2019 Volume 60 Issue 9 Pages 1954-1963
A commercial AA7075 (Zn: 5.4 mass%) and experimental Al–Zn–Mg alloys with 8.5 or 10.5 mass% of Zn were used to investigate the influence of Zn in the alloys on the composition of grain boundary precipitates and the susceptibility to the stress corrosion cracking (SCC) of the alloys in NaCl solutions. The stress-strain curves of the alloys under the slow strain rate test (SSRT) demonstrated that the susceptibility of the Al–Zn–Mg alloys to SCC increased as the Zn content increased from 5.4 to 8.5 mass% in the alloys. The susceptibility of the Al–10.5Zn–Mg alloy to SCC was not evaluated because of its brittleness. A coarse mainly Zn-containing precipitate was observed at the grain boundary only in the Al–8.5Zn–Mg alloy by STEM/EDS analysis. Calculated Al–Zn–Mg–Cu–Cr phase diagram predicts that MgZn2 is appeared in the Al–8.5Zn–Mg alloy but not in the Al–5.4Zn–Mg alloy. Anodic polarization measurements of the alloys demonstrated that the dissolution current density increased and the pitting corrosion potential decreased as the Zn content in the alloys increased. The generation of active grain boundary precipitates, such as the mainly Zn-containing precipitates, appears to lead to a higher susceptibility to SCC in Al–Zn–Mg alloys.
Fig. 10 (a) HAADF STEM image of grain boundary precipitates in Al–8.5Zn–Mg alloy. Enlarged (b) HAADF image and (c–e) EDS maps marked by Area 4 in Fig. 10(a).
Because high strength aluminum alloys, such as 7xxx series aluminum alloys based on the Al–Zn–Mg–(Cu) system, tend to suffer from stress corrosion cracking (SCC) in Cl−-containing environments, a much deeper understanding of the causes of SCC is required. Alloying elements added to improve the strength of aluminum alloys generate intermetallic precipitates both in the grains and at the grain boundaries. The SCC of the high strength aluminum alloys is commonly known to be intergranular,1,2) and the susceptibility of such alloys to SCC depends on the grain boundary microstructure and/or microchemistry.2) A considerable amount of information of the influence of the precipitation free zone (PFZ), the size and distribution of precipitates, and the composition of precipitates at the grain boundaries on the SCC of high strength aluminum alloys has been revealed over the course of decades of research work.1–6) Nowadays, the microchemistry at the grain boundaries related with the grain boundary precipitates is recognized as one of the most common causes of SCC in high strength aluminum alloys.5) The electrochemical characteristics of intermetallic precipitates appearing in 7xxx series aluminum alloys and the coupling effects of the intermetallic precipitates with the matrix have been analyzed in detail using microelectrochemical techniques with a glass capillary microcell and a scanning Kelvin prove force microscope (SKPFM).7–11) The noble intermetallic precipitates, which contain Cu, Fe, and/or Zr, have been reported to be cathodic with respect to the matrix, promoting the dissolution of the surrounding matrix. On the other hand, the active intermetallic precipitates, which mainly contain Mg, Zn, and/or Si, have been known to be anodic with respect to the matrix, resulting in the selective dissolution of the precipitates. The composition of the grain boundary precipitates is of key importance in addressing the corrosion problems, such as the SCC, of the high strength aluminum alloys.
