Corrosion Behavior of Carbon Steels in Sulfide-Containing Caustic

Ind. Eng. Chem. Res. 2006, 45, 7789-7794

7789

Corrosion Behavior of Carbon Steels in Sulfide-Containing Caustic Solutions Patrick E. Hazlewood School of Chemical and Biomolecular Engineering, Georgia Institute of Technology, Atlanta, Georgia 30332

Preet M. Singh* School of Materials Science and Engineering, Georgia Institute of Technology, Atlanta, Georgia 30332

Jeffery S. Hsieh School of Chemical and Biomolecular Engineering, Georgia Institute of Technology, Atlanta, Georgia 30332

The corrosion resistance of carbon steels in caustic-containing sulfide solutions depends on the formation of a stable passive film on the metal surface. As processes change chemical concentrations and temperature parameters, the corrosivity of the system also changes. New resistant materials such as duplex stainless steel are being selected for equipment in many processes, yet carbon steel continues to be utilized for a significant amount of new and existing equipment and can be subjected to conditions beyond the initial design parameters. The present study investigates the effects of temperature up to 170 °C and of the concentrations of sulfide and hydroxide on the corrosion rate to determine environmental limits for the reliable operation of carbon steel A516-Gr70. The results indicate that carbon steel A516-Gr70 can exhibit high corrosion rates as the concentration of hydroxide or sulfide increases at temperatures above 100 °C and changes the open-circuit potential to values below -1.00 V (SCE). Introduction Sulfide-containing caustic solutions are used or produced in a variety of chemical processes. These include modified Bayer processes used in alumina ore processing, the Gilder sulfide process used in heavy water production, hydrocarbon processing for the treatment of acidic impurities such as hydrogen sulfide and mercaptans, and various process streams in pulp mills utilizing the Kraft pulping process. During the Kraft pulping process, wood chips are cooked in a white liquor that contains sodium hydroxide and sodium sulfide at temperatures up to 180 °C. In this process, the wood’s lignin is fragmented into smaller segments whose sodium salts are soluble in the cooking liquor, leaving cellulose and hemicellulose in the form of intact fibers needed for papermaking. The concentrations of hydroxide and sulfide have increased over the years along with the pulping temperatures to improve product yield and quality.1,2 Although there are other process streams in which hydroxide and sulfide salts can exist with other chemicals, the present study focuses on environmental conditions encompassing and exceeding those typically found in different parts of a pulp mill.3,4 Because of the variability in white liquor concentration, with trends toward higher hydroxide concentrations, which can exceed 150 g/L sodium hydroxide, and higher levels of sulfide, which can exceed 50 g/L sodium sulfide, research in this area has not been comprehensive.5 The literature has typically focused on concentrations of 100 g/L sodium hydroxide and 30-40 g/L sodium sulfide. The dependence of white liquor corrosivity on the concentrations of hydroxide and sulfide, at approximately 100 g/L sodium hydroxide and 30 g/L sodium sulfide, was described as weak by Yeske.6 Instead, many researchers have focused on other ions that form in most pulping liquors and other industrial processes as a result of hydrolysis or oxidation reactions via * To whom correspondence should be addressed. E-mail: [email protected].

eqs 1 and 2.7 Although the reactions, specifically eq 2, have been shown to occur at very low rates when not in an oxidizing environment, they are often referred to in the literature as the mechanisms for corrosion rate control.7-9

S2- + H2O S HS- + OH-

(1)

S2- + (S2O3)2- S S22- + SO32-

(2)

