Regular Article
Formation Behavior of Fe–Zn Intermetallic Layers at the Interface between Fe–Mn and Pure Zn Melt at 460°C
2024 Volume 64 Issue 9 Pages 1457-1463
Details
2024 Volume 64 Issue 9 Pages 1457-1463
The effect of Mn on the alloying reaction during hot-dip galvanization was investigated. The microstructure of the Fe–Zn intermetallic layers consisted of ζ, δ, and Γ phases for both pure Fe and Fe–2Mn (wt.%) alloy. The intermetallic layers grew thicker with increasing dipping time, and the growth rate of each layer was similar for both substrates. In the case of Fe–2Mn, the formation of the δ1p phase was observed after dipping for 2 s. However, δ1p formation was delayed for pure Fe, indicating that Mn may promote nucleation of the δ1p phase. It is known that the δ1p phase nucleates in the Fe-saturated ζ phase. The Fe content at the ζ/δ1p interface was found to be lower for the Fe–2Mn alloy by electron probe microanalysis, suggesting that the supersaturation of Fe for the nucleation of δ1p is decreased by Mn addition and Mn may stabilize the δ1p phase. Once δ1p became a continuous layer, the growth rates of the δ1p layer on pure Fe and Fe–2Mn were similar. Mn could affect only the nucleation of δ1p during the initial stage of the alloying reaction.
Hot-dip galvanized steel sheets are extensively used in industries such as automobile manufacturing and construction owing to their good corrosion resistance.1,2,3,4,5) In an effort to reduce the weight of automobile bodies and improve fuel efficiency, the use of high-strength steel sheets has recently increased.3,5) Superior adhesion of the Zn coating is required to maintain sufficient formability of the hot-dip galvanized steel sheets.6,7,8,9,10,11,12,13) During the hot-dip galvanizing process, Fe–Zn intermetallic layers are formed by interfacial reaction between the steel and molten Zn.14,15,16,17) These intermetallic layers typically consist of ζ-FeZn13, δ-FeZn7–10, Γ1-Fe5Zn21, and Γ-Fe3Zn10.1,2) It has also been reported that there are two types of δ phases, namely, δ1k-FeZn7 and δ1p-FeZn10,14,15,16,17) where δ1k is an ordered phase of δ1p,18,19) and the phase diagram of Fe–Zn system was experimentally determined.15) The microstructures of the Fe–Zn intermetallic layers affect the adhesion at the steel/Zn interface.20) Thus, it is anticipated that achieving control over these microstructures could lead to improved adhesion.
High-strength transformation-induced plasticity steels contain Si and Mn to improve the mechanical properties.21,22,23,24) To successfully apply Zn coatings to such high-strength steels, it is crucial to understand the effects of these alloying elements on the interfacial reactions and the formation of the Fe–Zn intermetallic layers. Previous studies have reported that the Zn coatings on Fe–Si alloys were thicker than those on pure Fe,25,26) and the Fe–Si alloys were significantly consumed during dipping.26) It was suggested that the interfacial reactions in Fe–Si alloys may involve the decomposition of initially formed ζ-FeZn13 into liquid and FeSi phases due to an increase in the Si content in the ζ-FeZn13 phase by Si diffusion from the alloy substrate.26) However, the influence of the Mn content of steel sheets on the alloying reaction between Zn and steel during hot-dip galvanization is not yet fully understood. Previous research has focused on the effect of Mn addition to a Zn alloy bath,27) the formation of the intermetallic phases during the coating of steels with a Zn–Al alloy,28,29) and the alloying reaction of the hot-dip coating during the galvannealing process.29,30) The aim of this study was to investigate the influence of solute Mn in steel sheets on the alloying reaction during the hot-dip galvanizing process using Fe–2Mn (wt.%) alloy.
Pure Fe and Fe–2Mn alloy with a single-phase ferrite microstructure were used in the present work. The metal ingots were manufactured by Kobe Steel, Ltd., and the analytical compositions of ingots are shown in Table 1. The metal sheets with dimensions of 200 mm × 50 mm × 1 mm were hot dipped in molten pure Zn at 460°C. Hot-dip galvanization was performed using a hot-dip process simulator (RHESCA CO., LTD, HDPS). Figure 1 shows an example of the thermal profile during the hot dip galvanizing process. Prior to hot dipping, reduction treatment was performed at 850°C for 60 s for pure Fe, or at 700°C for 600 s for the Fe–2Mn alloy (corresponding to the single-phase ferrite region in this alloy system31)), followed by rapid cooling to 460°C under N2. Next, the sheets were dipped in molten Zn at 460°C for between 2 and 3600 s. The temperature of the metal/alloy surface during the hot dipping was recorded by a thermo-couple welded onto the surface. The microstructures of the Fe–Zn intermetallic layers after the hot dipping were observed by field-emission scanning electron microscope (FE-SEM). The thicknesses of the Zn coating and each of the Fe–Zn intermetallic layers were measured from the obtained FE-SEM images. The chemical compositions of the Zn coatings were also analyzed by energy-dispersive X-ray spectroscopy (EDS) and electron probe microanalysis (EPMA) to identify the Fe–Zn intermetallic phase.
