This study investigates the corrosion behavior of three stainless steel grades at two H2SO4 concentrations, namely 1 wt% and 10 wt%, with varying NaCl concentrations in the range from 500 mg/L to 10,000 mg/L. Dissociation of sulfuric acid yields the hydrogen (H+) and sulfate () ions, the former of which lowers the pH value of a solution while the latter increases the concentration of sulfate ions that act as a corrosion inhibitor. The equilibrium chemistry of the solutions was defined at the test temperatures of 22°C, 50°C, 90°C, and 130°C, and correlated with the observations on the electrochemical and microstructural examination of the materials. The results showed clear differences in the main corrosion form between the two H2SO4 concentrations. In 1 wt% H2SO4, pitting was the major form of corrosion attack in the presence of chlorides, whereas uniform corrosion dominated in 10 wt% H2SO4. The pitting corrosion tendency for the three stainless steel grades under various test conditions was consistent, but there were differences in their resistance to uniform corrosion. The chloride-to-sulfate activity ratio, , was found to be the key parameter in defining the occurrence of pitting corrosion for all three alloys. In H2SO4-NaCl systems, no pitting occurred at the activity ratio below 10, with higher values inducing pitting attack, particularly in 1 wt% H2SO4. The described novel results are presented and discussed in this paper.

Demand for metals is foreseen to increase globally in the coming years, putting pressure on to utilize more complex and heterogeneous raw materials.1-3  Hydrometallurgical processing is an essential part of sustainable metal production4-,5  and needs to accommodate varying raw material quality. Therefore, operating conditions in the hydrometallurgical processes are harsh. Sulfuric acid is a common leaching agent and introduces a low pH value, and the raw materials often contain chlorides and oxidizing ions, such as ferric cations, that end up in the leachant.6  Additionally, processes are run at elevated temperatures.6-8  When targeting efficient leaching of the raw materials, the construction materials used in the assets and hydrometallurgical equipment, such as tanks, containers, pipes, tubes, valves, and mixers, are equally subjected to the aggressive process environment with the risk of corrosion.

Stainless steels typically have excellent corrosion resistance in pure sulfuric acid at low or high concentrations (but not at intermediate concentrations), due to the passive film that develops on the surfaces.9,-10  Iso-corrosion curves for austenitic stainless steels show that, at room temperature, numerous grades exhibit decent uniform corrosion resistance in both dilute and concentrated sulfuric acid solutions. However, the increase in temperature and addition of chlorides at as low amounts as 200 ppm (mg/L) may change the situation drastically with respect to the corrosion rate (CR) and prevailing corrosion form.11  Indeed, it is well known that chloride ions can give rise to the breakdown of passivity, leading to localized corrosion of the alloy, essentially pitting corrosion.12  The possibility of localized corrosion attack is dependent on several factors related to both material and operation environment, such as alloying13-17  and microstructure,18-20  temperature,13,21,-22  the concentration of chloride ions,21,23-26  and the oxidation power of the solution.27  Generally, with an increase in the pitting resistance equivalent number (PREN), calculated based on the used alloying:13 
formula
the stainless steel becomes more resistant to localized corrosion. However, the practice has proven that under hydrometallurgical conditions, uniform corrosion also plays a significant role. For obtaining insights into the corrosion phenomena, which may occur in the materials, sulfuric acid as the environment must be understood properly. Sulfuric acid dissociates in two stages:
formula
that results in the release of a hydrogen ion, H+, and a bisulfate ion, , the latter of which then further dissociates following:
formula
that eventually yields a hydrogen ion and a sulfate () ion. Equilibrium constants for the Reactions (2) and (3) are Ka = 1 × 102 and Ka = 1 × 10−2, respectively.28  Both bisulfate and sulfate ions may interact with the passive film.29,-30  Marcus31  has explained this by the adsorption (chemisorption) of sulfur on the surfaces as a result of chemical and electrochemical reactions of the sulfur species, i.e., inhibitive action.32-38  Besides the release of inhibitive sulfur species, dissociation of 1 mol of sulfuric acid produces two hydrogen ions. In other words, an increase in sulfuric acid content not only adds the concentration of corrosion inhibitors but simultaneously decreases the pH value of a solution, making the passive film thermodynamically less stable.

Outokumpu has put plenty of effort into defining the uniform CR of stainless steels at different sulfuric acid concentrations and temperatures.11,39  Other than this, corrosion studies performed for stainless steels (with the PREN of approximately 25 to 26) in sulfuric acid solutions with NaCl are not too many. The work by Hren, et al.,40  has concentrated on the development of a corrosion sensor rather than providing a thorough understanding of the corrosion behavior of stainless steels. Loto41  has studied the corrosion behavior of two grades of austenitic stainless steel, SS2562 (UNS N08904(1)) (EN1.4539 and PREN 36) and S31603 (EN1.4404 and PREN 28), in 2.5 M H2SO4 solutions with 3.5% NaCl as a function of temperature. Particularly for the grade S31603 (EN1.4404 and PREN 28), pitting corrosion was the primary corrosion mechanism at all temperatures, concentrating on grain boundaries. Pitting potential and the width of the passive range for S31603 decreased from 35°C to 45°C but further temperature raises did not have a profound influence on them. Khedr, et al.,23  have investigated the influence of chloride additions to 5 N, 7 N, and 10 N (2.5 M, 3.5 M, and 5 M, respectively) H2SO4 solutions on the corrosion behavior of the stainless steel of grade 1.4301 (PREN 18). Their results showed that the presence of chlorides introduced pitting corrosion in all cases except for 10 N H2SO4, i.e., at the highest sulfuric acid concentration. Laitinen42  has examined the localized corrosion of the stainless steel 1.4301 in 55°C H2SO4 solutions of pH 5 using the constant chloride concentration of 300 mg/L but varying sulfate additions. The results indicated that pitting corrosion occurred in the absence of additional sulfate and at the two lowest sulfate concentrations, 50 mg/L and 300 mg/L, but not at the highest sulfate concentration of 1,800 mg/L. Where the pits formed, they were deep and partly closed but did not contain any visible deposits. The inhibiting effect of sulfate ions on the pitting corrosion of stainless steel grade 1.4301 was proved valid with the molar ratio of [Cl]/[SO42–]≤0.5. Except for the work by Laitinen,42  the inhibitive role of sulfate ions in the pitting corrosion of stainless steel has been demonstrated in neutral NaCl solutions supplemented with sulfates, typically added as Na2SO4, e.g., Pohjanne, et al.43  The challenge is that the critical [Cl]/[SO42–] ratio for pitting corrosion inhibition from near-neutral (like pH 5) or neutral solutions cannot be directly transferred to acidic systems, like the ones in hydrometallurgy. Therefore, in order to use stainless steel assets for the hydrometallurgical processes, it is essential to define the material’s behavior under the respective process conditions. The work was motivated by the need to determine the safe operating conditions for the traditional S31603 type of stainless steel, an equivalent newer austenitic alloy with dissimilar alloying, and the corresponding duplex counterpart of S32101.

This study examines the corrosion behavior of three stainless steel grades of similar PREN but with dissimilar alloying and microstructure: S31603 (EN1.4432; austenitic), S31655 (EN1.4420; austenitic), and S32101 (EN1.4162; duplex) in two H2SO4 concentrations, 1 wt% and 10 wt%. Varying amounts of sodium chloride (500 mg/L to 10,000 mg/L) corresponding to 300 mg/L to 6,000 mg/L (ppm) chloride ions were added to the H2SO4 solutions. The test temperatures ranged from room temperature 22°C to 130°C. The tests involved electrochemical measurements (open-circuit potential [OCP] measurements and potentiodynamic polarization scans) and immersion experiments for 28 d to provide an understanding of the corrosion behavior of the three stainless steel grades. The concentrations of H2SO4 and NaCl and the test temperatures represent the industrial operation conditions in the hydrometallurgical process and provide the uniqueness for this work along with the approach, in which the solution equilibrium chemistry was investigated under each set of test conditions to obtain the chloride-to-sulfate activity ratio, , and correlated with the observations about the corrosion attack. Being motivated by the need to understand the link between the test conditions and the corrosion behavior of the materials, the research aimed at defining the prevailing corrosion form and key electrochemical characteristics for the three alloys as a function of test conditions. Ultimately, the target is to use the results for the computational prediction of corrosion of stainless steels under the hydrometallurgical process conditions. Earlier, e.g., Sridhar and Anderko44  have predicted the corrosion processes of selected stainless steel grades in sulfuric acid mixtures.

Materials and Test Conditions

Three grades of stainless steel: austenitic S31603 (EN1.4432), S31655 (EN1.4420), and austenitic-ferritic (duplex) S32101 (EN1.4162), were studied. The grade names in this paper were systematically presented according to the UNS designation. The actual composition of the grades together with the calculated PREN values, all in the range from 25 to 26, were given in Table 1. The test materials were supplied by Outokumpu Stainless AB, Sweden, as sheets of 2 mm (S31603) and 3 mm (S31655 and S32101) in thickness for the examinations at temperatures up to 90°C, and of 10 mm in thickness for the tests at 130°C. The surfaces of the specimens were wet ground to the 320 grit surface finish, ultrasonically cleaned in acetone, rinsed with ethanol, dried, and then let to oxidize in atmospheric air for at least 18 h.

Table 1.

