Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Unmasking chloride attack on the passive film of metals

Unmasking chloride attack on the passive film of metals ARTICLE DOI: 10.1038/s41467-018-04942-x OPEN Unmasking chloride attack on the passive film of metals 1 1 1 1 1 1 1 1 2 1,3 B. Zhang , J. Wang ,B.Wu , X. W. Guo , Y. J. Wang , D. Chen , Y. C. Zhang ,K.Du , E. E. Oguzie &X.L.Ma Nanometer-thick passive films on metals usually impart remarkable resistance to general corrosion but are susceptible to localized attack in certain aggressive media, leading to material failure with pronounced adverse economic and safety consequences. Over the past decades, several classic theories have been proposed and accepted, based on hypotheses and theoretical models, and oftentimes, not sufficiently nor directly corroborated by experimental evidence. Here we show experimental results on the structure of the passive film formed on a FeCr Ni single crystal in chloride-free and chloride-containing media. We use 15 15 aberration-corrected transmission electron microscopy to directly capture the chloride ion accumulation at the metal/film interface, lattice expansion on the metal side, undulations at the interface, and structural inhomogeneity on the film side, most of which had previously been rejected by existing models. This work unmasks, at the atomic scale, the mechanism of chloride-induced passivity breakdown that is known to occur in various metallic materials. Shenyang National Laboratory for Materials Science, Institute of Metal Research, Chinese Academy of Sciences, Wenhua Road 72, 110016 Shenyang, China. Electrochemistry and Materials Science Research Laboratory, Department of Chemistry, Federal University of Technology Owerri, PMB, Owerri 1526, Nigeria. State Key Lab of Advanced Processing and Recycling on Non-ferrous Metals, Lanzhou University of Technology, 730050 Lanzhou, China. These authors contributed equally: B. Zhang, J. Wang. Correspondence and requests for materials should be addressed to X.L.M. (email: xlma@imr.ac.cn) NATURE COMMUNICATIONS | (2018) 9:2559 | DOI: 10.1038/s41467-018-04942-x | www.nature.com/naturecommunications 1 1234567890():,; ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04942-x orrosion is one of the major causes of material Results failure and hence leads to a huge cost to our society . Sample preparation under various chemical conditions. Passive CThe nanometer-thick passive film on metals films were formed on the (001) and (110) plane, respectively, of resists a general corrosion, but it is susceptible to severe FeCr Ni single crystal (Supplementary Note 1 and Note 2, 15 15 localized attack in certain aggressive media . The best-known Supplementary Figs 1–3). This enabled us to obtain a distinct inducer of localized passive film breakdown is the chloride ion. metal/passive film interface (Supplementary Fig. 4) and better Despite the enormous amount of experimental data and diverse characterize the structure of the interface region. On the other 1–13 hypotheses and models proposed till date , the breakdown of hand, the single crystal, which is free of any inclusions and grain the passive film is still not sufficiently understood and remains boundaries, yields a high-quality passive film with a continuous one of the most important and basic problems in corrosion coverage on the alloy matrix. It also effectively avoids the para- science. digm of the weakest sites breaking down the soonest, which The lack of agreement on the mechanism of passive film makes figuring out the intrinsic mechanism complex (Supple- breakdown is mainly due to the difficulty encountered in mentary Note 2). In order to monitor the transport and effect of obtaining precise experimental information. To clarify the chloride ions, passive films were formed under three designated exact nature of the chloride-induced breakdown, Cl incor- conditions: passivation in H SO electrolyte, passivation in 2 4 poration to the film has to be experimentally confirmed H SO + NaCl electrolyte, and initial passivation in H SO 2 4 2 4 and the accurate location needs to be identified. In the electrolyte and subsequent addition of NaCl into the H SO 2 4 meanwhile, chloride-induced modification to the film has to electrolyte (Supplementary Note 3, Supplementary Figs 5 and 6). be experimentally addressed as well. It is worthy of note The cross-sectional TEM specimen was prepared by the con- that these issues were extensively studied by X-ray photoelec- ventional method, that is, passivated surfaces of two samples were 7,14–21 tron spectroscopy (XPS) , Auger electron spectroscopy bonded face-to-face and then thinned by grinding and ion- 7,14,16,19,22–27 7,8,16,19,25 (AES) , secondary ion mass spectrometry , milling. During sample preparation and subsequent TEM and radiotracer techniques . Nonetheless, it is still very difficult observation, the extremely thin passive film was strictly ensured and challenging to guarantee the precision and accuracy of free of mechanical and beam-induced damage (Supplementary observed locations and concentrations of a very small amount of Note 4). chloride in an extremely thin passive film with a thickness of only a few nanometers. Much of evidence on the incorporation of Cl Structural evolution of passive film with aging in air. The TEM in the passive oxide film can be mainly classified into two groups: observation in the high-angle annular-dark-field (HAADF) mode 7,8,11,14–17,22–24 one is chloride incorporation , and the other is and Super-X EDS analysis on the passive film were performed 14,18–21,25–27 chloride absence in the passive film . In the case of and the results are shown in Figs. 1 and 2. Figure 1a is the incorporation, the location of chloride in the passive film is also HAADF scanning transmission electron microscopic (HAADF- controversy. Some investigators declare that Cl locate or con- STEM) image showing the passive film on FeCr Ni single 15 15 7,14–16,22,23 centrate in the outer layer of the film , whereas some crystal formed in H SO electrolyte (condition 1). According to 2 4 8,24 others claim it is in the inner layer . In reality, most of the the contrast difference, the film seems to be tri-layer structured. reported methods do not directly identify the presence of chloride Whereas the EDS mapping analysis (Fig. 2a) indicates a well- or its location within the passive film. Alongside questions defined bi-layer structure with the inner Cr-rich layer and the regarding chloride ion distribution in the passive film are issues 29,47–51 outer Fe-rich layer, as generally accepted . Interestingly, relating to the nature of atomic-scale interactions between the after aging the specimen in air for a few days, we also subjected passive film and chloride, i.e., issues regarding how, where, and the aged specimen again to TEM observation and we found that when the chloride modifies the passive film. To the best of our the contrast had become homogeneous, following disappearance knowledge, although the experimental advances, including of the middle darker-contrast layer (Fig. 1b). The implication 28–31 transmission electron microscopic (TEM) characterization , is that the outer layer of the passive film, formed by precipitation promote powerful evidence on the structure and chemistry of the of the hydrolyzed metal cations, is transformed to the metal 20,28,29,31–45 passive film as well as the evolution imparted by oxide by a dehydration reaction, which was confirmed by XPS 36,39,43,46 34,38,39,50,51 chloride , no technique has succeeded in directly fol- analysis . It is worthwhile to mention that the HAADF lowing the evolution of the passive film, in chloride-containing mode image provides an incoherent image using high-angle media, across the entire film ranging from the surface to the scattered electrons, where the contrast is strongly dependent on metal/film interface. the scattering ability of heavy atoms. Thus the much denser metal Without doubt, deciphering the interactions of the chloride matrix would show the brightest contrast, followed by the metal ion with the passive film, including chloride-induced modifica- oxide, with metal hydroxide displaying the darkest contrast. tions to the properties of the passive film at the atomic scale, From the foregoing and taking into consideration of is key to understanding the precise mechanism of passivity the EDS mapping results, the inner layer is a Cr-rich oxide, while breakdown. In the present work, using aberration-corrected the outer layer, with the darkest contrast should be Fe-rich TEM (Cs-corrected TEM) and a fast and precise super X-ray energy-dispersive spectrometer (Super-X EDS) analysis with four detectors, we simultaneously investigate the film and metal matrix as well as their interface via cross-sectioning in real Glue space. We find the chloride accumulation within the inner layer Film of the passive film and the associated fluctuations at the matrix/ passive film interface. We provide direct evidence on the Matrix location of chloride and the resultant phenomena of the lattice expansion on the metal side, undulations at the interface, and structural inhomogeneity on the film side. The present findings Fig. 1 HAADF-STEM image showing the contrast evolution of the passive allow for the atomic-scale mechanism of passivity breakdown film on FeCr Ni single crystal on exposure to air. Scale bar, 5 nm. a Image 15 15 to be revisited on the basis of real-space imaging in multi- immediately after the specimen was prepared. b Image after the specimen dimensions. has been exposed in air for about 6 days 2 NATURE COMMUNICATIONS | (2018) 9:2559 | DOI: 10.1038/s41467-018-04942-x | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04942-x ARTICLE Film Matrix Cl O Ni Cr Fe Film Matrix Cl O Cr Ni Fe Film Matrix Cl O Ni Fe Cr Fig. 2 Super-X EDS mapping showing chloride ion incorporated in and penetrating the passive film and accumulating at the matrix/passive film interface. −1 −1 −1 Element maps of the film formed in a 0.5 mol L H SO electrolyte at 640 mV/SHE for 30 min; b 0.5 mol L H SO + 0.3 mol L NaCl electrolyte at 2 4 2 4 −1 640 mV/SHE for 30 min, and c passivated in 0.5 mol L H SO electrolyte at 640 mV/SHE for 30 min with subsequent addition of NaCl. All scale bars in 2 4 a–c are 2 nm a b (110) (001) cd (110) (001) Fig. 3 Chloride-induced undulations at the metal/passive film interface, wherein the yellow line indicates the metal/passive film interface. Scale bar, 2 nm. a, b HRTEM images along the [001] and [110] axes of the austenitic matrix showing the passive film grown on (110) and (001) planes of FeCr Ni single 15 15 −1 crystal in 0.5 mol L H SO electrolyte, revealing a sharp and straight interface, even at the atomic scale. c HRTEM images along [001] axis showing the 2 4 −1 −1 passive film grown on (110) plane in 0.3 mol L NaCl + 0.5 mol L H SO electrolyte, revealing an indistinct and substantially undulating interface. d 2 4 −1 HRTEM images along [110] axis showing the passive film initially grown in 0.5 mol L H SO electrolyte for 30 min with subsequent addition of NaCl into 2 4 the H SO electrolyte. The interface is as well indistinct and substantially undulating 2 4 hydroxides. The unavoidable exposure in air going from passi- number of Fe and Cr is rather close, making the contrast of Cr- vation in the electrolyte to TEM observation allows the outmost rich oxide layer and Fe-rich oxide layer seems homogeneous. The hydroxide layer to be partially dehydrated yielding the brighter above results provide direct experiment evidence of a dehydration oxide, thus giving the film an apparent tri-layered structure in the model. HAADF-STEM image. With prolonged exposure, the hydroxide layer became completely dehydrated and the passive film corre- Structural evolution of the interface imparted by chloride. spondingly reverted to an oxide film. Although the oxide film has Prior to studying the interactions of chloride ions with the passive a bi-layered structure distinguished from the Cr-rich and Fe-rich film and austenitic matrix, the pristine passive film formed in layer (Fig. 2a), it can be hardly identified in the image taken with chloride-free electrolyte was analyzed, in which attention was the HAADF mode (Fig. 1b), since the contrast in the HAADF paid to the structure and the element distribution within the film image is strongly dependent on the scattering ability of heavy and near the interfaces. Cs-corrected TEM revealed a passive atoms that is associated with the atomic number. Here the atomic film with thickness of about 4–5 nm (see TEM image in NATURE COMMUNICATIONS | (2018) 9:2559 | DOI: 10.1038/s41467-018-04942-x | www.nature.com/naturecommunications 3 (010) (110) (110) F (112) (100) (001) (110) (110) ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04942-x Supplementary Fig. 4). The corresponding EDS maps are shown in Fig. 2a. A well-defined bi-layer passive film is seen, where the inner (barrier) layer adjacent to the metal is enriched in Cr and (110) depleted in Fe, while the outer (precipitated) layer is depleted in Cr and enriched in Fe. It is noteworthy that the film/matrix interface is sharp, well defined, and straight, even at the atomic level. On the matrix side, immediately adjacent to the passive film/matrix interface there exists a layer with Cr depletion and Ni enrichment, which is in agreement with the previous find- 29,31,48,51 ings . Cross-sectional high-resolution TEM (HRTEM) images, obtained along the [001] and [110] axes (Fig. 3a, b) of the austenitic matrix, respectively, show that