| (wt.%) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Mn | C | Si | P | S | Al | N | O | Cu | Ni | |
| Pure Fe | <0.01 | 0.001 | 0.01 | <0.001 | 0.001 | 0.002 | 0.001 | 0.008 | – | – |
| Fe–2Mn | 1.91 | 0.001 | 0.01 | <0.002 | 0.001 | 0.001 | 0.0014 | 0.020 | <0.01 | <0.01 |
Figure 2 shows the thickness change of the Zn-coatings during dipping. The thickness was similar for both samples and increased gradually with increasing dipping time. After hot dipping for 60 s, the growth rate of the Zn coating tended to increase. No significant effect of Mn was observed in terms of either the overall thickness change or the growth rate of the Zn coating. Figure 3 presents cross-sectional FE-SEM images of pure Fe after various dipping times. After dipping for 2 s, the intermetallic layers of Fe–Zn consisted of ζ, δ1, and Γ phases with thicknesses of approximately 8 μm, 1 μm, and 800 nm, respectively. Each layer became thicker with increasing dipping time, and the formation of a protruding compound phase was observed above the δ1 layer after dipping for 10 s, as shown in Fig. 3(b). The δ1 phase is known to have two subphases, namely, δ1k and δ1p, where the latter is reported to exhibit a protruding morphology during the early stage of formation.32) Thus, the protruding intermetallic phase was tentatively identified as δ1p above a δ1k layer. After dipping for 60 s, a continuous layer had developed and the unevenness of the δ1p/ζ interface had decreased. Figure 4 shows the corresponding cross-sectional FE-SEM images for Fe–2Mn after dipping in Zn melt. Similar to pure Fe, ζ, δ1, and Γ layers were observed after 2 s of dipping, although these were slightly thicker than those for pure Fe. The formation of a protruding δ1p phase was also observed at this point, which indicates that Mn promoted the nucleation of this phase. With longer dipping times, all of the intermetallic layers grew thicker, and the growth of δ1 seemed to be faster among the intermetallic layers.
Figures 5 and 6 show the dependence of the total thickness of the Fe–Zn intermetallic layers and the thickness of each individual layer, respectively on the dipping time. In general, a power-law growth equation can be used to evaluate the growth rate:2)
where x is the thickness of the intermetallic layer, K is the growth rate constant, t is the dipping time, and n is the growth rate time constant. When the value of n is 0.5, the growth of the intermetallic layer is controlled by diffusion and the layer thickness increases parabolically. However, numerous studies have reported that n does not reach 0.5 during the growth of Fe–Zn intermetallic layers. For example, Jordan and Marder33) reported n values of 0.31–0.37 for the hot dipping of various steels in a pure Zn melt at 450°C. As shown in Fig. 5, in this work the n value was less than 0.3 until about 60 s, then it increased to approximately 0.3 for both pure Fe and Fe–2Mn. Figure 6 shows the growth kinetics of the ζ, δ1, and Γ layers for the two samples. The growth rate of δ1 was found to be higher than those of ζ and Γ in both cases. The thickness of the δ1 layer refer to that of the continuous δ1 layer, disregarding the protrusion of the δ1p phase (the thickness changes distinguishing between δ1k and δ1p are discussed in Section 3.4). The growth rate of the δ1 layer increased after 10 s for both samples, which may account for the increase in the overall growth rate of the intermetallic layers observed in Fig. 5. After dipping for 10 s, the n value for the δ1 layer reached approximately 0.5, which indicates that the δ1 layer grew parabolically. The n values of the ζ and Γ layers were both less than 0.5, in accordance with the results reported in a previous work.2) Marder reported that the Γ layer was consumed by the growth of the δ1 layer,2) and Onishi et al. described the dissolution of the ζ layer in the Zn melt.34) These factors may have prevented the n value from reaching 0.5. In regard to the latter factor, a decrease in the thickness of the ζ layer was observed after dipping for 3600 s, indicating the dissolution of the ζ layer in the Zn melt. The growth rate of each layer was similar irrespective of Mn addition.
