Article pubs.acs.org/JPCC
Structure and Morphology Modifications of Silver Surface in the Early Stages of Sulfide Growth in Alkaline Solution Ning Li,†,‡ Vincent Maurice,*,† Lorena H. Klein,† and Philippe Marcus*,† †
Laboratory of Physical Chemistry of Surfaces (LPCS), Chimie ParisTech - CNRS (UMR 7045), 11 rue Pierre et Marie Curie, F-75005 Paris, France ‡ Thin films, Saint-Gobain Recherche, 39 quai Lucien Lefranc, F-93303 Aubervilliers, France ABSTRACT: Electrochemical scanning tunneling microscopy has been employed to investigate the growth mechanism of the 2D adlayer formed on Ag(111) surfaces in the potential range negative to 3D sulfide growth in HS−-containing alkaline aqueous solution. The results show that (2√3 × 2√3)R30°, (√3 × √3)R30°, and (√7 × √7)R19° superstructures are formed with potential-driven increasing uptake of sulfur up to saturation and that extraction of Ag adatoms from the terrace edges is induced by adsorption already at low surface coverage. Irreversible surface fragmentation is produced at the nanometer scale. It is concluded that S uptake occurs via a twoelectron oxidative reaction forming S adatoms in the whole range of potential. No extracted Ag adatoms participate to the construction of the (2√3 × 2√3)R30° and (√3 × √3)R30° superstructures. However, the formation of the (√7 × √7)R19° saturated adlayer involves surface reconstruction and the immobilization of the Ag adatoms extracted from the terrace edges at lower potential and diffusing on the surface to form “AgS” bilayers and “Ag2S” trilayers on the substrate terraces. The saturated adlayer can be considered as a structural precursor of the 3D growth of Ag2S formed at higher anodic potential. Upon reductive desorption of sulfur, the disruption of the Ag−S bonds in the adlayer destabilizes the Ag adatoms that become again mobile on the surface and cluster to form dispersed adislands of smaller dimensions than the initial terraces, thus amplifying surface roughening at the nanometer scale.
1. INTRODUCTION Ultrathin (∼10 nm) layers of silver, transparent in the visible range and reflecting infrared radiations, are increasingly used in the glass industry because of their high performance for thermal insulation. However, Ag is sensitive to tarnishing and ultrathin layers are prone to deteriorate rapidly due to atmospheric corrosion. The atmospheric corrosion of silver has been recognized and analyzed for a long time.1−5 The presence of Ag2S and AgCl, most often reported in silver corrosion layers, and the essential absence of sulfate, nitrate, carbonate, or organic salts of silver have been shown to be a consequence of the thin aqueous layer chemistry taking place in humid environments containing primarily H2S, COS, particulate chloride, and possibly HCl as corrosive agents.5 A nanometric film of moisture forms on the surface, in which Ag+ ions dissolve and react with corrosive agents (e.g., HS−). The study of the initial stages of electrochemical formation of silver sulfide on silver metal is thus relevant to understand and to develop strategies for the control the corrosion resistance of nanometric films. The electrochemical reaction of silver with hydrosulfide HS− ions has been studied6−13 as well as the influence of adsorbed sulfur and its desorption on the electrochemical surface properties.14 However, no consensus has been established on the details of the mechanisms. Briss and Wright7 were the first to report that silver sulfide 3D film growth occurs through the © 2012 American Chemical Society
initial formation of a monolayer of silver sulfide in a separate underpotential deposition (UPD) step. This UPD process was shown later to involve multiple adsorption steps8,9 that were further studied in detail on Ag(111) model surfaces combining electrochemistry with electrochemical quartz crystal microbalance (EQCM),10,11 electrochemical scanning tunneling microscopy (EC-STM),12 or X-ray photoelectron spectroscopy (XPS).13 According to White and coauthors,10,11,13 the UPD growth of the first layer of Ag2S occurs via a two-electron, twostep mechanism. First, the one-electron oxidative adsorption of HS− would form an adlayer noted Ag−SH (Ag + HS− → Ag− SH + e−), and second, at more anodic potential, a one-electron oxidative phase transition involving Ag atoms would transform the Ag−SH adlayer into the saturated Ag2S monolayer (Ag− SH + Ag + OH− → Ag2S + H2O + e−). This two-step mechanism was supported by EQCM and XPS data that showed that the second step is not accompanied by additional mass uptake nor sulfide adsorption, and thus that the adsorbed sulfide saturation coverage would be reached already after the first oxidative step. However, Aloisi et al.12 observed using ECSTM that these first and second steps of the UPD process result in the formation of a (√3 × √3)R30° superstructure (S Received: January 17, 2012 Revised: February 22, 2012 Published: February 29, 2012 7062