the passive film is mostly amorphous. c The passive film formed in chloride-containing H SO solution 2 4 (condition 2) is shown in Fig. 3c. It is immediately obvious that the previously sharp and straight interface in chloride-free (110) solution has become indistinct and substantially undulating. This (001) is indisputably a consequence of chloride ion attack. By means of Super-X EDS mapping experiments, we successfully obtained the distribution of elemental Cl in the passive film, as shown in Fig. 2b, which, interestingly, shows Cl to be concentrated within the inner layer, (average content ~1 atomic %). Supplementary Figure 7 also shows a distinctive Cl peak in the composition spectrum. Significantly lower amounts of Cl (0.1 atomic %) were Fig. 4 Cross-sectional HRTEM images along the [001] and [110] axes of detected in the outer layer. Evidently, these findings indicate that the austenitic matrix showing the interface between the passive film and the chloride ions incorporate in the passive film, permeate the steel matrix. It is seen that the passive film is mainly amorphous, with some outer and inner layers, and ultimately attack the interface, giving nanocrystals. A series of HRTEM images obtained from variant orientations rise to the undulating interfacial structure. Such an incorporation and locations indicate that the nanocrystals feature face-centered cubic of solution species in the inner layer has not been previously structure. In most cases, the nanocrystals have crystallographic observed and is actually considered non-feasible by some existing 5,10,13,14,52 orientations with austenitic matrix but occasionally not. This figure displays theories on passivity breakdown . typical configurations where the orientation relationships between the We designed another experimental procedure, wherein the nanocrystals in the passive film and the single-crystalline substrate are FeCr Ni single crystal was initially passivated in H SO 15 15 2 4 labeled. a Film (110) || FeCr Ni (110) and film (1–10) || FeCr Ni (1–10). 15 15 15 15 electrolyte for 30 min and subsequently NaCl was added into b The crystal is randomly oriented with no orientation relationship with the the H SO electrolyte (condition 3). This enabled us to monitor 2 4 austenitic matrix. c Film (110) || FeCr Ni (110) and film (1–10) || FeCr Ni 15 15 15 15 the attack of chloride ions on the as-grown passive film. By (1–12). d Film (001) || FeCr Ni (001) and film (1–10) || FeCr Ni (1–10). 15 15 15 15 careful examination of a series of interfaces, what we observed is a Scale bar, 1 nm reduced prevalence of the undulating interface, which only manifested at a few locations (Fig. 3d). These locations are expected to be terminal points for paths through which chloride austenitic matrix but occasionally not. Thus the grain boundary ions permeate. Correspondingly and maybe not unexpectedly, Cl between the NCs and amorphous phase can be readily identified was only detected at locations manifesting the undulating according to the HRTEM images. Nevertheless, a determination of interfaces, typically like that in Fig. 2c and was not detected atomic configurations at this kind of boundary is a challenge, (Supplementary Fig. 8) at the still distinct and unperturbed which should be more complex than that of two crystalline grains. interfaces (similar to that in Fig. 2a). It is thus obvious that We hypothesize that the interfaces between the NCs and the chloride ions only get to certain interfacial locations by amorphous zone assume the features of atomic irregularity and heterogeneously penetrating the as-grown film. thus provide ready paths for chloride ion transport. The selective permeation implied by our results arises from the inherently non-homogeneous microstructure of passive films and Nanocrystal-amorphous interface assisting Cl permeation.It depends on the nature of and interconnection between the paths is well known that grain boundaries exhibit an irregular atom created along the NCs/amorphous zone interface. When a array yielding a loose structure that usually provide tunnels for connected path traverses the entire thickness of the passive film, species diffusion and transport. In contrast, the amorphous phase chloride ions tunneling through those paths would eventually always features a random atom distribution, so the diffusion and arrive at the matrix/passive film interface. On the other hand, transport processes in amorphous materials are not yet well where no connected paths exist, or all the paths are abridged, the understood. It is noteworthy that, although the passive films in the matrix/passive film interface would remain unperturbed because present study are mostly amorphous as seen in the HRTEM chloride ions are unable to get through. images, some periodic two-dimensional lattices with a scale of 1–3 In order to provide a theoretical basis for the above assertions, nm are often visualized, that is to say, a small amount of nano- we performed first-principle computations to model the diffusion crystals (NCs) are embedded in the passive films. Figure 4a–d of chloride ions within the passive film. The supercells constructed display typical cross-sectional HRTEM images in which some NCs for the simulation contain the NC structure, the amorphous zone, inherently present in the amorphous passive film. A series of and the amorphous zone/NC interface, respectively. This enables a HRTEM images obtained from variant orientations and locations quantification of the diffusion barriers to chloride ion diffusion indicate that NCs feature face-centered cubic structure (also seen from one oxygen vacancy to the next, as illustrated in Fig. 5. in Supplementary Fig. 9). In our present study, the NCs often Details of the computation procedure can be found in display specific crystallographic orientation relationships with the Supplementary Methods. Actually, for a qualitatively comparison, 4 NATURE COMMUNICATIONS | (2018) 9:2559 | DOI: 10.1038/s41467-018-04942-x | www.nature.com/naturecommunications (110) M NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04942-x ARTICLE the trend of diffusion barriers in these three zones has nothing to In earlier studies, several techniques, including reciprocal space 32,53–55 do with the specific crystal structure. So, for simplicity in the identification by XRD analysis , electron diffraction in a 56–58 calculations, we selected the binary spinel Fe O representing the TEM , and real-space imaging upon the outmost surface 3 4 33,42,50,51,59–62 spinel-structured NC embedded in the passive film. By comparing by scanning tuneling microscopic technique , have the energy barriers for Cl ion diffusion from one oxygen vacancy been applied in determining structural information on the nano- to its neighboring one, our computations show the diffusion crystalline nature of the passive films, and the role barrier to be lowest at the interface region, thus confirming that of grain boundaries in passivity breakdown and initiation of 63–65 the interface between the NCs and the amorphous zone provides a localized corrosion has been widely discussed . It is worth- favorable channel for chloride ion diffusion. while to mention that, although the grain boundary was specified 0.8 Diffusion path 0.7 [010] 0.6 [100] Cl 0.5 a-Fe O a/c-Fe O c-Fe O 3 4 3 4 3 4 Fe 0.4 0.3 [010] [001] [100] Fig. 5 Energy barriers to Cl ion diffusion from one oxygen vacancy to a neighboring one in the three zones. Representative schematic diagram of the diffusion paths in c-Fe O is inserted. The diffusion barrier is lowest at the interface region, thus confirming that the interface between the nanocrystals and 3 4 the amorphous zone provides a favorable channel for chloride ion diffusion a b –0.15 (110) –0.11 –0.07 –0.04 0.00 0.04 0.07 0.11 0.15 (001) Fig. 6 LADIA simulation results showing the strain state in the matrix side near the metal/passive film interface. Elastic strain component measured from HAADF-STEM image intensity peaks (normal to the interface plane between the passive thin film and the alloy specimen) through the alloy crystal. Scale bar, 2 nm. a High-resolution HAADF-STEM image along the [001] direction of the austenitic matrix showing the passive film on (110) of FeCr Ni single 15 15 −1 crystal in 0.5 mol L H SO . The interface is sharp and straight at the atomic scale. b LADIA simulation map based on a, which shows no evidence of 2 4 lattice expansion and associated tension. The image shows the local expansion/contraction of next-neighbor atom column distances in the alloy along the −1 −1 [110] direction. c High-resolution HAADF-STEM image along the [110] axis showing a passive film formed in 0.5 mol L H SO + 0.3 mol L NaCl 2 4 electrolyte, with corresponding undulating interface. d LADIA simulating map based on c, revealing obvious lattice expansion, with associated induced tension. The image shows the local expansion/contraction of next-neighbor atom column distances in the alloy along the [100] direction. The color baron the right indicates the normal strain, where colors for positive values represent tensile strain and colors for negative values represent compressive strain NATURE COMMUNICATIONS | (2018) 9:2559 | DOI: 10.1038/s41467-018-04942-x | www.nature.com/naturecommunications 5 Energy barrier (eV) ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04942-x Electrolyte Electrolyte δ+ δ+ δ+ δ+ δ+ M M M M M Passive film Interface movement Interface movement Straight interface Metal Metal b Breakdown site δ+ δ+ δ+ M M M – – Cl – Cl – Cl Cl Cl Stress Cl Slow Convex Fast Fast Metal Concave Cl Cl δ+ δ+ δ+ δ+ δ+ M M M M M Passive film Cl Unperturbed Concave Metal Metal interface Fig. 7 Schematic maps illustrating interface evolution in the absence and presence of chloride ions. The film growth process involves transport of both injected metal ions from the matrix and oxygen in solution through the barrier layer, causing the metal/film (Me/BL) interface to move toward the metal −1 matrix side. a In chloride-free (0.5 mol L H SO ) electrolyte, the austenitic matrix experienced selective dissolution of metal ions (major Fe) in a 2 4 −1 −1 somewhat homogeneous manner, yielding a straight Me/BL interface at the atomic scale. b In chloride-containing (0.5 mol L H SO + 0.3 mol L NaCl) 2 4 electrolyte, the tendency of chloride ions to be preferentially adsorbed at defective sites yields non-homogeneous adsorption on the bare metal surface. High chloride ion concentration would induce faster dissolution of Fe, which leads to inhomogeneous interface-movement rates, hence differences in passive film growth rates. Such a process yields a passive film with an irregular and undulating Me/BL interface. c When chloride ions attack the as-grown passive film, chloride ions only get to certain interfacial locations by heterogeneously penetrating the as-grown film along the connected path provided by the interfaces between nanocrystals and the amorphous zone. This gives rise to an undulating interface to the boundary between nano-crystalline oxide grains in those pronounced tension is imparted to the passive film as a result of works, different from the boundary between NCs and the interface undulating induced by chloride ions. amorphous zone we propose, it is strongly implied that full amorphization has greater tendency to resist localized attack due Chloride-induced inhomogeneity to the passive film. According to the absence of the ready path (grain boundaries) for outward to the image contrast in Fig. 6 and Supplementary Fig. 10, the 60,66–68 migration of cations or inward migration of anions . passive film formed in chloride-free solution appears quite uniform and compact, going from the homogeneous contrast of Interface strain distribution modeling. The as-prepared single the HAADF-STEM image. In contrast, the inhomogeneous con- crystal matrix provides us an opportunity to examine the strain trast in the film obtained in chloride-containing solution corre- state within the matrix side using a well-established LADIA sponds to a non-uniform and rather loose passive film, which simulation method based on the high-resolution HAADF-STEM is indisputably resultant from the chloride ion incorporation to images, since the local lattice distortion immediately below the the passive film. Especially at the inner layer, the image contrast interface would be induced if a stress exists at the interface. The is much darker, which indicates that the film is much looser strain state within the matrix side corresponding to the two types (note that the image is taken in the HAADF mode, and the of metal/film interfaces visualized in our study (i.e., straight and contrast is strongly associated with the mass density of heavy undulating interfaces) was modeled using the LADIA simulation atoms). Based on a combination of HAADF imaging and EDS method (Supplementary Methods). The results are shown in analysis of chloride concentration to the inner layer aforemen- Fig. 6 and Supplementary Fig. 10. The straight interface formed in tioned, we propose that some Me (OCl) species might be formed 3 n chloride-free electrolyte revealed no signs of lattice distortion on at the inner layer of the passive film in chloride-containing the matrix side (Fig. 6a, b). Conversely, the undulating interface environments. formed in chloride-containing electrolyte shows clear evidence of strain-induced lattice expansion on the matrix side (Fig. 6c, d). The local strain state that we have extracted is along the direction Discussion parallel to the interface of metal/oxide and meanwhile