1 1 INTRODUCTION

Steel is widely used in building, automobiles, appliances and construction industry due to its high strength, good processing performance and low production cost [1–3]. However, steel is prone to corrosion, which has brought great losses to the national economy [4]. Hot-dip zinc coating can provide not only the barrier protection but also the galvanic protection to the steel substrate. With the further development of modern industry, traditional hot-dip galvanizing can no longer meet the demand for corrosion resistance [5, 6]. Recently, researchers have found that Zn–Al–Mg alloys have better corrosion resistance than traditional zinc coatings and Zn-Al alloys, so Zn–Al–Mg alloys have been extensively studied [7, 8]. For example, the corrosion resistance of Zn–6Al–3Mg alloy produced by Nisshin Steel Group is nearly 10 times higher than that of traditional galvanized sheet, and it is widely used in construction, automobile industries and so on [9, 10].

By adding other elements (such as Si, Ti, Zr, Re, etc.) to further improve the corrosion resistance of Zn–Al–Mg alloys. Li et al. [11] analyzed the effects of Re and Si on the microstructure and corrosion resistance of Zn–6Al–3Mg hot dip coating. The results indicated that the additions of Si and Re suppress the intergranular corrosion and pitting that occurred in the Zn–6Al–3Mg coating. Thereby, the corrosion resistance of Zn–6Al–3Mg alloy is improved. Zn–11Al–3Mg–0.2Si alloy designed by Nippon Steel has a series of advantages of super corrosion resistance, excellent black denaturation resistance and good formability. It is mainly used in household appliance industry [12, 13]. In addition, by adding proper amounts of Ti to the Zn–Al–Mg alloy, Ti can be used as heterogeneous nucleation particles to refine the alloy structure. The addition of Ti form a stable, strong bonding, protective and dense oxide film on the surface of the alloy coating to improve the corrosion resistance of the alloy [14, 15]. Xu et al. [16] studied the microstructure and properties of a new Al–Cu–Mg–Zn–Zr–Ti heat resistant aluminum alloy. It was found that the alloy with Ti added has good resistance to intergranular corrosion. Liu et al. [17] analyzed the effect of Ti addition on microstructure and corrosion property of Zn–5Al alloy. The results showed that a small amount of Ti addition in Zn–5Al alloy refines the primary Zn phase and increases the percentage of the eutectic structure, and finally improves the corrosion resistance of Zn–5Al alloy. Moreover, researchers have also found that adding an appropriate amount of Zr can improve the anti-corrosion characters of the alloy. Chai et al. [18] studied the effect of minor Zr addition on exfoliation corrosion resistance Al–Zn–Mg–Mn alloy sheet. The results showed that the exfoliation corrosion resistance of Al–Zn–Mg–Mn alloy sheet was significantly improved due to the addition of Zr. Zhou et al. [19] studied the effect of Zr content on microstructure and corrosion resistance of hot-dip galvanized Zn–0.1% Ni–Zr alloy. It can be found that the electrochemical impedance and polarization resistance of Zn–0.1% Ni–Zr alloy are increased, and then improve the corrosion resistance of the alloy.

Moreover, in the previous works, most researchers paid more attention to the effect of single element addition on the corrosion resistance of Zn–5Al–2Mg alloy. Few people have studied adding two elements with equal proportions to Zn–Al–Mg alloy. In this work, the effect of (Si, Ti), (Si, Zr) and (Ti, Zr) elements with equal proportions on the microstructures and corrosion resistance of Zn–5Al–2Mg alloy was studied.

2 2 EXPERIMENTAL

The alloy raw materials used in the experiment were Zn (99.99 wt %), pure Mg (99.95 wt %), pure Al (99.95 wt %) and Al–20Si, Al–10Zr, Mg–30Zr and Al–5Ti master alloys. Zn–5Al–2Mg–x(SiTi), Zn–5Al–2Mg-y(SiZr) and Zn–5Al–2Mg-z(TiZr) (x = y = z = 0, 0.1, 0.2) (wt %, same below) alloy were melted within a crucible in resistance furnace. Set the temperature of the resistance furnace to 720°C, added the treated alloy ingot, heated them to complete melting. Then set the furnace temperature to 550°C and kept at this temperature for a certain time. Finally, stirred evenly, and cast into the mold under atmospheric conditions to obtain the alloy ingot required for the experiment. The alloy elemental composition was shown in Table 1.

