The corrosion behavior of stainless steels and Ni-based alloys in nonoxidizing sulfuric acid mixtures at concentrations below approximately 30 mol/kg H2O is modeled. The redox potential in sulfuric acid across a broad concentration range, from 0 to 80 mol% (0 to 95.6 wt%), is determined by the proton reduction reaction. Thus, in the absence of other oxidizing species, sulfuric acid behaves as a nonoxidizing (reducing) acid. The calculated corrosion rates, using an electrochemical model up to about 30 mol/kg H2O (about 75 wt%), are in agreement with experimental values. The predicted polarization curves of anodic and cathodic processes show that the alloys in these environments are in active dissolution regime, consistent with experimental data. The model predictions of corrosion rates in H2SO4+HCl, H2SO4+HF, and H2SO4+HCl+HF mixtures are in agreement with weight-loss corrosion data. The corrosion rate of alloys in the nonoxidizing sulfuric acid mixtures correlated to an equivalent alloy composition given by (Ni0.7-Cr0.1+Mo+0.5 W). The effect of alloying elements under these conditions may be related to their beneficial effect on active dissolution and proton reduction reaction rates.

INTRODUCTION

Sulfuric acid is a large volume chemical, with the world production estimated to be about 270 million metric tons in 2018.1  The need to reduce sulfur emissions from fossil fuel usage may be a driver for increased sulfuric acid production in the future. Sulfuric acid mixtures occur in many applications, such as spent acid recovery, crude oil refining, chemical processing, agricultural and pharmaceutical chemical production, metal descaling, electropolishing, and hydrometallurgical treatment.2  The acid mixtures can be broadly classified as: (1) mixtures of two nonoxidizing (also referred to as “reducing”) strong acids, (2) nonoxidizing mixtures with one strong acid containing an aggressive species, such as a halide ion, (3) mixtures in which one of the acids or salts has an inhibitive species, such as a nitrate ion, and (4) mixtures in which one component generates a high redox potential, such as ferric or nitrate ions. In nonoxidizing acids, the main cathodic reaction is the reduction of protons and the redox potential generally follows the H+/H2 equilibrium. The presence of halides in the nonoxidizing solutions generally does not lead to localized corrosion of stainless steels and nickel-based alloys, but can lead to high rates of active corrosion.3  In oxidizing acid mixtures, the cathodic reactions generate a redox potential higher than that of the proton reduction.3  In the absence of an aggressive species, such as chloride, the oxidizing agent can have an inhibitive effect on corrosion if the resultant corrosion potential is in the passive dissolution range of the anodic polarization behavior of the alloy. On the other hand, if an alloy is not passive, the presence of an oxidizing agent can accelerate corrosion. The presence of halides in oxidizing acid mixtures can lead to localized corrosion, depending on the alloy composition, the halide species concentration, and other species in solution.4  In the absence of halides, the corrosion rates can be high, if the redox potential is in the transpassive regime. This paper is focused on nonoxidizing sulfuric acid mixtures with chlorides and fluorides.

The corrosion behavior of stainless steels and Ni-based alloys in sulfuric acid has been studied extensively over many decades.5-19  Although the electrochemical behavior of alloys in dilute sulfuric acid is reasonably well understood, the behavior in concentrated acids (above about 70 wt%) exhibits interesting features. For example, the redox potential measured in sulfuric acid on platinized Pt shows an inflection at about 68 wt%, above which the redox potential rises to very high values.3,20  However, the corrosion behavior of alloys suggests that sulfuric acid across its whole concentration range behaves as a reducing environment in the absence of oxidizing species, such as ferric salts. Furthermore, mixtures of concentrated sulfuric acid with other reducing acids, such as HCl and HF, also behave as reducing environments. While chloride accelerates the corrosion rate of alloys when mixed with sulfuric acid, HF exhibits a variable tendency, accelerating corrosion at low sulfuric acid concentrations and inhibiting corrosion at high sulfuric acid concentrations.20 

Despite significant experimental work, a general model to predict the corrosion behavior of stainless steels and Ni-based alloys in acid mixtures does not exist. Such an electrochemical model can be useful in three ways. First, it can be used to correlate the available experimental data as a function of mixture composition and temperature and to predict the corrosion rates at conditions that have not been experimentally investigated. Second, the observed corrosion behavior can be rationalized in terms of electrochemical processes and help in understanding the corrosion mechanisms. Third, the model can be used as an engineering tool for guiding the initial choice of alloys that could be tested in complex, multicomponent solutions often encountered in the chemical process industries. The model cannot replace the need for experimental data, but can provide a valuable starting point for process-specific testing. In this paper, the results of an electrochemical model, based on concentrated electrolyte speciation, are presented to predict the corrosion behavior of various stainless steels and Ni-based alloys in sulfuric acid mixtures.

EXPERIMENTAL DATA

The fully immersed corrosion experiments reported previously3,20  were performed on a range of corrosion-resistant alloys, whose nominal compositions are shown in Table 1. Most of the corrosion tests were conducted in solutions that were not intentionally deaerated (the solutions were exposed to still air during preparation and transfer to test vessels). Coupons, approximately 2.54 cm × 5 cm × 3.2 mm thick, were ground to 120-grit finish, cleaned and dried, and weighed before exposure to solutions at different temperatures. After exposure, they were cleaned and weighed again to determine average corrosion rate. Duplicate tests were performed and the average corrosion rates are reported. All samples were without welds and in the mill annealed or as-received condition. Tests were typically conducted for 24 h to 96 h. Tests in HF-containing solutions were performed in polytetrafluoroethylene (PTFE) vessels. Visual examination of corrosion was performed after the tests.

Table 1.

Nominal Compositions (wt%) of the Alloys Presented in This Paper(A)

Nominal Compositions (wt%) of the Alloys Presented in This Paper(A)
Nominal Compositions (wt%) of the Alloys Presented in This Paper(A)

The electrochemical test results were reported previously.3  The redox potentials were measured on platinized Pt in acid solutions that were purged with 1 atm of high-purity hydrogen against a saturated calomel electrode (SCE).20  The anodic and cathodic polarization curves were conducted using a scan rate of 0.167 mV/s. Potentiostatic tests were also conducted and weight-loss corrosion rates were measured on the potentiostated specimens.3  Additionally, corrosion data reported in the literature are used and, where necessary, those experimental conditions are briefly mentioned.

MODELING APPROACH

The speciation in concentrated sulfuric acid mixtures can be understood using the mixed solvent electrolyte (MSE) model.21-22  An essential aspect of modeling such systems is the computation of excess Gibbs energy and the activity coefficients through a sum of three terms: (i) the long-range electrostatic interactions between ions, (ii) the short-range interactions between neutral molecules, neutral molecules and ions, and between ion pairs, and (iii) ionic interactions between ion pairs and ions and molecules. The activity coefficient calculations are combined with the Helgeson-Kirkham-Flowers equation of state for the computation of standard-state properties of individual species in solution. The combination of standard-state property and activity coefficient formulations with mass and charge balances makes it possible to compute equilibrium speciation in solutions. The MSE model is applicable to systems ranging from dilute aqueous solutions to very concentrated mixtures up to the nonaqueous solute (e.g., pure acid) limit.

