2016 Volume 57 Issue 11 Pages 1922-1929
The optimal applied potential for the acceleration of the corrosion of hot-dip aluminized ferritic stainless steel in 5 mass% NaCl solution was investigated using electrochemical techniques (open-circuit potential measurement, electrochemical impedance spectroscopy, potentiodynamic- and potentiostatic-polarization tests). The feasibility of the use of the potentiostatic polarization test as a method for the acceleration of actual corrosion behavior was proved by the corrosion potential and the pitting potential of each layer in the potentiodynamic polarization test. The optimal applied potential for the potentiostatic polarization test was selected through comparison of the surface and electrochemical impedance spectroscopy analysis.
Ferritic stainless steels are widely used for the main materials that constitute the cold end parts of automobile exhaust systems, such as main muffler and tail pipe, because of a high durability, excellent thermal properties, and an effective corrosion resistance.1–5) Due to these advantages, the demand from the automobile industry for ferritic stainless steels consistently increases. Localized corrosion, however, can affect stainless steel when it is exposed to either inner-environment condensed water or outer-environment de-icing salts.6,7)
For protection against this problem, hot-dip aluminizing that protects the surface of stainless steel was developed. Because the corrosion potential of the aluminized layer is relatively lower than that of ferritic stainless steel in terms of general corrosive media,8) the aluminized layer acts as a sacrificial anode and the stainless steel substrate acts as a cathode. As far as the aluminized layer exists, ferritic stainless steel can be protected; therefore, the corrosion resistance of hot-dip aluminized ferritic stainless steel against condensed water and de-icing salts is higher than that of bare ferritic stainless steel.
Nevertheless, pitting corrosion occurs on hot-dip aluminized ferritic stainless steel due to the structure of the interdiffusion layer. During the hot-dipping process, an interdiffusion layer is formed from the reaction between the aluminized layer and the ferritic stainless steel.9–11) Generally, the interdiffusion layer consists of an upper FeAl3 layers and a lower Fe2Al5 layer and numerous cracks and cavities are present in the thin Fe2Al5 layer due to the orthogonal lattice structure.12,13) According to the previous research, the Al2O3 scales were formed on the Fe2Al5 cracks in an NaCl environment14), and the substrate could be corroded as the Al2O3 scales were destroyed.
From a study of the corrosion mechanism between the hot-dip aluminized layer and the steel substrate, Graeve15) found that the open-circuit potential of the interdiffusion layer is higher than that of the steel; therefore, the corroding of steel is preferential when the defect is deep enough to reach the steel substrate. In a similar context, pits can occur in ferritic stainless steel due to the difference of the open-circuit potential with the interdiffusion layer after the entire aluminized layer has dissolved in the solution. Once the pits are initiated, the exhaust system can be broken by the propagation of the pits.
To prevent the destruction of exhaust systems that is caused by pitting corrosion, the development of a method that can be used to judge the speed of the corrosion occurrence in a vehicle exhaust system that is composed of hot-dip aluminized ferritic stainless steel is required. Electrochemical acceleration is the most commonly used method for the identification of the characteristics of corrosion and the life span of a variety of coated panels.16–18) Streicher19) studied pitting corrosion with respect to type 316 stainless steel according to the current-controlled corrosion acceleration test for which the number of pits in the NaCl environment are counted; however, it is difficult for this study to determine the way that this acceleration test corresponds to the actual corrosion behavior.
A proper quantitative corrosion acceleration method for the evaluation of the multiple mixed layers in the de-icing salt environment at the outer surface of a vehicle exhaust system has not been developed. Therefore, the main purpose of this paper is the evaluation of the life span of hot-dip aluminized ferritic stainless steel through the use of electrochemical tests that were conducted in the synthetic de-icing salt environment.