7xxx series aluminum alloys contain Zn and Cu as precipitation hardening elements, generating Guinier-Preston (GP) zones and metastable phases in the grains and equilibrium phases at the grain boundaries. The influence of Cu on the corrosion resistance of 7xxx series aluminum alloys has been investigated for Al–Zn–Mg alloys with various Cu contents and under different heat-treatment regimes.5,6,12–16) The concentrations of Cu in equilibrium precipitates, such as Mg(Zn, Al, Cu)2, and the matrix have been demonstrated to change during heat-treatment (aging treatment).12) The Cu content in the precipitates was shown to increase as the solute content of the matrix decreases and the aging temperature increases (to the over-aged state). The benefit of Cu in the matrix is that it reduces pitting corrosion damage in salt-spray environments because of the galvanic relationship between Cu-containing precipitates and the matrix.13) A higher Cu content in the equilibrium MgZn2 precipitates, typically more than 17 at%, has been reported to decrease the activity of the precipitates and increase the pitting corrosion potential of the alloys during polarization measurements.14) It has been proposed that the Cu content of the grain boundary precipitates, such as MgZn(2−x)Cux, is the most important factor determining the resistance to intergranular corrosion (IGC) and SCC of the Al–Zn–Mg alloys.5,6,15) Decreasing the activity of grain boundary precipitates with Cu is likely to effectively reduce the susceptibility to the IGC and SCC of the high strength aluminum alloys. In contrast, the influence of Zn on the corrosion resistance of the Al–Zn–Mg alloys, especially resistance to the SCC, is not well understood. Since not only the Zn content but also the Cu content varies in commercial 7xxx series aluminum alloys, the influence of Zn and Cu on the corrosion resistance of the alloys cannot be discussed independently. In this context, the investigation of Al–Zn–Mg alloys with various Zn contents and a constant Cu content is required to understand the correlation between the amount of Zn content in the alloys, the susceptibility to the SCC of the alloys, and the composition of the grain boundary precipitates.
The susceptibility to the SCC of aluminum alloys has been successfully evaluated by the slow strain rate test (SSRT).17–22) The stress-strain curves under the SSRT carried out at a strain rate of around 1 × 10−6 s−1 measured at open circuit potential in the environments with and without Cl− ions, such as in air and NaCl solutions, have been reported to provide clear indications on the susceptibility of aluminum alloys to SCC. The susceptibility to the SCC of the alloys comes up as the decrease in the values of ultimate tensile strength and/or elongation to failure measured in the environments with Cl− ions compared with those values measured in environments without Cl− ions. The stress-strain curve measurements under the SSRT are considered suitable for the initial evaluation of the susceptibility of the Al–Zn–Mg alloys with various Zn contents and a constant Cu content to SCC.
The purpose of this study was to investigate the influence of Zn content in Al–Zn–Mg alloys on the composition of grain boundary precipitates and the susceptibility of the alloys to SCC. Three types of aluminum alloys with different Zn content were used in this study. The microstructural characterizations of the grain size and intermetallic precipitates of the three aluminum alloys were carried out. Stress-strain curves under the SSRT were measured to evaluate the susceptibility of these alloys to SCC. Besides, anodic polarization measurements were taken to clarify the influence of Zn content on electrochemical properties of the alloys.
A commercial AA7075 aluminum alloy (Zn: 5.4 mass%) bar with a diameter of 200 mm and two ingots of Al–Zn–Mg alloys (Zn: 8.5 or 10.5 mass%) with a weight of 7 kg produced by vacuum induction melting were used in this study. The three aluminum alloys are referred to as Al–5.4Zn–Mg, Al–8.5Zn–Mg, and Al–10.5Zn–Mg alloys according to their Zn content in this paper. The chemical compositions of the three aluminum alloys are listed in Table 1. The amount of Fe content in the Al–5.4Zn–Mg alloy was larger than that in the Al–8.5Zn–Mg and Al–10.5Zn–Mg alloys. The two ingots of the Al–8.5Zn–Mg and Al–10.5Zn–Mg alloys were heat-treated at 733 K for 24 h (homogenization treatment), and then were forged and rolled at 703 K to a thickness of 15 mm. The three aluminum alloys were heat-treated at 753 K for 2 h and quenched in water (solution treatment), and then were heat-treated at 393 K for 24 h and quenched in water (aging treatment).
The specimen surfaces were etched with NaOH solution to reveal the microstructures of the three aluminum alloys. The chemical compositions of micrometer-sized intermetallic compounds in the three aluminum alloys were analyzed using FEI Quanta 250 FEG field emission scanning electron microscope (FE-SEM) equipped with AMETEX EDAX Octane Elite energy-dispersive X-ray spectroscopy (EDS) system at an electron accelerating voltage of 20 kV. The chemical compositions of sub-nanometer-sized intermetallic compounds in the three aluminum alloys were obtained by JEOL JEM-2100F scanning transmission electron microscopy (STEM) equipped with an EDS system at an electron accelerating voltage of 200 kV. The samples for the STEM/EDS analysis were prepared with Hitachi NB5000 focused ion beam (FIB) instrument by milling the specimen surfaces with a gallium ion beam. The thickness of the STEM/EDS samples was about 100 nm.
