Info

Fig. 2. Section profile along [2-1-1] of the interface between the thin oxide film and the metal substrate constructed from bulk parameters in the case of the NiO(433) // Ni(l 11) with NiOfO-11] // Ni[0-ll] epitaxy (8.02° tilt between the (111) planes of the oxide and those of the substrate). The atomic planes and nodes are indicated. Different terminations of the oxide film are illustrated. The vector of the coincidence cell along [2-1-1] is times that of the substrate. Possible atomic displacements at the interface towards the nearest three-fold sites of the substrate are indicated.

atomic displacements (marked by arrows) at the interface towards the nearest threefold hollow sites of the substrate. The presence of terraces and steps at the surface of the passive film may also reflect a preferential dissolution at steps in the passive state. In addition to steps and kinks at the surface of the passive film, other crystalline defects such as point defects possibly related to vacancies have been imaged. One example is also shown in Fig. 1. The presence of crystalline defects at the surface of the passive film may play a key role in the resistance to breakdown. The bottoms of steps and kinks correspond to sites of reduced thickness of the passive film where the barrier property of the film is expected to be diminished. These defects may also constitute sites of preferential adsorption of aggressive ions such as chloride ions. The hydroxide layer present in the outer part of the film, which is about one monolayer thick [10-13], could not be observed in a distinct manner in the SIM images. The possible role of the hydroxyl groups in the tunneling mechanism is discussed below.

The crystalline nature of the passive film formed on Ni has been confirmed by in-situ STM for a different crystallographic orientation of the electrode and in a different electrolyte: Ni(100) in 1M NaOH [14]. At low potential (-0.7 to -0.5 V/NHE), oxidation resulted in the formation of a well-ordered rhombic structure resistant to reduction and assigned to the irreversible formation of the Ni/Ni(OH)2 interface. At increasing potential, a distortion of this rhombic structure was first observed, followed at higher potential (>0.18 V/NHE) by a quasi-hexagonal structure with a nearest-neighbor spacing consistent with either |3-Ni(OH)2(0001) or NiO(lll). These results suggest that independently of the substrate orientation (Ni(100) or Ni(lll)), the crystalline passive film is NiO (111) oriented and terminated by a p-Ni(OH)2(0001) hydroxide layer in (lxl) epitaxy on the oxide layer.

For the study of chromium passivation performed in our group [6], (110) oriented single crystal surfaces were used. Pretreaments by electrochemical polishing and annealing at high temperature in hydrogen revealed, as for Ni, atomic terraces of lateral dimensions of the order of tenths of micrometers. Passivation was performed in 0.5 M H2SO4 by potential steps from the corrosion potential to three different potentials in the passive region (+300, +500 and +700 mV/RHE). Different time periods of polarization were investigated (20 min, 2 and 22 h). The STM measurements were combined with X-ray photoelectron spectroscopy (XPS) analysis. The XPS analysis (in UHV) showed that the passive film contains trivalent chromic species only. The oxide inner part of the film varies from dispersed three dimensional (3D) islands to a complete layer about 0.9 nm thick. The thickness of the hydroxide outer layer varies from about 0.6 to about 1.3 nm. Composition and thickness were found to be stable upon exposure to air. STM measurements in air showed that unlike Ni, the surface is quite homogeneous on the submicroscopic scale after passivation. This is due to the quasi-absence of dissolution during the passivation treatment. On Cr, the surface topography is heterogeneous only on a nanoscopic scale. It is characterized by disordered protrusions of 1 - 4 nm lateral dimensions which induce vertical variations of 0.4 - 0.8 nm amplitude. A typical image is shown in Fig. 3. This topography is independent of the passivation conditions. On the atomic scale, small areas of limited lateral extension are detected.

Fig. 3. STM images of the passive film formed on Cr(110) in 0.5M H2S04 at +500 mV/RHE. The left image shows the disordered protrusions of nanoscopic dimensions. The right images show a small ordered domain assigned to a nanocrystal of oxide (a-Cr203(0001)) at two different magnifications.

1 nm

Fig. 3. STM images of the passive film formed on Cr(110) in 0.5M H2S04 at +500 mV/RHE. The left image shows the disordered protrusions of nanoscopic dimensions. The right images show a small ordered domain assigned to a nanocrystal of oxide (a-Cr203(0001)) at two different magnifications.

