2023 Volume 64 Issue 10 Pages 2496-2502
The corrosion behavior and resistance of hot-dip galvanized steel were studied in a saturated Ca(OH)2 aqueous solution containing chloride ions (Cl−) in concrete. In the saturated Ca(OH)2 aqueous solution, which simulated water in the concrete pores, Ca(Zn(OH)3)2·2H2O was formed on the surface of the hot-dip galvanized steel, acting as a protective film. However, the corrosion rate of zinc increased as the Cl− concentration increased. This is presumed to be owing to the formation of Zn5(OH)8Cl2·H2O and CaCO3 on the surface and the decrease in the coverage of the protective film Ca(Zn(OH)3)2·2H2O. However, in concrete, the corrosion of hot-dip galvanized steel was not promoted by the cyclic corrosion test. This may be because the Cl− that penetrated into the concrete did not reach the depth of the galvanized steel and the Ca(Zn(OH)3)2·2H2O that was formed during the curing of the concrete remained after the cyclic corrosion test.
This Paper was Originally Published in Japanese in Zairyo-to-Kankyo 71 (2022) 263–270. Abstract, captions of Figs. 1–13, and Figs. 2, 3, 12 were slightly modified.
Fig. 13 Diagram of the penetration range of the Cl− and scheme of the corrosion mechanism of galvanized steel in concrete.
In the field of building materials, hot-dip galvanized steel is expected to be applied to concrete structures. Since the corrosion rate of zinc generally increases in an alkaline environment such as concrete, the corrosion resistance of the hot-dip zinc coating in concrete with a pH of 12.6 has been a concern.1) However, previous studies1–6) have shown that in a solution simulating concrete (a saturated aqueous solution of Ca(OH)2), corrosion resistance is maintained because of Ca(Zn(OH)3)2·2H2O (calcium hydroxy zincate; hereinafter referred to as CHZ), which suppresses corrosion. According to the report by Maeda et al.,6) CHZ precipitates based on eq. (2) after the eluted zinc reacts with Ca contained in the concrete based on eq. (1).
| \begin{equation} \text{Zn} + \text{4OH$^{-}$} \to \text{Zn(OH)$_{4}{}^{2-}$} + \text{2e$^{-}$} \end{equation} | (1) |
| \begin{align} &\text{2Zn(OH)$_{4}{}^{2-}$} + \text{Ca$^{2+}$} + \text{2H$_{2}$O} \\ &\to \text{Ca(Zn(OH)$_{3}$)$_{2}$}{\cdot}\text{2H$_{2}$O} + \text{2OH$^{-}$} \end{align} | (2) |
However, it has been reported that the corrosion resistance of zinc decreases with an increase in the concentration of chloride ions (hereinafter referred to as Cl−) contained in the concrete-simulating solutions.7,8) Hence, it is necessary to investigate in detail the correlation between the concentration of Cl− and corrosion resistance of zinc.
As described above, most of the existing knowledge on the corrosion mechanism and resistance of zinc in concrete has focused on hot-dip zinc-coated steel1–5,7) or pure zinc.6,8) However, there is only one report on hot-dip galvanized steel sheets (JIS G 3302) manufactured on a continuous galvanizing line.9) Additionally, there have been no reports on the corrosion mechanism of hot-dip galvanized steel sheets in actual concrete containing gravel, admixtures, and aggregates. Therefore, it is necessary to investigate the corrosion mechanism of hot-dip galvanized steel sheets in actual concrete for the application of hot-dip galvanized steel sheets manufactured in a continuous galvanizing line into concrete structures.
In this study, hot-dip galvanized steel manufactured in a continuous galvanizing line was first used to investigate the effect of the Cl− concentration on the corrosion mechanism and resistance in a concrete-simulating solution. Additionally, hot-dip galvanized steel sheets manufactured on a continuous galvanizing line were embedded in actual concrete to evaluate the corrosion resistance, and the results were compared to the evaluation results in a concrete-simulating solution. Further, based on the corrosion mechanism estimated by the test in the concrete-simulating solution, the corrosion mechanism in the actual concrete was also considered.
