Вы находитесь на странице: 1из 17

EFFECTS OF SULFUR IN CORROSION

Philippe Marcus and Elie Protopopoff Laboratoire de Physico-Chimie des Surfaces CNRS-ENSCP (UMR 7045) Ecole Nationale Suprieure de Chimie de Paris Universit Pierre et Marie Curie 11, rue Pierre et Marie Curie, F-75005 Paris, France e-mail : philippe-marcus@enscp.fr

ABSTRACT The effects of atomic layers of adsorbed sulfur on corrosion of metals and alloys are reviewed and the conditions and mechanisms of corrosive effects of sulfur are discussed. The different aspects that are reviewed include potential-pH diagrams for adsorbed sulfur, enhanced dissolution, passivity breakdown and localized corrosion, promoted entry of hydrogen in metals. E-pH diagrams for adsorbed S phases are available for different metals (Fe, Ni, Cr, Cu) in water containing H2S, HS-, S2O32-. The E-pH domains where the adsorption of S may take place by electrooxidation or electroreduction of these species are discussed. Adsorbed atomic sulfur layers accelerate the metal dissolution, while remaining adsorbed on the surface. Sulfur present as impurity in the bulk metal accumulates at the surface during the course of the metal dissolution ("anodic segregation"). The anodically segregated sulfur may enhance the dissolution and inhibit the passivation. Similarly, sulfur present in the bulk can accumulate at the metal/passive film interface. This process involves a long incubation time, followed by local breakdown of the passive film. Adsorbed sulfur has a major effect on the H adsorption and H-H combination of hydrogen on metal surfaces. A model including the promoted entry of hydrogen in metals is presented.

INTRODUCTION The detrimental effect of sulfur-containing species on the resistance of iron, nickel and their alloys to localized corrosion is well known. Sulfur is found in gaseous or liquid environments, in the form of, e.g., hydrogen sulfide (H2S), sulfur dioxide (SO2), thiosulfate (S2O32-), as well as in materials themselves, as an atomic impurity dissolved in the bulk, or as sulfide inclusions. The mechanisms of the sulfur action (from all those sources) remained unclear until it was demonstrated that the major role was held by the adsorbed (or chemisorbed) state, where sulfur forms a monoatomic layer chemically bonded to the metal (1). These conclusions were drawn after comparing the effects of pre-adsorbed, well controlled monoatomic layers of sulfur on the metal dissolution and passivation to the effects of sulfur coming from the metal bulk or from a hydrogen sulfide aqueous solution. Here the conditions and mechanisms of sulfur-induced corrosion are discussed, with special emphasis on the effects of monoatomic layers of adsorbed sulfur on metal dissolution and passivation. First the E-pH conditions of adsorption of sulfur on metal surfaces in water are predicted from thermodynamic calculations. Then the basic effects of adsorbed sulfur on anodic dissolution, and on the formation or breakdown of passive films are presented. Examples of sulfur-induced corrosion in different areas of practical importance are given. In addition to their effects on the anodic reactions, sulfur species may also have a detrimental effect on the resistance of metallic materials via the action of sulfur on the H cathodic reactions. Sulfur species are known to promote the reaction of H absorption into the metal bulk (HAR), which may lead to embrittlement of metals, and H-induced cracking. They also modify (enhance or impede, depending on the metal) the rate of the molecular hydrogen evolution reaction (HER), with consequences on the corrosion potential and the corrosion rate. 1

The influence of preadsorbed layers of sulfur on the HER on electrodes of various metals was studied. The effect of adsorbed sulfur on H electroadsorption was studied from voltammograms on platinum electrodes. The knowledge gained from these studies allowed us to discuss and model the mechanisms of promotion of H entry by sulfur species

THERMODYNAMICS OF SULFUR ADSORPTION ON METAL SURFACES IN WATER The principle of potential-pH (Pourbaix) diagrams may be extended to the case of bidimensional layers of elements adsorbed on metal surfaces [2, 3]. It allows us to predict the conditions of stability of adsorbed sulfur on metals immersed in water containing dissolved sulfur species such as S, H2S, HS-, HS2O3-, S2O32-, HSO4-, and SO42-. The calculated E-pH diagrams show that the stability domain of adsorbed sulfur extends beyond the usually predicted range of stability of metal sulfides, and thus adsorbed sulfur layers can exist under conditions in which no bulk sulfide is stable. Potential-pH diagrams have been calculated for sulfur adsorbed on surfaces of iron, nickel and chromium, in water containing sulfides or sulfates (2-4), and in water containing sulfides and thiosulfates (5). They are also available for copper in thiosulfate or sulfite solutions [6].

Principle of E-pH Diagrams for Adsorbed Species An element A (which may be O, S, N, or H) is adsorbed on a metal M in the form of a monoatomic layer, with a valence state equal to zero [denoted Aads(M)]. The adsorption of A from a species dissolved in aqueous solution may be an electro-oxidation or an electro-reduction reaction, depending on the valence state of A in the dissolved species. In contrast to adsorption in ultra high vacuum (UHV) and in gas phase, the adsorption of an atom or a molecule on a metal surface in water involves replacement of adsorbed water molecules [H2Oads(M)] and competition with the adsorption of oxygen [Oads(M)] or hydroxyl [OHads(M)] resulting from the dissociation of H2O molecules on the metal surface. In a Langmuir model for adsorption, it is assumed that the two-dimensional phase is an ideal solution, where water, oxygen or hydroxyl, and sulfur adsorb competitively on the same surface sites and there are no interactions between adsorbed species. Under these conditions the chemical potential of each element A in the phase adsorbed on a metal M can be expressed as follows: Aads(M) = Aads(M) + RT ln A [1]

where A is the relative coverage of the surface of M by adsorbed A (0 A1; A = 1 for the complete monolayer of A); Aads(M) is the standard chemical potential of A, corresponding to saturation of the surface by A. The chemical potential of adsorbed water is given by: H2Oads(M) = H2Oads(M) + RT ln H2O with H2O = (1[2a] [2b]

Ai )

As an example of the method of calculation of the E-pH relations, let us consider the adsorption of oxygen from water on a metal: H2Oads(M) = Oads(M) + 2H+ + 2eThe equilibrium potential of this half-reaction is obtained by applying the Nernst law with the chemical potentials of adsorbed oxygen and water expressed as in Eqs. (1) and (2): E = E + (RT ln10 /2F) log [O/(1S-O)] - (RT ln10/F) pH [3a]

with the standard potential E given on the standard hydrogen scale by : E= [Oads(M) + H+ + e- - H2Oads(M) ]/2F with H+ + e- = H2 (g). Electrochemical experiments and surface analyses show that the adsorbed oxygen species in solution on most transition metals at 25C are likely to be hydroxyls. Thermodynamic data obtained from electrochemical experiments are presently available only for OHads on copper [7]. Therefore Oads is considered here on Fe, Ni and Cr, and OHads on Cu. The standard Gibbs energies of formation (chemical potentials) for sulfur and oxygen adsorbed on metal surfaces can be calculated [2-4] from literature thermodynamic data for reversible chemisorption at the metal-gas interface. E-pH relations for the equilibria between dissolved and adsorbed species In water containing thiosulfates, the sulfur species to be considered are the thermodynamically stable sulfide species H2S(aq), HS-, and the metastable thiosulfate species HS2O3- and S2O32-. The E-pH relations associated with the various equilibria between water, the dissolved sulfur species and adsorbed sulfur and oxygen or hydroxyl adsorbed on Fe, Ni, Cr and Cu are at 25C as follows : Sads(M) + H2O(l) = Oads(M) + H2S(aq) log (O/S) = - pK - log mS pK(Fe) = 18.8; pK(Ni) = 27.4; pK(Cr)= 13.6. [4] [3b]

Sads(M) + H2Oads(M) + H2O(l) = log [2OH /S (1S-OH) ] = - pK - log mS pK(Cu) = 19.7.