perpen- In chloride-free electrolyte, the austenitic matrix experienced dicular to the viewing direction (Supplementary Figure 11a and selective dissolution of both the major Fe and minor Cr com- 11b). It is proposed that the strain state along the viewing ponents in a somewhat homogeneous manner, promoting initial direction is also the case, since these two perpendicular directions formation of the inner Cr-rich oxides at the metal/solution (Me/ are crystallographically equivalent, as shown in Supplementary Sol) interface. The dissolved Fe and Cr cations were then Fig. 11c. So the local strain state is in two dimensions parallel to simultaneously hydrolyzed and subsequently re-deposited, to the metal/film interface. The lattice expansion in the matrix side yield the outer Fe-rich precipitation layer, as illustrated in Fig. 7a. means a tensile strain at the interface. Correspondingly, a At this stage, the film growth process involving transport of both 6 NATURE COMMUNICATIONS | (2018) 9:2559 | DOI: 10.1038/s41467-018-04942-x | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04942-x ARTICLE injected metal ions from the matrix and oxygen in solution This accounts for the few select locations at the interface, where through the barrier layer was faster than the film dissolution chloride ion accumulation was detected, with corresponding process, causing the metal/film (Me/BL) interface to move toward interface undulation. the metal matrix side. With time, dynamic equilibrium would be According to the large amount of observation in this study, we attained, where the velocity of film growth equaled that of film do find that the role of interface roughening is quite non-uniform. dissolution. Even though the interfaces of Me/BL can still keep In some locations, the roughening of interface is very consider- moving at dynamic equilibrium, the thickness of the passive film able, as shown in the HAADF-STEM image of Supplementary remained more or less constant. These processes thus yielded a Fig. 12. In the zoom-in image Supplementary Fig. 12(b), which is straight Me/BL interface at the atomic scale, as illustrated in an enlarged image of the area marked with a rectangular in (a), Fig. 7a. the convex site at the matrix side (featuring with the well-defined The propensity of chloride ions to be preferentially adsorbed at lattice images) is getting close to the outer surface of the passive certain non-uniformly distributed distinct defective locations film, wherein, accordingly, the thickness of the passive film is means that the chloride ions would be non-homogeneously quite thin. It is reasonable to propose that, with further rough- adsorbed on the bare metal surface and could modify interfacial ening, the film would become thinner and thinner. Under the process mechanisms in several ways. For one, chloride ions assistance of the stress, the film would be pulled apart and finally particularly promote Fe and to some extent Cr dissolution by a breakdown would occur. coordination: Pit nucleation, initiated at the surface of high purity metals or even of single crystals, is generally thought to be random and n1 Me þ Cl ¼ MeCl ð1Þ unpredictable. Our present experimental results and the analysis above indicate that convex sites where the nature of the passive The non-homogeneous nature of Cl adsorption on Fe should film facilitates its amplification would be the preferential sites for therefore also induce non-homogeneous and irregular Fe cation film breakdown and pit nucleation. injection from substrate. Since the passivity of stainless steels is Interestingly, it becomes obvious from our atomic-scale ana- controlled primarily by the selective dissolution of iron ,we lysis of the chloride attack mechanism that the film breakdown believe that the chloride-induced non-uniform cation injection of sites are not really locations with high chloride ion concentration iron would obviously yield a passive film with an irregular and (concave sites) as widely believed but are actually the adjacent undulating Me/BL interface, as illustrated in Fig. 7b, wherein a locations where the effect of chloride ions are relatively weak large amount of chloride adsorption yields a faster growth of film, (convex sites). This idea, which is based on our experimental in contrast, chloride-free or little chloride leads to a slower findings, introduces another dimension to understanding the growth. As a result, the concave sites, derived from the chloride mechanism of chloride attacking the passive film and implying attack, and the convex sites, with no chloride or little chloride, are the mechanism of chloride-induced passivity breakdown. formed. One plausible mechanism for the observed lattice expansion As the oxide film thickens progressively, whether or not the involves penetration and occupation of vacancies in the metal concave sites are more aggressively attacked (yielding a faster film lattice by chloride (either chloride ions and/or Cl atoms), which growth) or the convex sites more gently attacked (leading to a both possess larger radii than Fe and Cr and as such can induce slower film growth), is decided by the facility of chloride ion lattice expansion. Chloride can modify processes occurring at the permeation. That depends on two aspects: one is the electric field interface when Cl adatoms (from Cl ion reduction) penetrate and occupy the residual Fe vacancies from the cation injection, and the other is the intrinsic nature of the new oxide film. The 69,70 electric field in the oxide varies during film growth . The thereby causing lattice expansion. Correspondingly, no lattice concave interfaces in the film are notably thicker and possess expansion or slight lattice expansion happens at the convex sites, weaker electric field than the convex interface and hence pose while remarkable lattice expansion occurs at the concave sites greater resistance to chloride ion permeation. Correspondingly, (schematically illustrated in Supplementary Fig. 13). It is worth parts of the metal substrate directly beneath the concave positions mentioning that the lattice behaviors based on our LADIA in the undulating film will experience less severe cation injection simulation reflect the atomic projection along the [100] direction. process, which retards film thickening. In contrast, cation injec- Such a projection makes the outmost layer at the matrix side tion process of the metal substrate beneath the convex positions is actually correspond to the convex area with no or little lattice more pronounced, with the higher electric field speeding up the expansion, yielding the lattice expansion to be below the interface, thickening of the passive film. Evidently, the effect of the electric rather than at the interface (Fig. 6b and Supplementary Fig. 13). field is against further amplifying the amplitude between the Regarding the chloride presence below the interface at a depth convex and concave locations. where lattice expansion is measured, we did identify the chloride Nevertheless, if the intrinsic nature of the thicker passive film signal that is derived from the elemental maps, as shown in located at some concave interfaces facilitates the chloride per- Supplementary Figure 14, although its intensity is much lower meation, namely, it is less compact or exhibits more connected than that in the inner layer of the passive film. paths created along the NCs/amorphous interface, those concave All of our experimental results indicate clearly that chloride sites can be expected to still keep a faster film growth, and vice ions remarkably and unilaterally modify the interface zones via versa for the convex interface. The cumulative impact of the lattice expansion on the metal side and induced undulations at occurrences ought to amplify some undulations (concave/convex the interface and structural inhomogeneity on the film side. Such locations), and the convex interfaces have the tendency to move a series of events, by which chloride ions incorporate and attack closer and closer to the outer surface of the passive film, as illu- the passive film, were neither envisaged nor considered probable strated in Fig. 7b. in the available theories describing chloride-induced passivity Another mode of chloride attack, involving permeation breakdown. through pathways along the NCs/amorphous zone interface, as We have shown that the passive films are mostly amorphous mentioned earlier, is illustrated in Fig. 7c. Such a mode is only but with small amount of embedded NCs and that interfaces feasible when a connected path exists through the film. As shown between NCs and the amorphous zone assume the features of in Fig. 7c, chloride ions will permeate the passive film laterally grain boundaries providing ready paths for the chloride ion through the red tunnel and arrive at the matrix/ film interface. transport. We desire a passive film with full amorphous structure, NATURE COMMUNICATIONS | (2018) 9:2559 | DOI: 10.1038/s41467-018-04942-x | www.nature.com/naturecommunications 7 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04942-x so that tunnels for species diffusion and transport are not avail- 2. Hoar, T. P. Production and breakdown of passivity of metals. Corros. Sci. 7, 341–355 (1967). able. We could apply microalloying via an adding of certain 3. Sato, N. Theory for breakdown of anodic oxide films on metals. Electrochim. element(s), which enhance the degree of amorphization of the Acta 16, 1683–1692 (1971). passive film and in the meanwhile the properties of bulk steels 4. Galvele, J. R. Transport processes and mechanism of pitting of metals. remain unchanged. The possibility of this proposal is under J. Electrochem. Soc. 123, 464–474 (1976). consideration and there is no doubt that the newly reported 5. Lin, L. F., Chao, C. Y. & Macdonald, D. D. A point-defect model for anodic passive films. II. Chemical breakdown and pit initiation. J. Electrochem. Soc. mechanism will prompt scientists and engineers to reconsider the 128, 1194–1198 (1981). existing models and look for all the possible approaches to retard 6. Sato, N. Anodic breakdown of passive films on metals. J. Electrochem. Soc. the passivity breakdown. 129, 255–260 (1982). In summary, this study has provided new experimental insights 7. Macdougall, B., Mitchell, D. F., Sproule, G. I. & Graham, M. J. Incorporation on the well-known role of chloride ions in passivity breakdown, of chloride ion in passive oxide-films on nickel. J. Electrochem. Soc. 130, 543–547 (1983). where the existing ideas are mostly based on theoretical models, 8. Murphy, O. J., Bockris, J. O., Pou, T. E., Tongson, L. L. & Monkowski, M. D. with insufficient experimental evidence. Using spherical Cs- Chloride ion penetration of passive films on iron. J. Electrochem. Soc. 130, corrected TEM, in combination with computational modeling, we 1792–1794 (1983). have compared the passive films formed in chloride-free and 9. Di Quarto, F., Piazza, S. & Sunseri, C. Electrical and mechanical breakdown of -containing electrolytes and directly observed atomic-scale accu- anodic films on tungsten in aqueous electrolytes. J. Electroanal. Chem. Inter. Electrochem 248,99–115 (1988). mulation of chloride ions at the metal/film interface, including 10. Macdonald, D. D. The point-defect model for the passive state. J. Electrochem. the chloride-induced lattice expansion on the metal side, inter- Soc. 139, 3434–3449 (1992). facial undulations, and structural alterations to the film. We find 11. Marcus, P. & Herbelin, J. M. The entry of chloride ions into passive films that the passive film is mostly amorphous, with some NCs. We on nickel studied by spectroscopic (ESCA) and nuclear (Cl-36 radiotracer) directly visualize, from the in-plane and also the out-of-plane methods. Corros. Sci. 34, 1123–1145 (1993). 12. Xu, Y., Wang, M. H. & Pickering, H. W. On electric-field-induced breakdown direction, the size, distribution, and crystallographic orientation of passive films and the mechanism of pitting corrosion. J. Electrochem. Soc. of NCs and figure out the boundaries between NCs and the 140, 3448–3457 (1993). amorphous phase. Our experimental and computational results 13. Macdonald, D. D. Passivity - the key to our metals-based civilization. Pure suggest that the interface between NCs and the amorphous zone Appl. Chem. 71, 951–978 (1999). assume a special kind of grain boundaries and thus provide ready 14. Hubschmid, C., Landolt, D. & Mathieu, H. J. XPS and AES analysis of passive films on Fe-25Cr-X (X=Mo, V, Si and Nb) model alloys. Fresenius. J. Anal. paths for chloride ion transport. Our results show clearly that Chem. 353, 234–239 (1995). chloride ions actually permeate the outer and inner layers to 15. Brooks, A. R., Clayton, C. R., Doss, K. & Lu, Y. C. On the role of Cr in the attack the interface, rendering it indistinct and undulating. The passivity of stainless-steel. J. Electrochem. Soc. 133, 2459–2464 (1986). weakest site which is believed to be the preferential site for a 16. Mischler, S., Vogel, A., Mathieu, H. J. & Landolt, D. The chemical breakdown does not coincide, at atomic scale, with locations composition of the passive film on Fe-24Cr and Fe-24Cr-11Mo studied by AES, XPS and SIMS. Corros. Sci. 32, 925–944 (1991). having high chloride ion concentration as widely believed but 17. Ningshen, S., Mudali, U. K., Mittal, V. K. & Khatak, H. S. Semiconducting and occurs at adjacent locations with subdued chloride ion influence. passive film properties of nitrogen-containing type 316LN stainless steels. Corros. Sci. 49, 481–496 (2007). 