Controlling Atmospheric Corrosion of Weathering Steel Using Anodic Polarization Protection Technique

: The atmospheric corrosion of weathering steels varies as a function of geographic zone, season, and other environmental variables related to that region which the experiments have been done. Meanwhile, rusting is a continuous process, and it is the main corrosion product of atmospheric corrosion. The current study investigates the effects of rust on weathering steel in the localized region of Digha, a sea resort of West Bengal, India. The investigations have been performed by purposely accelerating the rusting of weathering steel in a laboratory within one week in order to simulate approximately 18 months of actual rusting that can be achieved at ﬁeld exposure. Anodic polarization of weathering steel comparable to potentiostatic passivation is obtained by shorting weathering steel with nobler metals, such as copper or graphite. The effect of rust formation on corrosion resistance after being immersed in 0.01 M KCl solutions for polished and unpolished samples has been investigated using electrochemical techniques, such as potentiodynamic polarization and electrochemical impedance spectroscopy (EIS). The rusted surfaces’ morphology and composition were characterized using ﬁeld emission scanning electron microscope (FE-SEM) and energy dispersive X-ray analysis (EDX). Based on the obtained results, it is concluded that the progressive rusting of weathering steel leads to a decrease in corrosion rate.


Introduction
Corrosion of substances and materials that are exposed to open air and its pollutants, as opposed to submerged in fluid, is called atmospheric or air corrosion [1]. The everincreasing dissatisfaction with atmospheric corrosion is leading in expense and weight loss of material above any other type of material degradation processes. Nevertheless, atmospheric corrosion is also accountable for half the corrosion damage every year [2]. However, some metals that form non-permeable layers give rise to the formation of oxides, which, in turn, achieve limited thickness through the dispersion of hardly noticeable oxide layers at room temperatures, unseen by the naked eye. In the case of metals such as chromium, titanium, and weathering steel, these limited thickness layers give an incredible shield over the metal surface.
Day-to-day enhancements in environmental contaminants have become a global concern. Different levels of pollutants entering the environment affect the corrosion rate. Few pollutants affecting the corrosion rate have been reported. Sulfur dioxide (SO 2 ) derived from the burning of petroleum products containing sulfur is estimated to be a leading factor in environmental corrosion [3,4]. The corrosion rates of non-coated metallic materials in the surrounding atmosphere are amplified by sulfur dioxide pollutants. Sulfur dioxide has higher dissolvability in water, giving rise to the formation of H 2 SO 4 over the surface in the form of surface dampness thin layer [5,6]. Comparatively, levels of NOx discharged from exhaust and combustion (street activities, traffic, and production of energy) are greater than emissions of sulfur dioxide [7,8]. Increasing environmental saltiness improves the electrolyte arrangement of the surface by using hygroscopic salts (e.g., MgCl 2 and NaCl), which, in turn, enhance atmospheric corrosion rates. Several studies have stated that the core reason behind environmental corrosion is chloride ions [9][10][11]. Nonetheless, hydrogen peroxide (H 2 O 2 ) is another pollutant available in the atmosphere, and it is the predisposition of hydrogen peroxide to be responsive to all iron alloys and composites, resulting in the production of HOX from its decomposition. Covering and steeping the surface of a target metal with a solution containing hydrogen peroxide results in the production of a passive layer [12,13]. Several studies [14,15] have revealed that most environmental changes and transformations of the atmosphere related to chlorinated hydrocarbons also contribute to fresh air purification levels.
Weathering steel (WS) comprises 1-2.5% alloying elements (Cr, Cu, Si, and P) and is the most common type of structural steel, widely used for bridges, buildings, etc. [16]. Moreover, it tends to form rust at a rate depending on the level of contact with oxygen in the presence of moisture and air. Researchers [17,18] demonstrated that the main products of environmental corrosion of steel are magnetite (Fe 3 O 4 ), amorphous ferric oxy-hydroxide (FeOx(OH) 3-2x ), and crystalline rust (α-FeOOH or G-FeOOH) as well as β-FeOOH in marine environments. According to Kamimura et al. [18], corrosion begins with the anodic dissolution of iron into ferrous ions (Fe 2+ ). These ions react with the moisture (hydrolysis) on the surface of the steel to form FeOH + . The FeOH + , in turn, reacts with the oxygen (oxidation) in the atmosphere to form G-FeOOH, which crystallizes and precipitates out of the system (the rate of crystallization and precipitation is increased if drying cycle occurs). Moisture mixed with pollutants such as sulfur dioxide (SO 2 ) has a relatively low pH. In contact with this mixture, the crystalline G-FeOOH dissolves to form amorphous FeOx (OH) 3-2x , which precipitates again. Lastly, the FeO x (OH) 3-2x undergoes a solid-state transformation (deprotonation hydroxyl ions from rainwater) to become α-FeOOH [19][20][21].
Using controlled passivation, the rust on weathering steel (WS), comparable to the rust formed on weathering steel due to a long period of exposure in Digha, was simulated. The simulation of the data was carried out by anodically polarizing the WS with a nobler metal, as the usage of a potentiostat would have been expensive and confined. Previous studies of Digha stated that 0.01 M KCl is found to be stable for simulating the data. Therefore, weathering steel was connected with nobler metals, such as Cu and graphite, and dipped in 0.01 M KCl. The corrosion rate of weathering steel that was anodically polarized with Cu or graphite showed a significant decrease in corrosion rate, confirming that the progressive rusting of weathering steel leads to a progressive decrease in the corrosion rate. The rust manufactured by this process was characterized by SEM and analyzed by EDX. The corrosion rate of WS obtained by this method was compared and rust characterized with data from 18-month field exposed WS in Digha. Thus, the study revealed that the protective layer of weathering that was produced after 18 months' exposure could easily be countered in the laboratory by connecting WS with nobler metals and dipping it in an appropriate solution. Using this inexpensive method, the corrosion rate of WS was also decreased.
Generally, this work aims to improve the atmospheric corrosion effect of commercial weathering steel in the localized region of Digha, a sea resort of West Bengal, India, using anodic polarization protection technique. The use of this technology ensures that the best anti-corrosion behavior of commercial weathering steel is obtained in the specified region, which has not been investigated before.