Peterman and Yeske5 investigated the effect of solution oxidation on chemical concentration for a sulfide-containing caustic solution, 100 g/L sodium hydroxide and 35 g/L sodium sulfide, in 4-week tests. In all cases, the concentration of the sodium salts analyzed showed change. Yet, testing performed immediately after mixing showed the most significant change in concentration. From their data, it appears evident that the initial variation in the concentrations of the tested sodium salts was driven by oxidation of the solution during initial mixing and not by the proposed equilibrium of the salts according to eq 2, supported by the equilibrium constant, 1.6 × 10-4, reported by Haegland and Roald.7 Wensley and Charlton10 investigated the role of inorganic sodium sulfide, sulfite, thiosulfate, and sulfate in the corrosivity of pulp mill liquors, through a potentiostatic polarization study of mild steel. They showed that the sulfide and thiosulfate ions act as corrosion activators, whereas the sulfite and sulfate ions do not affect corrosion behavior, a result previously reported by Kesler.8 Haegland and Roald demonstrated the ability of polysulfide, shown in eq 2 in its primary disulfide form, to passivate mild steel when in sufficiently high concentration, above 4 g/L, and activate corrosion at lower concentrations, 1-3 g/L.6,7 The complex relationship of inorganic salts was also realized in the investigation of thiosulfate and sulfide. Thiosulfate ions at low concentration and in the presence of sulfide have little impact on corrosion rates.6,7 However, when both thiosulfate and sulfide are present in sufficient concentrations,

10.1021/ie060358p CCC: $33.50 © 2006 American Chemical Society Published on Web 10/12/2006

7790

Ind. Eng. Chem. Res., Vol. 45, No. 23, 2006

Table 1. Initial Concentrations of Solutions Containing Sodium Sulfide and Sodium Hydroxide Tested at 25, 90, 100, and 170 °C test

NaOH (g/L)

Na2S (g/L)

1 2 3 4 5 6 7 8

100 200 0 100 200 0 100 200

0 0 37.5 37.5 37.5 75 75 75

they inhibit the formation of a passive layer on the metal surface, leading to higher corrosion rates.6-14 The behavior of polysulfide and thiosulfate ions in caustic sulfide solution is similar, as an increase in concentration increases the oxidation potential and the rate of corrosion below the passivation potential. Yet, for polysulfide ions, an increase in the oxidation potential above the passivation potential forms a stable passive film, resulting in low corrosion rates, whereas an increase in thiosulfate ions does not show the ability to increase the potential above that of the passivation potential.6 Recently, Singh et al.15 investigated the impact of temperature and hydroxide and sulfide concentration on liquor corrosivity. The results showed that an increase in the temperature or in the concentration of sulfide or hydroxide resulted in a decrease in the corrosion potential for austenitic stainless steels. Depending on the change in these parameters, the open-circuit potential, or potential at rest, of a steel could fall in the range where it is susceptible to stress corrosion cracking. The resultant potential shift might be due to the formation of hematite, Fe2O3, rather than magnetite, Fe3O4, on iron-based alloys. Further defining the relationship of the primary white liquor constituents, open-circuit potential (OCP), and corrosivity has shed some light on the corrosion susceptibility of materials used to construct process equipment, commonly carbon steel.8,16-18 In the present study, the effects of temperature and the hydroxide and sulfide concentrations on the corrosivity of liquors toward carbon steel A516-Gr70 and its OCP in solution were systematically investigated for a range of conditions beyond those encountered in the pulp and paper industry. Experimental Procedures Corrosion tests were carried out in 4-L duplex stainless steel autoclaves. All gravimetric (weight loss) corrosion tests in this study were carried out for 72 h at 25, 100, and 170 °C for all solutions described in Table 1. Mixing of the sulfide-containing caustic solutions was performed with gentle stirring of sodium hydroxide pellets into solution until dissolved, followed by addition of 9-hydrate sodium sulfide crystals, without formation of vortex. A minimum of two A516-Gr70 carbon steel corrosion coupons were exposed in each test to verify the reproducibility of the corrosion rate measurements in these tests; average rates of corrosion are reported. Samples not exhibiting reproducibility within 10% were discarded, and the experiment was repeated. Metal samples were polished to a 1.2-µm finish and cleaned with acetone. Each sample was weighed, and its area was measured before being mounting on an electrically isolated Teflon rack to eliminate galvanic effects (electrical contact between dissimilar metals). Samples were placed in the autoclave such that they did not touch the autoclave walls. Testing was performed under stagnant conditions with no mechanical stirring or agitation of the solution. Upon completion of each test, the samples were removed from the autoclave and placed on drying racks for further analysis. Dried coupons were

