Inhibition effect of underpotential deposition of metallic cations on aqueous corrosion of metals

4.1.1 Detailed study of the Pb-UPD effect on the anodic dissolution of Ni

Aiming at a more quantitative understanding of the relation between Pb-UPD and corrosion of Ni, Seo et al. (2008) investigated in detail the effect of Pb²⁺ (10⁻³m–10⁻⁵m) on the polarization behavior of Ni at room temperature in deaerated 0.1 m NaClO₄+10⁻²m HClO₄ solution. Figure 4 shows the voltammograms of Ni specimens at the first cycle in the wide potential region between −0.40 V and 1.20 V in the solutions without and with 10⁻³m Pb²⁺ (added as PbO to give the desired concentration of Pb²⁺ in solution). The solid and dotted lines in Figure 4 represent the voltammograms in the presence and absence of Pb²⁺, respectively. After E_corr attained a steady state (it takes about 15 min), the voltammograms were measured. The steady state value of E_corr of Ni in the presence of Pb²⁺ is 0.09 V, which is more positive by 0.15 V than that (−0.06 V) in the absence of Pb²⁺. The potentio-dynamic polarization (10 mV/s) is started from E_corr to the cathodic limit potential, E_c,l=−0.40 V, turned from −0.40 V to the anodic limit-potential, E_a,l=1.20 V, and returned to −0.40 V. The cathodic current peak observed at −0.27 V during the first cathodic potential sweep in the presence of Pb²⁺ results from the bulk deposition of Pb, while the anodic current peak observed at −0.20 V during the subsequent anodic potential sweep is caused by the anodic stripping of bulk-deposited Pb. Moreover, the anodic current peaks corresponding to the anodic dissolution of Ni are observed at 0.11 V in the absence of Pb²⁺ and at 0.15 V in the presence of Pb²⁺. It is remarked that the anodic current peak height of Ni-active dissolution decreases significantly in the presence of Pb²⁺. It is also noted that in the passive region between 0.40 V and 1.20 V, the passivity-maintaining current density in the presence of Pb²⁺ is almost the same as that in the absence of Pb²⁺.

Figure 4:

Voltammograms of Ni specimens at the first cycle in the wide potential region between −0.40 V and 1.20 V in deaerated 0.1 m NaCO₄+10⁻²m HClO₄ solutions without and with 10⁻³m Pb²⁺ (Seo et al., 2008). The solid and dashed lines represent the voltammograms in solutions with and without 10⁻³m Pb²⁺, respectively. PbO was added in solution to give 10⁻³m Pb²⁺. Potential sweep rate: 10 mV/s. Reproduced with permission from © Elsevier Ltd., 2008. All rights reserved.

Figure 5 shows the voltammograms of Ni specimens in the narrow potential region between −0.215 V and 0.80 V in which E_c,l is fixed at −0.215 V corresponding to E_eq(10⁻³m Pb²⁺/Pb) in order to avoid the influence of bulk-deposited Pb on the anodic dissolution of Ni. In Figure 5 as well as Figure 4, the presence of Pb²⁺ shifts the potential at the anodic current peak of Ni-active dissolution toward the positive direction and decreases the anodic current peak. In addition, in the presence of Pb²⁺ during the first cathodic potential sweep toward E_c,l, a cathodic current peak is observed at −0.12 V, which is associated with the electro-deposition of Pb (i.e. Pb-UPD) on Ni. On the other hand, during the subsequent anodic potential sweep from −0.215 V, no clear anodic current peak corresponding to the anodic stripping of electro-deposited Pb is observed up to 0.15 V at which the anodic current peak of Ni-active dissolution emerges, suggesting that the anodic current peak due to the anodic stripping of electro-deposited Pb is masked in the background current. The results of Figure 5 indicate that the inhibition effect of Pb²⁺ on the anodic dissolution of Ni is associated with Pb-UPD on Ni, which is consistent with those obtained in 0.1 m HClO₄ at 90°C by Radhakrishnan et al. (2005).

Figure 5:

Voltammograms of Ni specimens at the first cycle in the narrow potential region between −0.215 V and 0.80 V in deaerated 0.1 m NaCO₄+10⁻²m HClO₄ solutions without and with 10⁻³m Pb²⁺ (Seo et al., 2008). The cathodic limit potential is fixed at E_c,l=−0.215 V corresponding to E_eq(10⁻³m Pb²⁺/Pb) in order to avoid the influence of bulk-deposited Pb on the anodic dissolution of Ni. The solid and dashed lines represent the voltammograms in solutions with and without 10⁻³m Pb²⁺, respectively. Potential sweep rate: 10 mV/s. Reproduced with permission from © Elsevier Ltd., 2008. All rights reserved.