2.3 Stress corrosion cracking susceptibilityThe susceptibility of the three aluminum alloys to stress corrosion cracking (SCC) was evaluated by the slow strain rate test (SSRT). The geometry of the round bar type tensile specimens used for the SSRT is shown in Fig. 1. The long axis of the SSRT specimens was parallel to the rolling direction. The gauge dimensions of the SSRT specimens were 21 mm in length and 3.0 mm (for the Al–5.4Zn–Mg and Al–8.5Zn–Mg alloys) or 2.8 mm (for the Al–10.5Zn–Mg alloy) in diameter. The gauge area was ground with SiC paper of P-4000, and then the surface of the SSRT specimens with the exception of the gauge area was covered by a silicone resin (TSE3941, Momentive Performance Materials Japan). The SSRT was carried out at the strain rate of 6.0 × 10−6 s−1 (stroke rate: 7.6 µm min−1). The stress-strain curves of the three aluminum alloys under the SSRT were measured at open circuit potential in air, deionized water (pH 6.9), 0.1 M Na2SO4 (pH 5.6), 0.1 M NaCl (pH 5.3), and 1 M NaCl (pH 5.1) solutions at 298 K.
Geometry of the round bar type tensile specimen used for SSRT.
Anodic polarization curves of the three aluminum alloys were measured using an IVIUM pocketSTAT potentiostat. The three aluminum alloys were cut into ca. 25 mm × 15 mm coupons parallel to the rolling direction. The specimen surfaces were ground with SiC papers and cleaned with ethanol. With the exception of the electrode area (ca. 10 mm × 10 mm), the surface of the specimens was coated with an epoxy resin (AR-R30, Nichiban) and subsequently with paraffin. The measurements were performed in a conventional three electrode cell; the counter electrode was a platinum plate and the reference electrode was an Ag/AgCl (3.33 M KCl) electrode. All potentials cited in this paper refer to the Ag/AgCl (3.33 M KCl) electrode (0.206 V vs. standard hydrogen electrode at 298 K). The working electrode potential was scanned at a constant rate of 0.4 mV s−1. The size of the electrode areas was scaled accurately to convert the measured current value into current density after the measurements. The anodic polarization curves ware measured in 0.1 M Na2SO4 (pH 5.6), 0.1 M NaCl (pH 5.3), and 1 M NaCl (pH 5.1) solutions at 298 K under naturally aerated conditions. All the solutions were prepared from deionized water and analytical grade chemicals.
Figure 2 shows the optical microscope images of the three aluminum alloys etched with NaOH solution. The Al–5.4Zn–Mg and Al–8.5Zn–Mg alloys had a randomly mixed structure of fine and coarse grains. The average grain size of the Al–8.5Zn–Mg alloy (about 9 µm) was larger than that of the Al–5.4Zn–Mg alloy (about 6 µm). The grain size of the Al–10.5Zn–Mg alloy was around 20 µm and was coarser than that of the Al–5.4Zn–Mg and Al–8.5Zn–Mg alloys. Many micro-cracks were observed at the grain boundaries in the Al–10.5Zn–Mg alloy, which were likely induced during the forging and rolling processes.
Optical microscope images of (a) Al–5.4Zn–Mg, (b) Al–8.5Zn–Mg, and (c) Al–10.5Zn–Mg alloys etched with NaOH solution.
Figure 3 exhibits the optical microscope images of the three mirror polished aluminum alloys. Dot-like areas, a darker gray in appearance than the matrix, were dispersed in the three aluminum alloys. There were more darker gray dot-like areas in the Al–5.4Zn–Mg alloy than in the Al–8.5Zn–Mg and Al–10.5Zn–Mg alloys: the incidence of these areas appears to depend on the amount of Fe content in the aluminum alloys. Figures 4–6 display the SEM images and the corresponding EDS maps of the areas enclosed by the broken lines in Fig. 3. The relative compositions of Fe, Cu, Zn, Mg, Cr, Mn, Si, Ti, and Al, as determined by the EDS point analysis at the sites marked by Points 1–6 in Figs. 4–6, are listed in Table 2. The main composition detected on the Points 1–6 was Al, Fe, and Cu, indicating that the darker gray dot-like areas shown in Fig. 3 were Fe-containing intermetallic compounds.