Materials
For the present study, hot rolled weathering steel (WS) sheets (5 mm thick) sourced from a leading manufacturer was taken. The properties and chemical composition of WS are given in Table 1. Microstructural characterization and surface morphology determination of the samples were carried out by field emission scanning electron microscopy (FE-SEM) using a JEOL-7600F, Tokyo, Japan, microscope. Elemental analysis was performed using energy dispersive X-Ray analysis (EDX), which connected to the abovementioned SEM device.

Electrochemical Corrosion Studies
The electrochemical corrosion studies were carried out on samples of dimensions 10 × 10 mm with 5 mm depth, which were cut from hot rolled weathering steel sheets. These samples were polished using various grades (500-1200 grades) of silicon carbide emery papers to get a smooth surface. An ultrasonic cleaner was then used to clean the specimens.
A three-electrode corrosion cell was used for EIS and potentiodynamic polarization measurements using a standard polarization cell test, in which a sample was taken as a working electrode; the graphite rod represented the auxiliary electrode, while standard calomel represented the reference electrode (SCE). Samples were mirror polished before each current transient experiment. The impedance experiments were performed using AC voltage 100 mV and a DC voltage of 1.0 V in the frequency range of 0.01 to 100 kHz [2,4,6]. The absolute impedance and phase angles were measured at each frequency, and Nyquist and Bode plots were obtained. Using actual and imaginary data components in a simplex fit model, the impedance data were fitted into a suitable equivalent electrical circuit. Based on this equivalent electrical circuit, the impedance data were analyzed. In this study, 0.01 M KCl at pH 8.8 solution was taken as the electrolyte for the corrosion test, while in some comparisons, 3.5% NaCl was used.

Rusting Procedure
Rusting of the samples in the laboratory was performed on polished as well as on weather exposed samples in 0.01 M KCl solution by two methods: In the first set of experiments, weathering steel was connected with a nobler material (Cu or graphite) and dipped in a 0.01 M KCl solution until the polarized potential stabilized. The sample, in as received condition-i.e., in a rusted condition or in a polished conditionwas connected with the nobler material (sample numbers SP1-SU4 of Table 2). It took 7 days to attain the stabilized potential and steady-state currents.