For electrochemical modeling, a previously developed mixed-potential model is used to represent the corrosion behavior as a function of environment composition, temperature, and pressure.23-25  Unlike the case of the MSE model for speciation, the electrochemical kinetics is modeled using an aqueous chemistry thermophysical model, which is valid for electrolyte solutions with a total ionic strength up to only approximately 30 mol of solute/kg H2O. The thermophysical module predicts the speciation in the aqueous environment and yields the activities and transport properties of solution species that participate in interfacial reactions. This information is then utilized in the electrochemical module to simulate the kinetics of electrochemical reactions. The electrochemical module includes various partial reactions that may occur on the surface of the metal and the transport processes for the electrochemically active species. In particular, the model accounts for the active-passive transition and dissolution in the passive state. In the most general case, alloy dissolution in the active state is modeled using the Butler-Volmer equation written as:
formula
where the solution chemistry effects are accounted for by considering the adsorption of electrochemically active species and θi are the surface coverage fractions of species i. In Equation (1), iM is the metal dissolution current, n is the average number of electrons, ka is the rate constant, αa is the electrochemical transfer coefficient, xi are the reaction orders for species i, E is the potential, E0 is the equilibrium potential, T is the temperature in K, and F and R are the Faraday and gas constants, respectively. The reverse reaction of metal deposition is ignored in this formulation. The active-passive transition is introduced by considering a current, iMO, that leads to the formation of a passive layer in addition to the current that is associated with active dissolution, iM. In the steady-state limit, the expression for the total anodic current, iM,TOT, becomes:23-24,26 
formula
where ip is the passive current density, representing the ratio of the rates of formation and dissolution of passive film. It should be noted that in the absence of oxide formation, the active dissolution rate, iM, will be limited by the transport of dissolved metallic species (e.g., Ni2+) away from the electrode. However, the solubility of Ni2+ in H2SO4 as determined by the precipitation of NiSO4·1H2O is quite high (about 1.7 m in a 10 m H2SO4 solution). The diffusion limited current density can be estimated to be about 100 A/cm2 assuming a diffusion layer thickness of 0.01 cm. Thus, the anodic dissolution behavior is not limited by diffusion of dissolved products. In comparison, the FeSO4 solubility is much lower (about 0.03 m FeSO4 and 0.04 m Fe2+ in 10 m H2SO4) leading to a diffusion-limited current density of about 2 A/cm2. Thus, a diffusion-limited dissolution rate, influenced by agitation, may be expected.
With appropriate parameterization, Equation (2) yields the critical current density, Flade potential, and passive dissolution range. The passive dissolution and active-passive transition are related to solution chemistry by considering surface reactions between the oxide film and solution species. In acidic solutions, such a reaction can be expressed as:
formula
where the symbol “≡” denotes surface species. In addition to the acidity of the environment, some active ions (e.g., halides) may influence the dissolution of the oxide film. The effect of active species is modeled by considering specific surface reactions between these species and the metal oxide film, i.e.,
formula
where Xi is the ith reactive species in the solution and the subscripts a, b, ci, x, and ei represent the reaction stoichiometry. The surface species that forms as a result of Reaction (4) may subsequently undergo dissolution reactions such as:
formula
Mathematical analysis of such reactions yields a relationship between the passive current density and the activities of reactive species at the interface, i.e.,23,26 
formula
where is the passive dissolution rate in the absence of specific species and depends primarily on the solution’s acidity, li is the forward rate of Reaction (5), and Ki is the equilibrium constant of Reaction (4).
In addition to the alloy dissolution reactions, the model incorporates expressions for various cathodic reactions that may occur on the surface of the metal. In mixed acid environments considered here, the cathodic reactions include the reduction of protons, water, undissociated acid molecules, oxidizing anions, such as NO3, and oxidizing cations, such as Fe3+. In a general form, the current density due to a cathodic reaction is written as a combination of activation-controlled current density, ic,θ, and transport-controlled current density, ic,l (Equation [7]):
formula
The activation controlled current density is given by a Butler-Volmer expression (Equation [8]):
formula
The limiting cathodic current density, ic,l, is determined by the bulk and surface concentrations of species. The surface activities, aj*, which are necessary to compute the limiting cathodic and anodic current densities, are related to the bulk activities, aj, using the flow regime-dependent mass transfer coefficients, km:
formula
The expressions for the partial cathodic and anodic processes are combined into a total predicted polarization curve and the corrosion current density and potential are calculated based on the mixed-potential theory (assuming equal anodic and cathodic areas), i.e.,
formula
where ic,i and ia,j denote the ith cathodic and jth anodic process. The necessary parameters in the expressions for the partial current densities are determined on the basis of corrosion rate, corrosion potential, and polarization curve measurements. The model parameters are typically determined on the basis of data for alloys in single-solute systems. Then, they are used for predictions for mixed-acid systems. The only exception to this rule occurs when data for single-solute systems are not available or show discrepancies for a given alloy, which necessitates the use of mixed-acid data, if available. The speciation and electrochemical models were run using the OLI Studio Version 10.0.1.24(1).

RESULTS

Speciation in Sulfuric Acid Mixtures

The chemistry of sulfuric acid has been studied extensively and the speciation of sulfuric acid can be modeled reasonably well, including the oleum range.22  The modeling results for sulfuric acid across the concentration range are consistent with speciation determined by Raman spectroscopy.27  The mol% of various sulfur-containing species in H2SO4 is shown in Figure 1. The corresponding wt% of H2SO4 is also indicated in the figure. Sulfuric acid can be roughly divided into three chemical speciation regimes: (1) in dilute solutions below about 5 mol%, both bisulfate (HSO4) and sulfate (SO42–) are present, the latter in higher proportion below about 0.1 mol%(Figure 2); (2) the SO42– species reaches a maximum at about 10 mol% H2SO4, and beyond that concentration, the solution continues to be dominated by HSO4, which reaches a maximum at about 40 mol% H2SO4; and (3) beyond 40 mol% H2SO4, the H2SO40 (the superscript 0 indicates undissociated, dissolved species) concentration increases and it becomes the dominant species above 65 mol% (90 wt%) acid. At 99% acid, the dissolved SO3 species starts increasing and becomes dominant in oleum. The hydronium ion (H3O+) activity reaches a maximum at 45 mol% H2SO4, whereas the dielectric constant reaches a minimum. The decrease in the latter would imply that the concentration of fully dissociated ionic species will be low. The molar concentrations of HSO4 and H2SO40 determined from careful Raman spectroscopy by Walrafen, et al.,27  are in agreement with the calculated speciation. Raman spectroscopy27  indicates that below about 60 mol% H2SO4, the hydronium ion (H3O+) is associated with HSO4, whereas above this concentration, H3O+ is increasingly associated with undissociated H2SO40. The lowering of dielectric constant and the association of ionic species suggest that the concentrations of other species, such as chloride, depend on the H2SO4 concentration.

FIGURE 1.

Speciation in sulfuric acid at 25°C, using mixed solvent electrolyte (MSE) model.

FIGURE 1.

Speciation in sulfuric acid at 25°C, using mixed solvent electrolyte (MSE) model.

FIGURE 2.

Sulfuric acid speciation at low concentrations predicted by MSE model.

FIGURE 2.

Sulfuric acid speciation at low concentrations predicted by MSE model.