dx.doi.org/10.1021/jp300570a | J. Phys. Chem. C 2012, 116, 7062−7072
The Journal of Physical Chemistry C
Article
The Ag single-crystal sample (Mateck Material Technologie & Kristalle GmnH) was as-received oriented (111) with a precision of −0.75 V, for which the S coverage further increases, that reconstruction of the surface is initiated, building the (√7 × √7)R19° superstructure with the incorporation of the mobile Ag adatoms. Recent DFT calculations21 of S adlayers formed on Ag(111) confirm that it is indeed only when reaching a S coverage of 3/7 ML that reconstruction of the Ag(111) planes occurs, thus supporting the mechanism proposed here. In this DFT study, an in-plane upward movement above the S adlayer
observed superstructures and the measured charge density transfers are consistent with the formation of an ordered adlayer of adsorbed S species of increasing coverage. These data do not support the previously proposed adsorption of HS species on different energy sites in this potential range.11,13 The (√7 × √7)R19° superstructure observed at −0.64 V also covers the whole surface. The charge density transfer measured for peaks A + B + C is 178 μC cm−2 (Table 1), in good agreement with the values previously reported at saturation of the 2D sulfide adlayer.12,13 Again, this is slightly (6%) lower than the expected value (189 μC cm−2) corresponding to the formation of 3/7 ML of adsorbed S by a two-electron reaction, the difference being possibly also due to defects in the adlayer such as the boundaries observed between the different domains of the superstructure. These data confirm that, at saturation, the adlayer consists of S species and has a coverage of 3/7 ML as previously reported.11−13 One can then assume the presence of 3 S adatoms per unit cell (not resolved in the present EC-STM study). The NN distance between the S adatoms is then 0.44 nm if the adatoms are equidistant as shown in Figure 9c. The model shown in Figure 9c is an adaptation of models previously proposed for the sulfide/Ag(111) interface formed by dissociative adsorption of S-containing gaseous molecules19,20 and was constructed on the basis of the stacking sequence of the Ag2S trilayers in the (111)-oriented γ-Ag2S phase shown in Figure 10. In this fcc structure (a = 0.627 nm),
Figure 10. Top and side views of the (111)-oriented fcc structure of γAg2S. The larger and smaller spheres represent the Ag and S atoms, respectively. Unit cell, S−S, or Ag−Ag nearest-neighbor distances in the atomic planes and interplanar distances are indicated.
every (111) trilayer consists of 1 S plane sandwiched between 2 Ag planes. The distance between the S and Ag planes is 0.09 nm, and that between two consecutive S planes is 0.36 nm (reticular distance between 2 Ag2S trilayers). The three atomic planes in the trilayers have an hexagonal structure with a NN (S−S or Ag−Ag) distance of 0.443 nm in each plane, in excellent agreement with the expected NN distance of 0.44 nm in the (√7 × √7)R19° adlayer if one assumes equidistance between the S adatoms. Thus, the proposed model of the saturated 2D sulfide adlayer presented in Figure 9c includes 1 7070
dx.doi.org/10.1021/jp300570a | J. Phys. Chem. C 2012, 116, 7062−7072
The Journal of Physical Chemistry C
Article
which causes S-induced reconstruction of the Ag surface for the formation of the “AgS” bilayers and immobilization of the mobile Ag adatoms by strong bonding with S for the formation of the “Ag2S” bilayers. The saturated adlayer can be considered as a structural precursor of the 3D growth of Ag2S formed at higher anodic potential. Correlating the adlayer superstructures with the charge density transfers allows concluding that, from the initial stages up to saturation of the 2D adlayer, the adsorption of sulfur occurs via a two-electron oxidative reaction forming S adatoms. The formation of adsorbed SH groups by a one-electron oxidative reaction in the range of potential of formation of the (2√3 × 2√3)R30° and (√3 × √3)R30° superstructures is not consistent with our data. The surface terraces are fragmented by the mechanism of extraction, diffusion, and immobilization of Ag adatoms induced by the oxidative adsorption of sulfur, which roughens the surface morphology at the nanometer scale. Upon reductive desorption of sulfur, the disruption of the Ag−S bond in the adlayer destabilizes the Ag adatoms that become again mobile on the surface and cluster to form dispersed adislands of smaller dimensions than the initial terraces. This amplifies surface roughening at the nanometer scale.