Hot-dip galvanized steel sheets (thickness: 0.6 mm, coating weight: 80 g/m2 on one side) manufactured on a continuous galvanizing line were cut into 20 × 70 mm pieces, ultrasonically degreased in acetone, and subjected to the test. For the concrete-simulating solution, three kinds of solutions were used: (a) Ca(OH)2 saturated aqueous solution, (b) Ca(OH)2 saturated aqueous solution with 0.5 mass% NaCl (hereinafter referred to as Ca(OH)2 + 0.5% NaCl), and (c) Ca(OH)2 saturated aqueous solution with 5.0 mass% NaCl (hereinafter referred to as Ca(OH)2 + 5.0% NaCl). The pH of (a) was 12.6. For (b) and (c), NaOH was added to adjust the pH to 12.6. Figure 1 shows a diagram of the electrochemical measurement cell used in the test.
Diagram of the electrochemical measurement cell.
The specimen, counter electrode (Pt), and reference electrode (KCl–Ag/AgCl) were placed, and 400 mL of a concrete-simulating solution was poured. The holding plate of the specimen has a hole of 10 mm2, and only the exposed specimen surface comes into contact with the solution. Here, the specimens were immersed in the solution for 7 days, and the corrosion products formed at the surface of the specimen were analyzed. Polarization measurements were performed using the electrochemical measurement cell and a potentiostat (Biologic VSP). The sweep rate during polarization was 1.0 mV/s, and the potential was swept from the corrosion potential to +1.0 and −1.0 V during anodic and cathodic polarizations, respectively. As shown in Fig. 2, different specimens were used for the anodic and cathodic polarization.
Scheme of the immersion test.
Figure 3 shows a diagram of the specimen.
Diagram of the concrete test specimen.
Hot-dip galvanized steel sheets (thickness: 2.3 mm, coating weight: approximately 135 g/m2 on one side) were cut into 60 × 60 mm pieces, and the ends and back surface were sealed with epoxy resin paint to prevent corrosion. In this test, specimens with different plate thicknesses and coating weights from 2.1 were used. The sealed specimen was embedded in concrete of 100 × 100 × 120 mm (width, depth, and height, respectively) and cured indoors for 28 days before being subjected to the test until the concrete solidified. Table 1 shows the composition of the concrete. Ordinary Portland cement was used, and the water/cement ratio was 64.4%. After curing, the specimen was placed in a cyclic corrosion test (CCT) equipment. Further, it was subjected to a neutral salt spray cycle test based on JIS H8502 (5.0 mass% salt spray at 35°C for 2 h → drying at 60°C, <30% relative humidity, 4 h → humidifying at 50°C, >95% relative humidity, 2 h). This cycle was performed for 30 cycles. To investigate the influence of the concentration of Cl− contained in the sprayed salt water, two aqueous solutions containing 5.0 mass% and 0.5 mass% NaCl (hereinafter referred to as 0.5% NaCl aq. and 5.0% NaCl aq.) were used. It has been reported that the conditions for using 5.0% NaCl aq. are equivalent to those for hot-dip galvanized steel sheets exposed to air for three years in Okinawa, a salt-damaged environment.10) Additionally, for comparison, a specimen without CCT was also prepared and used for the test.
After the immersion test, the surface morphology of the specimens was observed using a scanning electron microscope (SEM; Hitachi SU6600), and elemental analysis was performed by energy dispersive X-ray analysis (EDX). Structural analysis of the corrosion products formed on the surface of the specimens was performed by X-ray diffraction (XRD; SmartLab manufactured by Rigaku) measurement (X-ray source: Co-Kα, scan range 5–110° (2θ)). Furthermore, after the specimen was embedded in resin and polished, elemental analysis was performed on the cross-section of the specimen using an electron probe microanalyzer (EPMA; JEOL JXA-8230).