The HCl speciates as fully dissociated Cl ions and undissociated HCl0, the proportion of dissociated Cl decreasing with increasing HCl concentration. In dilute HCl, the dissociated Cl dominates. When mixed with H2SO4, the ratio of dissociated Cl to undissociated HCl0 decreases with increasing H2SO4 (Figure 3). Beyond about 30 mol% to 40 mol% H2SO4, the undissociated HCl0 dominates. In the case of HF, the undissociated HF0 is the dominant species in pure HF and dilute mixtures of HF + H2SO4. However, fluosulfonic acid (HSO3F) and fluosulfonate ion (SO3F) assume greater significance at higher H2SO4 concentrations (Figure 4). These predicted species are consistent with ion chromatography, Raman spectroscopy, infrared spectroscopy, and nuclear magnetic resonance (NMR) data of a 9 vol% of 98 wt% H2SO4 + 1 vol% of 49 wt% HF mixture.28  For example, in the above solution, equivalent to 75 mol% H2SO4 + 25.4 mol% HF, Ford, et al.,28  reported a ratio of SO3F/HF0 of 0.25 from NMR, whereas the MSE model predicts a ratio of 0.09.

FIGURE 3.

Speciation of HCl in H2SO4 as predicted by the MSE model.

FIGURE 3.

Speciation of HCl in H2SO4 as predicted by the MSE model.

FIGURE 4.

Predicted speciation in 10 mol% HF + H2SO4 + balance H2O solutions at 25°C.

FIGURE 4.

Predicted speciation in 10 mol% HF + H2SO4 + balance H2O solutions at 25°C.

Redox Potential in Sulfuric Acid

The redox potential measured with platinized Pt in hydrogen-purged sulfuric acid is replotted in Figure 5 as a function of the calculated pH values. It can be seen that at concentrations of H2SO4 higher than 68 wt%, the redox potential rises steeply to high values.20  Kish and Ives29  and Kish, et al.,30  suggested that concentrated H2SO4 behaves as an oxidizing acid, likely due to redox reactions involving S, SO2, or H2S. Stypula and Banas31  and Renner32  suggested that the high redox potential was due to the redox reactions involving SO2:
formula
formula
formula
FIGURE 5.

Redox potential of platinized Pt in sulfuric acid purged with 1 atm H2 at 25°C. Original data from Sridhar.20 

FIGURE 5.

Redox potential of platinized Pt in sulfuric acid purged with 1 atm H2 at 25°C. Original data from Sridhar.20 

However, hydronium reduction reaction (Equation [14]) may also constitute the governing redox reaction at these concentrations:
formula
where . The equilibrium potentials depend on the activities of the species in Equations ([11] through [14]), which are dependent on the concentration of H2SO4. The calculated equilibrium redox or oxidation reduction potential (ORP) for H3O+ reduction is plotted in Figure 6 as a function of wt% of input H2SO4. In addition, the calculated ORP assuming various sulfur redox species in combination with S(6) species (H2SO4) are also shown. The PH2 = 1 atm was assumed in all of the calculations. The measured potentials on platinized Pt are in reasonable agreement with the values calculated using only H3O+ as the redox species (Figure 6), up to about 80 wt% H2SO4. At higher concentrations, the ORP predicted using the S(6) to S(5) (S2O62–) redox equilibrium matched the experimental data best. The S(6) to S(3) (S2O32–) ORP was in agreement at 99% acid, but not in more dilute acids. In more dilute acids, the assumption of sulfur redox equilibrium resulted in a significantly higher ORP than measured on platinized Pt. It should be noted further that the equilibrium concentrations of dissolved S(2) to S(6) species are extremely small (less than 10−14 mol%). Arvia and Carozza33  measured the gases and dissolved species from cathodic polarization of Pt in H2SO4 and found that hydrogen was the reaction product at 100% yield from 40 wt% to 100 wt% H2SO4, whereas SO2 was found to be the main reaction product in the oleum range and the H2 yield dropped considerably. Thus, sulfuric acid can be regarded as a nonoxidizing (reducing) medium across a broad concentration range, being governed by proton reduction reaction. The role of S2O62– in redox equilibrium is not clear but is likely not important for most of the concentrations of H2SO4.
FIGURE 6.

Calculated and measured redox potential as a function of sulfuric acid wt%. The thermodynamic calculations assume various sulfur redox species.

FIGURE 6.

Calculated and measured redox potential as a function of sulfuric acid wt%. The thermodynamic calculations assume various sulfur redox species.

Corrosion Model vs. Experimental Data

Depassivation pH

Chromium-containing alloys are typically spontaneously passive at ambient conditions above the depassivation pH (pHD) that depends mostly on alloy content and microstructure. Above pHD, the corrosion behavior is determined by the passive film characteristics and the effects of local damage of the passive film. However, below pHD, the corrosion behavior is essentially active dissolution. The experimentally determined pHD for a variety of stainless steels and Ni-based alloys in chloride solutions has been reported by Okayama, et al.,34  Crolet, et al.,35  Oldfield,36  Trasatti and Mazza,37  and Sridhar and Cragnolino.38  It has also been shown that the pHD thus determined were relatively independent of temperature and chloride concentration.34,39  Trasatti and Maza37  showed that the pHD decreases with an increase in the pitting resistance equivalent number (PREN = Cr+3.3Mo+17N). Okayama, et al.,34  suggest a more complex relationship to alloying elements. The experimental pHD for Alloys UNS N08825,(2) N06625, and N10276 are approximately 0.1 to 0.4, 0 to −0.6, and −0.7, respectively.

The modeling of the depassivation behavior in H2SO4 without chloride is illustrated for polarization behavior of Alloy N08825 in sulfuric acids of various concentrations in Figure 7. The temperature of all solutions was assumed to be 25°C. To accentuate the difference in corrosion potentials as a function of pH, the solutions were assumed to be exposed to air (0.79 m N2 + 0.21 m O2 at 1 atm). At H2SO4 concentrations below about 0.5 m (corresponding to a pH of 0.37), the corrosion potential of the alloy is in the passive region. At lower pH values, the corrosion potential shifts slowly to the active-passive region, and finally ending in the active region above about 2 m H2SO4 (or below a pH of −0.35). This pH is more negative than the measured values in chloride solutions and reflects the effect of chloride concentration. Indeed, the model predicts that if 0.5 m NaCl is added to the H2SO4, the pHD shifts to +0.035. The pHD of four alloys, including S31600, in H2SO4 solutions containing 0.5 m NaCl at 25°C is shown in Figure 8. The oscillations in the open-circuit potential (OCP) for Alloys N06625, N10276, and N06022 are because the oxygen reduction curves intersect the active passive currents at multiple locations in these solutions. The lowest pH values at which the OCP drops rapidly is consistent with experimentally measured pHD values mentioned earlier. Below the pHD, the OCP increases slightly because the effect of decreasing pH on the proton reduction reaction potential. From this discussion, it may be concluded that for sulfuric acid concentrations greater than about 10 molal, even the highly alloyed Ni-based alloys exhibit an active dissolution behavior. In the case of HCl, the calculated pHD for N06022 occurs between 0.8 m and 2.2 m HCl where significant oscillations in predicted OCP occurs. This is consistent with the experimental data presented by Bocher, et al.40 

FIGURE 7.

Calculated polarization curves and corrosion potentials of Alloy N08825 as a function of H2SO4 concentration at 25°C. Calculations assumed an aerated solution to clearly indicate corrosion potential decrease.

FIGURE 7.

Calculated polarization curves and corrosion potentials of Alloy N08825 as a function of H2SO4 concentration at 25°C. Calculations assumed an aerated solution to clearly indicate corrosion potential decrease.

FIGURE 8.

Calculated OCP as a function of pH values of H2SO4 solutions containing 0.5 m NaCl at 25°C and purged with 1 m O2. The measured depassivation pH are indicated.

FIGURE 8.

