**3. Results**

#### *3.1. Potentiodynamic Polarization Curves*

Figure 1 shows the polarization curves of the bare electrode and S8-coated electrode at different pH levels. In the pH range of 3.10 to 7.18, the polarization curves of the electrode in the corrosive solution were similar. It can be observed that the potential moved toward a more negative direction and the passive range became narrower with the increase in pH, and the Ecorr decreased from −550 to −611 mV (vs. SCE) owing to the deposition of S8 on the electrode. As shown in Figure 1a, the anode was actively dissolved. In contrast to the bare electrode, the Tafel slope (*b*a) of the electrode coated with S8 was smaller, and the cathode coated with S8 was greatly affected by diffusion (Figure 1b). The main reason is that when S8 contacts the metal, the electron transfer between them is more direct and rapid, and the corrosion rate is higher. Based on the Marcus theory [22], the deposition of S8 on the surface of steel will lead to the weakening of the metal–metal bonds, thereby reducing the activation energy barrier for the dissolution of the surface metal atoms.

**Figure 1.** Potentiodynamic polarization curves with 3.5% NaCl at different pH values: (**a**) bare steel, (**b**) S8-coated steel.

The parameters, such as the self-corrosion potential, corrosion current density, and anodic and cathodic Tafel slopes (*b*a, *b*c), are listed in Table 1. With an increase in pH, both the current density and potential decreased, and the current density of the electrode coated with S8 was significantly higher than that of the bare electrode at the same pH. At pH = 3.10, the corrosion current density of the S8-coated electrode was twice as high as that of the bare electrode and reached 14 times at pH = 7.18. The increase in the corrosion current may be related to the hydrolysis of S8. It was proven that the corrosion potential under acidic conditions was more positive than that under neutral conditions; because of the decrease in pH, the acidity of the solution increased, and more electropositive H+ accumulated on the surface of the metal, resulting in a positive shift of the potential. The Tafel slope shows that the *ba* is greater than the *bc*, and the corrosion process is controlled by the anode.


**Table 1.** Parameters calculated by potentiodynamic polarization curve.

#### *3.2. Electrochemical Impedance Spectroscopy*

To further explore the effect of pH on the electrode coated with S8, EIS measurements were systematically performed, and the impedance spectra of the bare electrode and S8 coated electrode at different pH values are presented in Figure 2. The Nyquist spectrum has a capacitance semi-circular arc in the high-frequency area, and the arc center pressure is below the x-axis, which is attributed to the interface charge transfer reaction [23]. As can be seen from Figure 2a,d, there were obvious differences in capacitance diameter at different pH values, and the radius of the arc increased significantly with the increase in pH. This shows that a high pH has larger resistance, which slows down the corrosion process of the electrode surface.

**Figure 2.** Nyquist plots and Bode plots of the bare electrode and S8-coated electrode at different pH values: (**<sup>a</sup>**–**<sup>c</sup>**) bare electrode, (**d**–**f**) S8-coated electrode.

From the Bode diagram (Figure 2b,c), the phase angle vs. frequency curves of the bare electrode conformed to a one-time constant model. For the S8-coated electrode (Figure 2e,f), there were two arcs on the electrode surface, which were distributed in low-frequency and high-frequency areas, respectively, and the high frequency was mainly related to the defects caused by S8 coated on the surface, and the low-frequency region was related to the electrochemical corrosion process of the electrode.

In order to fit the EIS data, all the data were fitted using the appropriate equivalent circuit as given in the insert, in which *R*s is the solution resistance, *R*ct is the charge transfer resistance, *Q*dl is the double-layer capacitance, n is the dispersion index, and *R*f and *Q*f represent the resistance and capacitance of the elemental sulfur deposition layer, respectively [24]. Table 2 lists the fitting results of electrochemical impedance spectroscopy; the Chi-Squared values were between 10−<sup>3</sup> and <sup>10</sup>−4, which suggests that the fitting results are reliable. With the increase in pH, both *R*ct and *R*f showed a rising tendency. The *R*ct of the bare electrode far exceeded the sum of *R*ct and *R*f of the S8-coated electrode. Generally, the *R*ct of the electrode is inversely proportional to the corrosion rate and can be used to characterize the corrosion rate [25]. At pH = 3.10, the *R*ct value of the bare steel electrode was 16 times higher than that of the S8-coated electrode. Therefore, EIS measurements further showed that the corrosion of the electrode coated with S8 was more severe under acidic conditions, and the EIS results were in good agreemen<sup>t</sup> with the results of the polarization curves.