18. Khalil, W., Haupt, S. & Strehblow, H. H. The thinning of the passive layer of Methods iron by halides. Werks. Korros. Mater. Corros. 36,16–21 (1985). Materials preparation. The FeCr Ni (wt.%) single crystal alloy was grown by 19. Landolt, D., Mischler, S., Vogel, A. & Mathieu, H. J. Chloride ion effects on 15 15 thermal-gradient directional solidification method. The orientation was determined passive films on FeCr and FeCrMo studied by AES, XPS and SIMS. Corros. Sci. by single-crystal X-ray diffractometry, and two low-index crystallographic orien- 31, 431–440 (1990). tations [001] and [110] were obtained. The (001) or (110) plane as exposure surface 20. Olsson, C. O. A. & Landolt, D. Passive films on stainless steels- chemistry, was potentiostatically passivated at a potential of 640 mV/SHE for 30 min in structure and growth. Electrochim. Acta 48, 1093–1104 (2003). chloride-free and -containing electrolyte. The cross-sectional TEM specimen was 21. Natishan, P. M. et al. Chloride interactions with the passive films on stainless prepared by the conventional method. Two passivated surfaces of two samples steel. J. Electrochem. Soc. 158,C7–C10 (2011). were bonded face-to-face and then thinned by grinding and ion-milling. 22. Mitchell, D. F. Quantitative interpretation of Auger sputter profiles of thin- layers. Appl. Surf. Sci. 9, 131–140 (1981). HAADF-STEM imaging. HAADF-STEM images were recorded using Cs-corrected 23. Olefjord, I., Brox, B. & Jelvestam, U. Surface-composition of stainless-steels during anodic-dissolution and passivation studied by ESCA. J. Electrochem. TEM (Titan Cubed 60–300 kV microscope (FEI) fitted with a high-brightness field- emission gun (X-FEG), double Cs corrector from CEOS, and a monochromator Soc. 132, 2854–2861 (1985). operating at 300 kV). The beam convergence is 25 mrad and thus yields a probe 24. Olefjord, I. & Wegrelius, L. Surface-analysis of passive state. Corros. Sci. 31, size of <0.1 nm. 89–98 (1990). 25. Goetz, R., Macdougall, B. & Graham, M. J. An AES and SIMS study of the influence of chloride on the passive oxide film on iron. Electrochim. Acta 31, LADIA calculation method. The LADIA package was used to analyze the elastic 1299–1303 (1986). strain state of the alloy underneath the passive thin film. This algorithm determines 26. Mitrovicscepanovic, V., Macdougall, B. & Graham, M. J. The effect of Cl ions the displacement of actual column image positions versus the position of a refer- on the passivation of Fe-26Cr alloy. Corros. Sci. 27, 239–247 (1987). ence lattice. From this information, we determined the local expansion/contraction 27. Hubschmid, C. & Landolt, D. Formation conditions, chloride content, and of next-neighbor atom column distances in the alloy, i.e., the local lattice para- stability of passive films on an iron-chromium alloy. J. Electrochem. Soc. 140, meters in the <110> directions orthogonal to the <001> viewing direction and the 1898–1902 (1993). <100> directions orthogonal to the <011> viewing direction. 28. Murayama, M., Makiishi, N., Yazawa, Y., Yokota, T. & Tsuzaki, K. Nano-scale chemical analyses of passivated surface layer on stainless steels. Corros. Sci. 48, Data availability. The data that support the findings of this study are available 1307–1318 (2006). from the corresponding author upon reasonable request. 29. Hamada, E. et al. Direct imaging of native passive film on stainless steel by aberration corrected STEM. Corros. Sci. 52, 3851–3854 (2010). 30. Soulas, R. et al. TEM investigations of the oxide layers formed on a 316L alloy Received: 30 September 2017 Accepted: 4 June 2018 in simulated PWR environment. J. Mater. Sci. 48, 2861–2871 (2013). 31. Oh, K., Ahn, S., Eom, K., Jung, K. & Kwon, H. Observation of passive films on Fe-20Cr-xNi (x=0, 10, 20 wt.%) alloys using TEM and Cs-corrected STEM- EELS. Corros. Sci. 79,34–40 (2014). References 32. Davenport, A. J., Oblonsky, L. J., Ryan, M. P. & Toney, M. F. The structure of 1. Hoar, T. P., Mears, D. C. & Rothwell, G. P. The relationships between anodic the passive film that forms on iron in aqueous environments. J. Electrochem. passivity, brightening and pitting. Corros. Sci. 5, 279–289 (1965). Soc. 147, 2162–2173 (2000). 8 NATURE COMMUNICATIONS | (2018) 9:2559 | DOI: 10.1038/s41467-018-04942-x | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04942-x ARTICLE 33. Rees, E. E., Ryan, M. P. & McPhail, D. S. An STM study of the nanocrystalline 62. Ryan, M. P., Newman, R. C. & Thompson, G. E. Atomically resolved STM structure of the passive film on iron. Electrochem. Solid State Lett. 5, B21–B23 of oxide film structures on Fe-Cr alloys during passivation in sulfuric-acid- (2002). solution. J. Electrochem. Soc. 141, L164–L165 (1994). 34. Keller, P. & Strehblow, H. H. XPS investigations of electrochemically formed 63. Hendy, S. C., Laycock, N. J. & Ryan, M. P. Atomistic modeling of cation passive layers on Fe/Cr-alloys in 0.5 M H SO . Corros. Sci. 46,1939–1952 (2004). transport in the passive film on iron and implications for models of growth 2 4 35. Rao, J. C., Zhang, X. X., Qin, B. & Fung, K. K. TEM study of the structural kinetics. J. Electrochem. Soc. 152, B271–B276 (2005). dependence of the epitaxial passive oxide films on crystal facets in polyhedral 64. Marcus, P., Maurice, V. & Strehblow, H. H. Localized corrosion (pitting): a nanoparticles of chromium. Ultramicroscopy 98, 231–238 (2004). model of passivity breakdown including the role of the oxide layer nanostructure. 36. Seyeux, A., Maurice, V., Klein, L. H. & Marcus, P. In situ STM study of the Corros. Sci. 50, 2698–2704 (2008). effect of chloride on passive film on nickel in alkaline solution. J. Electrochem. 65. Seyeux, A., Maurice, V. & Marcus, P. Breakdown kinetics at nanostructure Soc. 153, B453–B463 (2006). defects of passive films. Electrochem. Solid State Lett. 12, C25–C27 (2009). 37. Macdonald, D. D. The history of the point defect model for the passive state: 66. Okamoto, G. Passive film of 18-8 stainless-steel structure and its function. A brief review of film growth aspects. Electrochim. Acta 56, 1761–1772 (2011). Corros. Sci. 13, 471–489 (1973). 38. Padhy, N., Paul, R., Mudali, U. K. & Raj, B. Morphological and compositional 67. Leach, J. S. L. The role of surface films in corrosion and oxidation. Surf. Sci. analysis of passive film on austenitic stainless steel in nitric acid medium. 53, 257–271 (1975). Appl. Surf. Sci. 257, 5088–5097 (2011). 68. Bertocci, U. & Kruger, J. Studies of passive film breakdown by detection and 39. Jung, R. H., Tsuchiya, H. & Fujimoto, S. XPS characterization of passive films analysis of electrochemical noise. Surf. Sci. 101, 608–618 (1980). formed on type 304 stainless steel in humid atmosphere. Corros. Sci. 58,62–68 69. Diawara, B., Beh, Y.-A. & Marcus, P. Nucleation and growth of oxide layers on (2012). stainless steels (FeCr) using a virtual oxide layer model. J. Phys. Chem. C 114, 40. Maurice, V. & Marcus, P. Passive films at the nanoscale. Electrochim. Acta 84, 19299–19307 (2010). 129–138 (2012). 70. Seyeux, A., Maurice, V. & Marcus, P. Oxide film growth kinetics on metals 41. Leistner, K., Toulemonde, C., Diawara, B., Seyeux, A. & Marcus, P. Oxide film and alloys I. Physical model. J. Electrochem. Soc. 160, C189–C196 (2013). growth kinetics on metals and alloys II. Numerical simulation of transient behavior. J. Electrochem. Soc. 160, C197–C205 (2013). Acknowledgements 42. Massoud, T., Maurice, V., Klein, L. H. & Marcus, P. Nanoscale morphology This work is supported by the National Natural Science Foundation of China (Nos. and atomic structure of passive films on stainless steel. J. Electrochem. Soc. 51771212, 11327901, 51390473), the Key Research Program of Frontier Sciences CAS 160, C232–C238 (2013). (QYZDJ-SSW-JSC010), the Innovation Fund in IMR (2017-ZD05). E.E.O. is grateful to 43. Fajardo, S. et al. Low energy SIMS characterization of passive oxide films formed the Chinese Academy of Sciences (CAS) for award of the CAS President’s Fellowship. on a low-nickel stainless steel in alkaline media. Appl. Surf. Sci. 288,423–429 The authors are grateful to Professor H. Wei for the help with the single crystal growth, (2014). Dr. B. J. Wang at Thermo Fisher Scientific Shanghai Nanoport for Super-X EDS 44. Macdonald, D. D. Passivity: enabler of our metals based civilisation. Corros. mapping, Professor X. P. Song for the orientation determination of the single crystal, Eng. Sci. Technol. 49, 143–155 (2014). Mr. T. F. Du for specimens cutting, and Miss X. X. Wei for some experimental 45. Strehblow, H. H. Passivity of metals studied by surface analytical methods, a supplement during addressing the reviewers’ comments. review. Electrochim. Acta 212, 630–648 (2016). 46. Vignal, V. et al. Influence of long-term ageing in solution containing chloride ions on the passivity and the corrosion resistance of duplex stainless steels. Author contributions Corros. Sci. 53, 894–903 (2011). X.L.M. and B.Z. conceived the project of transmission electron microscopy in corrosion 47. Calinski, C. & Strehblow, H. H. ISS depth profiles of the passive layer on Fe/Cr science and designed the experiments and simulations. J.W. and B.Z. conducted the TEM alloys. J. Electrochem. Soc. 136, 1328–1331 (1989). observations, and B.Z. carried out the electrochemical experiments. B.W. provided the 48. Castle, J. E. & Qiu, J. H. The application of ICP-MS and XPS to studies of imaging technique support on the Titan G2 60–300 platform of the aberration-corrected ion selectivity during passivation of stainless-steels. J. Electrochem. Soc. 137, scanning transmission electron microscope. B.Z., J.W., X.L.M., and E.E.O. analyzed the 2031–2038 (1990). data and co-wrote the manuscript. X.W.G., Y.J.W., and D.C. performed the first- 49. Yang, W. P., Costa, D. & Marcus, P. Resistance to pitting and chemical- principle calculations, and Y.C.Z and K.D. carried out the LADIA simulation. All authors composition of passive films of a Fe-17-percent-Cr alloy in chloride-containing contributed to the discussions and manuscript preparation. acid-solution. J. Electrochem. Soc. 141, 2669–2676 (1994). 50. Maurice, V., Yang, W. P. & Marcus, P. XPS and STM study of passive films formed on Fe-22Cr (110) single-crystal surfaces. J. Electrochem. Soc. 143, Additional information 1182–1200 (1996). Supplementary Information accompanies this paper at https://doi.org/10.1038/s41467- 51. Maurice, V., Yang, W. P. & Marcus, P. X-ray photoelectron spectroscopy 018-04942-x. and scanning tunneling microscopy study of passive films formed on (100) Fe-18Cr-13Ni single-crystal surfaces. J. Electrochem. Soc. 145, 909–920 (1998). Competing interests: The authors declare no competing interests. 52. Szklarska-Smialowska, Z. Mechanism of pit nucleation by electrical breakdown of the passive film. Corros. Sci. 44, 1143–1149 (2002). Reprints and permission information is available online at http://npg.nature.com/ 53. Evans, U. R. The passivity of metals. Part I: the isolation of the protective film. reprintsandpermissions/ J. Chem. Soc. 1, 1020–1040 (1927). 54. Vernon, W. H. J., Wormwell, F. & Nurse, T. J. The thickness of air-formed Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in oxide films on iron. J. Chem. Soc. 0, 621–632 (1939). published maps and institutional affiliations. 55. Toney, M. F., Davenport, A. J., Oblonsky, L. J., Ryan, M. P. & Vitus, C. M. Atomic structure of the passive oxide film formed on iron. Phys. Rev. Lett. 79, 4282–4285 (1997). 56. Mayne, J. E. O. & Pryor, M. J. The mechanism of inhibition of corrosion of Open Access This article is licensed under a Creative Commons iron by chromic acid and potassium chromate. J. Chem. Soc. 0, 1831–1835 (1949). Attribution 4.0 International License, which permits use, sharing, 57. Foley, C. L., Kruger, J. & Bechtoldt, C. J. Electron diffraction studies of active adaptation, distribution and reproduction in any medium or format, as long as you give passive and transpassive oxide films formed on iron. J. Electrochem. Soc. 114, appropriate credit to the original author(s) and the source, provide a link to the Creative 994–1001 (1967). Commons license, and indicate if changes were made. The images or other third party 58. McBee, C. L. & Kruger, J. Nature of passive films on iron-chromium alloys. material in this article are included in the article’s Creative Commons license, unless Electrochim. Acta 17, 1337–1341 (1972). indicated otherwise in a credit line to the material. If material is not included in the 59. Ryan,M.P., Newman,R.C.&Thompson,G. E. AnSTM study of the article’s Creative Commons license and your intended use is not permitted by statutory passive film formed on iron in borate buffer solution. J. Electrochem. Soc. regulation or exceeds the permitted use, you will need to obtain permission directly from 142,L177–L179 (1995). the copyright holder. To view a copy of this license, visit http://creativecommons.org/ 60. Maurice, V., Yang, W. P. & Marcus, P. XPS and STM investigation of the licenses/by/4.0/. passive film formed on Cr (110) single-crystal surfaces. J. Electrochem. Soc. 141, 3016–3027 (1994). 61. Ryan,M. P., Newman,R.C. & Thompson, G.E. A scanning tunneling © The Author(s) 2018 microscopy study of structure and structural relaxation in passive oxide- films on Fe-Cr alloys. Philos. Mag. B 70, 241–251 (1994). NATURE COMMUNICATIONS | (2018) 9:2559 | DOI: 10.1038/s41467-018-04942-x | www.nature.com/naturecommunications 9 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Nature Communications Springer Journals

Unmasking chloride attack on the passive film of metals

Loading next page...