Hot-dip aluminized ferritic stainless steel is the specimen that was used in this study and the chemical composition of the corresponding 409L stainless steel is given in Table 1. To produce the hot-dip aluminized ferritic stainless steel, the 409L stainless steel was immersed in a molten Al-10 mass% Si bath under a temperature that is above 600℃; the addition of silicon is for the flatting of the interface of the intermetallic layer.20,21) To improve the adhesion of the Al-Si coating layer and to prevent the oxidation of the stainless steel on the surface, H2 gas was injected into the stainless steel surface during the hot-dipping process.
| C | N | Si | Mn | P | S | Cr | Ni | Ti | Fe |
|---|---|---|---|---|---|---|---|---|---|
| 0.006 | 0.008 | 0.55 | 0.25 | 0.02 | 0.001 | 11.2 | 0.12 | 0.2 | Bal. |
The hot-dip aluminized ferritic stainless steel is composed of the following three layers: aluminized layer, interdiffusion layer, and 409L stainless steel substrate. The interdiffusion layer was exposed by a dipping in a 1 M NaOH solution for 1 hour that removed the aluminized layer. The interdiffusion layer was selectively removed using nitric acid and the substrate of the aluminized stainless steel was exposed.
All of the experiments were examined using the same specimen of the exposed 1 cm × 1 cm area. The 5 mass% NaCl solution was used as the test solution, and a water bath was used to continually maintain the solution temperature of 60℃.
2.2 Potentiodynamic polarization testTo observe the corrosion behaviors of the three layers and to determine the applied potential for the electrochemical acceleration, potentiodynamic polarization tests for the aluminized layer, interdiffusion layer, and 409L stainless steel substrate were progressed. The aluminized stainless steel was used to obtain the corrosion characteristics of the aluminized layer because it is positioned at the most outer site of the aluminized stainless steel.
Before the potentiodynamic polarization tests were commenced, the aluminized stainless steel and the bare stainless steel were immersed for 2 hours to reach the stable open-circuit potential (OCP); likewise, the interdiffusion layer was immersed for 6 hours. Each experiment was carried out using the EG&G Model 273A potentiostat, and the potential was swept from an initial potential of −0.25 V vs. OCP to a final potential of 1.6 V vs. saturated calomel electrode (SCE) at a scan rate of 0.166 mV/s.
2.3 Real-time and electrochemical acceleration monitoring testsThe corrosion potential of the hot-dip aluminized ferritic stainless steel was measured to investigate the real-time corrosion tendency in a 5 mass% NaCl solution. A VMP2 multichannel potentiostat was used to measure the corrosion potential every 5 hours in a three-electrode electrochemical system. An SCE and two pure graphite rods were used as the reference and the counter electrodes, respectively. The specimen was immersed in a 1 L pyrex cell filled with the 5 mass% NaCl solution while electrochemical impedance spectroscopy (EIS) was simultaneously conducted, and the duration is 400 hours. The frequency range is from 100 kHz to 10 mHz, and the sinusoidal amplitude is 0.03 V. Using the ZSimpWin fitting program, the surface characteristics and the variation of the polarization resistance over time were obtained from the EIS results.
To accelerate the corrosion of the aluminized stainless steel, another potentiostatic polarization test was carried out. The three different potentials −0.4 V vs. SCE, −0.3 V vs. SCE, and −0.2 V vs. SCE that were selected from the results of the potentiodymanic polarization test were applied to the aluminized stainless steel using the VMP2 multichannel potentiostat. For the electrochemical acceleration, each potential was applied until 10 C was reached. The area under the current-time curves is the applied charge and can be obtained by the following equation:
| \[Q = \int Idt\] | (1) |
The chemical composition of the interdiffusion layer was investigated using energy dispersive X-ray spectroscopy (EDS) after the aluminized layer was removed by the NaOH solution. Scanning electron microscopy (SEM) was used for the surface morphology analysis that was conducted after the electrochemical tests. To find out the optimal applied potential, the pit depth that had formed on each surface was measured according to the cross-sectional image. After the acceleration test, a corrosion product that had formed on the surface of the specimen was analyzed by X-ray diffraction (XRD) and the scan rate is 3 degrees/min.