Calculated OCP as a function of pH values of H2SO4 solutions containing 0.5 m NaCl at 25°C and purged with 1 m O2. The measured depassivation pH are indicated.

H2SO4 Solutions

In sulfuric acid solutions without any oxidizing species, the corrosion potential of stainless steels and Ni-based alloys lies in the active region of the polarization curve and the corrosion rate is determined mostly by an active dissolution process.20  This can be true even if the acid is exposed to static ambient air as the proton reduction reaction is the major contributor to the cathodic curve. The calculated polarization curve for alloy UNS N06625 is compared in Figure 9 to the experimental polarization curves from Crum and Adkins41  in 30 wt% H2SO4 solution that was exposed to still air at 49°C, and for alloy UNS N06022, the calculated curve is compared to experimental curve from Mishra42  in a deaerated 30 wt% H2SO4 at room temperature (assumed to be 25°C). The experimental polarization curves for N0662541  were determined by a rapid scan technique at a scan rate of 13.9 mV/s and a higher anodic current density is to be expected. The polarization curve for N0602242  was generated using a scan rate of 0.167 mV/s. Although the calculated anodic curves for these alloys are lower in current density and potential than the experimental measurements, both indicate that the corrosion behavior of these alloys is in the active region. The lower calculated current density is related to the assumption of steady-state behavior in the model.

FIGURE 9.

Comparison of predicted and experimental polarization curves on Alloys N06625 and N06022. The experimental curves for N06625 were replotted from Crum and Adkins41  and for N06022 from Mishra.42 

FIGURE 9.

Comparison of predicted and experimental polarization curves on Alloys N06625 and N06022. The experimental curves for N06625 were replotted from Crum and Adkins41  and for N06022 from Mishra.42 

The calculated corrosion rates for Alloy N08825 in boiling sulfuric acid solutions (dashed line) are compared to experimental data (symbols) from several sources43-45  in Figure 10. The agreement with experimental data is relatively good below about 70 wt% H2SO4. However, significant deviations between the experimental data and model are observed beyond about 70 wt% H2SO4, i.e., beyond the practical applicability range of the aqueous corrosion model. In the high concentration region, the active-passive peak current densities increase and the corrosion potential shifts to higher values.45  Alloy N08825 and some stainless steels show an oscillating corrosion potential in sulfuric acid due to the intersection of cathodic polarization curve across the anodic active-passive peak current density.46-47  This is not captured in the model. In all of these cases, the alloy does not exhibit a truly passive behavior—the current densities in the passive region are quite high, suggesting a nonprotective pseudo passive film.

FIGURE 10.

Comparison of predicted and experimental corrosion rates for Alloy N08825 in boiling sulfuric acid solutions.43-45 

FIGURE 10.

Comparison of predicted and experimental corrosion rates for Alloy N08825 in boiling sulfuric acid solutions.43-45 

The predicted corrosion rates in sulfuric acid for Alloy N06022 as a function of H2SO4 concentration and temperature in comparison to experimental data reported in the literature2-3,48-52  are shown in Figure 11. Model calculations were performed at the temperature ranges represented by the literature data and shown as color-coded lines with similar colors representing the data. It must be noted that some of the experimental data extracted from the literature are from iso-corrosion diagrams and refer to maximum corrosion rate at a given acid concentration. The corrosion rate increases with H2SO4 concentration and reaches a maximum at intermediate concentration values. Because the model is valid only up to about 30 molal, experimental data are limited to this concentration. There is general agreement with the experimental data. A comparison of model results with experimental data in dilute, nitrogen-deaerated, H2SO453  from 1 wt% to 5 wt% at temperatures from 107°C to 204°C is shown in Figure 12. The R2 value for the linear correlation is 0.5, but the model predicts the right order of magnitude of corrosion rate.

FIGURE 11.

Comparison of model to experimental data for N06022 in H2SO4 solutions at different temperatures.2-3,48-50 

FIGURE 11.

Comparison of model to experimental data for N06022 in H2SO4 solutions at different temperatures.2-3,48-50 

FIGURE 12.

Predicted vs. measured corrosion rates of N06022 at 107°C to 204°C in dilute H2SO4 from 1 wt% to 5 wt%.53 

FIGURE 12.

Predicted vs. measured corrosion rates of N06022 at 107°C to 204°C in dilute H2SO4 from 1 wt% to 5 wt%.53 

Table 2.

Measured Corrosion Rates of Various Alloys in H2SO4 + HCl Mixtures(A)

Measured Corrosion Rates of Various Alloys in H2SO4 + HCl Mixtures(A)
Measured Corrosion Rates of Various Alloys in H2SO4 + HCl Mixtures(A)

H2SO4 + HCl Mixtures

The corrosion rates of various alloys in H2SO4 + HCl mixtures were reported previously.3  These data are reproduced in Table 2 for 79.4°C. A comparison of calculated corrosion rates for N06022 with experimental weight-loss measurements (Figure 13) for various H2SO4 + HCl mixtures indicates that the model predicts the observed trends in the corrosion behavior (R2 = 0.55). The measured corrosion rates for S30600 and N06022 show a dual slope with calculated pH of these solutions (bottom of Figure 14). In the pH range from about 1 to 0, there is steep increase in corrosion rate with a decrease in pH, whereas at lower pH values the slope is much shallower (Figure 14[b]). Similar behavior is observed for the other alloys. A plot of free chloride and undissociated HCl0 in these mixtures as a function of pH (Figure 14[a]) shows a similar behavior with pH as the corrosion rate. However, the maximum in the free chloride concentration occurs at a higher pH than that of the HCl0 and corresponds to the maximum in corrosion rate. This suggests that the adsorption of free Cl is important for the corrosion process.

FIGURE 13.

Comparison of measured and calculated corrosion rates for Alloy N06022 in various H2SO4 + HCl mixtures at 79.1°C (see Table 2).

FIGURE 13.

Comparison of measured and calculated corrosion rates for Alloy N06022 in various H2SO4 + HCl mixtures at 79.1°C (see Table 2).

FIGURE 14.

Corrosion rate of two alloys as a function of pH in H2SO4 + HCl mixtures at 79.4°C. Also shown are the corresponding free chloride and undissociated HCl concentrations. The lines are not fit curves.

FIGURE 14.

Corrosion rate of two alloys as a function of pH in H2SO4 + HCl mixtures at 79.4°C. Also shown are the corresponding free chloride and undissociated HCl concentrations. The lines are not fit curves.

An alloy equivalent number defined as (Ni0.7-Cr0.1+Mo+0.5W) was developed through a trial-and-error method to represent the corrosion behavior of various alloys in these mixtures (Figure 15). It should be noted that the corrosion rate did not correlate with the traditional pitting resistance equivalent (PREN = Cr+3.3(Mo+0.5W)+16N) because Fe-based alloys with the same PRE as Ni-based alloys corroded at a considerably higher rate. However, this alloy equivalent number is not based on a rigorous regression analysis of the data. The exponent on Ni indicates the importance of Ni in imparting corrosion resistance in these environments and conversely, the detrimental effect of Fe. The negative coefficient for Cr indicates that in nonoxidizing environments, Cr may be slightly detrimental. This is consistent with the findings by Rebak and Srivastava54  in nonoxidizing acids, although they examined only three Ni-based alloys with a more limited composition variation. The most corrosion resistant alloy in this environment was the Ni-Mo alloy. The role of Ni is also consistent with other published data of these alloys in sulfuric acid environments (see Davies,2  figure 9.5).

FIGURE 15.