Actual Composition of Stainless Steel Grades Included in the Research

Actual Composition of Stainless Steel Grades Included in the Research
Actual Composition of Stainless Steel Grades Included in the Research

Two sulfuric acid concentrations, 1 wt% and 10 wt%, were studied. Varying amounts of NaCl, namely 500 mg/L, 1,000 mg/L, 2,000 mg/L, 5,000 mg/L, and 10,000 mg/L, were added to the sulfuric acid solutions. Chemicals of analytical grade purity and ion-exchanged water were used in the preparation of test solutions. Solution characteristics at the four test temperatures: 22°C , 50°C, 90°C, and 130°C, were shown in Table 2. The equilibrium chemistry was calculated using HSC Chemistry 10 software.45  Chloride concentrations varied from 300 mg/L to 6,000 mg/L, giving the equilibrium chloride activities, a(Cl), in the range from 0.0001 to 0.0020. Equilibrium sulfate concentrations and, hence, sulfate activities, , showed a strong temperature dependency, yielding the ratio of chloride and sulfate activities, , in the range from 1 to 20 at room temperature and from 20 to 350 at the highest test temperature, 130°C. Calculated pH values ranged from 1.0 to 1.2 at the H2SO4 concentration of 1 wt% and from 0.01 to 0.2 in 10 wt% solution.

Table 2.

Targeted Conditions in the Electrochemical Measurements(A)

Targeted Conditions in the Electrochemical Measurements(A)
Targeted Conditions in the Electrochemical Measurements(A)

Methods

The test materials were subjected to microstructural characterization in order to obtain information on the overall material microstructure and composition of the passive film. For optical microscopy (OM) examinations, conducted using a Zeiss Axio Observer 7 optical microscope, austenitic alloys in the plane were chemically etched with V2A, while the duplex grade S32101 was electrolytically etched in oxalic acid and NaOH. The microstructure of the duplex grade was further imaged by scanning electron microscopy (SEM) using a Zeiss ULTRAplus field emission (FE) SEM equipped with energy dispersive spectroscopy (EDS) capabilities. X-ray photoelectron spectroscopy (XPS) measurements were conducted for wet-ground specimens of all three stainless steel grades that were then ultrasonically cleaned in acetone, rinsed with ethanol, dried, and left to oxidize for at least 18 h (treatment like for the specimens for electrochemical measurements) using a ULVAC-Phi Quantum 2000 instrument. The spectra were obtained using a monochromated Al Kα beam with a 100 µm spot size. In the case of duplex-grade S32101, such a spot size produced combined data for the austenitic and ferritic phases. The elemental analyses were performed for O (O1s), Fe (Fe2p3), Cr (Cr2p3), Ni (Ni2p3), Mo (Mo3d), and Mn (Mn2p3). The analysis depth was in the range of 3 nm to 5 nm, enabled by the sputtering with Ar+ with the acceleration voltage of 500 V. The measurements were followed by depth profiling for the same elements, at the sputtering rate of 1 nm/min. The depth profiles were collected every 30 s.

Electrochemical measurements up to 90°C were performed in an electrochemical glass cell with double walls, enabling the circulation of heated media in the casing for maintaining the desired test temperature. Heating was provided by an external heating unit. The counter electrode was either a platinum grit or a glassy carbon electrode, while a silver-silver chloride (Ag/AgCl electrode with saturated KCl solution) was used as a reference electrode. The studied test material as a plate specimen was located in the bottom of the cell with the help of flat rubber gaskets (that did not yield crevice corrosion, in contrast to conventional Avesta cell). The exposed area of the working electrode was 1 cm2 to 2 cm2, with the electrolyte volume being approximately 0.35 L. The measurements at 130°C were performed in a Cormet Hastelloy C276 autoclave using cylindrical specimens with a diameter of 10 mm and height of 15 mm. The cylinder was of the material of interest, and a wire of grade S31603 (EN 1.4404) stainless steel was welded on the cylinder to provide an electrical connection. Special care was taken during the welding step to avoid the propagation of a heat-affect zone in the material. The weld area and the connector wire were isolated from the test medium using a polytetrafluoroethylene (PTFE, brand name Teflon) sheath, leaving only the material of interest (3.2 cm2 to 4.5 cm2) exposed to the test solution of the volume of 1 L. Prior to the tests, the solution was warmed to the test temperature with an external resistance jacket. In the autoclave tests, platinum was used as a counter electrode and a silver-silver chloride (Ag/AgCl with saturated KCl) as a reference electrode. All potential values in this paper are reported against Ag/AgCl reference electrode (potential E = 0.210 V at 22°C).

The electrochemical measurements involved OCP measurements and cyclic potentiodynamic polarization measurements. In all cases, the electrolyte was purged with nitrogen gas (N2) before and during the measurements, in order to remove air. In 130°C experiments, in addition to purging the test solution with N2, the autoclave was closed and filled with N2 up to five bars and then released, the cycle which was repeated at least five times to deaerate the autoclave and ensure that there was no leakage. The electrochemical measurements under each set of test conditions were performed sequentially for one specimen. First, OCP measurements were performed for at least 1 h or until OCP stabilized (ΔE/Δt was less than 20 mV/20 min). Second, cyclic potentiodynamic polarization measurements were conducted from the cathodic potential of −0.25 VOCP up to the anodic value of +1.4 V at a scan rate of 0.167 mV/s. The cut-off current density value of 0.1 mA/cm2 was defined for the measurements. The measurements were performed using either a Gamry Reference 600 or a Princeton Applied Research VersaSTAT 3 potentiostat. The following electrochemical parameters were then extracted from the measurement data: OCP at the end of the measurement period, corrosion potential (Ecorr), and corrosion current density (icorr) that were extrapolated from Tafel fits, critical passivation current density (icrit), passivation potential (Epass), passive current density (ipass), and breakdown potential (Eb). Where a positive hysteresis was detected in the cyclic potentidynamic curve, repassivation potential (Erep) also was defined. Finally, CR was calculated from the value of icorr according to Faraday’s law. The reliability of the results was verified by checking the consistency of observed trends. Where the positive hysteresis was detected in the cyclic potentiodynamic curve, at least two parallel measurements (often more) were run (with the average values being presented).

For validation of the results from electrochemical measurements, 28-d immersion experiments were done. The aim of the test was to examine, whether the same phenomena could be introduced by immersion experiments that were revealed by electrochemical measurements. The identical immersion test was conducted at four test sites, thus only one combination of material and test conditions was chosen for investigations. The stainless steel grade included in the immersion tests was S31603, with a surface area of at least 20 cm2. The specimen surfaces were treated similarly to the specimens for electrochemical tests. The electrolyte in the immersion experiments was 1 wt% H2SO4 with 2,000 mg/L NaCl. The solution was further supplemented by Fe2(SO4)3, in order to rise the redox of the solution high enough to induce pitting corrosion. In total, 5,000 mg/L Fe3+ was added as in Lindgren, et al.,27  The immersion experiments were performed at 90°C, thus the prepared solutions were heated, either in a heating cabinet or using a hotplate. In each case, at least two parallel weighed specimens were exposed to the test solution with the help of a PTFE specimen holder located at the bottom of a container.

After the experiments, the specimens were examined visually to detect the occurrence, form, and extent of corrosion. Surfaces of selected specimens were further investigated with SEM/EDS, using a Zeiss ULTRAplus SEM to define the corrosion form and investigate the details of corrosion attack. After the immersion tests, weight changes for the specimens were also defined.

Microstructural Analysis

Metallographic images, showing the surface microstructure of the three grades of stainless steel, are presented in Figure 1. The austenitic grades, S31603 and S31655, varied in grain size. The average grain diameter on the surface of grade S31603 was approximately 20 µm, yet the diameter of the grains varied from 10 µm up to 40 µm (Figure 1[a]). The microstructure of grade S31655 was relatively much finer, with the average grain size on the surface approximately half compared to that in grade S31603 (Figure 1[b]). The third grade, S32101, had a duplex austenitic-ferritic microstructure (Figures 1[c] and [d]). SEM/EDS analyses indicated that the unetched phase (seen in slightly darker gray contrast in the images) in the duplex microstructure, Figures 1(c) and (d), contained slightly greater Ni contents than the etched phase, being likely austenite. Ni content in the austenite phase was 2% while that in the ferrite phase was clearly less than 1%, and the chromium contents were 20% and 23%, respectively. However, EDS analyses provided suggestive rather than absolute values for the phases, as a majority of the measurement data originated underneath the surface (interaction volume) and likely contained both phases. Based on the microstructural studies, austenite islands in the duplex grade were from 5 µm to 50 µm in width and up to 250 µm in length. Sections of ferrite between the austenite islands, seen in slightly lighter gray contrast, were relatively thinner and elongated when compared to the corresponding austenite regions.
FIGURE 1.

Results from microscopy examinations for the three stainless steel grades. (a) OM image, grade S31603, (b) OM image, S31655, (c) OM image, S32101, and (d) SEM image, S32101.

FIGURE 1.

Results from microscopy examinations for the three stainless steel grades. (a) OM image, grade S31603, (b) OM image, S31655, (c) OM image, S32101, and (d) SEM image, S32101.

Close modal
FIGURE 2.

XPS spectra for grade S31603. (a) Depth profile for the detected elements, sputtering rate 1 nm/min, (b) depth profile for Cr, and (c) depth profile for Fe. In (b) and (c), the sputtering time between the sequential spectra is 30 s.

FIGURE 2.

XPS spectra for grade S31603. (a) Depth profile for the detected elements, sputtering rate 1 nm/min, (b) depth profile for Cr, and (c) depth profile for Fe. In (b) and (c), the sputtering time between the sequential spectra is 30 s.

Close modal
FIGURE 3.

XPS spectra for grade S31655. (a) Depth profile for the detected elements, sputtering rate 1 nm/min, (b) depth profile for Cr, and (c) depth profile for Fe. In (b) and (c), the sputtering time between the sequential spectra is 30 s.