 
/lp/springer-journals/unmasking-chloride-attack-on-the-passive-film-of-metals-b4nkcXp2IP

References (75)

Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s)
Subject
Science, Humanities and Social Sciences, multidisciplinary; Science, Humanities and Social Sciences, multidisciplinary; Science, multidisciplinary
eISSN
2041-1723
DOI
10.1038/s41467-018-04942-x
Publisher site
See Article on Publisher Site

Abstract

ARTICLE DOI: 10.1038/s41467-018-04942-x OPEN Unmasking chloride attack on the passive film of metals 1 1 1 1 1 1 1 1 2 1,3 B. Zhang , J. Wang ,B.Wu , X. W. Guo , Y. J. Wang , D. Chen , Y. C. Zhang ,K.Du , E. E. Oguzie &X.L.Ma Nanometer-thick passive films on metals usually impart remarkable resistance to general corrosion but are susceptible to localized attack in certain aggressive media, leading to material failure with pronounced adverse economic and safety consequences. Over the past decades, several classic theories have been proposed and accepted, based on hypotheses and theoretical models, and oftentimes, not sufficiently nor directly corroborated by experimental evidence. Here we show experimental results on the structure of the passive film formed on a FeCr Ni single crystal in chloride-free and chloride-containing media. We use 15 15 aberration-corrected transmission electron microscopy to directly capture the chloride ion accumulation at the metal/film interface, lattice expansion on the metal side, undulations at the interface, and structural inhomogeneity on the film side, most of which had previously been rejected by existing models. This work unmasks, at the atomic scale, the mechanism of chloride-induced passivity breakdown that is known to occur in various metallic materials. Shenyang National Laboratory for Materials Science, Institute of Metal Research, Chinese Academy of Sciences, Wenhua Road 72, 110016 Shenyang, China. Electrochemistry and Materials Science Research Laboratory, Department of Chemistry, Federal University of Technology Owerri, PMB, Owerri 1526, Nigeria. State Key Lab of Advanced Processing and Recycling on Non-ferrous Metals, Lanzhou University of Technology, 730050 Lanzhou, China. These authors contributed equally: B. Zhang, J. Wang. Correspondence and requests for materials should be addressed to X.L.M. (email: xlma@imr.ac.cn) NATURE COMMUNICATIONS | (2018) 9:2559 | DOI: 10.1038/s41467-018-04942-x | www.nature.com/naturecommunications 1 1234567890():,; ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04942-x orrosion is one of the major causes of material Results failure and hence leads to a huge cost to our society . Sample preparation under various chemical conditions. Passive CThe nanometer-thick passive film on metals films were formed on the (001) and (110) plane, respectively, of resists a general corrosion, but it is susceptible to severe FeCr Ni single crystal (Supplementary Note 1 and Note 2, 15 15 localized attack in certain aggressive media . The best-known Supplementary Figs 1–3). This enabled us to obtain a distinct inducer of localized passive film breakdown is the chloride ion. metal/passive film interface (Supplementary Fig. 4) and better Despite the enormous amount of experimental data and diverse characterize the structure of the interface region. On the other 1–13 hypotheses and models proposed till date , the breakdown of hand, the single crystal, which is free of any inclusions and grain the passive film is still not sufficiently understood and remains boundaries, yields a high-quality passive film with a continuous one of the most important and basic problems in corrosion coverage on the alloy matrix. It also effectively avoids the para- science. digm of the weakest sites breaking down the soonest, which The lack of agreement on the mechanism of passive film makes figuring out the intrinsic mechanism complex (Supple- breakdown is mainly due to the difficulty encountered in mentary Note 2). In order to monitor the transport and effect of obtaining precise experimental information. To clarify the chloride ions, passive films were formed under three designated exact nature of the chloride-induced breakdown, Cl incor- conditions: passivation in H SO electrolyte, passivation in 2 4 poration to the film has to be experimentally confirmed H SO + NaCl electrolyte, and initial passivation in H SO 2 4 2 4 and the accurate location needs to be identified. In the electrolyte and subsequent addition of NaCl into the H SO 2 4 meanwhile, chloride-induced modification to the film has to electrolyte (Supplementary Note 3, Supplementary Figs 5 and 6). be experimentally addressed as well. It is worthy of note The cross-sectional TEM specimen was prepared by the con- that these issues were extensively studied by X-ray photoelec- ventional method, that is, passivated surfaces of two samples were 7,14–21 tron spectroscopy (XPS) , Auger electron spectroscopy bonded face-to-face and then thinned by grinding and ion- 7,14,16,19,22–27 7,8,16,19,25 (AES) , secondary ion mass spectrometry , milling. During sample preparation and subsequent TEM and radiotracer techniques . Nonetheless, it is still very difficult observation, the extremely thin passive film was strictly ensured and challenging to guarantee the precision and accuracy of free of mechanical and beam-induced damage (Supplementary observed locations and concentrations of a very small amount of Note 4). chloride in an extremely thin passive film with a thickness of only a few nanometers. Much of evidence on the incorporation of Cl Structural evolution of passive film with aging in air. The TEM in the passive oxide film can be mainly classified into two groups: observation in the high-angle annular-dark-field (HAADF) mode 7,8,11,14–17,22–24 one is chloride incorporation , and the other is and Super-X EDS analysis on the passive film were performed 14,18–21,25–27 chloride absence in the passive film . In the case of and the results are shown in Figs. 1 and 2. Figure 1a is the incorporation, the location of chloride in the passive film is also HAADF scanning transmission electron microscopic (HAADF- controversy. Some investigators declare that Cl locate or con- STEM) image showing the passive film on FeCr Ni single 15 15 7,14–16,22,23 centrate in the outer layer of the film , whereas some crystal formed in H SO electrolyte (condition 1). According to 2 4 8,24 others claim it is in the inner layer . In reality, most of the the contrast difference, the film seems to be tri-layer structured. reported methods do not directly identify the presence of chloride Whereas the EDS mapping analysis (Fig. 2a) indicates a well- or its location within the passive film. Alongside questions defined bi-layer structure with the inner Cr-rich layer and the regarding chloride ion distribution in the passive film are issues 29,47–51 outer Fe-rich layer, as generally accepted . Interestingly, relating to the nature of atomic-scale interactions between the after aging the specimen in air for a few days, we also subjected passive film and chloride, i.e., issues regarding how, where, and the aged specimen again to TEM observation and we found that when the chloride modifies the passive film. To the best of our the contrast had become homogeneous, following disappearance knowledge, although the experimental advances, including of the middle darker-contrast layer (Fig. 1b). The implication 28–31 transmission electron microscopic (TEM) characterization , is that the outer layer of the passive film, formed by precipitation promote powerful evidence on the structure and chemistry of the of the hydrolyzed metal cations, is transformed to the metal 20,28,29,31–45 passive film as well as the evolution imparted by oxide by a dehydration reaction, which was confirmed by XPS 36,39,43,46 34,38,39,50,51 chloride , no technique has succeeded in directly fol- analysis . It is worthwhile to mention that the HAADF lowing the evolution of the passive film, in chloride-containing mode image provides an incoherent image using high-angle media, across the entire film ranging from the surface to the scattered electrons, where the contrast is strongly dependent on metal/film interface. the scattering ability of heavy atoms. Thus the much denser metal Without doubt, deciphering the interactions of the chloride matrix would show the brightest contrast, followed by the metal ion with the passive film, including chloride-induced modifica- oxide, with metal hydroxide displaying the darkest contrast. tions to the properties of the passive film at the atomic scale, From the foregoing and taking into consideration of is key to understanding the precise mechanism of passivity the EDS mapping results, the inner layer is a Cr-rich oxide, while breakdown. In the present work, using aberration-corrected the outer layer, with the darkest contrast should be Fe-rich TEM (Cs-corrected TEM) and a fast and precise super X-ray energy-dispersive spectrometer (Super-X EDS) analysis with four detectors, we simultaneously investigate the film and metal matrix as well as their interface via cross-sectioning in real Glue space. We find the chloride accumulation within the inner layer Film of the passive film and the associated fluctuations at the matrix/ passive film interface. We provide direct evidence on the Matrix location of chloride and the resultant phenomena of the lattice expansion on the metal side, undulations at the interface, and structural inhomogeneity on the film side. The present findings Fig. 1 HAADF-STEM image showing the contrast evolution of the passive allow for the atomic-scale mechanism of passivity breakdown film on FeCr Ni single crystal on exposure to air. Scale bar, 5 nm. a Image 15 15 to be revisited on the basis of real-space imaging in multi- immediately after the specimen was prepared. b Image after the specimen dimensions. has been exposed in air for about 6 days 2 NATURE COMMUNICATIONS | (2018) 9:2559 | DOI: 10.1038/s41467-018-04942-x | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04942-x ARTICLE Film Matrix Cl O Ni Cr Fe Film Matrix Cl O Cr Ni Fe Film Matrix Cl O Ni Fe Cr Fig. 2 Super-X EDS mapping showing chloride ion incorporated in and penetrating the passive film and accumulating at the matrix/passive film interface. −1 −1 −1 Element maps of the film formed in a 0.5 mol L H SO electrolyte at 640 mV/SHE for 30 min; b 0.5 mol L H SO + 0.3 mol L NaCl electrolyte at 2 4 2 4 −1 640 mV/SHE for 30 min, and c passivated in 0.5 mol L H SO electrolyte at 640 mV/SHE for 30 min with subsequent addition of NaCl. All scale bars in 2 4 a–c are 2 nm a b (110) (001) cd (110) (001) Fig. 3 Chloride-induced undulations at the metal/passive film interface, wherein the yellow line indicates the metal/passive film interface. Scale bar, 2 nm. a, b HRTEM images along the [001] and [110] axes of the austenitic matrix showing the passive film grown on (110) and (001) planes of FeCr Ni single 15 15 −1 crystal in 0.5 mol L H SO electrolyte, revealing a sharp and straight interface, even at the atomic scale. c HRTEM images along [001] axis showing the 2 4 −1 −1 passive film grown on (110) plane in 0.3 mol L NaCl + 0.5 mol L H SO electrolyte, revealing an indistinct and substantially undulating interface. d 2 4 −1 HRTEM images along [110] axis showing the passive film initially grown in 0.5 mol L H SO electrolyte for 30 min with subsequent addition of NaCl into 2 4 the H SO electrolyte. The interface is as well indistinct and substantially undulating 2 4 hydroxides. The unavoidable exposure in air going from passi- number of Fe and Cr is rather close, making the contrast of Cr- vation in the electrolyte to TEM observation allows the outmost rich oxide layer and Fe-rich oxide layer seems homogeneous. The hydroxide layer to be partially dehydrated yielding the brighter above results provide direct experiment evidence of a dehydration oxide, thus giving the film an apparent tri-layered structure in the model. HAADF-STEM image. With prolonged exposure, the hydroxide layer became completely dehydrated and the passive film corre- Structural evolution of the interface imparted by chloride. spondingly reverted to an oxide film. Although the oxide film has Prior to studying the interactions of chloride ions with the passive a bi-layered structure distinguished from the Cr-rich and Fe-rich film and austenitic matrix, the pristine passive film formed in layer (Fig. 2a), it can be hardly identified in the image taken with chloride-free electrolyte was analyzed, in which attention was the HAADF mode (Fig. 1b), since the contrast in the HAADF paid to the structure and the element distribution within the film image is strongly dependent on the scattering ability of heavy and near the interfaces. Cs-corrected TEM revealed a passive atoms that is associated with the atomic number. Here the atomic film with thickness of about 4–5 nm (see TEM image in NATURE COMMUNICATIONS | (2018) 9:2559 | DOI: 10.1038/s41467-018-04942-x | www.nature.com/naturecommunications 3 (010) (110) (110) F (112) (100) (001) (110) (110) ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04942-x Supplementary Fig. 4). The corresponding EDS maps are shown in Fig. 2a. A well-defined bi-layer passive film is seen, where the inner (barrier) layer adjacent to the metal is enriched in Cr and (110) depleted in Fe, while the outer (precipitated) layer is depleted in Cr and enriched in Fe. It is noteworthy that the film/matrix interface is sharp, well defined, and straight, even at the atomic level. On the matrix side, immediately adjacent to the passive film/matrix interface there exists a layer with Cr depletion and Ni enrichment, which is in