Optical microscope images of mirror polished (a) Al–5.4Zn–Mg, (b) Al–8.5Zn–Mg, and (c) Al–10.5Zn–Mg alloys.
SEM backscattered electron (B. S. E) image and EDS maps of Al–5.4Zn–Mg alloy marked by the broken line in Fig. 3(a).
SEM backscattered electron (B. S. E) image and EDS maps of Al–8.5Zn–Mg alloy marked by the broken line in Fig. 3(b).
SEM backscattered electron (B. S. E) image and EDS maps of Al–10.5Zn–Mg alloy marked by the broken line in Fig. 3(c).
Figure 7 shows the stress-strain curves of the three aluminum alloys obtained during the SSRT. The stress-strain curves of the aluminum alloys under the SSRT were measured at open circuit potential in the five environments: air, deionized water, 0.1 M Na2SO4, 0.1 M NaCl, and 1 M NaCl solutions. It was considered that the environments without Cl− ions, such as air, deionized water, and 0.1 M Na2SO4 solution, would not result in the SCC in the aluminum alloys, and the environments with Cl− ions, such as 0.1 M and 1 M NaCl solutions, would result in the SCC in the aluminum alloys in the experimental conditions. The SCC susceptibilities of the aluminum alloys can be evaluated by comparing the values of ultimate tensile strength and elongation to failure measured in the environments with and without Cl− ions. The loss in the ultimate tensile strength and/or elongation to failure caused by the environments with Cl− ions is used as an index of the susceptibility to SCC. Table 3 displays the values of ultimate tensile strength and elongation to failure of the three aluminum alloys. The Al–5.4Zn–Mg alloy reached almost the same value of ultimate tensile strength in the five environments. The average value of elongation to failure of the Al–5.4Zn–Mg alloy measured in 0.1 M and 1 M NaCl solutions decreased to about 95% of that measured in air, deionized water, and 0.1 M Na2SO4 solution. The about 5% decrease in elongation to failure of the Al–5.4Zn–Mg alloy was attributed to the SCC caused by the environments with Cl− ions. The Al–8.5Zn–Mg alloy had almost the same value of ultimate tensile strength of the five environments; however, the average value of elongation to failure measured in 0.1 M and 1 M NaCl solutions decreased to about 70% of that measured in air, deionized water, and 0.1 M Na2SO4 solution. The about 30% decrease in elongation to failure of the Al–8.5Zn–Mg alloy, which was attributed to the SCC caused by the environments with Cl− ions, indicates the higher susceptibility of the Al–8.5Zn–Mg alloy to SCC than that of the Al–5.4Zn–Mg alloy. Almost no plastic deformation of the Al–10.5Zn–Mg alloy was observed in any of the five environments. The susceptibility of the Al–10.5Zn–Mg alloy to SCC was not evaluated with the values of elongation to failure because of the brittleness of the alloy. From the stress-strain curves measured during the SSRT, it is clear that the susceptibility of the Al–Zn–Mg alloys to SCC increased (the resistance to the SCC of the alloys decreased) with increasing Zn content from 5.4 mass% to 8.5 mass% in the alloys.
Stress-strain curves of (a) Al–5.4Zn–Mg, (b) Al–8.5Zn–Mg, and (c) Al–10.5Zn–Mg alloys during SSRT (strain rate: 6.0 × 10−6 s−1) measured at the open circuit potential in air, deionized water, 0.1 M Na2SO4, 0.1 M NaCl, and 1 M NaCl solutions.