Potentiodynamic Polarization Tests
From Table 3, it can be seen that when polished samples were shorted with copper or graphite, very low steady-state current values were obtained. Current transient data are given in Figure 1. Slightly higher current values were obtained with unpolished samples. However, steady-state polarized potentials are in the passive range, as can be seen from Figure 2. It is interesting to note that when current transient experiments were done by applying the same polarized potentials, the steady-state currents obtained were very close to those obtained in the first set of experiments performed with polished specimens (6 against 3.5 and 8 against 4.2 mA/cm 2 , respectively). Thus, while it took 3600 s to attain 3.5 or 4.2 mA/cm 2 in the current transient experiment, it took 7 days to attain the steadystate currents in the first set of experiments where the steel samples were only shorted with copper/graphite and dipped in the solution. The effect in both the experiments was the same; the samples were both anodically polarized to the passive region of the steel. To cross check whether shorting the steel with nobler metal indeed gives the same effect as the passivation experiment done with the help of a potentiostat, weathering steel was shorted with nobler lead metal and dipped in a 0.01 KCl solution. The stabilized polarized potential and current was found to be 0.2 mV vs. SCE and 1.0 A/cm 2 , respectively. When the current transient experiment of the same steel was carried out in 0.01 M KCl by subjecting it to a polarized potential of 0.2 mV vs. SCE, the steady-state current after 7 days was found to be 20.82 micro A/cm2. This conclusively established the fact that shorting the steel with nobler metals gives the same effect as is obtained by anodic passivation by potentiostat. These experiments also confirmed that the progressive rusting of weathering steel led to a progressive decrease in corrosion rate. Rust on weathering steel was also formed anodically using a standard potentiostat. Samples were polarized anodically in a potentiostat by applying polarized potential obtained from the first set of experiments. Incidentally, these potentials are in the passive range of the steel, as can be seen from Figure 2. Results of set 1 as well as current transient data are given in Table 3.
Corrosion rates of samples SP1-SU4 after the first set of experiments and samples SP5 and SU6 as per Table 2 in 0.01 M KCl solutions were measured by potentiodynamic polarization tests (Figure 2, Table 4). The corrosion rates of samples SP1-SU4 were also measured in 3.5% NaCl solution using a potentiodynamic polarization test (Figure 3, Table 5).      Figure 2, it can be seen that the potentiodynamic curve of samples SP1-SU4 almost merges with that of the weathering steel exposed in Digha. The I corr values expressed in µA/cm 2 and µm/yr as well as E corr values are listed in Table 4. It can be noticed that the values corroborate the statement made above. The I corr values expressed in µm/yr show remarkable correlation with the corrosion rates of the field exposed corrosion rates given in Table 6. In fact, the corrosion rate of polished sample shorted with copper is almost same as that of the 24 months field exposed sample. It is interesting to note that, with increased exposure time, the corrosion rate decreased. From Table 4, it can also be inferred that shorting weathering steel with different nobler metals yields different corrosion rates. Therefore, corrosion rates similar to those of longer period field exposed samples can be obtained by shorting WS with other nobler metals or by exposing it beyond 7 days.

Electrochemical Impedance Spectroscopy (EIS)
From Figure 3 and Table 5, it can be seen that, while the unpolished sample shorted with copper gives the most active E corr , the unpolished sample shorted with graphite gives the most noble E corr . Table 6 gives a comparative study of corrosion rates of weathering steel in 3.5% NaCl solution. Since the experimental corrosion rates in 3.5% NaCl solution also correlate well with the field exposed sample, it can be stated that a satisfactory simulation of 18-month Digha exposed rusting by shorting with copper was achieved.
A three-electrode corrosion cell was constructed using Auto Lab to generate electrochemical impedance spectroscopy measurements. Bode and Nyquist plots were obtained. In order to choose the respective models and to find the value of all fitting models, curve fitting was performed. A software installed on the instrument was used to calculate parameters including R p , R u , and Y 0 . Fit/simplex model program was used in order to fit the experimental data (Figures 4 and 5 and Table 6) in order to interpreted impedance data ( Figure 6) on the basis of an equivalent electric circuit (Figure 7). EIS was performed for a better understanding of the corrosion properties. The experiment was done for WS. All the EIS spectra complied with CPE with diffusion.

