The material used in this work is T91 steel. Its chemical compositions are as follows (wt.%): C 0.08, Mn 0.30, Si 0.20, S < 0.010, P < 0.020, Cr 8.00~9.50, Ni < 0.40, Mo 0.85, Nb 0.06~0.10, V 0.18, N 0.03, Fe balance. The material was processed into a size of 10 mm × 10 mm × 3 mm, and then sealed with epoxy resin to expose an area of 10 mm × 10 mm for electrochemical measurements. According to the national standard GB5776-86, the working surfaces of the specimens were polished into a mirror using a series of waterproof abrasive papers, and cleaned with acetone and alcohol in sequence. After that, the specimens were stored in a desiccator for use.

The experimental solution used in this work is a series of NaHSO₃ solutions with different concentrations, with concentrations and corresponding pH values showing in Table 1. It can be seen that the pH of the solutions are between 4.1 and 5.1, which are suitable for simulating the moderate acid rains.

In the experiment, electrochemical impedance spectroscopy (EIS) and Tafel polarization were used to study the electrochemical performance of T91 steel in the solutions. All electrochemical measurements were performed in a classic three-electrode system, with a saturated calomel electrode (SCE) as a reference electrode, a platinum electrode as an auxiliary electrode, and a T91 specimen as a working electrode. All measurements were performed on a CS350 electrochemical test system produced by Wuhan Coster Equipment Co., Ltd. The frequency range of EIS measurement was from 100 kHz to 10 mHz, and the excitation signal was a sine wave voltage signal with an amplitude of 10 mV. The EIS results were modeled and fitted by a ZSimpWin software. The scanning range of Tafel polarization curve was ± 250 mV vs. open circuit potential (OCP), and its scan rate was 0.25 mV s⁻¹. The experimental results were simulated by Origin software package. The EIS and Tafel polarization tests in this experiment were repeated once, and the average value or typical value was written into the paper. The potentials reported in the paper were all relative to SCE.

After immersed for 504 hours, the specimens were taken out from the solutions, washed with absolute ethanol for 10 minutes, dried with cold air, and stored in a vacuum desiccator. The morphologies and compositions of the corrosion products on the steel surfaces were analyzed by a SEM (JSM-6360LV) with an energy disperse spectroscopy (EDS, NECA 250). After that, the surface corrosion products were removed by the rust remover to expose the corrosion morphologies, and the steel surface were observed again by the SEM. The rust remover used is composed of 500 ml hydrochloric acid (37 wt.%) + 500 mL deionized water + 3.6 g hexamethylenetetramine.

Fig. 1 shows SEM micrographs of corrosion products on the steel surface after 504 h immersion in the NaHSO₃ solutions, and the corresponding EDS results are listed in Table 2. After immersed in 0.01% NaHSO₃ solution for 504 h, the steel surface has been completely covered with dense corrosion products (Fig. 1a), and there are obvious cracks in the inner layer of the corrosion products (Fig. 1b), which may stem from the dehydration process during the post-treatment of the specimens. The color of the corrosion products in this layer is deep, containing elements such as O, Fe, Cr and S (Table 2), and its main components are oxides and sulfides of iron and chromium. The outer layer of the corrosion product is composed of some light-color materials, with the similar main components as the inner layer, but its oxidation content is slightly higher than the later. As for in 0.1% NaHSO₃ solution, the corrosion products have completely covered the steel surface and cracked (Fig. 1c). Compared with the corrosion products in 0.01% NaHSO₃ solution, the number and width of the cracks in the corrosion products are increased (Fig. 1c), indicating that the mechanical properties of the corrosion products are poorer than the former. A layer of mantle like materials is distributed on the corrosion products (Fig. 1d), and it is composed of oxides and sulfides of iron and chromium (Table 2). When the NaHSO₃ concentration increased to 1.0%, massive cracks appears in the corrosion products, and some corrosion products fall off from the steel surface, which indicates that the mechanical properties of the corrosion products are the worst in the 1.0% NaHSO₃ solution (Fig. 1e). Intact granular corrosion products exist in the cracks and the fall-off region (Fig. 1f). The composition of granular corrosion products is close to other corrosion products on the surface, with high contents of chromium and sulfur (Table 2).

The above results show that after immersed in NaHSO₃ solutions for 504 h, complete corrosion products with high contents of chromium and sulfur generate on the steel surface. The compactness and mechanical properties of corrosion products decreases with the concentration increase of NaHSO₃ in the solutions.

Fig. 2 shows the corrosion attack morphologies of the steel after removing the corrosion product. In the 0.01% and 0.1% NaHSO₃ solutions, corrosion morphologies of the steel surface are relatively uniform (Figs. 2a and 2c), with many shallow corrosion spots inlaying in it. The corrosion spots look like bird’s nests, and their diameters are about 50 μm (Figs. 2b and 2d). In the above solutions, the steel tends to take place uniform corrosion only with slight corrosion spots.

When NaHSO₃ concentration increased to 1.0%, localized corrosion dominates the steel surface (Fig. 2e). The corrosion spots connect with each other, forming the contiguous localized corrosion spots. Besides, out of the localized corrosion spots, some small pits seem to be forming (Fig. 2f). From the above, it can be concluded that the T91 steel tends to occur uniform corrosion with slight corrosion pits in the low-concentration NaHSO₃ solutions, while the corrosion mode changes to localized corrosion with the increase of NaHSO₃ concentration.