**Table 2.** EIS data fitted electrochemical parameters.

#### *3.3. Potential and Current Distribution*

To further study the localized corrosion behavior under S8 deposition, a WBE was used to reveal the corrosion difference between the bare steel electrode and the S8-coated electrode. Tan et al. [26] first adopted the WBE to study localized corrosion and obtained localized potential and current information. A new parameter, the localized corrosion intensity index (*L*CII), is proposed to quantify the degree of localized corrosion using the following equation:

$$L\text{CII} = \frac{i\_{\text{max}}}{i\_{\text{tot}}} \tag{1}$$

where *imax* is the maximum anode current, and *itot* is the positive galvanic current density of the WBE. Tan et al. [26] presumed that localized corrosion was not serious when the *L*CII index was lower than 0.1.

When sulfuric acid and sodium hydroxide were added to the medium solution as a simple pH modifier to adjust the pH of the solution, the corrosion mechanism changed and the corrosion distribution became highly different, as reflected by the color and scale. For the bare electrode, as shown in Figure 3, all the point current densities were positive, and the potential distribution was uneven and random. The potential difference was relatively small, indicating that the localized corrosion tendency of the WBE was relatively weak. The maximum anodic currents of 0.607, 0.279, and 0.279 A/cm<sup>2</sup> were present at the electrodes (8, 4), (6, 6), and (6, 6), respectively (X-column, Y-row), at pH of 3.10–7.18.

**Figure 3.** Potential (E) and galvanic current density (ig) distribution maps of WBE bare electrode at different pH values ((**<sup>a</sup>**,**b**): pH 3.10), ((**<sup>c</sup>**,**d**): pH 5.15), ((**<sup>e</sup>**,**f**): pH 7.18).

Figure 4 shows the coupling potential and current density distribution maps of the WBE coated with S8. It was found that with the increase in pH, the anode region gradually enlarged, and its position changed. The non-uniformity of the potential distribution reflects the heterogeneity of the corrosion. As shown in Figure 4e, the upper half of the electrode potential is small. In general, the anode position causes the OCP of the electrode to move negatively; therefore, the electrode in the upper half region is used as the anode region, and the corresponding Figure 4f shows a high position current value, which can also be reflected by the color. The other electrode potentials presented similar results.

In the pH range of 3.10–7.18, Figure 5 shows the calculation results of some key parameters (*itot, imax*, *L*CII, and *N*A) based on Figure 3, where *N*A is the number of anodes. When the pH was adjusted from 3.10 to 7.18, both the *ito*t and *imax* decreased, all *N*A values were 100, and a small *L*CII value (*L*CII < 0.1) was obtained, indicating that all bare electrodes experienced homogeneous corrosion, and localized corrosion did not occur. As shown in Figure 6c, as the pH increased, the surface of the S8-coated electrode produced new anode sites, and the *N*A value increased from 57 to 99; therefore, the anode current density increased significantly, and *imax* also increased. This also means that the current distribution became less concentrated on a small number of anode sites, and then generated more active corrosion activities, and corrosion on the electrode surface became more homogeneous. After the pH increased from acidic to neutral, the corrosion pattern changed significantly. The distribution of the anode and cathode sites became random, and the *L*CII value also decreased rapidly from 0.3061 to 0.0295. Obviously, higher *L*CII values correspond to concentrated anode regions and intensive corrosion locations. As shown in Figure 5c, the *L*CII (*L*CII = 0.3061 > 0.1) of the electrode coated with S8 was higher than that of the bare electrode at pH = 3.10, which indicates that the localized corrosion trend of the electrode coated with S8 was enhanced. This result agrees with the polarization curves and EIS results.

An optical picture of the WBE is shown in Figure 7. For the bare electrode (Figure 7a), the current density in the anode region was extremely small, and the corresponding electrode was covered with a negligible amount of black corrosion products, causing mild corrosion. For the S8-coated electrode (Figure 7b), part of the electrode was covered with black corrosion products, and the other was still bright. S8 accelerated the corrosion of the electrode, and the electrode suffered from localized corrosion, which agrees with the calculated results shown in Figures 5 and 7.