Table 1.   The elemental composition of alloys
Full size table

Alloy microstructures were studied after grinding and polishing by scanning electron microscopy (SEM; FEG-250) with back scattered electron (BSE) imaging. While the phase composition was analysed by X‑ray diffraction spectrometry (XRD; D8-advanY) and energy dispersive X-ray spectrometry (EDS), respectively. The ingot was cut into thin pieces, 0.3 mm thick, by electrospark. Sample preparations for TEM (JEM-2100) with EDS observation included mechanical polishing and ion-beam thinning (GATAN-691). In addition, the hardness of alloy samples was measured by the Rockwell hardness tester (HBRV-187.5).

All the electrochemical tests were carried out using IM6d Zahner-Elektrik workstation at room temperature. A standard three-electrode system was used to test samples. Standard saturated calomel electrode (SCE) was used as reference electrode, metal platinum electrode as the counter electrode and the sample with a test surface of 1 cm2 as the working electrode. The samples were subjected to dynamic polarization after stabilizing the open circuit potential (OCP) for 30 min in all the experiments. Electrochemical impedance spectra (EIS) were performed under the sinusoidal potential perturbation of ±10 mV after stabilizing the open circuit potential in the frequency range of 10–1 ~ 105 Hz. Polarization curves were recorded at the voltage range of –2 to 0 V and the scanning rate of the polarization curve was 10 mV/s. The corroding medium of all electrochemical tests was 3.5% NaCl solution. Besides, in order to further test the corrosion resistance of alloy samples, total immersion weight loss method was carried out. The weight loss of the samples was gravimetrically calculated by measuring the weight difference before and after removing the corrosion products by chemical cleaning in NH4Cl solution according to GB/T16545-1996. Immersed the test surface with 1 cm2 in 3.5% NaCl solution, every three days as a cycle, 7 cycles in total. The corrosion products of the alloy samples were analyzed by X‑ray photoelectron spectroscopy (XPS) of Thermo Fisher Scientific ESCALAB Xi+.

3 3 RESULTS AND DISCUSSION

3.1 3.1 Microstructure and Phase Constituent of Zn–Al–Mg Alloys

In order to identify the phase composition of alloy samples, XRD of four Zn–Al–Mg alloys were tested. The results were presented in Fig. 1. As can be seen from the results, that Zn, Al and MgZn2 phases were identified in all alloys. However, the diffraction peaks of Si-rich phase, Al3Ti phase and Al3Zr phase appeared in the alloy samples, by adding Si, Ti and Zr in binary and equal proportion, respectively [17, 18]. In addition, it is not enough to meet the conditions of phase quantitative analysis by XRD because the sample was as-casted and had a strong preferred orientation.

Fig. 1.
figure 1

XRD spectrum of Zn–5Al–2Mg and Zn–5Al–2Mg–0.2(Si/Ti/Zr) alloys.