agreement with the previous find- 29,31,48,51 ings . Cross-sectional high-resolution TEM (HRTEM) images, obtained along the [001] and [110] axes (Fig. 3a, b) of the austenitic matrix, respectively, show that the passive film is mostly amorphous. c The passive film formed in chloride-containing H SO solution 2 4 (condition 2) is shown in Fig. 3c. It is immediately obvious that the previously sharp and straight interface in chloride-free (110) solution has become indistinct and substantially undulating. This (001) is indisputably a consequence of chloride ion attack. By means of Super-X EDS mapping experiments, we successfully obtained the distribution of elemental Cl in the passive film, as shown in Fig. 2b, which, interestingly, shows Cl to be concentrated within the inner layer, (average content ~1 atomic %). Supplementary Figure 7 also shows a distinctive Cl peak in the composition spectrum. Significantly lower amounts of Cl (0.1 atomic %) were Fig. 4 Cross-sectional HRTEM images along the [001] and [110] axes of detected in the outer layer. Evidently, these findings indicate that the austenitic matrix showing the interface between the passive film and the chloride ions incorporate in the passive film, permeate the steel matrix. It is seen that the passive film is mainly amorphous, with some outer and inner layers, and ultimately attack the interface, giving nanocrystals. A series of HRTEM images obtained from variant orientations rise to the undulating interfacial structure. Such an incorporation and locations indicate that the nanocrystals feature face-centered cubic of solution species in the inner layer has not been previously structure. In most cases, the nanocrystals have crystallographic observed and is actually considered non-feasible by some existing 5,10,13,14,52 orientations with austenitic matrix but occasionally not. This figure displays theories on passivity breakdown . typical configurations where the orientation relationships between the We designed another experimental procedure, wherein the nanocrystals in the passive film and the single-crystalline substrate are FeCr Ni single crystal was initially passivated in H SO 15 15 2 4 labeled. a Film (110) || FeCr Ni (110) and film (1–10) || FeCr Ni (1–10). 15 15 15 15 electrolyte for 30 min and subsequently NaCl was added into b The crystal is randomly oriented with no orientation relationship with the the H SO electrolyte (condition 3). This enabled us to monitor 2 4 austenitic matrix. c Film (110) || FeCr Ni (110) and film (1–10) || FeCr Ni 15 15 15 15 the attack of chloride ions on the as-grown passive film. By (1–12). d Film (001) || FeCr Ni (001) and film (1–10) || FeCr Ni (1–10). 15 15 15 15 careful examination of a series of interfaces, what we observed is a Scale bar, 1 nm reduced prevalence of the undulating interface, which only manifested at a few locations (Fig. 3d). These locations are expected to be terminal points for paths through which chloride austenitic matrix but occasionally not. Thus the grain boundary ions permeate. Correspondingly and maybe not unexpectedly, Cl between the NCs and amorphous phase can be readily identified was only detected at locations manifesting the undulating according to the HRTEM images. Nevertheless, a determination of interfaces, typically like that in Fig. 2c and was not detected atomic configurations at this kind of boundary is a challenge, (Supplementary Fig. 8) at the still distinct and unperturbed which should be more complex than that of two crystalline grains. interfaces (similar to that in Fig. 2a). It is thus obvious that We hypothesize that the interfaces between the NCs and the chloride ions only get to certain interfacial locations by amorphous zone assume the features of atomic irregularity and heterogeneously penetrating the as-grown film. thus provide ready paths for chloride ion transport. The selective permeation implied by our results arises from the inherently non-homogeneous microstructure of passive films and Nanocrystal-amorphous interface assisting Cl permeation.It depends on the nature of and interconnection between the paths is well known that grain boundaries exhibit an irregular atom created along the NCs/amorphous zone interface. When a array yielding a loose structure that usually provide tunnels for connected path traverses the entire thickness of the passive film, species diffusion and transport. In contrast, the amorphous phase chloride ions tunneling through those paths would eventually always features a random atom distribution, so the diffusion and arrive at the matrix/passive film interface. On the other hand, transport processes in amorphous materials are not yet well where no connected paths exist, or all the paths are abridged, the understood. It is noteworthy that, although the passive films in the matrix/passive film interface would remain unperturbed because present study are mostly amorphous as seen in the HRTEM chloride ions are unable to get through. images, some periodic two-dimensional lattices with a scale of 1–3 In order to provide a theoretical basis for the above assertions, nm are often visualized, that is to say, a small amount of nano- we performed first-principle computations to model the diffusion crystals (NCs) are embedded in the passive films. Figure 4a–d of chloride ions within the passive film. The supercells constructed display typical cross-sectional HRTEM images in which some NCs for the simulation contain the NC structure, the amorphous zone, inherently present in the amorphous passive film. A series of and the amorphous zone/NC interface, respectively. This enables a HRTEM images obtained from variant orientations and locations quantification of the diffusion barriers to chloride ion diffusion indicate that NCs feature face-centered cubic structure (also seen from one oxygen vacancy to the next, as illustrated in Fig. 5. in Supplementary Fig. 9). In our present study, the NCs often Details of the computation procedure can be found in display specific crystallographic orientation relationships with the Supplementary Methods. Actually, for a qualitatively comparison, 4 NATURE COMMUNICATIONS | (2018) 9:2559 | DOI: 10.1038/s41467-018-04942-x | www.nature.com/naturecommunications (110) M NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04942-x ARTICLE the trend of diffusion barriers in these three zones has nothing to In earlier studies, several techniques, including reciprocal space 32,53–55 do with the specific crystal structure. So, for simplicity in the identification by XRD analysis , electron diffraction in a 56–58 calculations, we selected the binary spinel Fe O representing the TEM , and real-space imaging upon the outmost surface 3 4 33,42,50,51,59–62 spinel-structured NC embedded in the passive film. By comparing by scanning tuneling microscopic technique , have the energy barriers for Cl ion diffusion from one oxygen vacancy been applied in determining structural information on the nano- to its neighboring one, our computations show the diffusion crystalline nature of the passive films, and the role barrier to be lowest at the interface region, thus confirming that of grain boundaries in passivity breakdown and initiation of 63–65 the interface between the NCs and the amorphous zone provides a localized corrosion has been widely discussed . It is worth- favorable channel for chloride ion diffusion. while to mention that, although the grain boundary was specified 0.8 Diffusion path 0.7 [010] 0.6 [100] Cl 0.5 a-Fe O a/c-Fe O c-Fe O 3 4 3 4 3 4 Fe 0.4 0.3 [010] [001] [100] Fig. 5 Energy barriers to Cl ion diffusion from one oxygen vacancy to a neighboring one in the three zones. Representative schematic diagram of the diffusion paths in c-Fe O is inserted. The diffusion barrier is lowest at the interface region, thus confirming that the interface between the nanocrystals and 3 4 the amorphous zone provides a favorable channel for chloride ion diffusion a b –0.15 (110) –0.11 –0.07 –0.04 0.00 0.04 0.07 0.11 0.15 (001) Fig. 6 LADIA simulation results showing the strain state in the matrix side near the metal/passive film interface. Elastic strain component measured from HAADF-STEM image intensity peaks (normal to the interface plane between the passive thin film and the alloy specimen) through the alloy crystal. Scale bar, 2 nm. a High-resolution HAADF-STEM image along the [001] direction of the austenitic matrix showing the passive film on (110) of FeCr Ni single 15 15 −1 crystal in 0.5 mol L H SO . The interface is sharp and straight at the atomic scale. b LADIA simulation map based on a, which shows no evidence of 2 4 lattice expansion and associated tension. The image shows the local expansion/contraction of next-neighbor atom column distances in the alloy along the −1 −1 [110] direction. c High-resolution HAADF-STEM image along the [110] axis showing a passive film formed in 0.5 mol L H SO + 0.3 mol L NaCl 2 4 electrolyte, with corresponding undulating interface. d LADIA simulating map based on c, revealing obvious lattice expansion, with associated induced tension. The image shows the local expansion/contraction of next-neighbor atom column distances in the alloy along the [100] direction. The color baron the right indicates the normal strain, where colors for positive values represent tensile strain and colors for negative values represent compressive strain NATURE COMMUNICATIONS | (2018) 9:2559 | DOI: 10.1038/s41467-018-04942-x | www.nature.com/naturecommunications 5 Energy barrier (eV) ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04942-x Electrolyte Electrolyte δ+ δ+ δ+ δ+ δ+ M M M M M Passive film Interface movement Interface movement Straight interface Metal Metal b Breakdown site δ+ δ+ δ+ M M M – – Cl – Cl – Cl Cl Cl Stress Cl Slow Convex Fast Fast Metal Concave Cl Cl δ+ δ+ δ+ δ+ δ+ M M M M M Passive film Cl Unperturbed Concave Metal Metal interface Fig. 7 Schematic maps illustrating interface evolution in the absence and presence of chloride ions. The film growth process involves transport of both injected metal ions from the matrix and oxygen in solution through the barrier layer, causing the metal/film (Me/BL) interface to move toward the metal −1 matrix side. a In chloride-free (0.5 mol L H SO ) electrolyte, the austenitic matrix experienced selective dissolution of metal ions (major Fe) in a 2 4 −1 −1 somewhat homogeneous manner, yielding a straight Me/BL interface at the atomic scale. b In chloride-containing (0.5 mol L H SO + 0.3 mol L NaCl) 2 4 electrolyte, the tendency of chloride ions to be preferentially adsorbed at defective sites yields non-homogeneous adsorption on the bare metal surface. High chloride ion concentration would induce faster dissolution of Fe, which leads to inhomogeneous interface-movement rates, hence differences in passive film growth rates. Such a process yields a passive film with an irregular and undulating Me/BL interface. c When chloride ions attack the as-grown passive film, chloride ions only get to certain interfacial locations by heterogeneously penetrating the as-grown film along the connected path provided by the interfaces between nanocrystals and the amorphous zone. This gives rise to an undulating interface to the boundary between nano-crystalline oxide grains in those pronounced tension is imparted to the passive film as a result of works, different from the boundary between NCs and the interface undulating induced by chloride ions. amorphous zone we propose, it is strongly implied that full amorphization has greater tendency to resist localized attack due Chloride-induced inhomogeneity to the passive film. According to the absence of the ready path (grain boundaries) for outward to the image contrast in Fig. 6 and Supplementary Fig. 10, the 60,66–68 migration of cations or inward migration of anions . passive film formed in chloride-free solution appears quite uniform and compact, going from the homogeneous contrast of Interface strain distribution modeling. The as-prepared single the HAADF-STEM image. In contrast, the inhomogeneous con- crystal matrix provides us an opportunity to examine the strain trast in the film obtained in chloride-containing solution corre- state within the matrix side using a well-established LADIA sponds to a non-uniform and rather loose passive film, which simulation method based on the high-resolution HAADF-STEM is indisputably resultant from the chloride ion incorporation to images, since the local lattice distortion immediately below the the passive film. Especially at the inner layer, the image contrast interface would be induced if a stress exists at the interface. The is much darker, which indicates that the film is much looser strain state within the matrix side corresponding to the two types (note that the image is taken in the HAADF mode, and the of metal/film interfaces visualized in our study (i.e., straight and contrast is strongly associated with the mass density of heavy undulating interfaces) was modeled using the LADIA simulation atoms). Based on a combination of HAADF imaging and EDS method (Supplementary Methods). The results are shown in analysis of chloride concentration to the inner layer aforemen- Fig. 6 and Supplementary Fig. 10. The straight interface formed in tioned, we propose that some Me (OCl) species might be formed 3 n chloride-free electrolyte revealed no signs of lattice distortion on at the inner layer of the passive film in chloride-containing the matrix side (Fig. 6a, b). Conversely, the undulating interface environments. formed in chloride-containing electrolyte shows clear evidence of strain-induced lattice expansion on the matrix