Marcus P. Corrosion mechanisms in theory and practice

Подождите немного. Документ загружается.

environment in the large pits formed on the Mo-free alloy changes, on the other

hand, noticeably because the production of the solvated H

ions is high, the

drop in pH at the pit surface becomes more pronounced, and the ingress migra-

tion of Cl

–

ions is more effective. For the N-bearing steels the lowering of pH

due to hydrolysis is compensated by formation of NH

, as suggested by

Osozawa and Okato [26]. The pH value of the solution in the pits will be above

a critical value and thereby repassivation will occur in an early stage of the

pitting process. Hence, this model [31] concludes that the role of Mo is to lower

the dissolution rate of the initiated pits and thereby limit the increase in acidity

in the pitted volume.

It seems therefore that Mo plays a multifold role in corrosion protection of

stainless steels: (a) it lower the chloride content in the passive film during passivation

and repassivation and thereby makes the film more resistant to breakdown; (b) it

lowers the dissolution rate of the alloy during initiation of pitting; (c) it forms

insoluble salt at the bottom of pits, which is favorable to the formation of passive

film; and (d) during the repassivation process it forms molybdate, which governs the

deprotonation reaction of the poorly protective hydroxide and thereby a more

protective barrier layer is formed. According to the thick-membrane studies [17],

molybdate has been shown to promote ionic rectification, which retards chloride

ingress and promotes deprotonation of the hydroxide. The synergistic effect of N,

Ni, and Mo can be understood from the work of Jack [46], who reported that

whereas single nitrides of Mo and Ni are thermodynamically relatively unstable, with

238 Clayton and Olefjord

Figure 16 A model for the growth and repassivation of pits in Mo-free, Mo and Mo-N

containing alloys. (From Ref. 15.)

Ni being the less stable of the two, a mixed Mo and Ni nitride of the form Ni

exhibited very high thermodynamic stability. Although this suggests the likely

involvement of Ni in such mixed surface nitrides formed during anodic segregation

of N-bearing steels, it also suggests strong intermetallic bonding. This is not contrary

to the nature of bonding in interstitial nitrides formed by transition elements. For

example, although N is the electron acceptor in the mixed nitride phase, one would

expect strong interaction between Mo and Ni and somewhat weaker interaction

between Cr and Ni.

Finally, it is evident that less than one monolayer of surface nitride is capa-

ble of increasing the pitting resistance of 304 stainless steel [33] and therefore it

is probable that this mixed nitride phase is situated at previously high-energy sites

such as grain boundaries and surface dislocations, which are preferential sites for

both pitting and surface nitride formation. In the case of Ni-bearing steels these

sites would be initially the most active and therefore the sites where anodic segre-

gation would tend to be first developed. Therefore, as a consequence of anodic

segregation at high-energy sites during the active stage of repassivation, a more

homogeneous passive film may be generated that would be less vulnerable to

localized breakdown.

SUMMARY

The passivity of austenitic stainless steels is a field of study that will continue to

develop as new service environments are found for this important class of steels.

Many of the studies reviewed here provide a basis for new testing and analytical

strategies that must be applied to these and similar materials to develop new stainless

steels for such new markets as the bioprocessing industry and facilities for

atmospheric scrubbing of industrial emmisions. In this chapter we have emphasized

the importance of barrier oxide layers, salt deposit layers, and alloy surface layers in

the overall process of passivity and breakdown of passivity. Finally, it has been shown

that by combining electrochemical and surface analytical methodologies, both

corrosion parameters and structure-property relationships can be combined to model

passivation processes in a way that provides new insights into alloy performance and

new alloy design concepts.

ACKNOWLEDGMENTS

This work has been supported by the National Swedish Board for Technical and

Industrial Development (NUTEK), U.S.O.N.R. (A. J. Sedriks, Contract Officer)

contract N0001485K0437, and the U.S.N.S.F. International Program Division

(Christine Glenday, Contract Officer).

REFERENCES

1. I. Olefjord, B. Brox, and U. Jelvestam, J. Electrochem. Soc. 132:2854 (1985).

2. J. Horvath and H. H. Uhlig, J. Electrochem. Soc. 115:791 (1968).

Passivity of Austenitic Stainless Steels 239

3. C. R. Clayton and Y. C. Lu. J. Electrochem. Soc. 133:2465 (1986).

4. G. Okamoto, Corros. Sci. 13:471 (1973).

5. H. Saito, T. Shibata, and G. Okamoto, Corros. Sci. 19:693 (1979).

6. K. Tachibana, K. Miya, K. Furuya, and G. Okamoto, Corros. Sci. 31:527 (1990).

7. A. R. Brooks, C. R. Clayton, K. Doss, and Y. C. Lu, J. Electrochem. Soc. 133:2459

(1986).