Nyquist plots of the steel immersed in the NaHSO₃ solutions with time are shown in Fig. 3. At the beginning of the experiment, the impedance of the steel firstly increases with the increase of NaHSO₃ concentration, while it decreases after the NaHSO₃ concentration reaching 0.1% (Fig. 3a). What’s more, the impedances of the steel in all solutions remain within one order of magnitude. After 72 h immersion, the impedance of the steel is still very high in the solutions with NaHSO₃ concentration higher than 0.1%, but it becomes very low in the low-concentration solutions. At 144 h, the impedance increases slightly in the low-concentration solutions, decreases sharply in the 0.1% NaHSO₃ solution, and continues to increase in the high-concentration solutions. At 216 h, the impedance of the steel in low concentration solution continues to rise, while that in the high-concentration solution slightly decreased. At 336 h, the impedance in the low-concentration solution is higher than that in the high-concentration solution. At the end of the experiment, the impedance of the steel shows a negative correlation with the NaHSO₃ concentration in solutions.

From the above discussion, it is reasonable to infer that the impedance of the steel is relatively high once the steel is immersed in a NaHSO₃ solution, and it will experience a drop at a certain time during the immersion process. The concentration of NaHSO₃ mainly affects the time of the impedance drop and the subsequent trend of the impedance. The lower the concentration of NaHSO₃ is, the earlier the impedance drop will occur. For example, the impedance drop occurs at 72^nd h in 0.05% and 0.1% NaHSO₃ solutions. In the low-concentration solutions, the impedance increases immediately after the drop, which may be related to the better mechanical properties of the corrosion products forming on the steel surface in the solution.

In order to further study the electrochemical behavior of the steel in NaHSO₃ solution, equivalent circuits were used to simulate the EIS data. Two-constant model is adopted for fitting as shown in Fig. 4. R_s is the solution resistance. R_f and Q_f are the resistance and capacitance of surface corrosion products, respectively. R_ct is the charge transfer resistance, and C_dl is the double electric layer capacitance. W is the Weber impedance, and the corresponding fitting results are shown in Tables 3,–7.

In the following, the evolutions of R_f and R_ct of the steel are used to discuss the experimental results. R_f is the resistance of the surface corrosion products, which can represent the integrity of the products and its blocking effect to corrosive ions. R_f evolution of the steel with time in the NaHSO₃ solutions is shown in Fig. 5. R_f variations can be divided into two types. When the NaHSO₃ concentration is less than or equal to 0.1%, R_f of the steel gradually decreases in the early stage of the experiment, and then slightly increases with time. During the whole immersion process, R_f always decreases with the increase of NaHSO₃ concentration. However, when the NaHSO₃ concentration is 0.5% or 1.0%, R_f of the steel increases slightly in the early stage of the experiment, and then it decreases gradually. What’s more, R_f of the steel in the 1.0% NaHSO₃ solution is higher than that in the 0.5% NaHSO₃ solution in the early stage, while it decreases sharply from 144^th h and is lower than the later in the end stage of the experiment.

Generally, R_ct is inversely related to corrosion rate of metals. Thus, R_ct evolution with time is represented as the R_ct⁻¹ change with time, as shown in Fig. 6. When NaHSO₃ concentration is lower than 0.1%, R_ct⁻¹ of the steel sharply increases, and gradually decreases from the 72^nd or 144^th h to the end of the experiment. However, R_ct⁻¹ of the steel firstly decreases and then rapidly increases from 144^th h in the 0.50% and 1.0% NaHSO₃ solution. Obviously, the corrosion behavior of the steel present almost the opposite modes in low- and high-concentration NaHSO₃ solutions. However, at the end of the experiment, there is a positive correlation between the corrosion rate and the NaHSO₃ concentration. Comparing the evolutions of R_f and R_ct, they display the same evolution pattern, which indicates that the change of corrosion products in the corrosion system plays an important role in corrosion rate.

Tafel curves of the steel immersed in the NaHSO₃ solutions for 504 h are shown in Fig. 7, and the corresponding fitting results are shown in Table 8. The steel shows the characteristics of active anodic dissolution in all solutions, and the cathodic and anodic polarization behaviors are all directly affected by the NaHSO₃ concentration. As shown in Table 8, the corrosion potential E_corr of the steel decreases gradually but the corrosion current density i_corr increases gradually with the increase of NaHSO₃ concentration, indicating that NaHSO₃ accelerates the corrosion process of the steel during the immersion. The cathodic Tafel slope β_c of the steel is in the range of 260~320 mV/dec in the solutions, which is a limited difference. However, the anodic Tafel slope β_a decreases from 390 mV/dec to 110 mV/dec with the increase of NaHSO₃ concentration, and then increases sharply. The polarization resistance R_p of the system can be calculated by combining i_corr and Tafel slopes:

The calculated R_p is also shown in Table 8. The polarization resistance of the system decreases and the corrosion rate increases with the increase of NaHSO₃ concentration at the end stage of the experiment. In order to compare R_p and R_ct, the evolutions of R_p and R_ct as a function of NaHSO₃ concentration after 504 h immersion are presented in Fig. 8, and the results are fitted as shown in a black solid line. It can be seen from the figure that the results of R_p and R_ct are similar, and they are scattered near the model results. The facts indicate that the result of Tafel curve is consistent with that of EIS.