Effect of alloying elements on the corrosion rate of Ni-based alloys in selected H2SO4+HCl mixtures at two temperatures.

FIGURE 15.

Effect of alloying elements on the corrosion rate of Ni-based alloys in selected H2SO4+HCl mixtures at two temperatures.

Table 3.

Weight-Loss Corrosion Rates of Alloys in H2SO4 + HF Mixtures at 79.4°C

Weight-Loss Corrosion Rates of Alloys in H2SO4 + HF Mixtures at 79.4°C
Weight-Loss Corrosion Rates of Alloys in H2SO4 + HF Mixtures at 79.4°C

H2SO4 + HF Mixtures

It has been shown previously55  that HF accelerates corrosion in low to moderate concentrations of H2SO4, but inhibits corrosion at concentrations higher than about 60 wt% H2SO4. The speciation calculations suggest that at the higher H2SO4 concentrations, the fluosulfonate ion and fluosulfonic acid concentrations increase (Figure 4). The measured corrosion rates of several alloys is plotted as a function of the ratio of fluosulfonic species to undissociated HF in Figure 16. It can be seen that the corrosion rates of alloys, especially the ferrous alloys, decrease considerably as the fluosulfonic species concentrations increase. It is possible that the adsorption of fluosulfonic species on the stainless steel surface is not as strong as the HF species. Interestingly, for the polishing of Nb in a H2SO4 + HF solution, Ford, et al.,28  observed a reduction in the peak intensity of HSO3F, which they hypothesized as due to the consumption of F (or HF0) during the anodic polarization and the resulting equilibrium with HF and HSO3F driving the concentration of the latter down (F and HF are not Raman active). This would again suggest that HF0 adsorption determines the corrosion rate and the increase in the concentration of fluosulfonic species reduces the HF0 adsorption and, therefore, the corrosion rate.

FIGURE 16.

Effect of fluoride speciation in sulfuric acid on measured corrosion rates. All tests conducted at 79°C for 24 h.

FIGURE 16.

Effect of fluoride speciation in sulfuric acid on measured corrosion rates. All tests conducted at 79°C for 24 h.

The predicted corrosion rates for Alloy N06022 agree reasonably well with measured corrosion rates (Figure 17). However, the agreement between model and experiment is much poorer than for H2SO4 + HCl mixtures (R2 = 0.18) and also poor for the other alloys. As in the case of HCl mixtures, the measured corrosion rates correlate with (Ni0.7-Cr0.1+Mo+0.5W), in similarity to H2SO4 + HCl mixtures (Figure 18). The importance of Ni in imparting corrosion resistance in HF environments is consistent with experience that shows even Ni-Cu alloys show corrosion resistance in these environments.

FIGURE 17.

Calculated vs. measured corrosion rates for Alloy N06022 in H2SO4 + HF mixtures at 79.4°C (see Table 3).

FIGURE 17.

Calculated vs. measured corrosion rates for Alloy N06022 in H2SO4 + HF mixtures at 79.4°C (see Table 3).

FIGURE 18.

Effect of alloying elements on the corrosion rates of stainless steels and Ni-based alloys in H2SO4 + HF mixtures at 79.4°C.

FIGURE 18.

Effect of alloying elements on the corrosion rates of stainless steels and Ni-based alloys in H2SO4 + HF mixtures at 79.4°C.

H2SO4 + HCl + HF Mixtures

A good test of the model is its ability to predict the corrosion rate of alloys in a multi-component mixture that has not been used in parameterizing the model. This is shown in Table 4, where the predicted corrosion rate is compared to weight-loss data in a mixture of 10 wt% H2SO4 + 10 wt% HCl + 10 wt% HF at 80°C. The agreement is reasonable for the two alloys. Furthermore, the trend in corrosion rate with alloying is also consistent with expectation from Figures 15 and 18. Alloy N10276 has an alloy equivalent (Ni0.7-Cr0.1+Mo+0.5W) of 32.87, whereas Alloy N06022 has an alloy equivalent of 29.88.

Table 4.

Comparison of Model vs. Experimental Corrosion Rates for a Three Component Mixture Consisting of 10 wt% H2SO4 + 10 wt% HCl + 10 wt% HF at 80°C(A)

Comparison of Model vs. Experimental Corrosion Rates for a Three Component Mixture Consisting of 10 wt% H2SO4 + 10 wt% HCl + 10 wt% HF at 80°C(A)
Comparison of Model vs. Experimental Corrosion Rates for a Three Component Mixture Consisting of 10 wt% H2SO4 + 10 wt% HCl + 10 wt% HF at 80°C(A)

DISCUSSION

It has long been recognized that the activities of water, protons, and sulfur species are affected by sulfuric acid concentration. The effect of speciation on corrosion of stainless steels and Ni-based alloys in highly concentrated sulfuric acid has been mentioned previously.32  The solution modeling results shown in this paper provide a quantitative understanding of the chemical speciation that forms the backdrop for corrosion. Redox potentials calculated by the model and compared to experimental results are consistent with the idea that sulfuric acid redox potential is primarily determined by the proton reduction reaction up to about 80 wt% acid. Thus, it acts as a nonoxidizing medium across a broad range of concentrations. Beyond 80 wt%, other sulfur species may contribute to the redox potential, but their mole fractions are exceedingly low and, thus, they may be kinetically limited. Both the modeling and experimental results highlighted in this paper indicate that the corrosion is driven by active dissolution and not determined by passive film formation. In this scenario, the parameters that govern corrosion kinetics under open-circuit conditions are the active dissolution kinetics and cathodic reduction kinetics of protons (or hydronium ions). The electrochemical model predictions based on the solution speciation are in general agreement with weight-loss experimental data in a variety of alloys in nonoxidizing sulfuric acid mixtures. Another support for the conclusion that the corrosion behavior in sulfuric acid is governed by active dissolution is that in the Cr-containing alloys, the addition of oxidizing agents, such as ferric sulfate, shifts the corrosion potential to the pseudo passive region and results in lower corrosion rates.20  Under these conditions, Cr is a beneficial alloying element. However, the addition of such oxidizing agents to sulfuric acid increases the corrosion rate of Ni-Mo alloys that do not show any passive behavior.20 

The predicted effect of alloying elements on corrosion rate in nonoxidizing acids from the model is shown in Figure 19. A wide range of stainless steels and Ni-based alloys are included in the calculations. The environment is considered to be deaerated by 1 atm N2 and is static. The predicted trends with alloying elements are consistent with those shown for experimental data in Figures 15 and 18. However, considerably more scatter is observed with modeled results, suggesting that improved parameterization of the model is needed through a combination of additional experimental data and theoretical modeling.

FIGURE 19.

Relationship between alloying and corrosion rates predicted by the model for two solutions. The symbols are model predictions and the lines are logarithmic fit.

FIGURE 19.

Relationship between alloying and corrosion rates predicted by the model for two solutions. The symbols are model predictions and the lines are logarithmic fit.