This is also shown in Fig. 3. These areas are characteristic of small ordered domains assigned to nanocrystals of oxide emerging at or near the film surface. The quasi-hexagonal arrangement and the nearest neighbor distances of the corrugations are consistent with the arrangement of the O2" ligands in the basal plane of the oxide (a-Cr203(0001)). Larger ordered domains are observed by STM when more oxide is formed in the inner part of the film, according to XPS measurements. These ordered domains are surrounded by areas where no structural periodicity is evident which are assigned to the hydroxide outer part of the film. No crystalline defects are detected at the boundaries of the ordered domains. In terms of resistance to breakdown, the topography variations of the hydroxide outer part also induce sites of reduced thickness of the passive film where the barrier property of the film is expected to be diminished. However, crystalline defects are not evident in these sites, probably because they are cemented by the amorphous structure of the hydroxide. Therefore, they are expected to offer higher resistance to film breakdown. In addition, the amorphous structure of the hydroxide is expected to minimize the variations of coordination of the surface atoms at ciystalline defects and therefore to induce a higher chemical passivity at these sites. Hence, the nanocrystalline structure of the oxide and the role of cement played by the hydroxide would be responsible for the higher passivity of the film formed on Cr.

Moffat et al. [15] also reported an ex-situ STM investigation of Cr(l 10) passivated in 1M H2SO4. Their images confirmed the presence of ordered domains consistent with the structure of a-Cr203 oriented (0001) and parallel to Cr(l 10).

The structure of the passive film on a Fe-22Cr alloy [7, 8] was investigated with (110) oriented single crystal surfaces. The passivation was performed in 0.5 M H2SO4 at potential values of +300, +500 and +700 mV/RHE. Different time periods of polarization were investigated (20 min, 2, 22 and 63 h). The XPS analysis (in UHV) showed that the passive film is well described by a bilayer model with a mixed trivalent oxide inner layer enriched in C^Cb (from 88 to 95% and 0.7 to 1.4 nm thick with increased aging) and a chromium hydroxide outer layer whose thickness varies from a 1 nm-thick 3D layer to 2D islands with increased aging. Upon exposure to air, only the films aged for 22 and 63 h do not show a significant evolution of the (much thinner) hydroxide outer layer. STM measurements in air showed that the terrace topography of the substrate surface is maintained after passivation. Similar protrusions (up to 10 nm across) to those recorded on Cr(l 10) are observed on the alloy. However, in the case of the alloy, a fusing process has been observed as a function of aging under polarization. This process is attributed to the coalescence of islands of the passive film occurring during film growth and aging (in a similar way to that observed on Fe [2] but on a much longer time scale). On the atomic scale, the passive film has been found to be noncrystalline after polarization for 2 h at +500 mV/RHE. Aging under polarization favors a crystallization process evidenced by the presence of epitaxial crystalline areas consistent with the structure of a-Cr203. This is illustrated in Fig. 4. Similarly to Cr(llO), the basal plane (0001) of Cr203 is found to be parallel to the (110) plane of the substrate. Deviations from a quasi-hexagonal arrangement of the corrugations are measured. This is possibly due to the presence of OH groups, as discussed below. On

Fig. 4. STM images of the passive film formed on Fe-22Cr(l 10) in 0.5M H2S04 at +500 mV/RHE and aged 63 h under polarization. The left image shows a crystalline area with (possibly) the emergence of a screw dislocation. The right image shows an ordered area at higher magnification. The quasi-hexagonal lattice is consistent with the basal plane of chromium oxide: <x-Cr203(0001). Deviations from a perfect periodicity of the corrugations are observed in these crystalline areas.

Fig. 4. STM images of the passive film formed on Fe-22Cr(l 10) in 0.5M H2S04 at +500 mV/RHE and aged 63 h under polarization. The left image shows a crystalline area with (possibly) the emergence of a screw dislocation. The right image shows an ordered area at higher magnification. The quasi-hexagonal lattice is consistent with the basal plane of chromium oxide: <x-Cr203(0001). Deviations from a perfect periodicity of the corrugations are observed in these crystalline areas.

the alloy, three different azimuthal orientations of the oxide islands have been found. However, the crystallization is not complete in these conditions and the topography of the passive film is intermediate between that recorded on passivated Ni(l 11) (complete crystallization with large crystals) and that recorded on passivated Cr(110) (nanocrystals cemented by noncrystalline areas). It shows the presence of both crystalline defects and noncrystalline areas.

Ryan et al. [16, 17] studied ex-situ and in-situ microcrystalline Fe-Cr surfaces prepared by sputter deposition and passivated at a low potential in the passive region where mostly Cr is expected to participate in film formation. The noncrystalline character of the passive film formed on Fe-Cr alloys has been confirmed in-situ after lh of passivation [17]. In this study, it was observed that after 1 h of passivation the amount of disorder increased sharply for a Cr content of the alloy between 14.7 and 16.5%. The effect of aging was tested only in ex-situ conditions and it was observed that it favors the (re)crystallization of the passive film for 18 and 21% Cr content [15]. The structural order observed in the passive film was consistent with the (0001) orientation of a-Cr203. In the case of pure Fe sputter-deposited thin films polarized at high anodic potentials in borate buffer solution, the same group reported long-range crystalline order in both ex-situ and in-situ conditions of examination [18].

It appears from these studies that both the Cr content of the alloy and the conditions of aging of the film are critical factors ruling the crystallization of the passive film.

0 0

Post a comment