Effect of Cu addition on microstructure and corrosion behavior of Al–15Mg₂Si composite

3.1 Microstructural characterization

Figure 2 shows the deep-etched microstructure of the Al–15Mg₂Si composite. Based on the liquidus projection of the Al–Mg–Si equilibrium phase diagram (Figure 3a) [11], the solidification of the base hypereutectic alloy begins with the formation of Mg₂Si_P phase. Following this transformation, due to the consumption of Si and Mg atoms, the liquid composition moves along lines SA and then AB in which the binary eutectic reaction occurs, i.e. L → L₁ + Mg₂Si_P → L₂ + (α-Al + Mg₂Si)_E + Mg₂Si_P → (α-Al + Mg₂Si)_E + Mg₂Si_P [11]. Therefore, the microstructure of the base composite (0.01 wt% Cu) consists of coarse Mg₂Si_P polyhedrals and eutectic cells comprising α-Al and Mg₂Si_E particles (Figure 2). The EDS analysis of Mg₂Si_p and Mg₂Si_E particles in Figure 2 showed a respective Mg to Si content of 68 to 32 and 66.3 to 33.7 (compositions in wt%).

Figure 2:

Deep-etched SEM microstructure of the base composite.

Figure 3:

a) Al–Mg–Si liquidus projection [11], b) Al–Mg–Si–Cu partial liquidus projection depicting quaternary liquidus lines [12].

The microstructure of Cu-added composites is shown in Figure 4. It is obvious that the size and area fraction of bright Cu-rich phases increased as the Cu content increased. According to the partial liquidus projection of the Al–Mg–Si–Cu phase diagram (Figure 3b), after Mg₂Si_P particles form, the liquid composition in the Cu-added composites moves towards the ternary eutectic point E1 in which the liquid transforms into blocky-shape θ-Al₂Cu and honeycomb Q-Al₅Mg₈Si₆Cu₂ phases [12]. The detailed morphology of these intermetallics is shown in Figure 5 and the corresponding EDS point analysis could be seen in Table 1.

Figure 4:

Microstructure of Cu-added Al–15Mg₂Si composites, a) Al–15Mg₂Si–1Cu, b) Al–15Mg₂Si–3Cu, c) Al–15Mg₂Si–5Cu.

Figure 5:

SEM micrograph showing Q-phase (Area A) and θ-phase (Area B) in the microstructure of Al–15Mg₂Si composite.

Cu affected the composite’s microstructure by (i) increasing the content of Cu-rich particles (brighter contrast phases in Figure 3), (ii) refining, and (iii) changing the morphology of Mg₂Si particles. The graphical representation of the effects of Cu on the average size, area, and perimeter of the Mg₂Si particles based on the image analysis results of Figure 4 denotes that finer Mg₂Si were obtained by increasing the Cu content (Figure 6a). The fraction of Mg₂Si_P particles also experienced significant growth of about 40 % (Figure 6b).

Figure 6:

Effect of Cu on a) average size and area, and perimeter of Mg₂Si_P particles, b) fraction of Mg₂Si_P and Cu-rich phases and porosity percentage in Al–15Mg₂Si composite.

According to the proposed mechanisms [9], the modification/refinement of Mg₂Si particles in Cu-added composites may be explained by the nucleation of Mg₂Si on the Cu-rich compounds, poisoning the preferred growth planes of Mg₂Si phase by absorption of Cu atoms, or promoting isotropic growth of Mg₂Si particles by the copper-induced multiple twinning mechanism. However, the crystal structure and lattice parameters of Mg₂Si (anti-fluorite face-centered cubic, a = 0.6338 nm) [13] are quite different from those of Cu-rich compounds (θ-phase: tetragonal, a = b = 0.428 nm and c = 0.24 nm) [14] and Q-phase (hexagonal: a = 1.04 nm and c = 0.405 nm) [15]. Moreover, the Cu-rich phases have a lower melting point than the Mg₂Si phase. Therefore, these compounds are unlikely to serve as the heterogeneous nucleation sites for Mg₂Si particles. On the other hand, the ratio of r_Cu/r_Si is 1.23 which is lower than the value of 1.54–1.85 reportedly required for twinning [16]. Therefore, poisoning appears to be the predominant modification mechanism in which the absorption of Cu atoms at the growing interfaces of the Mg₂Si particles changes the interfacial energy and suppresses the anisotropic growth [15, 16].

According to Figure 6b, higher Cu additions caused a higher volume percentage of microporosities to form. The porosity content, in this regard, increased from ∼0.6 vol% in the Cu-free structure to about 1.67, 3.95, and 5.9 vol% in 1, 3, and 5 wt% Cu-containing composites, respectively. Based on the cooling curve analysis of Cu-containing composites (Figure 7a) augmenting the porosity content by Cu addition is seemingly due to the increased freezing range of the composite where its solidus and nucleation points were progressively lowered and raised, respectively (Figure 7b). According to [9], in wide freezing range alloys, at the last stages of solidification, the pockets of molten metal are isolated in the mushy zone and therefore deprived of liquid metal that would compensate for the solidification shrinkage. Under this circumstance, shrinkage micropores are likely to be formed in the microstructure (Figure 4).

Figure 7:

Thermal analysis data of the Al–15Mg₂Si alloy, a) cooling curves, b) freezing range, nucleation, and solidus temperatures at various Cu contents.

In addition, the enrichment of Cu atoms in front of the solidification front enhances the activity coefficient of hydrogen, giving rise to a substantial reduction in its solubility in the molten portion of samples [17, 18] increasing its porosity content.

3.2 Corrosion behavior

The potentiodynamic polarization curves of the base, Al–15Mg₂Si–3Cu, and Al–15Mg₂Si–5Cu samples could be seen in Figure 8. As seen, these alloys showed almost similar behaviors with a small move toward more active potentials. This was ascribed to the more positive potential of Al₂Cu than the surrounding matrix which increases the anodic reactivity [19]. Also, a reduction in the corrosion current density (i_corr) by increasing the Cu content was observed. Both Cu-added alloys showed a very narrow region of almost constant i_corr which may show an unstable passivity. The passive potential was about −0.75 V and −0.72 V and the breakdown potential was about −0.67 V and −0.62 V for Al–15Mg₂Si–3Cu and Al–15Mg₂Si–5Cu composites, respectively. A summary of the corrosion data including E_corr, i_corr, corrosion rate, β_a, and β_c is shown in Table 2. Al–15Mg₂Si–5Cu with an i_corr of 0.164 mA cm⁻² showed the lowest corrosion rate, while the base alloy was the lowest corrosion-resistant alloy with an i_corr of about 0.3 mA cm⁻². These results suggest that adding Cu to the composition of the Al–Mg₂Si alloy improved its corrosion behavior.

Figure 8:

Potentiodynamic polarization curves for the base, Al–15Mg₂Si–3Cu, and Al–15Mg₂Si–5Cu samples in 3.5 wt% NaCl solution.

The Bode and Nyquist diagrams obtained from EIS of the base, Al–15Mg₂Si–3Cu, and Al–15Mg₂Si–5Cu samples after 1 h immersion in a 3.5 wt% NaCl solution and the corresponding equivalent circuit could be seen in Figure 9. The Bode diagram (Figure 9a) showing the modulus of impedance (|Z|) versus frequency could be divided into three regions showing alloys polarization resistance at low frequencies (0.01–1 Hz), double layer behavior at intermediate frequencies (1–100 Hz), and the ohmic resistance of 3.5 wt% NaCl solution at high frequencies (100–10,000 Hz). The higher |Z| value in low frequencies signifies a higher corrosion resistance [20]. Thus, agreeing with polarization results, it is observed that Cu addition considerably enhanced the corrosion resistance of the alloy.

Figure 9:

EIS test results of the base, Al–15Mg₂Si–3Cu, and Al–15Mg₂Si-5Cu samples in 3.5 wt% NaCl solution, a) Bode diagram, b) Nyquist diagram, c) suggested equivalent circuit.

The Nyquist diagrams (Figure 9b) show two distinct arcs (time constants), a smaller one at higher frequencies representing the response of the corrosion products (oxide films) and an arc with a larger diameter at intermediate to low frequencies that shows the charge transfer behavior of the alloy. The larger diameter of the capacitive arcs of the Al–15Mg₂Si–5Cu sample verifies its higher corrosion resistance in accordance with the polarization test results. The EIS data were fitted by the equivalent circuit shown in Figure 9c using the Iviumsoft software simulation. This circuit that fitted well with the impedance data of the present experiments has been widely accepted in the case of Al alloys with and without second-phase particles [20], [21], [22], [23]. Some researchers have proposed more complicated circuits to take into account the role of intermetallics, separately [24]. However, a good fit was not obtained by those circuits in the present study. The relevant EIS parameters are shown in Table 3. In this Table, R _s shows the solution resistance, Q_ox, n₁, and R_ox mean the capacitance of the surface oxide, the dispersion coefficient, and the oxide film resistance, respectively. Q_dl, n₂, and R_ct denote the capacitive response of the double layer of oxide film-electrolyte interface, the dispersion coefficient, and the charge transfer resistance at the oxide-metal interface. The charge transfer resistance (R_ct) describes the electron transfer across the surface which is a measure of the integrity and protectiveness of the surface films and the subsequent corrosion resistance of the alloy [20]. The sum of R_ct and R_ox denotes the corrosion resistance of the alloy [21]. According to Table 3, the R_ox and R_ct for Al–15Mg₂Si–5Cu were higher than those for Al–15Mg₂Si–3Cu and base materials which verify the highest corrosion resistance of Al–15Mg₂Si–5Cu. The base material, on the other hand, showed the lowest corrosion resistance.

Constant phase elements of Q_ox and Q_dl denote a deviation from the ideal capacitor [20] which is generally caused by the presence of surface heterogeneities/discontinuities such as roughness and porosity [25]. Here, Q_ox might be attributed to the discontinuity of the oxide film caused by second phases and porosities which will be discussed later.

SEM micrographs of the corroded surfaces of the base, Al–15Mg₂Si–3Cu, and Al–15Mg₂Si–5Cu samples after 24 h immersion in the 3.5 wt% NaCl solution are shown in Figure 10. The comparison of the corroded surfaces shows that in the base Al–Mg₂Si composite, the Mg₂Si particles were severely attacked (Figure 10a–c) by the solution based on the mechanism describes above and pits were formed in the vicinity of the particles. However, in Cu-containing alloys, the pitting corrosion was seemingly concentrated at the periphery of the Al₂Cu particles (Figure 10f) and no pronounced damage was observed in the matrix around the Mg₂Si particles. This behavior could be explained as follows:

Figure 10:

SEM micrographs of the corroded samples after 24 h immersion in 3.5 wt% NaCl solution, a) Base, b) Base, c) an enlarged view of the Base, d) Al–15Mg₂Si–3Cu, e) Al–15Mg₂Si–5Cu, f) an enlarged view of Al–15Mg₂Si–5Cu.

The formation of various compounds with different electrochemical potentials in the as-cast microstructure of Al alloys makes them quite susceptible to localized corrosion, especially in chloride solutions. Because, in this condition, the surface oxide film cannot provide complete protection [26]. It is well accepted that due to the presence of considerable amounts of Mg in the composition, Mg₂Si is attacked preferentially and acts as an anodic phase when coupled with Al in neutral saline solutions [5, 27, 28]. The dissolution starts with the preferential oxidation of Mg in the compound through the anodic reaction: Mg → Mg²⁺ + 2e⁻ [29]. The loss of Mg leads to the enrichment of the more noble Si in the compound and the subsequent polarity transformation of the remaining Mg₂Si to a cathodic one. Once this happens, the anodic oxidation of the Al matrix starts via the reaction Al → Al³⁺ + 3e⁻ and the release of Al³⁺ to the electrolyte causes Al dissolution/corrosion [30]. A cathodic reaction corresponding to neutral pH values happens on more noble phases as follows: 2H₂O + O₂ + 4e⁻ → 4OH⁻ [29]. The dissolution of the Al surrounding the Si-enriched Mg₂Si particles could form corrosion damages called peripheral pits owing to their morphology (Figure 10a). A bigger cathode size leads to a higher anodic current density and more intense corrosion of the surrounding Al matrix. Chloride ions are reported to disrupt the protective film and promote pit formation in Al alloys. Al³⁺ ions may react with Cl⁻ ions in the solution through the reaction Al³⁺ + Cl⁻ + H₂O → Al(OH)₃ + 3HCl and form aluminum hydroxide where the surface is not protected [26]. Intermetallic particles are sites for the initiation of localized corrosion because of the integrity loss of the oxide film at these sites [31].