Figure 1 shows the cross-sectional SEM image of the aluminized stainless steel. This result allowed for an observation of the aluminized layer, interdiffusion layer, and stainless steel, and the boundaries between the three layers in the figure are indicated with white lines. With the employment of the EDS analysis, Table 2 shows the composition of the intermetallic layers as a function of the position from line 1 to line 8 that are given in Fig. 1. The results of Fig. 1 and Table 2 suggest that two intermetallic layers exist between the aluminized layer and the 409L stainless steel. During the hot-dipping process, the Cr and the Fe in the stainless steel are diffused to the aluminized layer, while the Al and the Si in the aluminized layer are diffused to the substrate. For this reason, the interdiffusion layer was formed and the various compositions were distributed between the aluminized layer and the bare 409L stainless steel. The effect that the interdiffusion layer exerts on corrosion behavior can therefore be expected. El-Mahallawy22) reported that various AlFeSi phases are formed at different compositions in Al-Si baths (Al20Fe7Si, Al19Fe8Si, Al7Fe2Si, and Al3Fe2Si). Cheng23) found that the intermetallic layers are composed of outer and inner layers, which are τ5(H)-Al7(Fe,Cr)2Si and FeAl3/τ1-(Al,Si)5Fe3/Fe2Al5, respectively, due to the diffusion of the Cr during the hot-dipping process in an Al-10 mass% Si bath. Wu24) explained that Al5(Fe,Cr)2 and other intermetallic compounds were formed between the aluminized layer and FeCrAl foils due to the infiltration of the Al.
Cross-sectional SEM image of hot-dip aluminized stainless steel.
| Line | Al | Si | Cr | Fe |
|---|---|---|---|---|
| 1 | 12.63 | 87.37 | ||
| 2 | 26.42 | 4.13 | 9.06 | 60.39 |
| 3 | 37.79 | 5.65 | 6.11 | 50.45 |
| 4 | 62.75 | 11.57 | 3.69 | 21.98 |
| 5 | 69.03 | 10.27 | 3.25 | 17.46 |
| 6 | 71.68 | 9.93 | 2.48 | 15.91 |
| 7 | 70.51 | 11.24 | 1.33 | 16.91 |
| 8 | 71.73 | 12.55 | 15.72 |
Figure 2 shows the intermetallic outer layer and inner layer of the aluminized stainless steel that appears after the dipping in the 1 M NaOH solution that removed the aluminized layer. Because the average ratio of Al:Fe:Si from line 4 to line 7 in Fig. 1 is ca. 7.0:1.8:1.1, the intermetallic outer layer consists of a τ5-Al7(Fe,Cr)2Si dendritic structure; moreover the inner layer is below the outer layer in Fig. 2. As a result of the average ratio of Al:Cr (5.8:2.0) in line 2 of Fig. 1, it is expected that the inner layer compound is Al5(Fe,Cr)2. Since a number of cracks are formed on the inner layer, the substrate can be exposed to the solution through these cracks.
SEM image of the two different intermetallic layer.
Figure 3 shows the potentiodynamic polarization curves of the three different layers in the 5 mass% NaCl solution. The corrosion potential (Ecorr) and the pitting potential (Epit) of the three layers are listed in Table 3.
Potentiodynamic polarization curves of three different layers in 5 mass% NaCl solution.
| Ecorr (V vs. SCE) | Epit (V vs. SCE) | |
|---|---|---|
| Aluminized stainless steel | −0.788 | - |
| Interdiffusion layer | −0.423 | −0.041 |
| 409L stainless steel | −0.426 | −0.204 |
Active behavior appeared in the aluminized stainless steel until the entire aluminized layer was dissolved into the solution. Due to the exposure of the interdiffusion layer and the stainless steel, the current density decreases at −0.638 V to −0.321 V vs. SCE. The current density then increased rapidly, indicating that pits had been generated on the stainless steel.
This study found that the passive behavior and the corrosion potentials of the interdiffusion layer and stainless steel are similar, while the pitting potential of the interdiffusion layer is higher than that of stainless steel; therefore, pits can be preferentially generated on stainless steel when the interdiffusion layer and the stainless steel are simultaneously exposed to a 5 mass% NaCl solution.
3.3 Immersion testThe corrosion potential and the polarization resistance of the immersed aluminized stainless steel are presented in Fig. 4. Through a comparison with the potential levels in Fig. 3 and Fig. 4, it is possible to identify the layer that is exposed on the surface in each section of the immersion test.