side (Fig. 6c, d). The local strain state that we have extracted is along the direction Discussion parallel to the interface of metal/oxide and meanwhile perpen- In chloride-free electrolyte, the austenitic matrix experienced dicular to the viewing direction (Supplementary Figure 11a and selective dissolution of both the major Fe and minor Cr com- 11b). It is proposed that the strain state along the viewing ponents in a somewhat homogeneous manner, promoting initial direction is also the case, since these two perpendicular directions formation of the inner Cr-rich oxides at the metal/solution (Me/ are crystallographically equivalent, as shown in Supplementary Sol) interface. The dissolved Fe and Cr cations were then Fig. 11c. So the local strain state is in two dimensions parallel to simultaneously hydrolyzed and subsequently re-deposited, to the metal/film interface. The lattice expansion in the matrix side yield the outer Fe-rich precipitation layer, as illustrated in Fig. 7a. means a tensile strain at the interface. Correspondingly, a At this stage, the film growth process involving transport of both 6 NATURE COMMUNICATIONS | (2018) 9:2559 | DOI: 10.1038/s41467-018-04942-x | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04942-x ARTICLE injected metal ions from the matrix and oxygen in solution This accounts for the few select locations at the interface, where through the barrier layer was faster than the film dissolution chloride ion accumulation was detected, with corresponding process, causing the metal/film (Me/BL) interface to move toward interface undulation. the metal matrix side. With time, dynamic equilibrium would be According to the large amount of observation in this study, we attained, where the velocity of film growth equaled that of film do find that the role of interface roughening is quite non-uniform. dissolution. Even though the interfaces of Me/BL can still keep In some locations, the roughening of interface is very consider- moving at dynamic equilibrium, the thickness of the passive film able, as shown in the HAADF-STEM image of Supplementary remained more or less constant. These processes thus yielded a Fig. 12. In the zoom-in image Supplementary Fig. 12(b), which is straight Me/BL interface at the atomic scale, as illustrated in an enlarged image of the area marked with a rectangular in (a), Fig. 7a. the convex site at the matrix side (featuring with the well-defined The propensity of chloride ions to be preferentially adsorbed at lattice images) is getting close to the outer surface of the passive certain non-uniformly distributed distinct defective locations film, wherein, accordingly, the thickness of the passive film is means that the chloride ions would be non-homogeneously quite thin. It is reasonable to propose that, with further rough- adsorbed on the bare metal surface and could modify interfacial ening, the film would become thinner and thinner. Under the process mechanisms in several ways. For one, chloride ions assistance of the stress, the film would be pulled apart and finally particularly promote Fe and to some extent Cr dissolution by a breakdown would occur. coordination: Pit nucleation, initiated at the surface of high purity metals or even of single crystals, is generally thought to be random and n1 Me þ Cl ¼ MeCl ð1Þ unpredictable. Our present experimental results and the analysis above indicate that convex sites where the nature of the passive The non-homogeneous nature of Cl adsorption on Fe should film facilitates its amplification would be the preferential sites for therefore also induce non-homogeneous and irregular Fe cation film breakdown and pit nucleation. injection from substrate. Since the passivity of stainless steels is Interestingly, it becomes obvious from our atomic-scale ana- controlled primarily by the selective dissolution of iron ,we lysis of the chloride attack mechanism that the film breakdown believe that the chloride-induced non-uniform cation injection of sites are not really locations with high chloride ion concentration iron would obviously yield a passive film with an irregular and (concave sites) as widely believed but are actually the adjacent undulating Me/BL interface, as illustrated in Fig. 7b, wherein a locations where the effect of chloride ions are relatively weak large amount of chloride adsorption yields a faster growth of film, (convex sites). This idea, which is based on our experimental in contrast, chloride-free or little chloride leads to a slower findings, introduces another dimension to understanding the growth. As a result, the concave sites, derived from the chloride mechanism of chloride attacking the passive film and implying attack, and the convex sites, with no chloride or little chloride, are the mechanism of chloride-induced passivity breakdown. formed. One plausible mechanism for the observed lattice expansion As the oxide film thickens progressively, whether or not the involves penetration and occupation of vacancies in the metal concave sites are more aggressively attacked (yielding a faster film lattice by chloride (either chloride ions and/or Cl atoms), which growth) or the convex sites more gently attacked (leading to a both possess larger radii than Fe and Cr and as such can induce slower film growth), is decided by the facility of chloride ion lattice expansion. Chloride can modify processes occurring at the permeation. That depends on two aspects: one is the electric field interface when Cl adatoms (from Cl ion reduction) penetrate and occupy the residual Fe vacancies from the cation injection, and the other is the intrinsic nature of the new oxide film. The 69,70 electric field in the oxide varies during film growth . The thereby causing lattice expansion. Correspondingly, no lattice concave interfaces in the film are notably thicker and possess expansion or slight lattice expansion happens at the convex sites, weaker electric field than the convex interface and hence pose while remarkable lattice expansion occurs at the concave sites greater resistance to chloride ion permeation. Correspondingly, (schematically illustrated in Supplementary Fig. 13). It is worth parts of the metal substrate directly beneath the concave positions mentioning that the lattice behaviors based on our LADIA in the undulating film will experience less severe cation injection simulation reflect the atomic projection along the [100] direction. process, which retards film thickening. In contrast, cation injec- Such a projection makes the outmost layer at the matrix side tion process of the metal substrate beneath the convex positions is actually correspond to the convex area with no or little lattice more pronounced, with the higher electric field speeding up the expansion, yielding the lattice expansion to be below the interface, thickening of the passive film. Evidently, the effect of the electric rather than at the interface (Fig. 6b and Supplementary Fig. 13). field is against further amplifying the amplitude between the Regarding the chloride presence below the interface at a depth convex and concave locations. where lattice expansion is measured, we did identify the chloride Nevertheless, if the intrinsic nature of the thicker passive film signal that is derived from the elemental maps, as shown in located at some concave interfaces facilitates the chloride per- Supplementary Figure 14, although its intensity is much lower meation, namely, it is less compact or exhibits more connected than that in the inner layer of the passive film. paths created along the NCs/amorphous interface, those concave All of our experimental results indicate clearly that chloride sites can be expected to still keep a faster film growth, and vice ions remarkably and unilaterally modify the interface zones via versa for the convex interface. The cumulative impact of the lattice expansion on the metal side and induced undulations at occurrences ought to amplify some undulations (concave/convex the interface and structural inhomogeneity on the film side. Such locations), and the convex interfaces have the tendency to move a series of events, by which chloride ions incorporate and attack closer and closer to the outer surface of the passive film, as illu- the passive film, were neither envisaged nor considered probable strated in Fig. 7b. in the available theories describing chloride-induced passivity Another mode of chloride attack, involving permeation breakdown. through pathways along the NCs/amorphous zone interface, as We have shown that the passive films are mostly amorphous mentioned earlier, is illustrated in Fig. 7c. Such a mode is only but with small amount of embedded NCs and that interfaces feasible when a connected path exists through the film. As shown between NCs and the amorphous zone assume the features of in Fig. 7c, chloride ions will permeate the passive film laterally grain boundaries providing ready paths for the chloride ion through the red tunnel and arrive at the matrix/ film interface. transport. We desire a passive film with full amorphous structure, NATURE COMMUNICATIONS | (2018) 9:2559 | DOI: 10.1038/s41467-018-04942-x | www.nature.com/naturecommunications 7 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04942-x so that tunnels for species diffusion and transport are not avail- 2. Hoar, T. P. Production and breakdown of passivity of metals. Corros. Sci. 7, 341–355 (1967). able. We could apply microalloying via an adding of certain 3. Sato, N. Theory for breakdown of anodic oxide films on metals. Electrochim. element(s), which enhance the degree of amorphization of the Acta 16, 1683–1692 (1971). passive film and in the meanwhile the properties of bulk steels 4. Galvele, J. R. Transport processes and mechanism of pitting of metals. remain unchanged. The possibility of this proposal is under J. Electrochem. Soc. 123, 464–474 (1976). consideration and there is no doubt that the newly reported 5. Lin, L. F., Chao, C. Y. & Macdonald, D. D. A point-defect model for anodic passive films. II. Chemical breakdown and pit initiation. J. Electrochem. Soc. mechanism will prompt scientists and engineers to reconsider the 128, 1194–1198 (1981). existing models and look for all the possible approaches to retard 6. Sato, N. Anodic breakdown of passive films on metals. J. Electrochem. Soc. the passivity breakdown. 129, 255–260 (1982). In summary, this study has provided new experimental insights 7. Macdougall, B., Mitchell, D. F., Sproule, G. I. & Graham, M. J. Incorporation on the well-known role of chloride ions in passivity breakdown, of chloride ion in passive oxide-films on nickel. J. Electrochem. Soc. 130, 543–547 (1983). where the existing ideas are mostly based on theoretical models, 8. Murphy, O. J., Bockris, J. O., Pou, T. E., Tongson, L. L. & Monkowski, M. D. with insufficient experimental evidence. Using spherical Cs- Chloride ion penetration of passive films on iron. J. Electrochem. Soc. 130, corrected TEM, in combination with computational modeling, we 1792–1794 (1983). have compared the passive films formed in chloride-free and 9. Di Quarto, F., Piazza, S. & Sunseri, C. Electrical and mechanical breakdown of -containing electrolytes and directly observed atomic-scale accu- anodic films on tungsten in aqueous electrolytes. J. Electroanal. Chem. Inter. Electrochem 248,99–115 (1988). mulation of chloride ions at the metal/film interface, including 10. Macdonald, D. D. The point-defect model for the passive state. J. Electrochem. the chloride-induced lattice expansion on the metal side, inter- Soc. 139, 3434–3449 (1992). facial undulations, and structural alterations to the film. We find 11. Marcus, P. & Herbelin, J. M. The entry of chloride ions into passive films that the passive film is mostly amorphous, with some NCs. We on nickel studied by spectroscopic (ESCA) and nuclear (Cl-36 radiotracer) directly visualize, from the in-plane and also the out-of-plane methods. Corros. Sci. 34, 1123–1145 (1993). 12. Xu, Y., Wang, M. H. & Pickering, H. W. On electric-field-induced breakdown direction, the size, distribution, and crystallographic orientation of passive films and the mechanism of pitting corrosion. J. Electrochem. Soc. of NCs and figure out the boundaries between NCs and the 140, 3448–3457 (1993). amorphous phase. Our experimental and computational results 13. Macdonald, D. D. Passivity - the key to our metals-based civilization. Pure suggest that the interface between NCs and the amorphous zone Appl. Chem. 71, 951–978 (1999). assume a special kind of grain boundaries and thus provide ready 14. Hubschmid, C., Landolt, D. & Mathieu, H. J. XPS and AES analysis of passive films on Fe-25Cr-X (X=Mo, V, Si and Nb) model alloys. Fresenius. J. Anal. paths for chloride ion transport. Our results show clearly that Chem. 353, 234–239 (1995). chloride ions actually permeate the outer and inner layers to 15. Brooks, A. R., Clayton, C. R., Doss, K. & Lu, Y. C. On the role of Cr in the attack the interface, rendering it indistinct and undulating. The passivity of stainless-steel. J. Electrochem. Soc. 133, 2459–2464 (1986). weakest site which is believed to be the preferential site for a 16. Mischler, S., Vogel, A., Mathieu, H. J. & Landolt, D. The chemical breakdown does not coincide, at atomic scale, with locations composition of the passive film on Fe-24Cr and Fe-24Cr-11Mo studied by AES, XPS and SIMS. Corros. Sci. 32, 925–944 (1991). having high chloride ion concentration as widely believed but 17. Ningshen, S., Mudali, U. K., Mittal, V. K. & Khatak, H. S. Semiconducting and occurs at adjacent locations with subdued chloride ion influence. passive film properties of nitrogen-containing type 316LN stainless steels. Corros. Sci. 49, 481–496 (2007). 