FIGURE 3.

XPS spectra for grade S31655. (a) Depth profile for the detected elements, sputtering rate 1 nm/min, (b) depth profile for Cr, and (c) depth profile for Fe. In (b) and (c), the sputtering time between the sequential spectra is 30 s.

Close modal
FIGURE 4.

XPS spectra for grade S32101. (a) Depth profile for the detected elements, sputtering rate 1 nm/min, (b) depth profile for Cr, and (c) depth profile for Fe. In (b) and (c), the sputtering time between the sequential spectra is 30 s.

FIGURE 4.

XPS spectra for grade S32101. (a) Depth profile for the detected elements, sputtering rate 1 nm/min, (b) depth profile for Cr, and (c) depth profile for Fe. In (b) and (c), the sputtering time between the sequential spectra is 30 s.

Close modal

XPS spectra with elemental depth profiles were measured to define the structure and thickness of passive films on the three materials, Figures 2 through 4. The passive films on all investigated stainless steels were of equal thickness of 2 nm, as estimated based on the depth where the concentration of oxygen was halved from the initial value. The passive film thickness is consistent with the literature, e.g., Gardin, et al.46  In all cases, the relative concentration of chromium in the vicinity of the surface was much greater than that at greater depths, suggesting chromium was enriched in the outer layer of the passive film. For the grade S31603, the maximum Cr/Fe ratio of 0.45 was detected at the depth of approximately 1 nm, in comparison to the corresponding value of 0.25 in the bulk of the alloy. For the two other grades, S31655 and S32101, the maximum Cr/Fe ratio was slightly higher, 0.55 (in comparison to 0.30 in the bulk of the alloys), and it occurred at the depths of 1 nm (S31655) and 0.5 nm (S32101). Chromium depth profiles (Figures 2[b], 3[b], and 4[b]) revealed that down to the depth of 1.5 nm, the highest-intensity peak was located at 577 eV to 576 eV, referring to Cr(OH)3 and Cr2O3, with the respective binding energies of 577.3 eV and 576.3 eV.47-51  Indeed, the experience has proved that Cr(OH)3 often coexists with Cr2O3 on wet-ground surfaces. At greater depths, the metallic state of chromium prevailed, as demonstrated by a major peak at the binding energy of 574 eV (574.1 eV).46,50  Iron depth profiles (Figures 2[c], 3[c], and 4[c]) disclosed the maximum peak at 711 eV to 710 eV until the depth of 1 nm, suggesting Fe2O3 (Fe3+) with the binding energy of 711.0 eV.52  In the austenitic grades S31603 and S31655, nickel and molybdenum were also detected in the passive film, but these existed at lower amounts than in the bulk alloy. It is emphasized that, although nitrogen is included in the alloys, its detection by XPS is challenging, due to the overlapping of the N1s and the Mo 3p3/2 signals. Altogether, the main finding was that in the Alloys S31655 and S32101 with higher chromium contents in the bulk alloy than in grade S31603, chromium was also enriched to a greater degree in the passive film.

Electrochemical Measurements

As stainless steels are passive alloys, in which localized corrosion is typically the major issue, cyclic potentiodynamic polarization measurements were chosen as the main method to define the corrosion behavior and yield electrochemical parameters that can be compared. Besides pitting corrosion, the measurements enable theevaluation of uniform CR via the parameters, particularly icorr derived by extrapolating the electrochemical data. In most cases, the uniform corrosion occurred as the uniform corrosion in active state (uniform active state corrosion), but there were also cases in which the uniform corrosion took place despite the passive state (uniform passive state corrosion); these are described in the following. We acknowledge that the extrapolation requires first-order kinetics in the system; this was met in many cases, but not in all. It is also emphasized that as polarization always deviates the system from the equilibrium, the CR defined by this approach is likely higher than under equilibrium conditions. However, the obeyed methodology enables thedefining of the dominating corrosion form in each case and to make an overall comparison between the stainless steel grades and test conditions, keeping in mind the large test matrix of the study.

1 wt% Sulfuric Acid

Examples of potentiodynamic polarization curves for the three alloys in 1 wt% H2SO4 are shown in Figure 5. Due to the high number of measured polarization curves and their partial overlapping when plotted, and in order to provide novel insights into the corrosion behavior, our approach is to define the key electrochemical parameters from the curves and evaluate the trends observed in them as a function of test conditions. The thermodynamic and kinetic parameters describing the corrosion behavior of the three test materials in 1 wt% H2SO4 solution are shown in Figures 6 through 9. The results from OCP measurements, Figure 6(a), revealed differences between the alloys but not so clearly between the chloride contents or test temperatures. The lowest overall OCP values were detected for grade S31603, with the average value being −260 mVAg/AgCl and the minimum as low as −310 mVAg/AgCl. The OCP values for the grade S31655 were slightly higher than for the other austenitic grade S31603, on average −210 mVAg/AgCl. By far the highest OCP values were obtained for the grade S32101. At temperatures up to 90°C, the values for the grade S32101 fell mostly in the range from 0 to 100 mVAg/AgCl, which means that they were approximately 200 mV higher than the corresponding values for the austenitic grades. These results indicate that, up to 90°C, the austenitic grades S31603 and S31655 were likely in an active state, whereas the grade S32101 was essentially in a passive state; these observations are supported by the results from polarization measurements shown below. At the highest test temperature, OCP values for all three stainless steel grades were of the same magnitude, around −250 mVAg/AgCl to −200 mVAg/AgCl, suggesting an active behavior consistent with austenitic alloys at the lower temperatures. It is also worth mentioning that in some cases, for example, for the grade S31603 at NaCl concentrations in the range from 500 mg/L to 5,000 mg/L, the OCP values at the four temperatures were all within 15 mV range, thus the respective data points in Figure 6(a) overlapped.
FIGURE 5.

Examples of potentiodynamic polarization curves for the three grades of stainless steel in 1 wt% H2SO4 at 90°C at two NaCl concentrations. (a) 0 mg/L NaCl, i.e., pure 1 wt% H2SO4 and (b) 5,000 mg/L NaCl in 1 wt% H2SO4.

FIGURE 5.

Examples of potentiodynamic polarization curves for the three grades of stainless steel in 1 wt% H2SO4 at 90°C at two NaCl concentrations. (a) 0 mg/L NaCl, i.e., pure 1 wt% H2SO4 and (b) 5,000 mg/L NaCl in 1 wt% H2SO4.

Close modal
FIGURE 6.

(a) OCP values and (b) correlation between corrosion potential, Ecorr, and OCP for the alloys in 1 wt% sulfuric acid.

FIGURE 6.

(a) OCP values and (b) correlation between corrosion potential, Ecorr, and OCP for the alloys in 1 wt% sulfuric acid.

Close modal

The values for Ecorr were lower than those for OCP for the austenitic grades at all temperatures and duplex specimens tested at the highest test temperature, Figure 6(b). In general, the main observation was that the data points could be categorized into three groups. For the austenitic grades S31603 and S31655 at temperatures up to 90°C, both Ecorr and OCP values fell around the range from −300 mVAg/AgCl to −200 mVAg/AgCl. For the duplex-grade S32101, both potential values were much higher than the corresponding values for the austenitic grades (Figure 6[b]) and Ecorr values were typically higher than OCP values; this is likely due to the passive film that may have evolved during the OCP measurement period. The third group of data points included the measurements performed at 130°C, with the values of Ecorr for all three grades being in the range from −450 mV to −300 mV, thus clearly lower than the OCP values. The obtained results reflect that at temperatures up to 90°C, the duplex-grade S32101 was more easily passivated than the austenitic grades S31603 and S31655, but at the highest test temperature, the potential values indicated an active rather than passive behavior for all three alloys. The behavior up to 90°C is consistent with the findings by Helmersson.53 

Uniform corrosion rates for the three alloys under the studied conditions, calculated based on the values for corrosion current density, icorr, are presented in Figure 7. The obtained results showed that the corrosion rates varied clearly between the materials, test temperatures, and NaCl concentrations of the solution. The highest overall corrosion rates were detected for the grade S31603, consistent with the lowest Cr content, while the lowest corrosion rates were conversely associated with the grade S32101 with the highest Cr content and duplex microstructure. The test temperature had a drastic influence on the CR: the higher the temperature, the greater the CR for all materials. At the highest test temperature, the CR for all three materials in 1 wt% H2SO4 solution was 1 mm/y or higher, and in solutions with any NaCl at minimum 3 mm/y. The CRs for all three alloys also increased with the increase in NaCl content of the solution, yet not as radically as with the rise in temperature. It is emphasized that as the corrosion rates are derived based on the dynamic polarization curves, they may be higher than those under (thermodynamic equilibrium conditions of) real applications.
FIGURE 7.

CR for the alloys in 1 wt% sulfuric acid, defined based on the corrosion current density, icorr.

FIGURE 7.

CR for the alloys in 1 wt% sulfuric acid, defined based on the corrosion current density, icorr.