Alloy designers have long known that in nonoxidizing acids, the formation of an oxide film is not important.43,56-57  The alloying elements can reduce corrosion rate by reducing the active anodic or cathodic kinetics, or both. The research to date on the effect of Ni does not provide a clear delineation of the rate controlling processes involved in the beneficial effect of Ni. For example, Ni in Fe-Ni binary alloys has been shown to decrease the corrosion rates and corrosion current densities as a function of potential in sulfuric acid solutions.5,58  The corrosion behavior in these alloys is determined by active dissolution, although adsorption of a reaction intermediate, Ni(OH)2, is postulated.58  The passive current densities of Ni and Fe-Ni alloys are high (of the order of 1 mA/cm2 or higher) and may be considered porous or pseudo passive films. At high dissolution rates, the anodic curves of Ni exhibit a diffusion-limited current density governed by the transport of NiSO4.30  Condit8  showed that the increase in Ni in binary Ni-Fe alloys increased the exchange current density and decreased the cathodic Tafel slope for proton reduction in a 1 N H2SO4 solution that was purged with 1 atm H2. The Ecorr increased with an increase in Ni content, while the anodic Tafel slope did not vary significantly. The peak anodic current density decreased dramatically above about 50 at% Ni. Sayano and Nobe,59  on the other hand, found that both the anodic and cathodic current densities in 1 N H2SO4 decreased significantly with the addition of Ni to Fe in binary alloys, while the Tafel slopes were not affected significantly. They also showed that the Ecorr increased with an increase in Ni content. It is possible that the deactivation of the electrode prior to testing by Sayano and Nobe contributed to the contradictory finding of the cathodic polarization. However, all of the studies of Fe-Ni binary alloys in H2SO45,8,58-59  agree that increasing Ni in Fe-Ni alloys reduces corrosion rate and increases the Ecorr, suggesting that Ni inhibits anodic dissolution rates much more than it affects cathodic reduction kinetics.

The addition of Cu to stainless steels and Ni-based alloys has been known to reduce corrosion rates in low and moderate concentrations of H2SO420,60-61  and commercial alloys have been developed with Cu as a small alloying addition (2% to 4%) to Ni-Fe-Cr-Mo alloys. However, in duplex stainless steels the effect does not appear to be significant, if the alloy is previously activated by exposure to HCl or contacted with Zn.60  Kim and Park61  found an enrichment of Cu on duplex stainless steels exposed to sulfuric acid, which resulted in a decrease in cathodic current density, critical current density for active-passive transition, and the pseudo passive current density. Other studies have indicated that Cu increases the passive dissolution rate in H2SO4, but decreases the active dissolution rate and Cu enrichment of the passive film was found.62-63  In the current model, the effect of Cu is not included because there is insufficient data to parameterize the Cu effect in the model.

It is well known that Cr is beneficial to corrosion in oxidizing environments through the promotion of passivity. However, in nonoxidizing conditions, the addition of Cr to Ni-based alloys and Fe-Cr binary has been shown to be detrimental to corrosion in a variety of sulfuric acid solutions.32,54,64  In binary Fe-Cr alloys containing 12 wt% to 35 wt% Cr, Yau and Streicher64  found that Cr increases corrosion rate in hydrogen-purged, 1 N H2SO4 at temperatures ranging from 30°C to 101°C. The addition of Cr increased the cathodic current density significantly, although their extrapolated exchange current density did not show any trend with Cr content. They concluded that Cr increased the anodic current density in these alloys. In the case of Ni-Cr binary alloys, the effect of Cr on increasing the active anodic dissolution rate in 1 N H2SO4 is only observed when the Cr content is above approximately 90 wt%.65-66  In the high concentration H2SO4, Cr is somewhat beneficial, although the combination of Cr and Si appears to be essential for corrosion resistance.20 

It is well known that Mo and W improve the localized corrosion resistance in oxidizing halide environments, as is evident by the PREN index and its variants. The role of Mo has been shown to be related to improved passive film formation,68-69  but may also be related to alteration of localized chemistry during stable growth.70  Although native oxide films cannot be ruled out in nonoxidizing environments, the dissolution in these environments at OCPs appear to proceed in an active manner. Furthermore, no localized corrosion is observed. In binary Ni-Mo alloys with up to 15 wt% Mo, the anodic portion of the polarization curves in pH 2.8 Na2SO4 solution shows a variable effect, but generally decreasing current density at a given potential in the active regime.9  Brooks, et al., in their classic review of Ni-Mo alloys,71  showed that the commercial Ni-29%Mo alloy showed a lower anodic dissolution rate than Ni in 1 N H2SO4 and 1 N HCl. However, they assumed that Mo did not affect the cathodic current density. Studies of hydrogen evolution exchange current densities for transition metals show a periodicity with atomic number72  and a characteristic volcano shape when plotted against metal-hydrogen adsorption energies.73-74  Mo has a much lower exchange current density for hydrogen evolution than Ni, Cr, and Fe. This suggests that the beneficial effect of Mo on corrosion behavior in nonoxidizing acids may also arise from its inhibition of hydrogen evolution reaction.

There has recently been an extensive effort to predict corrosion behavior of complex alloys from fundamental principles.75-78  The focus of these papers is on passive film formation and identifying regions of passivity through a combination of thermodynamic and density functional theory (DFT) models. The competitive formation of oxides versus chloride is computed through DFT calculations as an approach to predicting the localized corrosion resistance.79-80  These investigations provide a path for the alloy designers to develop novel alloys through a coupling of fundamental modeling with experimental data. In nonoxidizing acids, passive film formation is not important. This suggests that atomistic modeling in these environments should focus on the anodic dissolution process in the active region of the polarization behavior and on the proton reduction kinetics on alloy surfaces.

In addition to the effect of alloying elements, the effect of electrolyte speciation must be considered. In industrial systems, the environmental chemistry is complex, requiring the consideration of the activity of species, including water. In this interpretive paper, we focus on electrolyte speciation effects on corrosion. The corrosion model does not consider atomistic details and the model parameters are regressed using experimental data in selected sub-sets of environments and alloys. Thus, the electrolyte-based model is useful as a method to extrapolate corrosion data from a small sub-set of environmental chemistry to a larger set of environments. However, the parameters in the model can be further developed using more fundamental atomistic calculations, thus providing a possible link to future multiscale modeling. The effect of oxidizing species on corrosion is included in the current model and will be reported in a companion paper. The electrochemical model is, at present, limited to aqueous solution with an approximate solute limit of 30 mol/kg H2O. The extension of the electrochemical model to higher concentrations, including nonaqueous systems, using the MSE chemistry model is being pursued and will be the subject of a future paper.

CONCLUSIONS

This paper provides an interpretive modeling of literature data pertaining to nonoxidizing, acidic solutions involving sulfuric acid.

  • The mixed solvent electrolyte (MSE) model shows that the high redox potential in concentrated sulfuric acid beyond 68 wt% acid is not due to the presence of an oxidizing species, but the result of low activity of water and hydronium ions. The hydronium reduction reaction controls the redox potential up to about 80 wt% H2SO4. Beyond this concentration, other sulfur species may participate in determining the redox potential, but their concentrations are exceedingly low, thus limiting their kinetics. Sulfuric acid redox potential is consistent with it being a nonoxidizing environment across essentially the whole concentration range.

  • A general electrochemical model has been developed to predict the corrosion rates as a function of environment composition and temperature. The model represents the anodic dissolution of the alloy in the active and passive states, partial cathodic processes associated with the reduction of water, protons and oxidizing species, and the effect of solution species on passive dissolution and active-passive transition. The model reproduces the observed corrosion rates in environments with a total solute concentration up to 30 molal and helps interpret the corrosion behavior using speciation information and simulated polarization curves. The model predictions were in reasonable agreement with data in a three component mixture of H2SO4 + HCl + HF.

  • In concentrated H2SO4 + HCl and HF acid mixtures, the corrosion may be controlled by the free halide species. Thus, increasing the concentration of H2SO4 in these mixtures reduces the free halide species and the corrosion rate.

