3.1. Microstructure of the Weld Joint
Three locations (weld center, aluminum weld heat-affected zone and steel-aluminum interface) were selected to analyze the microstructure of the weld joint obtained by SEM, as shown in
Figure 1, where
Figure 1a–c represent the micromorphology of the different positions of the welding joint without high-temperature oxidation, and
Figure 1d–f show the micromorphology of the different positions of the welding joint with high-temperature oxidation. The weld microstructure is mainly composed of α-Al solid solution and Al-Si eutectic phases in
Figure 1a,d. However, petal structures are sparsely distributed in the weld center after high-temperature oxidation treatment, as shown in the yellow circle in
Figure 1d.
Figure 2 and
Table 2 show the surface scan and point scan results of the local magnified view in the upper right corner of
Figure 1a,d, respectively. The EDS analysis of point 1 and point 4 showed that the composition in this area was mainly composed of aluminum substrate; point 2 indicates that the area is mainly Al-Si eutectic phases, and point 3 indicates that iron forms a petal-like compound with aluminum, which may be caused by the diffusion of iron in steel into the weld at high temperature. The results of weld surface scanning in
Figure 2 show that iron diffuses to the central structure of the weld after high-temperature oxidation compared with that without high-temperature oxidation. The microstructure of the weld heat-affected zone is similar in
Figure 1b,e, indicating that high-temperature oxidation has no obvious effect on this area.
Figure 1c,f show the structure at the interface between steel and aluminum. There are intermetallic compounds (IMCs) due to the physical diffusion and chemical reaction between different elements at the interface [
10]. The width of the intermetallic compounds at the interface of the sample after high-temperature oxidation treatment was wider (approximately 7–8 μm) than that of the sample (2–3 μm) without high-temperature oxidation treatment. This result indicated that the IMCs at the interface obtained heat energy and further grew during HTO treatment, which means that the thickness of the IMCs is related to the heat input (as reported by Honggang Dong et al. [
11]).
3.2. Morphology of High-Temperature Oxidation (HTO) Film
Figure 3a depicts the surface macroscopic morphology of the HTO film of the weld joint. It can be seen from the figure that the 304 stainless steel side is yellow, and no metallic luster can be seen on the side of the weld and 6061 aluminum alloy. Three locations were selected as shown in
Figure 3a to analyze the microstructure of the HTO film by SEM, as shown in
Figure 3b–d, in which the upper right corners are partially enlarged views.
Figure 3b shows that the surface of the IMC side is smooth without an obvious oxidation film, the metal surface of 304 is smooth at a magnification of 2000× and there is an obvious discontinuous oxidation film at a magnification of 10,000×. Some oxide convex and concave areas can be clearly observed on the surface of the 6061 side at 2000× or 10,000×. EDS surface scanning analysis was carried out, as shown in
Figure 3c–e, and the analysis results are shown in
Figure 4. It was found that there were oxygen-containing compounds in these three surface scans (IMC, 304 and 6061) or in the three EDS point scans, indicating that the sample substrate was covered with the high-temperature oxidation film, and their atomic percentages of oxygen from the surface scan results were 4.12%, 7.65% and 20.51%. Fe and Cr oxides were mainly formed on the side of 304 as well as Al oxides on the side of 6061 during the oxidation process, as shown in
Table 3. According to the EDS surface scanning analysis (
Table 3), more oxygen elements were found on the 6061 aluminum alloy side of the welded joint compared with the side of IMC and 304 which underwent high-temperature oxidation treatment at a temperature of 500 °C because at the same temperature, the Gibbs free energy of the Al
2O
3 is lower than the Cr
2O
3, and the ability of aluminum to reduce oxygen is stronger. In short, the driving force for the formation of Al
2O
3 is greater than that of Cr
2O
3, as demonstrated in the diagrams of ∆Gv⊝of the formation of oxides [
12]. At the same time, Chenglong Yu et al. [
13] found that the oxidation initiates at 140 °C for aluminum, and the oxidation resistance of austenitic stainless steel is still very good at 900 °C, according to the oxide thickness scale for oxidation of 304 [
14].