Full size image

The surface microstructure and phase composition of the alloys were analysed using SEM and EDS. As showed in Fig. 2, it is clearly observed that the typical structure of Zn–5Al–2Mg alloy consisted of the black Al phase, the white Zn phase attached around the Al phase dendrite, and lamellar Zn/Al/MgZn2 ternary eutectic structure. In Fig. 3, the microstructure of six alloys (Zn–5Al–2Mg–0.1/0.2(SiTi), Zn–5Al–2Mg–0.1/0.2(SiZr) and Zn–5Al–2Mg–0.1/0.2(TiZr)) was studied. Firstly, as indicated in Figs. 3a, 3b, the Al dendrite arm was greatly short and Zn dendrite was slightly increased by adding Si and Ti elements. In addition, it can be found that when the addition of (Si, Ti) was 0.2 wt %, the lamellar spacing of ternary eutectic structure increased obviously. Secondly, as can be seen in Figs. 3c, 3d, the size of Al dendrites had not obviously changed with the addition of (Si, Zr) element. However, the percentage of Zn rich phase increased and the lamellar spacing of ternary eutectic structure also increased with the increasing of (Si, Zr) element. Finally, it is indicated that Al and Zn dendrites decreased with the increasing of (Ti, Zr) element addition in Figs. 3e, 3f. Nevertheless, compared with that of Zn–5Al–2Mg alloy, the lamellar spacing of ternary eutectic structure of Zn–5Al–2Mg–0.1/0.2(TiZr) alloys first increased and then decreased with the increase of addition. In particular, it is worth noticing that the Zn–5Al–2Mg–0.2(TiZr) alloy had the shortest dendrite arm and the largest percentage of ternary eutectic structure among them.

Fig. 2.
figure 2

The SEM image of Zn–5Al–2Mg alloy.

Full size image
Fig. 3.
figure 3

SEM images of Zn–5Al–2Mg–x(SiTi), Zn–5Al–2Mg–y(SiZr) and Zn–5Al–2Mg–z(TiZr) alloys: (a, c,e ,) x = y = z = 0.1; (b, d, f) x = y = z = 0.2.

Full size image

Zn–5Al–2Mg–0.2(SiTi), Zn–5Al–2Mg–0.2(SiZr) and Zn–5Al–2Mg–0.2(TiZr) alloys were selected to test EDS. The selected points were marked in Fig. 4, and the results of point analysis were listed in Table 2. According to the results of point analysis, it can be found that the percentages of Si atoms at points 1 and 4 were 93.17 and 81.47%, respectively, which corresponded to the Si rich phase. However, Point 2 was located at the black block phase in Fig. 4a, and the ratio of Ti and (Al,Si) was closed to 1 : 3, which was considered as the Ti(Al1 – xSix)3 phase. Ti(Al1 – xSix)3 phase still maintained the crystal structure of Al3Ti phase. Due to the atomic radius of Al and Si were similar, Si atom can replaced Al ions in the lattice structure of Al3Ti phase to form Ti(Al1 – xSix)3 phase [20]. At point 5, it showed that the primary Al phase contained up to 53% Zn, because compositional segregation occurred in the Al-rich phase during solidification. The point analysis results of the eutectic region of the alloy structures at points 6 and 8 were very closed to the atomic ratio of MgZn2 phase, and the content of Al element in the eutectic region was low. It is indicated that there should be coexistence of Zn and MgZn2 phases here [21]. The ratio of Al and (Ti, Zr) was close to 3 : 1 at point 7, which was supposed to Al3(Ti, Zr) phase. Owing to the Al3Zr phase will precipitated preferentially with the addition of Ti and Zr elements when the temperature was above the peritectic reaction temperature of Al–Ti. When Al3Zr just precipitated and not aggregated, the Ti atom can replaced the Zr atom in Al3Zr to form the Al3(TixZr1 – x) phase [22]. At the same time, combined with Table 2, it can be known that this was the coexistence of Zn, Al and MgZn2 phases at points 3 and 9 in Fig. 4.

Fig. 4.
figure 4

EDS spectra of alloys: (a) Zn–5Al–2Mg–0.2(SiTi); (b) Zn–5Al–2Mg–0.2(SiZr); (c) Zn–5Al–2Mg–0.2(TiZr).