**Figure 4.** Potential (*E*) and galvanic current density (*i*g) distribution maps of WBE with S8-coating at different pH values ((**<sup>a</sup>**,**b**): pH 3.10), ((**<sup>c</sup>**,**d**): pH 5.15), ((**<sup>e</sup>**,**f**): pH 7.18).

**Figure 5.** Bare electrode statistical and calculated results for (**a**) the sum of anode current density *itot*, (**b**) the maximum anodic current density *imax*, (**c**) the localized corrosion intensity index *L*CII, and the number of anodes *N*A.

**Figure 6.** S8-coated electrode statistical and calculated results for (**a**) the sum of anode current density *itot*, (**b**) the maximum anodic current density *imax*, (**c**) the localized corrosion intensity index *L*CII, and the number of anodes *N*A.

**Figure 7.** Optical picture of WBE at pH = 3.10: (**a**) bare electrode, (**b**) S8−coated electrode.

#### *3.4. Surface and Component Characterization*

The micromorphology of the corrosion products formed on the bare steel and S8- coated steel at different pH values with an autoclave is shown in Figure 8. The entire surface was uniformly covered by a thick and dense corrosion scale (Figure 8a–c), and there was no obvious localized corrosion. The dense corrosion scale provided protection for the substrate by hindering the corrosive species from penetrating the interface between the corrosion scale and the steel. The surface of the S8-coated steel was covered with an incomplete and loose corrosion scale and dispersed solid particles, and some small holes and recessed areas were observed at pH = 3.10, indicating that the steel suffered from serious localized corrosion attacks. Therefore, the presence of S8 changed the structure of the corrosion scale, leading to a reduction in the resistance of corrosion scale to the corrosive medium. The corrosion of the S8-coated steel was more serious than that of the bare steel, which was consistent with the results of potential polarization.

**Figure 8.** SEM surface morphologies of the samples at different pH values: (**a**) bare steel, 3.10, (**b**) bare steel, 5.15, (**c**) bare steel, 7.18, (**d**) S8−coated steel, 3.10, (**e**) S8−coated steel, 5.15, (**f**) S8−coated steel, 7.18.

To further evaluate the corrosion difference between the bare electrode and S8-coated electrode at pH = 3.10, an OLYMPUS DSX500 optical digital microscope was used to observe and calculate the maximum corrosion depth of the metal substrate after removing the corrosion scale in the bright field (BF) mode. Representative 3D images are shown in Figure 9. To comprehend the severity of the localized corrosion, a pitting factor was proposed to characterize the severity of pitting corrosion [15]:

$$\text{PF} = \frac{p}{d} \tag{2}$$

where *P*F is the pitting factor, p is the deepest metal penetration (μm), and d is the average metal penetration (μm). The *P*F value was positively correlated with the destructiveness of pitting, and the pitting corrosion threat was obvious at *P*F > 5. According to the results in Table 3, the *P*F value of the S8-coated electrode was above 5, whereas that of the bare electrode was at a low level.

**Figure 9.** Three-dimensional profilometry measurement of corroded samples after removing corrosion products at pH = 3.10 for (**<sup>a</sup>**−**<sup>c</sup>**) the bare electrode, (**d**−**f**) S8−coated electrode.


**Table 3.** Three−dimensional profilometry measurement depth at pH = 3.10.

Only a small number of pores and scratches were observed on the bare electrode's surface, and the substrate was relatively flat. Many large pores were found on the surface of the S8 deposition electrode, and the metal substrate was uneven. Therefore, the characteristics of deeper holes and surface roughness indicated that the corrosion of the S8-coated electrode was more serious, which further confirmed the results mentioned above.

Furthermore, Figure 10 shows the corresponding chemical composition of the corrosion scale. The main elements in the deposits were C, O, Fe, and S. Figure 11 illustrates the XRD spectra of the corrosion products of the S8-coated and bare electrodes at pH 3.10. The corrosion scale on the bare electrode was composed of FeCO3, which is a typical product in CO2 systems. FeCO3 is deposited on the electrode surface when the concentration of ferrous ions and carbonate ions surpasses the solubility product of FeCO3 at a certain temperature. The product on the S8-coated electrode was FeS. Therefore, the sediments on the substrate were mainly ferrous carbonate and ferrous sulfide.