Corrosion potential and polarization resistance as a function of the immersion test.
In section (a), the aluminized layer was preferentially dissolved for 180 hours. The average corrosion potential of the aluminized layer during the 180 hours is −0.71 V vs. SCE, which corresponds to the corrosion potential of the aluminized layer in the potentiodynamic test; therefore, the aluminized layer was dissolved in the 5 mass% NaCl solution, while the corrosion potential is constant.
In section (b), the potential increased rapidly to approximately −0.52 V vs. SCE after the 180 hours. It is suggested that section (b) is the initiation point of the exposing of the interdiffusion layer, while a high current density at a certain defective site caused the dissolving of the aluminized layer. The remaining aluminized layer was consistently dissolved, whereas the exposed interdiffusion layer was not corroded because the corrosion potential of the interdiffusion layer is higher than that of the aluminized layer as shown in Fig. 3.
In section (c), the potential was maintained at approximately −0.5 V vs. SCE, which is near the corrosion potentials of the interdiffusion layer and 409L stainless steel. It is further suggested that most of the coated layer was dissolved while the interdiffusion layer remained on the surface. After 315 hours, the corrosion potential had fluctuated widely because a passive film had formed on the stainless steel.25) The maximum point of the corrosion potential near the 350 hours is −0.372 V vs. SCE, which is higher than that of the corrosion points near the interdiffusion layer and the 409L stainless steel; at that time, the interdiffusion layer could be corroded and the stainless steel was exposed. Lastly, the pits were generated on the stainless steel during the potential drop in section (d).
Figure 5 (a) shows the Nyquist plots as a function of the immersion time that are fitted to the equivalent circuit of Fig. 6. The elements of the equivalent circuit are expressed as follows: Rs is the solution resistance, Rel is the electrical resistance of a disturbing ion movement of the exposed layer, CPE1 is the dielectric strength of the exposed layer surface, Rct is the resistance of metal dissolution reaction and CPE2 is the capacitance that originated from the electric double layer between the solution and the metal. CPE (constant phase element) is usually used to express the frequency dependence of a nonideal capacitor that is caused by a distribution of the current due to the surface inhomogeneity.26) Rel and Rct changed during the immersion time because different layers were exposed on the surface. The values of Rs, Rel, Rct, and Rp as a function of the immersion time are presented in Table 4.
(a) Nyquist plots and (b) Bode plots as a function of the immersion time.
Equivalent circuit model for hot-dip aluminized stainless steel in 5 mass% NaCl solution.
| Immersion time (hours) | Rs | Rel | Rct | Rp |
|---|---|---|---|---|
| 100 | 0.9 | 23.8 | 284.7 | 308.5 |
| 200 | 2.1 | 128.9 | 214.5 | 343.4 |
| 300 | 3.2 | 120.4 | 337.6 | 458.0 |
| 350 | 3.2 | 144.8 | 2106.0 | 2250.8 |
| 400 | 1.0 | 131.1 | 656.3 | 787.4 |
Figure 5 (b) presents the Bode plots (phase vs. frequency) that are useful for an investigation of the phase change. The low frequency spectra detected the change of the processes between the interfaces of the exposed metal or the oxide layer, while the high frequency spectra detected local surface defects.27) The corrosion mechanism of the aluminized stainless steel in the 5 mass% NaCl solution can be interpreted by using the Nyquist plots and the Bode plots. Figure 7 shows the corrosion mechanism processes of hot-dip aluminized stainless steel as six steps.
Schematics of corrosion mechanism of hot-dip aluminized stainless steel as a function of time in 5 mass% NaCl solution: (a) 0 hour, (b) 100 hours, (c) 200 hours, (d) 300 hours, (e) 350 hours, and (f) 400 hours.