  • Based on the model and the experimental data, approximate relationships have been established between the corrosion rates and the Ni, Cr, Mo, and W content of the alloys. The corrosion rates correlate with the (Ni0.7-Cr0.1 + Mo + 0.5W) content at fixed environment conditions in nonoxidizing acids.

(1)

OLI Systems, Inc., www.olisystems.com.

(2)

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.

References

1.
Tyagi
A.K.
, “
A World of Sulfur
,”
in
World Fertilizer Magazine
,
December 24, 2018
, https://www.worldfertilizer.com/special-reports/24122018/a-world-of-sulfur/.
2.
Davies
M.
,
Materials Selection for Sulfuric Acid
(
St. Louis, MO
:
Materials Technology Institute of the Chemical Process Industries, Inc
,
2005
).
3.
Sridhar
N.
,
Wu
J.B.C.
,
Corey
S.M.
,
Mater. Perform.
24
,
10
(
1987
):
p
.
17
.
4.
Sridhar
N.
,
Wu
J.B.C.
,
Manning
P.E.
,
J. Met.
37
,
11
(
1985
):
p
.
51
53
.
5.
Economy
G.
,
Speiser
R.
,
Beck
F.H.
,
Fontana
M.G.
,
J. Electrochem. Soc.
108
,
4
(
1961
):
p
.
337
.
6.
Sato
N.
,
Okamoto
G.
,
J. Electrochem. Soc.
110
6
(
1963
):
p
.
605
.
7.
Trabanelli
G.
,
Zucchi
F.
,
Felloni
L.
,
Corros. Sci.
5
(
1965
):
p
.
211
224
.
8.
Condit
D.O.
,
Corros. Sci.
12
,
5
(
1972
):
p
.
451
462
.
9.
Mitrovic-Scepanovic
V.
,
Ives
M.
,
J. Electrochem. Soc.
127
,
9
(
1980
):
p
.
1903
.
10.
Misra
R.D.K.
,
Burstein
G.T.
,
Corros. Sci.
24
,
4
(
1984
):
p
.
305
323
.
11.
Danielson
M.J.
,
Baer
D.R.
,
Corros. Sci.
29
,
11
12
(
1989
):
p
.
1265
1274
.
12.
Marcus
P.
,
Moscatelli
M.
,
J. Electrochem. Soc.
136
,
6
(
1989
):
p
.
1634
1637
.
13.
Cortes
R.
,
Froment
M.
,
Hugot-Le Goff
A.
,
Joiret
S.
,
Corros. Sci.
31
(
1990
):
p
.
121
127
.
14.
Gopaliengar
R.
,
Oriani
R.A.
,
Corros. Sci.
30
,
11
(
1990
):
p
.
1101
1119
.
15.
Itagaki
M.
,
Nakazawa
H.
,
Watanabe
K.
,
Noda
K.
,
Corros. Sci.
39
,
5
(
1997
):
p
.
901
911
.
16.
Abdulsalam
M.I.
,
Pickering
H.W.
,
Corros. Sci.
41
,
2
(
1998
):
p
.
351
372
.
17.
Gregori
J.
,
García-Jareño
J.J.
,
Giménez-Romero
D.
,
Vicente
F.
,
J. Electrochem. Soc.
153
,
6
(
2006
):
p
.
B206
.
18.
Iida
M.
,
Ohtsuka
T.
,
Corros. Sci.
49
3
(
2007
):
p
.
1408
1419
.
19.
Trompette
J.L.
,
Massot
L.
,
Vergnes
H.
,
Corros. Sci.
74
(
2013
):
p
.
187
193
.
20.
Sridhar
N.
,
Mater. Perform.
25
,
3
(
1988
):
p
.
40
46
.
21.
Wang
P.
,
Anderko
A.
,
Young
R.D.
,
Fluid Phase Equilibria
203
,
1
(
2002
):
p
.
141
176
.
22.
Wang
P.
,
Anderko
A.
,
Springer
R.D.
,
Young
R.D.
,
J. Mol. Liq.
125
,
1
(
2006
):
p
.
37
44
.
23.
Anderko
A.
,
Young
R.
,
Corrosion
56
,
5
(
2000
):
p
.
543
555
.
24.
Anderko
A.
,
McKenzie
P.
,
Young
R.D.
,
Corrosion
57
,
3
(
2001
):
p
.
202
213
.
25.
Anderko
A.
, “
Modeling of Aqueous Corrosion
,”
in
Shreir’s Corrosion
,
ed.
Richardson
T.J.A.
(
Amsterdam, the Netherlands
:
Elsevier Oxford
,
2010
),
p
.
1585
1629
.
26.
Sridhar
N.
,
Brossia
C.S.
,
Dunn
D.S.
,
Anderko
A.
,
Corrosion
60
,
10
(
2004
):
p
.
915
936
.
27.
Walrafen
G.E.
,
Yang
W.H.
,
Chu
Y.C.
,
Hokmabadi
M.S.
,
J. Solut. Chem.
29
,
10
(
2000
):
p
.
905
936
.
28.
Ford
D.C.
,
Cooper
C.
,
Cooley
L.D.
,
Thompson
C.
,
Bouchard
D.
,
Albee
B.
,
Bishnoi
S.
,
J. Electrochem. Soc.
160
,
6
(
2013
):
p
.
H398
H403
.
29.
Kish
J.
,
Ives
M.
,
Can. J. Chem. Eng.
89
,
5
(
2011
):
p
.
1311
1319
.
30.
Kish
J.
,
Ives
M.
,
Rodda
J.
,
J. Electrochem. Soc.
147
,
10
(
2000
):
p
.
3637
.
31.
Stypula
B.
,
Banas
J.
,
Electrochim. Acta
38
,
15
(
1993
):
p
.
2309
2314
.
32.
Renner
M.
,
Mater. Corros.
47
(
1996
):
p
.
246
260
.
33.
Arvia
A.J.
,
Carrozza
J.S.W.
,
Electrochim. Acta
11
,
11
(
1966
):
p
.
1641
1649
.
34.
Okayama
S.
,
Tsujikawa
S.
,
Kikuchi
K.
,
Corros. Eng.
36
(
1987
):
p
.
631
638
.
35.
Crolet
J.
,
Defranoux
J.
,
Seraphin
L.
,
Tricot
R.
,
Mem. Sci. Rev. Met.
71
,
12
(
1974
):
p
.
797
805
.
36.
Oldfield
J.W.
,
Int. Mater. Rev.
32
,
1
(
1987
):
p
.
153
172
.
37.
Trasatti
S.P.
,
Mazza
F.
,
Brit. Corros. J.
30
,
4
(
1995
):
p
.
275
282
.
38.
Sridhar
N.
,
Cragnolino
G.A.
,
Corrosion
49
,
11
(
1993
):
p
.
885
894
.
39.
Crolet
J.-L.
,
Seraphin
L.
,
Tricot
R.
,
Memoires Sci. Rev. Metall.
73
(
1976
):
p
.
1
8
.
40.
Bocher
F.
,
Huang
R.
,
Scully
J.R.
,
Corrosion
66
,
5
(
2010
):
p
.
055002-01
to
055002-15
.
41.
Crum
J.R.
,
Adkins
M.E.
,
Mater. Perform.
25
,
2
(
1986
):
p
.
27
32
.
42.
Mishra
A.
, “