Close modal
Build-up and breakdown of passivity were investigated via key potential values obtained upon polarization: passivation potential, Epass, breakdown potential, Eb, and repassivation potential, Erep, Figure 8. Examination of the values for Epass disclosed differences between the test materials, but not so clearly between temperatures or solution NaCl concentrations (Figure 8[a]). The lowest Epass values were detected for the grade S31655. In turn, clearly, the highest Epass values were related to grade S32101, consistent with the highest OCP and Ecorr values. However, the relatively lower Epass values for the grade S31655 than for S31603 reflect the role of key alloying elements: Cr, Ni, Mo, and N in the passivation capability, particularly the higher Cr and N contents in S31655 likely contribute to the relatively easier passivation. Further insights into the passive film build-up were obtained via examining the difference between passivation and corrosion potentials, Epass-Ecorr (Figure 8[b]). These results revealed that the grade S31603 presented the highest Epass-Ecorr values among the three alloys, indicating that the build-up of its passive film requires a greater deviation from the equilibrium in comparison with the other two grades, thus greater oxidation capacity of the environment. In other words, the development of passive film on grade S31603 was not as easy as in the case of Alloys S31655 and S32101. Another key feature was the significant increase in Epass-Ecorr upon temperature increase from 90°C to 130°C, particularly for the grade S32101, reflecting the higher difficulty in reaching the passive state at the highest test temperature. The analysis of Eb values showed the dependence of the parameter on temperature, NaCl concentration, and alloy composition (Figure 8[c]). At the lowest temperature, all materials behaved in a consistent manner irrespective of the NaCl content of the solution: Eb values were around 750 mVAg/AgCl to 900 mVAg/AgCl, corresponding to transpassivity rather than actual passivation breakdown. However, Eb values decreased with an increase in temperature. Already at 50°C at the highest NaCl concentration of the solution, Eb values for the grades S31603 and S31655 decreased to the level below 600 mVAg/AgCl. With further temperature increase to 90°C, Eb values at the highest NaCl concentration of the solution for all three grades were clearly lowered, in the case of austenitic grades S31603 and S31655 down to the level of 200 mVAg/AgCl and in the case of duplex grade S32101 down to 700 mVAg/AgCl. Additionally, at 90°C, the lowered Eb values were detected at lower NaCl concentrations than at lower temperatures, even at 2,000 mg/L. At the highest test temperature, Eb values for all three alloys were significantly lowered as compared to the lower temperatures, both in the solutions without NaCl and particularly in the ones involving NaCl. A detailed analysis of Eb as a function of solution chloride activity, aCl, Figure 8(d), revealed some linearity between the values of Eb and aCl (grade S31603 at 90°C, grades S31655 and S32101 at the temperature of 130°C) but not a truly consistent behavior for all materials. This observation may be partly explained by the fact that the values of Eb also include the cases of transpassivity, i.e., the points that do not represent pitting corrosion. Eventually, the study of values for Erep revealed a decreasing trend with increases in temperature and NaCl concentration, Figure 8(e). Up to 90°C, the values for Erep were, on average, lower for the austenitic grades S31603 and S31655 than for the duplex grade S32101. However, at the highest test temperature, no significant differences between the grades could be observed, with all Erep values falling in the range from −150 mVAg/AgCl to −50 mVAg/AgCl.
FIGURE 8.

Important potential values for the passivity of the alloys in 1 wt% sulfuric acid. (a) Passivation potential, Epass, (b) Epass-Ecorr (c) breakdown potential, Eb, (d) Eb as a function of chloride activity, aCl, and (e) repassivation potential, Erep.

FIGURE 8.

Important potential values for the passivity of the alloys in 1 wt% sulfuric acid. (a) Passivation potential, Epass, (b) Epass-Ecorr (c) breakdown potential, Eb, (d) Eb as a function of chloride activity, aCl, and (e) repassivation potential, Erep.

Close modal
Critical current density for passivation, icrit, is a parameter characterizing the ease of passivation in such cases where a clear shift from active to passive state occurs upon polarization. In the analysis of the results, Figure 9(a), it should be kept in mind that the grade S32101 did not exhibit a clear icrit in the polarization curve in the temperature range of 22°C to 90°C, thus 130°C was the only temperature where active-passive transformation involved an evident icrit. This is consistent with the observations for the grade S32101 about much higher values for Epass-Ecorr at the temperature of 130°C than at lower temperatures, reported above. The reason why the duplex grade did not involve a clear icrit in the active-passive transition up to 90°C is likely the considerably easy passivation which did not require the triggering by icrit. Helmersson53  earlier reported that passivation in duplex grades is relatively easier than in austenitic grades. Consequently, the behavior of critical current density within the temperature range from 22°C to 90°C is investigated only for the austenitic grades. Figure 9(a) revealed that the lowest icrit values were detected at the lowest test temperature and the highest icrit values were consistently related to the highest test temperature. At the temperatures of 22°C and 50°C, the lowest icrit values were associated with the grade S31603 up to the NaCl content of 5,000 mg/L. Upon a temperature increase to 90°C and NaCl concentration rise to 10,000 mg/L, the lowest icrit values were detected for the grade S31655. Furthermore, at the temperature of 130°C, the lowest icrit values were related to the Alloy S32101. These results indicate that under the mildest test conditions, the passive film development was easiest for the austenitic alloy with the lowest Cr content (S31603), but as the test conditions became more extreme, the passive film formation became relatively earlier for the alloys with higher Cr alloying (S31655 and S32101). As for the passive current density ipass for the alloys, Figure 9(b), the main observation was that at temperatures up to 90°C, the values for all test materials and NaCl concentrations were of the magnitude of 1 µA/cm2 to 2 µA/cm2 or less. Indeed, these values are in line with the data that has been reported earlier for the passive state in stainless steels, e.g., Nicic and Macdonald.54-55  The only exception was the grade S31603 at 90°C, where occasional higher values were detected (e.g., at NaCl concentration of 2,000 mg/L, 7 µA/cm2). At the highest test temperature, the values increased significantly beyond what was observed at lower temperatures, up to some tens of µA/cm2. Such values indicate the greater rate of dissolution of the passive film, i.e., uniform corrosion. This implies that the high CR that were detected for the three materials at the temperature of 130°C were related to uniform passive state corrosion (as the distinction from the uniform active state corrosion occurring at lower temperatures). As pointed out by, e.g., the point defect model (PDM),56  the passive film features a steady-state thickness, which results from the balance between the dissolution and growth of the passive film (barrier film), with the transmission of ions occurring through vacancy motion. The uniform corrosion relates to the dissolution (and related formation of the passive film), with high dissolution rate causing either thinning of the barrier layer (decrease in the steady-state thickness), sustaining the steady-state thickness by equally increasing the barrier film formation rate (by the expense of loss of the metal at the film/metal interface), or a combination of these. In turn, the localized breakdown (pitting corrosion) of the film is associated with the capability of aggressive ionic species, like chlorides, to contribute to the vacancy formation, travel across the barrier film and annihilation at the film/metal interface, thus two corrosion forms are not exclusive but may occur simultaneously like here detected at the temperature of 130°C.
FIGURE 9.

Current density values important for passivation for the three grades of stainless steel in 1 wt% H2SO4. (a) Critical current density for passivation, icrit and (b) passive current density, ipass.

FIGURE 9.

Current density values important for passivation for the three grades of stainless steel in 1 wt% H2SO4. (a) Critical current density for passivation, icrit and (b) passive current density, ipass.

Close modal
Surface examinations of the specimens after the polarization measurements revealed that, indeed, pitting corrosion was detected in the alloys under the test condition where a positive hysteresis was observed in the polarization curves. At the temperatures of 50°C and 90°C, only few pits were detected in the exposed surfaces by SEM in each case, but these were large with the diameter up to some hundreds of µm, Figures 10(a) through (c). This indicates that pit growth was favored over pit nucleation, thus nucleation was the critical step in the pitting process. It is possible that the competitive adsorption between the chloride and sulfate ions may have contributed to the low pit frequency on the surfaces; it has earlier been reported in the case of pure aluminum.57  Additionally, individual pits often contained several nuclei. This may be explained either by the fact that some of the pits may have repassivated after initiation or that nucleation of new pits in the existing pits was easier than on pristine surfaces due to the differences in the crevice and bulk chemistries, or a combination of these. The pit morphology in the austenitic alloys showed a correlation with temperature. At 50°, the pits were open and did not contain any deposit or cover, while at 90°C, the pits contained a lacy cover (Figures 10[a] through [c]). At the highest test temperature, the attack in austenitic grades occurred through elongated pits that contained several nuclei (Figures 10[d] and [e]). Here, the edge areas also contained the lacy cover. Thus, at the lowest temperatures where pitting occurred, the hemispherical pits were open, but toward higher test temperatures, pits in austenitic stainless steels systematically involved a lacy cover. Ernst, et al.,58  have described the lacy cover formation mechanism as a sequential process, where the formation of a hemispherical pit is followed by passivation near the pit mouth due to shorter diffusion length to the bulk solution, further dissolution within the pit which undercuts the passivated material on the surface and enables ions to diffuse in and within the hole, while the material around the hole remains passivated. The proposed mechanism reflects that pitting has progressed further in such cases where the lacy cover exists as compared to open hemispherical pits, in agreement with the observation of lacy cover cases toward higher test temperatures. In the duplex grade, numerous pit nuclei were essentially grown together to form very large pits (Figure 10[f]), the diameter up to mm length scale (not completely shown in the image). Furthermore, surface studies of all grades at the highest temperature clearly revealed the contribution of uniform corrosion to the morphology, consistent with high current density values in the passive state. In the duplex grade, the pit growth was so extensive that it was not possible to detect whether pit initiation favored the austenite or ferrite phases or their interface. Deposit was not observed to cover the pits or inside the pits in any of the cases, likely due to the overall low pH of the solutions (in which the corrosion deposit-forming compounds may not be thermodynamically stable) and the lack of suitable cations in the solution. It is emphasized that SEM examinations for the specimens exposed to 130°C revealed the uniform loss of material on the surfaces. This indicates that the uniform passive film corrosion occurred via accelerated passive film dissolution, compensated by the accelerated film formation at the film/metal interface, which eventually resulted in the uniform loss of the metal. As described above, the PDM explains why the uniform passive state corrosion and pitting corrosion may occur simultaneously, as demonstrated here.
FIGURE 10.