**Figure 10.** Elemental analysis (EDS) spectra: (**a**) bare electrode, (**b**) S8-coated electrode.

**Figure 11.** XRD patterns of the bare electrode and S8-coated electrode at pH = 3.10.

#### **4. Corrosion Mechanism**

The presence of S8 and CO2 accelerated corrosion by reacting directly with the metal substrate or forming acid by hydrolysis to further promote corrosion. The corrosion process under S8 deposition mainly included the hydrolysis of S8, formation of the corrosion product (ferrous sulfide), and further catalytic corrosion caused by ferrous sulfide in addition to Cl− [27]. The following equations are proposed to explain the corrosion formation of steel caused by S8 deposition on the electrode surface.

Cathodic reaction:

$$2\text{H}\_2\text{CO}\_3(\text{aq}) + 2\text{e}^- \rightarrow \text{H}\_2(\text{g}) + 2\text{HCO}\_3^- \text{(aq)}\tag{3}$$

$$2\text{HCO}\_3^-\text{(aq)} + 2\text{e}^- \rightarrow \text{H}\_2\text{(g)} + 2\text{CO}\_3^{2-} \text{(aq)}\tag{4}$$

$$2\text{H}\_2\text{O(l)} + 2\text{e}^- \rightarrow \text{H}\_2\text{(g)} + 2\text{OH}^- \text{(aq)}\tag{5}$$

Sy−1·S2<sup>−</sup> + <sup>2</sup>(x − 1) e<sup>−</sup> + 2xH<sup>+</sup> *FeS* → xH2S + Sy−<sup>x</sup> (6)

$$\text{H}^+ + \text{e}^- \rightarrow \text{H}\_{\text{ads}} \rightarrow 1/2\text{H}\_2\tag{7}$$

Anodic reaction process:

$$\text{Fe}(\text{x}-1)\text{Fe} \rightarrow (\text{x}-1)\text{Fe}^{2+} + 2(\text{x}-1)\text{e}^{-} \tag{8}$$

$$\text{Fe}(\text{x}-1)\text{Fe}^{2+} + (\text{x}-1)\text{H}\_2\text{S} \rightarrow (\text{x}-1)\text{FeS} + 2(\text{x}-1)\text{H}^\* \tag{9}$$

$$\text{Fe}^{2+}\text{(aq)} + \text{CO}\_3^{2-} \text{ (aq)} \rightarrow \text{FeCO}\_3\text{(s)}\tag{10}$$

Overall corrosion reaction:

$$\rm{H\_8} + 8H\_2O \rightarrow 6H\_2S + 2H\_2SO\_4 \tag{11}$$

$$\text{H}\_2\text{S} \Longleftrightarrow \text{H}^+ + \text{HS}^- \tag{12}$$

$$\text{H}\text{S}^- \Leftrightarrow \text{H}^+ + \text{S}^{2-} \tag{13}$$

$$\rm{Fe}(x-1)Fe + S\_{y-1} \cdot S^{2-} + 2H^+ \stackrel{Fe\\$}{\rightarrow} (x-1)FeS + H\_2S + S\_{y-x} \tag{14}$$

In the above reaction equations, Fe2+ generated by the substrate entered the solution, and initially combined with CO3<sup>2</sup>− to form FeCO3 or with S2− and SO4<sup>2</sup>− to form FeS and FeSO4. In the CO2-saturated acidic solution, the direct reduction of carbonic acid was the main cathodic reaction, and other possible cathodic reactions, including H+ reduction, were ignored. Sy−1·S2<sup>−</sup>, as a reduced substance under the catalysis of ferrous sulfide, was considered to be formed by the chemical adsorption of sulfide ions on the surfaces of S8 particles, and corrosive substances might also involve adsorbed polysulfate ions, which may be formed by the reaction of S8 with water [14]. It has been reported that acidification occurs in sulfur-containing aqueous systems, and H2SO2, H2SO3, H2SO4, H2S, and polysulfides are species for the hydrolysis products of S8.

At the beginning of corrosion, the surface of the metal substrate was covered by S8, and the corrosion process was a self-catalytic process. The direct contact between the metal substrate and S8 was a necessary factor for corrosion, which would lead to the corrosion potential moving in the positive direction, and it was easy to induce pitting corrosion with defect areas [27]. Then, the Cl− and acid solution contacted the substrate, further aggravating the corrosion. However, some scholars believe that, in the presence of S8, the hydrolysis of S8 to generate acid is the main factor controlling corrosion, and sulfidecatalyzed cathodic sulfur reduction is considered to be the most important step [28]. To better demonstrate the corrosion mechanism of the S8-coated electrode, a pattern was proposed, as shown in Figure 12.