Regarding the 100 hours of immersion that were measured, Rel is the resistance of Al2O3 and Rct is the charge transfer resistance of the aluminized layer. Unlike other immersion time, only 100 hours of immersion represents the inductance at low frequency. The inductance loop is associated with the weakening of the protective effectiveness of the Al2O3 and aluminized layer due to the anodic dissolution reaction.28) At 200 hours, the increase of Rel is significant while the decrease of Rct is slight when compared to the 100 hours data, and the sign of the phase angle changed to positive at low frequencies. The above findings indicate that the outer interdiffusion layer was partially exposed on the surface, as shown in Fig. 7 (c) and this makes it difficult to move the ion on the surface; this is the reason for the increase of Rel. Meanwhile, the maximum phase angle decreased at high frequencies with an increasing of the immersion time from 100 hours to 200 hours, indicating that a galvanic couple was formed between the exposed interdiffusion layer and the remaining aluminized layer. Consequently, the electric charge on the remaining aluminized layer was increased and Rct was reduced.
At 300 hours, Rct increased slightly because the interdiffusion layer became exposed on the entire surface after the whole aluminized layer was dissolved in the solution, as shown in Fig. 7 (d). The maximum phase angle increased rapidly at low frequencies when the immersion time increased from 300 hours to 350 hours, indicating that the Rct increase is due to the formation of the stable passive film on the stainless steel25), as shown in Fig. 7 (e).
At 400 hours, the Rct value is much lower than that at 350 hours because of the generated pits on the stainless steel, as shown in Fig. 7 (f). The chloride ions in the solution caused the corrosion of the inner interdiffusion layer, and this was followed by the formation of the passive film on the exposed surface of the stainless steel. The sufficient absorption of the chloride ions on the passive film, then acts to break down the passive film.29) Because the pitting potential of stainless steel is lower than that of the interdiffusion layer, as shown in Fig. 4 and Table 3, most of the pits that occurred on the stainless steel are exposed under the cracked lines of the interdiffusion layer.
3.4 Potentiostatic polarization testFigure 8 shows the results of the corrosion potential and the polarization resistance from the potentiostatic polarization test. The corrosion potential and the polarization resistance decreased dramatically because the generation of the pits on the surface of the stainless steel is common. However, the total amount of the charge that generates the pits, QT, is different according to each applied potential, and when the higher potential was applied, a lower QT was needed.
Corrosion potential and polarization resistance as a function of integrated charge: (a) −0.2 V vs. SCE, (b) −0.3 V vs. SCE, and (c) −0.4 V vs. SCE.
Figure 8 (a) could not be divided into four sections like the immersion test of Fig. 4. Because the applied potential of −0.2 V vs. SCE is higher than the pitting potential of stainless steel, the underlying stainless steel beneath the interdiffusion layer was attacked by the force of the electric field even though the aluminized layer persisted on the surface. As a consequence, the aluminized layer could not serve as the sacrificial anode during the corroding of the substrate. Therefore, the applied potential of −0.2 V vs. SCE is inappropriate for corrosion acceleration.
Alternatively, Figs. 8 (b) and (c) could be divided into four sections. Table 5 shows the charge for the simulation of the immersion test during the potentiostatic polarization test. The required charge increased with the decrease of the applied potential in section (c), indicating that a higher charge is needed to break down the passive film of the stainless steel as the lower voltage is applied. Because the applied potentials of −0.3 V vs. SCE and −0.4 V vs. SCE are higher than the corrosion potentials of the interdiffusion layer and 409L stainless steel, the anodic polarization occurred on the surface of the specimen.29) The higher the applied voltage, the greater the quantity of electrons that flow into the surface, and a more positively-charged double layer was therefore formed on the surface. The applied potential of −0.3 V vs. SCE is dragging more chloride ions than that of −0.4 V vs. SCE by a high electric force. Lastly, a large amount of chloride ions are absorbed on the passive film, and the passive film was quickly broken when the higher voltage was applied.30–33); therefore, if a high potential is applied by a potentiostatic polarization test, pits could be generated by just a small amount of charge. These results show that an application of the −0.3 V vs. SCE and the −0.4 V vs. SCE of the potentiostatic-polarization test could reduce the time of the immersion test by approximately 88% and 77%, respectively, as shown in Table 6.