2 OHads(M) + H2S(aq) [5]

Sads(M) + H2O(l) = log (O/S) = pH - pK - log mS

Oads(M) + HS- + H+ [6]

pK(Fe) = 25.8; pK(Ni) = 34.4; pK(Cr)= 20.6.

Sads(M) + H2Oads(M) + H2O(l) = log [2OH /S (1S-OH) ] = pK(Cu) = 26.7. H2Oads(M) = Oads(M) + 2H+ + 2e-

2 OHads(M) + HS- + H+ [7]

pH - pK - log mS

E25 = E25 + 0.030 log [O/(1S-O) ] - 0.059 pH

[8]

E25(Fe) = -0.11 V; E25(Ni) = 0.04 V; E25(Cr) H2Oads(M) = OHads(M) + H+ + e-

= -0.52 V.

E25 = E25 + 0.059 log [OH /(1S-OH) ] - 0.059 pH E25(Cu) = 0.095V.

[9]

H2Oads(M) + H2S(aq)

Sads(M) + H2O(l) + 2H+ + 2e-

E25 = E25 - 0.030 log mS + 0.030 log [S/H2O] - 0.059 pH E25(Fe) = -0.67 V; E25(Ni) = -0.77V; E25(Cr) = -0.92 V; E25(Cu) = -0.488 V

[10]

H2Oads(M) + HS- =

Sads(M) + H2O(l) + H+ + 2e-

E25 = E25 - 0.030 log mS + 0.030 log [S/H O] -0.030 pH 2

[11]

E25(Fe) = -0.88 V; E25(Ni) = -0.97 V; E25(Cr) = -1.12 V; E25(Cu) = -0.695V. 2Sads(M) + 5H2O(l) = 2Oads(M) + HS2O3- + 9H+ + 8e-

E25 = E25 + 0.007 log(mS/2) + 0.015 log (O/S) - 0.067 pH E25(Fe) = 0.58 V; E25(Ni) = 0.71 V; E25(Cr) = 0.50 V.

[12]

2Sads(M) + 5H2O(l) = 2OHads(M) + HS2O3- + 7H+ + 6e-

E25 = E25 + 0.010 log(mS/2) + 0.020 log (OH/S) - 0.069 pH E25(Cu) = 0.76V. 2Sads(M) + 5H2O(l) = 2Oads(M) + S2O32- + 10H+ + 8eE25 = E25 + 0.007 log(mS/2) + 0.015 log (O/S) - 0.074 pH E25(Fe) = 0.60 V; E25(Ni) = 0.72 V; E25(Cr) = 0.52 V.

[13]

[14]

2Sads(M) + 5H2O(l) = 2OHads(M) + S2O32- + 8H+ + 6e-

E25 = E25 + 0.010 log(mS/2) + 0.020 log (OH/S) - 0.079 pH E25(Cu) = 0.78 V.

[15]

Potential -pH diagrams The above equations have been used to construct the potential-pH diagrams for sulfur and oxygen or hydroxyl adsorbed in water containing sulfides (H2S or HS- or thiosulfates (HS2O3- or S2O32-) on Fe, Ni, Cr and Cu [5, 6]. The diagrams are shown in Figs. 1 to 4 at 25C for different sulfur coverages (S = 0.01; 0.5; 0.99) and a molality of dissolved sulfur mS= 10-4 mol kg-1. The diagrams are superimposed on the S-M-H2O diagrams (M = Fe, Ni, Cr, Cu), calculated for a molality of dissolved metal mM = 10-6 mol kg-1. These diagrams allow us to predict the E-pH conditions in which sulfur is adsorbed on a surface of Fe, Ni or Cr in water by oxidation of sulfides or reduction of thiosulfates . In aqueous solution, when the potential is increased in the anodic direction, the adsorbed water molecules are replaced by sulfur atoms adsorbed from H2S(aq) and HS. At higher potentials, sulfur is oxidized in HS2O3- or S2O32- and replaced by adsorbed oxygen or hydroxyl. The replacement reaction is completed within a very narrow range of potential ( 0.06 V for Oads and 0.08 V for OHads). The domains of stability of the adsorbed S monolayer (for mS= 10-4 mole kg-1 and mM= 10-6 mole kg-1) overlap the domains of stability of the metals (Fe, Ni, Cr, Cu), the dissolved cations (Fe2+, Ni2+, Cr2+, Cr3+, Cu+ and Cu2+) and the oxides or hydroxides (Fe3O4, Fe2O3, Ni(OH)2, NiO, Cr2O3, Cu2O, CuO). The stability domains of Sads are significantly wider than the stability domains of the bulk metal sulfides, and thus Sads is stable in E-pH regions where there is no stable metal sulfide, which reflects the excess of stability of the chemisorbed state with respect to the corresponding 3D compound. On this basis, detrimental effects of sulfur on the corrosion resistance of metals are predicted even under potential and pH conditions where the metal sulfides are not thermodynamically stable. The comparison of the behaviors of the different metals in the presence of dissolved sulfur species shows that the extent of the overlap of Sads with the stable metals decreases in the sequence Ni, Fe, Cr, so the effect of thiosulfates on corrosion of these metals is expected to decrease in the same order Ni > Fe > Cr. Prediction of S adsorption in the active domains (i. e. during anodic dissolution) is important because this is the condition in which S-enhanced dissolution is experimentally observed. Another effect of thiosulfates expected from the stability of Sads in the passive domains is the blocking or retarding of passivation (or repassivation) of stainless steels. Such diagrams are useful to assess the risk of corrosion of metals and alloys induced by adsorbed sulfur produced by electro-oxidation of sulfides or electro-reduction of thiosulfates.

FUNDAMENTAL ASPECTS OF SULFUR ADSORPTION EFFECTS ON CORROSION The basic concepts and the associated mechanisms underlying the effects of sulfur on the corrosion of metals and alloys are presented here. They have been derived to a large extent from studies performed on well-defined metal and alloy surfaces. In many cases, oriented single crystals have been used to control the structure of the surface on which the corrosion reactions were investigated. Controlled amounts of sulfur in the range of a monolayer were produced prior to the electrochemical measurements, by dosing the surface with gaseous H2S. Surface science techniques, including Auger electron spectroscopy (AES), electron spectroscopy for chemical analysis (ESCA), also called X-ray photoelectron spectroscopy (XPS), low35 energy electron diffraction (LEED), and a more specific radiochemical technique utilizing the S radioisotope, were combined to analyze the surface prior to and after the electrochemical tests, in order to relate the electrochemical and 5

corrosion behavior with the chemical composition and the structure of the surface. This approach has allowed us to determine the mechanisms of the sulfur-assisted corrosion reactions on an atomic scale.

Activation (or Acceleration) of Anodic Dissolution by Adsorbed Sulfur The first studies were conducted on Ni [1, 8, 9] and Ni-Fe [10] single crystals with well-controlled coverages of radioactive sulfur. They clearly demonstrated that adsorbed sulfur accelerates the anodic dissolution of the metal and the alloys. This effect is seen in Figure 5, where the polarization diagrams for a nickel-iron single-crystal alloy [Ni-25Fe(100)(at %)] with and without adsorbed sulfur are shown. Two major findings are apparent from the measurements: (a) a significant increase in anodic current density is observed when the surface is covered by a monolayer of adsorbed sulfur, showing that the dissolution is accelerated by the presence of sulfur, and (b) the surface sulfur is highly stable, since it remains adsorbed during dissolution of the metal. This result was established by the measurements of the sulfur coverage, denoted in Figure 5. This gives evidence of the catalytic nature of the accelerating effect of adsorbed sulfur on the dissolution. In this way, a microscopic amount of adsorbed sulfur can stimulate the dissolution of a significant amount of material (about 1 m of alloy is dissolved in the active region of Figure 5 under the catalytic action of a monoatomic layer of adsorbed sulfur). The catalytic effect of a monolayer of adsorbed sulfur on anodic dissolution of metals and alloys is summarized in the schematic i-E diagram of Figure 6 with, in inset, potential shifts measured with adsorbed sulfur on different metals and alloys. Since the first direct demonstration of the accelerating effect of dissolution by adsorbed sulfur, which was for nickel and nickeliron alloys, the more general nature of this effect has been revealed by similar experiments on nickel alloy 600 [11,12] and on austenitic stainless steels [13]. The acceleration factor R (ratio of the currents at the maximum of the active peak with and without adsorbed sulfur) for nickel-base and iron-base alloys is reported in Table 1. Values in the range 5 to 12 have been observed. This striking effect has been attributed to a weakening of the metal-metal bonds induced by adsorbed sulfur, which leads to a lowering of the activation energy barrier for the passage of a metal atom from the surface to the solution. This mechanism is shown schematically in Figure 7 [14]. The existence of a dipole (M+ S-) associated with the presence of sulfur can also promote the anodic dissolution. TABLE 1 Ratio (R) of the Currents at the Maximum of the Active Peaks With and Without Adsorbed Sulfura Metal or alloy R Reference Ni-Sads/Ni (average on the three low-index faces) 7.7 [1] Ni-25Fe-Sads/Ni-25Fe (average on the three low-index faces) 5.5 [10] 7 [13] Fe-17Cr-13Ni-Sads/Fe- 17Cr-13Ni [(100) face] 12 [11] Ni-15Cr-10Fe-Sads/Ni-15Cr-10Fe a Coverages of the surfaces by preadsorbed sulfur are (in 10 -9g/cm2): 42 on Ni(100), 47 on Ni(111), 44 on Ni(110), 40 on Ni-25Fe (100), 42 on Ni-25Fe(111), 41 on Ni-25Fe (110), 40 on Fe-17Cr-13Ni(100), and 40 on Ni-15Cr-10Fe. The influence of sulfur can be strongly localized if the sulfur is not homogeneously distributed over the surface but adsorbed in specific sites, such as surface defects, where sulfur atoms are more tightly bonded. The major conclusion of the results reviewed in this section is that a monoatomic layer of sulfur can promote the dissolution of macroscopic amounts of material. This indicates a direct link between the atomic-level interactions of adsorbed sulfur atoms with metal surfaces and their macroscopic manifestations in corrosion. Blocking or Retarding Effect of Adsorbed Sulfur on Passivation Effect on the Growth of the Passive Film Another important effect of adsorbed sulfur which was observed first on Ni [1, 8, 9] and later on Ni-Fe alloys [10] is the poisoning of passivation. This effect takes place above a critical coverage of the surface by sulfur which was found to be 0.7-0.8 monolayer. This effect of sulfur is observed in the i-E diagram shown previously in Figure 5, where the passivation potential of the surface covered by a complete monolayer of sulfur ( = 1) is shifted (by 100 mV) to a more anodic value with respect to the sulfur-free surface. In using the 35S radiotracer, it was demonstrated that the desorption (or electrooxidation) of 20% of the full monolayer is necessary to allow the passive film to be formed. Sulfur retards the growth of the passive film by blocking the sites of adsorption of hydroxyl ions, which are the precursors in the formation of the passive layer. X-ray photoelectron spectroscopy (XPS) measurements have revealed that in the presence of adsorbed sulfur the adsorption of OH groups is not totally inhibited, but the OH groups on the surface are more diluted and thus the 6

disproportionation reaction of adjacent OH groups is prevented. A schematic representation of this mechanism is given in Figure 8. In this way the oxide film, which would normally passivate the surface, is not formed. Effect on the Passivation Kinetics Measurements of the passivation kinetics by means of potential steps of surfaces without or with adsorbed sulfur have revealed that the time to complete passivation increases when the surface is covered by adsorbed sulfur. This effect is represented in the i-t diagram of Figure 9. A rapid decrease of the current immediately after the potential step is observed for the clean surface, whereas for the sulfur-covered surface the current density decreases slowly, indicative of a slow process of lateral growth of the film on top of the fraction of a monolayer of adsorbed sulfur. This is consistent with the view that a complete monolayer of sulfur inhibits the nucleation of the oxide, as observed on nickel, whereas a fraction of a monolayer does not inhibit the nucleation but may diminish the density of nucleation sites for the oxide and make the lateral growth of the film more difficult. These two effects can slow down the passivation kinetics. Such kinetics effects have been observed on sulfur-contaminated nickel [14] and nickel alloy 600 [12]. Effect on the Structure and Properties of the Passive Film Below the critical sulfur coverage of 0.7-0.8 monolayer on Ni, the passive film grows on top of the remaining adsorbed sulfur. However, the structure and the properties of the film formed on the sulfur-contaminated surface are modified. A study [1] by reflection high-energy electron diffraction (RHEED) of the structure of the passive films formed on nickel single crystals [Ni(111)] has shown that, in the absence of sulfur, a crystalline film epitaxial with the substrate is formed (a fact which has been confirmed by atomic resolution imaging of the film by scanning tunneling microscopy [15-17]), whereas in the presence of sulfur the epitaxial growth is disrupted and a more defective polycrystalline film is obtained. Accordingly, the current density in the passive state is about four times larger for a film formed with sulfur at the metal-oxide interface. XPS measurements show that the binding energy of sulfur (S 2p) is close to that of sulfur bonded to the metallic surface (~ 162 eV) hence provide evidence that sulfur remains at the metal-oxide interface. Anodic Segregation of Sulfur The concept of anodic segregation was first proposed after the discovery that sulfur present as an impurity in the bulk of Ni and of Ni-25Fe alloys is enriched on the surface during anodic dissolution [1,10,18]. This phenomenon was also observed on Fe-50 Ni [19] and on alloy 600 [11,20]. The experiments were performed on Ni and Ni-Fe alloys which had been doped with sulfur prior to the electrochemical treatments. The introduction of radioactive S into the bulk was performed by exposing the samples to an H2S-H2 gas mixture (with 35S-labeled H2S) at high temperature (1000 to 1200C) under thermodynamic conditions in which sulfur is in solid solution in the metal or the alloy. Rapid quenching to room temperature after the high-temperature treatment avoids sulfide precipitation. The i-E curves recorded on a sulfur-free (i.e., annealed in pure hydrogen) Ni electrode and on Ni containing 50 ppm sulfur are shown in Figure 10. The sulfur-free Ni sample exhibits the normal active-passive transition observed in 0.05 M H2SO4. The sample with bulk sulfur exhibits a completely different behavior: there is no active-passive transition, and the current density increases with increasing potential in the whole range of potentials corresponding to the passive state of sulfur-free Ni. The results of the measurements of the sulfur concentration on the surface, using radioactive sulfur, are shown in Figure 10. They demonstrate that sulfur segregates on the surface during dissolution of the electrode containing bulk sulfur. The measured values of the surface sulfur concentration are denoted Ssegregated in Fig. 10. The values denoted Sexposed in Figure 10 represent the total amount of sulfur made available on the surface by the dissolution of the metal matrix, calculated by integration of the i-E curve. Comparison of the data for Ssegregated and Sexposed reveals that all the sulfur initially present in the bulk remains on the surface during the electrode 2 dissolution, up to a surface concentration of 40 x 10-9 g/cm , which corresponds to a complete monolayer of sulfur on Ni(100). Above this surface coverage, nickel sulfide precipitates and then a thin layer of nickel sulfide is formed. The 2 amount of sulfur in this layer is ~40 x 10-8 g/cm , which corresponds to ~25 of Ni3S2. Sulfur in excess of this amount is dissolved. The formation, before reaching the passivation potential, of a complete monolayer of sulfur and the subsequent growth of a sulfide layer preclude the formation of the passive oxide film. The sulfide layer is not protective and thus high corrosion rates are obtained in the range of potentials in which pure Ni is normally passivated. This phenomenon was called anodic segregation [9,10]. The rate of anodic segregation depends on the sulfur content of the material and on the rate of anodic dissolution. Because of differences in dissolution rates for single-crystal surfaces of different crystallographic orientation, the sensitivity to sulfur increases in the order (111) (100) < (110), which is the order of increasing dissolution rates observed for Ni in 0.05 M H2SO4 [1]. The mechanism of anodic segregation is represented in Figure 11. 7

Passivity Breakdown Caused by Interfacial Sulfur After the formation of the passive film on a bare metal surface, slow dissolution of the metal cations continues, involving dissolution at the passive film surface and transport of ions through the oxide. The question then arises as to where the bulk impurities, e.g., sulfur, go during this process. This question was addressed in a detailed investigation of Ni and Ni-Fe alloys which were doped with radioactive sulfur in order to trace the path taken by the sulfur atoms during the slow dissolution in the passive state [21]. The radiochemical results gave direct evidence that sulfur present in the bulk accumulates at the metal-passive film interface. The enrichment rate was shown to be proportional to the sulfur content in the metal and to the dissolution current density. Above a critical concentration of sulfur at the metal-oxide interface, breakdown of the passive film was observed. This critical concentration of interfacial sulfur was measured with radioactive sulfur and found to be close to one monolayer of sulfur (i.e. ~40 x 10-9 g/cm2). The loss of adherence (or decohesion) at the interface for this critical coverage has been attributed to a weakening of the bonding of the oxide to the substrate caused by sulfur. The nucleation of the sulfide which is expected to take place when the sulfur concentration exceeds a complete monolayer may be responsible for the observed local breakdown of the film and pitting. The defects which are likely to exist at the interface could serve as specific sites for the nucleation of the sulfide, once a quasi-complete monolayer of sulfur has been accumulated at the interface. A schematic representation of the proposed mechanism of the sulfur-induced breakdown of the passive film is shown in Figure 12. It is to be noted that the three following factors are in favor of sulfur remaining at the metal-passive film interface: (a) the sulfur-metal chemical bond is very strong, (b) the solubility of S in nickel oxide (which constitutes the inner part of the passive film on Ni and Ni-Fe alloys) is very low, and (c) the electric field across the passive film, which assists the passage of cations from the metal-passive film interface to the passive film-solution interface, should impede the transport of negatively charged sulfur. The Joint Action of sulfur and chloride Ions In the preceding parts we have seen that S alone can enhance the dissolution, block or retard the growth of the passive film, and cause passivity breakdown and pitting by enrichment at the metal-film interface. Sulfur species dissolved in water may cause directly the breakdown of the film if the metallic sulfide is thermodynamically more stable than the metallic oxide, but this case requires a high concentration of the sulfur species in solution to provide a reaction of the type : MxOy + zH2Saq = MxSz +y H2O + 2(z-y)H+ +2(z-y)e(16)

In most cases, passivity breakdown and pitting requires the presence of chloride ions. The combined action of thiosulfate (S2O32-) and chloride ions on a Fe-17Cr alloy was investigated (22). First, it was demonstrated that thiosulfates are not reduced on the surface of the passive film formed in neutral solution, whereas the reduction to adsorbed sulfur or sulfide does take place on the bare alloy surface. Then it was shown that in a 0.02M chloride solution at pH7 with 30 ppm of thiosulfate, the initial breakdown of the passive film is mostly caused by the chlorides, whereas the thiosulfates play a major role in the subsequent step, the occurence of stable pits. With chlorides alone, unstable pits (i.e., pits that are repassivated) are formed over a wide range of potentials below the classical pitting potential corresponding to the formation of stable pits, whereas with added thiosulfates, the production of adsorbed sulfur or sulfide by reduction on the depassivated surfaces precludes repassivation and causes a marked increase of stable pits at low potential. This effect is shown in Figure 13. The formation of adsorbed sulfur by reduction of thiosulfates on the bare metal surfaces was demonstrated by XPS measurements of the binding energy of the sulfur core level electrons (S2p), which was found to be 169 eV for thiosulfate adsorbed on the passive film surface, and 162 eV after reduction of the thiosulfate on the bare Fe-17%Cr alloy surface. In previous work on the effects of addition of S2O32- on corrosion of stainless steels [23, 24], such a mechanism had been hypothesized, but direct experimental evidence was lacking. To summarize, a likely mechanism for the synergistic effect of chlorides and sulfur is the local breakdown of the film by chlorides, which then permit all the sulfur effects on metal and alloy surfaces, described above, to take place. A major consequence is the stabilization of otherwise unstable (or metastable) pits. This mechanism is consistent with the lower pitting potential and/or shorter incubation time observed experimentally. IMPLICATIONS IN AREAS OF PRACTICAL IMPORTANCE There is much evidence for the damaging effect of sulfur species in a wide range of corrosion-related service failures. The relation between the sulfur-induced corrosion mechanisms presented in the preceding section and the implications in areas of practical importance can be rationalized on the basis of (a) the source of sulfur, (b) the transport process to the metal 8

surface, and (c) the conditions of the reduction (or oxidation) of the sulfur species into the harmful chemical state of sulfur, i.e., adsorbed (chemisorbed) sulfur or metallic sulfide if the concentration of the sulfur species is high. Table 2 summarizes these considerations and gives a few examples of practical areas in which detrimental effects of sulfur have been identified, e.g., oil and gas, pulp and paper industries, power plants (high-temperature water reactors), and atmospheric corrosion. Two categories of sulfur species may be considered according to the origin of the species: (a) sulfur species (molecules or ions) in the environment, e.g., SO2 (gaseous), H2S(gaseous or aqueous), HS-, HS2O3-, S2O32, S4O62-, HSO4-, SO42-, (b) sulfur in the material: sulfur in solid solution, sulfur segregated at the surface or in the grain boundaries, and sulfur in sulfide inclusions. After considering the sources of sulfur, the processes by which the sulfur species are transported to the surface have to be identified. For environmental sulfur species it is generally diffusion in a liquid (e.g., aqueous solution). For sulfur present in the bulk metal or alloy, it may be surface segregation by solid-state diffusion at high temperature, or anodic segregation (a process that has been defined in the preceding section). Higher anodic dissolution rates of grain boundaries, further accelerated by sulfur (according to the mechanism of sulfur-enhanced dissolution described earlier), result in rapid surface enrichment of sulfur at grain boundaries exposed to the electrolyte. In the case of sulfide inclusions, e.g., MnS, attempts to investigate the transport of sulfur from the inclusions to the surrounding surface led to the conclusion that sulfur is cathodically deposited after dissolution of the sulfide, producing rings of sulfur around the sulfide inclusion [25]. Although the detrimental effect of sulfide inclusions on corrosion of steels and stainless steels has been extensively studied [25-32], the exact mechanism is not fully understood. On the basis of the data reviewed here, some authors have invoked the role of adsorbed sulfur to explain the detrimental effects of sulfide inclusions in stainless steels [(25, 30, 31)]. An earlier study of H2S addition on corrosion of stainless steels [32] has also revealed the equivalence of the effects of sulfur added in the form of H2S or present in sulfide inclusions. . The effects of S2O32- observed in the pulp and paper industry [33], and of sulfur species in the steam generators of pressurized water reactors [34] have also been interpreted on the basis on the data presented here. The nature of the surface reaction producing adsorbed sulfur must be considered. Sulfur may be adsorbed by electro-oxidation of sulfides (H2S(aq) or HS-), or by electro-reduction of sulfates (HSO4- or SO42-) , or thiosulfates (HS2O3-, S2O32-). These reactions and the conditions (E, pH) in which they may take place have been detailed in a preceding section.

TABLE 2 Sulfur species that may be present in the environment or the material, with the process of transport to the surface, nature of the surface reaction producing adsorbed sulfur, and examples of areas of practical importance Surface reaction Examples of Transport process to producing adsorbed areas of Sulfur species the surface sulfur practical importance Environmental H2S Diffusion in liquid Oxidative reaction Oil and gas HS Diffusion in liquid Oxidative reaction Diffusion in liquid Reductive reaction Pulp and paper S2O32 Diffusion in the Reductive reaction Atmospheric SO2 atmosphere and in liquid Diffusion in liquid Reductive reaction HSO4 Diffusion in liquid Reductive reaction High-temperature SO42 water Material Sulfur in solid Solid-state diffusion (at high temperature) solution or anodic segregation Sulfur segregated Solid-state diffusion (at high temperature) All areas at grain or anodic segregation boundaries Sulfide inclusions Dissolution and readsorption

Once sulfur is present on the surface in its active chemical state, i.e. the adsorbed state, it has the same effects (which have already been described) irrespective of its origin. The stage at which the process becomes localized is not exactly known. It is probably strongly related to the stage at which surface defects are crucial, which is the case for the surface reactions which produce adsorbed sulfur.

EFFECT OF SULFUR ON THE H CATHODIC REACTIONS

We have undertaken a study of the influence of preadsorbed sulfur monoatomic layers on the H electroadsorption on platinum and the hydrogen evolution reation (HER) on Ni, Fe, Ag and Pt. The metal surfaces were pre-covered with chemisorbed sulfur labeled with radioactive sulfur using the same method as described in the first section. Effect of adsorbed sulfur on H electroadsorption on platinum

There is no desorption of sulfur after immersion of the pre-covered platinum sample and during potential cycling in the hydrogen adsorption region. Partial S coverages were obtained by electrooxidation of the complete monolayer in the anodic region (35). On metals of the Pt group, H is electroadsorbed at equilibrium at potentials anodic to the reversible hydrogen electrode potential at the given pH and at a pressure of 1 atm (RHE1), up to a monolayer. This reaction (called H underpotential deposition) can be studied from cyclic linear sweep voltammetry The hydrogen coverage is measured for each potential by integrating the voltammograms recorded between 0V and 0.4V/RHE1. The experiments were performed on Pt(110) (35) and Pt(111) (36) in 0.05M H2SO4. Fig.14 shows the H desorption voltammograms obtained during the positive going scan on Pt(111) with various S coverages. Integration gives the corresponding electroadsorption isotherms, in absolute coverage or in relative coverage (normalized with respect to the maximum value on the given surface). The absolute isotherm provides the degree of filling of the H adsorption sites as a function of potential, for a given S coverage. The normalized isotherm, restricted to the sites which are not blocked by S; brings information on the energetics of H adsorption in the available sites in the presence of S (Fig.15). From the absolute isotherms, the saturation H coverage Hsat, proportional to the number of available (non-blocked) sites, can be plotted versus the sulfur coverage S (Fig.16). Adsorbed sulfur (Sads) drastically reduces the capacity of the Pt surface to adsorb hydrogen; H adsorption is totally inhibited when the surface is covered by a complete monolayer of Sads;the Hsat versus S curve is not linear, it is best fitted by Hsat = (1-S )7 , which from a statistical model of poisoning of H adsorption sites indicates a deactivation of 7 H adsorption sites in nearest neighbour position with respect to one isolated S atom. Hsat falls to zero for S =0.36, i.e. before complete S saturation (Ssat = 0.5 on Pt(111)). The normalized isotherms of Fig.15 provide information on the effect of Sads on the energetics of H adsorption in the H sites not blocked by Sads. The H adsorption Gibbs free energy from 1/2H2(g) is related to the potential related to RHE1 by the Nernst relation, Gads(H)= -FE. From Fig.15, Gads on the clean Pt surface is increased (is less negative) of about 2kJmol-1 at half-saturation H coverage for a S-covered surface at S=0.33, which indicates that Sads weakens the Pt-Hads bond. The slope of the isotherms at half-H coverage increases with S, which indicates that the repulsive interactions between the adsorbed H atoms are reduced in the presence of Sads . On corroding or passivating metal electrodes, no underpotential H electroadsorption is detected from voltammograms because it is masked by the metal dissolution and oxidation reactions. Effect of adsorbed sulfur on the HER The effects of adsorbed sulfur on the HER on single crystal surfaces of single crystal surfaces of Ni , Fe, Ag and Pt were investigated. a) HER on platinum: 10

The experiments were performed on Pt(111) (37) and Pt(110) (38) in 0.05M H2SO4. Stationary I-E curves were recorded between +0.1 and 0.1V/RHE(1atm). Preadsorbed sulfur is stable in this potential range. On both faces the I-E curves are shifted to higher cathodic overpotentials when the coverage increases, which indicates a strong poisoning effect by adsorbed . The Tafel slope increases from 35mV on the clean surfaces to 44mV for S=0.36 on Pt(111) and 48mV for S =080 on Pt(110). This indicates that the mechanism of the HER at low overpotentials deviates from the mechanism on the clean Pt surface governed by the chemical combination step (Hads +Hads = H2) ,with low H coverage, when S is adsorbed. The decrease of the current density at 0V/RHE1 with S increase can be attributed to a poisoning by S of the sites of the HER. It is to be noted that at a value of S = 0.36 for which the H underpotential electroadsorption is totally inhibited on the same face (see Fig.16), the HER rate at 0V still has a significant value. A careful analysis with a statistical model of poisoning of pairs of H adsorption sites, taking into account the slight increase of the Tafel slope with S indicates that one S atom deactivates three sites of the overpotentially adsorbed H species which is the HER intermediate (39). Fig. 17 compares the blocking ranges of one isolated S atom on the H sites of underpotential H adsorption and HER respectively. Two kinds of sites are possible for the HER, but in any case they are different from the sites of underpotential adsorption.

b)

HER on Ag

Experiments were performed on Ag(110)(40). The S coverage in the preadsorbed monolayer is S =4/5. A partial desorption (~ 30% of a monolayer) takes place between +0.09 and 0.01 V/RHE1. After further polarization up to 1.01 V/RHE1, nearly all S is desorbed, with a residual coverage of 1% of a monolayer. There is nevertheless a catalytic effect of since the HER current is higher on the S covered surface. This effect was attributed to a change of the H adsorption energy or to a change in the orientation of the adsorbed water molecules induced by Sads, as evidenced by measurements of double-layer capacity (40).

c)

HER on Ni

Measurements were performed on the three low-index faces (41). On Ni(100), Sads is stable up to ~-0.5V/RHE1, whereas on the (111) and (110) faces partial S desorption is observed in the same potential range. On Ni(100) covered with a monolayer of Sads (S ~ 0.5), the Tafel slope is unchanged with respect to the clean surface (105mV); the exchange current density is lower by a factor 2; the HER current density measured at a significant cathodic overpotential (-0.38V/RHE) is lower by a factor 2.7. The HER mechanism on Ni is most likely governed by the ion+atom reaction, with S ~1. The current decrease in the presence of Sads can be explained either by an increase of the Ni-Hads bond energy of ~ 4kJ mol-1, or by a blocking effect on the H adsorption sites. d) HER on Fe i-E curves were recorded for the three low-index faces up to 1.13 V/RHE (42). A partial desorption (10-20% of a monolayer) was observed after immersion of the electrodes at open circuit. The remaining coverage was found to be constant over the whole range of cathodic potentials investigated. Three effects were evidenced: i) the HER current is higher (ii) the sensitivity of the HER to the surface structure is greatly lowered (iii) the Tafel slope decreases from 105 mV on clean iron to 90mV on S-covered iron. The exchange current density increases by a factor 1.6 on (110) and (111) and remains constant on (100). The HER current density at 1.03V/RHE (at least 0.12V from the equilibrium potential for all faces) increases by a factor ~2 on (100), 2.5 on (110) and 3.5 on (111). This increase can be explained by a change of the H adsorption energy, increase or decrease, depending on the exact HER mechanism on iron. However, the HER on iron is likely to occur on a surface still covered by an oxide film (43). As Sads blocks sites for O or OH adsorption and inhibits the formation of the passive film, the increase of the HER current could be due to a removal of the passivated layer by Sads. Mechanisms of the promotion of H entry into metals by Sulfur On the basis of the above results, we now discuss the mechanisms by which S species may promote H absorption by metals while poisoning H adsorption. In aqueous medium, the reaction of H atomic absorption (HAR) into an electrode proceeds in parallel to the reaction of H2 gas evolution at the surface (HER) . Classic explanations for the action of H absorption promoters 11

From the many suggestions made in the past to explain the mechanism of action of promoters of H entry into ferrous metals (mainly compounds of elements from columns 15 and 16 in the periodic Table), the more realistic ones are summarized below. It was often considered that the promoter is adsorbed on the surface, although it was not clearly stated if it is under elemental or hydride form. (a) the promoter strengthens the M-Hads bond, thus decreasing the rate of H-H combination and increasing the surface H coverage which thereby increases the permeation rate (44). (b) the promoter weakens the M- Hads bond. This decreases the height of the energy barrier for H surface-bulk transfer (45, 46, 47). (c) the promoter inhibits the combination step of the HER, so that it increases the H entry rate (48, 49, 50, 51). (d) the promoter is adsorbed in the sites of Hads, lowering the probability of finding a pair of H adsorbed at contiguous sites, necessary for H2 formation; it thus poisons the H-H chemical combination step, so a high concentration of atomic H is built up on the surface which increases the rate of H entry into the metal (52, 53). Explanation a) is valid only for a Volmer-Tafel path with H close to unity. Furthermore, the promoters are known to be poisons for adsorption of H in gas phase (54), adsorption of UPD H for Pt (see above), and also adsorption of the HER intermediate (OPD H) (see above). They lower the H adsorption capacity, and are likely to lower the M-Hads bond strength (see above). Explanation (b) was proposed in ref. (45) for a coupled Volmer-Tafel path with small H, and leads to an increase of H entry only for a constant current density, assumed to be proportional to the H-H combination rate. However, a lowering of the rate constant for surface-bulk transfer cannot explain by itself why H permeation is increased under potentiostatic conditions (55). At fixed potential, the weakening of the M-Hads bond strength, associated to a strong reduction of the number of surface adsorption sites should lead to a reduction of the H coverage which would probably be the determining factor. Explanation (c) is often given for an increase of permeation current under gavanostatic conditions, for a HER mechanism assumed to occur only by the Volmer-Tafel path. In this case, the constant cathodic current is related only to the electroadsorption rate, which is equal to the sum of the chemical combination rate plus the H entry rate. If the promoter inhibits the chemical combination step, thus the permeation rate is increased (48, 51). This trivial explanation fails to explain promoter effects occuring under potentiostatic conditions. Explanation (d) is valid if the reaction of consumption of Hads is solely the chemical combination step: in addition to assuming the Volmer-Tafel path for the HER, a lowering of the combination rate leads to an increase of the H coverage only if the H entry rate is negligible. This classic explanation is apparently not self-consistent since it proposes that a high H coverage may appear whereas the H sites are occupied by poison, unless the sites for electroadsorption are supposed to be different from sites for combination, which would be conflicting with the present view. Moreover, three of these explanations involve a lowering of the H-H combination rate by the promoter, whereas according to some authors (56, 51, 52), dissolved H2S enhances the HER on iron. This is consistent with the increase of the HER rate observed on Fe single crystals precovered by a fraction of monolayer of Sads (42). Classic explanations are that in acid medium H2S molecules adsorbed on the cathode catalyse the proton discharge step by being proton-transfer centers (56, 52, 57). We have proposed above that the increase of the HER current in the presence of Sads could be due to a removal of a remaining passivated layer by Sads Modeling of the action of sulfur species on the HER and HAR The observed effects of Sads on the H adsorption and the HER on Pt can be summarized as follows: An electronegative element like sulfur is strongly adsorbed in the potential region of the H reactions, irreversibly compared to Hads. Its coverage is thus independent of potential, while the H coverage is dependent both on potential and on the sulfur coverage. Adsorbed sulfur has two main effects on H adsorption: (i) a short range blocking effect: an adsorbed atom induces a weakening of the M-Hads bond into all nearest-neighbour sites, so drastic that adsorption of H atoms in these sites is prevented; this effect diminishes the H adsorption capacity of the surface, leading to total inhibition for a surface covered by a complete monolayer and (ii) a longer range effect on the energy of the M-Hads bond in the sites left active (not 12

blocked). It was shown that for S chemisorbed on Pt, the latter effect may be a weakening of the metal-H bond under 0V/RHE1, and a strengthening of the metal-H bond in the HER, i.e. over 0V/RHE1 (39). The effects of adsorption site blockers (ASB) like sulfur on the rates of the surface reactions involving adsorbed hydrogen may be modeled simply by taking into account simultaneously but separately the blocking effect which reduces the number of sites for H adsorption and the effect on the M-Hads bond in the sites not blocked, without any a priori on the direction of the latter effect (39, 58, 59). On a surface irreversibly covered by ASB, the H coverage at a given potential, H , defined as the ratio of the density of H sites occupied by adsorbed H atoms to the total density of H sites existing on the clean surface (s) can be expressed as: H(E) =Hsat H(E) (17)