Figure 6 shows the anodic polarization curves of Ni specimens measured starting from E_corr at a potential sweep rate of 1.0 mV/s in deaerated 0.1 m NaClO₄+10⁻²m HClO₄+x M Pb²⁺ (x=0, 10⁻⁵, 10⁻⁴ and 10⁻³) solutions. The steady-state value of E_corr shifts toward the positive direction with increasing Pb²⁺ concentration. The values of E_corr are listed in Table 2. The potential, E_a,p, at the anodic current peak of Ni-active dissolution also shifts toward the positive direction with increasing Pb²⁺ concentration. Moreover, the height of the anodic current peak decreases with increasing Pb²⁺ concentration. In Figure 7, the logarithm of the anodic current density, i_a, in the potential range between E_corr and E_a,p is plotted versus the applied potential, E, at various Pb²⁺ concentrations. The anodic Tafel slope, b⁺, obtained from the linear relation between log i_a and E is listed in Table 2. The corrosion current density, i_corr, obtained with the extrapolation of the Tafel line at E_corr is also listed in Table 2. The addition of 10⁻⁴m–10⁻³m Pb²⁺ decreases i_corr by a factor of 3.

Figure 6:

Anodic polarization curves of Ni specimens in deaerated 0.1 m NaClO₄+10⁻²m HClO₄+x M Pb²⁺ (x=0, 10⁻⁵, 10⁻⁴ and 10⁻³) solutions (Seo et al., 2008). The anodic polarization is started from E_corr at a potential sweep rate of 1.0 mV/s. Reproduced with permission from © Elsevier Ltd., 2008. All rights reserved.

Figure 7:

Logarithm of anodic current density, i_a, versus applied potential, E, at various Pb²⁺ concentrations: (A) without Pb²⁺, (B) with 10⁻⁵m Pb²⁺, (C) with 10⁻⁴m Pb²⁺ and (D) with 10⁻³m Pb²⁺ (Seo et al., 2008). The data in Figure 7 are taken from the anodic polarization curves in Figure 6. Reproduced with permission from © Elsevier Ltd., 2008. All rights reserved.

The XPS and glow discharge-optical emission spectroscopy (GDOES) analyses (Seo et al., 2008) of Ni specimens polarized at different potentials in deaerated 0.1 m NaClO₄+10⁻²m HClO₄ solution with 10⁻³m Pb²⁺ have provided for the first time the evidence that Pb species is present on the Ni surfaces at potentials more positive than E_eq(10⁻³m Pb²⁺/Pb)=−0.215 V and the surface coverage of Pb species, θ_Pb, decreases with increasing applied potential. Figure 8 shows the Pb 4f XPS spectra of Ni specimens polarized for 30 min at three different potentials in the presence of 10⁻³m Pb²⁺. It has been generally accepted that the amount of UPD species on noble metals is saturated at the equilibrium potential of bulk deposition (Herrero et al., 2001). Assuming θ_Pb=1 at −0.215 V, θ_Pb=0.43 at 0.07 V is estimated from the area ratio of the Pb 4f_7/2 XPS peak at 0.07 V to that at −0.215 V. The value of θ_Pb corresponding to E_corr=0.091 V at 10⁻³m Pb²⁺ (Table 2) would be less than θ_Pb=0.43 at 0.07 V since θ_Pb decreases due to the anodic stripping of electro-deposited Pb with increasing potential. As shown in Figure 8, no Pb species is detected on the Ni surface at 0.80 V in the passive potential region of Ni. The GDOES results have also indicated that Pb species is absent on the Ni surface at potentials more positive than 0.30 V (Seo et al., 2008). The absence of Pb species in the passive region of Ni is consistent with the results obtained by Costa et al. (1995).

Figure 8:

Pb 4f XPS spectra of the Ni specimens polarized for 30 min at three different potentials in deaerated 0.1 m NaClO₄+ 10⁻²m HClO₄ solution with 10⁻³m Pb²⁺: (A) −0.215 V, (B) 0.07 V and (C) 0.80 V (Seo et al., 2008). Reproduced with permission from © Elsevier Ltd., 2008. All rights reserved.