Full size image
Table 2.   The EDS results at the point in Fig. 4
Full size table

In order to determine the phase composition of the ternary eutectic structure in the alloy more accurately, Zn–5Al–2Mg–0.1(SiTi) alloy was detected by TEM and EDS. The bright-filed phase of the Zn–5Al–2Mg–0.1(SiTi) alloy was showed in Figs. 5a, 5d, 5g. Figs. 5b, 5c, 5e, 5h, 5f, 5i were respectively the selected area electron diffraction and energy-dispersive spectroscopy analysis corresponding to Figs. 5a, 5d, 5g. Through the bright field phase and the SAED pattern in Figs. 5a, 5b, it is determined that there was an Al phase in the eutectic region of the alloy, which corresponded to the result of EDS analysis. Besides, the phase was indexed as [1\(\overline{1}\) \(\overline{1}\)]Al, so the particle was identified to be fcc-Al. According to the SAED pattern (Fig. 5e) and the result of EDS analysis (Fig. 5f), this phase was considered to be [\(\overline{1}\)101]Zn. There were two forms of Zn in the eutectic structure, lamellar Zn and Zn particles wrapped in Al ring, showing in Fig. 6. In Fig. 5i, it can be seen that the ratio was very closed to the atomic ratio of MgZn2, and the SAED pattern (Fig. 5h) was consistent with [\(\overline{1}\) 2\(\overline{1}\) \(\overline{3}\) ]MgZn2. Therefore, the ternary eutectic structure was composed of Al, Zn and MgZn2 phase.

Fig. 5.
figure 5

The TEM analyses of ternary eutectic structure in Zn–5Al–2Mg–0.1(SiTi) alloy: (a, d, g) bright field phases, (b, e, h) diffraction patterns and (c, f, i) energy-dispersive spectroscopy analyses.

Full size image
Fig. 6.
figure 6

TEM image of Zn/Al/MgZn2 ternary eutectic structure.

Full size image

3.2 3.2 Rockwell Hardness

The addition of Si, Ti and Zr elements to the Zn–5Al–2Mg alloy contributed to a substantial increase in Rockwell hardness values as compared to the Zn–5Al–2Mg alloy, as showed in Fig. 7. It is clearly observed that with the increase of the addition of (Si, Ti) and (Si, Zr), the hardness values of the alloys first increased and then decreased, and the maximum values were 79.95 HRB and 72.03 HRB, respectively. However, the hardness value of Zn–5Al–2Mg–z(TiZr) alloy was increasing all the time. The Rockwell hardness value of Zn–5Al–2Mg–z(TiZr) alloy reached the maximum value of 77.79 HRB by the content of the addition (Ti, Zr) reached to 0.2 wt %. Combined with the analysis of microstructure, the hardness value was improved for the following reasons: (1) The structure of the alloy was refined by the addition of Si, Ti and Zr elements. According to the fine grain strengthening theory, the more the number of grains, the finer the grains, and the plastic deformation caused by external force can be dispersed in more grains. The plastic deformation was more uniform and the stress concentration was smaller, so the alloy sample had higher hardness. (2) The thickness of the Zn/Al/MgZn2 ternary eutectic structure played a significant role in the crack resistance and comprehensive mechanical properties of the galvanized alloy [23]. Combined with the appeal analyses, adding an appropriate amount of Si, Ti and Zr elements to Zn–5Al–2Mg alloy could improve the hardness value of the alloy.

Fig. 7.
figure 7

The Rockwell hardness diagram of Zn–5Al–2Mg–x(SiTi), Zn–5Al–2Mg–y(SiZr) and Zn-5Al–2Mg–z(TiZr) alloys.

Full size image

3.3 3.3 Electrochemical Impedance Spectroscopy (EIS)

Changed in OCP values as a function of immersion time were monitored for all the alloys in 3.5 wt % NaCl solution and the results were given in Fig. 8. The change of OCP at different immersion periods, showed a continuous change in the morphology of metal surfaces and corrosion products [24]. As reported in [25, 26], a relatively stable OCP value indicated a stable state between the advance of corrosion and the deposition of corrosion products. These samples exhibited a similar tendency with the immersion time, indicating that all alloys experienced a similar corrosion process. In addition, the curve of the open potential gradually tended to be stable after 1500 s.

Fig. 8.
figure 8

Change in OCP values of cast alloys as a function of time in 3.5 wt % NaCl solution.