2.3.2 Analysis of specimens after the CCTAfter removing the concrete from the specimen to be analyzed, the steel plate was extracted and used as a specimen. First, the coating thickness before and after the CCT was compared to evaluate of reduction in the coating thickness owing to corrosion. Here, the initial coating thickness before the CCT was the value obtained by observing the cross sections of 11 arbitrary fields of view using an optical microscope for specimens cut from a identical steel plate as the hot-dip galvanized steel plate embedded in concrete. For the coating thickness after CCT, the values measured similarly on the specimen extracted from the concrete were used. Furthermore, structural analysis was performed on the central part of the surface of the specimen using XRD measurement. Additionally, for the specimens without the CCT and those after the CCT with 5.0% NaCl aq., the section where the concrete adhered to the surface was scraped with a width and depth of approximately 50 and 30 µm, respectively by a focused ion beam-scanning electron microscope combined system (FIB-SEM; Hitachi NB5500). A part of the machined cross-section was observed using a bright-field scanning transmission electron microscope (BF-STEM), and a line analysis from the coating to concrete layer was performed by EDX. In a portion of the field of view of the BF-STEM image, the coating/concrete interface was observed using a field emission transmission electron microscope (FE-TEM; JEOL JEM-1200F) at an accelerating voltage of 200 kV. Elemental analysis was performed using a JED-2300T, and electron diffraction measurements were performed using a µ-diffraction beam (beam diameter: approximately 10 nm).
Figure 4 shows the results of the polarization measurements of each specimen.
Polarization curves of the test specimens after the immersion test for 6.0 × 105 s in (a) Ca(OH)2, (b) Ca(OH)2 + 0.5%NaCl, and (c) Ca(OH)2 + 5.0%NaCl.
In solutions (b) and (c) containing Cl−, the anodic current increased with an increase in the Cl− concentration. For the cathodic current, there was no difference between solutions (a) and (b), although it increased in solution (c). Figure 5 shows the results of calculating the corrosion rate from the polarization curve in Fig. 4 using the Tafel extrapolation method with reference to the report by Tsujita et al.7)
Corrosion rate of the test specimens in (a) Ca(OH)2, (b) Ca(OH)2 + 0.5%NaCl, and (c) Ca(OH)2 + 5.0%NaCl.
The corrosion rate values increased in the order of (a), (b), and (c), thereby revealing that the corrosion resistance decreased as the Cl− concentration increased.
3.1.2 Morphology of corrosion productsFigure 6 shows the SEM image of the cross-section of each specimen after the immersion test and the results of the elemental analysis by EDX at an arbitrary point (indicated as “+” in the figure) in the corrosion product.
Cross-section SEM images and EDX analysis (atomic%) of the corrosion products formed in (a) Ca(OH)2, (b) Ca(OH)2 + 0.5%NaCl, and (c) Ca(OH)2 + 5.0%NaCl.
Corrosion products with a thickness of approximately 5 µm were formed on the surface of each specimen. In (c), in addition to the corrosion products with a thickness of approximately 5 µm, corrosion products with a thickness of approximately 10 µm were also observed locally. At points (a), (b), and (c) ii, Zn, O, and Ca were detected at similar concentrations and at point (c) i, only Ca and O were detected. Further, Cl was not detected at points (b) and (c). Figure 7 shows the results of the elemental mapping using EPMA on the cross-section observed in Fig. 6.
Cross-section SEM images and EDX analysis (atomic%) of the corrosion products formed in (a) Ca(OH)2, (b) Ca(OH)2 + 0.5%NaCl, and (c) Ca(OH)2 + 5.0%NaCl.
In all cases (a) to (c), Zn, Ca, and O were distributed in the corrosion products with a thickness of approximately 5 µm. In (c), among the corrosion products with a thickness of approximately 10 µm, the distribution of the elements in the dark and bright parts differ, with Ca and O in the former and Zn, Ca, and O in the latter. Additionally, in (b) and (c), Cl was only slightly distributed locally.
