a large amount of Al₂Y phase is distributed around the corrosion pits on the surface of the Mg–2Al–2Y alloy.

Figure 9:

Corrosion morphology of immersed in 3.5 wt%NaCl solution for 30 min and 7, respectively (after corrosion products removed).

4 Discussion

From the backscatter images, it is observed that The Mg–2Al and Mg–2Y alloys contain a small number of secondary phases, indicating that Al and Y are primarily present in the form of solid solution within the alloys, with only a very small amount existing in an enriched form. In Mg–2Al–2Y alloy, a large amount of second phase is present, exhibiting a reticular distribution. EDS results confirm that this second phase is the Al₂Y phase, indicating the formation of the Al₂Y phase during the solidification process of the alloy with composite additions of Al and Y elements. XRD results show that only peaks corresponding to Mg are present in Mg, Mg–2Al, and Mg–2Y, which may be due to the secondary phase content being too low to be detected. In the Mg–2Al–2Y alloy, however, an Al₂Y peak appears, further confirming the presence of the Al₂Y phase in the alloy. The presence of secondary phases can generate a potential difference between them and the matrix, which may lead to galvanic corrosion of the material. In general, t the occurrence of galvanic corrosion not only requires the satisfaction of thermodynamic conditions but also the fulfillment of kinetic conditions; both are indispensable. The thermodynamic condition for galvanic corrosion requires a significant potential difference between metals of different potentials, which drives electrons from the more active metal (anode) to the less active metal (cathode). This potential difference leads to the dissolution of the metal at the anode, causing galvanic corrosion. Compared to Mg, the potential of most secondary phases is more positive, and they often act as cathodes, where hydrogen evolution reactions occur, while the surrounding magnesium matrix acts as the anode and undergoes dissolution. Therefore, the second phase in the alloy provides the thermodynamic conditions for galvanic corrosion.

Electrochemical methods are among the most commonly used techniques for studying corrosion dynamics, as they can directly reveal the electrochemical behavior of metal surfaces. Electrochemical tests provide information such as the OCP, E_corr, I_corr, and EIS of materials. OCP is a key parameter for evaluating the corrosion tendency of metallic materials. In general, a more positive OCP indicates higher stability and corrosion resistance of the alloy (Feng et al. 2023). Zhao et al. (2008) reported that an increase in OCP reflects the onset and progression of corrosion, while a relatively stable OCP suggests a steady state between corrosion progression and the deposition of corrosion products. Figure 3(a) shows that LP–Mg and Mg–2Al–2Y alloys achieve surface stability within a short period of immersion, compared to Mg–2Al and Mg–2Y alloys. This indicates a faster establishment of equilibrium between surface corrosion and the deposition of corrosion products. The OCP of Mg–2Al and Mg–2Y alloys increases gradually over time, possibly due to the presence of a protective film on the material surface, which delays the onset of localized corrosion beyond 1800 s. In contrast, the earlier stabilization of OCP for the Mg–2Al–2Y alloy, compared to LP–Mg, may be attributed to the presence of a network-distributed second phase (Li et al. 2016). Additionally, fluctuations in the OCP of the Mg–2Al–2Y alloy over time may result from competition between localized corrosion and passivation caused by the corrosion product layer (Liu et al. 2012).

Electrochemical polarization curves are a critical tool for analyzing the cathodic and anodic corrosion behavior of alloys, providing a wealth of information. For passivating metals, the pitting potential (E_pt) and corrosion current density (i_coor) are crucial parameters for assessing corrosion resistance. The E_pt marks the abrupt current increase on the curve due to the breakdown of the passivation film on the alloy surface and is often referred to as the E_b. In this experiment, corrosion pits were observed on the surface after electrochemical testing. Therefore, we believe that for the three alloys in this study, the E_b corresponds to their E_pt. Song et al. (2004) suggested that a sharp current decrease on the cathodic polarization curve also indicates the E_b, as shown by the arrows in Figure 3(b). A higher E_b indicates a lower likelihood of localized corrosion. Compared to LP–Mg, the E_b for binary and ternary alloys appear on the anodic curve. At the E_b, during the measurement of the polarization curve, the initiation of localized corrosion is accompanied by significant and visible hydrogen evolution from the corroding sites on the electrode surface. E_b is an important electrochemical parameter that indicates the tendency for localized corrosion. A more positive E_pt implies a lower likelihood of localized corrosion. Therefore, compared to pure magnesium, alloys are less prone to localized corrosion.