Figure 1. Plot of the corrosivity of carbon steel A516-Gr70 in sulfidecontaining caustic solutions at 25, 100, and 170 °C.

examined for visible corrosion features. Surface corrosion products on exposed samples were air-dried and characterized by X-ray diffraction (XRD). Subsequently, surface corrosion products were removed from the sample prior to sandblasting, and the final weight of each coupon was used to calculate the corrosion rate under the given test conditions. Electrochemical tests were also carried out in 4-L duplex stainless steel autoclaves at 25 and 90 °C for all solutions described in Table 1. Cylindrical metal samples were polished to a 1.2-µm finish, cleaned with acetone, and weighed, and the surface area was measured. Test samples were mounted on a polymer-coated steel rod, which was protected against corrosion. Carbon steel samples were further insulated from the autoclave material with Teflon fittings. Electrochemical tests were started when the autoclave reached the test temperature, and the conditions were maintained throughout the test. Potentiodynamic polarization tests were carried out using a platinum foil counter electrode and a double-junction saturated Ag/AgCl reference electrode with a saturated KCl salt bridge. All values are reported as saturated calomel electrode (SCE), 1 V (SHE) ) 1.222 V (AgCl) ) 1.242 V (SCE) at 25 °C.19,20 Samples were conditioned at cathodic potential for 5 min to reduce the oxide film on the surface. A potential scan rate of 1 mV/s was used for this study. Electrochemical tests at or above 170 °C were conducted using a commercial external-pressure-balanced reference electrode that uses saturated KCl in its salt bridge and a porous ceramic plug to hold the salt bridge solution. There were problems with the ceramic plug clogging at higher temperatures after extended exposure, so long-term electrochemical test data are not reported here. Results and Discussion Corrosion Dependence on Temperature. To determine the effects of temperature on the corrosion rate of carbon steels in sulfide-containing caustic solutions, tests were carried out at different temperatures ranging from 25 to 170 °C. The solutions used for these studies varied in caustic concentration and in the addition of sodium sulfide. The corrosion rates determined by weight change reported in this study represent an average corrosion rate over the entire test duration and not the instantaneous corrosion rate, which can change during the test duration. The rate of corrosion for carbon steel was found to increase with increasing temperature. In Figure 1, a plot of the temperature-corrosion relationship shows that the corrosion rate (CR) of carbon steel A516-Gr70 in the tested solutions increased

Ind. Eng. Chem. Res., Vol. 45, No. 23, 2006 7791 Table 2. Corrosion Rates (mm/year) for Carbon Steel A516-Gr70 in Sulfide-Containing Caustic Solutions at 25 °C Determined by the Tafel Slope Extrapolation and Gravimetric Test Methods NaOH (g/L)

Na2S (g/L)

Tafel extrapolation

gravimetric (avg)

100 200 0 100 200 0 100 200

0 0 37.5 37.5 37.5 75 75 75

0.09 0.13 0.44 2.21 1.74 2.79 8.14

0.03 0.01 0.01 0.11 0.02 0.11 0.12 0.15

Table 3. Corrosion Rates (mm/year) for Carbon Steel A516-Gr70 in Sulfide-Containing Caustic Solutions at 90 °C Determined by the Tafel Slope Extrapolation and Gravimetric Test Methods NaOH (g/L)

Na2S (g/L)

Tafel extrapolation

gravimetric (avg)