SEM images, showing the examples of corrosion attack in the alloys exposed to 1 wt% sulfuric acid solution at various temperatures and NaCl concentrations. (a) S31603, 90°C, and 5,000 mg/L, (b) S31655, 50°C, and 5,000 mg/L, (c) S31655, 90°C, and 5,000 mg/L, (d) S31603, 130°C, and 500 mg/L, (e) S31655, 130°C, and 500 mg/L, and (f) S32101, 130°C, and 500 mg/L.

FIGURE 10.

SEM images, showing the examples of corrosion attack in the alloys exposed to 1 wt% sulfuric acid solution at various temperatures and NaCl concentrations. (a) S31603, 90°C, and 5,000 mg/L, (b) S31655, 50°C, and 5,000 mg/L, (c) S31655, 90°C, and 5,000 mg/L, (d) S31603, 130°C, and 500 mg/L, (e) S31655, 130°C, and 500 mg/L, and (f) S32101, 130°C, and 500 mg/L.

Close modal

10 wt% Sulfuric Acid

Examples of potentiodynamic polarization curves for the three grades of stainless steel in 10 wt% are shown in Figure 11. Differences in the corrosion behavior of the alloys could be detected in comparison to the lower H2SO4 concentration. In particular, in 10 wt% H2SO4 the test materials had a lower tendency to passive film breakdown and pitting corrosion as compared to 1 wt% solution. A detailed analysis (Figure 12[a]) revealed that for the austenitic grades S31603 and S31655 independent of test temperature, OCP values fell in the range from −150 mVAg/AgCl to −400 mVAg/AgCl. For the duplex Alloy S32101 at temperatures up to 90°C, the OCP values were positive, in the range from 0 to 300 mVAg/AgCl, while the OCP values measured at the highest test temperature were clearly negative, −200 mVAg/AgCl and lower. In comparison between the alloys, the lowest overall OCP values in 10 wt% were detected for the grade S31603 and the highest OCP levels for S32101, consistent with what was observed at the lower H2SO4 content. Altogether, slightly lower OCP values for the materials were measured in 10 wt% sulfuric acid as compared to 1 wt% solution. The exception was the solutions with no added NaCl, in which the OCP values for the three alloys in 10% H2SO4 solution were somewhat higher than in 1 wt% H2SO4, and higher than in the 10 wt% H2SO4 solutions with NaCl, yet no consistent development trend in OCP values as a function of NaCl content of the solution could be detected. On average, the difference in OCP values between the two H2SO4 concentrations was greater for grades S31655 and S32101 than for the grade S31603, approximately 50 mV. This indicates that the alloys behaved in a slightly more active manner and thus had a greater driving force for corrosion.
FIGURE 11.

Examples of potentiodynamic polarization curves for the three grades of stainless steel in 10 wt% H2SO4 at 90°C at two NaCl concentrations. (a) 0 mg/L NaCl, i.e., pure 10 wt% H2SO4 and (b) 5,000 mg/L NaCl in 10 wt% H2SO4

FIGURE 11.

Examples of potentiodynamic polarization curves for the three grades of stainless steel in 10 wt% H2SO4 at 90°C at two NaCl concentrations. (a) 0 mg/L NaCl, i.e., pure 10 wt% H2SO4 and (b) 5,000 mg/L NaCl in 10 wt% H2SO4

Close modal
FIGURE 12.

(a) OCP values and (b) correlation between Ecorr and OCP for the alloys in 10 wt% sulfuric acid.

FIGURE 12.

(a) OCP values and (b) correlation between Ecorr and OCP for the alloys in 10 wt% sulfuric acid.

Close modal
As for the values of Ecorr, Figure 12[b], a one-to-one dependency on OCP was detected for all test materials under the majority of conditions up to the temperature of 90°C. At 130°C, the values of Ecorr were lower than for those of OCP, yet a linear relationship between the two was detected also at this temperature. What do these results then indicate? In most cases, the OCP values were retained at a relatively constant level as a function of time (Figure 13), indicating the presence of passive film on the surfaces during the measurements. This is likely due to the air-formed passive layer that was retained on the surfaces and developed further during the exposure until the steady-state thickness was reached. However, as shown above by somewhat lower OCP values in comparison to 1 wt% solution, it is possible that the steady-state thickness of the passive film for the alloys in 10 wt% H2SO4 was relatively thinner than in the case of the 1 wt% H2SO4. Figure 13 shows that, indeed, there were occasional cases in 10 wt% sulfuric acid where OCP did not develop consistently but featured evident fluctuation. Earlier, such fluctuations have been connected to the activation of the passivated surface.59  We interpret the fluctuations as the competition between the build-up and dissolution of the film, which in passive films that have not yet reached the steady-state thickness. This suggests that the passive layer on the alloys in 10 wt% sulfuric acid was not as kinetically stable and protective as in 1 wt% solution. Such a situation resembles the material’s behavior under cathodic polarization and may partly explain why one-to-one correspondence between the measured Ecorr and OCP values was observed in many cases for the alloys in 10 wt% sulfuric acid or even situations with lower Ecorr values compared to OCP values were detected, consistent to what has been reported in the literature.60  It is emphasized that the overall potential (OCP and Ecorr) differences were relatively small, thus it is clear that the surface was not activated. Minor differences were detected between the three alloys. The essentially higher Ecorr and OCP values as compared to the austenitic grades suggest the greater thermodynamic stability of the passive film on the duplex alloy.
FIGURE 13.

Example of OCP records for the Alloy S31655 in 10 wt% sulfuric acid solution at 90°C as a function of NaCl concentration.

FIGURE 13.

Example of OCP records for the Alloy S31655 in 10 wt% sulfuric acid solution at 90°C as a function of NaCl concentration.

Close modal
Figure 14 shows the CR of the three alloys under the studied test conditions, calculated based on the values of icorr. The key observation was that in most cases, the values were higher than in 1 wt% H2SO4 solution, occasionally as significantly as by two orders of magnitude. In 10 wt% H2SO4 solutions, the highest overall corrosion rates were associated with the grade S31603, while the lowest corrosion rates were measured for the grade S32101, consistent with observations in 1 wt% H2SO4 solutions and the increasing level of Cr alloying in the material. For the grades S31603 and S31655, the corrosion rates were, on average, by one magnitude higher in 10 wt% solution than in 1 wt% H2SO4 solution, but there were occasional conditions at the highest temperature where the difference was almost by two orders of magnitude. The values increased by several orders of magnitude with an increase in temperature: they were in the range of 0.01 mm/y to 1 mm/y at the lowest temperatures but increased first up to 10 mm/y at 90°C and further up to 2,000 mm/y at 130°C. For the grade S32101 up to 90°C, the degradation rates in 10 wt% H2SO4 were equal to those detected in 1 wt% H2SO4 solutions, of the magnitude of 0.001 mm/y and below, but at the temperature of 130°C the rates in 10 wt% H2SO4 solutions were by one magnitude higher than in the more diluted solution, at a maximum 300 mm/y. In all cases, the corrosion rates of the materials increased significantly with an increase in temperature and systematically also with an increase in the solution NaCl concentration, similar to 1 wt% H2SO4 solution.
FIGURE 14.

CR for the alloys in 10 wt% sulfuric acid, defined based on the corrosion current density, icorr.

FIGURE 14.

CR for the alloys in 10 wt% sulfuric acid, defined based on the corrosion current density, icorr.

Close modal
The build-up and breakdown of passivity were again evaluated through three parameters, Epass, Eb, and Erep. Values for Epass fell in the range from −300 mVAg/AgCl to 350 mVAg/AgCl, Figure 15(a), thus the range was somewhat wider than in 1 wt% H2SO4 solution (from −250 mVAg/AgCl to 300 mVAg/AgCl). There were clear differences in the values of Epass between the alloys. The lowest values were detected for the grade S31655 (from −300 mVAg/AgCl to 100 mVAg/AgCl), while the highest values were related to the grade S32101 (from −100 mVAg/AgCl to 350 mVAg/AgCl). The values showed a systematic increase with the rise in test temperature, yet no clear dependency on the solution NaCl concentration could be detected. The parameter that more clearly indicates the ease of passivation is Epass-Ecorr, Figure 15(b), revealing differences between the three alloys. Up to the temperature of 90°C, the values for Epass-Ecorr were lowest for the duplex-grade S32101 (ranging approximately between 20 mV and 100 mV), indicating the easiest passivation, while the highest values were obtained for the grade S31603 (from 100 mV to 250 mV), conversely suggesting that passivation required significant boost by the oxidation capacity of the environment. Furthermore, up to 90°C, there was no clear trend between the values for Epass-Ecorr and test conditions (temperature and NaCl concentration), whereas temperature increase to 130°C significantly increased the values for Epass-Ecorr for all three grades, reflecting that reaching the passive state became significantly more difficult than at lower temperatures. Eb values for the three alloys, Figures 15(c) and (d), revealed transpassivity in the potential range 850 mVAg/AgCl to 1,000 mVAg/AgCl and evident passivity breakdown only at the highest NaCl concentrations at the temperatures of 90°C and 130°C. In such cases where passivity breakdown occurred, a linear dependency on chloride ion activity was observed (Figure 15[d]). As for the Erep, Figure 15(e), the values at 90°C represented the grade S31655 and were in the range from −30 mVAg/AgCl to −70 mVAg/AgCl. At 130°C, the values fell in the range from −150 mVAg/AgCl to 400 mVAg/AgCl, but no clear dependency on the test conditions (alloy composition and NaCl concentration) could be disclosed. The reason for this could be the simultaneously occurring uniform corrosion.
FIGURE 15.