Assuming a Langmuir type of adsorption isotherm and employing the value of θ_Pb=0.43 at −0.07 V reported by Seo et al. (2008), Protopopoff and Marcus (2013) calculated the standard free energy of formation (defined at θ_Pb=0.5) for Pb-UPD on Ni, ΔG^o_f (298.15 K)=−54.4 kJ/mol, and drew the stability domains of electro-deposited Pb on polycrystalline Ni in potential-pH diagrams at 25°C and 300°C in order to point out the importance of Pb-UPD on Ni for understanding the Pb-SCC mechanism of Ni-base alloys used in nuclear power plants. Figure 9 shows the potential-pH diagram for electro-deposited Pb on polycrystalline Ni at 300°C. It is remarked that the Ni²⁺/Ni or NiO/Ni equilibrium potential line is located within the stability domain of electro-deposited Pb on Ni. Besides, Protopopoff and Marcus (2017) derived the adsorption energies of Pb on Fe and Cr from density functional theory calculations and drew the stability domains of electro-deposited Pb on Fe and Cr in potential-pH diagrams at 25°C and 300°C. The stability domains of electro-deposited Pb on Fe, Cr and Ni are overlapped, suggesting that Pb-UPD participates in Pb-SCC of Ni-base alloys. According to Staehle (2004, 2005), Pb-SCC of Ni-base alloys in high-temperature water takes place in the vicinity of the NiO/Ni equilibrium potential boundary associated with the active/passive transition of Ni. Consequently, it is very probable that Pb-UPD plays a vital role in Pb-SCC as suggested by Newman (2005) although more detailed experiments on atomistic levels are needed in future to confirm the direct participation of Pb-UPD in the Pb-SCC mechanism.

Figure 9:

Potential-pH diagram for electro-deposited Pb on polycrystalline Ni at 300°C (Protopopoff & Marcus, 2013). Both molalities of dissolved Pb and Ni species are fixed to 10⁻⁶ mol/kg. The stability domains are limited by dotted (A, B) lines for the H₂O system, thick-dashed lines for the Pb-H₂O system, thin-solid lines for the Ni-H₂O system, and the thick-solid lines for the Pb_ads(Ni)-Pb-H₂O system (θ_Pb=0.5), respectively. Reproduced with permission from The Electrochemical Society.

Furthermore, Seo et al. (2016) investigated the effect of Pb-UPD on the anodic dissolution and passivation of Fe, Fe-30Ni and Fe-70Ni binary alloys in acidic perchlorate solution at room temperature in relation to Pb-SCC of Ni-base alloys. The anodic polarization curves of pure Fe, Fe-30 Ni and Fe-70 Ni alloys were measured at a potential sweep rate of 1.0 mV/s from corrosion potential, E_corr, in deaerated 0.1 m NaClO₄+10⁻²m HClO₄ solutions without and with 10⁻³m Pb²⁺. The order of E_corr(Fe)=−0.377 V<E_corr(Fe-30Ni)=−0.145 V<E_corr(Fe-70Ni)=−0.095 V in the absence of Pb²⁺ is consistent with the order of corrosion resistivity of pure Fe and Fe-Ni binary alloys (Economy et al., 1961; Condit, 1972). It is expected that ΔE_UPD widens with increasing Ni content in the alloy because of Φ_Ni=5.15 eV>Φ_Fe=4.5 eV. It has been found that the addition of 10⁻³m Pb²⁺ shifts E_corr toward the positive direction, irrespective of alloy composition, and the positive shift in E_corr increases with increasing Ni content in the alloy, suggesting that there is a correlation between E_corr and ΔE_UPD (Seo et al., 2016).

4.1.2 In situ XAS study of Pb-UPD on Ni

Although the presence of Pb species adsorbed on the Ni surface was experimentally verified by XPS, the peak position (138.0 eV) of the Pb 4f_7/2 spectrum for the Ni specimens at −0.215 V and 0.07 V in Figure 8 is closer to that (137.95 eV) for Pb(OH)₂ than to that (136.8 eV) for metallic Pb (Wagner, 1983), indicating that the Pb species is not metallic but in an oxidation state. However, surface in situ XAS, using synchrotron radiation, has provided clear evidence that the Pb species adsorbed on Ag or Pt is metallic (Samant et al., 1987; McBreen & Sansone, 1994). Consequently, it appears that the Pb species has been oxidized in air before the introduction of the Ni specimen into the XPS apparatus. In fact, surface in situ XAS is a powerful tool for the in situ characterization of the electrode/solution interface (Abruna, 1991; Lützenkirchen-Hecht & Strehblow, 2006; Nagy, 2011). The X-ray absorption near-edge structure (XANES) provides information on the chemical state of adsorbed species and the surface extended X-ray fine structure (EXAFS) provides information on the local atomic structure of the adsorption layer such as coordination number, inter-atomic distance and local order. In general, XAS is not sensitive to a surface since X-rays of high energy have long propagation length in condensed matters.