100 200 0 100 200 0 100 200

0 0 37.5 37.5 37.5 75 75 75

12.79 20.93 12.79 36.04 3.37 8.14 32.55

0.05 0.49 0.05 0.54 0.61 0.48 0.79 1.33

with increasing temperature. Although Figure 1 indicates the control of temperature independent of environment for the scope of chemical environments examined, the range of corrosion rates at each temperature suggests a dependence on the concentrations of sodium hydroxide and sodium sulfide. (Also included in Figure 1 is the average corrosion rate, 6.51 mm/year, for carbon steel tested in 150 g/L NaOH and 50 g/L Na2S at 170 °C.) Gravimetric Evaluation and Electrochemical Studies. To investigate how the temperature and the concentrations of hydroxide and sulfide affect the corrosivity of the liquor, electrochemical tests were performed at 25 and 90 °C. Tafel slope extrapolation was utilized in simple systems for the rapid determination of the corrosion rate, often in less than 1 h, in comparison to gravimetric tests, which can last longer than 2 weeks. In potentiodynamic polarization tests, the potential is scanned in a predetermined range at a fixed rate, typically 1 mV/s, and the current flowing between the working electrode (carbon steel) and the counter electrode (platinum electrode) is measured. The resulting plots of the potential versus the logarithm of the current density can be used to calculate the corrosion rate by the Tafel extrapolation method. Corrosion rate measurements are generally more reliable when measured using gravimetric testing, which allows for reactions to occur at the rest potential. However, this method typically takes a long time for measuring corrosion rates, especially when the corrosion rates are low. Potentiodynamic polarization curves for carbon steel samples were analyzed, and Tafel regions of the curves were used to calculate corrosion rates, as described elsewhere.21 The results from these measurements are reported in Tables2 and 3. The corrosion rates calculated from the Tafel extrapolation method in sulfide-containing caustic solutions did not correspond to the corrosion rates measured from gravimetric tests. However, general trends of increasing corrosion rates with increased temperature are seen with the electrochemical tests. The corrosion rates for carbon steel A516-Gr70 measured by the Tafel extrapolation method correlated better with the corrosion rates from gravimetric measurements in simple caustic solutions, without addition of sulfide, at both 25 and 90 °C. However, in solutions containing both hydroxide and sulfide, the corrosion

Figure 2. Open-circuit potentials for tested concentrations of caustic and sulfide solutions at 25 and 90 °C.

Figure 3. Open-circuit potentials for tested concentrations of sulfidecontaining caustic solutions at 25 and 90 °C.

rates predicted from the electrochemical measurements did not correlate well with the gravimetrically measured corrosion rates. At 25 and 90 °C, as shown in Tables 2 and 3, the corrosion rates calculated from Tafel extrapolations did not follow the trend of the gravimetric data. Instead, at 90 °C, the Tafel data anticipate a reduction in the corrosion rate with an increase in the sulfide content at both 100 and 200 g/L sodium hydroxide. This might be due to the presence of sulfide, which can participate in redox reactions at the steel surface, but which interferes with the use of polarization methods to calculate corrosion rates of metals accurately. Comparison of Open-Circuit Potential to Corrosivity of Environment. To understand the variation in corrosion rate measurements between the electrochemical and gravimetric methods, OCPs were measured for carbon steel A516-Gr70 samples exposed to different test conditions at 25 and 90 °C. The value of the stable OCP is plotted versus temperature in Figures2 and 3. As shown in Figure 2, an increase in hydroxide concentration at 25 and 90 °C and an increase in temperature at constant concentration correspond to a decrease in the OCP of carbon steel A516-Gr70 in solution. As also seen in Figure 2, a decrease in potential was not observed for increases in sulfide content at 25 °C. In fact, an increase in sulfide increased the potential. Yet, at 90 °C, this trend reversed itself, as an increase in sulfide did yield a decrease in measured OCP. For all compositions tested, the OCP decreased with increasing temperature. In sulfide-containing caustic solutions, polarization curves result from a number of possible redox reactions on the electrode surface, making it difficult to extract information about metal

7792

Ind. Eng. Chem. Res., Vol. 45, No. 23, 2006

Figure 4. Open-circuit potentials and corresponding measured corrosion rates for carbon steel A516-Gr70 in tested concentrations of sulfidecontaining caustic solutions at 25 and 90 °C.