Important potential values for the passivity of the alloys in 10 wt% sulfuric acid. (a) Epass, (b) Epass-Ecorr, (c) Eb, (d) Eb as a function of aCl, and (e) Erep.

FIGURE 15.

Important potential values for the passivity of the alloys in 10 wt% sulfuric acid. (a) Epass, (b) Epass-Ecorr, (c) Eb, (d) Eb as a function of aCl, and (e) Erep.

Close modal
Values of icrit for the alloys in 10 wt% H2SO4 solution are shown in Figure 16(a). Similar to 1 wt% H2SO4 solution, the active-passive transition in grade S32101 did not exhibit a clear icrit in the temperature range from 22°C to 90°C, hence the analysis of results within this temperature range concentrates on the austenitic grades. In general, the values for the austenitic grades were at least one magnitude higher than the corresponding values in 1 wt% H2SO4 solution (Figure 9[a]), ranging from 10 µA/cm2 up to 1.4×105 µA/cm2, yet there were occasional data points in pure H2SO4 solution (with no added NaCl) that were slightly lower. This is an indication that the formation of the passive film became more challenging at the more concentrated H2SO4. The values of icrit were on average lower for the grade S31603 than for S31655, consistent with the observations in 1 wt% H2SO4 solution. The consistent trend for all cases was that the values increased significantly with an increase in temperature, with the highest values being detected at the highest test temperature. The values also increased with an increase in NaCl concentration of the solution, yet in such a systematic was as in the case of temperature. Most of the measured current density values in the passive state, ipass, Figure 16(b), fell in the range of 1 µA/cm2 to 4 µA/cm2, thus on average, they were slightly higher than in 1 wt% H2SO4. Nevertheless, at the temperatures of 90°C and particularly 130°C, the values were much higher than in 1 wt% H2SO4, reaching occasionally the level of almost 3 mA/cm2. At 90°C, the data points around or above 10 µA/cm2 represented the austenitic grades S31655 and S31603, while the values of ipass for the duplex grade S32101 were essentially lower. At 130°C, where the polarization curves for the grade S32101 also involved a clear icrit, the values of ipass for all three alloys were above 10 µA/cm2 (except for at the NaCl content of 500 mg/L for S31603), yet the average values for the grade S32101 were one magnitude lower than those for the grades S31603 and S31655. These observations further confirm the findings that in 10 wt% sulfuric acid, the passive films on austenitic Alloys S31603 and S31655 were not as protective as on duplex grade S32101, as their dissolution rate in the electrolyte was essentially higher. As indicated above, this means uniform passive state corrosion, in which the enhanced passive film dissolution at the solution/film interface is compensated by the corresponding increase in the film formation at the film/metal interface, to maintain the steady-state film thickness. This process, although progressing via the passive film, consumes the metal underneath, enabled by a simultaneous vacancy formation, travel across the film, and annihilation at the film/metal interface. Figure 17 shows examples of the material surfaces examined with SEM after the experiments in 10 wt% H2SO4 solution. The dominating mode of materials degradation was evidently uniform corrosion (Figure 17[a]), consistent with high current density values in the passive area where several of the data points indicated active corrosion. Grain boundaries within the microstructure were often clearly distinguishable in microscopy investigations, indicating these were preferentially etched and that the internal areas of the grains dissolved homogeneously (Figures 17[b]through [f]). The attack along the grain boundaries was detected particularly for the austenitic grades S31603 and S31655.
FIGURE 16.

Current density values important for passivation of the three grades of stainless steel in 10 wt% sulfuric acid. (a) icrit and (b) ipass.

FIGURE 16.

Current density values important for passivation of the three grades of stainless steel in 10 wt% sulfuric acid. (a) icrit and (b) ipass.

Close modal
FIGURE 17.

SEM images, showing the examples of corrosion attack in the alloys exposed to 10 wt% sulfuric acid solution. (a) S31603, 90°C, and 5,000 mg/L; (b) S31655, 50°C, and 5,000 mg/L; (c) and (d) S31655, 90°C, and 5,000 mg/L; (e) S31603, 130°C, and 500 mg/L; and (f) S31655, 130°C, and 500 mg/L.

FIGURE 17.

SEM images, showing the examples of corrosion attack in the alloys exposed to 10 wt% sulfuric acid solution. (a) S31603, 90°C, and 5,000 mg/L; (b) S31655, 50°C, and 5,000 mg/L; (c) and (d) S31655, 90°C, and 5,000 mg/L; (e) S31603, 130°C, and 500 mg/L; and (f) S31655, 130°C, and 500 mg/L.

Close modal

In such occasional cases where the polarization curves featured a positive hysteresis with passivity breakdown and repassivation, areas with a pit-like appearance, likely formed by the initiation of several pits and their growth together (Figure 17[c]), were detected in the surfaces together with the uniform corrosion mode (Figure 17[d]). At the highest test temperature, corrosion of the alloys was severe and took place uniformly on the surfaces (Figures 17[e] and [f]), as uniform passive state corrosion.

Immersion Tests

The results from electrochemical measurements were supplemented by immersion tests for the grade S31603 stainless steel in 1 wt% H2SO4 solution containing 2,000 mg/L NaCl. The solution was further supplemented with Fe3+ in order to enhance its oxidation capacity. Weight losses for the specimens during the 28-d-experiment are shown in Figure 18. The identical tests were conducted by four laboratories, with weight losses being detected in all specimens. In one of the test sites (site 1), weight losses for the specimens were higher than in the three other cases, approximately 120 mg/cm2. In the remaining three test sites (sites 2 to 4), the weight losses for the specimens were of the same order of magnitude, in the range from 10 mg/cm2 to 30 mg/cm2. In all cases, pitting was detected in the specimens, Figure 19. However, also signs of uniform were evident in most cases, for example, via the surfaces becoming dull (Figure 19[a]), featuring coloring (Figure 19[b], like in the case of thermal aging61 ) or experiencing edge corrosion (Figure 19[a]). The latter occurred primarily in the specimens of site 1, in which the edges were rounded. The pits were often irregular in shape, reflecting they contained several nuclei. Additionally, the observation of the lacy cover supported the findings from microscopy examinations of the specimens from electrochemical measurements (Figure 10). Hence, the observations from immersion tests seamlessly supported the results from electrochemical measurements.
FIGURE 18.

Weight loss for the specimens of grade S31603 in the immersion test.

FIGURE 18.

Weight loss for the specimens of grade S31603 in the immersion test.

Close modal
FIGURE 19.

(a,b) Photographs and (c) through (f) microscopy images of the specimens from the immersion experiments. (c,e) Stereomicroscopy images, and (f) SEM image.

FIGURE 19.

(a,b) Photographs and (c) through (f) microscopy images of the specimens from the immersion experiments. (c,e) Stereomicroscopy images, and (f) SEM image.

Close modal

Table 3 presents a summary of the findings from the electrochemical measurements concerning the occurrence of major corrosion forms: pitting corrosion and uniform corrosion, and the rate of the latter. In Table 3, P in the cell indicates the susceptibility of the material to pitting corrosion under the specific combination of test conditions. In the categorization of the uniform CR, we have followed Outokumpu Corrosion Handbook11  in that the green color means that the material is compatible with the conditions and has a CR below 0.1 mm/y, yellow indicates that the material is expected to corrode at a rate between 0.1 mm/y and 1 mm/y, so it may be useable in some applications, but it is not completely resistant and red denotes that the material will corrode heavily at a rate above 1 mm/y. The results revealed that there were no significant differences between the test materials with respect to the tendency of pitting corrosion, which was expected as the materials have a very similar PREN number. However, there were differences between the test conditions, in particular between the two H2SO4 concentrations, in this respect: pitting in the materials was clearly more frequent in 1 wt% than in 10 wt% sulfuric acid solution. Only in 10 wt% H2SO4 solution minor dissimilarities in the pitting corrosion tendency between the materials were found: the duplex grade S32101 exhibited pitting corrosion only at the highest test temperature, while the two austenitic grades underwent pitting corrosion already at 90°C at least at the highest NaCl contents.

Table 3.

Summary of Detected Corrosion Forms(A)

Summary of Detected Corrosion Forms(A)
Summary of Detected Corrosion Forms(A)

Although the presented corrosion rates were likely conservative due to the nature of the methodology, e.g., use of polarization, they enable the comparison between the three alloys. Indeed, the analysis revealed evident differences between the uniform corrosion rates of the materials. The highest overall corrosion rates were related to the austenitic grade S31603 consistent with the lowest Cr alloying in the bulk alloy and in the passive film, revealed by the XPS results. In 1 wt%, the uniform CR of S31603 was low (<0.1 mm/y) only at room temperature, but already at 50°C at the highest NaCl content was the uniform CR of the alloy at a moderate level (between 0.1 mm/y and 1 mm/y). At 90°C at the highest NaCl content and at 130°C, the uniform CR of the alloy was high (>1.0 mm/y). Based on the determined ipass values, uniform corrosion at the temperatures of 90°C and lower represented uniform active state corrosion, whereas uniform corrosion at the temperature of 130°C progressed as the uniform passive state corrosion of the material. The results from immersion tests confirmed not only the presence of pitting corrosion but also the simultaneous uniform corrosion at 90°C at the NaCl concentration of 2,000 mg/L detected in the electrochemical tests. In 10 wt%, the uniform CR of the alloy was low only occasionally at room temperature and high CR was detected already at the temperature of 50°C at the highest NaCl concentration. Again, at 130°C, uniform corrosion progressed as uniform passive state corrosion. Furthermore, at the H2SO4 concentration of 10 wt%, the most corrosion-resistant alloy was the duplex-grade S32101, which contained the highest chromium alloying. Although in 1 wt% sulfuric acid the corrosion rates for the Alloys S31655 and S32101 were equal and low up to 90°C, the differences between these alloys became evident in 10 wt% sulfuric acid. For the austenitic S31655, the corrosion rates were low only at room temperature and 50°C at low NaCl concentrations, and moderate or high under all other conditions. For the duplex grade, uniform corrosion rates were retained at a low level up to 90°C, the highest NaCl concentration. It is worth mentioning that the experience has shown that although the duplex grades of stainless steel passivate easier than the austenitic grades if the duplex grades for some reason start to corrode actively, the corrosion rates may be much higher than for the austenitic grades.