Performance of Corrosion-Resistant Alloys in Individual and Mixed Acids
,”
CORROSION 2019, paper no. 9409
(
Houston, TX
:
NACE International
,
2019
).
43.
Friend
W.Z.
,
Corrosion of Nickel and Nickel-Base Alloys
(
New York, NY
:
John Wiley & Sons
,
1980
),
p
.
459
.
44.
Brill
U.
,
Alves
H.
, “
Alloy Selection for Dilute and Medium Concentration Sulfuric Acid
,”
CORROSION 2004, paper no. 04233
(
Houston, TX
:
NACE
,
2004
).
45.
Sivieri
E.
,
Mazza
F.
,
Cristina Reggiani
R.
,
Int. J. Hydrogen Energy
8
,
11
(
1983
):
p
.
913
921
.
46.
Morris
P.E.
,
Scarberry
R.C.
,
Corrosion
26
,
7
(
2013
):
p
.
269
279
.
47.
Jones
S.
,
Li
Y.
,
Coley
K.S.
,
Kish
J.R.
,
Ives
M.B.
,
Corros. Sci.
52
,
1
(
2010
):
p
.
250
254
.
48.
Craig
B.D.
,
Anderson
D.S.
,
Handbook of Corrosion Data
(
Materials Park, OH: ASM International
,
1994
).
49.
Kreysa
G.
,
Schütze
M.
,
DECHEMA Corrosion Handbook - Revised and Extended
, 2nd Edition (
Frankfurt am Main, Germany: DECHEMA
,
2008
).
50.
Manning
P.
,
Schobel
J.-D.
,
Werkst. Korros.
37
,
3
(
1986
):
p
.
137
145
.
51.
Crook
P.
,
Mater. Corros.
56
,
9
(
2005
):
p
.
606
610
.
52.
Agarwal
D.C.
,
Kloewer
J.
, “
Nickel Base Alloys: Corrosion Challenges in the New Millennium
,”
CORROSION 2001
(
Houston, TX
:
NACE
,
2001
).
53.
Deodeshmukh
V.P.
,
Manning
P.E.
,
Crook
P.
, “
Corrosion Performance of Nickel Alloys in Dilute Hydrochloric and Sulfuric Acids Near and Above Boiling Point
,”
CORROSION 2013, paper no. 2194
(
Houston, TX
:
NACE
,
2013
).
54.
Rebak
R.B.
,
Srivastava
S.K.
,
Corrosion
55
,
4
(
1999
):
p
.
412
421
.
55.
Sridhar
N.
, “
The Effect of Velocity on Corrosion in H2SO4 + HF Mixtures
,”
CORROSION 1990
(
Houston, TX
:
NACE
,
1990
).
56.
Hodge
F.G.
,
Kirchner
R.W.
, “
An Improved Ni-Mo Alloy for Hydrochloric Acid Service
,”
CORROSION 1975, paper 60
(
Houston, TX
:
NACE
,
1975
).
57.
Hodge
F.G.
,
JOM
58
(
2006
):
p
.
28
31
.
58.
Altura
D.
,
Nobe
K.
,
Corrosion
32
,
2
(
2013
):
p
.
41
47
.
59.
Sayano
R.R.
,
Nobe
K.
,
Corrosion
25
,
6
(
2013
):
p
.
260
266
.
60.
Kivisäkk
U.H.
,
Corrosion
61
,
6
(
2005
):
p
.
602
608
.
61.
Kim
S.-T.
,
Park
Y.-S.
,
Corrosion
63
,
2
(
2007
):
p
.
114
126
.
62.
Lin
H.T.
,
Tsai
W.T.
,
Lee
J.T.
,
Huang
C.S.
,
Corros. Sci.
33
,
5
(
1992
):
p
.
691
697
.
63.
Jiangnan
Y.
,
Lichang
W.
,
Wenhao
S.
,
Corros. Sci.
33
,
6
(
1992
):
p
.
851
859
.
64.
Yau
Y.-H.
,
Streicher
M.A.
,
Corrosion
47
,
5
(
1991
):
p
.
352
359
.
65.
Hodge
F.G.
,
Wilde
B.E.
,
Corrosion
26
,
6
(
2013
):
p
.
246
250
.
66.
Myers
J.R.
,
Beck
F.H.
,
Fontana
M.G.
,
Corrosion
21
,
9
(
2013
):
p
.
277
287
.
67.
Gerard
A.Y.
,
Lutton
K.
,
Lucente
A.
,
Frankel
G.S.
,
Scully
J.R.
,
Corrosion
,
76
,
5
(
2020
):
p
.
485
499
.
68.
Cwalina
K.L.
,
Ha
H.M.
,
Ott
N.
,
Reinke
P.
,
Birbilis
N.
,
Scully
J.R.
,
J. Electrochem. Soc.
166
,
11
(
2019
):
p
.
C3241
C3253
.
69.
Wang
Z.
,
Carrière
C.
,
Seyeux
A.
,
Zanna
S.
,
Mercier
D.
,
Marcus
P.
,
J. Electrochem. Soc.
168
,
4
(
2021
):
article 041503
.
70.
Jakupi
P.
,
Noël
J.
,
Shoesmith
D.
,
Corros. Sci.
54
(
2012
):
p
.
260
269
.
71.
Brooks
C.R.
,
Spruiell
J.E.
,
Stansbury
E.E.
,
Int. Met. Rev.
29
(
1984
):
p
.
210
248
.
72.
Kita
H.
,
J. Electrochem. Soc.
113
,
11
(
1966
):
p
.
1095
.
73.
Jaksic
J.M.
,
Ristic
N.M.
,
Krstajic
N.V.
,
Jaksic
M.M.
,
Int. J. Hydrogen Energy
23
,
12
(
1998
):
p
.
1121
1156
.
74.
Nørskov
J.K.
,
Bligaard
T.
,
Logadottir
A.
,
Kitchin
J.R.
,
Chen
J.G.
,
Pandelov
S.
,
Stimming
U.
,
J. Electrochem. Soc.
152
,
3
(
2005
):
p
.
J23
.
75.
Taylor
C.D.
,
Lu
P.
,
Saal
J.
,
Frankel
G.S.
,
Scully
J.R.
,
npj Mater. Degrad.
2
,
1
(
2018
):
article 6
.
76.
Scully
J.R.
,
Inman
S.B.
,
Gerard
A.Y.
,
Taylor
C.D.
,
Windl
W.
,
Schreiber
D.K.
,
Lu
P.
,
Saal
J.E.
,
Frankel
G.S.
,
Scr. Mater.
188
(
2020
):
p
.
96
101
.
77.
Lu
P.
,
Saal
J.E.
,
Olson
G.B.
,
Li
T.
,
Swanson
O.J.
,
Frankel
G.S.
,
Gerard
A.Y.
,
Quiambao
K.F.
,
Scully
J.R.
,
Scr. Mater.
153
(
2018
):
p
.
19
22
.
78.
Xie
Y.
,
Artymowicz
D.M.
,
Lopes
P.P.
,
Aiello
A.
,
Wang
D.
,
Hart
J.L.
,
Anber
E.
,
Taheri
M.L.
,
Zhuang
H.
,
Newman
R.C.
,
Sieradzki
K.
,
Nat. Mater.
20
(
2021
):
p
.
789
793
.
79.
Ke
H.
,
Frankel
G.S.
,
Taylor
C.D.
,
J. Electrochem. Soc.
167
,
13
(
2020
):
article 131510
.
80.
Taylor
C.D.
,
Li
S.
,
Samin
A.J.
,
Electrochim. Acta
269
(
2018
):
p
.
93
10
.