of the 3 Ag atoms forming the hcp adsorption site of the S adatoms was observed, thus suggesting that the Ag adlayer of the “AgS” bilayers could be formed by direct reconstruction of the substrate planes and disruption of the Ag−Ag bonds. The second Ag adlayer could then be formed by mobile Ag adatoms trapped by the construction of the “Ag2S” trilayers. The mobile Ag adatoms are also thought to feed the observed lateral growth of the adislands that were not consumed in the initial stage of formation of the adlayer (Figures 5 and 6). Some of them could also dissolve in the electrolyte. The mechanism proposed by White and coauthors11,13 for the reaction occurring at peak C of the voltammogram was referred to as a phase transition involving the addition of Ag atoms and the transfer of one electron (Ag−SH + Ag + OH− → Ag2S + H2O + e−). Our study confirms that the addition of Ag adatoms indeed occurs in this later stage of formation of the 2D sulfide adlayer. It shows in addition that the added adatoms originate from the consumption of the terraces occurring in the earlier stages of growth of the adlayer and leading to marked modifications of the surface morphology. However, our results do not support the view of White and coauthors of an unchanged S coverage since an increase from 1/3 to 3/7 ML is associated with the formation of the (√7 × √7)R19° superstructure, in agreement with the study of Aloisi et al.12 An explanation of this discrepancy could be that the related increase of the S coverage (0.1 ML) was too small to be detected by EQCM measurements11 and could have been altered by ex situ transfer in the case of XPS measurements13 that led to the conclusion of no additional S uptake.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected] (V.M.),
[email protected] (P.M.); Fax: +33 (0)1 46 34 07 53; Tel: +33 (0)1 44 27 67 38. Notes
5. CONCLUSION This EC-STM study of the Ag(111)/1 mM Na2S + 0.1 M NaOH(aq) interface in the potential range negative to 3D sulfidation of silver shows that not only different surface phases are formed during the growth of a 2D sulfide adlayer up to saturation but also that irreversible surface morphology modifications occur and that surface reconstruction takes place. At −1.00 V, a (2√3 × 2√3)R30° superstructure has been observed for the first time at this interface. It is associated with a local S coverage of 1/6 ML in the islands forming this superstructure. A step flow mechanism occurs locally due to the extraction of Ag adatoms from the retracting terraces edges. These adatoms, most likely mobile on the surface, do not contribute to the formation of the superstructure. At −0.75 V, at which a fully covering (√3 × √3)R30° superstructure is formed with a S coverage of 1/3 ML, the consumption of the terraces is generalized on the surface. However, the grown superstructure still does not involve Ag adatoms because the S adatoms retain the same low-energy adsorption site (fcc hollow) than at lower coverage. The Ag adatoms remain mobile on the surface, and it cannot be excluded that some fraction dissolves in the electrolyte. At −0.64 V, at saturation of the 2D adlayer (S coverage of 3/ 7 ML), the formation of (√7 × √7)R19° superstructure is confirmed. The proposed model for the saturated adlayer includes 1 plane of S adatoms and 1 or 2 planes of Ag adatoms, forming “AgS” bilayers and “Ag2S” trilayers, respectively, in adjacent islands as observed by EC-STM. It is also proposed that the Ag adatoms involved in the formation of the saturated adlayer are those extracted from the terraces edges at lower potential and diffusing on the surface. In this superstructure, the additional uptake of S requires the occupation of new adsorption sites of higher energy (hcp hollow and atop sites)
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS Saint Gobain Recherche is gratefully acknowledged for financial support and for technical assistance in the preparation of the thin films.