Full size image

The EIS spectra of Zn–5Al–2Mg–x(SiTi), Zn–5Al–2Mg–y(SiZr), Zn–5Al–2Mg–z(TiZr) alloys were tested in the frequency range of 10–1–105 Hz, showed in Fig. 9. As seen in Figs. 9a, 9d, the Nyquist plots were composed of two capacitive loops: high frequency and low frequency, indicating the existence of two time constants. The first time constant at high frequency was attributed to electrolyte penetration including water uptake and salt intrusion, and the second time constant at low frequency was related to the charge transfer of the corrosion process [27, 28]. Moreover, the larger the diameter of the semi-circle is, the higher the resistance is and the lower the corrosion rate is [29]. It is clear that all these alloy samples had exhibited significantly higher charge transfer resistance than the Zn–5Al–2Mg alloy, besides, the Zn–5Al–2Mg–0.1(SiTi) alloy. B-P diagrams (Bode phase angle curve) in Figs. 9b, 9e and B-|Z| diagrams (Amplitude-frequency characteristic curve) in Figs. 9c, 9f were plotted and compared. Figs. 9b, 9e showed the wave crest at a frequency around the range of 102–103 Hz for tested specimens. Based on the observation of Bode impedance |Z| diagrams in Figs. 9c, 9f, it is interesting to notice that the |Z| value of Zn–5Al–2Mg alloy was improved with the addition of Si, Ti and Zr, indicating that improved the corrosion resistance of Zn–5Al–2Mg alloy. However, the |Z| value and the semi-circle diameter of Zn–5Al–2Mg–0.1(SiTi) alloy were slightly lower than that of Zn–5Al–2Mg alloy, the result showed that when the content of Si and Ti was 0.1 wt %, the resistance of the alloy to corrosive medium was weakened and the corrosion resistance of the alloy was reduced at the initial stage of corrosion.

Fig. 9.
figure 9

EIS spectra of alloys in 3.5 wt % NaCl solution: (a, d) Nyquist diagrams; (b, e) Bode phase angle diagrams and (c, f) Bode impedance |Z| diagrams.

Full size image

An equivalent electrical circuit(EEC), used to fit the experimental data was given in Fig. 10. Rs represented the solution resistance, Rf represented the resistance of the corrosion product layer, Rct was the charge transfer resistance, Q1 and Q2 were constant phase angle elements of corrosion product layer and electric double layer capacitance, respectively; n represented the dispersion coefficient, and the dimensionless constant range was 0–1 [30]. For n = 1, the constant phase element represented a pure capacitor, n = 0 represented a pure resistor, and n = 0.5 was a Warburg impedance with a diffusion character [31, 32].

Fig. 10.
figure 10

Equivalent circuit model for EIS data fitting of the experimental materials.

Full size image

The fitted EIS parameters were summarized in Table 3. As compared in Table 3, the Rs of all tests were much lower than Rf and Rct, indicating that all tests were in a stable environment. The EEC model did not changed after adding Si, Ti and Zr, showing that the corrosion behaviour of the Zn–5Al–2Mg alloy did not changed at the initial stage of corrosion. Independent of the proposed EEC model, the corrosion performance of the samples can be characterized by considering the value of Rp [33]. The value of polarization resistance Rp was equal to the resistance of the corrosion product layer Rf plus the charge transfer resistance Rct [34]. In principle, Rp can be considered as being inversely proportional to the corrosion rate, the greater the polarization resistance, the smaller the corrosion rate [28]. The Rp of the alloys by adding to Ti and Zr elements were much higher than Zn–5Al–2Mg alloy, demonstrating that the addition of Ti and Zr greatly reduced the corrosion rate of the alloy. The reason was that Ti and Zr elements, as commonly used refiners in Al alloy, had good grain refining effect. Ti and Zr elements were inherently good corrosion resistance. Hence, the addition of Ti and Zr could significantly improve the corrosion resistance of the Zn–5Al–2Mg alloy, making corrosion difficult to occur.

Table 3.   Equivalent circuit parameters obtained via fitting the EIS data results of alloys
Full size table

3.4 3.4 Potentiodynamic Polarization

The polarization curve tests were conducted in 3.5 wt % NaCl solution, and corresponding plots were shown in Fig. 11. The corrosion potential (Ecorr) and corrosion current density (Icorr) had been estimated using Tafel extrapolation, and obtained values were listed in Table 4. In Fig. 11, The Ecorr of these alloys by addition to (Si, Ti), (Si, Zr) and (Ti, Zr) were more higher than for Zn–5Al–2Mg alloy, indicating that the addition of (Si, Ti), (Si, Zr) and (Ti, Zr) made the alloy more difficult to corrosion. The Ecorr and Icorr of Zn–5Al–2Mg and Zn–5Al–2Mg–0.2(TiZr) alloys were –1.526 V and 2.461 E–4 A/cm2, –1.416 V and 1.176E–4 A/cm2 in Table 4, respectively. The Ecorr and Icorr of Zn–5Al–2Mg–0.2(TiZr) alloy increased by 0.110 V and decreased by 52.21% compared with Zn–5Al–2Mg alloy separately. Hence, the addition of Ti and Zr elements significantly increased corrosion resistance of the alloy. The reason was that the microstructure uniformity of Zn–5Al–2Mg–0.2(TiZr) alloy was improved and the lamellar spacing of ternary eutectic structure was reduced. A dense corrosion product layer was formed at the initial stage of corrosion, which hinders further the diffusion of corrosion ions. Relevant studies also showed that [35] the dense Zn/Al/MgZn2 ternary eutectic structure reduced the possibility of intergranular corrosion and improved the corrosion resistance of the alloy.

Fig. 11.
figure 11

Polarization curves of alloys in 3.5 wt % NaCl solution.

Full size image
Table 4.   Test results of alloys from polarization curves
Full size table

3.5 3.5 Total Immersion Test

The total immersion test was regarded as one of the most accurate methods to characterize the corrosion rate. The corrosion rate was calculated based on mass loss of all alloys immersed in 3.5 wt % NaCl solution, and the results were plotted in Fig. 12. In Fig. 12a, all the curves showed similar characteristic, with the immersion time increased, the corrosion rate first increased slowly, then decreased, then increased sharply and then decreased, and finally showed a slow rising trend. It is obvious that the corrosion rate was the highest on the 15th day. The Average corrosion rate histogram of alloys was showed in Fig. 12b. The corrosion rate of all alloys was lower than Zn–5Al–2Mg alloy, and the order was Zn–5Al–2Mg > Zn–5Al–2Mg–(SiTi) > Zn–5Al–2Mg–(SiZr) > Zn–5Al–2Mg–(TiZr). It is clearly observed that the corrosion rate of Zn–5Al–2Mg–0.2(TiZr) alloy was the lowest in these alloys, including that the corrosion resistance of Zn–5Al–2Mg–0.2(TiZr) alloy was best. This result was in line with the above EIS results, demonstrating further that the adding of Ti and Zr elements can significantly improved the corrosion resistance of the alloy.

Fig. 12.
figure 12

Total immersion test diagram of alloys: (a) corrosion weight loss rate curves; (b) average corrosion rate histogram.

Full size image

3.6 3.6 X-ray Photoelectron Spectroscopy Analysis

In order to identify corrosion products and analyze the corrosion mechanism, the Zn–5Al–2Mg and Zn–5Al–2Mg–0.2(TiZr) alloys were tested by XPS, and the results were presented in Figs. 13 and 14. The surface corrosion products of Zn–5Al–2Mg alloy mainly contained Zn, Al, Mg, C, O and other elements in Fig. 13a. The appearance of C mainly came from the C adsorbed in the test process and the corrosion products formed in the corrosion process, the element O mainly came from the corrosion products on the sample surface, the elements Na and Cl came from the NaCl solution, and the elements Zn, Al and Mg mainly came from the alloy sample itself. Afterwards, these elements were used to fit by software, were showed in Figs. 13b–13f. The C 1s spectrum can be divided into two peaks at the binding energy of 284.90 and 288.96 eV, corresponding to C–C and O–C=O, respectively. The O 1s spectrum had a peak at the binding energy of 532.03 eV, which corresponded to surface hydroxides and oxygen containing organic adsorbates on the surface [36]. The binding energy of 1303.48 eV of Mg 1s corresponded to the corrosion product MgO, the Zn 2p spectrum contained two peaks, at the binding energy of 1045.11 and 1022.29 eV, all corresponded to the corrosion product ZnO. The Al 2p spectrum contained a peak, at the binding energy of 74.25 eV, corresponding to the corrosion product Al2O3 [37].