**Figure 12.** Description of corrosion mechanism of L360 pipeline steel under S8 deposition in NaCl solution containing CO2: (**a**) S8 non-uniform deposition, (**b**) under S8 deposition.

The deposition of S8 on the metal surface was non-uniform for the electrode coated without S8, which directly contacted the NaCl solution containing CO2, mainly resulting in the anodic dissolution of the metal and the reduction of H2CO3. The typical corrosion scale of FeCO3 was formed, covering the surface of the metal and preventing Cl− from contacting the metal substrate, which is consistent with the results of XRD and SEM. In addition, with an increase in pH, the solubility of CO2 in the solution decreased, and the stability of the corrosion product FeCO3 was also enhanced. Finally, the corrosion rate slowed with an increase in pH, as shown in the left half of Figure 12a. S8 reacted with the metal substrate and formed the corresponding sulfide, such as FeS (Equation (9)). At the same time, hydrolysis of S8 occurred, and the solution was acidified. With the passage of time, the hydrolysis of S8 and the dissolution of H2CO3 occurred slowly, and the local environment between the S8 deposition layer and metal substrate changed, which was mainly manifested in the excessive pH shifting from neutral to acidic and activated the dissolution of metal. In addition, the hydrolysis products of S8, H2SO4 and H2S, contacted the metal and combined with Fe2+ to form FeS and FeSO4, as shown in Figure 11.

Crevice corrosion was also formed due to the non-uniform deposition of S8, and localized corrosion occurred first at the edge of the deposition of S8. There was a potential difference between the steel substrate and the corrosion products with electronic conductivity, such as FeS and FeCO3, and the substrate covered by S8. Therefore, the potential of the electrode covered by S8 was significantly lower than that of the bare electrode or corrosion products. The metal under S8 deposition was used as the anode corrosion area, and the bare electrode or corrosion products were used as the cathodic protection area, forming a large cathode and a small anode, which caused galvanic corrosion and further accelerated the corrosion under deposits [29,30], as shown in Figure 12a.

At the same time, the environment under the S8 deposition was different from that of the surface of the bare electrode, and the corrosion behavior changed. The self-catalytic cathodic reaction accelerated the corrosion of the substrate during S8 deposition. A selfcatalytic cathodic reaction occurred when the electrode was covered with S8 in an aqueous solution. Ferrous sulfide is an electronic conductor with structural defects and low hydrogen overpotential, so it has a catalytic effect on the cathodic reaction [14], as shown in Figure 12b.

In addition, the chloride concentration had a significant influence on the overall corrosion rate (Equations (13) and (14)). When the chloride content was high, the concentration of dissolved Fe2+ ions was much higher than the solubility constant of FeS [28].

Anodic metal dissolution in the presence of Cl<sup>−</sup>:

$$\text{Fe} + 2\text{Cl}^- + \text{H}\_2\text{O} \rightarrow \text{[Fe(OH)}^+ + \text{Cl}^-\text{]} + \text{HCl} + 2\text{e}^- \tag{15}$$

Cathodic sulfur reduction in the presence of Cl<sup>−</sup>:

$$\text{[FeS}\_{\text{x+1}}\text{]} + 2\text{e}^- + 2\text{Na}^+ + \text{H}\_2\text{O} \rightarrow \text{[FeS}\_{\text{x}}\text{]} + \text{(Na}^+ + \text{SH}^-) + \text{NaOH} \tag{16}$$

With a decrease in pH, especially at the tips of pits, higher concentrations of H+ (lower pH) accumulated under S8 deposition due to the hydration of Fe2+, forming an occluded corrosion chamber, which aggravated pitting corrosion. Cl− with a negative charge in the solution was adsorbed into the corrosion chamber with a positive charge to maintain the charge balance, which accelerated the anodic dissolution of the metal and negatively shifted the cathode potential, further deteriorating the environment under the S8 deposition layer. Furthermore, pitting began to expand outward, increasing the range of pitting and further accelerating the corrosion of the L360 pipeline steels.
