Effects of Sulfur Addition on the Formation of Inclusionsand the Corrosion Behavior of Super Duplex Stainless Steelsin Chloride Solutions of Different pH

Soon-Hyeok Jeon, Soon-Tae Kim+, Jun-Seob Lee, In-Sung Lee and Yong-Soo Park

Department of Materials Science and Engineering, Yonsei University, Seoul 120-749, Korea

Effects of sulfur addition on the formation of inclusions and the corrosion behavior of super duplex stainless steels in chloride solutions ofdifferent pH were investigated using potentiodynamic and potentiostatic polarization techniques. As sulfur content increased, the corrosionresistance decreased due to the formation of numerous manganese sulfides deteriorating the corrosion resistance. The solutions used for thisstudy are NaCl and NaCl + HCl solutions. The corrosion behavior is dominated by the pitting corrosion in NaCl solution of pH 7 and by bothuniform and pitting corrosion in NaCl + HCl solution of pH 1. The pitting corrosion of the alloys was initiated at all types of inclusions ofmanganese sulfides, manganese oxy-sulfides and manganese oxides and then propagated to the ferrite phases, and finally propagated to theaustenite phases. [doi:10.2320/matertrans.M2012194]

(Received May 21, 2012; Accepted June 27, 2012; Published August 8, 2012)

Keywords: polarization, duplex stainless steel, pitting corrosion, inclusion, sulfur

1. Introduction

Super duplex stainless steels have been increasingly usedfor various applications such as desalination facilities andchemical plants due to the high resistance to pitting, crevicecorrosion and stress corrosion cracking and good mechanicalproperties.1­3)

Manganese sulfide (MnS) is well known to provideinclusions favorable for machining. Given adequate con-ditions, such inclusions build a protective layer on the toolsurface, termed a built-up layer (BUL), which promotes wearreduction of the tool.4­6) The manganese sulfide inclusionsgenerate good results with respect to costs and productivity,motivating constant studies aimed at machinability improve-ments.7)

As referred to Fig. 1, Jeon and Kim elucidated that the toollife for an alloy with 0.053mass% S was about 3 times longerthan that of an alloy with 0.001mass% S at a cutting speedof 100mmin¹1, due to the lubricating films of manganesesulfides adhering to the tool surface.

However, the intentional formation of sulfide inclusionssuch as MnS for the purpose of improving machinability, canlead to a deterioration of the resistance of the steels to pittingcorrosion.8,9) In general, MnS in commercial stainless steelshas been known to act as the initiation site of pittingcorrosion in environments exposed to chloride ions.10­14)

While the detrimental role of MnS has been recognized, theexact mechanisms of the pitting corrosion related to itspresence have not been elucidated. In spite of a decrease incorrosion resistance due to addition of sulfur, in considerationof the machining cost and productivity, it is certainlyworthwhile for industrial aspects if the machinability canbe improved through optimizing alloy design by the additionof sulfur to super duplex stainless steels.

The corrosion behavior of duplex stainless steel is greatlyaffected by the difference in chemical composition betweenferrite (¡) and austenite (£) phases. In general, Cr and Mo

contents are higher in the ¡-phase whereas Ni and N aremuch higher in the £-phase. Nitrogen (N) is a strong £-stabilizer, and it increases the temperature at which thetransformation of the ¡-phase to £-phase occurs.15­17)

Accordingly, by controlling such alloying elements toachieve a balance of corrosion resistance between the ¡-phase and £-phase, excellent corrosion resistance can beachieved. The pitting resistance equivalent number (PREN)has been introduced to estimate the relative pitting corrosionresistance in chromium (Cr), molybdenum (Mo), tungsten(W) and nitrogen (N) concentrations. Lorenz and Medawarfound the following relationship between the PREN andthe alloy composition.18) They reported that the PREN ofstainless steels, a criterion for the resistance to pittingcorrosion, is defined by the following experimental equationand is calculated according to the chemical compositions(mass%) of the major alloying elements such as Cr, Mo and

1.95 2.00 2.05 2.10 2.15 2.20 2.25

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0 BASE DX5S

log V (Cutting Speed, m/min.)

log

T (

Too

l Lif

e, m

in.)

Tailor's equation

BASE : log T = 6.82 - 3.30 log V DX5S : log T = 4.78 - 2.05 log V

Fig. 1 Effects of sulfur addition and cutting speed on the tool life of thealloys. (feeding rate of 0.3mm/rev., cutting depth of 1mm and no cuttingfluid).

+Corresponding author, E-mail: stkim78@yonsei.ac.kr

Materials Transactions, Vol. 53, No. 9 (2012) pp. 1617 to 1626©2012 The Japan Institute of Metals

N, which greatly affect the resistance to pitting corrosion ofstainless steels.

PREN ¼ ½mass% Cr� þ 3:3ð½mass% Mo�þ 0:5½mass% W�Þ þ 16½mass% N� ð1Þ

However, Bernhardsson et al.19) reported that while theformula is relevant for austenitic stainless steels (ASS) andthe austenitic phase in duplex stainless steel (DSS), it is notvalid for DSS for the following reasons. First, nitrogen isnearly completely solutionized in the austenite phase in DSSwhereas it is rarely solutionized in the ferrite phase in DSS.The solubility of nitrogen in the ferrite phase in DSS hasa maximum 0.05mass%. Second, the addition of nitrogenchanges the partitioning coefficients for chromium andmolybdenum.20) Therefore, the relationship between thePREN and the alloying content of DSS is more complexthan that of ASS. However, it is reasonable to nearly doublethe coefficients for the N content;

PREN ¼ ½mass% Cr� þ 3:3ð½mass% Mo�þ 0:5½mass% W�Þ þ 30½mass% N� ð2Þ

The PREN formula in eq. (2) for DSS has been used byseveral researchers to investigate DSS’s resistance tolocalized corrosion.21,22) However, the reasonable nitrogenfactor in the PREN equation is still disputed. In particular, itis necessary to quantitatively verify the mechanisms of theinitiation and propagation of pitting corrosion due to thePREN differences (PREN£ ¹ PREN¡) between the £-phaseand ¡-phase in duplex stainless steels and the pittingcorrosion behavior in chloride solutions of different pH.

To elucidate the effects of sulfur addition on the formationof inclusions and the corrosion behavior of super duplexstainless steels in chloride solutions of different pH and thereasonable nitrogen factor in the PREN equation, a metallo-graphic examination, potentiodynamic & potentiostatic polar-ization tests, scanning electron microscope and energydispersive spectroscope (SEM-EDS) analyses of inclusionand austenite (£) phase and ferrite (¡) phase, a scanningAuger multi-probes (SAM) analysis of two phases, anda thermodynamic calculation using Thermo-Calc softwarewere conducted.

2. Experimental Procedures

2.1 Material and heat treatmentThe experimental alloys were prepared by atmospheric

melting in a high frequency induction furnace. For the rawmaterial, commercial high-grade pure iron, chromium (Cr),molybdenum (Mo), nickel (Ni), ferro-chromium nitride ofFe­57.5mass% Cr­6.5mass% N, and ferro-sulfide of Fe­45mass% S were used. Table 1 shows the chemical

composition of the experimental alloys designated as BASEand DX5S. The alloys were solution heat-treated for 30minat 1403K and then cooled in furnace to 1353K, and thenquenched in water.

2.2 Thermodynamic equilibrium calculationTo predict the phase diagram and PREN values of ferrite

and austenite phases in experimental alloy, thermodynamiccalculation using Thermo-Calc software was conducted. Thesteel database TCFE5 available with the Thermo-Calcsoftware was used to perform the calculation and it shouldbe stressed that such calculation give the equilibrium state ofsystem.

2.3 Corrosion testsEffects of sulfur addition on the corrosion behavior of

super duplex stainless steels in chloride solutions of differentpH were investigated using a potentiodynamic anodicpolarization technique. The corrosion potential (Ecorr), thepitting potential (Epit), the passive region (¦Ep), thepassivation current density (Ip) and the corrosion currentdensity (Icorr) were obtained from the potentiodynamicpolarization curves. The corrosion current density (Icorr)was commonly obtained by the extrapolation of the cathodicand anodic curve between 50 and 100mV away from thecorrosion potential. The pitting potential marked the end ofthe passive potential region and the transition from passiveto transpassive behavior. The potentiodynamic anodic polar-ization test was conducted in a deaerated 3.0M NaCl ofpH 7 and a deaerated 0.5M HCl + 2.5M NaCl of pH 1 at338K according to the ASTM G 5.23) The pH value wasregularly checked using a Mettler pH meter.

Test specimens were joined with copper wire throughsoldering (95mass% Sn­5mass% Sb), and then mountedwith an epoxy resin. One side of the sample was ground to600 grit using SiC abrasion paper. After defining the exposedarea of the test specimen as 0.5 © 10¹4m2, the remainder waspainted with a transparent lacquer. The test was conductedat a potential range of ¹0.65 to +1.1V vs. SCE (saturatedcalomel electrode) and at a scanning rate of 1 © 10¹3V s¹1,using a SCE. The current transients through the potentiostatictest were measured in deaerated 3.0M NaCl and deaerated0.5M HCl + 2.5M NaCl solution at 338K with an appliedpotential of 0V vs. SCE in the passive region of thepotentiodynamic anodic polarization curves at which meta-stable pitting can occur.24,25) The current transients wererecorded for a duration of 3600 s.

A SEM-EDS was used to observe the pitting sites on thespecimen after taking the potentiodynamic tests in deaerated3.0M NaCl and deaerated 0.5M HCl + 2.5M NaCl solu-tions at 338K.

Table 1 Chemical composition of the experimental alloys (mass%).

AlloyDesignation

C Cr Ni Mo Si Mn S N Fe PREN*

BASE 0.015 24.36 7.09 3.98 0.52 0.82 0.001 0.29 Bal. 46.19

DX5S 0.016 24.74 7.12 3.99 0.50 0.91 0.053 0.28 Bal. 46.30

*PREN (Pitting Resistance Equivalent Number) = mass% Cr + 3.3 © mass% Mo + 30 © mass% N

S.-H. Jeon, S.-T. Kim, J.-S. Lee, I.-S. Lee and Y.-S. Park1618

2.4 Micro-structural characterizationTo observe the optical microstructures of the alloys, the

alloys were ground to 2000 grit using SiC abrasive papers,polished with diamond paste, and then electrolytically etchedusing 10mass% KOH. The chemical compositions of variousinclusions, £-phase and ¡-phase were analyzed using ascanning electron microscope and energy dispersive spectro-scope (SEM-EDS). The nitrogen content was analyzed usinga scanning Auger multi-probes (SAM).

3. Results and Discussion

3.1 Effect of sulfur addition on the distribution ofinclusions

Figure 2 shows the SEM microstructures of the alloyswhich were solution heat treated. It seemed that the globularparticles in black shade are inclusions like sulfides and oxy-sulfides. No secondary phases such as sigma (·) and chi (»)

were observed in the alloys. As S content in the alloyincreased, the number and the areas of inclusions increased(Fig. 2).

Figure 3 presents the effect of the sulfur addition on thedistribution of inclusions per frame area in the alloys. As Scontent in the alloy increased, the number and the total areasof inclusions per frame area increased (Figs. 3(a) and 3(b)).The number of inclusions per frame area in the alloy-DX5Swas increased by 3.1 times, compared with that of the alloy-BASE. As S content in the alloy increased, the number ofinclusions of various sizes ranging from fine inclusions tocoarse inclusions increased (Fig. 3(c)).

Figure 4 shows back-scattered electron (BSE) images ofthe inclusions in the experimental alloys. The chemicalcomposition and morphology for the inclusions wereanalyzed using a SEM-EDS. As presented in Fig. 4(a), theinclusions in the alloy-BASE were composed of the maintype of (Mn, Cr) sulfides and (Mn, Cr, Fe, Al) oxides and the

mμ50 mμ50

: inclusion

(a) (b)

Fig. 2 Back-scattered electron (BSE) images of the inclusions in the alloys: (a) the BASE alloy and (b) the DX5S alloy.

BASE DX5S0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

Th

e n

um

ber

of

Incl

usi

on

s

/Fra

me

Are

a (7

.28x

105 μ m

2 )

Experimental AlloysBASE DX5S

0

50

100

150

200

250

300

350

400

450

500

Experimental Alloys

Th

e ar

ea o

f In

clu

sio

ns

(μm

2 )

/Fra

me

Are

a (7

.28x

105 μm

2 ) (b)(a)

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 160

100

200

300

400

500

Th

e n

um

ber

of

Incl

usi

on

s

/Fra

me

Are

a (7

.28x

105 μ m

2 )

Size of inclusions (μm)

BASE DX5S(c)

Fig. 3 Effects of sulfur addition on the number, the total area and the distribution of inclusions per frame area of the alloys: (a) the numberof inclusions, (b) the total area of inclusions and (c) the distribution of inclusions.

Effects of Sulfur Addition on the Formation of Inclusions and Corrosion Behavior of Super Duplex Stainless Steels 1619

mixed type of (Mn, Cr, Fe) oxy-sulfides and (Mn, Cr, Fe, Al)oxy-sulfides. As presented in Fig. 4(b), the inclusions in theS added alloy-DX5S were composed of the main type of(Mn, Cr) sulfides and (Mn, Cr, Al) oxy-sulfides and themixed type of (Mn, Cr, Fe, Al) sulfides and (Mn, Cr, Fe, Al)oxy-sulfides. The chemical compositions of inclusions in theS added alloy-DX5S were similar to those in the alloy-BASE.

3.2 Effect of sulfur addition on the resistance to pittingcorrosion

Figure 5(a) shows the effect of sulfur addition on thepotentiodynamic polarization behavior of the experimentalalloys in a deaerated 3.0M NaCl solution at 338K accordingto ASTM G 5. In general, the pitting potential (Ep) is definedas the breakdown potential destroying a passive film. As theEp of an alloy increases, the resistance to pitting corrosion ofthe alloy increases. As S content in the experimental alloysincreased, the resistance to pitting corrosion decreased dueto a decrease in the Ep (Table 2). The passivation behaviorappeared unstable with an increase in S content because thecurrent transients of the S added alloys in the passive regionare much more frequent than those of the base alloy.

The reasons that the resistance to pitting corrosion with anincrease in S content decrease are as follows: the deterio-ration in the resistance to pitting corrosion as a result of the S

addition seems to be associated with the inclusions in thealloy. In particular, the pitting potential of the alloy decreasedwith an increase in the area of inclusions per frame area. Thatis, as the interface areas between the inclusion and the matrixin the S added alloy-DX5S increased compared to thosein the alloy-BASE (Fig. 3), the preferential sites for theinitiation of pitting corrosion in alloy-DX5S increasedcompared to those in the alloy-BASE. The interface areabetween the inclusion and the metallic matrix appears toaffect the resistance to pitting corrosion by supplying thepreferential area as a role of pit initiation site. In this way, theresistance to pitting corrosion of the S added alloy wasdecreased.

Figure 5(b) presents the potentiostatic polarization behav-ior (the current transient behavior) for the experimental alloysin a deaerated 3.0M NaCl solution at 338K with an appliedpotential of 0V vs. SCE in the passive region according toASTM G 5. The potentiostatic test was conducted to observethe formation and the repassivation of meta-stable pits. AsS content in the experimental alloys increased, the currentspikes for the formation and the repassivation of the meta-stable pits increased.

The number of the current spikes corresponding to theinitiation of pitting corrosion and the repassivation of themeta-stable pits of the alloy-DX5S was 46 whereas that of the

mμ5

(a)

mμ3 mμ4

(a) (a)

(b)

Type 1

Type of inclusion

Chemical compositions (wt.%)

S Cr Mn Fe Al O

1 43.08 22.03 34.89 - - -2 - 26.18 11.66 6.57 1.35 54.24

3-(a) 47.56 7.16 39.03 1.86 - 4.393-(b) 16.74 5.29 16.99 5.42 1.20 54.36

(a)

(a)

mμ3

(a)

mμ3mμ3

Type 1

Type 3

Type 3

Type 2

Type 2

Type of inclusion

Chemical compositions (wt.%)

S Cr Mn Fe Al O

1 46.33 14.48 39.19 - - -2 9.09 33.11 28.07 - 2.74 26.98

3-(a) 17.80 15.94 27.62 31.73 6.913-(b) 9.47 9.20 8.82 18.22 25.45 28.84

(b)

(a)(b)

Fig. 4 SEM-EDS analysis of inclusions in the alloys: (a) the BASE alloy and (b) the DX5S alloy.

S.-H. Jeon, S.-T. Kim, J.-S. Lee, I.-S. Lee and Y.-S. Park1620

alloy-BASE was 11. It was assumed that there are twodistinct processes before stable pit formation occurs: pitnucleation and growth of the meta-stable pit.26,27) Aspresented in Fig. 5(b), the degree of meta-stable pit growthin the S added alloy-DX5S seemed to increase compared tothose in the alloy-BASE.

Figure 6(a) shows the effect of sulfur addition on thepotentiodynamic polarization behavior of the experimentalalloys in a deaerated 0.5M HCl + 2.5M NaCl at 338Kaccording to ASTM G 5. The resistance to pitting corrosionof the alloy-DX5S was inferior to that of the alloy-BASEbecause the Ep (0.05V vs. SCE) of the alloy-DX5S is lowerthan that Ep (0.18V vs. SCE) of the alloy-BASE and thecritical passive current density of the alloy-DX5S was higherthan that of the alloy-BASE in active region (Table 2).

Figure 6(b) presents the potentiostatic polarization behav-ior (the current transient behavior) for the experimental alloysin a deaerated 0.5M HCl + 2.5M NaCl solution at 338Kwith an applied potential of 0V vs. SCE in the passive regionaccording to ASTM G 5. As S content in the experimentalalloys increased, the current spikes for the formation and therepassivation of the meta-stable pits increased.

The corrosion potential (Ecorr), the pitting potential (Epit),the passive region (¦Ep), the passivation current density (Ip)and the corrosion current density (Icorr) were obtained fromthe anodic polarization curves for experimental alloys. Thevalues of the pitting potential (Epit), the corrosion potential(Ecorr), the passive region (¦Ep), the passivation currentdensity (Ip) and the corrosion current density (Icorr) for theexperimental alloys are listed in Table 2. The Epit, Ecorr and

10-8 10-7 10-6 10-5 10-4 10-3-600

-400

-200

0

200

400

600

800

1000

1200

Po

ten

tial

(m

V v

s. S

CE

)

BASE DX5S

Current Density (A/cm2)

BASE

DX5S

(a)

0 600 1200 1800 2400 3000 3600

0

10

20

30

40

50 BASE DX5S

Cu

rren

t D

ensi

ty (

μA/c

m2 )

Time (s)

DX5S

BASE

(b)

Fig. 5 Effects of sulfur addition on potentiodynamic and potentiostaticpolarization behavior of the alloys in deaerated 3.0M NaCl solution at338K: (a) potentiodynamic polarization behavior and (b) potentiostaticpolarization behavior at an applied potential of 0V vs. SCE.

Table 2 Electrochemical parameters for the experimental alloys in chloride solutions of the different pH obtained from the potentiodynamic anodicpolarization curves.

Alloys SolutionCl¹ concentration

(M)pH

Epit

(mVSCE)Ecorr

(mVSCE)¦Ep

(mV)Ip

(µA/cm2)Icorr

(µA/cm2)

BASENaCl 3.0 7

943 ¹261 1011 1.4 0.28

DX5S 247 ¹270 389 2.0 0.39

BASE NaCl+ HCl

3.0 1183 ¹353 394 1007.3 628.53

DX5S 54 ¹358 245 1528.5 935.87

10-6 10-5 10-4 10-3 10-2 10-1-600

-400

-200

0

200

400

600

800

1000

1200 BASEDX5S

Po

ten

tial

(m

V v

s. S

CE

)

Current Density (A/cm2)

BASE

DX5S

(a)

0 600 1200 1800 2400 3000 36000.0

0.1

0.2

0.3

Cu

rren

t D

ensi

ty (

A/c

m2 )

Time (s)

BASE DX5S

BASE

DX5S(b)

Fig. 6 Effects of sulfur addition on potentiodynamic and potentiostaticpolarization behavior of the alloys in deaerated 0.5M HCl + 2.5M NaClsolution at 338K: (a) potentiodynamic polarization behavior and (b)potentiostatic polarization behavior at an applied potential of 0V vs. SCE.

Effects of Sulfur Addition on the Formation of Inclusions and Corrosion Behavior of Super Duplex Stainless Steels 1621

¦Ep decrease with the increase of the acidity of chloridesolution. The passivation current density (Ip) and corrosioncurrent density (Icorr) increases when the acidity of chloridesolution increases. It can be stated that a higher acidity (lowerpH) imply lower Epit and more difficult repassivation process.As the acidity increased, the intersecting point of the anodicand cathodic curve (Ecorr) shifted to more negative potentialand larger current density. The change in slope of thecathodic curve is due to the increase of hydrogen ionconcentration equivalent to a change in pH of the solution.Acidity of solution normally influences the corrosion behav-ior of iron and its alloys by changing the stability of the oxidefilm, as predicted by the Pourbaix diagram.28)

3.3 Mechanism on the effects of sulfur addition on thepitting or uniform corrosion behavior in chloridesolutions of different pH

Figure 7 presents scattered electron images (SEI) andoptical microstructure in the alloys after the potentiodynamicpolarization test in a deaerated 3.0M NaCl at 338K. Thecorrosion behavior is dominated by the pitting corrosion inNaCl solution. In the case of the 3.0M NaCl solution,chloride ions (Cl¹) had a deleterious effect on the pitting

corrosion. Chloride ions (Cl¹) penetrate the passive filmand attack the exposed bare metal. As S content in theexperimental alloys increased, the sites of pitting corrosionincreased. The pitting corrosion of the base alloy and the Sadded alloy was initiated at the interface areas betweeninclusions and the matrix, and propagated to the inclusions,and then propagated to the ¡-phases, and finally propagatedto the £-phases, based upon the not pitted £-phases whichexisted at the edge of the large pit. It is necessary to verifythe mechanism of propagation of pitting corrosion for highperformance duplex stainless steel.

As presented in Table 3, to clarify the difference in theresistance to pitting corrosion between the £-phase and ¡-phase, the content of Cr and Mo in the £-phase and the ¡-phase in the alloys was quantitatively measured using aSEM-EDS and the N content was measured using a SAM.Then, the pitting resistance equivalent number (PREN)values of the £-phase and ¡-phase by the formula (1) and(2) using a nitrogen factor of 16 and 30 were calculated.When the PREN values are calculated by the formula (1)using a nitrogen factor of 16, the PREN differences(PREN£ ¹ PREN¡) show negative values; thus, the PRENvalues of the ¡-phases are higher than those of the £-phases.

OM OM

SEMSEM

(Step 2) Propagation

of pitting corrosion

(Step 1) Initiation of pitting corrosion

Alloys

OMOM

BASE DX5S

mμ50

γ γ

γ

α

α

γ

γ

γ α

γ

γ γ

α

γ

mμ1

Not pitted

pitted

mμ100

: pitted

Fig. 7 SEM micrograph and optical microstructures of the alloys after the potentiodynamic anodic polarization test in deaerated 3.0MNaCl of pH 7 at 338K.

S.-H. Jeon, S.-T. Kim, J.-S. Lee, I.-S. Lee and Y.-S. Park1622

When the PREN values are calculated by the formula (2)using a nitrogen factor of 30, the PREN differences(PREN£ ¹ PREN¡) show positive values. The PREN valuesof the £-phase and ¡-phase versus temperature for the BASEalloy were calculated using a Thermo-Calc software package(Fig. 8). This computing result indicates that PREN valuesusing factor of 30 are more reasonable to explain thedifference of corrosion resistance between these two phases,which is in good agreement to the present findings shown inFig. 7. Hence, it is reasonable to explain the difference ofpitting corrosion resistance between these two phases intested super duplex stainless steels through the PREN valuescalculated by the formula (2) using a nitrogen factor of 30.

Based upon the calculated PREN£ and PREN¡ valuescalculated using a N factor of 30, the pitting corrosion mustselectively generate at the ¡-phases in a highly concentratedchloride (Cl¹) environment because the PREN value of the £-phase is much larger than that of the ¡-phase, irrespective ofthe experimental alloys. The pitting corrosion was finallypropagated from the ¡-phase to the £-phase.

Figure 9 shows the corrosion morphologies of the alloysafter the potentiodynamic polarization test in deaerated 0.5MHCl + 2.5M NaCl at 338K. The corrosion behavior isdominated by the pitting corrosion and uniform corrosion inNaCl + HCl solution. In the case of the 0.5M HCl + 2.5MNaCl of pH 1, both chloride ions (Cl¹) and hydrogen ions(H+) had deleterious effects on the pitting and uniformcorrosion. Both hydrogen and chloride ions stimulate thedissolution of most metals and alloys, and the entire processaccelerates with time. McCaffertty and Heckermann showedthat in HCl solution, a synergetic interaction exists betweenthe Cl¹ ions chemisorbed on the metallic surface and H+ ionsfrom the solution, which are attracted electrostatically by

Cl¹ ions that cover the metallic surface and catalyze thedissolution reactions.29)

Chloride ions (Cl¹) penetrate the passive film and attackthe exposed bare metal. Hydrogen ions accelerated corrosionprocess due to an increase of acidity. As presented in Table 4,Hydrogen electrode potential (H2/H+) is higher thanelectrode potential of Fe/Fe+2, Cr/Cr+3, Ni/Ni+2 (Table 4).In HCl solution, the corrosion of the experimental alloyproceeds via two partial reactions. The partial anodic reactioninvolves the oxidation of the metal and formation of solubleFe2+ ions according to equation as follows.

Fe ! Fe2þ þ 2e� ð3ÞWhile the partial cathodic reaction involves the evolution

of hydrogen gas.

2Hþ þ 2e� ! H2 ð4ÞIt is generally accepted that in acidic chloride media, the

corrosion rate is mainly controlled by the partial cathodicreaction.30) The overall surface of alloys in 0.5M HCl +2.5M NaCl was strongly corroded due to active hydrogenions (H+). The ferrite phases were uniformly corroded muchmore than the austenite phases due to the difference ofcorrosion resistance between two phases. With the progressof uniform corrosion, the pitting corrosion was initiated atthe interface areas between inclusions and the matrix, andpropagated to the inclusions, and then propagated to the¡-phases, and finally propagated to the £-phases.

In summary, the effects of solution of different pH on thecorrosion behavior of the experimental alloys are schemati-cally presented in Fig. 10. The corrosion behavior isdominated by the pitting corrosion in NaCl solution ofpH 7. In the case of the NaCl solution of pH 7, chloride ions

30

35

40

45

50

55

60

PR

EN

16

1000 1050 1100 1150 1200

Temperature (°C)

ferrite

austenite

30

35

40

45

50

55

60

PR

EN

30

1000 1050 1100 1150 1200

Temperature (°C)

ferrite

austenite

(a) (b)

Fig. 8 PREN values of two phases in the BASE alloy calculated usingThermo-Calc software: (a) nitrogen factor: 16 and (b) nitrogen factor: 30.

Table 4 Standard electromotive force (EMF) series of metals.

Metal-metal ion equilibrium(unit activity)

Electrode potential vs. normal hydrogenelectrode at 25°C, volts

Cu­Cu+2 +0.337

H2­H+ 0.000

Ni­Ni+2 ¹0.250

Fe­Fe+2 ¹0.440

Cr­Cr+3 ¹0.744

Zn­Zn+2 ¹0.763

Na­Na+ ¹2.714

Table 3 PREN values of austenite phases and ferrite phases in the experimental alloys.

AlloyFerrite content

(vol%)Phase

Chemical compositions (mass%)PREN 16*1 PREN 30*2

Cr Mo N

BASE 47Ferrite (¡) 25.80 5.21 0.05 43.79 44.49

Austenite (£) 23.93 3.26 0.51 42.85 49.99

DX5S 48Ferrite (¡) 25.92 5.18 0.05 43.81 44.51

Austenite (£) 24.01 3.28 0.52 43.15 50.43

*1PREN (Pitting Resistance Equivalent Number) 16 = mass% Cr + 3.3 © mass% Mo + 16 © mass% N*2PREN (Pitting Resistance Equivalent Number) 30 = mass% Cr + 3.3 © mass% Mo + 30 © mass% N

Effects of Sulfur Addition on the Formation of Inclusions and Corrosion Behavior of Super Duplex Stainless Steels 1623

(Cl¹) had a deleterious effect on the pitting corrosion.Chloride ions (Cl¹) penetrate the passive film and attack theexposed bare metal. The pitting corrosion was initiated atthe interface areas between inclusions and the matrix, andpropagated to the inclusions, and then propagated to the ¡-phases, and finally propagated to the £-phases.

However, the corrosion behavior is dominated by thepitting corrosion and uniform corrosion in NaCl + HClsolution of pH 1. In the case of the HCl + NaCl of pH 1,both chloride ions (Cl¹) and hydrogen ions (H+) haddeleterious effects on the pitting and uniform corrosion.Hydrogen ions accelerated corrosion process due to increasesof acidity. The overall surface of alloys in HCl + NaClwas strongly corroded due to active hydrogen ions. Theferrite phases were uniformly corroded much more than theaustenite phases due to the difference of corrosion resistancesbetween two phases. With the progress of uniform corrosion,the pitting corrosion was initiated at the interface areasbetween inclusions and the matrix, and propagated to theinclusions, and then propagated to the ¡-phases, and finallypropagated to the £-phases.

The effects of sulfur addition on the formation ofinclusions and resistance to pitting corrosion of the

OMOM

SEMSEM

(Step 2)Uniform corrosion

&propagation

of pitting corrosion

(Step 1) Uniform corrosion

&initiation of pitting corrosion

Alloys

OMOM

BASE DX5S

α

γ

: pitted

γ

α

α γ

αγ

α

γ

α

γγ

γ

mμ50

mμ100

mμ25

mμ4 Not pitted

pitted

Fig. 9 SEM micrograph and optical microstructures of the alloys after the potentiodynamic polarization test in deaerated 0.5MHCl + 2.5M NaCl of pH 1 at 338K.

ααγ αγ γ

Step 1

Step 2

NaCl solution of pH 7

Initial state

HCl + NaCl solution of pH 1

αγ αγ γ

inclusions

αγ αγ γ

Pit propagation from inclusions to ferrite phases

Pitting corrosion at inclusions

αγ αγ γ

Uniform corrosion at two phases

Pitting corrosion at inclusions

αγ αγ γ

Uniform corrosion at two phases Pit propagation from

inclusions to ferrite phases

Fig. 10 Schematic of the effects of pH on the corrosion behavior of thealloys in deaerated chloride solutions of different pH.

S.-H. Jeon, S.-T. Kim, J.-S. Lee, I.-S. Lee and Y.-S. Park1624

experimental alloys are schematically presented in Fig. 11.Compared to the base alloy, as the interface areas between theinclusions and the matrix in the sulfur added alloy increase,the preferential sites for the initiation of pitting corrosion inthe sulfur added alloys also increase. Hence, as sulfur contentincreased, the resistance to pitting corrosion decreased due tothe formation of numerous manganese sulfides deterioratingthe corrosion resistance. The pitting corrosion of the alloywas initiated at the interface areas between inclusions andthe matrix, and then propagated to the ¡-phases, and finallypropagated to the £-phases.

4. Conclusions

To elucidate the effects of sulfur addition on the formationof inclusions and the corrosion behavior of super duplexstainless steels in chloride solutions of different pH and thereasonable nitrogen factor in the PREN equation, a metallo-graphic examination, potentiodynamic & potentiostatic polar-ization tests, scanning electron microscope and energydispersive spectroscope (SEM-EDS) analyses of inclusionand austenite (£) phase and ferrite (¡) phase, a scanningAuger multi-probes (SAM) analysis of two phases, anda thermodynamic calculation using Thermo-Calc softwarewere conducted. These various tests led to the followingconclusions.(1) As sulfur content in the alloys increased, the pitting

corrosion resistance decreased due to the formationof numerous manganese sulfides and manganese oxy-sulfides deteriorating the corrosion resistance and anincrease in the areas of preferential interfaces for theinitiation of the pitting corrosion.

(2) The corrosion behavior is dominated by the pittingcorrosion in neutral chloride solution and by the pittingcorrosion and uniform corrosion in acidic chloridesolution. In the case of the NaCl solution of pH 7,chloride ions (Cl¹) had a deleterious effect on thepitting corrosion. The pitting corrosion was initiated atalmost all types of inclusions of MnS sulfides, MnOSoxy-sulfides and MnO oxides deteriorating the corro-sion resistance, and then propagated to the ¡-phases,and finally propagated to the £-phases. However, in the

case of the HCl + NaCl solution of pH 1, both chlorideions (Cl¹) and hydrogen ions (H+) had deleteriouseffects on the pitting and uniform corrosion. The ferritephases were uniformly corroded more than the austenitephases due to the difference of corrosion resistancesbetween two phases.

(3) The resistance to pitting corrosion of the tested superduplex stainless steel was affected by the preferentialinterface areas between inclusions and the matrix, andthe PREN difference between the £-phase and the ¡-phase for the initiation and propagation of the pittingcorrosion. It is reasonable to explain the difference ofpitting corrosion resistance between these two phases intested super duplex stainless steels through the PRENvalues calculated using a nitrogen factor of 30 in thisstudy.

REFERENCES

1) J. Olson and S. Nordin: Proc. Duplex Stainless Steel ’86, (1986)pp. 219­225.

2) E. Perteneder, J. Tosch, P. Reiterer and G. Rabensteiner: Proc. DuplexStainless Steel ’86, (1986) pp. 48­56.

3) J.-O. Nilsson: Mater. Sci. Technol. 8 (1992) 685­700.4) J. M. Brion, B. Saqnder and G. Pierson: Wear 154 (1992) 225­239.5) X. D. Fang and D. Zhang: Wear 197 (1996) 169­178.6) G. Poulachon, M. Dessoly, J. L. Lebrun, C. Le Calvez, V. Prunet and

I. S. Jawahir: Wear 253 (2002) 339­356.7) A. A. Pereira, L. Boehs and W. L. Guesser: J. Mater. Process. Technol.

179 (2006) 165­171.8) G. Wranglen: Corros. Sci. 14 (1974) 331­349.9) G. Karlberg: Scand. J. Met. 3 (1974) 46­48.10) J. R. Scully, N. D. Budiansky, Y. Tiwary, A. S. Mikhailov and J. L.

Hudson: Corros. Sci. 50 (2008) 316­324.11) P. Schmuki, H. Hildebrand, A. Friedrich and S. Virtanen: Corros. Sci.

47 (2005) 1239­1250.12) T. L. Sudesh, L. Wijesinghe and D. J. Blackwood: Corros. Sci. 49

(2007) 1755­1764.13) M. A. Bakers and J. E. Castle: Corros. Sci. 34 (1993) 667­682.14) A. Pardo, M. C. Merino, A. E. Coy, F. Viejo, R. Arrabal and E.

Martykina: Corros. Sci. 50 (2008) 1796­1806.15) S. Atamert and J. E. King: Mater. Sci. Technol. 8 (1992) 896­911.16) J. C. Lippold, W. A. Baselack and I. Varol: Weld. J. 71 (No. 1) (1992)

1­14.17) I. Varol, J. C. Lippold and W. A. Baeslack, III: Key Eng. Mater. 69

(1992) 217­252.

γα

Initiation of pitting corrosion at (Mn, Cr) S, (Mn, Cr, Fe, Al) S and (Mn, Cr, Al) OS

Propagation of pitting corrosion from α-phase to the γ-phase

α

A few inclusions a few pit initiation

sites high pitting corrosion resistance

A lot of inclusions a lot of pit initiation sites low pitting corrosion resistance

αcathode

γcathode

(anode)

γcathode

γcathode

α(anode)

Corrosion resistance ( γ > α > Mn inclusions )

Mn Inclusion (anode) Mn Inclusion (anode)

Corrosion resistance ( γ > α > Mn inclusions )

Base alloy S added alloy

Fig. 11 Schematic of the effect of sulfur addition on the resistance to pitting corrosion of the alloys.

Effects of Sulfur Addition on the Formation of Inclusions and Corrosion Behavior of Super Duplex Stainless Steels 1625

18) K. Lorenz and G. Medawar: Thyssen Forschung 1 (1969) 97­108.19) S. Bernhardsson: Proc. Duplex Stainless Steel ’91, Vol. 1, (1991)

pp. 185­210.20) S. Herzman, W. Roberts and M. Lindenmo: Proc. Duplex Stainless

Steel ’86, (1986) pp. 257­262.21) M. Barteri, M. G. Mecozzi and I. Nembrini: Proc. Duplex Stainless

Steels 94, Vol. 3, (1994) paper 60.22) U. Heubner and M. Rockel: Werkst. Korros. 37 (1986) 7­12.23) Annual Book of ASTM Standards, ASTM G 5.24) P. C. Pistorius and G. T. Burstein: Corros. Sci. 33 (1992) 1885­1897.

25) G. T. Burstein, P. C. Pistorius and S. P. Mattin: Corros. Sci. 35 (1993)57­62.

26) P. C. Pistorius and G. T. Burstein: Corros. Sci. 36 (1994) 525­538.27) S. T. Kim, S. H. Jeon, I. S. Lee and Y. S. Park: Corros. Sci. 52 (2010)

1897­1904.28) M. Pourbaix: Atlas of Electrochemical Equilibrium Diagram in

Aqueous Solutions, (NACE, Houston, Texas, 1966) paper 499.29) E. McCaffertty and N. Heckermann: J. Electrochem. Soc. 119 (1972)

146­154.30) S. A. M. Refaey: Appl. Surf. Sci. 240 (2005) 396­404.

S.-H. Jeon, S.-T. Kim, J.-S. Lee, I.-S. Lee and Y.-S. Park1626