**3. Results and Discussion**

Figure 1 shows the potentiodynamic polarization behaviors of Alloy 600 in the solutions of MRI 0.1, 1, and 10 at a total sodium and chloride concentration of 0.011 M. The corrosion potentials (*Ecorr*) and corrosion rates (*icorr*) of Alloy 600 were significantly decreased as the MRI increased from 0.1 to 10. The active–passive transition appeared at the MRI 0.1, but the alloy was passivated without active dissolution at the MRI 1 and 10.

Figure 2 shows the SEM micrographs of corroded surfaces after the anodic polarization scans. The surfaces exposed to the MRI 0.1 and 1 solutions showed extensive pitting, whereas the surface at the MRI 10 was uniformly corroded without pitting corrosion. Therefore, the transpassivity showing an abrupt increase of current at about 0.260 and 0.390 V in the solutions of the MRI 0.1 and 1, respectively, was due to pitting, whereas the current increase at 0.600 V in the solution of the MRI 10 was owing to oxygen evolution, the reaction of which can be expressed by the following equation [20]:

$$\text{H}2\text{H}\_2\text{O} = \text{O}\_2 + 4\text{H}^+ + 4\text{e} \tag{2}$$

**Figure 1.** Polarization curves of Alloy 600 in the solutions of the MRI 0.1, 1 and 10 at a total sodium and chloride ion concentration of 0.011 M.

**Figure 2.** Scanning electron microscopy (SEM) micrographs showing the corroded surfaces of Alloy 600 after polarization scans at (**a**) the MRI 0.1, (**b**) the MRI 1, (**c**) magnification of the pit denoted by the white arrow in (**b**), and (**d**) the MRI 10.

As shown in Figure 3, the corrosion potentials of SA508 were also significantly decreased as the MRI increased from 0.1 to 10. SA508 actively dissolved at high corrosion rates without any passivation at the MRI 0.1 and 1, whereas the alloy showed the lowest corrosion rate with a passive behavior in a potential range of −0.190 to 0.600 V at the MRI 10. SA508 also showed an abrupt increase of current density due to oxygen evolution near a potential of 0.600 V at the MRI 10, as did Alloy 600.

**Figure 3.** Polarization curves of SA508 in the solutions of the MRI 0.1, 1 and 10 at a total sodium and chloride ion concentration of 0.011 M.

Figure 4 shows the SEM micrographs of SA508 surfaces after the anodic polarization scans. The surface exposed to the MRI 0.1 was severely and uniformly corroded enough to dissolve out the grinding marks, which was made by emery paper during the surface finishing process. The surface at the MRI 1 was also corroded uniformly, but less severely than at the MRI 0.1. On the contrary, it can still clearly be seen the grinding marks on the surface exposed to the MRI 10, indicating that the anodic dissolution rate was very low in the MRI 10 solution. Therefore, the morphologies of these corroded surfaces were in good agreement with the polarization behaviors shown in Figure 3.

**Figure 4.** SEM micrographs showing the corroded surfaces of SA508 after polarization scans at (**a**) the MRI 0.1, (**b**) the MRI 1, and (**c**) the MRI 10.

The important corrosion parameters from Figures 1 and 3 are summarized in Table 4. Alloy 600 showed the highest corrosion rate at the MRI 0.1, while the corrosion rates at the MRIs 1 and 10 were nearly similar. In case of SA508, this alloy also showed the highest corrosion rate at the MRI 0.1, but the corrosion rate at the MRIs 10 was rather smaller than that at the MRI 1. Consequently, this result indicates that the molar ratio control method is beneficial only when the crevice chemistry has a low MRI and pH.