oxidation reactions from the net current measurement. Sulfur can oxidize into various states including thiosulfate and sulfate. Thus, OCPs in these systems represent mixed potentials and include possible sulfur redox reactions. It is difficult to use a simple Tafel approach to extract information on metal oxidation reactions in such systems. However, OCPs can indicate thermodynamic tendencies for the metal to form more stable products.14 As shown in Figure 4, the determination of measured OCP values does not allow for a linear correlation of corrosion rate for carbon steel A516-Gr70 as expected for activationcontrolled reactions, such as iron oxidizing to ferrous or ferric ions. Although a similar range of OCP values, from approximately -0.40 to -1.15 V (SCE), was measured, a significant variation in corrosion rate was observed with increased temperature. Depending on whether the metal has a tendency to form a protective passive film on the surface or a soluble product, the corrosion rate at a given potential can vary significantly. Figure 4 shows the OCP for carbon steel A516-Gr70 versus gravimetrically measured corrosion rate, in millimeters per year, at different solution concentrations and at temperatures of 25 and 90 °C. It is clear from the results that, above approximately -1.00 V (SCE), the corrosion rates were relatively low, whereas below -1.00 V (SCE), the rate of corrosion for carbon steel increased significantly as temperature increased from 25 to 90 °C. This behavior is comparable to the potentiodynamic polarization of carbon steels in high-pH environments where passivity is generally stable above a critical potential. Further, these results are in agreement with previous studies on stress corrosion cracking of carbon steel in sulfide-containing caustic solutions by Mueller,22 Kesler and Bakken,23 and Singbeil.24 Surface Characterization Analysis. Figures5 and 6 show potential-pH diagrams for an Fe-S-H2O system at 25 and 170 °C, respectively, derived from a chemical equilibrium and reaction software database.25 These diagrams show the thermodynamically favored phases that exist for Fe-S-H2O systems at different pHs and potentials. At the pH range tested, solid iron oxide formations of Fe2O3 and Fe3O4 and soluble anion compounds FeO2(-a) and HFeO2(-a) are expected. A comparison of Figure 4 with the Pourbaix diagrams in Figures 5 and 6 indicates that, in a pH range of 12.5-14, typical for most pulping liquors, the stable phase at the carbon steel surface can be an Fe2O3 or Fe3O4 film or it can be soluble HFeOO-

ions. Environmental conditions such as temperature or changes in solution concentration can stabilize the OCP in the region of a soluble product, which can lead to higher corrosion rates. If an insoluble film is stable, the corrosion rates are low and depend on the stability and ionic or electronic conductivity properties of the passive film. As shown in Figures 5 and 6, the formation of FeS is not thermodynamically favorable above a pH of 11.5. Ionic species of soluble sulfur compounds HS2O8(-a), HS2O6(-a), HS(-a), and S(-2a) are also not thermodynamically favored above a pH of 10.5. However, work by Tromans14 has shown that sulfur is incorporated into Fe3O4 films in the region of critical current density, below the active-passive transition region of the polarization curve. Similar results were found in this study. XRD analysis was performed on the surface corrosion products of carbon steel A516-Gr70 samples tested in solution at 170 °C. The results, reported in Table 4, are limited to crystalline compounds readily identified in the diffraction pattern; further quantification of identified compounds was not performed. The inclusion of multiple crystalline layers and the incorporation of multiple compounds within the same layer were observed within the tested surface corrosion product samples. It is clear from the results that the surface films, in solutions with sulfide present, contained FeS along with Fe2O3 and Fe3O4. In the presence of sulfide, the incorporation of FeS into the Fe3O4 lattice structure can change the conductivity of the surface film and destabilize the surface corrosion product, leading to increased corrosion rates for carbon steel A516-Gr70.14 At the lower concentrations of hydroxide and sulfide tested here, an indication of iron sulfide compounds along with iron oxide did not correlate with an increase in corrosion rate. However, at increased concentrations of sulfide and hydroxide above 0 g/L sodium hydroxide and 37.5 g/L sodium sulfide, incorporation of FeS into the lattice structure did correlate with an increased rate of corrosion. Because of surface film instabilities, thermodynamic and XRD characterization of surface corrosion products in sulfide-containing caustic solutions is not sufficient in the evaluation of passivation and activation of carbon steel A516Gr70. Concentration Effects at Increased Temperature. To further investigate the effects of increasing temperature on liquor corrosivity toward carbon steel A516-Gr70, the corrosion rate at 170 °C was plotted as a function of increasing hydroxide concentration at constant sulfide content, as shown in Figure 7. An increase in sulfide content from 37.5 to 75 g/L at 170 °C and constant initial hydroxide concentration increased the rate of corrosion experienced by carbon steel A516-Gr70. Yet, an increase in added hydroxide from 100 to 200 g/L reduced the corrosion rate when in the presence of initial sulfide at both 37.5 and 75 g/L. The reduction in corrosion rate that occurs at increased hydroxide levels in the presence of sulfide might be due to the formation of a more stable film on the metal surface, such as Fe3O4, and a corresponding passivation of carbon steel A516-Gr70. In the presence of sulfide, the diffusion of iron ions from the metal surface at higher concentrations of hydroxide might be limited as a result of plugging of the passages within the lattice structure by sulfur compounds. The local plugging of pits has been suggested by Shoesmith et al.26 as a mechanism of film passivation. At 100 g/L sodium hydroxide and both 37.5 and 75 g/L sodium sulfide, the OCP can stabilize in a region of a soluble product, whereas an increase in hydroxide shifts the potential to a region of greater stability and a reduction in the corresponding corrosion rate. In the presence of sulfide, the incorporation of FeS into the lattice structure of the surface corrosion

Ind. Eng. Chem. Res., Vol. 45, No. 23, 2006 7793

Figure 5. Potential-pH diagram, Pourbaix diagram, of the Fe-S-H2O system at 25 °C calculated with a standalone database.25

Figure 6. Potential-pH diagram, Pourbaix diagram, of the Fe-S-H2O system at 170 °C calculated with a standalone database.25 Table 4. Crystalline Surface Corrosion Products of Carbon Steel A516-Gr70 in Sulfide-Containing Caustic Solutions at 170 °C NaOH (g/L)

Na2S (g/L)

CR (mm/year)

productsa

100 200 0 100 200 0 100 200

0 0 37.5 37.5 37.5 75 75 75

0.28 1.05 0.06 8.68 6.23 2.24 11.71 8.70

Fe2O3, FeO*Fe2O3, Fe3O4 Fe2O3, FeOOH, Fe3O4 FeS, Fe3S4, Fe2O3, Fe3O4 FeS, Fe3S4, Fe2O3, Fe3O4 Fe, FeS, Fe2O3, Fe3O4 FeS, Fe3S4, Fe3O4 FeS, Fe2O3, Fe3O4 FeS, Fe9S8, Fe3O4

a

Peaks identified by XRD.

product changes the conductivity of the more stable Fe3O4. With the addition of hydroxide, FeS can act as a destabilizing force to the surface film, enabling a significant increase in corrosion rate at all levels of hydroxide concentration in the presence of sulfide. Additionally, the formation of multiple crystalline compounds in the surface corrosion product can result in

distinctly different conductivity properties across the metal surface and nonuniform corrosion of the metal. Most of the published work in this area3,5-12,14-16,20 has studied corrosion rate of carbon steels as a function of specific environmental variables such as temperature or the concentration of hydroxide and/or sulfide and other oxidized sulfur species. Conclusions from these studies were contradictory and cannot be compared, as the exact compositions of experimental solutions were not similar. The present work relates the corrosion rate of carbon steel A516-Gr70 to its OCP, which can change with changes in specific environmental variables in the system, individually or in combination. The change in OCP further determines the thermodynamic tendency for a metal either to form a soluble product and corrode or to form an insoluble product that can form a passive film on the surface. Pourbaix diagrams for the selected system, shown in Figures 5 and 6, show thermodynamically stable products at a given pH and potential for iron or carbon steel. A comparison of the experimental results

7794

Ind. Eng. Chem. Res., Vol. 45, No. 23, 2006

Figure 7. Gravimetrically determined corrosion rate of A516-Gr70 carbon steel at 170 °C as a function of increasing sodium hydroxide concentration at constant sodium sulfide concentration, given as sodium hydrosulfide, Na2S + H2O f NaSH + NaOH.

with the Pourbaix diagrams in Figures 5 and 6 indicates that, under the tested conditions, when the OCP was in a region where soluble products were thermodynamically stable, carbon steel A516-Gr70 experienced higher corrosion rates. However, when the potential was in the region of insoluble products, carbon steel A516-Gr70 had relatively low rates of corrosion. Conclusions The measured ranges of corrosion rates, namely, 0.01-0.15 mm/year at 25 °C, 0.05-1.33 mm/year at 90 °C, and 0.0611.71 mm/year at 170 °C, suggest a dependence on the temperature and the concentrations of sodium hydroxide and sodium sulfide. At 170 °C and a constant initial hydroxide concentration, an increase in sulfide content from 37.5 to 75 g/L increased the rate of corrosion experienced by carbon steel A516-Gr70. Yet, an increase in added hydroxide from 100 to 200 g/L reduced the corrosion rate in the presence of initial sulfide at both 37.5 and 75 g/L. At the active-passive transition, a shift in OCP due to a change in concentration or temperature can result in carbon steel A516-Gr70 moving to a region of greater stability, with a reduction in corrosion rate, or to a region of greater solubility, which can increase the corrosion rate. In the tested environments, OCP values above approximately -1.00 V (SCE) corresponded to relatively low corrosion rates, below ∼0.2 mm/year, whereas below -1.00 V (SCE), the rate of corrosion for carbon steel A516-Gr70 could increase significantly, above ∼1 mm/year. Literature Cited (1) Yeske, R. A.; Garner, A. Processing Changes and Materials Engineering Challenges in the Pulp and Paper Industry. In 7th International Symposium on Corrosion in the Pulp and Paper Industry; TAPPI Press: Atlanta, GA, 1992; p 1. (2) Lopez, I.; Chang, H.-M.; Jameel, H.; Wizani, W. Effect of Sodium Sulfide Pretreatment on Kraft Pulping. In TAPPI Pulping Conference; TAPPI Press: Orlando, FL, 1999; Vol. 1, p 135.

(3) Smook, G. A. Handbook for Pulp & Paper Technologists, 2nd ed.; Angus Wilde Publications: Vancouver, BC, Canada, 1992. (4) Arpalahti, O.; Engdahl, H.; Ja¨ntti, J.; Kiiskila¨, E.; Liiri, O.; Pekkinen, J.; Puumalainen, R.; Sankala, H.; Vehmaan-Kreula, J. Papermaking Science and Technology, White Liquor Preparation. In Book 6B, Chemical Pulping; Gullichsen, J., Fogelholm, C.-J., Eds.; TAPPI Press: Atlanta, GA, 1999; p B135. (5) Peterman, L.; Yeske, R. A. Thiosulfate Effects on Corrosion in Kraft White Liquor. In CORROSION/87; NACE: San Francisco, CA, 1987; No. 201, p 1. (6) Yeske, R. A. Closure Effects on Pulp Mill Corrosion. In 1985 TAPPI EnVironmental Conference; TAPPI Press: Atlanta, GA, 1985; p 327. (7) Haegland, B.; Roald, B. The Corrosion of Steel in White Liquor. Nor. Skogind. 1955, 9, 351. (8) Kesler, R.B; Bakken, J. F. Corrosion of Mild Steel in Alkaline Pulping Liquors. Part 1. Kraft White Liquor. TAPPI J. 1958, 41, 3, 97. (9) Stockman, L. Sulfate Digester Corrosion as a Function of WhiteLiquor Polysulfide Content. SVen. Papperstidn. 1960, 63 (13), 425. (10) Wensley, D. A.; Charlton, R. S. Corrosion Studies in Kraft White Liquor: Potentiostatic Polarization of Mild Steel in Caustic Solutions Containing Sulfur Species. Corrosion 1980, 36, 8, 385. (11) Tonsi-Eldakar, N. Corrosivity of Kraft Liquors. In CORROSION/ 80; NACE: Chicago, IL, 1980; No. 258; p 1. (12) Laliberte´, L. H. Corrosion Problems in the Pulp and Paper Industry. In CORROSION/77; NACE: San Francisco, CA, 1970; No. 165, p 1. (13) Yasuda, M; Okada, M; Hine, F. Corrosion of Carbon Steel in Hot NaOH Solutions Under Heat Transfer Conditions. In CORROSION/82; NACE: Houston, TX, 1982; No. 5, p 256. (14) Tromans, D. Anodic Polarization Behavior of Mild Steel in Hot Alkaline Sulfide Solutions. J. Electrochem. Soc. 1980, 127 (6), 1253. (15) Singh, P. M.; Ige, O.; Mahmood, J. Stress Corrosion Cracking of Type 304L Stainless Steel in Sodium Sulfide-Containing Caustic Solutions. Corrosion 2003, 59, 10, 843. (16) Kesler, R. B. Corrosion of Mild Steel in Alkaline Pulping Liquors. Part 2. Neutral Sulfite Cooking Liquor. TAPPI J. 1958, 41, 3, 102. (17) Kesler, R. B. Corrosion of Mild Steel in Alkaline Pulping Liquors. Part 3. Special Effect of Sulfite Ion in Kraft White Liquor. TAPPI J. 1960, 43, 4, 355. (18) Von Essen, C. G. Corrosion Problems in Sulfate Pulp Mills. TAPPI J. 1950, 33, 7, 14. (19) Janz, G. J. Silver-Silver Halide Electrodes. In Reference Electrodes, Theory and Practice; Ives, D. J. G., Janz, G. J., Eds.; Academic Press: New York, 1961; p 189. (20) Hills, G. J.; Ives, D. J. G. The Calomel Electrode and Other Mercury-Mercurous Salt Electrodes. In Reference Electrodes, Theory and Practice; Ives, D. J. G., Janz, G. J., Eds.; Academic Press: New York, 1961; p 159. (21) Kelly, R. G.; Scully, J. R.; Shoesmith, D. W.; Buchheit, R. G. Electrochemical Techniques in Corrosion Science and Engineering; Schweitzer, P. A., Ed.; Marcel Dekker, Inc.: New York, 2003. (22) Mueller, W. A. Mechanism and Prevention of Corrosion of Steels Exposed to Kraft Liquors. In CORROSION/74; NACE: Houston, TX, 1974; p 109. (23) Kesler, R. B.; Bakken, J. F. Corrosion of Mild Steel in Alkaline Pulping Liquors. Part 1. Kraft White Liquor. TAPPI J. 1958, 41 (3), 97. (24) Singbeil, D. Kraft Continuous Digester Cracking: Causes and Prevention. Pulp Pap. Can. 1986, 87 (3), 93. (25) Chemical Reaction and Equilibrium Software with ExtensiVe Thermochemical Database, version 5.1; Outokumpu Research Oy: Pori, Finland, 2002. (26) Shoesmith, D. W.; Taylor, P.; Bailey, M. G.; Ikeda, B. Electrochemical Behavior of Iron in Alkaline Sulfide Solutions. Electrochim. Acta 1978, 23, 903.

ReceiVed for reView March 23, 2006 ReVised manuscript receiVed September 5, 2006 Accepted September 6, 2006 IE060358P