8. A. J. Sedriks, Corrosion of Stainless Steels, Wiley, New York, 1979.

9. K. Sugimoto and Y. Sawada, Corros. Sci. 17:425 (1979).

10. K. Hashimoto, K. Asami, and K. Teramoto, Corros. Sci. 19:3 (1970).

11. H. Ogawa, H. Omata, I. Itoh, and H. Okada, Corrosion 34:53 (1978).

12. M. Urquidi and D. D. MacDonald, J. Electrochem. Soc. 132:533 (1985).

13. L. Wegrelius and I. Olefjord, Mater. Sci. Forum 185–188:347 (1995).

14. L. Wegrelius, F. Falkenberg, and I. Olefjord, J. Electrochem. Soc. 146:1397 (1999).

15. F. Falkenberg and I. Olefjord, Proceedings, International Symposium in Honor of

Professor Norio Sato: Passivity and Localized Corrosion (Eds. M. Seo, B.

MacDougall, H. Takahashi and R. G. Kelly), Electrochemical Society Pennington

New Jersey 404 (1999).

16. F. Falkenberg, Metastable pits on austenitic stainless steels, thesis for the

degree Licentiate of Engineering, Chalmers University of Technology,

Gothenburg, 1998.

17. M. Sakashita and N. Sato, Corros. Sci. 17:473 (1977).

18. C. R. Clayton and Y. C. Lu, Corros. Sci. 29:881 (1989).

19. Y. C. Lu, C. R. Clayton, and A. R. Brooks, Corros. Sci. 29:863 (1989).

20. J. R. Ambrose, Passivity of Metals (R. P. Frankenthal and J. Kruger, eds.), Electrochemical

Society, Pennington, NJ, 1978, p. 740.

21. A. Schneider, D. Kuron, S. Hofmann, and R. Kircheim, Proceedings, Sixth International

Symposium Passivity, PART I (N. Sato and K. Hashimoto, eds.), Corros. Sci. 31:191

(1990).

22. F. Falkenberg and I. Olefjord, Proceedings, Eight International Symposium on

Passivity, Jasper, Alberta, Canada, 1999 (in press).

23. G. P. Halada, C. R. Clayton, H. Herman, S. Sampath, and R. Tiwari, J. Electrochem.

Soc. 142:74 (1995).

24. Y. Hisamatsu, Passivity and Its Breakdown on Iron Based Alloys., USA–Japan Seminar

(R. Staehle and H. Okada, eds.), National Association of Corrosion Engineers, Houston,

1976, p. 99.

25. Z. Szklaska-Smialowska, Pitting Corrosion of Metals, National Association of Corrosion

Engineers, Houston, 1986, p. 358.

26. K. Osozawa and N. Okato, Passivity and Its Breakdown on Iron and Iron Based

Alloys, USA–Japan Seminar, National Association of Corrosion Engineers, Houston,

1976, p. 135.

27. J. E. Truman, M. J. Coleman, and K. R. Pirt, Br. Corros. J. 12:236 (1977).

28. J. Eckenrod and C. W. Kovack, ASTMSTP 679, Philadelphia, 1977, p. 17.

29. R. Bandy and D. van Rooyen, Corrosion 39:227 (1983).

30. Y. C. Lu, and R. Bandy, C. R. Clayton, and R. C. Newman, J. Electrochem. Soc.

130:1774 (1983).

31. I. Olefjord and L. Wegrelius, Corros. Sci. 38:1203 (1996).

32. I. Olefjord, Surface Characterization (E. Brune, R. Hellborg, H. J. Whitlow, and

O. Hunderi, eds.), Wiley-VCH, Weinheim, 1997, p. 291.

33. R. D. Willenbruch, C. R. Clayton, M. Oversluizen, D. Kim, and Y. C. Lu, Proceedings,

Sixth International Symposium on Passivity, PART 1 (N. Sato and K. Hashimoto, eds.),

Corros. Sci. 31:179 (1990).

34. J. Kolts, J. B. C. Wu, and A. I. Asphahani, Met. Prog. September:25 (1983).

240 Clayton and Olefjord

35. C. R. Clayton, G. P. Halada, and J. R. Kearns, Proceedings, US-Japan Seminar,

Timberline Lodge, Mount Hood, OR, June 1994 (R. Latanision and K. Hashimoto,

eds.), Mater. Sci. Eng. A 198:135 (1995).

36. Y. C. Lu, M. B. Ives, C. R. Clayton, and D. Kim, Corros. Sci. 35:89 (1993).

37. G. P. Halada and C. R. Clayton, J. Va t. Sci. Technol. A11:2342 (1993).

38. I. Olefjord, Mater. Sci. Eng. 42:161 (1980).

39. I. Olefjord and B.-O. Elfström, Corrosion (Houston) 38:46 (1982).

40. B. Brox and I. Olefjord, Proceedings, Stainless Steel 84, Institute of Metals, London,

1985, p. 134.

41. R. C. Newman, Corros. Sci. 25:331 (1985).

42. B. E. Clark, S. J. Thorpe, and K. T. Aust, Corros. Sci. 31:551 (1990).

43. L. Brewer, Science 161:115 (1968).

44. S. Börjesson and I. Olefjord, private communication.

45. I. Olefjord and C. R. Clayton, ISIJ Int. 31(2):134 (1991).

46. K. H. Jack, High Nitrogen Steels (J. Fact and A. Henry, eds.), Institute of Metals,

London, 1989, p. 117.

Passivity of Austenitic Stainless Steels 241

Mechanisms of Pitting Corrosion

Hans-Henning Strehblow

Heinrich-Heine-Universität Düsseldorf, Düsseldorf, Germany

INTRODUCTION

Pitting corrosion occurs at passivated metal surfaces during the access of so-called

aggressive anions. Halides very effectively attack passivating thin oxide layers,

leading to an intense localized dissolution of the metal surface, which is otherwise

protected by the passive layer against general dissolution. Chlorides cause the most

serious problems due to their presence in many environments such as seawater and

salt on roads, in food, and in the chemical industry. Many metals and their alloys

are subject to this type of corrosion (e.g., iron, nickel, copper, aluminium, steels),

whereas chromium is one of the few exceptions that resists pitting in aggressive

environments. The restriction of the dissolution to pits within a large passivated

metal surface, which may serve as a large cathode for the reduction of oxidants

such as dissolved oxygen, leads to fast perforation of the metal, which weakens the

construction and thus causes large economic losses and safety problems.

Pitting at passivated metal surfaces is a complex process with a sequence of

steps. For each stage of the development and growth of a corrosion pit, one has to

study the mechanistic details in order to understand the process as a whole. For a

detailed mechanistic discussion the following stages are usually distinguished:

1. Processes leading to breakdown of passivity

2. Early stages of pit growth

3. Late stages of pit growth

4. Repassivation phenomena

The details of the mechanisms also depend on the metal or the composition of

the alloys as well as on the electrolyte and other environmental conditions. Metallic

and nonmetallic inclusions often play a decisive role in the start of a corrosion pit.

In most cases the presence of aggressive anions is necessary for breakdown of

passivity and stable pit growth. The discussion in this chapter explains the effect of

these anions by their tendency to form complexes with metal ions. It concentrates

on the behavior of some pure metals, such as pure Fe and Ni, in simple electrolyte

solutions. Some basic concepts are the center of interest, although it is known from

the technical applications of the different materials that the appearance and the

243

details of the corrosion mechanisms may involve many complicating factors.

Thus, this contribution intends to describe the main mechanistic aspects rather

than the details of complicating factors.

Some Basic Details of Passivity

In order to get a better basis for the discussion and understanding of the corrosion

mechanisms, a brief summary of the structure and the properties of passive layers is

given. For more detailed information one should refer to the chapter on passivity in

this book. Passive layers form on many reactive metals. If these films have semicon-

ducting properties, as for Fe, Cr, Ni, and Cu, they will grow only up to a few nanome-

ters in thickness to the potential for oxygen evolution. Valve metals such as Al, Nb, Ta,

Ti, Zr, and Hf form insulating oxide films that may grow up to more than 100 V with-

out oxygen evolution with thicknesses reaching several tens of nanometers. These

films generally show a potential drop at the metal-oxide and the oxide-electrolyte

interface as well as within the passive layer (see Fig. 2a in the section Breakdown of