3.1.3 Structure of corrosion productsFigure 8 shows the results of the XRD measurement for each specimen. In (a), only the CHZ peak was detected. In (b), the Zn5(OH)8Cl2·H2O peak was detected in addition to CHZ. In (c), the CaCO3 peak was detected in addition to CHZ and Zn5(OH)8Cl2·H2O. However, the diffraction intensity of the peaks of the compounds other than CHZ is weak, and they are believed to exist only slightly in the product.
XRD patterns of the corrosion products formed in (a) Ca(OH)2, (b) Ca(OH)2 + 0.5%NaCl, and (c) Ca(OH)2 + 5.0%NaCl.
Figure 9 shows the results of measuring the initial coating thickness before and after the CCT. Since the coating thickness of hot-dip galvanized steel sheets produced in the continuous galvanizing line varies, it is thought that the initial coating thickness also varies, and the actual value was 15.0 to 21.7 µm (represented by the dotted line). At a depth of 35 to 85 mm from the concrete surface, the coating thickness after the CCT also varied, which was within the range of the initial coating thickness. This suggests that there was no corrosion.
Correlation between the thickness of Zn and distance from the surface of the concrete after the CCT with 0.5%NaCl aq. and 5.0%NaCl aq., respectively.
Figure 10 shows the results of the XRD measurement for each specimen taken out of the concrete before and after the CCT. The CHZ and ZnO peaks were detected in all the specimens. The CHZ peak is broader than the peak of the sample after the immersion test shown in Fig. 8, suggesting lower crystallinity. Additionally, the Zn5(OH)8Cl2·H2O and CaCO3 peaks detected after the immersion test in Fig. 8 were not detected.
XRD patterns of the corrosion products formed before and after the CCT with 0.5% and 5.0%NaCl aq., respectively.
Figures 11(i) and (ii) show the FE-TEM images of the plating/concrete interface and electron beam diffraction patterns at multiple points in the images for the specimens before and after the CCT. Elements were identified by qualitative analysis using EDX. In the FE-TEM images, both before and after the CCT, a coating and concrete layer were observed in the lower and upper part, respectively (black and gray part, respectively). In addition, a layer with different contrast was observed at the interface between them. The yellow line indicates the boundary with the concrete. At points a, b, and e located in the concrete layer, patterns of the amorphous structure containing Zn in addition to the concrete components (Ca, Al, Si, and O) were detected. Points c and f located at the coating/concrete interface have a structure where the amorphous and microcrystals containing Ca, Zn, and O are mixed, and these layers correspond with CHZ. Additionally, in the specimen after the CCT, point g showed the crystal structure of ZnO.
FE-TEM images, electron diffraction patterns of spots, and EDX line scanning profile of the interface between Zn and concrete; (i) and (iii) before, and (ii) and (iv) after the CCT with 5.0%NaCl aq.
Figures 11(iii) and (iv) show the results of the line analysis using EDX in the area indicated by the dotted line in the FE-TEM images. Zn was detected in all the areas of the coating layer (0 to 2.5 µm), coating/concrete interface (2.5 to 2.8 µm, gray area in the figure), and concrete layer (2.8 to 4.5 µm). It diffused from the coating to the concrete layer. Additionally, no Cl was detected before or after the CCT. From the above, the structure and composition of the coating/concrete interface were similar before and after the CCT.
Figure 12 schematically shows the presumed corrosion mechanism of the hot-dip galvanized steel sheets in concrete simulation solutions (a)–(c).
Scheme of the corrosion mechanism of galvanized steel in (a) Ca(OH)2, (b) Ca(OH)2 + 0.5%NaCl, and (c) Ca(OH)2 + 5.0%NaCl.