Figure 5a depicts the cross-section SEM image of the HTO film of the weld joint. Three locations were selected as shown in
Figure 5a to analyze the microstructure of the weld joint, as shown in
Figure 5b–d.
Figure 5e is the cross-section SEM image of the 6061 side of the weld joint, which is the same sample as in
Figure 5a.
Figure 5f is the cross-section SEM image of the 6061 side of the weld joint without HTO treatment. The thicknesses of the HTO film on the 304 and 6061 sides are 449.35 nm and 2401 nm, respectively, while the thickness of the 6061 side of the weld joint without HTO treatment is 181.1 nm. That is, the film thickness of the sample increases after HTO treatment. At the same time, we found that there were cracks in some parts of the oxide film on the 6061 side of the weld joint, as shown in the red circle in
Figure 5a,e.
Based on the EDS point scanning analysis results from
Table 4, the oxygen content on the surface of the interface intermetallic compound is 4.29 at.% (point 2), and the oxygen content of the substrate is 1.92 at.% (point 4); the oxygen content on the surface of the 304 side is 6.00 at.% (point 5), while the substrate is 1.61 at.% (point 6); the oxygen content on the surface of the 6061 side is 9.82 at.% (point 7) and 7.04 at.% (point 9), while the substrate is 1.69 at.% (point 8). The oxygen content at point 10 is 7.45 at.%, which is similar to that of the sample oxidized at high temperature. Therefore, it can be inferred from the EDS point scanning results that different parts of the 304/6061 dissimilar metal welding–brazing joint treated by HTO can react with oxygen to generate oxides.
3.3. Corrosion Resistance of the HTO-Treated 304/6061 Joint
To investigate corrosion properties such as corrosion potential, potentiodynamic polarization tests were conducted, as shown in
Figure 6. The corrosion potential (Ecorr) and corrosion current density (Icorr), which were obtained from the potentiodynamic polarization curves, are summarized in
Table 5. The corrosion resistance of samples with HTO treatment was lower than that of samples without HTO treatment. Whether oxidized at high temperature or not, the corrosion potential of the welding samples was situated between those of the two base materials. This result can be explained by the mix potential theory [
15], where a new potential is determined by the intersection of the anodic reaction (6061) and cathodic reaction (304) of the unjoined metals, which is similar to the report by Bosung Seo et al. [
3]. The sequence of the corrosion current density was weld (HTO) >weld>6061 (HTO) >6061>304 (HTO) >304. The higher the corrosion current density is, the worse the corrosion resistance of the materials. Furthermore, from
Figure 6 (the purple curve), it is seen that the anodic and cathodic processes for the weld sample are both under activation control. Nevertheless, a short active–passive transition for the weld (HTO) sample (the yellow curve), which is evidenced with a decrease in corrosion current density as potential increased from −0.8 V to −0.5 V vs. SCE, indicates the occurrence of a passivation process that is related to the formation of the protective layer [
16,
17] on the surface of the weld (HTO) sample.
The corrosion morphologies after polarization are shown in
Figure 7. Some deep corrosion pits were observed in the base metals without HTO treatment as well as the base metals with HTO treatment corroded in a large area. There was intergranular corrosion (
Figure 7b) in 304 stainless steel (HTO). This is because sensitization occurred in 304 stainless steels during oxidation treatment at 500 °C. Sensitization is a deleterious phenomenon for austenitic stainless steel when they are heated to a certain temperature around 500 and 800 °C [
18]. During the exposure of austenitic stainless steel in this high temperature range, chromium carbides precipitate near the grain boundary, resulting in chromium deficiency [
19,
20]. According to the EDS spot scanning results in
Table 6, the Cr element in point 1 (10.63 at.% or 14.45 wt.%) of the severely corroded area is slightly lower than that in point 2 of the non-corroded area, and the Cr element content in point 2 (17.52 at.% or 18.03 wt.%) is similar to that of the base material (17.51 wt.%). Due to reduced Cr, the corrosion resistance of the grain boundaries was reduced. Then, intergranular corrosion will occur when they are exposed to corrosive media. Corrosion cracks were found in the 6061 alloy (
Figure 7d).
Figure 7e,f show the corrosion morphology of the interface after polarization of the weld joint, while
Figure 7g,h show the corrosion morphology of the 6061 side of the weld after polarization. Obvious galvanic corrosion was found in the 304/6061 joints (whether HTO treated or not). The 6061 aluminum alloy and weld were corroded as anodes, while 304 stainless steel was protected as cathodes. In addition, the corrosion of the sample after high-temperature oxidation was more serious than that of the weld joint sample without high-temperature oxidation. The corrosion phenomenon is similar to that of a single base metal; there were some local corrosion pits on the welded joint sample without the HTO film (
Figure 7g), while large corrosion areas and local corrosion pits were found on the welded joint sample with the HTO film (
Figure 7h). The above phenomena indicate that the HTO treatment will reduce the corrosion resistance or even accelerate the corrosion of the base metals and weld joints. Because there were tiny cracks in the oxide film (
Figure 5a,e), the corrosion solution was immersed in it and reacted with the base metal to corrode the metal.
The composition of the base metals treated by HTO was measured by an EDS point scan and is listed in
Table 6. Point 1 has an atomic composition of 39.23% Fe, 10.63% Cr and 41.04% O, while point 2 has an atomic composition of 62.66% Fe, 17.52% Cr and 11.43% O. The results of point 4 show that it is an aluminum matrix, and the results of points 3, 5 and 6 show that it is aluminum oxide. From the results of the EDS point scan, it is confirmed that the corrosion products of the 304 stainless steel base metal are mainly iron and chromium oxides, while the corrosion products of the aluminum alloy base metal are mainly aluminum oxides.
Variations in the aluminum, iron and oxygen contents at the interface of the weld joint were measured by an EDS line scan along the yellow line marked in
Figure 7e,f, as shown in
Figure 8. This change indicates that the oxygen element distribution on the aluminum side of the weld treated by HTO is uneven, and its fluctuation is larger than that of the sample without HTO treatment. The uniformity of the oxide distribution is related to the corrosion resistance of the sample; the more uneven its distribution is, the worse its corrosion resistance is.
A total of 360,000 data points were obtained through electrochemical noise measurements both in current and potential vs. time.
Figure 9 displays the time dependence of the galvanic potential (a) and current density (b) for the 304-6061 and 304-6061 (HTO) couples in 3.5 wt.%NaCl solution. From these plots, both the galvanic potential and current density of the 304-6061 couple oscillate sharply; the maximum amplitude for potential is 0.6 V, while that for current density is 0.32 × 10
−6 A/cm
2. The galvanic potential of the 304-6061 (HTO) couple has no obvious fluctuation, except that the potential fluctuation amplitude is approximately 0.01 V between 24,374 s and 31,254 s. High frequency fluctuation is observed for the galvanic current density of the 304-6061 (HTO) couple after 30,000 s and then trends toward stable. From the above analysis, it can be inferred that the galvanic corrosion of the 304-6061 couple was more active than that of the 304-6061 (HTO) couple. The average galvanic current density of the 304-6061 couple (1.571 × 10
−5 A/cm
2) is slightly higher than that of the 304-6061 (HTO) couple (1.044 × 10
−5 A/cm
2), which means that the samples after HTO treatment have a lower tendency of galvanic corrosion. However, according to the data analysis after polarization corrosion (
Figure 6 and
Figure 7), the samples with HTO treatment have more serious corrosion, which may be due to the following two reasons: on the one hand, the HTO treatment is carried out in air, and the oxygen content is not sufficient, resulting in an uneven oxide film (
Figure 3). On the other hand, there were small cracks in local areas of the surface oxide film (
Figure 5), easily leading to local corrosion. The samples with HTO treatment are more prone to many local pits, thus increasing the corrosion rate, as shown in
Figure 6 and
Figure 7.