■
REFERENCES
(1) Sharma, S. P. J. Electrochem. Soc. 1978, 125, 2005. (2) Rice, D. W.; Peterson, P.; Rigby, E. B.; Phipps, P. B. P.; Cappell, R. J.; Tremoureux, R. J. Electrochem. Soc. 1981, 128, 275. (3) Franey, J. P.; Kammlott, G. W.; Graedel, T. E. Corros. Sci. 1985, 25, 133. (4) Volpe, L.; Peterson, P. J. Corros. Sci. 1989, 29, 1179. (5) Graedel, T. E. J. Electrochem. Soc. 1992, 139, 1963. (6) Power, G. P.; Ritchie, I. M.; Wylie, M. T. Electrochim. Acta 1981, 26, 1633. (7) Birss, V. I.; Wright, G. A. Electrochim. Acta 1982, 27, 1. (8) Horanyi, G.; Vertes, G. Electrochim. Acta 1986, 31, 1663. (9) Hepel, M.; Bruckenstein, S.; Tang, G. C. J. Electroanal. Chem. 1989, 261, 389. (10) Hatchett, D. W.; White, H. S. J. Phys. Chem. 1996, 100, 331. (11) Hatchett, D. W.; White, H. S. J. Phys. Chem. 1996, 100, 9854. (12) Aloisi, G. D.; Cavallini, M.; Innocenti, M.; Forsti, M. L.; Pezzatini, G.; Guidelli, R. J. Phys. Chem. B 1997, 101, 4774. (13) Conyers, J. L.; White, H. S. J. Phys. Chem. B 1999, 103, 1960. (14) Nguyen Van Huong, C.; Parsons, R.; Marcus, P.; Montes, S.; Oudar, J. J. Electroanal. Chem. 1981, 119, 137. (15) Maurice, V.; Klein, L. H.; Strehblow, H.-H.; Marcus, P. J. Phys. Chem. C 2007, 111, 16351. (16) Maurice, V.; Klein, L. H.; Marcus, P. Surf. Sci. 2000, 458, 185. (17) Kunze, J.; Maurice, V.; Klein, L. H.; Strehblow, H.-H.; Marcus, P. J. Phys. Chem. B 2001, 105, 4263. (18) Yu, M.; Woodruff, D. P.; Bovet, N.; Satterley, C. J.; Lovelock, K.; Jones, R. G.; Dhanak, V. R. J. Phys. Chem. B 2006, 1101, 2164. 7071
dx.doi.org/10.1021/jp300570a | J. Phys. Chem. C 2012, 116, 7062−7072
The Journal of Physical Chemistry C
Article
(19) Yu, M.; Woodruff, D. P.; Satterley, C. J.; Jones, R. G.; Dhanak, V. R. J. Phys. Chem. C 2007, 111, 3152. (20) Window, A. J.; Hentz, A.; Sheppard, D. C.; Parkinson, G. S.; Woodruff, D. P.; Noakes, T. C. Q.; Bailey, P. Surf. Sci. 2010, 604, 1254. (21) Alvarez Soria, L.; Zampieri, G.; Martiarena, M. L. J. Phys. Chem. C 2011, 115, 9587. (22) Schwaha, K.; Spencer, N. D.; Lambert, R. M. Surf. Sci. 1979, 81, 273. (23) Rovida, G.; Pratesi, F. Surf. Sci. 1981, 104, 609. (24) Heinz, R.; Rabe, J. P. Langmuir 1995, 11, 505. (25) Garcia, S. G.; Salinas, D. R.; Mayer, C. E.; Lorenz, W. J.; Staikov, G. Electrochim. Acta 2003, 48, 1279. (26) Brunetti, V.; Blum, B.; Salvarezza, R. C.; Arvia, A. J. Langmuir 2003, 19, 5336. (27) Wang, D.; Xu, Q. M.; Wan, L. J.; Wang, C.; Bai, C. L. Surf. Sci. 2002, 499, L159. (28) Sugimasa, M.; Inukai, J.; Itaya, K. J. Electrochem. Soc. 2003, 150, 210. (29) Spänig, A.; Broekmann, P.; Wandelt, K. Electrochim. Acta 2005, 50, 4289. (30) Domange, J.-L.; Oudar, J. Surf. Sci. 1968, 11, 124. (31) Perdereau, M.; Oudar, J. Surf. Sci. 1970, 20, 80. (32) Edmons, T.; McCarrou, J. J.; Pitkethly, R. C. J. Vac. Sci. Technol. 1971, 8, 68. (33) Stickney, L. L.; Rosasco, S. D.; Salaita, G. N.; Hubbard, A. T. Langmuir 1985, 1, 66. (34) Hayek, K.; Glassl, H.; Gutmann, A.; Leonhard, H.; Prutton, M.; Tear, S. P.; Welton-Cook, M. R. Surf. Sci. 1985, 152, 419. (35) Koestner, R. J.; Salmeron, M.; Kollin, E. B.; Gland, J. L. Surf. Sci. 1986, 172, 668. (36) Maca, F.; Scheffler, M.; Berndt, W. Surf. Sci. 1985, 160, 467. (37) Patterson, C. H.; Lambert, R. M. Surf. Sci. 1987, 187, 339. (38) Chan, C.-M.; Weinberg, W. H. J. Chem. Phys. 1979, 71, 3988. (39) Wong, P. C.; Zhou, M. Y.; Hui, K. C.; Mitchell, K. A. R. Surf. Sci. 1985, 163, 172. (40) Gao, X.; Zhang, Y.; Weaver, M. J. J. Phys. Chem. 1992, 96, 4156.
7072
dx.doi.org/10.1021/jp300570a | J. Phys. Chem. C 2012, 116, 7062−7072