Fig. 13.
figure 13

XPS analysis of Zn–5Al–2Mg alloy: (a) survey spectrum; (b) C; (c) O; (d) Mg; (e) Zn; (f) Al.

Full size image
Fig. 14.
figure 14

XPS analysis of Zn–5Al–2Mg–0.2(TiZr) alloy: (a) survey spectrum; (b) C; (c) O; (d) Mg; (e) Zn; (f) Al; (g) Ti; (h) Zr.

Full size image

The survey spectrum of Zn–5Al–2Mg–0.2(TiZr) alloy mainly had Zn, Al, Mg, C, O and trace amounts of Ti elements, was shown in Fig. 14a. Nevertheless, Zr element was not detected and only a very small amount of Ti was detected in the Zn–5Al–2Mg–0.2(TiZr) alloy. In addition, there was not found the corresponding compound of the Ti 2p spectrum at the binding energy of 476.77 and 473.98 eV in Fig. 14g. This may be due to the low content of Ti and Zr elements, or the strong corrosion resistance of Ti and Zr elements. In Figs. 14b–14f. The C 1s spectrum contained three peaks, at the binding energy of 284.8, 285.68 and 288.98 eV, corresponding to C–C, C–O–C and O–C=O, respectively. The binding energy of O 1s corresponded to oxide. The binding energy of 1304.12 eV of Mg 1s corresponded to Mg2+, indicating that there was Mg(OH)2 in the corrosion product layer. Al element mainly existed in the form of Al2O3, Zn element mainly existed in the form of Zn5(CO3)2(OH)6 [38].

Combined the above analysis, the schematic model of Zn–5Al–2Mg–z(TiZr) was shown in Fig. 15. It is known to all that the open circuit potentials of MgZn2 and Zn were both low in Zn–Al–Mg alloys. Hence, they were generally used as anodes to participate in the electrode reaction. The oxygen in the air acted as the cathode to participate in the electrode reaction. The reaction equation was as follows:

$${\text{Zn}} \to {\text{Z}}{{{\text{n}}}^{{2 + }}} + 2{{{\text{e}}}^{ - }},$$
(1)
$${\text{MgZ}}{{{\text{n}}}_{2}} \to 2{\text{Z}}{{{\text{n}}}^{{2 + }}} + {\text{M}}{{{\text{g}}}^{{2 + }}} + 6{{{\text{e}}}^{ - }},$$
(2)
$${{{\text{O}}}_{2}} + 2{{{\text{H}}}_{2}}{\text{O}} + 4{{{\text{e}}}^{ - }} \to 4{\text{O}}{{{\text{H}}}^{ - }},$$
(3)

Zn2+ and OH– in the solution reacted to form Zn(OH)2, while Zn(OH)2 will further decomposed into ZnO. In addition, Zn2+ also reacted with OH– and \({\text{CO}}_{3}^{{2 - }}\) in the solution to form Zn5(CO3)2(OH)6. As the reaction Eq. (3) progresses, the pH value increased, ZnO can originate from previously precipitated Zn5(CO3)2(OH)6, and its reaction was as follows equation 5, 6 and 7 [28, 38, 39].