Figure 7 shows the EPMA concentration profiles of each element after hot dipping for 2 s. The results confirmed the formation of Γ, δ1, and ζ layers from the substrate side. Among the intermetallic layers, the Γ layer exhibited the highest Mn content of approximately 0.5 at.%. The δ1 layer contained about 0.1–0.2 at.%Mn, while the ζ layer contained almost no Mn. As explained in Section 3.1, the nucleation of δ1p occurred with increasing dipping time, followed by the formation of a continuous layer. However, the continuous δ1k and δ1p layers were difficult to distinguish by microstructural observations. Figure 8 presents the EPMA concentration profiles for Fe–2Mn after dipping for 3600 s. In the δ1 layer, sudden step-like changes in the concentrations of Fe and Zn were observed at a distance of approximately 40 μm, and such changes have been reported previously for the δ1k/δ1p interface.14,32) Thus, the position of these changes observed during EPMA analysis was defined as the δ1k/δ1p interface. As shown in Fig. 8, the Mn content in the δ1k phase was found to be higher than that in the δ1p phase. Notably, the observed Mn enrichment in the δ1k or δ1p phases is in good agreement with a homogeneous δ1-phase region extending toward the equi-Zn concentration up to approximately 8 at.% Mn (stabilizing the δ1k or δ1p phases by Mn in solution) at a temperature of 450°C in the Fe–Mn–Zn ternary system.35)
Figure 9 shows the schematic illustration of the formation behavior of the Fe–Zn intermetallic layer on pure Fe and Fe–2Mn. The presence of Mn appeared to exert no significant effect on the microstructures or growth rates of the Fe–Zn intermetallic layers. As described above, the intermetallic layers initially consisted of Γ, δ1k, and ζ layers from the substrate side and the formation of a protruding δ1p phase was observed with increasing dipping time. The only difference observed between pure Fe and Fe–2Mn was the dipping time required to form the δ1p phase, where protrusions of the δ1p phase were observed above the δ1k layer in the Fe–2Mn alloy after only 2 s, compared with 10 s for pure Fe. This finding indicates that Mn promoted the nucleation of the δ1p phase. δ1p is known to nucleate in the Fe-saturated ζ phase.14) According to the EPMA results after dipping for 2 s (Fig. 7), the Fe content at the ζ/δ1k interface was approximately 10.6 at.% for pure Fe and 9.6 at.% for Fe–2Mn, and a small amount of Mn was found to be contained in the δ1p layer. Thus, it is suggested that the supersaturation of Fe for the nucleation of the δ1p phase is decreased by Mn addition, further Fe necessary for δ1p nucleation in pure Fe.
Figure 10 presents schematic Gibbs free energy curves for the ζ and δ1p phases in the Fe–Zn system. When the Fe content required for the nucleation of δ1p is Xζpure-Fe, the driving force for the nucleation of this phase in pure Fe, ΔGnclpure-Fe, is represented by the line AB. In the case of Fe–2Mn, the Fe content required for the nucleation of δ1p could decrease to Xζ2Mn. Under the assumption that the driving force for δ1p nucleation is same for pure Fe and Fe–2Mn, the δ1p phase be stabilized by Mn addition as shown by the dot curve in Fig. 10. The EPMA results shown in Fig. 8 reveal that the Mn content in the δ1p phase was approximately 0.15 at.%, which suggests that Mn dissolution stabilized this phase. However, the details regarding the stability of the δ1p phase are not understood at present, and further evaluation is needed.
Figure 11 shows the change in the thickness of the δ1k and δ1p layers with dipping time. The time when δ1p formation was observed is indicated by the black arrows. The δ1k layer initially grew slowly for shorter dipping times, but both the δ1p and δ1k layers grew more rapidly after δ1p had become a continuous layer. The n values were almost 0.5 for both δ1p and δ1p. However, the growth rate constant, K, was apparently larger for δ1p, causing this phase to grow to a greater thickness than δ1k in both pure Fe and Fe–2Mn. Wakamatsu and Onishi36) reported that the diffusion coefficient in the δ1 phase at 460°C increased with decreasing Fe content. The Fe content in the δ1p phase was lower than that in the δ1k phase, suggesting rapid growth of the δ1p layer. No significant difference was observed between the n and K values for pure Fe and Fe–2Mn. Therefore, Mn could affect only the nucleation of δ1p during the initial stage of the alloying reaction.
The influence of Mn on the alloying reaction during hot dip galvanization was investigated using Fe–2Mn alloy. The obtained results can be summarized as follows:
(1) Similar formation behavior of the Fe–Zn intermetallic layers was observed for pure Fe and Fe–2Mn. The growth rates of ζ and Γ were found to be lower than that of δ1, and there was no significant difference between pure Fe and Fe–2Mn.
(2) Mn promoted the nucleation of the δ1p phase, suggesting that the supersaturation of Fe for the nucleation of δ1p is decreased by Mn addition, and Mn may stabilize the δ1p phase.
Authors are requested to declare any conflicts of interest related to the conduct of this research.
The authors are grateful for the technical support Kobe Steel, Ltd. gave with the sample preparation. The support of The Iron and Steel Institute of Japan was gratefully acknowledged.