As shown in Table 3, the corrosion potentials of the four materials show no significant differences. However, the addition of Al and Y causes a slight shift to more negative potentials, attributed to the suppression of the cathodic hydrogen evolution reaction, as explained by the mixed potential theory. The polarization curves also show that the corrosion current density of the alloys is reduced by one order of magnitude compared to LP–Mg. Combined with the pitting potential, it can be inferred that a protective film forms on the alloy surfaces, effectively suppressing corrosion.

Electrochemical impedance is an important parameter in electrochemical research, used to describe the response characteristics of electrochemical systems to alternating current (AC) electrical signals, i.e., the impedance behavior of the system. It refers to the degree of hindrance to the flow of current in an electrochemical system under an applied AC electric field. It is generally accepted that the high-frequency capacitive arc in the Nyquist plot represents the double-layer reaction at the electrode-solution interface, while the diameter of the capacitive arc corresponds to the magnitude of the charge transfer resistance (Cao et al. 2019; Li et al. 2016). The mid-frequency or low-frequency capacitive arc is mainly associated with the protective nature of the corrosion product film on the alloy surface (Cao et al. 2019; Li et al. 2016). The low-frequency inductive arc is primarily related to the relaxation process of local breakdown and detachment of corrosion products (Li et al. 2016; Zhang et al. 2011). In comparison to alloys, pure magnesium’s Nyquist plot exhibits an inductive arc in the low-frequency region, indicating significant localized corrosion. The Nyquist plot of magnesium alloys mainly consists of two capacitive arcs, indicating uniform corrosion of magnesium alloys. This observation contrasts with the conclusions drawn from polarization curves. This discrepancy arises because the pitting process is highly localized and rapid. In EIS measurements, low-frequency responses primarily reflect long-term electrochemical behavior in a stable state. If pitting occurs only in a small localized area, it may not significantly affect the overall impedance, particularly at lower frequencies. This can be confirmed through post-electrochemical test surface analysis (Figure 6).

The inverse of polarization resistance (1/R_p) is directly proportional to the corrosion rate of the alloy. The polarization resistance of LP–Mg is significantly lower than that of the alloys, indicating a higher corrosion rate. Among the alloys, Mg–2Al–2Y exhibits lower R_p than Mg–2Al and Mg–2Y, suggesting a faster corrosion rate. This trend is evident from the corrosion rates listed in Table 2. Electrochemical polarization and impedance experiments demonstrate that a protective film forms on the alloy surface, significantly reducing the corrosion rate in NaCl solution compared to LP–Mg. Among the alloys, Mg–2Al–2Y exhibits higher corrosion current density and lower R_p, indicating faster corrosion. However, its higher pitting potential suggests that the protective film resists further degradation, making it more resistant to advanced corrosion. These findings align with the results of hydrogen evolution and weight loss experiments.

Bode and Nyquist plots show the same data in different ways. In Bode plots, higher impedance modulus values in the low-frequency range generally indicate stronger corrosion protection, suggesting the formation of a stable passivation layer or protective film on the material’s surface. Conversely, lower impedance modulus values imply poorer integrity of the protective layer, potentially resulting in higher corrosion rates. As shown in Figure 4(b), the impedance of the alloys increases with decreasing frequency, indicating the possible formation of a stable protective film on the surface. In the Bode phase plots, an increase in the peak phase angle and a broader peak shape also suggests enhanced corrosion resistance of the alloys.

The EIS test is performed at potential below the E_b and only reflects the interfacial reaction kinetics of the material in the passivated film stable state. The polarization curve covers a wide potential range, which may stimulate the passivation film rupture (E > E_b) and capture the pitting initiation behavior. When E < E_b, EIS cannot directly detect pitting initiation, but can indirectly infer film defects through low-frequency impedance characteristics. The induced reactance arc of LP–Mg in the low-frequency region indicates that the surface film of LP–Mg is loose and not play a protective role. The polarization curve can clearly identify the E_b of the material and judge the difficulty of the passivation film rupture on the surface of the material. Therefore, the combination of EIS and polarization curve can better analyze the electrochemical behavior of materials.