| Section (a) | Section (b) | Section (c) | Section (d) | ||
|---|---|---|---|---|---|
| Immersion test (hours) | 180 | 240 | 380 | 400 | |
| Potentiostatic polarization test |
−0.3 V vs. SCE (C) |
50 | 120 | 200 | 220 |
| −0.4 V vs. SCE (C) |
70 | 120 | 370 | 400 | |
| Immersion test | Potentiostatic polarization test | ||
|---|---|---|---|
| −0.3 V vs. SCE | −0.4 V vs. SCE | ||
| Time (hour) | 400 | 47 | 92 |
Figures 9 and 10 show a comparison of the surface morphologies of the 200 hours and 300 hours immersed specimens to the −0.3 V vs. SCE and the −0.4 V vs. SCE of the potentiostatic polarization test, until the proper charge is applied according to those that are required for 200 hours and 300 hours, respectively, as shown in Table 7. Since the applied potential of −0.2 V vs. SCE is unsuitable for corrosion acceleration, it was excluded from the surface analysis.
SEM images of specimen surface after (a) immersion for 200 hours, (b) applied charge of 60 C with −0.3 V vs. SCE, and (c) applied charge of 80 C with −0.4 V vs. SCE.
| Immersion time (hour) | −0.3 V vs. SCE (C) | −0.4 V vs. SCE (C) |
|---|---|---|
| 200 | 60 | 80 |
| 300 | 180 | 280 |
| 400 | 220 | 380 |
In Fig. 9, all of the compounds on the surface of the three specimens seemed to be of a dendritic structure, showing that the applied voltages of −0.3 V vs. SCE and −0.4 V vs. SCE are appropriate for a simulation of the 200 hours immersion test. These results correspond to the corrosion progress of the immersion test in Fig. 7 (d).
Figure 10 shows the structure of the inner interdiffusion layer. From these results, it is obvious that the applied voltage of −0.3 V vs. SCE is more compatible with the 300 hour-immersed specimen than that of −0.4 V vs. SCE because the surface of the specimen that was applied to −0.4 V vs. SCE was partially covered with oxide film.
SEM images of specimen surface after (a) immersion for 300 hours, (b) applied charge of 180 C with −0.3 V vs. SCE, and (c) applied charge of 280 C with −0.4 V vs. SCE.
The XRD patterns of the surface after the application of 280 C to −0.4 V vs. SCE, as shown in Fig. 10 (c) are presented in Fig. 11; here, it is suggested that the surface is mainly composed of Fe-Cr and AlFeO3. The AlFeO3 is formed by the reaction of Al and Fe ions in solution with oxygen, which needed the sufficient amount of ions in solution. In Fig. 3, the aluminized stainless steel represents the passive behavior at −0.4 V vs. SCE but trans-passive behavior at −0.3 V vs. SCE. Thus, the localized corrosion occurs at −0.4 V vs. SCE but uniform corrosion occurs at −0.3 V vs. SCE. Based on the results of Fig. 1 and Table 2, the concentration of Fe ion is increased as the corrosion occurs in the depth direction but not in the plane direction. Since the localized corrosion accelerates the corrosion in the depth direction, the applied potential of −0.4 V vs. SCE solution may contain enough Fe ions than −0.3 V vs. SCE. However, the applied potential of the −0.4 V vs. SCE with 80 C did not form AlFeO3. This is because the small amount of Fe ions exist when the outer interdiffusion layer are formed on surface. Thus, the AlFeO3 is only formed at the applied potential of the −0.4 V vs. SCE with 280 C.
XRD analysis of specimen surface shown in Fig. 10 (c).
Figure 12 shows the SEM images of the cross section around the pits on the stainless steel and the pit depth of each specimen. The pit depth at the applied potential of −0.3 V vs. SCE is similar to that of the immersion test. These results indicate that, for the acceleration of the corrosion of hot-dip aluminized stainless steel, the potential of −0.3 V vs. SCE is more suitable than the potential of −0.4 V vs. SCE.
Cross-sectional SEM images of specimens after (a) immersion for 400 hours, (b) applied charge of 220 C with −0.3 V vs. SCE, and (c) applied charge of 380 C with −0.4 V vs. SCE.
In this study, an accelerated-corrosion testing method regarding hot-dip aluminized ferritic stainless steel in 5 mass% NaCl solution was investigated through the performance of electrochemical experiments. The results of the electrochemical tests and the surface analysis led to the formulation of the following conclusions:
