Acta Chim. Slov. 2003, 50, 287-299. 287 ON THE ROLE OF TUNGSTEN OR MOLYBDENUM IN IMPROVED PASSIVATION TO AUSTENITIC STAINLESS STEEL IN PHOPSPHORIC ACID POLLUTED BY SULFIDES Abdellah Guenbour," Mohammed Essahli," Ali Benbachir," Lucien Aries,6 and Rachid Boulifc Laboratoire de Corrosion - Electrochimie, Faculte des Sciences ,Avenue Ibn Battouta, B.P 1014, Rabat, Morocco. E-mail: guenbour@fsr.ac.ma Laboratoire de Traitement de Surface, Faculte des Sciences, Universite Paul Sabatier 31062, Route de Narbonne, Toulouse - Cedex, France c Centre de Recherche des Phosphates Mineraux (CERPHOS), Avenue Moulay Ismail -Casablanca(Maroc) Received 30-05-2002 Abstract The study passive state behavior of 18Cr-10Ni, 16Cr-14Ni-4W and 18Cr-10Ni-1.5Mo austenitic stainless steels has been performed in -phosphoric acid polluted by sulfides ions using electrochemical techniques, X ray photoelectron spectroscopy (XPS) and secondary ion mass spectroscopy (SIMS). The addition of W or Mo. to austenitic SS induced a decrease of dissolution of rate and affects the nature of passive film. The passive film formed in an overlap of two layers with the external layer rich in W (VI), Mo (VI) and non-metallic elements (O and P). The improved resistance of SS alloys with W is related to formation the film rich in oxide of chromium and tungsten with a little concentration of sulfide. The role of Mo is attributed to formation the film of oxide chrome and molybdenum sulfide. It assumed that Mo prevents the formation of ferrous sulfide. Intoduction The corrosive effect of sulfide ions occurs in various industrial installations. Recently the exploitation of black phosphates (new mineš in south Morocco) containing sulfides induced degradation in the plant used for the production of phosphoric acid by a vvetprocess.1"6 Several studies have shown the aggressive character of sulfide ions towards metal and alloys in different media. In previous paper, we showed that sulfide ion increase the corrosion rate of stainless steel " and nickel and this harmful effect is attributed to adsorption of sulfide ions on the material surface. Similar results were obtained in other sulfuric acid for nickel and nickel based alloys.7"8 The effect of sulfide ions is linked to 7 & formation of a monolayer that presents passivity, " or to the formation of ferrous sulfide (FeS) with the specific adsorption of SFF.9"11 A. Guenbour, M. Essahli, A. Benbachir, L. Aries, R. Boulif: On the Role of Tungsten or Molybdenum... 288 Acta Chim. Slov. 2003, 50, 287-299. To improve the corrosion resistance of stainless steels in various media, elements are frequently added to the alloy. In phosphoric acid without sulfide ions, the beneficial effect of molybdenum addition has been demonstrated; " this element promotes the establishment of passivity. Tungsten additions also improve the corrosion resistance of stainless steel, particularly in the presence of abrasive.14"15 Our purpose in this paper is to study the influence of additions of molybdenum or tungsten to austenitic stainless steel on their behavior in phosphoric acid polluted by sulfide ions. With this in view, films grown in the passive state were analyzed using XPS and SDVIS techniques. Experimental Methods The experimental stainless steels investigated in the present study were prepared in a vacuum furnace. Solution treatment was performed by heating in nitrogen at 1150 °C for 30 min followed by water quenching. Their chemical compositions are given in Table 1. Table 1. Chemical composition (wt %) of the stainless steels tested. Alloy C Mn Si Ni Cr Mo W S 18-10 0.021 1.78 0.58 9.60 18.59 - - 0.0034 16-14-4 0.005 0.30 0.41 14.10 16.50 - 4.81 0.002 18-10-1.5 0.023 1.77 0.63 9.56 18.72 1.52 - 0.0032 Stainless steels have a fully austenitic structure and contain above 0.2% ferrite. The material was shaped as a thin, cylindrical disk with a section of 1 cm . Prior to polarization, the specimen surfaces were polished mechanically with wet silicon carbide (SiC) paper to 1000 grade, digressed with ethyl alcohol, washed in distilled water, and dried with blowing warn air. This study was carried out in 40% wt H3PO4 containing impurities: 500 ppm Cl" and 5 ppm S"" at 60 °C. Sulfides were added to the solution using sodium sulfide of high purity. To accelerate the formation of passive film, potentiostatic curves were obtained by applying a potential step of +600 mV/S.C.E, and registering the resulting current intensity for testing tirne of 2 h, with an AMEL electrochemical set. Potentials were measured with respect to a saturated calomel electrode (S.C.E). A. Guenbour, M. Essahli, A. Benbachir, L. Aries, R. Boulif: On the Role of Tungsten or Molybdenum... Acta Chim. Slov. 2003, 50, 287-299. 289 Valence of the alloy elements in the passive film was investigated by means of XPS spectroscopy. The instrument was an ESCA LAB MK2 (V. G. Scientific). The samples were irradiated with a Mg Ka radiation. Successive sputtering by means of argon ion bombardment did analysis at different depths. The photoelectron spectra on the 2pi/2 and 2p3/2 levels for Fe, Cr, Ni, P, and 3d5/2 levels for Mo, W, and Is level for O and C were measured. The distribution profiles of elements in the passive films were obtained by SIMS using an IMS 300 Camera analyzer. The diameter of the circular zone analyzed was 25 [xm. Profiles of non metallic elements were obtained in an inert atmosphere by bombardment with an argon ion gun. Metal signals were maximized by keeping the oxygen in the analysis chamber under ultrahigh vacuum (The system was evacuated to the 10"9 torr range). Results and discussion 1. Electrochemical behavior on the passive Passive films were grown by polarization at +600 mV/ S.C.E for immersion of samples in 40 wt % H3PO4 + 500 ppm Cl" + 5 ppm S ". Figure 1 shows the current-time curves of the three steels studied. The curves have approximately the same features: current densities decrease quickly and take low values after a length of tirne, which depends on the steel composition. I(»a/cm2) 50 \ \ \ \ \ \ -^- H6Cr-14Nl -4W 40 30 20 'O \ ^-----A-_ S. 1 _^_ ^z^ \ Cl 1 1 so ^o 190 Time (min) Figure 1. Plot of current against immersion tirne in phosphoric medium for different alloys; applied potential +600mV7S.C.E. A. Guenbour, M. Essahli, A. Benbachir, L. Aries, R. Boulif: On the Role ofTungsten or Molybdenum... 290 Acta Chim. Slov. 2003, 50, 287-299. The presence of molybdenum in the steel accelerates a great deal the establishment of the stationary state, achieved after 50 min for 18-10-Mo steel, 120 min for 16-14-W steel, and 130 min for 18-10 steel. In addition Mo and W decrease the dissolution rate at the stationary passive state: 2.8 \xAlcm for 18-10-Mo steel, 5 \xAlcm for 16-14-W steel and 14.4 \xAlcm for 18-10 steel. It can be seen that alloying with molybdenum is much efficient that alloying with tungsten in improved passivation. The results are in good agreement with previous measurement obtained in corrosion-abrasion condition.14"15 So, in this very corrosive medium, where sulfide ions enter into competition with hydroxyl groups, which usually induce passivity, and lead to a structure with defects, the presence of tungsten and especially of molybdenum in the alloy facilitates the establishment of passivity and induces layers which are more protective. 2. Spectroscopy study Passive films grown after 4 hours at +600 mV/S.C.E on the three stainless steels were analyzed using XPS and SIMS. The layer was analyzed at various depths after bombardment with argon ions. Composition of the passive film: 18O-10M SS The XPS spectra thus obtained are shown in the same figure to make comparison easier (figure 2). The Cr 2p peaks (figure 2a) observed at 577 eV (Cr 2p3/2), and 587 eV (Cr 2p1/2) in the passive film are generally attributed to Cr203.In the deep zone peaks appears at 574 eV corresponding to the metallic chromium of the substrate. Iron compounds are minor constituents of the passive film. In the superficial zone the intensity of the Fe 2p peaks is too weak to be analyzed accurately (figure 2b); nevertheless the wide peaks observed about 707-709 eV (Fe 2p3/2) and 722-724 eV (Fe 2pi/2) can be attributed to sulfide and probably sulfate of iron rather than oxides which has peaks at higher energies. In the deep zone the spectrum of iron 2p presents two main Ti peaks at 711 eV (Fe 2p3/2) and 725 eV (Fe 2py2) characteristic of Fe probably in the form of Fe2C>3, Fe304 3>16 could also be present in the layer; these peaks coexist with two others at 707 eV and 721 eV attributed, in the superficial zone, to ferrous sulfide.3"5'16'19 The XPS signal of nickel was too weak to be analyzed exactly; the nickel content of the film is very low, less than a few percent in the superficial zone. The oxygen Is A. Guenbour, M. Essahli, A. Benbachir, L. Aries, R. Boulif: On the Role of Tungsten or Molybdenum... Acta Chirn. Slov. 2003, 50, 287-299. 291 spectrum (figure 2c) presents two broad peaks at 531 eV and 532.5 eV. The second one is preponderant on the surface, and may be attributed to the oxygen of adsorbed molecules such as CO (532 eV) and H2O (533 eV) because of contamination, or to the OH group (531.9 eV) in hydroxides. " The first one is representative of oxygen bound to a metal in an oxide,3-5,18"19 in the deep zone the intensity of the peak at 532.5 eV decreases and only that corresponding to oxides exists. The spectrum of S 2p (figure 2d) in the surface zone exhibits two peaks: the main one at 169 eV indicates the presence of sulfates and the other at 162.2 eV sulfides4-5,16,19 in small quantities. In the deep zone the sulfate form disappears and sulfur is present only as sulfide. Arbitrory Units -------0 j Cr(2p3/2) ............ Srnin A ¦ // v. ¦: // V. •: /' \ Cr(2p1/2) I ^ J' / / J____. _i- 570 575 580 585 590 595 Binding Energy/eV ---------3min ............ G min 528 530 532 534 536 538 540 542 544 Binding Energy/eV (c) Arbitrary Units Fe(2p3/2) 700 705 710 715 720 725 730 Binding Energy/eV (b) Arbitrary Units ¦— ¦¦'/ V-i i V II \ S(2p) /A \ / / x v. / v x A v K lili v^ 160 iG2 164 166 170 172 174 176 178 Binding Enerqy/eV (d) Figure 2. ESCA spectra at various depths of film grown on 18Cr-10Ni alloy in phosphoric acid polluted by sulfides; at the surface (o), after bombardment with argon for 3 min and 6 min. (a) Chromium spectrum. (b) Iron spectrum. (c) Oxygen spectrum. (d) Sulfur spectrum. A. Guenbour, M. Essahli, A. Benbachir, L. Aries, R. Boulif: On the Role ofTungsten or Molybdenum... 292 Acta Chim. Slov. 2003, 50, 287-299. So, the passive film grown on the 18Cr-10Ni steel, in this very corrosive medium, is mainly constituted by Cr2C>3 with a little iron sulfide in the superficial zone. SIMS analysis (figure 3) shows that the profiles iron, chromium and nickel have approximately the same features: the signal intensity of these elements is low at the surface and increases in the deep zone. 40000 - 30000 >- % 20000 on lOOOO- HOO 2 SO 3CO Time (sec) •iOOOOO 75 000 g 50000 25COO -lOO -iSO 200 250 300 Time (sec) Figure 3. SIMS distribution profiles of film grown on 18Cr-10Ni stainless steel after 4 h in phosphoric acid medium with sulfide as impurity. A. Guenbour, M. Essahli, A. Benbachir, L. Aries, R. Boulif: On the Role ofTungsten or Molybdenum... Acta Chim. Slov. 2003, 50, 287-299. 293 The superficial (external) layer is rich in non-metallic elements and particularly in oxygen whose concentration decreases from the surface to the film/metal interface (figure3). Sulfur is also present in the layer; which indicates the formation of sulfides, oxysulphides or sulfates as well as metallic oxides. Likewise the superficial layer contains chlorides and phosphates. This analysis confirms that, the nickel is not present at a significant level in the passive film. These results suggest that the passive film consist of a mixture of Cr and Fe oxide, iron sulfide and to a small extent metallic phosphate. The high corrosion rate observed 0 9091 for 18Crl3Ni can be associated with the amount of ferrous sulfide. ' " The presence of iron sulfide in the passive layer mixed with chromium oxide, leads to a structure with defects because of the non-stoichiometric character of this sulfide. Role of tungsten: 16Cr-16Ni-4W S S No change is observed in the spectra of chromium and oxygen by comparison to those presented for 18Cr-10Ni steel. The spectrum of iron (figure 4a) presents, in the surface zone, peaks characteristic of oxide form (711 eV and 725 eV) and these attributed previously to Fe2C>3 and Fe3C>4 form with a possible contribution of metallic iron (substrate) at 707 eV and 721 eV. The tungsten Adm spectrum (figure 4b) shows the peaks at 250 eV and 263.5 eV characteristic of W6+ probably in WC>315'23'24 at the surface. So, the presence of tungsten in the alloy leads to WC>3 in the passive layer in addition to the constituents observed in the čase of the 18Cr-10Ni stainless steel. The presence of WC>3 reinforces the stability of passive oxide. Figure 5 shows the depth-composition profile (SHVIS analysis) of the film grown on the 16Cr-14Ni-4W: the external layer is enriched in tungsten and chromium, with very low nickel content. The most interesting feature of profile constituent from stainless steel containing W is the existence of the very low amount of sulfur. In addition, the chromium profile suggests that the chromium content of the superficial layer is high. It is well that for given chromium content in SS, the addition of W has strongly beneficial influence of passivity. Tungsten addition to stainless steel, causes the presence of WC>3 in the superficial zone of the passive layer; this compound, insoluble in acid medium, reinforces the protection of Cr2C>3. A. Guenbour, M. Essahli, A. Benbachir, L. Aries, R. Boulif: On the Role of Tungsten or Molybdenum... 294 Acta Chim. Slov. 2003, 50, 287-299. r"e(2p1/2) 715 720 Binding Energy/eV (a) 7?. 5 W{4d3/2) 233 243 253 263 Jinding Energy/eV 273 (b) Figure 4. ESCA spectrum of film grown on 16Cr-14Ni-4W alloy in phosphoric medium polluted by sulfides: (a) Iron spectrum. (b) Tungsten spectrum. Role of molybdenum: 18Cr-10Ni-1.5Mo S S The spectrum of Cr2p3/2 is the same as that observed for 18Cr-10Ni steel: Cr2C>3 is the main component of the passive film. In the superficial zone, the spectrum of iron (figure 6a) is different from that described for the stainless steel without molybdenum; it does not present peaks attributed to iron sulfide, only peaks at 711 eV (Fe 2p3/2) and 725 eV (Fe 2p3/2) characteristics of Fe2C>3. A. Guenbour, M. Essahli, A. Benbachir, L. Aries, R. Boulif: On the Role of Tungsten orMolybdenum... Acta Chim. Slov. 2003, 50, 287-299. 