The surface XAS with grazing incidence excitation and fluorescence detection has to be employed in order to get the information limited to a surface. In application of surface XAS to the electrochemical UPD systems, a specimen electrode with a smooth surface and a thin electrolyte layer are indispensable for grazing angle excitation and for fluorescence detection. In the case of a Ni electrode subjected to corrosion in acidic solution, however, the above technique is not acceptable because of the changes in surface morphology and the precipitation of corrosion products in the thin electrolyte layer.

Seo et al. (2012) developed the periodical emersion method using a surface-roughened (sr-)Ni plate with a surface roughness (ratio of the real surface area to the geometrical surface area) of S_r=78.3 for the in situ XAS study of Pb-UPD on Ni in deaerated acidic perchlorate solution and succeeded in measuring the XAS spectra of Pb species on Ni under potentiostatic polarization. Figure 10 shows the normalized Pb L_III XANES spectra of the sr-Ni plate (solid curve) polarized at −0.185 V in deaerated 0.1 m NaClO₄+10⁻²m HClO₄ solution with 10⁻⁴m Pb²⁺, of the metallic Pb foil (dotted curve) and of the Pb²⁺ solution (dashed curve). The location of photon energy, E_p=13.037 keV, at μ=0.5 in the normalized Pb L_III XANES spectra of the sr-Ni plate at −0.185 V coincides with that of the metallic Pb foil. Moreover, the absorption peak at E_p=13.052 keV typical for the Pb²⁺ solution is not observed for the sr-Ni plate at −0.185 V. The results of Figure 10 have provided the first confirmative evidence that the Pb species on Ni at potentials more positive than E_eq(10⁻⁴m Pb²⁺/Pb) is metallic. Assuming that the uppermost surface of the Ni plate exposed to solution is mainly oriented to the Ni (111) plane, the EXAFS analysis was made to obtain the information on the surface atomic structure of the Pb-UPD layer on Ni. The most suitable surface atomic structure model for Pb-UPD on Ni obtained eventually by the EXAFS analysis is depicted in Figure 11, which represents the coexistence of the surface alloy and the overlayer with the same p (2×2) structure. Some Pb atoms (black circles) occupy a three-fold hollow site of Ni atoms (white circles), while other Pb atoms (gray circles) are embedded into the Ni lattice, surrounded by six nearest-neighbor Ni atoms in the first Ni layer and coordinated with three Ni atoms (not drawn in Figure 11) in the second Ni layer. Moreover, the inter-atomic distances, d_Pb-Ni=2.64 Å, between Pb (black or gray circles) and Ni (white circles) atoms, and, d_Pb-Pb=3.33 Å, between Pb (black circles) and Pb (gray circles) atoms were obtained by the EXAFS analysis. The value of d_Pb-Ni=2.64 Å is significantly small as compared to that (d_Pb-Ni=3.00 Å) estimated from a hard sphere model based on the metallic radii (r_Pb=1.75 Å and r_Ni=1.25 Å) of bulk polycrystalline Pb and Ni, indicating that the bonding between the Pb and Ni atoms is close to covalent (Brown et al., 2000). The effective radius of the Pb atom, r_Pb=1.39 Å, described in Figure 11 was determined from d_Pb-Ni=2.64 Å, assuming that the effective radius of the Ni atom, r_Ni=1.25 Å, is unchanged.

Figure 10:

Normalized Pb L_III XANES spectra of the sr-Ni plate (solid curve) polarized at −0.185 V in deaerated 0.1 m NaClO₄+10⁻²m HClO₄ solution with 10⁻⁴m Pb²⁺, of the metallic Pb foil (dotted curve) and of the Pb²⁺ solution (dashed curve) (Seo et al., 2012). Reproduced with permission from © Elsevier B.V., 2012. All rights reserved.