In the concrete-simulating solutions (a) to (c), the formation of the CHZ was confirmed regardless of the Cl− concentration. In solution (b), Zn5(OH)8Cl2·H2O was formed in addition to the CHZ. In solution (c), CaCO3 was formed in addition to the CHZ and Zn5(OH)8Cl2·H2O, and the corrosion resistance decreased with an increase in the Cl− concentration. The Zn5(OH)8Cl2·H2O formed on the Zn surface in the concrete-simulating solution has a lower protective property than CHZ.8) Further, from the results of the SEM observation, there is a high probability that CaCO3 precipitated as a porous film. From the aforementioned, it is presumed that the proportion of CHZ in the total corrosion products decreased, thereby decreasing the corrosion resistance.
Further, Zn5(OH)8Cl2·H2O cannot exist thermodynamically under the Cl− concentration and pH conditions of this study.11) Although the formation of CaCO3 was not confirmed in solution (b), it is believed that CO2 in the atmosphere dissolves the concrete-simulating solution, thereby causing a formation reaction of CaCO3 as expressed by eq. (3), which locally reduces the pH and forms Zn5(OH)8Cl2·H2O.
| \begin{equation} \text{Ca$^{2+}$} + \text{HCO$_{3}{}^{-}$} \to \text{H$^{+}$} + \text{CaCO$_{3}$} \end{equation} | (3) |
The corrosion mechanism of the hot-dip galvanized steel sheets in actual concrete was discussed based on the corrosion mechanism in the concrete-simulating solution shown in 4.1. Figure 13 shows a diagram. It is presumed that Zn was eluted into the concrete as Zn(OH)42− and CHZ was formed at the coating/concrete interface during the solidification process of the concrete (Figs. 13(i)–(iii)). This is similar to the case where the hot-dip galvanized steel sheet was immersed in the concrete-simulating solution.
Diagram of the penetration range of the Cl− and scheme of the corrosion mechanism of galvanized steel in concrete.
Additionally, since Zn was also distributed in the concrete layer, part of the eluted Zn diffused into the concrete layer and incorporated into compounds14) generated by hydration reactions12,13) during the concrete solidification process, such as SiO2, Al2O3, and CaO.
Further, even after the CCT, CHZ existed at the coating/concrete interface regardless of the Cl− concentration (Fig. 13(iv)), thereby suggesting that CHZ formed during the concrete solidification process remained after the CCT.
Furthermore, the corrosion products containing Cl such as Zn5(OH)8Cl2·H2O were not formed, and the distribution of Cl was not confirmed. According to a report by Fujiwara et al., even when a concrete piece with a thickness of 150 mm was immersed for 9.1 × 106 s in a 10 mass% NaCl aqueous solution with a higher concentration than in this test, the penetration depth of the Cl− remained at approximately 20 mm.15) Hence, as shown in Fig. 13, Cl− did not reach the depth range of 30 to 90 mm from the concrete surface even after the CCT, and the corrosion products containing Cl were not formed.
4.3 Corrosion resistance in concreteIn the concrete-simulating solution, the corrosion resistance decreased with an increase in the Cl− concentration in the solution. However, no decrease in the corrosion resistance was observed with an increase in the Cl− concentration in actual concrete. One of the factors is that the actual concrete is an environment in which corrosion is less likely to proceed than in the simulated solution. However, as aforementioned, it is more likely that the corrosion resistance was maintained because Cl− penetrated the concrete after the CCT did not reach the steel plate surface, thereby causing CHZ to remain.
Based on the above-mentioned report,15) the penetration depth of the Cl-increases with the extension of the immersion period in salt water. Hence, it is expected that the correlation between the Cl− penetration depth and corrosion resistance will be clarified by increasing the number of CCT cycles and actual exposure tests. Additionally, the penetration depth of the Cl− changes depending on the water/cement ratio and curing method of the concrete.16) In the future, it is important to evaluate the corrosion resistance in concrete by clarifying these effects.
Based on the investigation of the corrosion mechanism and resistance of hot-dip galvanized steel sheets manufactured in a continuous galvanizing line in a concrete-simulating solution and actual concrete, the following conclusions were obtained:
