Corrosion Science Section

CORROSION SCIENCE SECTION

Sigma Phase Formation and Polarization Response of UNS S31803 in Sulfuric Acid R. Magnabosco‡,* and N. Alonso-Falleiros**

ABSTRACT For a better understanding of the relationship between the microstructure of UNS S31803 duplex stainless steel (DSS) and the shape of the polarization curves, this study evaluated the influence of the microstructure on the potentiodynamic polarization of the 850°C isothermal-aged UNS S31803 DSS in 0.5 M sulfuric acid (H2SO4). In the transpassive region, selective corrosion of chromium- and molybdenum-rich phases occurred. In the solution-treated sample, ferrite was selectively corroded, and in all aged samples, the sigma phase was the selectively corroded phase. Five current density maxima in the passive region were found during potentiodynamic polarization, and they can be related to the microstructures formed. The current density maximum at 564 mV vs. saturated calomel electrode (SCE) can be related to secondary ferrite, impoverished in chromium and molybdenum, that was formed during direct precipitation of sigma phase from the original ferrite. Secondary austenite, impoverished in chromium and molybdenum and formed together with the sigma phase during eutectoid decomposition of the original ferrite, can be related to the current density maxima at –85 mVSCE and –40 mVSCE. The austenite phase, present in all heat-treatment conditions, can be related to the current density maxima at –155 mVSCE and 111 mVSCE. Submitted for publication November 2004; in revised form, December 2004. Corresponding author. E-mail: [email protected]. * Department of Mechanical Engineering, Ignatian Educational Foundation – FEI, Av. Humberto A.C. Branco, 3972, 09850-901, São Bernardo do Campo, SP, Brazil. ** Department of Metallurgical and Materials Engineering, Polytechnic School, University of São Paulo, Av. Prof. Mello Moraes, 2463, 05508-900, São Paulo, SP, Brazil. (1) UNS numbers are listed in Metals and Alloys in the Unified Numbering System, published by the Society of Automotive Engineers (SAE International) and cosponsored by ASTM International. ‡

CORROSION—Vol. 61, No. 8

KEY WORDS: duplex stainless steel, microstructure, sigma phase, potentiodynamic polarization, sulfuric acid, UNS S31803

INTRODUCTION One of the earliest studies concerning the microstructural influence on the electrochemical behavior of duplex stainless steels (DSS) was carried out with 19%Cr-11%Ni steel, solution-treated to obtain a ferriteaustenite duplex structure.1 During potentiostatic polarization at –200 mV vs. saturated calomel electrode (SCE) in 20% sulfuric acid (H2SO4), conducted after prepolarization in the cathodic region (–600 mVSCE, to remove the passive film), selective corrosion of the austenite phase occurred. The potentials used were determined after potentiodynamic polarization in the same solution of an austenitic stainless steel (18%Cr9%Ni) and a ferritic one (27%Cr); the results showed that the anodic active region of the austenitic sample is located between –300 mVSCE and –100 mVSCE, and the anodic active region of the ferritic sample is located between –450 mVSCE and –300 mVSCE. However, potentiostatic polarization of the duplex sample in the ferrite active region did not lead to selective etching of this phase.1 Potentiodynamic polarization curves of DSS similar to UNS S31803,(1) in 2 M H2SO4 solutions with 1 M and 2 M hydrochloric acid (HCl) additions, showed selective corrosion of ferrite in potentials near the corrosion potential, and selective corrosion of austenite near the passivation potentials.2-3 Heat treatments conducted to change the volume fraction of austenite

0010-9312/05/000165/$5.00+$0.50/0 © 2005, NACE International

807

CORROSION SCIENCE SECTION

TABLE 1 Chemical Composition of the Investigated Steel (wt%) Cr

Ni

Mo

Mn

Si

V

N

C

S

22.2

5.7

2.98

1.60

0.44

0.07

0.161

0.016

0.001

and ferrite led to changes in the shape of the polarization curves, which was attributed to the rise of anodic areas (related to the ferrite phase) and the reduction of cathodic areas (related to the austenite phase).3 Selective corrosion of the ferrite phase occurred during potentiodynamic polarization at 60°C and 80°C of UNS S31803 DSS in 1 M H2SO4, if sodium chloride (NaCl) additions were higher than 0.2 M. NaCl additions of 0.05 M and 0.1 M led to generalized corrosion of the two phases. However, in the absence of chlorides the austenite phase was selectively corroded at the corrosion potential.4 The presence of the sigma phase reduced the corrosion potential measured before the potentiodynamic polarization of UNS S31803 (SAF 2205) in 10% H2SO4 solution; the higher the sigma phase content, the lower the corrosion potential, which varied from 299 mVSCE in the solution-treated sample (without the sigma phase) to 78 mVSCE in samples with 21% of the sigma phase. No active peaks were observed in the polarization curves, and the transpassive potential remained relatively constant (around 1,040 mVSCE).5 Potentiodynamic polarization of the solutiontreated UNS S31803 (SAF 2205) in 1 M H2SO4, at 25°C, yields polarization curves that exhibit two current density maxima: at –200 mVSCE in the active-passive transition and at 650 mVSCE in the passive region. The transpassive potential was around 1,050 mVSCE.6 To better understand the relationship between the microstructure of DSS and the polarization curves, this work evaluated 850°C isothermal-aged UNS S31803 DSS in 0.5 M H2SO4.

EXPERIMENTAL PROCEDURES The studied material had the chemical composition given in Table 1, and was received as a 3-mmthick sheet, solution-treated at 1,120°C and waterquenched. Specimens of 20 mm length and 15 mm width were machined, with its length parallel to the rolling direction. After machining, specimens were isothermically aged for 10 min, 30 min, 1 h, 5 h, or 100 h at 850°C, in a tubular electric furnace that controlled the aging temperature to within 1°C. All heat treatments were conducted in a 99.9% N2 atmosphere to suppress oxidation of the sample surfaces and prevent nitrogen loss from the DSS. Following heat treatments, the material was water-quenched. After the heat treatments the specimens were abraded using silicon carbide (SiC) papers to a

808

600-grit finish before mounting in thermosetting plastic, leaving an exposed surface area of ~0.5 cm2, parallel to the rolling direction. The mounted samples were metallographic polished in a semi-automatic grounding and polishing machine, with final polishing provided by 1-µm diamond abrasive. Samples were then etched for 40 s in modified Behara reagent, composed of 20 mL HCl, 80 mL distilled water, and 1 g potassium metabisulfide (K2S2O5); to this stock solution, 2 g of ammonium bifluoride (NH4F·HF) were added just before the etching. This etching procedure clearly distinguished the ferrite, austenite, and sigma phases. Selective etching of the sigma phase was obtained through electrolytic etching in 10% potassium hydroxide (KOH) aqueous solution, using 2 Vdc for 1 min. Quantitative metallography of the sigma phase was performed with an automated image analysis system attached to an optical microscope. The ferrite content of the samples was obtained using magnetic measurements; the austenite content was calculated using Equation (1): % γ = 100 – %α – % σ

(1)

Potentiodynamic polarization tests were conducted in a 0.5-M H2SO4 aqueous solution (pH 2), exposed to laboratory air, at a controlled temperature of 22°C ± 2°C. The test cell had a platinum wire as a counter electrode and an SCE as a reference electrode. The electrochemical tests were repeated 20 times for each heat-treatment condition. Immediately after polishing, sample surfaces were washed with ethyl alcohol (C2H5OH) and dried with hot blown air. Within 6 s, the sample was immersed in the test solution, and the electrode potential (E) was continuously changed from a potential 300 mV below the open-circuit potential to 1,200 mVSCE, at a scan rate of 1 mV/s. The corroded sample surfaces were examined using optical microscopy and scanning electron microscopy using secondary electron imaging (SEI); the identification of corroded phases was possible using energy-dispersive spectroscopy (EDS).

RESULTS Figure 1 shows the microstructure evolution of the aging process at 850°C. The solution-treated material (Figure 1[a]) presented 40.9 ± 1.9% of ferrite and 59.1 ± 1.9% of austenite, distributed as alternated

CORROSION—AUGUST 2005

CORROSION SCIENCE SECTION

(a)

(b)

(c)

(d)

(e)

(f)

FIGURE 1. Typical microstructures found in UNS S31803 DSS studied: (a) solution-treated, showing ferrite (dark) and austenite, and aged at 850°C for (b) 10 min, (c) 30 min, (d) 1 h, and (e) 5 h, showing ferrite (dark), austenite (gray), and sigma. In (f), sample aged for 100 h, with austenite (gray) and sigma. Optical microscopy. Etchant: modified Behara.

bands throughout the sheet thickness. The sigma phase was nucleated in the interfaces between ferrite and austenite (Figure 1[b]) in the initial period of aging, and grew preferentially, consuming the ferrite phase (Figures 1[c] through [e]). Indications of the eutectoid decomposition of ferrite could be noted in

CORROSION—Vol. 61, No. 8

samples aged 30 min at 850°C, with lamellar growth of sigma together to the formation of austenite, called secondary austenite (γs), which is probably impoverished in chromium and molybdenum compared to the original austenite present in the solution-treated sample. However, the eutectoid decomposition of fer-

809

CORROSION SCIENCE SECTION

(a)

(b)

(c)

(d)

FIGURE 2. Typical microstructures found in 850°C aged UNS S31803 DSS after selective etching to sigma phase. Aging times: (a) 10 min, (b) 30 min, (c) 1 h, and (d) 100 h. Optical microscopy. Electrolytic etching in KOH.

FIGURE 3. Phase content of UNS S31803 DSS as a function of aging time at 850°C.

rite could also generate divorced microstructures. After total ferrite consumption, the sigma phase could be formed inside the original austenite (Figure 1[f]). Figure 2 shows examples of aged samples after electrolytic etching in KOH; this procedure allowed the determination of the sigma phase content. Those

810

results, together with the magnetic measurements of the ferrite content and the use of Equation (1), allowed the determination of the volume fraction of the phases as a function of aging time (Figure 3). Figures 4 through 6 show the potentiodynamic polarization curves for the material in all heattreatment conditions. The corrosion potential of all samples remained constant at –200 mVSCE. The transpassive potential also remained unchanged at about 900 ± 25 mVSCE. In the passive region, reproducible current density maxima occurred at specific potentials for each sample, identified from E1 to E5 in Figure 5. For each heat-treatment condition, the mean values of potentials where the current density maxima occurred, and the standard deviation associated with the results obtained in the 20 tests, are listed in Table 2. After potentiodynamic polarization, sample surfaces were observed in a scanning electron microscope. Figure 7(a) shows the solution-treated sample, and EDS analysis confirmed that the ferrite phase was selectively corroded. In aged samples (Figures 7[b] through [f]), the sigma phase was preferably cor-

CORROSION—AUGUST 2005

CORROSION SCIENCE SECTION

roded. Some polarization curves were interrupted at the transpassive potential (900 mVSCE); in those cases, the sample surfaces remained in the as-polished condition, confirming that the selective corrosion of ferrite in the solution-treated material, and the selective corrosion of the sigma phase in the aged samples, occurred in the transpassive region.

DISCUSSION In aging times up to 10 min, the sigma phase was formed by consuming the original ferrite phase, and the austenite content remained unchanged, as showed in Figure 3. This is an indication that direct precipitation from ferrite is the primary mechanism of the sigma phase formation at aging times up to 10 min, and this probably leads to an impoverishment in chromium and molybdenum in the remaining ferrite, which could be called secondary ferrite (αs). The progress of the aging process leads to the continuous growth of sigma phase content (Figure 3), but there were modifications in the mechanism of the sigma phase formation. If aging was conducted for 30 min, a reduction in the volume fraction of ferrite and austenite was observed, indicating growth of the sigma phase (rich in chromium and molybdenum) through the consumption of the ferrite and austenite. After 30 min of aging, the microstructure probably had larger amounts of αs and γs. The presence of lamellar microstructures after 30 min of aging (Figure 1[c]) is an indication of sigma phase formation from the eutectoid decomposition of ferrite, with the formation of γs. This is clearly observed after 1 h of aging, when the austenite content was higher than after 30 min of aging (Figure 3). After 5 h of aging the ferrite consumption was almost complete, but the sigma content continued to grow up until 100 h of aging. This, and the formation of the sigma phase inside the austenite (Figure 1[f]), are other indications of sigma phase nucleation and growth from austenite, and the formation of the γs. Table 3 summarizes the major mechanisms of sigma phase formation and the probable related phases formed as a function of aging time. From the analysis of polarization curves shown in Figures 4 through 6 it was concluded that the microstructural complexity of the samples did not affect the

FIGURE 4. Potentiodynamic polarization curves in 0.5 M H2SO4 of UNS S31803 DSS solution-treated or aged 10 min at 850°C.

FIGURE 5. Potentiodynamic polarization curves in 0.5 M H2SO4 of UNS S31803 DSS aged 30 min or 1 h at 850°C. Current density maxima are identified from E1 to E5.

FIGURE 6. Potentiodynamic polarization curves in 0.5 M H2SO4 of UNS S31803 DSS aged 5 h or 100 h at 850°C.

TABLE 2 Identification of Current Density Maxima Potentials for All Heat-Treatment Conditions Identification (mVSCE)

SolutionTreated

10 min

30 min

E1 E2 E3 E4 E5

–171 ± 17 — — 113 ± 34 —

–137 ± 35 — — 129 ± 45 542 ± 31

— –85 ± 37 — — 562 ± 26

CORROSION—Vol. 61, No. 8

Aging Time at 850°C 1h –161 ± 13 — –40 ± 18 86 ± 18 603 ± 37

5h

100 h

Mean Potential

–149 ± 37 — — 108 ± 16 550 ± 31

–158 ± 12 — — 117 ± 31 —

–155 –85 –40 111 564

811

CORROSION SCIENCE SECTION

(a)

(b)

(c)

(d)

(e)

(f)

FIGURE 7. Scanning electron micrographs (SEI) of sample surfaces after polarization tests: (a) solution-treated, showing selective corrosion of ferrite. Specimens aged at 850°C for (b) 10 min, (c) 30 min, (d) 1 h, (e) 5 h, and (f) 100 h, showing selective corrosion of sigma. α and γ identify ferrite and austenite, respectively.

transpassive potential, which was always about 900 ± 25 mVSCE, or 1,142 ± 25 mV vs. hydrogen electrode (H); this behavior had already been observed in an earlier work.5 In fact, the reaction described in Equation (2),7 associated with the beginning of the transpassive region, has an equilibrium potential of

812

1,074 mVH (832 mVSCE) in the 0.5-M H2SO4 solution (pH 2), confirming that the transpassive region begins near 900 mVSCE: Cr 3+ + 4 H2O = HCrO 4– + 7H + + 3 e E o = 1, 350 – 137.9pH (mVH )

(2)

CORROSION—AUGUST 2005

CORROSION SCIENCE SECTION

TABLE 3 Mechanisms of Sigma Phase Formation and Probable Related Phases Formed as a Function of Aging Time at 850°C Aging Time at 850°C

Major Mechanism of Sigma Formation

10-min Direct Precipitation from α

Related phases

αs

30-min Growth from α and γ

1-h Eutectoid Decomposition of α and Growth from α and γ

5-h Growth from α and γ

100-h Nucleation and Growth from γ

α s, γ s

α s, γ s

α s, γ s

γs

Selective corrosion of chromium- and molybdenum-rich phases was observed in the transpassive region: in the solution-treated sample, ferrite was selectively corroded, and in all aged samples, the sigma phase was selectively corroded. This is associated with an increase in the current density in the transpassive region. Another reason for the increase in current density in the transpassive regions is O2 evolution (Equation [3]), whose equilibrium potential in a solution of pH 2 is 1,110 mVH (868 mVSCE): 2 H 2O = O 2 + 4H + + 4 e E o = 1, 228 – 59.1pH (mVH )

Identification (mVSCE)

Mean Potential

Related Microstructure

E1 E2 E3 E4 E5

–155 –85 –40 111 564

γ γs γs γ αs

(3)

The current density maximum identified in Table 2 as E5 was seen in all aged samples that contained secondary ferrite (αs), and was related to the presence of this phase in the microstructure. Between E5 (564 mVSCE) and the corrosion potential (–200 mVSCE), two current density maxima were observed in all samples, except for the one aged 30 min. These two current density maxima, identified in Figure 5 and Table 2 as E1 and E4, are probably related to the austenite phase, the only one present in all samples. Note that E1 (–155 mVSCE) occurs in the anodic region of 18%Cr9% Ni austenitic steel in 20% H2SO4.1 However, the association of E1 and E4 with austenite is valid only if the electrochemical behavior of the sample aged 30 min at 850°C can be explained. In those samples, these two current density maxima are not present, and a new current density maximum occurs at –85 mVSCE (identified as E2 in Figure 5 and Table 2). This maximum could be explained as an increase in the corrosion rates of the reactions that occur at E1, E3, and E4, which are the current density maxima found in the sample aged 1 h at 850°C (Figure 5). Besides, the occurrence of the E3 maximum only in the sample aged for 1 h suggests that the same electrochemical reactions occur in this sample and in the sample aged for 30 min, with differences in corrosion rates, which are lower in the 1-h-aged sample. The E3 current density maxima are related to anodic electrochemical reactions over γs, which are present in the microstructure after 30 min of aging, and probably become prominent in the microstructure after 1 h of aging, as a result of the eutectoid decomposition of ferrite in sigma and γs. Those anodic electrochemical reactions are more intense if the impoverishment in

CORROSION—Vol. 61, No. 8

TABLE 4 Relation Between Current Density Maxima Potentials and Microstructure of Solution-Treated or 850°C Aged UNS S31803 DSS

chromium and molybdenum of γs is greater, and this happens in the sample aged 30 min at 850°C. After 1 h of aging, together with the rise in the γs content, a chromium and molybdenum redistribution could occur in the earlier formed γs: the anodic reaction occurs with lower current density, and E1, E3, and E4 could be observed separately. For aging times of 5 h or more, despite the higher sigma phase and γs contents, the chromium and molybdenum redistribution reduced the probability of those anodic reactions, leading to the absence of E3 current density maximum. The relation between current density maxima and the microstructure of 850°C aged UNS S31803 DSS is summarized in Table 4.

CONCLUSIONS ❖ Selective corrosion of chromium- and molybdenumrich phases occurred in the transpassive region of UNS S31803 DSS in 0.5 M H2SO4 aqueous solution: in the solution-treated sample, ferrite was selectively corroded, and in all aged samples, the sigma phase was selectively corroded. ❖ Five current density maxima in the passive region were found during potentiodynamic polarization in 0.5 M H2SO4 of 850°C aged UNS S31803 DSS, and those maxima were related to the microstructures formed. ❖ Current density maximum occurred at 564 mVSCE and can be related to secondary ferrite, impoverished in chromium and molybdenum, which was formed during direct precipitation of the sigma phase from the original ferrite. ❖ Secondary austenite, impoverished in chromium and molybdenum and formed together with the sigma

813

CORROSION SCIENCE SECTION

phase during eutectoid decomposition of the original ferrite, is responsible for the current density maxima occurring at –85 mVSCE and –40 mVSCE.5 ❖ The austenite phase, present in all heat-treatment conditions, is responsible for the current density maxima that occurred at –155 mVSCE and 111 mVSCE. REFERENCES 1. E. Mor, A. Carlini, V. Scotto, E. Traverso, Metall. Ital. 64, 6 (1972): p. 261.

814

2. E. Symniotis-Barrdahl, “Selective Corrosion of Duplex Stainless Steels,” Proc. Stainless Steels ’87 (York, U.K.: The Institute of Metals, 1988), p. 176. 3. E. Symniotis, Corrosion 46, 1 (1990): p. 2. 4. J.W. Fourie, F.P.A. Robinson, “Mechanistic Aspects of Selective Corrosion of a 22% Cr Duplex Stainless Steel in Acid Chloride Mixtures,” Proc. Int. Conf. on Stainless Steels (Chiba, Japan: ISIJ, 1991), p. 11. 5. J.H. Potgieter, Br. Corros. J. 27 (1992): p. 219. 6. J.H. Potgieter, J. Mater. Sci. Lett. 15 (1996): p. 1,408. 7. M. Pourbaix, Atlas of Electrochemical Equilibria in Aqueous Solutions, trans. J.A. Franklin (Houston, TX: NACE International, 1974), p. 256.

CORROSION—AUGUST 2005