Hydrogen evolution and immersion experiments are the most effective methods for reflecting the long-term corrosion behavior of alloys in an electrolytic environment. From the results of the hydrogen evolution experiment, the hydrogen evolution amount of LP–Mg is significantly higher than that of the binary and ternary alloys, with the ternary alloy exhibiting the lowest hydrogen evolution amount. This indicates that the Mg–2Al–2Y alloy has the lowest hydrogen evolution rate, primarily due to its stronger surface film protection capability. From the average hydrogen evolution rate in the preceding n days, it can be observed that the hydrogen evolution rate of LP–Mg decreases with time, indicating that the Mg(OH)₂ formed on the surface, although not dense, still provides a certain hindrance to corrosion. In addition, the decrease in corrosion rate is also related to the change of the solution from neutral to alkaline. During the corrosion process of LP–Mg, (HO)^- is generated, which increases the pH of the solution. In an alkaline environment, Mg(OH)₂ deposits on the magnesium surface to form a protective film. Although this film is not dense, it still plays a role in hindering further corrosion of the magnesium. The corrosion rate of Mg–2Al on the first day is slightly lower than that of Mg–2Y alloy, but with time, the hydrogen evolution rate of Mg–2Y alloy gradually decreases, while that of Mg–2Al alloy slightly increases. This suggests that compared to Mg–2Y alloy, Mg–2Al alloy is more prone to forming a protective film, resulting in a lower corrosion rate in the initial stages of corrosion. This can also be confirmed from the polarization curves. As time progresses, the protective film forms on the surface of Mg–2Y alloy, reducing its corrosion rate. Moreover, the protective film on the surface of Mg–2Y alloy exhibits stronger protection capabilities than that of Mg–2Al alloy. The hydrogen evolution rate of Mg–2Al–2Y alloy shows little variation with time and is lower than that of other alloys, indicating that the protective film on the alloy surface can form quickly and has stronger protection capabilities. The reason for the rapid formation of the protective film on the surface of Mg–2Al–2Y alloy compared to binary alloys is attributed to the composite addition of Al and Y elements, where the Al element promotes the formation of Y element film on the alloy surface. Figure 7(c) shows hydrogen evolution rate and weight loss rate in 3.5 wt%NaCl solution for 7 days. Comparing the hydrogen evolution rate with the weight loss rate of the materials reveals that the hydrogen evolution rate is lower than the weight loss rate, primarily because a small amount of hydrogen gas is not collected during the corrosion process (Wang et al. 2023).

Observation of the surface after immersion reveals that, compared to the alloy, LP–Mg exhibits deep corrosion grooves, indicating severe surface corrosion. The addition of small amounts of Al and Y elements allows the formation of a protective passive film on the alloy surface during the corrosion process, thereby significantly reducing the alloy’s corrosion rate. The composite addition of Al and Y elements results in a mesh-like distribution of second phases in the alloy, further enhancing its corrosion resistance. After immersion for 7 days, the predominant corrosion morphology on the alloy surface is pitting corrosion. In comparison, the corrosion pits on the surface of the Mg–2Al–2Y alloy are shallower. This is mainly due to the stronger protective ability of the passive film on the alloy surface. In addition, a large number of network-distributed Al₂Y phases were observed around the corrosion pits on the surface of the Mg–2Al–2Y alloy. Due to the more negative potential of the Mg matrix, corrosion pits initiated in the Mg matrix and expanded outward. The network-distributed Al₂Y phases isolated the Mg matrix into “island-like” regions, effectively hindering the further propagation of the corrosion pits.

5 Conclusions

1. In the as-cast binary alloy, element A and Y primarily exist in solid solution, while in Mg–2Al–2Y alloy, there is a significant presence of a network-distributed Al₂Y phase. The networked distribution of Al₂Y phase hinders the expansion of corrosion on the Mg–2Al–2Y alloy surface.

2. The polarization curves reveal a distinct pitting potential in the anodic region, with the Mg–2Al–2Y alloy exhibiting the most positive pitting potential (−1.55 V_SEC). The impedance spectrum characteristics of the alloy (two capacitive arcs) indicate that its corrosion behavior is dominated by uniform corrosion. In contrast, the EIS of LP–Mg shows an inductive arc, suggesting that its corrosion behavior is primarily localized corrosion. The presence of the inductive arc indicates localized current fluctuations and instability during the corrosion process of LP–Mg. These results demonstrate that a stable protective film forms on the surface of the alloy, significantly reducing its corrosion rate in NaCl solution compared to LP–Mg. This protective film is a key factor in the improved corrosion performance of the Mg–2Al–2Y alloy.

3. After 7 days of immersion, the Mg–2Al–2Y alloy demonstrates the lowest hydrogen evolution rate (0.57 mm/y) and mass loss rate (1.01 mm/y), attributed to the synergistic effect of Al and Y in promoting a stable protective film.

4. This study provides new insights into the role of the Al2Y phase and the synergistic effect of Al and Y in enhancing the corrosion resistance of magnesium alloys. However, despite these advancements, the precise mechanisms governing the formation and stability of the protective film remain unclear. Future studies should focus on the atomic-scale interactions between Al, Y, and the Mg matrix, as well as the role of the Al₂Y phase in film formation.