Passivity). The electrical field strength within the passive layer of some nanometers is

in the order of some 10

V/cm, which enables the migration of ions through the film

at room temperature at a measurable level in a region of corresponding current densi-

ties of some few μA/cm

or less. According to detailed studies using surface analyti-

cal methods such as X-ray photoelectron spectroscopy (XPS), the passive film has a

complicated chemical structure that is described by a multilayer or at least by a bilay-

er model. Oxides and hydroxides are present, sometimes separated in different layers,

sometimes as hydroxy-oxides. The size and structure of the related sublayers change

with the electrode potential and with other factors such as the composition of the bulk

metal and the electrolyte. At least one part of the passive layer acts as a barrier to cor-

rosion, to which most of the details of the following discussion refer.

At appropriate pH the oxide films may be in equilibrium with the electrolyte

according to the data of pH-potential diagrams of M. Pourbaix. In strongly acidic

electrolytes the passive layer on metals such as Fe, Ni, Cr, and steels is far from

its dissolution equilibrium, and passivity for these conditions is a consequence of

the slow dissolution kinetics of the oxide. As this situation is not related to thermo-

dynamics of layer formation, it is in contradiction to the results of Pourbaix

diagrams. In other words, Pourbaix diagrams are not applicable to these situations.

The phase diagram of Figure 2a shows the reactions occurring on a passive metal

which are influenced by the potential drops E

1,2

at the metal-oxide and E

2,3

at the

oxide- electrolyte interface and Δφ within the oxide layer. E

2,3

influences the

kinetics of the dissolution reaction (1) of the metal ions and the related passive

corrosion current density, as well as the kinetics of O

2–

or OH

–

formation of reac-

tion (2) corresponding to the growth of the layer. Both reactions require in addition

a field-assisted ion transport through the passive layer. Last but not least, electron

transfer occurs as a third reaction (3) that is of interest for corrosion reactions

under open-circuit conditions, where the reduction of oxidants as dissolved

oxygen serves as a cathodic compensation for the anodic corrosion process.

Redox reactions, of course, do not occur on insulating passive layers. For techni-

cally applied valve metals with insulating passive layers, inclusions with high

electronic conductivity may provide sufficient electronic transfer from the passivated

244 Strehblow

metal to a redox system within the electrolyte. They may serve as cathodic sites within

the passivating oxide film on the metal, which enables a slow corrosion rate in the passive

state or even the intense metal dissolution during localized corrosion. Copper inclusions

in aluminum are a classical example. Here the differences in the electronic properties at the

surface have been visualized with the scanning Kelvin probe, which tests the change of

the work function with the material at the metal surface with lateral resolution [1].

A high electrical field is established at least in the part of a more complicated

layer structure that serves as a barrier for metal dissolution. This in turn leads to

logarithmic or inverse logarithmic growth of the film thickness and linear decay

of the logarithm of the current density with time for potentiostatic passivation

transients. The transfer of metal and O

2–

ions through the passive layer increases

exponentially with the electrical field strength. The transfer of aggressive anions

through the film according to reaction (4) (Fig. 2a) is controlled similarly by the

electrical field strength, which is important for the discussion of the penetration

mechanism for breakdown of passivity. The potential drop E

2,3

influences the

adsorption of anions including aggressive anions, the formation of O

2–

and OH

–

anions from the electrolyte, and the transfer of metal cations at the oxide-electrolyte

interface. It thus rules layer formation and the passive corrosion current density.

2,3

changes when the composition of the oxide changes at the surface in contact

with the electrolyte, especially when approaching the transpassive potential range

and intermediately for nonstationary conditions when the electrode potential is

altered. It also changes with the pH via reaction 2 of Figure 2a. The electrochemical

equilibrium for this reaction leads to a –0.059 V/pH dependence of the potential

drop E

2,3

at the oxide-electrolyte interface and thus to a decrease of the passive

current density with increasing pH. The transfer of metal ions into the electrolyte

is also influenced by complexing anions, which will be discussed in detail in the

chapter on the adsorption mechanism of pit nucleation.

Pitting Potentials and Inhibition Potentials

Pitting occurs when the potential exceeds a critical value E

in the passive range

of a polarization curve. E

decreases with the logarithm of the concentration of the

aggressive anions [A] according to Eq. (1). The presence of inhibiting anions I

may shift it to more positive values. An interesting observation is the inhibition

potential E

for the additional presence of an inhibitor that causes an upper limit

to pitting in the passive potential range. E

increases with the logarithm of the

concentration of the inhibitor [I] and decreases with that of the aggressive anion

[A] according to Eq. (2) [2].

Mechanisms of Pitting Corrosion 245

Nitrate and perchlorate may act as inhibitors. Figure 1 gives an example of

the dependence of E

and E

on the electrolyte composition for Ni in chloride- and

nitrate-containing solutions. The gap between the two lines is the potential range

of pitting corrosion. Similar results have been obtained for ClO

–

and Fe [2]. A very

peculiar property is observed for Fe in ClO

–

-containing solutions. Although

this anion is an inhibitor for chloride pitting at more negative potentials, it also

causes pitting of Fe by itself with a very positive critical potential [2,3]. Its E

follows

Eq. (1), is pH independent, and may be measured with an accuracy of some few mV.

Usually critical potentials may be measured with an uncertainty of several 10mV

only. Surface analytical studies have shown that the aggressive properties of

perchlorate are related to its decomposition close to the pitted surface, which yields

aggressive chloride [3]. This is very special as one expects a ClO

–

reduction at more

negative potentials only where it is kinetically stable. As a possible explanation, the

decomposition of ClO

–

in the increased electrical field of the oxide-electrolyte

interface has been proposed when approaching the transpassive potential range [3].

BREAKDOWN OF PASSIVITY

Three main mechanisms are discussed by most authors for the processes leading to

breakdown of passivity: the penetration mechanism, the film breaking mechanism,

and the adsorption mechanism [4]. Figure 2a to c presents diagrams for their

explanation. The penetration mechanism (Fig. 2a), first discussed by Hoar et al. [5],

involves the transfer of anions through the oxide film to the metal surface, where they

start their specific action. The film breaking mechanism (Fig. 2b), proposed by Vetter

and Strehblow [6] and Sato et al. [7,8], requires breaks within the film that give direct

access of the anions to the unprotected metal surface. The adsorption mechanism,

discussed first by Kolotyrkin [9] and Hoar and Jacob [10], starts with adsorption

of aggressive anions at the oxide surface, which enhances catalytically the

246 Strehblow

Figure 1 Pitting potential E

and inhibition potential E

as a function of the chloride

concentration and the concentration ratio of chloride to nitrate in phthalate buffer, pH 5.0:

(U) E

, (S) E

, 1 M KNO

; () E

,() E

, 0.1 M KNO

; (z) galvanostatic experiments,

1 M KNO

. (From Ref. 2.)

Mechanisms of Pitting Corrosion 247

Figure 2 Phase diagram of a passive metal demonstrating the processes leading to pit

nucleation. (a) Penetration mechanism and phase diagram of a passive layer with the related

processes of ion and electron transfer within the oxide and at its phase boundaries including

schematic potential diagram (φ). (b) Film breaking mechanism and related competing

processes, (c) Adsorption mechanism with increased local transfer of metal ions and related

corrosion current density i

caused by complexing aggressive anions leading to thinning

of the passive layer and increases in field strength and final free corrosion current density

c,h

within the pit. (From Ref. 4.)

transfer of metal cations from the oxide to the electrolyte (Fig. 2c). This effect

leads to thinning of the passive layer with possible final total removal and the start

of intense localized dissolution. Strict separation of these mechanisms might not

always be appropriate, as penetration of a passive layer via some very small defect

is not necessarily very different from the occurrence of fissures in the film that

permit easy access of the aggressive anions to the metal surface as in the film

breaking mechanisms. In the following discussion of the mechanism of breakdown

of passivity, a distinction is proposed between stationary and nonstationary

conditions of the passive layer. These different experimental or environmental

conditions may have a strong influence on the effective mechanism leading to

breakdown of passivity.

Penetration Mechanism

The penetration mechanism (Fig. 2a) requires transfer of the aggressive anions

through the passive layer to the metal-oxide interface, where they cause further

specific action. The high electrical field strength and a high defect concentration

within the presumably severely disordered structure of the passivating oxide layer

may explain this transfer. Many authors even postulate an amorphous structure of

the oxide layer [11]. Direct evidence is given by a few extended X-ray absorption

fine structure (EXAFS) measurements [12]. Systematic studies with this method

are, however, difficult and still missing. For a highly disordered structure it is not

clear why the breakdown event occurs locally.

248 Strehblow

Figure 2 (Continued)