Until now, the dual role of sulfuric acid in the corrosion behavior of passive alloys has received only a little research effort. Nevertheless, Yang and Macdonald62  have demonstrated that the addition of anionic corrosion inhibitors in the electrolyte shifts the Eb values toward higher activity-potential combinations. As the H2SO4 environment always contains the corrosion inhibitor, we cannot make parallel comparisons. Instead, we can evaluate the inhibitive effect of the sulfate ion on the passivity breakdown via plotting all of the Eb values for the three alloys as a function of activity ratio , as presented in Figure 20. The linear dependency demonstrates that Eb is dependent on , thus not only the presence of chloride ions but also that of sulfate ions contributes to the pitting tendency of the materials. This approach allows for a critical activity ratio to be defined. Table 4 presents a summary of the detected corrosion forms and uniform corrosion rates as a function of the activity ratio of chlorides to sulfates, . It can be seen that in 1 wt% H2SO4 solution, no pitting corrosion was detected under the conditions with the of 10 or lower, except for the austenitic grades at 130°C for which also uniform corrosion was detected. However, as the rate of uniform corrosion was very high at the highest test temperature (uniform passive state corrosion), the relative importance of pitting corrosion and uniform corrosion is challenging to define. Upon sifting to higher H2SO4 content, the overall values of decreased due to the enhanced contribution of hydrogen ions, H+, to the equilibrium chemistry. However, due to pH decrease, the passive films on the alloys became thermodynamically less stable and their dissolution was accelerated as compared to the 1 wt% sulfuric acid. The results in Table 4 revealed that no pitting corrosion in 10 wt% occurred at lower than 30, yet many of the cases in 10 wt% sulfuric acid, categorized here as pitting corrosion, also involved uniform corrosion (Figure 17). It is worth mentioning that if the solution is further supplemented with additional constituents, such as Fe2(SO4)3 that was added in the electrolyte in the immersion tests to introduce Fe3+ for the solution redox increase,27  it simultaneously adds the concentration of the sulfate ion [with of 2.8]. In such cases, the critical does not hold anymore. Nevertheless, the results from immersion tests that were used as a fast validation of the results from electrochemical tests were able to replicate the key corrosion mechanisms.
FIGURE 20.

Eb as a function of the ratio of chloride and sulfate activities, . Black data points refer to the datasets in 1 wt% sulfuric acid solutions, while light gray and white datasets represent 10 wt% sulfuric acid solutions.

FIGURE 20.

Eb as a function of the ratio of chloride and sulfate activities, . Black data points refer to the datasets in 1 wt% sulfuric acid solutions, while light gray and white datasets represent 10 wt% sulfuric acid solutions.

Close modal
Table 4.

Summary of Detected Corrosion Forms as the Function of Activity Ratio of Chlorides to Sulfates, (the Values Given in the Cells)(A)

Summary of Detected Corrosion Forms as the Function of Activity Ratio of Chlorides to Sulfates,  (the Values Given in the Cells)(A)
Summary of Detected Corrosion Forms as the Function of Activity Ratio of Chlorides to Sulfates,  (the Values Given in the Cells)(A)

The analysis of the contribution of pH value to the corrosion phenomena is complex. Values for icorr and ipass were proved to be at least one order of magnitude higher in 10 wt% H2SO4 than in 1 wt% H2SO4 solution, which is likely related to the lower pH of the more concentrated solution. However, the increase in sulfate ion content also adds to the solution conductivity, thereby facilitating the dissolution. In this respect, we likely observed the combined effect of pH and conductivity in the values for icorr and ipass. Another approach to analyze the contribution of pH to the corrosion of the alloys is via investigation of corrosion morphologies. It is known that localized corrosion results in the acidification of the solution within the pit (or crevice) as a result of the hydrolysis of metal cations, which then leads to the further acceleration of the localized attack.63  In 1 wt% H2SO4 solution, pitting corrosion was essentially the dominant corrosion mechanism. The detected large pit size and their low number may be linked with the contribution of sulfate ions through the inhibition of pit initiation and growth, but it may equally be associated with the overall low pH of the solution, which supports the dissolution of the metal in the formed attack site. In contrast to the lower H2SO4 content, uniform corrosion was the prevailing corrosion form in 10 wt% H2SO4 solution. Earlier, Szklarska-Smialowska13  has concluded that at very low pH values, pitting transfers into uniform corrosion. It is therefore expected that the uniform dissolution of the passive film undermines the importance of individual passive film breakdown sites. As indicated above, both low pH and conductivity likely play a role here.

Concerning the reliability of the results, at least two parallel measurements were systematically performed at least in such cases where the polarization curve contained a hysteresis loop. The average value is given above. The scatter in the results was very low, as can be seen in the sequential data points that were often overlapping. For example, in the case of OCP of austenitic grades, the values under all test conditions for each type of test environment (1 wt% or 10 wt% H2SO4) typically fell within the 50 mV range and the differences detected between the test materials were greater essentially greater than 50 mV. Independent of which electrochemical parameter was examined, the observations could be explained and were supported by both literature and microstructural investigations.

The findings presented in this paper enable the following conclusions to be drawn:

  • Sulfuric acid concentration influences the corrosion performance of stainless steel via two mechanisms: inhibitive effect which makes the ratio of chloride and sulfate activities rather than solely the chloride activity an important parameter for pitting initiation, and pH effect, which contributes to the dissolution rate of the steel and, in particular, the passive film on the surface. In H2SO4-NaCl systems, no pitting corrosion was detected below the chloride to sulfate activity ratio of 10.

  • Pitting corrosion was the dominant corrosion form in 1 wt% sulfuric acid solution for all three studied alloys. The susceptibility of alloys to pitting corrosion increased with increase in temperature and NaCl concentration of the solution, consistent with raising chloride-to-sulfate activity ratio, and no significant differences in pitting tendency between the alloys could be detected. In each case, pits were few in number on the surfaces, and often contained several pit nuclei, resulting in a large pit size. At the lowest test temperatures, the pits were open and did not contain a cover. At higher test temperatures, the pits contained a lacy cover and were formed by several pits growing together into worm-like attack sites and craters. Pitting with the formation of the lacy cover was also evidenced in 28-d immersion experiments.

  • In 10 wt% sulfuric acid solutions, uniform corrosion was the prevailing corrosion form. There were significant differences in the uniform corrosion rates between the materials estimated from the potentiodynamic polarization curves, with the lowest rates being systematically related to the grade S32101 with the highest Cr content and duplex microstructure, and the highest rates being conversely associated with the grade S31603 with the lowest Cr concentration and austenitic microstructure. The observations strongly correlate with the levels of ipass measured under the test conditions. In such cases where the uniform CR was moderate, the grain boundaries were preferentially etched.

  • At the highest test temperature, 130°C, where the uniform CR was high for all test materials, the degradation progressed as the uniform passive state corrosion. This was justified by much higher ipass values than at the lower temperatures.

  • The results revealed the importance of understanding the equilibrium solution chemistry in the analysis of the material’s behavior in multispecies electrolytes. They also demonstrated that despite the nominal identical corrosion resistance, as evaluated based on PREN values, differences in the corrosion performance may exist and they become more evident when the operation environment becomes more demanding.

(1)

UNS numbers are listed in Metals & Alloys in the Unified Numbering System, published by the Society of Automotive Engineers (SAE International) and cosponsored by ASTM International.

Trade name.

The financial support by EIT Raw Materials and the partner organizations: VTT Technical Research Centre of Finland Ltd., Metso Outotec Finland Oy, ZAG Slovenian National Building, and Civil Engineering Institute, Tecnalia, Outokumpu Stainless AB, Boliden Harjavalta Oy, Ferritico AB, and Data Measuring Systems DMS, for the CORTOOLS project under the acceleration program (Project No. 18158) is gratefully acknowledged.