**Table 4.** Corrosion potentials and corrosion rates of Alloy 600 and SA508 obtained from the polarization tests.


As shown in Tables 1 and 2, Alloy 600 is a high-alloyed steel containing 15.7 wt.% Cr and 73.7 wt.% Ni. Thus, this alloy has an excellent resistance to corrosion in overall pH ranges from acidic to alkaline. Therefore, the difference between the corrosion rates (*icorr*) at the acidic MRI 0.1 and at the alkaline MRI 10 is not so large. However, SA508 is an iron-based steel containing only 0.23 wt.% Cr and 0.58 wt.% Mo and thus has a basically poor corrosion resistance, especially in acidic solutions. From Table 4, the corrosion rates (*icorr*) of Alloy 600 were always significantly lower than those of SA508 in all the test conditions. In addition, there was a significant decrease of *icorr* for SA508 at the MRI 10 in comparison with the MRI 1 as well as the MRI 0.1. The reason for this can be attributed to the fact that the solubility of magnetite is significantly dependent on the pH of a solution [21,22]. The solubility of magnetite at pH 3 is about 5 <sup>×</sup> <sup>10</sup><sup>4</sup> times higher than that at pH 12 in water at 100 ◦C [21]. Therefore, the corrosion rate of SA508 increases significantly in low pH solutions (i.e., low MRI solutions) with a high solubility of magnetite because the corroding surface of the alloy cannot be protected by the magnetite film. Conversely, the alloy showed a passive behavior with a low corrosion current in a potential range of −0.190 to 0.600 V at the MRI 10 solution of pH 12, owing to a significantly low solubility of magnetite.

Figure 5 shows the potentiodynamic polarization behaviors of Alloy 600 and SA508 in the solutions with total sodium and chloride ion concentration of 0.011 M and 0.11 M at a constant MRI 1. The corrosion potentials of Alloy 600 and SA508 were approximately −0.450 V and −0.730 V at both concentrations, respectively, indicating that the corrosion potentials of the two materials were not affected by a change in the total ion concentration. The cathodic and anodic current density of Alloy 600 was also little affected by an increase of the ion concentration from 0.011 M to 0.11 M. However, the pitting potentials of Alloy 600 decreased from 0.390 V in 0.011 M to 0.170 V in 0.11 M. In case of SA508, the polarization current density was nearly same in the solutions of 0.011 M and 0.11 M, when this alloy was polarized around the corrosion potential. The above results mean that the electrochemical corrosion behavior of these materials in a region near the corrosion potentials does not depend on the total sodium and chloride ion concentrations if the sodium to chloride molar ratio in a solution is the same. Similar behaviors were also observed at the MRI 0.1 and 10.

**Figure 5.** Polarization curves of Alloy 600 and SA508 in the 0.011 and 0.11 M solutions at a constant MRI 1.

From Figure 1, Figure 3, and Figure 5, it is clear that the corrosion potential of SA508 is always lower than that of Alloy 600 in each test condition. The anodic curve of SA508 also intersects with the cathodic curve of Alloy 600. This result demonstrates that SA508 is an anodic member of the galvanic couple and its corrosion rate is accelerated, when SA508 and Alloy 600 are electrically contacted. When the two materials are coupled in equal area, the galvanic current density (*icouple*) of SA508, acting as an anode, is determined at the intersection of the anodic curve of SA508 and the cathodic curve of Alloy 600. Figure 6 shows the effect of the MRIs on the galvanic corrosion of SA508, based on the polarization curves. Upon coupling to an equal area of Alloy 600, the current density (*icouple*) of the

coupled SA508 was increased by about 2~6 times compared to that (*iSA508*) before coupling. However, the area of the tubesheet around a tube is much smaller than that of the tube in actual SGs, because the tubes are densely inserted into the tubesheet to increase the heat transfer area. Consequently, the corrosion rate of SA508 would be more accelerated by the effect of small anode (SA508) and large cathode (Alloy 600). In addition, the galvanic corrosion rate of SA508 was little changed by the total ion concentration at a fixed MRI as shown in Figure 6. This is because the polarization current density of the two materials was not affected by an increase in the total ion concentration from 0.011 M to 0.11 M, as shown in Figure 5.

**Figure 6.** Effect of the MRIs on the galvanic corrosion rate of SA508 coupled to an equal area of Alloy 600.

An SG tube can be slowly deformed by volume expansion of corrosion products due to corrosion of the tube support materials adjacent to and around the tube, which is called denting. The denting was attributed to concentration of chlorides and oxidants such as copper, in the crevices, leading to rapid corrosion of the tube support materials [2,23]. SG tubes are expanded into the tubesheet of SA508, a dissimilar metal. Therefore, based on the results obtained in this work, it is worth mentioning that corrosion of the tubesheet is accelerated by the galvanic coupling itself without concentration of chemical impurities in the crevices.