Figure 8 represents the {2.85Mg–1.6Cu–0.2Cr}-section of the Al–Zn–Mg–Cu–Cr alloy calculated using open Al-based alloy database23) on PANDAT in order to clarify the difference in microstructure between the Al–5.4Zn–Mg and Al–8.5Zn–Mg alloys. Because η′-MgZn2 phase, which is known to distribute in grains as a hardening precipitate,24,25) is metastable, it is not appeared in the equilibrium phase diagram. The Al–5.4Zn–Mg alloy was expected to form T1 (Mg3Cr2Al18) and T2 (CuMgZnAl) phases. In addition to these two kinds of phases, an equilibrium MgZn2 phase, which differs from the metastable η′-MgZn2 phase appearing in grains, was expected to be formed in the Al–8.5Zn–Mg and Al–10.5Zn–Mg alloys. The difference between the Al–5.4Zn–Mg and Al–8.5Zn–Mg alloys was likely to be the formation of equilibrium MgZn2 or not. The SCC in high strength aluminum alloys is characteristically intergranular. Since precipitates mainly containing Zn were not found by the FE-SEM/EDS analysis focusing on micrometer-sized intermetallic compounds, a STEM/EDS analysis was carried out focusing on sub-nanometer-sized grain boundary precipitates.
Calculated phase diagram of Al–Zn–2.85Mg–1.6Cu–0.2Cr (mass%).
Figures 9–11 show the high-angle annular dark field (HAADF) STEM images and corresponding EDS maps of the areas marked by Areas 1, 2, 4, and 6 in Figs. 9–11. The relative compositions of Cr, Cu, Zn, Mg, Fe, Mn, Si, Ti, and Al, as determined by the EDS point analysis at the sites marked by Points 7–13 in Figs. 9–11, are listed in Table 4. The main composition of the grain boundary precipitates found in the Al–5.4Zn–Mg alloy was Cr or Cu. No precipitate mainly containing Zn were detected in the Al–5.4Zn–Mg alloy. The main composition of the grain boundary precipitates found in the Al–8.5Zn–Mg alloy was Cr, Cu, or Zn. A relatively coarse Zn-containing precipitate, which contained 91 at% of Zn at Point 11 (Table 4), was detected at the grain boundary in the Al–8.5Zn–Mg alloy. The Al–10.5Zn–Mg alloy formed precipitates continuously lying along the grain boundaries. The grain boundary precipitates mainly contained Cr, Cu, Zn, and Mg. The precipitates generated at the grain boundaries in the three aluminum alloys detected by the STEM/EDS analysis were as follows: Cr-containing and Cu-containing precipitates in the Al–5.4Zn–Mg alloy, Zn-containing, Cr-containing, and Cu-containing precipitates in the Al–8.5Zn–Mg alloy, precipitates containing Cr, Cu, Zn, and Mg in the Al–10.5Zn–Mg alloy. The precipitates found by the STEM/EDS analysis corresponded to the equilibrium phases described in the phase diagram (Fig. 8). The difference between the Al–5.4Zn–Mg and Al–8.5Zn–Mg alloys demonstrated here was the precipitate mainly containing Zn at the grain boundaries in the Al–8.5Zn–Mg alloy. The generation of the mainly Zn-containing grain boundary precipitates is expected to result in the higher susceptibility of the Al–8.5Zn–Mg alloy to SCC than that of the Al–5.4Zn–Mg alloy.
(a) HAADF STEM image of grain boundary precipitates in Al–5.4Zn–Mg alloy. Enlarged HAADF images and EDS maps marked by (b–e) Area 1 and (f–i) Area 2 in Fig. 9(a).
(a) HAADF STEM image of grain boundary precipitates in Al–8.5Zn–Mg alloy. Enlarged (b) HAADF image and (c–e) EDS maps marked by Area 4 in Fig. 10(a).
(a) HAADF STEM image of grain boundary precipitates in Al–10.5Zn–Mg alloy. Enlarged (b) HAADF image and (c–e) EDS maps marked by Area 6 in Fig. 11(a).
It is possible that the increase in Zn content from 5.4 mass% to 8.5 or 10.5 mass% in the Al–Zn–Mg alloys results in an increase in the amount of the metastable η′-MgZn2 phases appearing in grains. The fine precipitates of the η′-MgZn2 phase dispersed in grains are known as a hardening precipitate: these contribute to the improved strength of Al–Zn–Mg alloys. Figure 12 describes the engineering stress-strain curves of the Al–5.4Zn–Mg, Al–8.5Zn–Mg, and Al–10.5Zn–Mg alloys. The geometry of the samples used for the tensile test was the same as that used for the SSRT (Fig. 1). The tensile tests were carried out with a video extensometer at 298 K in air at the strain rate of 6.0 × 10−6 s−1. The values of ultimate tensile strength and elongation to failure of the Al–5.4Zn–Mg alloy were 605 MPa and 12.3%, respectively. The values of ultimate tensile strength and elongation to failure of the Al–8.5Zn–Mg alloy were 689 MPa and 11.1%, respectively, resulting in about 80 MPa increase in the ultimate tensile strength and only about 1% decrease in the elongation to failure compared with those of the Al–5.4Zn–Mg alloy. The elongation to failure of the Al–10.5Zn–Mg alloy drastically decreased. The values of ultimate tensile strength and elongation to failure of the Al–10.5Zn–Mg alloy were 632 MPa and 1.8%, respectively. The Al–Zn–Mg alloy became brittle with 10.5 mass% of Zn content.
Engineering stress-strain curves of Al–5.4Zn–Mg, Al–8.5Zn–Mg, and Al–10.5Zn–Mg alloys. Sample deformation was measured by a video extensometer (strain rate: 6.0 × 10−6 s−1).
Figure 13 exhibits the HAADF STEM images and EDS line analysis (Zn intensity) of the areas marked by Areas 3, 5, and 7 in Figs. 9–11. A lot of nanometer-sized particles were observed in the grains of the three aluminum alloys. The EDS line analysis demonstrated the increase in Zn intensity measured on the nanometer-sized particles, implying that the nanometer-sized particles were the η′-MgZn2 hardening phase. The density of the η′-MgZn2 nanometer-sized particles in the Al–8.5Zn–Mg alloy was higher than that in the Al–5.4Zn–Mg alloy, which corresponded to the result of the about 80 MPa improved ultimate tensile strength of the Al–8.5Zn–Mg alloy compared with the Al–5.4Zn–Mg alloy (Fig. 12). The density of the η′-MgZn2 nanometer-seized particles in the Al–10.5Zn–Mg alloy seemed to be in the same range of that in the Al–5.4Zn–Mg alloy. From the results and the analysis of the Al–5.4Zn–Mg and Al–8.5Zn–Mg alloys, it is possible that Zn alloying, which prevents mainly Zn-containing precipitates from forming at the grain boundaries and generates high-density η′-MgZn2 nanometer-sized particles in the grains, improves the strength of the Al–Zn–Mg alloys and inhibits the susceptibility of the alloys to SCC.
HAADF STEM images and EDS line analysis (Zn intensity) of fine precipitates in grains: (a, b) Al–5.4Zn–Mg alloy marked by Area 3 in Fig. 9(a), (c, d) Al–8.5Zn–Mg alloy marked by Area 5 in Fig. 10(a), and (e, f) Al–10.5Zn–Mg alloy marked by Area 7 in Fig. 11(a). Broken lines indicate the locations of the EDS line analysis.
Figure 14 shows the anodic polarization curves of the three aluminum alloys measured with ca. 10 mm × 10 mm electrode area. The three aluminum alloys were passivated in 0.1 M Na2SO4 solution (Fig. 14(a)). As the amount of Zn content in the alloys increased, the corrosion potentials tended to be less noble, and the dissolution current densities increased. The dissolution current densities of the alloys at 0.5 V measured in 0.1 M Na2SO4 solution were as follows: ca. 0.015 A m−2 for the Al–5.4Zn–Mg alloy, ca. 0.04 A m−2 for the Al–8.5Zn–Mg alloy, and ca. 0.3 A m−2 for the Al–10.5Zn–Mg alloy. Since the formation of grain boundary precipitates likely meant that the grain boundaries of the Al–Zn–Mg alloys were more difficult to be passivated than the grains, the anodic polarization curves were considered to reflect the electrochemical property of the grain boundaries more strongly than that of the grains. The results measured in 0.1 M Na2SO4 solution, therefore, seems to suggest that the Al–Zn–Mg alloys generated grain boundaries which dissolve more easily as the amount of Zn content increased in the alloys. Figures 14(b) and 14(c) show the anodic polarization curves of the three aluminum alloys in 0.1 M and 1 M NaCl solutions, respectively. A peak in the dissolution current density was observed during the anodic polarization with the three aluminum alloys. Many researches have reported that the anodic polarization behavior is attributed to the breakdown and the dissolution of an active surface layer formed during surface preparation by polishing with grinding papers.26–30) The pitting corrosion potentials of the three aluminum alloys measured in 0.1 M and 1 M NaCl solutions, defined as the potential at which the current density exceeds 1 A m−2, became less noble as the Zn content in the alloys increased. The difference in the pitting corrosion potentials between the Al–5.4Zn–Mg and Al–8.5Zn–Mg alloys or the Al–8.5Zn–Mg and Al–10.5Zn–Mg alloys measured in 0.1 M or 1 M NaCl solutions was ca. 30 mV. This result indicates that the pitting corrosion resistance of the Al–Zn–Mg alloys in NaCl solutions decreased as the Zn content in the alloys increased. The anodic polarization measurements demonstrated the differences in the macroscopic polarization behavior between the three aluminum alloys with different Zn content; the grain boundaries of the alloys seemed harder to be passivated, and the pitting corrosion occurred more easily in Cl−-containing environments with as the Zn content in the alloys increased. These results are further evidence that the grain boundary precipitates