The addition of Cu increased the size and fraction of Al₂Cu particles in the alloys (Figure 4). This compound has a more positive potential than the matrix and provokes cathodic reactions and localized corrosion [32, 33]. Also, the microporosity percentage was increased by increasing Cu content (Figure 6). This in turn could have negative effects on the corrosion properties of the alloy probably because of increasing the contact area of the alloy and electrolyte [34, 35]. So, in an Al matrix, the Cu addition seems to be deleterious for the corrosion resistance of the alloy. Conversely, in this study, Cu played a remarkably beneficial role in the corrosion behavior of the alloy. The stability of passive oxide film in Al alloys depends on factors like grain size and distribution of particles [36]. Alloys Al–15Mg₂Si–3Cu and Al–15Mg₂Si–5Cu were characterized by much smaller Mg₂Si particles than the base alloy (Figures 2 and 5). According to Hesami et al. [9], Cu addition also refined the morphology of eutectic Mg₂Si. This is inferred that a finer matrix structure could also be obtained, accordingly. The high reactivity and diffusivity of the grain boundaries promote the diffusion of oxygen and therefore the formation of passive oxide films is expectedly accelerated. Rao et al. [37] suggested that in the case of an Al–30Si alloy, pitting potential decreased, and passivity improved by refining the microstructure and Si particles. The effect of finer second phases on the improvement of the corrosion behavior of Al alloys was also reported by Mitelea et al. [38]. Based on the preceding discussion, faster formation of an alumina film on the surface of the Cu-containing alloys is expected. This mechanism could contribute at least partly to the corrosion resistance of Al–15Mg₂Si–3Cu and Al–15Mg₂Si–5Cu alloys.

It was argued that intermetallic particles act against the formation of a perfect passive film. However, this effect is depressed when the particles are reduced in size and distributed more uniformly. Although a clear explanation was not presented in the literature, it was argued that the propagation of corrosion pits is hampered in such conditions [36, 37]. It seems that in the case of smaller particles, passive film is more integrated which manifested itself in the EIS results (Table 3). It could be stated that in Al–15Mg₂Si–3Cu and Al–15Mg₂Si–5Cu with finer Mg₂Si particles (Figure 5), the specific surface area of the particles in the matrix increases. So, the ratio of the cathodic to anodic surface area is reduced compared to the base alloy. This may slow down the dealloying of Mg from the anodic Mg₂Si. Thus, the later polarity change and pit formation in the Al matrix around Mg₂Si particles are retarded. Therefore, the alloy was less damaged and peripheral pits more or less formed around the more cathodic Cu intermetallics (which are sparse compared with Mg₂Si) as shown in Figure 10d–f. This supports the higher corrosion resistance of Al–15Mg₂Si–3Cu and Al–15Mg₂Si–5Cu observed in the polarization and EIS experiments.

Based on the above discussion, it could be said that the Cu addition on the one hand increases the chance of localized corrosion by the formation of less active Al₂Cu particles and increasing micropores, and on the other hand its effect on the microstructure refinement aids in the rapid formation and healing the passive film. It is anticipated that the latter effect works much better to provide the alloys with higher corrosion resistance than the base alloy.