Figure 11:

The most suitable surface atomic structure model for Pb-UPD on Ni representing the coexistence of the surface alloy and the overlayer with the same p (2×2) structure (Seo et al., 2012). Some Pb atoms (black circles) occupy a three-fold hollow site of Ni atoms (white circles), while other Pb atoms (gray circles) are embedded into the Ni lattice, surrounded by six nearest-neighbor Ni atoms in the first Ni layer and coordinated with three Ni atoms (not drawn in this figure) in the second Ni layer. The inter-atomic distances, d_Pb-Ni=2.64 Å, between the Pb (black or gray circles) and Ni (white circles) atoms, and, d_Pb-Pb=3.33 Å, between the Pb (black circles) and Pb (gray circles) atoms were also obtained by the EXAFS analysis. The effective radius of the Pb atom, r_Pb=1.39 Å, is determined from d_Pb-Ni=2.64 Å, assuming that the effective radius of the Ni atom, r_Ni=1.25 Å, is unchanged. Reproduced with permission from © Elsevier B.V., 2012. All rights reserved.

Despite the immiscibility of Pb in bulk Ni, the coexistence of the surface-constrained alloy phase with the overlayer in Figure 11 may be associated with the compressive surface stress produced by the large difference between the radii of the Pb and Ni atoms as pointed out by Ibach (1997). The transition between surface alloy and overlayer has been discussed by Nagl et al. (1994) for Pb-UPD on Cu (111) and by Shin et al. (2010) for Pb-UPD on Au (111). It is known that certain liquid metals embrittle certain pure or alloyed solid metals, e.g. Hg embrittles Zn and brass (Latanision & Westwood, 1970). The liquid-metal embrittlement (LME) has been discussed in terms of a decrease in surface energy associated with the adsorption of liquid-metal atoms at a crack tip. According to Staehle (2004), there has been no clear evidence that the Pb-LME is operative for Pb-SCC. Nevertheless, the above in situ XAS results (Seo et al., 2012) would give new insight of the Pb-SCC mechanism in which the electro-deposition of metallic Pb, followed by surface alloy formation, may promote the crack initiation of Ni-base alloys due to evolution of large internal compressive stress (equivalent to external tensile stress) in addition to a decrease in surface energy.

4.2.1 Possibility for participation of Sn-UPD in corrosion of iron and steels

The inhibition effect of Sn²⁺ on corrosion of iron and steel has been investigated in connection with the development of Sn-containing steels (Buck & Leidheiser, 1957; Kudo & Okamoto, 1968; Osozawa, 1971; Sherlock, 1975; Morad, 2007; Nam et al., 2010; Kamimura et al., 2012). Osozawa (1971) investigated not only the corrosion resistances of austenitic stainless steels (Fe-18Cr-20Ni-2Mo) containing up to 0.26% Sn at 98°C in 5% sulfuric acid but also the effect of 10⁻⁴m Sn²⁺ (added in solution) on corrosion of stainless steel without Sn and reported that Sn²⁺ added in solution as well as Sn added in stainless steel inhibits both metal dissolution and hydrogen evolution at potentials more negative than −0.04 V. The inhibition of metal dissolution and hydrogen evolution was ascribed to the presence of precipitation film formed on the steel by the reaction of Sn²⁺ with metal ions dissolved from the steel, as suggested by Kudo and Okamoto (1968). Osozawa (1971) did not refer to the participation of Sn-UPD in corrosion of the steel. The potential window of Sn-UPD on austenitic stainless estimated from the work functions of Sn (Φ_Sn=4.42 eV) and steel (Φ_steel=4.61 eV) is ΔE_UPD=0.095 V. Consequently, it is deduced that Sn-UPD proceeds on the steel in the potential region between −0.256 V and −0.161 V. According to the results by Osozawa (1971), the corrosion rate of the steel takes a maximum at about −0.15 V, while the corrosion rate decreases by a factor of more than 100 in the presence of 10⁻⁴m Sn²⁺ at the same potential. Therefore, the participation of Sn-UPD in corrosion inhibition of steel at potentials more negative than −0.15 V may not be excluded.