where Hsat is the saturation coverage on the ASB-covered surface, equal to the fraction of H adsorption sites left active, which depends only on the ASB coverage, and H(E) is the "local" coverage in the active sites, equal to the fraction of the sites not blocked which are occupied by H atoms, which depends on the potential and on the energetics of H adsorption in the sites not blocked. The combination step require two H surface sites in nearest-neighbor position, so the rate of this step is proportional to the density of pairs of nearest-neighbour H sites occupied by H atoms, which is 1/2 s H 2 . The fraction of pairs occupied by H atoms, equal to the probability of finding two nearest-neighbour sites occupied by H atoms, H on a clean surface, is replaced for a ASB -covered surface by p H, where p is the fraction of pairs of nearest neighbour sites left active (not blocked by S ). p is the equivalent of Hsat for a two-site reaction and depends only on the ASB coverage. The variations of Hsat and p (p Hsat ) with the coverage, which both reflect the strength of the site blocking effect, are interrelated (39). They depend on (i) the structure and the orientation of the substrate plane, (ii) the nature of the adsorption sites for H and ASB, (iii) the blocking range of the ASB, i.e. the number of H sites blocked by one isolated ASB adatom, and (iv) the degree of order in the overlayer. These variations have been modeled for low-index single crystal faces of fcc metals (39). In the particular case where one ASB atom blocks one H site (geometric blocking effect), Hsat = (1ASB) and p is equal to (1-ASB)2 for random ASB adsorption or ( 1- 2ASB) if the ASB layer is ordered (39). On the basis of this model, it is possible to predict the effects of adsorption site blockers (ASB) as sulfur on the bulk H concentration (39, 58, 59), taking the current assumption that the surface-bulk transfer step is in quasi-equilibrium. Applying the steady state equations for the HER and the HAR to the rate equations modified to include Hsat , H , p described above, the H bulk fractional concentration beneath the surface, XH, is expressed by the following equilibrium equation: XH/(1-XH) H/(1-H) exp -Gdiss/RT exp Gads/RT (18)

where Gdiss and Gads are respectively the Gibbs free energies of H absorption (dissolution) and adsorption from 1/2H2(g). This equation does not involve sat , which means that the reduction of the number of active H adsorption sites has no direct influence on XH and that the dependence of XH on surface effects occurs only via the local coverage H and Gads the adsorption free energy in the sites not blocked. If the Hads intermediate involved in the HAR is also the HER intermediate, (which is assumed for most metals), the expression of the ratio H /(1-H ) and hence the ASB effects on the bulk H concentration depend upon the mechanism of the HER (39, 58, 59); Let us consider the two main cases: a. One of the steps of the HER is in quasi-equilibrium: This situation is the aqueous solution analogue to gas-phase charging from H2. Let us consider the case where the electroadsorption (Volmer) step is in quasi-equilibrium (this mechanism is operating on noble metals at low overpotentials). The following adsorption isotherm is obtained: 13

H / (1-H) exp -Gads/RT exp -FE(RHE1))/RT

(19)

Because both the H electroadsorption and surface-bulk transfer steps are in quasi-equilibrium, the HAR is also in quasiequilibrium. Combining Equs. 18 and 19 leads to the following electroabsorption isotherm: XH /(1- XH) = exp -Gdiss/RT exp FE(RHE1)/RT (20)

XH is dependent only on the overpotential and on Gdiss which is insensitive to the ASB surface effects. In this case, no promoting or inhibiting effects induced by ASB can exist in the steady-state, as shown in Figure 18, where H is to be replaced by H in the presence of ASB This conclusion can be extended to all the HER mechanisms where one step is in quasi-equilibrium b. The electroadsorption and the chemical combination steps are coupled (coupled Volmer-Tafel mechanism): (this mechanism has been reported for iron and ferrous alloys, at low cathodic overpotentials (45). In the steady-state an adsorption isotherm is obtained:

H2 /(1-H ) = kea/kcc aH+ Hsat /p exp-(+2)Gads/RT exp-FE/RT


where kea and kcc are kinetic constants for the two steps.

(21)

Combining this expression with Equ.18 leads to a dependence of XH on Gads and Hsat /p, indicating that ASB surface effects have influence on XH. The curve in Fig.19 where H is to be replaced by H in the presence of ASB shows the variation of XH with Gads , for ==1/2. XH decreases or increases with Gads, according to the value of the local coverage H in the sites not blocked. For low values of H (at low overpotentials), an H entry increase may result from a weakening of the M-Hads bond (decrease of -Gads ) in the sites not blocked, as considered previously (45). In the particular case of non-activated adsorption from H2 ( ~ 1), there is no minimum of XH and the H concentration can only increase if the M-Hads bond is strengthened (increase of -Gads ) by the ASB. The influence of the blocking effects on xH via the parameter Hsat/p is considered now. Modeling of the blocking effects shows that this ratio always increases with the ASB coverage (39), the magnitude of the increase depending on the ASB blocking range and the degree of order in the ASB overlayer. Fig. 20 shows the variation of XH with ASB for the lowest increase of Hsat/p , when one ASB atom blocks only one H site and the ASB overlayer is disordered, so Hsat/p )-1 (39, 59). (More drastic variations are expected in other conditions). It is remarkable that the ASB effect, which =(1-
ASB

reduces the number of H adsorption sites and decreases the combination current leads to an increase of the local coverage H in the sites left active according to Eq. 20 and hence to an increase of the H bulk concentration XH , although the overall coverage H is decreased. In this case, H entry promotion is induced by the reduction of the number of active H adsorption sites (39,55, 58,59), a conclusion close to what was postulated previously (53). As a conclusion, the promoting effects of adsorption site blockers on H entry can be understood not by considering the effective diminution of H the overall coverage of the HER intermediate, but by considering instead the variation of H the local H coverage in the sites not blocked (obtained by normalization of H or from the Tafel and permeation slopes. If the surface-bulk transfer step is in equilibrium, a significant increase of the bulk H concentration can be induced by ASB surface effects only for the HER mechanism where the electroadsorption and chemical combination steps are coupled. This analysis provides a quantitative explanation of the effects of sulfur-induced promotion of H absorption into iron and ferrous alloys, though it is only valid in the limited range of potentials where the coupled mechanism applies. However, the above analysis provides no explanation for sulfur promoting effects on H absorption observed in cases where the HER is not controlled by a coupled Volmer-Tafel mechanism. An alternative explanation of the promoter effects exists for iron and steels. As reported above, the HER is likely to occur not on a bare surface, but instead on a surface 14

still covered by an oxide film (43). The reported increase of the HER current could be due to a removal of the passivating layer by the sulfur species. Similarly, by removing the passivating oxide layer, which is a barrier to H penetration (43, 6062), sulfur species might allow H entry into the bare metal (51), which would be a promotion compared to the entry in the passivated metal.

CONCLUSION The fundamental aspects of sulfur-induced corrosion have been reviewed. The mechanisms have been derived from data obtained on chemically and structurally well-defined surfaces using electrochemical and surface analysis techniques (35S radiotracer and surface spectroscopies). The data obtained show the direct link which exists between atomic-scale surface reactions of sulfur and macroscopic manifestations (enhanced dissolution, passivation blocking or retarding, and passivity breakdown, blocking of H adsorption, change of the HER and H entry promotion). The data provide the fundamental basis required to rationalize the detrimental effects of sulfur species encountered in a large number of service conditions.

REFERENCES J. Oudar and P. Marcus, Appl. Surf. Sci. 3: 48 (1979). P. Marcus and E. Protopopoff, J. Electrochem. Soc. 137:2709 (1990). P. Marcus and E. Protopopoff, J. Electrochem. Soc. 140: 1571 (1993). P. Marcus and E. Protopopoff, J. Electrochem. Soc. 144: 1586 (1997). P. Marcus and E. Protopopoff, Corrosion Science, 39: 1741 (1997). P. Marcus and E. Protopopoff, Corrosion Science, 45: 1191 (2003). V. Maurice, H. H. Strehblow and P. Marcus, Surf. Sci. 458: 185 (2000). P. Marcus, N. Barbouth, and J. Oudar, C.R. Acad. Sci. Paris 280: 1183 (1975). P. Marcus and J. Oudar, Fundamental Aspects of Corrosion Protection by Surface Modification (E. McCafferty, C.R. Clayton, and J. Oudar, eds.), The Electrochemical Society, Pennington, NJ, 1984, p. 173. 10. P. Marcus, A. Teissier, and J. Oudar, Corros. Sci. 24: 259 (1984). 11. P. Combrade, M. Foucault, D. Vanon, P. Marcus, J. M. Grimal, and A. Gelpi, Proceedings of the 4th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems--Water Reactors (D. Cubicciotti, ed.), NACE 1990, section 5, p. 79. 12. D. Costa and P. Marcus, Proceedings of the European Symposium on Modifications of Passive Films (P. Marcus, B. Baroux, and M. Keddam, eds.), The Institute of Materials (EFC 12), 1994, p. 17. 13. A. Elbiache and P. Marcus, Corros. Sci. 33: 261 (1992). 14. P. Marcus, Advances in Localized Corrosion (H. S. Isaacs, U. Bertocci, J. Kruger, and S. Smialowska, eds.), NACE, 1990, p. 289. 15. V. Maurice, H. Talah, and P. Marcus, Surf. Sci. 284: L431 (1993). 16. V. Maurice, H. Talah, and P. Marcus, Surf. Sci. 304: 98 (1994). 17. D Zuili, V. Maurice, and P. Marcus, J. Electrochem. Soc. 147: 1393 (2000). 18. P. Marcus, I. Olefjord, and J. Oudar, Corros. Sci. 24: 269 (1984). 19. P. Marcus and I. Olefjord, Corrosion (NACE) 42: 91 (1986). 20. P. Marcus and J. M. Grimal, Corros. Sci. 31: 377 (1990). 21. P. Marcus and H. Talah, Corros. Sci. 29: 455 (1989). 22. C. Duret-Thual, D. Costa, W. P. Yang, and P. Marcus, Corros. Sci. 39: 913 (1997). 23. R. C. Newman, H. S. Isaacs, and B. Alman, Corrosion (NACE) 38: 261 (1982). 24. R. C. Newman and K. Sieradski, Corros. Sci. 23: 363 (1983). 25. J. E. Castle and R. Ke, Corros. Sci. 30: 409 (1990). 26. Z. Szklarska-Smialowska, Corrosion (NACE) 28: 388 (1972). 27. G. Wranglen, Corros. Sci. 14: 331 (1974). 28. G. S. Eklund, J. Electrochem. Soc. 121: 467 (1974). 29. A. Szummer and M. Janick-Czachor, Br. Corros. J. 9: 216 (1974). 30. J. Stewart and D. E. Williams, Corros. Sci. 33: 457 (1992). 31. M. Janik-Czachor, Modifications of Passive Films (P. Marcus, B. Baroux, and M. Keddam, eds.), The Institute of Materials (EFC 12), 1994, p. 280. 32. J. L. Crolet, L. Seraphin, and R. Tricot, Metaux Corrosion Industries, no. 616 (1976). 15 1. 2. 3. 4. 5. 6. 7. 8. 9.

33. 34.

35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59.

60. 61. 62.

R. C. Newman, Corrosion (NACE) 41: 450 (1985). P. Combrade, O. Cayla, M. Foucault, D. Vanon, A. Gelpi, and G. Slama, Proceedings of the 3rd International Symposium on Environmental Degradation of Materials in Nuclear Power Systems--Water Reactors, Traverse City, (G.J. Theus and J.R. Weeks, eds.), The Metallurgical Society, 1987, p. 525. P. Marcus and E. Protopopoff, Surface Sci., 161, 533 (1985). E. Protopopoff and P. Marcus, Surface Sci., 169, L.237 (1986). E. Protopopoff and P. Marcus, J. Vac. Sci. Technol. A, 5 , 944 (1987). J. Electrochem. Soc., 135, 3073 (1988). E.Protopopoff and P. Marcus, J. Chim. Phys., 88, 1423 (1991). C. Nguyen Van Huong, R. Parsons, P. Marcus, S. Montes and J. Oudar, J. Electroanal. Chem. 119 (1981) 137. P. Marcus and J. Oudar, C. R. Acad. Sci. Paris 284 (1977) 959. P. Marcus, S. Montes and J. Oudar, in Atomistics of Fracture, ed. R. M. Latanision and J.R. Pickens (Plenum, N.Y., 1983) p. 909. A. M. Brass and J. R. Collet-Lacoste, Acta Mater. 46 (1998) 869. K. E. Shuler and K. J. Laidler, J. Chem. Phys. 17 (1949) 212. . J. O' M. Bockris, J. Mc Breen, L. Nanis, J. Electrochem. Soc. 112 (1965) 1025. I. A. Bagotskaya, Zh. Fiz. Khim. 36: 2667 (1962). J. F. Newman and L. L. Shreir, Corrosion Science, 9: 631 (1969). E. G. Dafft, K. Bohnenkamp, H. J. Engell, Corros. Sci. 19: 591 (1979). M. Enyo, In: B.E.Conway, J. O'M. Bockris, E. Yeager, S. U. M. Khan, R. E. White eds, Comprehensive Treatise Of Electrochemistry, Vol.7, New York: Plenum Press, 1983, pp 241-300. T. P. Radhakrishnan and L. L. Shreir, Electrochimica Acta, 11:1007 (1966). B. J. Berkowitz and H.H. Horowitz, J. Electrochem. Soc. 129: 468 (1982). R. N. Iyer, I. Takeuchi, M. Zamanzadeh and H. W. Pickering, Corrosion, 46: 460 (1990). U. R. Evans, The Corrosion and Oxidation Of Metals, London: Edward Arnold Publ. (1961), p. 397. M. P. Kiskinova, Surface Sci. Rep. 8, n8 (1988). P. Marcus and E. Protopopoff, In: P. Azou, N. Chen eds. Proc. 4th. Int. Conf. on Hydrogen and Materials, Beijing,China, 9-13 may 1988, t. I, p. 168. T. Zakroczymski, in: R. A. Oriani, J. P. Hirth, M. Smialowski, eds., Hydrogen Degradation of Ferrous Alloys, Noyes Publications, Park Ridge, NJ, 1985, pp. 215. A. Kawashima, K. Hashimoto, and S. Shimodaira, Corrosion 32: 321 (1976). P. Marcus and E. Protopopoff, C.R. Acad. Sci. Paris, 308, II: 1127 (1989). E. Protopopoff and P. Marcus, in B. E. Conway, G. Jerkiewicz, eds., Electrochemistry and Materials Science of Cathodic Hydrogen Absorption and Adsorption, PV 94-21, The Electrochemical Society, Pennington, NJ (1995) p. 374. A. M. Brass, Ann. Chim. Fr. 14: 273-300 (1989). A. M. Brass, J. Collet-Lacoste, in: A. Turnbull ed., Hydrogen Transport and Cracking in Metals, J. Inst. Mat. 142 (1995). A. M. Brass, J. Collet-Lacoste, M. Garet, J. Gonzalez, in: La Revue de Mtallurgie-CIT/Science et Gnie des Matriaux, Fvrier 98: 197.

16

17

Вам также может понравиться