differ depending on the amount of Zn content in the alloys as was revealed by the STEM/EDS analysis applied to the quite limited areas (Fig. 10). These results lead to the conclusion that the higher susceptibility of the Al–8.5Zn–Mg alloy to SCC than that of the Al–5.4Zn–Mg alloy (Fig. 7) was due to the formation of the mainly Zn-containing precipitates at the grain boundaries, such as the precipitates observed by the STEM/EDS analysis (Fig. 10) and as predicted by the phase diagram, as MgZn2 (Fig. 8). Knight et al. demonstrated that the susceptibility to SCC appeared to be mainly controlled by the Cu content in MgZn(2−x)Cux grain boundary precipitates with AA7075 and AA7079 alloys.5) The authors proposed that higher Cu content in the MgZn(2−x)Cux grain boundary precipitates is beneficial to increase the resistance against the SCC of the alloys. This result implies that precipitates with a higher Zn content have a negative effect on the susceptibility of aluminum alloys to SCC: the finding in the present study is in good agreement with this.
Anodic polarization curves of Al–5.4Zn–Mg, Al–8.5Zn–Mg, and Al–10.5Zn–Mg alloys measured in (a) 0.1 M Na2SO4, (b) 0.1 M NaCl, and (c) 1 M NaCl solutions.
Figure 15 displays the potential-pH diagrams for Al–H2O, Cu–H2O, Cr–H2O, and Zn–H2O systems in order to discuss the role of mainly Zn-containing precipitates generated at the grain boundaries in the susceptibility of the Al–Zn–Mg alloys to SCC. The standard chemical potentials of the species were obtained from the HSC Chemistry thermochemical database.31) The concentrations of soluble species were set to 1 × 10−5 mol/kg (H2O). The range of the corrosion potentials collected during the anodic polarization measurements shown in Fig. 14 and the range of the solution pH values used to measure the anodic polarization curves were marked in the potential-pH diagrams. Aluminum is predicted to form an Al2O3 passivation film in the marked range of the corrosion potentials and the solution pHs (Fig. 15(a)). Copper, which has been reported to make grain boundary precipitates less active,7,14) is considered to be stable as metallic copper in the marked range (Fig. 15(b)). Chromium, which is known to improve the corrosion resistance of Al–Zn–Mg alloys,32,33) is believed to form a Cr2O3 passivation film in the marked range (Fig. 15(c)). The Al–8.5Zn–Mg alloy formed the coarse Zn-containing precipitate at the grain boundary (Fig. 10), resulting in a higher susceptibility to SCC (Fig. 7). The potential-pH diagram for the Zn–H2O system indicates that metallic zinc is predicted to dissolve into Zn2+ ions in the marked range (Fig. 15(d)). Birbilis and Buchheit measured the corrosion potentials, the pitting corrosion potentials, and the anodic polarization curves of Zn-containing precipitates, such as Al32Zn49 and MgZn2, using microelectrochemical techniques with a glass capillary microcell, indicating these Zn-containing precipitates dissolved freely above their corrosion potentials, whereas the aluminum matrix was passivated.7,8) Wloka and Virtanen explained that the equilibrium phase of MgZn2 grain boundary precipitates were electrochemically active, and proposed that the dissolution of the MgZn2 precipitates leads to dissolution in the depth direction at the grain boundaries.33,34) According to the results in the present study and these findings in the literature, the higher susceptibility to the SCC induced by the mainly Zn-containing precipitates at the grain boundaries of the Al–8.5Zn–Mg alloy in Cl−-containing environments is assumed to be caused by the following two steps: 1) the Zn-containing precipitates selectively dissolve, resulting in the exposure of the bare aluminum matrix at the grain boundaries; 2) Cl− ions prevent the passivation film from being formed on the exposed bare aluminum matrix, leading to the continuous dissolution of the aluminum matrix along the grain boundaries. Finally, the alloy fails as the values of ultimate tensile strength and/or elongation to failure decrease and falls below the values measured in environments without Cl− ions. By avoiding the formation of grain boundary precipitates which dissolve selectively in the environments leading the passivation of aluminum matrix, the susceptibility of Al–Zn–Mg alloys to SCC can likely be reduced.