Kamimura et al. (2012) have developed Sn-bearing steel (Fe-0.1–0.6% Sn) that exhibits a superior atmospheric corrosion resistance, particularly in chloride environment, and have reported that the addition of 5×10⁻⁴m Sn²⁺ in deaerated 3% NaCl (pH 1.0) solution at 30°C inhibits both anodic dissolution of Fe and hydrogen evolution. Kamimura et al. (2012) explained the inhibition of hydrogen evolution in terms of the bulk deposition of Sn on Fe. However, no clear explanation was made on the inhibition effect of Sn²⁺ on the anodic dissolution of Fe. According to the polarization results by Kamimura et al. (2012), E_corr of Fe in the presence of 5×10⁻⁴m Pb²⁺ is about −0.30 V and an anodic current peak of 60 μA/cm² appears at −0.27 V; then the anodic current decreases to about 25 μA/cm² and stays constant in the narrow potential region between −0.24 and −0.20 V. Afterward, the anodic current increases rapidly with increasing potential. E_eq(5×10⁻⁴m Pb²⁺/Pb)=−0.235 V is more positive than E_corr=−0.30 V, indicating that the bulk deposition of Sn on Fe proceeds at E_corr. The potential window of Sn-UPD on Fe estimated from the work functions of Sn (Φ_Sn=4.42 eV) and Fe (Φ_Fe=4.5 eV) is ΔE_UPD=0.04 V. Consequently, it is deduced that Sn-UPD proceeds on Fe in the narrow potential region between −0.235 V and −0.195 V. The potential region of Sn-UPD on Fe coincides with the narrow potential region of anodic current arrest in the presence of 5×10⁻⁴m Sn²⁺, suggesting that the participation of Sn-UPD in inhibition of anodic dissolution is likely.

4.2.2 Study of the Sn-UPD effect on the anodic dissolution of Ni

Seo et al. (2014b) investigated the effect of Sn²⁺ on the anodic dissolution of Ni in deaerated 0.2 m HClO₄ solution with Sn²⁺ (added as SnO) at room temperature. Figure 12 shows the cyclic voltammograms of Ni specimens measured in the potential region between E_c,l=−0.226 V and E_a,l=1.0 V at a potential sweep rate of 10 mV/s in deaerated 0.2 m HClO₄ solutions without and with 10⁻³m Sn²⁺. E_c,l=−0.226 V corresponds to E_eq(10⁻³m Sn²⁺/Sn). E_corr=0. 22 V in the presence of 10⁻³m Sn²⁺ is more positive by 0.17 V than E_corr=0.05 V in the absence of 10⁻³m Sn²⁺. The first potential cycle was initiated from E_corr to E_c,l and then from E_c,l to E_a,l. The anodic current peak, i_a,p, at about 0.2 V in the absence of 10⁻³m Sn²⁺ increases with increasing cycle number, while no anodic current peak in the presence of 10⁻³m Sn²⁺ appears, irrespective of the cycle number, indicating that the addition of 10⁻³m Sn²⁺ inhibits completely the anodic dissolution of Ni in 0.2 m HClO₄. The potential window of Sn-UPD on Ni estimated from the work functions of Sn (Φ_Sn= 4.42 eV) and Ni (Φ_Ni=5.15 eV) is ΔE_UPD=0.365 V. It is thus deduced that Sn-UPD proceeds on Ni in the potential region between −0.226 V and 0.139 V. E_corr=0.05 V in the absence of 10⁻³m Sn²⁺ is located in the potential region of Sn-UPD on Ni, indicating that the shift of E_corr toward the positive direction due to the addition of 10⁻³m Sn²⁺ is associated with Sn-UPD on Ni. However, E_corr=0.22 V in the presence of 10⁻³m Sn²⁺ is more positive than the potential region of Sn-UPD on Ni. Figure 13 shows the Sn 3d XPS spectra of the Ni specimens polarized for 30 min at various potentials in deaerated 0.2 m HClO₄ solution with 10⁻³m Sn²⁺ (Seo et al., 2014b). The Sn 3d_3/2 peak area in Figure 13 decreases with increasing potential due to the anodic stripping of electro-deposited Sn. It is noted that Sn species is still present on the Ni surface at potentials more positive than the potential region of Sn-UPD on Ni and even in the passive region of Ni. The presence of Sn species on Ni means that the anodic stripping of Sn species does not proceed completely and the Sn species in the oxidation state remains partly on the Ni surface. A rotating ring-disk study by Vicente and Bruckenstein (1972) confirmed the incomplete anodic stripping of electro-deposited Sn on Au and the presence of Sn residue on the Au surface. It should be noted that Sn species in the metallic state is not observed in the potential region of Sn-UPD except for −0.20 V in 3d XPS spectra shown in Figure 13. The Sn species electro-deposited on Ni under potential control in solution may be oxidized in air since the Ni specimens after polarization were exposed to air and stored in a desiccator prior to the XPS analysis.