18. Khalil, W., Haupt, S. & Strehblow, H. H. The thinning of the passive layer of Methods iron by halides. Werks. Korros. Mater. Corros. 36,16–21 (1985). Materials preparation. The FeCr Ni (wt.%) single crystal alloy was grown by 19. Landolt, D., Mischler, S., Vogel, A. & Mathieu, H. J. Chloride ion effects on 15 15 thermal-gradient directional solidification method. The orientation was determined passive films on FeCr and FeCrMo studied by AES, XPS and SIMS. Corros. Sci. by single-crystal X-ray diffractometry, and two low-index crystallographic orien- 31, 431–440 (1990). tations [001] and [110] were obtained. The (001) or (110) plane as exposure surface 20. Olsson, C. O. A. & Landolt, D. Passive films on stainless steels- chemistry, was potentiostatically passivated at a potential of 640 mV/SHE for 30 min in structure and growth. Electrochim. Acta 48, 1093–1104 (2003). chloride-free and -containing electrolyte. The cross-sectional TEM specimen was 21. Natishan, P. M. et al. Chloride interactions with the passive films on stainless prepared by the conventional method. Two passivated surfaces of two samples steel. J. Electrochem. Soc. 158,C7–C10 (2011). were bonded face-to-face and then thinned by grinding and ion-milling. 22. Mitchell, D. F. Quantitative interpretation of Auger sputter profiles of thin- layers. Appl. Surf. Sci. 9, 131–140 (1981). HAADF-STEM imaging. HAADF-STEM images were recorded using Cs-corrected 23. Olefjord, I., Brox, B. & Jelvestam, U. Surface-composition of stainless-steels during anodic-dissolution and passivation studied by ESCA. J. Electrochem. TEM (Titan Cubed 60–300 kV microscope (FEI) fitted with a high-brightness field- emission gun (X-FEG), double Cs corrector from CEOS, and a monochromator Soc. 132, 2854–2861 (1985). operating at 300 kV). The beam convergence is 25 mrad and thus yields a probe 24. Olefjord, I. & Wegrelius, L. Surface-analysis of passive state. Corros. Sci. 31, size of <0.1 nm. 89–98 (1990). 25. Goetz, R., Macdougall, B. & Graham, M. J. An AES and SIMS study of the influence of chloride on the passive oxide film on iron. Electrochim. Acta 31, LADIA calculation method. The LADIA package was used to analyze the elastic 1299–1303 (1986). strain state of the alloy underneath the passive thin film. This algorithm determines 26. Mitrovicscepanovic, V., Macdougall, B. & Graham, M. J. The effect of Cl ions the displacement of actual column image positions versus the position of a refer- on the passivation of Fe-26Cr alloy. Corros. Sci. 27, 239–247 (1987). ence lattice. From this information, we determined the local expansion/contraction 27. Hubschmid, C. & Landolt, D. Formation conditions, chloride content, and of next-neighbor atom column distances in the alloy, i.e., the local lattice para- stability of passive films on an iron-chromium alloy. J. Electrochem. Soc. 140, meters in the <110> directions orthogonal to the <001> viewing direction and the 1898–1902 (1993). <100> directions orthogonal to the <011> viewing direction. 28. Murayama, M., Makiishi, N., Yazawa, Y., Yokota, T. & Tsuzaki, K. Nano-scale chemical analyses of passivated surface layer on stainless steels. Corros. Sci. 48, Data availability. The data that support the findings of this study are available 1307–1318 (2006). from the corresponding author upon reasonable request. 29. Hamada, E. et al. Direct imaging of native passive film on stainless steel by aberration corrected STEM. Corros. Sci. 52, 3851–3854 (2010). 30. Soulas, R. et al. TEM investigations of the oxide layers formed on a 316L alloy Received: 30 September 2017 Accepted: 4 June 2018 in simulated PWR environment. J. Mater. Sci. 48, 2861–2871 (2013). 31. Oh, K., Ahn, S., Eom, K., Jung, K. & Kwon, H. Observation of passive films on Fe-20Cr-xNi (x=0, 10, 20 wt.%) alloys using TEM and Cs-corrected STEM- EELS. Corros. Sci. 79,34–40 (2014). References 32. Davenport, A. J., Oblonsky, L. J., Ryan, M. P. & Toney, M. F. The structure of 1. Hoar, T. P., Mears, D. C. & Rothwell, G. P. The relationships between anodic the passive film that forms on iron in aqueous environments. J. Electrochem. passivity, brightening and pitting. Corros. Sci. 5, 279–289 (1965). Soc. 147, 2162–2173 (2000). 8 NATURE COMMUNICATIONS | (2018) 9:2559 | DOI: 10.1038/s41467-018-04942-x | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04942-x ARTICLE 33. Rees, E. E., Ryan, M. P. & McPhail, D. S. An STM study of the nanocrystalline 62. Ryan, M. P., Newman, R. C. & Thompson, G. E. Atomically resolved STM structure of the passive film on iron. Electrochem. Solid State Lett. 5, B21–B23 of oxide film structures on Fe-Cr alloys during passivation in sulfuric-acid- (2002). solution. J. Electrochem. Soc. 141, L164–L165 (1994). 34. Keller, P. & Strehblow, H. H. XPS investigations of electrochemically formed 63. Hendy, S. C., Laycock, N. J. & Ryan, M. P. Atomistic modeling of cation passive layers on Fe/Cr-alloys in 0.5 M H SO . Corros. Sci. 46,1939–1952 (2004). transport in the passive film on iron and implications for models of growth 2 4 35. Rao, J. C., Zhang, X. X., Qin, B. & Fung, K. K. TEM study of the structural kinetics. J. Electrochem. Soc. 152, B271–B276 (2005). dependence of the epitaxial passive oxide films on crystal facets in polyhedral 64. Marcus, P., Maurice, V. & Strehblow, H. H. Localized corrosion (pitting): a nanoparticles of chromium. Ultramicroscopy 98, 231–238 (2004). model of passivity breakdown including the role of the oxide layer nanostructure. 36. Seyeux, A., Maurice, V., Klein, L. H. & Marcus, P. In situ STM study of the Corros. Sci. 50, 2698–2704 (2008). effect of chloride on passive film on nickel in alkaline solution. J. Electrochem. 65. Seyeux, A., Maurice, V. & Marcus, P. Breakdown kinetics at nanostructure Soc. 153, B453–B463 (2006). defects of passive films. Electrochem. Solid State Lett. 12, C25–C27 (2009). 37. Macdonald, D. D. The history of the point defect model for the passive state: 66. Okamoto, G. Passive film of 18-8 stainless-steel structure and its function. A brief review of film growth aspects. Electrochim. Acta 56, 1761–1772 (2011). Corros. Sci. 13, 471–489 (1973). 38. Padhy, N., Paul, R., Mudali, U. K. & Raj, B. Morphological and compositional 67. Leach, J. S. L. The role of surface films in corrosion and oxidation. Surf. Sci. analysis of passive film on austenitic stainless steel in nitric acid medium. 53, 257–271 (1975). Appl. Surf. Sci. 257, 5088–5097 (2011). 68. Bertocci, U. & Kruger, J. Studies of passive film breakdown by detection and 39. Jung, R. H., Tsuchiya, H. & Fujimoto, S. XPS characterization of passive films analysis of electrochemical noise. Surf. Sci. 101, 608–618 (1980). formed on type 304 stainless steel in humid atmosphere. Corros. Sci. 58,62–68 69. Diawara, B., Beh, Y.-A. & Marcus, P. Nucleation and growth of oxide layers on (2012). stainless steels (FeCr) using a virtual oxide layer model. J. Phys. Chem. C 114, 40. Maurice, V. & Marcus, P. Passive films at the nanoscale. Electrochim. Acta 84, 19299–19307 (2010). 129–138 (2012). 70. Seyeux, A., Maurice, V. & Marcus, P. Oxide film growth kinetics on metals 41. Leistner, K., Toulemonde, C., Diawara, B., Seyeux, A. & Marcus, P. Oxide film and alloys I. Physical model. J. Electrochem. Soc. 160, C189–C196 (2013). growth kinetics on metals and alloys II. Numerical simulation of transient behavior. J. Electrochem. Soc. 160, C197–C205 (2013). Acknowledgements 42. Massoud, T., Maurice, V., Klein, L. H. & Marcus, P. Nanoscale morphology This work is supported by the National Natural Science Foundation of China (Nos. and atomic structure of passive films on stainless steel. J. Electrochem. Soc. 51771212, 11327901, 51390473), the Key Research Program of Frontier Sciences CAS 160, C232–C238 (2013). (QYZDJ-SSW-JSC010), the Innovation Fund in IMR (2017-ZD05). E.E.O. is grateful to 43. Fajardo, S. et al. Low energy SIMS characterization of passive oxide films formed the Chinese Academy of Sciences (CAS) for award of the CAS President’s Fellowship. on a low-nickel stainless steel in alkaline media. Appl. Surf. Sci. 288,423–429 The authors are grateful to Professor H. Wei for the help with the single crystal growth, (2014). Dr. B. J. Wang at Thermo Fisher Scientific Shanghai Nanoport for Super-X EDS 44. Macdonald, D. D. Passivity: enabler of our metals based civilisation. Corros. mapping, Professor X. P. Song for the orientation determination of the single crystal, Eng. Sci. Technol. 49, 143–155 (2014). Mr. T. F. Du for specimens cutting, and Miss X. X. Wei for some experimental 45. Strehblow, H. H. Passivity of metals studied by surface analytical methods, a supplement during addressing the reviewers’ comments. review. Electrochim. Acta 212, 630–648 (2016). 46. Vignal, V. et al. Influence of long-term ageing in solution containing chloride ions on the passivity and the corrosion resistance of duplex stainless steels. Author contributions Corros. Sci. 53, 894–903 (2011). X.L.M. and B.Z. conceived the project of transmission electron microscopy in corrosion 47. Calinski, C. & Strehblow, H. H. ISS depth profiles of the passive layer on Fe/Cr science and designed the experiments and simulations. J.W. and B.Z. conducted the TEM alloys. J. Electrochem. Soc. 136, 1328–1331 (1989). observations, and B.Z. carried out the electrochemical experiments. B.W. provided the 48. Castle, J. E. & Qiu, J. H. The application of ICP-MS and XPS to studies of imaging technique support on the Titan G2 60–300 platform of the aberration-corrected ion selectivity during passivation of stainless-steels. J. Electrochem. Soc. 137, scanning transmission electron microscope. B.Z., J.W., X.L.M., and E.E.O. analyzed the 2031–2038 (1990). data and co-wrote the manuscript. X.W.G., Y.J.W., and D.C. performed the first- 49. Yang, W. P., Costa, D. & Marcus, P. Resistance to pitting and chemical- principle calculations, and Y.C.Z and K.D. carried out the LADIA simulation. All authors composition of passive films of a Fe-17-percent-Cr alloy in chloride-containing contributed to the discussions and manuscript preparation. acid-solution. J. Electrochem. Soc. 141, 2669–2676 (1994). 50. Maurice, V., Yang, W. P. & Marcus, P. XPS and STM study of passive films formed on Fe-22Cr (110) single-crystal surfaces. J. Electrochem. Soc. 143, Additional information 1182–1200 (1996). Supplementary Information accompanies this paper at https://doi.org/10.1038/s41467- 51. Maurice, V., Yang, W. P. & Marcus, P. X-ray photoelectron spectroscopy 018-04942-x. and scanning tunneling microscopy study of passive films formed on (100) Fe-18Cr-13Ni single-crystal surfaces. J. Electrochem. Soc. 145, 909–920 (1998). Competing interests: The authors declare no competing interests. 52. Szklarska-Smialowska, Z. Mechanism of pit nucleation by electrical breakdown of the passive film. Corros. Sci. 44, 1143–1149 (2002). Reprints and permission information is available online at http://npg.nature.com/ 53. Evans, U. R. The passivity of metals. Part I: the isolation of the protective film. reprintsandpermissions/ J. Chem. Soc. 1, 1020–1040 (1927). 54. Vernon, W. H. J., Wormwell, F. & Nurse, T. J. The thickness of air-formed Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in oxide films on iron. J. Chem. Soc. 0, 621–632 (1939). published maps and institutional affiliations. 55. Toney, M. F., Davenport, A. J., Oblonsky, L. J., Ryan, M. P. & Vitus, C. M. Atomic structure of the passive oxide film formed on iron. Phys. Rev. Lett. 79, 4282–4285 (1997). 56. Mayne, J. E. O. & Pryor, M. J. The mechanism of inhibition of corrosion of Open Access This article is licensed under a Creative Commons iron by chromic acid and potassium chromate. J. Chem. Soc. 0, 1831–1835 (1949). Attribution 4.0 International License, which permits use, sharing, 57. Foley, C. L., Kruger, J. & Bechtoldt, C. J. Electron diffraction studies of active adaptation, distribution and reproduction in any medium or format, as long as you give passive and transpassive oxide films formed on iron. J. Electrochem. Soc. 114, appropriate credit to the original author(s) and the source, provide a link to the Creative 994–1001 (1967). Commons license, and indicate if changes were made. The images or other third party 58. McBee, C. L. & Kruger, J. Nature of passive films on iron-chromium alloys. material in this article are included in the article’s Creative Commons license, unless Electrochim. Acta 17, 1337–1341 (1972). indicated otherwise in a credit line to the material. If material is not included in the 59. Ryan,M.P., Newman,R.C.&Thompson,G. E. AnSTM study of the article’s Creative Commons license and your intended use is not permitted by statutory passive film formed on iron in borate buffer solution. J. Electrochem. Soc. regulation or exceeds the permitted use, you will need to obtain permission directly from 142,L177–L179 (1995). the copyright holder. To view a copy of this license, visit http://creativecommons.org/ 60. Maurice, V., Yang, W. P. & Marcus, P. XPS and STM investigation of the licenses/by/4.0/. passive film formed on Cr (110) single-crystal surfaces. J. Electrochem. Soc. 141, 3016–3027 (1994). 61. Ryan,M. P., Newman,R.C. & Thompson, G.E. A scanning tunneling © The Author(s) 2018 microscopy study of structure and structural relaxation in passive oxide- films on Fe-Cr alloys. Philos. Mag. B 70, 241–251 (1994). NATURE COMMUNICATIONS | (2018) 9:2559 | DOI: 10.1038/s41467-018-04942-x | www.nature.com/naturecommunications 9

Journal

Nature CommunicationsSpringer Journals

Published: Jul 2, 2018

There are no references for this article.