Potential-pH diagrams for (a) Al–H2O, (b) Cu–H2O, (c) Cr–H2O, and (b) Zn–H2O systems at 298 K (concentrations of soluble species: 1 × 10−5 mol/kg (H2O)). The range of corrosion potential obtained during the anodic polarization shown in Fig. 14 and the range of solution pH value used to measure the anodic polarization curves is marked in the diagrams.
The susceptibility of the Al–Zn–Mg alloys to SCC increased, or, in other words, the resistance of the alloys to SCC decreased, with increasing Zn content from 5.4 mass% to 8.5 mass% in the alloys. The stress-strain curves of the Al–Zn–Mg alloys under the SSRT demonstrated that the average value of elongation to failure of the Al–8.5Zn–Mg alloy measured in the environments with Cl− ions decreased by about 30% compared to that measured in the environments without Cl− ions, whereas the average value of elongation to failure of the Al–5.4Zn–Mg alloy measured in the environments with Cl− ions decreased by only about 5% compared to that measured in the environments without Cl− ions. The susceptibility of the Al–10.5Zn–Mg alloy to SCC was not evaluated because of the brittleness of the alloy.
The relatively coarse Zn-containing precipitate was observed at the grain boundary of the Al–8.5Zn–Mg alloy by the STEM/EDS analysis. The Al–5.4Zn–Mg alloy formed Cr-containing and Cu-containing precipitates at the grain boundaries. In addition to these two kinds of precipitates, the Al–8.5Zn–Mg alloy generated a mainly Zn-containing precipitate at the grain boundary. The grain boundary precipitates found by the STEM/EDS analysis corresponded to the equilibrium phases described in the calculated phase diagram of the Al–Zn–2.85Mg–1.6Cu–0.2Cr system.
The increase in Zn content from 5.4 mass% to 8.5 mass% in the Al–Zn–Mg alloys increased the amount of the metastable η′-MgZn2 nanometer-sized particles in the grains, which are known as a hardening precipitate and to contribute to the improvement of the strength of the alloys. The value of ultimate tensile strength of the Al–8.5Zn–Mg alloy obtained from the engineering stress-strain curves measured during the tensile test was about 80 MPa higher than that of the Al–5.4Zn–Mg alloy.
The anodic polarization measurements in 0.1 M Na2SO4 solution demonstrated that the corrosion potentials tended to be less noble, and the dissolution current densities increased as the amount of Zn content in the Al–Zn–Mg alloys increased. These results suggest that the grain boundaries of the alloys were harder to be passivated as the Zn content in the alloys increased. The pitting corrosion potentials of the Al–Zn–Mg alloys measured in 0.1 M and 1 M NaCl solutions became less noble as the Zn content in the alloys increased, indicating that the pitting corrosion occurred more easily on the alloys with a higher Zn content in Cl−-containing environments.
The higher susceptibility of the Al–8.5Zn–Mg alloy to SCC than that of the Al–5.4Zn–Mg alloy is considered to result from the formation of Zn-containing precipitates at the grain boundaries, such as the precipitates observed by the STEM/EDS analysis and predicted by the phase diagram. It is assumed that the dissolution of the Zn-containing precipitates results in the exposure of the bare aluminum matrix at the grain boundaries, leading the continuous dissolution of the aluminum matrix along the grain boundaries.
A part of this work was supported by NIMS TEM Station.