1.
Watari
T.
,
Nansai
K.
,
Nakajima
K.
,
Resour. Conservat. Recycl.
164
(
2021
):
p
.
105107
.
2.
Tokimatsu
K.
,
Höök
M.
,
McLellan
B.
,
Wachtmeister
H.
,
Murakami
S.
,
Yasuoka
R.
,
Nishio
M.
,
Appl. Energy
225
(
2018
):
p
.
1158
1175
.
3.
Habib
K.
,
Hansdottir
S.T.
,
Habib
H.
,
Resour. Conservat. Recycl.
154
(
2020
):
p
.
104603
.
4.
Spooren
J.
,
Binnemans
K.
,
Björkmalm
J.
,
Breemersch
K.
,
Dams
Y.
,
Folens
K.
,
Gonzalez-Moya
M.
,
Horckmans
L.
,
Komnitsas
K.
,
Kurylak
W
,
Lopez
M.
,
Mäkinen
J.
,
Onisei
S.
,
Oorts
K.
,
Peys
A.
,
Pietek
G.
,
Pontikes
Y.
,
Snellings
R.
,
Tripiana
M.
,
Varia
J.
,
Willqvist
K.
,
Yurramendi
L.
,
Kinnunen
P.
,
Resour. Conservat. Recycl.
160
(
2020
):
p
.
104919
.
5.
Mikula
K.
,
Izydorczyk
G.
,
Skrzypczak
D.
,
Moustakas
K.
,
Witek-Krowiak
A.
,
Chojnacka
K.
,
J. Hazard. Mater.
403
(
2021
):
p
.
123602
.
6.
Watling
H.R.
,
Hydrometallurgy
146
(
2014
):
p
.
96
110
.
7.
Kobylin
P.M.
,
Sippola
H.
,
Taskinen
P.
,
CALPHAD
40
(
2013
):
p
.
41
47
.
8.
Kobylin
P.M.
,
Sippola
H.
,
Taskinen
P.
,
CALPHAD
38
(
2012
):
p
.
185
193
.
9.
Okamoto
G.
,
Corros. Sci.
13
(
1973
):
p
.
471
489
.
10.
Sato
N.
,
Corros. Sci.
31
(
1990
):
p
.
1
19
.
11.
Outokumpu Handbook of Stainless Steel
(
Espoo, Finland
:
Outokumpu Oyj
,
2015
).
12.
Hoar
T.P.
,
Jacob
W.R.
,
Nature
216
(
1967
):
p
.
1299
1301
.
13.
Szklarska-Smialowska
Z.
,
Pitting and Crevice Corrosion
(
Houston, TX
:
NACE International
,
2005
).
14.
Pardo
A.
,
Merino
M.C.
,
Coy
A.E.
,
Viejo
F.
,
Arrabal
R.
,
Matykina
E.
,
Corros. Sci.
50
(
2008
):
p
.
780
794
.
15.
Metikos-Hukovic
M.
,
Babic
R.
,
Grubac
Z.
,
Petrovic
Z.
,
Lajci
N.
,
Corros. Sci.
53
(
2011
):
p
.
2176
2183
.
16.
Loable
C.
,
Vicosa
I.N.
,
Mesquita
T.J.
,
Mantel
M.
,
Nogueira
R.P.
,
Berthome
G.
,
Chauveau
E.
,
Roche
V.
,
Mater. Chem. Phys.
186
(
2017
):
p
.
237
245
.
17.
Urquidi-Macdonald
M.
,
Macdonald
D.D.
,
J. Electrochem. Soc.
136
(
1989
):
p
.
961
967
.
18.
Devine
T.M.
,
J. Electrochem. Soc.
126
(
1979
):
p
.
374
385
.
19.
Peguet
L.
,
Malki
B.
,
Baroux
B.
,
Corros. Sci.
51
(
2009
):
p
.
493
498
.
20.
Pradhan
S.K.
,
Bhuyan
P.
,
Mandal
S.
,
Corros. Sci.
158
(
2019
):
p
.
108091
.
21.
Park
J.O.
,
Matsch
S.
,
Böhni
H.
,
J. Electrochem. Soc.
149
(
2002
):
p
.
B34
B39
.
22.
Laylock
N.J.
,
Newman
R.C.
,
Corros. Sci.
40
(
1998
):
p
.
887
902
.
23.
Khedr
A.A.
,
Sanad
S.H.
,
El-Sobki
K.M.
,
Surf. Technol.
23
(
1984
):
p
.
151
158
.
24.
Hong
T.
,
Nagumo
M.
,
Corros. Sci.
39
(
1997
):
p
.
285
293
.
25.
Ernst
P.
,
Newman
R.C.
,
Corros. Sci.
49
(
2007
):
p
.
3705
3715
.
26.
Malik
A.U.
,
Mayan Kutty
P.C.
,
Siddiqi
N.A.
,
Andijani
I.N.
,
Ahmed
S.
,
Corros. Sci.
33
(
1992
):
p
.
1809
1827
.
27.
Lindgren
M.
,
Huttunen-Saarivirta
E.
,
Peltola
H.
,
Romu
J.
,
Sarikka
T.
,
Hänninen
H.
,
Pohjanne
P.
,
Corrosion
74
(
2018
):
p
.
225
240
.
29.
Varga
K.
,
Maleczki
E.
,
Horanyi
G.
,
Electrochim. Acta
33
(
1988
):
p
.
1775
1778
.
30.
Varga
K.
,
Baradlai
P.
,
Barnard
W.O.
,
Myburg
G.
,
Halmos
P.
,
Potgieter
J.H.
,
Electrochim. Acta
42
(
1997
):
p
.
25
35
.
31.
Marcus
P.
,
Electrochim. Acta
43
(
1998
):
p
.
109
118
.
32.
Leckie
H.P.
,
Uhlig
H.H.
,
J. Electrochem. Soc.
113
(
1966
):
p
.
1262
1267
.
33.
Pistorius
P.C.
,
Burnstein
G.T.
,
Corros. Sci.
33
(
1992
):
p
.
1885
1897
.
34.
Hong
T.
,
Nagumo
M.
,
Corros. Sci.
39
(
1997
):
p
.
961
967
.
35.
Ernst
P.
,
Newman
R.C.
,
Corros. Sci.
44
(
2002
):
p
.
943
954
.
36.
Moayed
M.H.
,
Newman
R.C.
,
Corros. Sci.
40
(
1998
):
p
.
519
522
.
37.
Moayed
M.H.
,
Newman
R.C.
,
Corros. Sci.
48
(
2006
):
p
.
3513
3530
.
38.
Deng
B.
,
Jiang
Y.
,
Liao
J.
,
Hao
Y.
,
Zhong
C.
,
Li
J.
,
Appl. Surf. Sci.
253
(
2007
):
p
.
7369
7375
.
39.
Ekman
S.
, “
Selecting Specialty Stainless Steel Grades for Sulfuric Acid Leaching Environment
,”
Alta 2011 Nickel/Cobalt/Copper Conference
,
Perth, Australia
.
40.
Hren
M.
,
Kosec
T.
,
Lindgren
M.
,
Huttunen-Saarivirta
E.
,
Legat
A.
,
Sensors
21
(
2021
):
p
.
1449
.
41.
Loto
R.T.
,
J. Mater. Res. Technol.
8
(
2019
):
p
.
5415
5421
.
42.
Laitinen
T.
,
Corros. Sci.
42
(
2000
):
p
.
421
441
.
43.
Pohjanne
P.
,
Carpén
L.
,
Hakkarainen
T.
,
Kinnunen
P.
,
J. Construct. Steel Res.
64
(
2008
):
p
.
1325
1331
.
44.
Sridhar
N.
,
Anderko
A.
,
Corrosion
77
(
2021
):
p
.
935
948
.
46.
Gardin
E.
,
Zanna
S.
,
Seyeux
A.
,
Allion-Maurer
A.
,
Marcus
P.
,
Corros. Sci.
155
(
2019
):
p
.
121
133
.
47.
Olefjord
I.
,
Wegrelius
L.
,
Corros. Sci.
31
(
1990
):
p
.
89
98
.
48.
Olefjord
I.
,
Wegrelius
L.
,
Corros. Sci.
38
(
1996
):
p
.
1203
1220
.
49.
Wegrelius
L.
, “
Passivation of Austenitic Stainless Steel
,” (
Ph.D. thesis diss.
,
Department of Engineering Metals
,
Göteborg
1995
),
ISBN 91-7197-182-3
.
50.
Brooks
A.R.
,
Clayton
C.R.
,
Doss
K.
,
Lu
Y.C.
,
J. Electrochem. Soc.
133
(
1986
):
p
.
2459
2464
.
51.
Liu
C.T.
,
Wu
J.K.
,
Corros. Sci.
49
(
2007
):
p
.
2198
2209
.
52.
Clayton
C.R.
,
Lu
Y.C.
,
J. Electrochem. Soc.
133
(
1986
):
p
.
2465
2473
.
53.
Helmersson
B.
, “
Effect of Acid Concentration and Temperature on Active Corrosion Rate of Stainless Steel and Nickel Base Materials
,”
CORROSION 2019, paper no. 13330
(
Houston, TX
:
NACE
,
2019
).
54.
Nicic
I.
,
Macdonald
D.D.
,
J. Nucl. Mater.
379
(
2008
):
p
.
54
58
.
55.
Metikos-Hukovic
A.
,
Babic
R.
,
Grubac
Z.
,
Petrovic
Z.
,
Lajci
N.
,
Corros. Sci.
53
(
2011
):
p
.
2176
2183
.
56.
MacDonald
D.D.
,
J. Electrochem. Soc.
139
(
1992
):
p
.
3434
3449
.
57.
Pyun
S.-I.
,
Moon
S.-M.
,
J. Solid State Electrochem.
3
(
1999
):
p
.
331
336
.
58.
Ernst
P.
,
Laylock
N.J.
,
Moayed
M.H.
,
Newman
R.C.
,
Corros. Sci.
39
(
1997
):
p
.
1133
1136
.
59.
Li
Y.
,
Ives
M.B.
,
Coley
K.S.
,
Rodda
J.R.
,
Corros. Sci.
46
(
2004
):
p
.
1969
1979
.
60.
Jegdic
B.
,
Drazic
D.M.
,
Popovic
J.P.
,
Corros. Sci.
50
(
2008
):
p
.
1235
1244
.
61.
Huttunen-Saarivirta
E.
,
Kuokkala
V.-T.
,
Pohjanne
P.
,
Corros. Sci.
87
(
2014
):
p
.
344
365
.
62.
Yang
S.
,
Macdonald
D.D.
,
Electrochim. Acta
52
(
2007
):
p
.
1871
1879
.
63.
McCafferty
E.
,
Introduction to Corrosion Science
(
New York, NY
:
Springer
,
2009
).