Figure 12:

Cyclic voltammograms of Ni specimens measured in the potential region between E_c,l=−0.226 V and E_a,l=1.0 V at a potential sweep rate of 10 mV/s in deaerated 0.2 m HClO₄ solutions without and with 10⁻³m Sn²⁺ (Seo et al., 2014b). The vertical arrows represent the changes in polarization behavior with cycle number (1→5). SnO was added in solution to give 10⁻³m Sn²⁺. Reproduced with permission from The Electrochemical Society.

Figure 13:

Sn 3d XPS spectra of the Ni specimens polarized for 30 min at various potentials in deaerated 0.2 m HClO₄ solution with 10⁻³m Sn²⁺ (Seo et al., 2014b). The vertical bars marked with Sn (m) and Sn (ox) in the Sn 3d_5/2 peak represent the energy positions of Sn in metallic and oxidized states, respectively. Reproduced with permission from The Electrochemical Society.

4.2.3 In situ XAS study of Sn-UPD on Ni

The in situ XAS study combined with the periodical emersion method using an sr-Ni plate with a surface roughness of 78.3 has been performed by Seo et al. (2014a) in order to investigate the chemical state and local structure of electro-deposited Sn on Ni under potential control in deaerated 0.2 m HClO₄ solution with 10⁻³m Sn²⁺. Figure 14 shows the normalized XANES spectra of Sn K-edge for the sr-Ni plates polarized at −0.15 V and 0.05 V in deaerated 0.2 m HClO₄ solution with 10⁻³m Sn²⁺. The normalized Sn K-edge XANES spectrum of the metallic β-Sn foil is also shown in Figure 14 for comparison. The normalized Sn K-edge XANES spectrum of the sr-Ni plate at −0.15 V or 0.05 V shifts to the high-energy side by about 5 eV as compared to that of the metallic β-Sn foil, indicating that the Sn species electro-deposited on the sr-Ni plate is in the oxidized state rather than the metallic state. For the EXAFS analysis of the Sn-UPD layer on Ni, the oscillatory part, χ(κ), of the normalized Sn K-edge XAS spectrum above the absorption edge for the sr-Ni plate was extracted as a function of the wave number, κ, by removing the smooth background to obtain the κ³-weighted EXAFS functions (κ³χ vs. κ curves).

Figure 14:

Normalized XANES spectra of the Sn K-edge for the sr-Ni plates polarized at −0.15 V and 0.05 V in deaerated 0.2 m HClO₄ solution with 10⁻³m Sn²⁺ (Seo et al., 2014a). The normalized Sn K-edge XANES spectrum of a metallic β-Sn foil is also shown for comparison. Reproduced with permission from The Electrochemical Society.

Figure 15 shows the radial structure functions, RSFs (|χ(R)| vs. R curves), obtained by Fourier transforming the κ³-weighted EXAFS functions in the range of κ=2.5 Å⁻¹–12.5 Å⁻¹ for the sr-Ni plates at various potentials. The RSFs in Figure 15 have two major peaks, P_A at R=1.7 Å and P_B at R=2.2 Å. The peak height of P_A increases with increasing potential, while the peak height of P_B decreases reversely. These two peaks correspond to the first few coordination shells in the vicinity of absorbing Sn atoms. It has been clarified from the comparison with RSFs of Sn standard materials (β-Sn foil, SnO and SnO₂) that the peaks, P_A and P_B, are assigned to the Sn-O and Sn-Ni interactions, respectively. The RSFs of the surface atomic structure models proposed for the Sn-UPD layer on Ni are simulated by using the program FEFF8.2 (Ankudinov et al., 1998, 2002) to compare wth the real RSFs in Figure 15 by curve fitting. As a result, a Sn-incorporation model is the most suitable in which Sn atoms are substituted like in a surface alloy at face-centered cubic sites in the first Ni layer and further coordinated with oxygen atoms (Seo et al., 2014a).