Morphology and Surface Characterization
The morphology of the rust surface rusted potentiostatically when subjected to a constant potential of −607 mV vs. SCE and by forming a closed circuit with copper, which, as described earlier, was studied using field emission scanning electron microscopy (FE-SEM) and energy dispersive X-ray analysis (EDX). EDX analysis shows that the hematite (α-Fe 2 O 3 ) represented the main corrosion product, which forms a protective layer against corrosion [22][23][24][25][26]. The micrographs and composition are given in Figures 8 and 9, respectively.
From Figures 2 and 3 and Table 6, it can be seen that, for sample SP1, the polarization resistance (R p ) is the highest (Rp = 729 ohm) and its corrosion rate is also lowest among all the samples. Similarly, the Rp is lowest for sample SU6 (R p = 68.35 ohm) and its corrosion rate is high. Therefore, as the polarization resistance of the metal increases, the corrosion rate decreases. This is because an increase in polarization resistance means an increase in the thickness of the passive layer. SampleSP1 formed the most protective layer, so its resistance is high and hence its corrosion rate is at minimum. Sample SU6 is weather rusted, and its layer is not so well protected; therefore, its resistance is low, which is why its corrosion rate is high. Y 0 is inversely proportional to corrosion rate. It is highest for sample SP1 (Y 0 = 1.125 × 10 −6 ) and lowest for sample SU6 (24.16 × 10 −10 ). All curves are fit using the CPE model. It is also seen from the Nyquist plot that all data have a single time constant. Sample SP1 had the largest diameter, so it has the lowest corrosion rate; sample SU6 has lowest polarization resistance. The modulus curves also show the same pattern. Sample SP1 behaves like capacitance as its angle is approximately 90 • , and sample SP1 acts as a resistor, as its angle is close to 0 • . SEM and EDX analysis were performed for rust morphology studies of the steels passivated by potentiostat as well as by the method of dipping samples in a solution and connecting them to nobler metals. From Figures 8 and 9, it is seen that, in both cases, the rust morphology and composition were of a globular, non-compact type and correlated with field exposed rust.

Conclusions
Weathering steel was connected with nobler metals in order to simulate the production of rust that can occur during long periods of exposure in Digha, a sea resort of West Bengal, India. The following qualitative and quantitative conclusions can be made:

1.
Anodic passivation of weathering steel comparable to potentiostatic passivation is obtained by shorting weathering steel with nobler metals (copper or graphite). However, the lower corrosion rate of the weathering steel further revealed that, besides the progressive rusting of weathering steel, other factors such as the binding force between the rust and the substrate also affect the corrosion rate significantly.

2.
A simulation of 18 months field exposed weathering steel in Digha is achieved by shorting weathering steel with copper and dipping it in 0.01 M KCl solutions. Meanwhile, from SEM images and EDX analysis, the morphology of the rust obtained by two weeks' laboratory simulation correlates well with the rust obtained in 18-month field exposed samples. 3.
The corrosion rate of weathering steel that was anodically polarized with Cu or graphite showed a significant decrease in corrosion rate, confirming that the progressive rusting of weathering steel leads to a progressive decrease in corrosion rate. Specifically, anodic polarization using copper increased the corrosion resistance by 80 and 89% for the polished and unpolished samples, respectively; using graphite in polarization increased the corrosion resistance by only 13.6 and 20% for the polished and unpolished samples, respectively.

4.
From the EDX analysis of the rust layer, the main corrosion product of iron oxide (hematite α-Fe 2 O 3 ) represents a very beneficial protective layer initiated by the mechanism of chemical interaction between the exposed steel surface and the solution intensified by the formation of micro galvanic pairs, which reduce the corrosion rate.