295 The S 2p spectrum (figure 6b) shows a single peak at 162.2 eV attributed to sulfide form MoS2.23'24 >¦ < :— IT) a: 'i 40000 30000 20000 - lOOOO H 25 iSO Time (sec) HOOOOO - 125 HSO Time (sec) Figure 5. SIMS distribution profiles of film grown on 16Cr-14Ni-4W stainless steel after 4 h in phosphoric acid medium with sulfide as impurity. Several molybdenum compounds are present in the layer. The different peaks of Mo3d5/2 and 3d3/2 (figure 6c) can be attributed: 232 eV and 235 eV to Mo03,25'26 228.5 77 7R 7^ 74 eV and 232 eV to M0O2 " and /or to M0S2 " in keeping with that of sulfides shown by the S 2p spectrum. Thus, the passive film grown on the 18Cr-10Ni-1.5Mo steel is constituted mainly by Cr203 with minor constituents like iron oxide, and of molybdenum oxide or sulfide. A. Guenbour, M. Essahli, A. Benbachir, L. Aries, R. Boulif: On the Role ofTungsten or Molybdenum... 296 Acta Chim. Slov. 2003, 50, 287-299. (a) (b) (c) Figure 6. ESCA spectrum of film grown on 18Cr-10Ni-1.5Mo alloy in phosphoric medium polluted by sulfides a) Iron spectrum. b) Sulfur spectrum. c) Molybdenum. A. Guenbour, M. Essahli, A. Benbachir, L. Aries, R. Boulif: On the Role ofTungsten or Molybdenum... Acta Chim. Slov. 2003, 50, 287-299. 297 Figure 7. SIMS distribution profiles of film grown on 18Cr-10Ni-1.5Mo stainless steel after 4h in phosphoric acid medium with sulfide as impurity. The element profiles of the 18Cr-10Ni-1.5Mo passive film (figure 7) are not very different from those observed on the stainless steel without molybdenum. However, the oxygen content is height in the internal layer of the passive film and that external layer is enriched in molybdenum and the nickel content is practically equal to zero. This alloyed element induces the formation of molybdenum oxides, which are very protective as well as insoluble molybdenum sulfide rather than iron sulfide. The thickness of the passive films was estimated from the abrasion tirne by sputtering the surface with argon ions (SIMS analysis) assuming that the erosion rate was constant. Films are very thin, their thickness are of the order of 10 nm, and are A. Guenbour, M. Essahli, A. Benbachir, L. Aries, R. Boulif: On the Role ofTungsten or Molybdenum... 298 Acta Chim. Slov. 2003, 50, 287-299. difficult to measure with precision; it seems that the one fonned on the 18Cr-10Ni-1.5Mo is a little thicker than the two others. Conclusion In order to explain the effect of improved passivation by added W or Mo to austenitic stainless steel in phosphoric acid, electrochemical measurement, SIMS and XPS analysis were perfonned: -Passivity is improved by addition of W or Mo in steel. Molybdenum addition to stainless steel appears to be more efficient than tungsten addition in reducing the steel dissolutionrate. -The presence in the solution of sulfides appreciably modifies the composition of passive layers. A superficial layer rich in iron sulfide characterizes the passive film. -The alloyed elements affects to nature and composition of passive film: the passive film formed is an overlap of two layer with the external layer rich in Mo (VI) or W (VI) and metallic phosphate (probably iron phosphate) -The protective role of the film formed on W alloyed SS may be results from the formation of mixed oxide of chrome and tungsten with very little amount of sulfur The addition of W have strongly effect on sulfur content in passive film -The role of molybdenum it is attributed to formation of oxide chrome and molybdenum sulfide. This alloyed element induces the formation of molybdenum oxides, which are very protective as well as insoluble molybdenum sulfide rather than iron sulfide. Acknowledgements The authors would like to thank the Unieux Research Centre (France) for the alloy samples and to the Phosphor Study and Research Center (Cerphos, Morroco) for fmancial support. Grateful thanks are also expressed to Dr Chatainier and Dr. Amnand (ENSECT, France) for their assistance in the XPS and SIMS analysis. References 1. A. Bellaouchou, A. Guenbour, and A. Benbachir, Corrosion 1993, 49, 656-662. 2. A. Bellaouchou, A. Guenbour, and A. Benbachir, L. Aries, F. Dabosi, Mat. Tech.1998, 8, 21-29. 3. A. Guenbour, S. Zeggaf, and A. Benbachir, Corrosion 1999, 44, 576-581. A. Guenbour, M. Essahli, A. Benbachir, L. Aries, R. Boulif: On the Role of Tungsten or Molybdenum... Acta Chim. Slov. 2003, 50, 287-299. 299 4. S. E. L. Hajjaji, L. Aries, J. P. Audoward, and F. Dabosi, Corros. Sci. 1995, 37, 927-939. 5. S. El Hajjaji, J. Roy, L. Aries, and F. Dabosi, Br. Corr. Jr. 1993, 28, 201-208. 6. H. Idriss, J. Mat. Sci. 1996, 31, 4599-4607. 7. P. Marcus and O. Audouard, Mem. Scien. Revue Met. 1979, 715-718. 8. P. Marcus and H. Talah, Corr.Sci. 1989, 29, 455-463. 9. Z. A. lofa, V. V. Batrakov, and C. N. Ba, Electrochim. Acta 1964, 9, 645-654. 10. B. C. Syrett, Corr. Sci. 1981, 21, 187-193. 11. K. Schawabe, Corr. Se/1964, 4, 156-162. 12. A. Guenbour, J. Faucheu, and A. Benbachir, Corrosion 988, 44, 214-221. 13. A.Guenbour, N. Bui, J. Faucheu, Y. Segui, A. Benbachir, and F. Dabosi, Corr. Sci. 1990, 30, 189-199. 14. A. Guenbour, N. Bui, J. Faucheu, A. Benbachir, and F. Dabosi, Br. Corr. Jr. 1988, 23, 234-238. 16. G. L. Ogundele and W. E. White, Corrosion 1986, 42, 398-408. 17. P. Marcus and J. Oudar, Advanced Research Workshop on Chemistry and Physic of Fracture, (editors: R. M. Latanision and R. H. Jones) Hague 1987, 670-678. 19. C. R. Bredle, T. T. Chung, W. Wandelt, Suf. Sci. 1977, 68, 459-468. 20. I. Ikemoto, K. Tshti, S. Kinoshita, H. Kusada, M. A. Alasio Franco, J. M. Thmas, J. Solide State Chem. 1976, 17, 425-430. 21. A. C. Mackrides and N. Hackerman, Ind. Eng. Chem. 1955, 47, 1773-1778. 22. C. A. Acosta, R. C. S. Solvarezza, H. A. Videla, and A. J. Arvia, Corrosion 1982, 22, 215-221. 23. S. O. Grim, L. J. Matienzo, Inorg. Chem. 1975, 14, 1014-1019. 24. Y. Okamoto, A. Meazawa, and T. Imanaka, J. Catalysis 1989,120, 29-36. 25. N. Bui, A. Irzho, F. Dabosi, and Y. Limouzine-maire, Corrosion 1983, 39, 491-497. 26. W. Yoshioka, H. Habazaki, A. Kawashima, K. Assami, and K. Hashimoto, Electrochem. Acta 1991, 1227-1232. 27. C. Tenret-Noel, J. Verbist, and Y. Golbillon, J. Micro. Spectr. 1976, 1, 255-260. 28. W. E. Shwartz, O. M. Hercules, Analytical Chemistry 1971, 43, 1114-1119. Povzetek Opravljene so bile raziskave o obnašanju 18Cr-10Ni, 16Cr-14Ni-4W in 18Cr-10Ni-1.5Mo pasiviziranih austenitnih jekel v fosforjevi(V) kislini onesnaženi s sulfidnimi ioni. Dodatek volframa in molibdena povzroči počasnejše raztapljanje in vpliva na naravo pasivnega sloja. Izboljšano odpornost povzroča nastanek sloja, bogatega s kromovim(VI) in volframovim(VI) oksidom. Molibden povzroča nastajanje sloja kromovega oksida in molibdenovega sulfida ter preprečuje nastanek železovega sulfida. A. Guenbour, M. Essahli, A. Benbachir, L. Aries, R. Boulif: On theRole ofTungsten or Molybdenum...