Figure 15:

Radial structure functions, RSFs (|χ(R)| vs. R curves), obtained by Fourier transforming the κ³-weighted EXAFS functions (κ³χ vs. κ curves) in the range of κ=2.5 Å⁻¹–12.5 Å⁻¹ for the sr-Ni plates kept at various potentials in deaerated 0.2 m HClO₄ solution with 10⁻³m Sn²⁺ (Seo et al., 2014a). The peaks, P_A and P_B, are assigned to the Sn-O and Sn-Ni interactions, respectively. Reproduced with permission from The Electrochemical Society.

The EXAFS parameters (including coordination number, inter-atomic distance, etc.) obtained with curve fitting for two shells of Sn-O and Sn-Ni based on the Sn-incorporation model are listed in Table 3. The inter-atomic distances between Sn and O, d_SnO=2.03–2.04 Å, and between Sn and Ni, d_Sn-Ni=2.55–2.57 Å, are almost independent of potential, while the coordination number of Sn-O, N_Sn-O, increases from 1.5 to 2.9 with increasing potential and the coordination number of Sn-Ni, N_Sn-Ni, decreases reversely from 7.1 to 4.6. The values of d_SnO and N_Sn-O at −0.22 V are not listed because of lack of a clear peak corresponding to the Sn-O interaction. The small values of R-factor (goodness of fit), R_f=0.004–0.012, in Table 3 exhibit the excellent match between real and simulated RSFs. Although the Sn K-edge XAS spectra on the sr-Ni plate are not measured at potentials more positive than 0.05 V, the values of N_Sn-O≈3.8 and N_Sn-Ni≈4.0 at the boundary potential (0.139 V) of Sn-UPD are estimated from the extrapolation of N_Sn-O and N_Sn-Ni at various potentials in Table 3. The EXAFS data in Table 3 support the presence of Sn residue on the Ni surface at potentials more positive than the potential region of Sn-UPD on Ni.

The value of d_Sn-Ni=2.55–2.57 Å in Table 3 is significantly small as compared to that (d_Sn-Ni=2.65 Å) predicted from a hard sphere model based on the metallic radii (r_Sn=1.40 Å and r_Ni=1.25 Å) of bulk crystalline α-Sn and Ni. If the effective radius of Ni atoms surrounding central Sn atoms is unchanged from the bulk, the effective radius of the Sn atom is estimated to be r_Sn=1.31 Å, which is less by 0.09 Å than that (r_Sn=1.40 Å) of bulk crystalline α-Sn, suggesting that the bond between the Sn and Ni atoms is nearly covalent (Woodruff & Robinson, 2003). The significant reduction in the effective radius of the adsorbate at formation of surface alloy was reported for the vapor-deposited Sn monolayer (Woodruff & Vlieg, 2002) and Pb monolayer (Brown et al., 2000; Woodruff & Vlieg, 2002) on Ni (111). When Sn atoms are electro-deposited on the Ni surface, a large compressive surface stress could be produced due to the large difference between the radii of the Sn and Ni atoms, as pointed out by Ibach (1997). The reduction in r_Sn at the surface alloy formation in spite of the immiscibility of Sn in bulk Ni is energetically preferable since the relaxation of compressive surface stress is brought by the reduction in r_Sn. The value of d_SnO=2.03–2.04 Å for the sr-Ni plate in Table 3 is slightly less than that (d_SnO=2.05 Å) for bulk SnO₂, which may be associated with the reduction in r_Sn at the surface alloy formation. It has been finally drawn from the above EXAFS results that the bonding of Sn atoms with oxygen atoms in addition to the covalent bonding of Sn atoms with Ni atoms suppresses the progress in anodic stripping of electro-deposited Sn with increasing potential to leave the Sn residue in the oxidized state on the Ni surface which inhibits completely the anodic dissolution of Ni. As described in the preceding section, the Pb-UPD behavior on Ni is different from the Sn-UPD behavior on Ni. The in situ XAS study of Pb-UPD on Ni (Seo et al., 2012) has indicated that Pb species on Ni is completely metallic and not oxygenated. Although the Pb atoms bonded with Ni atoms block the active sites of the Ni surface to inhibit the anodic dissolution of Ni, the anodic stripping of metallic Pb proceeds completely without leaving any Pb species on the Ni surface with increasing potential to induce the anodic dissolution from the bare surface sites prior to passivation. The difference between Pb-UPD and Sn-UPD on Ni may result from the strong affinity of Sn with oxygen as compared to Pb since the standard free energy (−251.8 kJ/mol) of SnO formation is more negative than that (−188.9 kJ/mol) of PbO formation (Sakai & Yokokawa, 2000).