$${\text{M}}{{{\text{g}}}^{{2 + }}} + 2{\text{O}}{{{\text{H}}}^{ - }} \to {\text{Mg}}{{\left( {{\text{OH}}} \right)}_{2}},$$
(4)
$${\text{Z}}{{{\text{n}}}^{{2 + }}} + 2{\text{O}}{{{\text{H}}}^{ - }} \to {\text{Zn}}{{\left( {{\text{OH}}} \right)}_{2}} \to {\text{ZnO}} + {{{\text{H}}}_{2}}{\text{O}},$$
(5)
$$5{\text{Z}}{{{\text{n}}}^{{2 + }}} + 6{\text{O}}{{{\text{H}}}^{ - }} + 2{\text{CO}}_{3}^{{2 - }} \to {\text{Z}}{{{\text{n}}}_{5}}{{\left( {{\text{C}}{{{\text{O}}}_{3}}} \right)}_{{\text{2}}}}{{\left( {{\text{OH}}} \right)}_{6}},$$
(6)
$$\begin{gathered} {\text{Z}}{{{\text{n}}}_{5}}{{\left( {{\text{C}}{{{\text{O}}}_{3}}} \right)}_{2}}{{\left( {{\text{OH}}} \right)}_{6}} + 2{\text{O}}{{{\text{H}}}^{ - }} \\ \to 5{\text{ZnO}} + 2{\text{HCO}}_{3}^{ - } + 3{{{\text{H}}}_{2}}{\text{O}}. \\ \end{gathered} $$
(7)
Fig. 15.
figure 15

The schematic model of corrosion mechanism of Zn–5Al–2Mg–z(TiZr) alloy.

Full size image

The corrosion products of Zn–5Al–2Mg alloy were ZnO, MgO and Al2O3, while the corrosion products of Zn–5Al–2Mg–0.2(TiZr) alloy were Mg(OH)2, Al2O3 and Zn5(CO3)2(OH)6. It is noteworthy that Zn–5Al–2Mg–0.2(TiZr) alloy did not had ZnO which was loose, porous and easy to fall off. Instead, it produced Zn5(CO3)2(OH)6 with better structure compactness and good adhesion [40]. The reason that may be the addition of Ti and Zr elements, promoted the formation of Zn5(CO3)2(OH)6, inhibited the formation of ZnO and the progress of reaction Eq. (7). Zn5(CO3)2(OH)6 can adhere to the surface of the sample to form a viscous protective layer, which hindered the transfer of electric charges, thereby reduced the corrosion rate of the sample to a certain extent.

4 CONCLUSIONS

The effects of the addition of Si, Ti and Zr elements to the Zn–5Al–2Mg alloy on the microstructure and corrosion resistance were studied. Major conclusions drawn from the work were summarized below:

(1) The Zn–5Al–2Mg alloy mainly contained Zn phase, Al phase and Zn/Al/MgZn2 ternary eutectic structure. When Si, Ti and Zr were added in binary and equal proportion respectively, Si rich phase, Ti(Al1 – xSix)3 phase, Al3Zr phase and Al3(Ti, Zr) phase appeared in the alloys.

(2) The Rockwell hardness value of these alloys by the addition of Si, Ti and Zr elements were higher than Zn–5Al–2Mg alloy. Among these alloys, Zn–5Al–2Mg–0.1(SiTi) alloy had the highest hardness, reached to 79.95 HRB.

(3) Combined with the results and analyses of EIS and total immersion test, the corrosion rate of the alloy was: Zn–5Al–2Mg alloy> Zn–5Al–2Mg–(SiTi) alloy > Zn–5Al–2Mg–(SiZr) alloy > Zn–5Al–2Mg–(TiZr) alloy. The corrosion resistance of Zn–5Al–2Mg–0.2(TiZr) alloy was best in the middle of these alloys.

(4) The results of XPS showed that the corrosion products of Zn–5Al–2Mg alloy were ZnO, MgO and Al2O3, while the corrosion products of Zn–5Al–2Mg–0.2(TiZr) alloy included Al2O3, Mg(OH)2 and Zn5(CO3)2(OH)6. The addition of Ti and Zr promoted the formation of Zn5(CO3)2(OH)6 and hindered the formation of loose, porous and easy to fall off ZnO. Zn5(CO3)2(OH)6 can adhere to the surface of the sample, which hindered the transfer of electric charges, thereby reduced the corrosion rate of the sample to a certain extent.