Thank you for visiting Nature.com. The browser version you are using has limited CSS support. For the best experience, we recommend that you use an updated browser (or disable Compatibility Mode in Internet Explorer). In the meantime, to ensure continued support, we will render the site without styles and JavaScript.
The widely used stainless steel and its wrought versions are resistant to corrosion in ambient conditions due to the passivation layer consisting of chromium oxide. Corrosion and erosion of steel is traditionally associated with the destruction of these layers, but rarely at the microscopic level, depending on the origin of the surface inhomogeneity. In this work, the nanoscale surface chemical heterogeneity detected by spectroscopic microscopy and chemometric analysis unexpectedly dominates the decomposition and corrosion of cold rolled cerium modified super duplex stainless steel 2507 (SDSS) during its hot deformation behavior. other side. Although X-ray photoelectron microscopy showed relatively uniform coverage of the natural Cr2O3 layer, cold rolled SDSS showed poor passivation results due to the localized distribution of Fe3+ rich nanoislands on the Fe/Cr oxide layer. This knowledge at the atomic level provides a deep understanding of stainless steel corrosion and is expected to help combat corrosion of similar high-alloy metals.
Since the invention of stainless steel, the corrosion resistance of ferrochromium alloys has been attributed to chromium, which forms a strong oxide/oxyhydroxide exhibiting passivating behavior in most environments. Compared to conventional (austenitic and ferritic) stainless steels, super duplex stainless steels (SDSS) with better corrosion resistance have superior mechanical properties1,2,3. Increased mechanical strength allows for lighter and more compact designs. In contrast, the economical SDSS has high resistance to pitting and crevice corrosion, resulting in a longer service life and broader applications in pollution control, chemical containers, and the offshore oil and gas industry4. However, the narrow range of heat treatment temperatures and poor formability hinder its wide practical application. Therefore, SDSS has been modified to improve the above properties. For example, Ce modification and high additions of N 6, 7, 8 were introduced in 2507 SDSS (Ce-2507). A suitable concentration of 0.08 wt.% rare earth element (Ce) has a beneficial effect on the mechanical properties of the DSS, as it improves grain refinement and grain boundary strength. Wear and corrosion resistance, tensile strength and yield strength, and hot workability have also been improved9. Large amounts of nitrogen can replace expensive nickel content, making SDSS more cost-effective10.
Recently, SDSS has been plastically deformed at various temperatures (low temperature, cold and hot) to achieve excellent mechanical properties6,7,8. However, the excellent corrosion resistance of SDSS is due to the presence of a thin oxide film on the surface, which is affected by many factors, such as the presence of many phases with different grain boundaries, unwanted precipitates and different reactions. the internal inhomogeneous microstructure of various austenitic and ferritic phases is deformed 7 . Therefore, the study of the microdomain properties of such films at the level of the electronic structure is of crucial importance for understanding SDSS corrosion and requires complex experimental techniques. So far, surface-sensitive methods such as Auger electron spectroscopy11 and X-ray photoelectron spectroscopy12,13,14,15 as well as the hard X-ray photoelectron photoelectron system distinguish, but often fail to separate, the chemical states of the same element in different points in space on the nanoscale. Several recent studies have linked local oxidation of chromium to the observed corrosion behavior of 17 austenitic stainless steels, 18 martensitic stainless steels, and SDSS 19, 20. However, these studies have mainly focused on the effect of Cr heterogeneity (e.g., Cr3+ oxidation state) on corrosion resistance. Lateral heterogeneity in the oxidation states of elements can be caused by different compounds with the same constituent elements, such as iron oxides. These compounds inherit a thermomechanically processed small size closely adjacent to each other, but differ in composition and oxidation state16,21. Therefore, revealing the destruction of oxide films and then pitting requires an understanding of surface inhomogeneity at the microscopic level. Despite these requirements, quantitative assessments such as lateral oxidation heterogeneity, especially of iron on the nano/atomic scale, are still lacking and their significance for corrosion resistance remains unexplored. Until recently, the chemical state of various elements, such as Fe and Ca, was quantitatively described on steel samples using soft X-ray photoelectron microscopy (X-PEEM) in nanoscale synchrotron radiation facilities. Combined with chemically sensitive X-ray absorption spectroscopy (XAS) techniques, X-PEEM enables XAS measurement with high spatial and spectral resolution, providing chemical information about the elemental composition and its chemical state with spatial resolution down to the nanometer scale 23 . This spectroscopic observation of the site of initiation under a microscope facilitates local chemical experiments and can spatially demonstrate previously unexplored chemical changes in the Fe layer.
This study extends the advantages of PEEM in detecting chemical differences at the nanoscale and presents an insightful atomic-level surface analysis method for understanding the corrosion behavior of Ce-2507. It uses K-means cluster chemometric data24 to map the global chemical composition (heterogeneity) of the elements involved, with their chemical states presented in a statistical representation. Unlike the traditional case of corrosion caused by chromium oxide film breakdown, the current poor passivation and poor corrosion resistance are attributed to localized Fe3+ rich nanoislands near the Fe/Cr oxide layer, which may be the result of protective oxides. In the place of breakdown, a film is formed that causes corrosion.
The corrosive behavior of deformed SDSS 2507 was first evaluated using electrochemical measurements. On fig. Figure 1 shows the Nyquist and Bode curves for selected samples in acidic (pH = 1) aqueous solutions of FeCl3 at room temperature. The selected electrolyte acts as a strong oxidizing agent, characterizing the tendency of the passivation film to break down. Although the material did not undergo stable room temperature pitting, these analyzes provided insight into potential failure events and post-corrosion processes. The equivalent circuit (Fig. 1d) was used to fit electrochemical impedance spectroscopy (EIS) spectra, and the corresponding fitting results are shown in Table 1. Incomplete half circles appeared when testing the solution treated and hot worked samples, while the corresponding compressed half circles were cold rolled (Fig. 1b). In the EIS spectrum, the semicircle radius can be considered as the polarization resistance (Rp)25,26. The Rp of solution treated SDSS in Table 1 is about 135 kΩ cm-2, however for hot worked and cold rolled SDSS we can see much lower values ​​of 34.7 and 2.1 kΩ cm–2 respectively. This significant decrease in Rp indicates a detrimental effect of plastic deformation on passivation and corrosion resistance, as shown in previous reports 27, 28, 29, 30.
a Nyquist, b, c Bode impedance and phase diagrams, and an equivalent circuit model for d, where RS is the electrolyte resistance, Rp is the polarization resistance, and QCPE is the constant phase element oxide used to model the non-ideal capacitance (n) . The EIS measurements were carried out at no-load potential.
The first order constants are shown in the Bode diagram and the high frequency plateau represents the electrolyte resistance RS26. As the frequency decreases, the impedance increases and a negative phase angle is found, indicating capacitance dominance. The phase angle increases, retaining its maximum value in a relatively wide frequency range, and then decreases (Fig. 1c). However, in all three cases this maximum value is still less than 90°, indicating a non-ideal capacitive behavior due to capacitive dispersion. Thus, the QCPE constant phase element (CPE) is used to represent the interfacial capacitance distribution derived from surface roughness or inhomogeneity, especially in terms of atomic scale, fractal geometry, electrode porosity, non-uniform potential, and surface dependent current distribution. Electrode geometry31,32. CPE impedance:
where j is the imaginary number and ω is the angular frequency. QCPE is a frequency independent constant proportional to the active open area of ​​the electrolyte. n is a dimensionless power number that describes the deviation from the ideal capacitive behavior of a capacitor, i.e. the closer n is to 1, the closer CPE is to pure capacitance, and if n is close to zero, it is resistance. A small deviation of n, close to 1, indicates the non-ideal capacitive behavior of the surface after polarization testing. The QCPE of cold rolled SDSS is much higher than similar products, which means that the surface quality is less uniform.
Consistent with most corrosion resistance properties of stainless steels, the relatively high Cr content of SDSS generally results in superior corrosion resistance of SDSS due to the presence of a passive protective oxide film on the surface17. This passivating film is usually rich in Cr3+ oxides and/or hydroxides, mainly integrating Fe2+, Fe3+ oxides and/or (oxy)hydroxides 33 . Despite the same surface uniformity, passivating oxide layer, and no visible fracture on the surface, as determined by microscopic images,6,7 the corrosion behavior of hot-worked and cold-rolled SDSS is different and therefore requires in-depth study of the deformation microstructure and structural characteristic of steel.
The microstructure of deformed stainless steel was quantitatively investigated using internal and synchrotron high-energy X-rays (Supplementary Figures 1, 2). A detailed analysis is provided in the Supplementary Information. Although this mostly corresponds to the type of main phase, differences were found in the volume fractions of the phases, which are listed in Supplementary Table 1. The difference may be due to the heterogeneous phase fraction at the surface and the volume fraction (XRD) subject to different depth of detection with using X-ray diffraction with various energy sources of incident photons. The relatively higher proportion of austenite in cold rolled specimens, determined by XRD from a laboratory source, indicates better passivation and subsequently better corrosion resistance35, while more accurate and statistical results indicate opposite trends in phase proportions. In addition, the corrosion resistance of steel also depends on the degree of grain refinement, grain size reduction, increase in microdeformations and dislocation density that occur during thermomechanical treatment36,37,38. The hot-worked specimens exhibit a more grainy nature, indicative of micron-sized grains, while the smooth rings observed in the cold-rolled specimens (Supplementary Fig. 3) indicate significant grain refinement to nanoscale in previous work6, which should contribute to film passivation. formation and increase of corrosion resistance. Higher dislocation density is usually associated with lower resistance to pitting, which agrees well with electrochemical measurements.
Changes in the chemical states of microdomains of elementary elements have been systematically studied using X-PEEM. Despite the abundance of alloying elements, Cr, Fe, Ni, and Ce39 were chosen here because Cr is the key element for the formation of the passivation film, Fe is the main element in steel, and Ni enhances passivation and balances the ferrite-austenitic phase structure and the purpose of modifying Ce. By adjusting the energy of the synchrotron radiation, the RAS was coated from the surface with the main features of Cr (edge ​​L2.3), Fe (edge ​​L2.3), Ni (edge ​​L2.3) and Ce (edge ​​M4.5). hot forming and cold rolling Ce-2507 SDSS. Appropriate data analysis was performed by incorporating energy calibration with published data (eg XAS 40, 41 on Fe L2, 3 edges).
On fig. Figure 2 shows X-PEEM images of hot-worked (Fig. 2a) and cold-rolled (Fig. 2d) Ce-2507 SDSS and corresponding XAS edges of Cr and Fe L2,3 at individually marked locations. The L2,3 edge of the XAS probes the unoccupied 3d states after electron photoexcitation at the spin-orbit splitting levels 2p3/2 (L3 edge) and 2p1/2 (L2 edge). Information about the valence state of Cr was obtained from XAS at the L2,3 edge in Fig. 2b, e. Comparison with judges. 42,43 showed that four peaks were observed near the L3 edge, named A (578.3 eV), B (579.5 eV), C (580.4 eV) and D (582.2 eV), reflecting octahedral Cr3+, corresponding to the Cr2O3 ion. The experimental spectra agree with the theoretical calculations shown in panels b and e, obtained from multiple calculations of the crystal field at the Cr L2.3 interface using a crystal field of 2.0 eV44. Both surfaces of hot-worked and cold-rolled SDSS are coated with a relatively uniform layer of Cr2O3.
a X-PEEM thermal image of thermally deformed SDSS corresponding to b Cr L2.3 edge and c Fe L2.3 edge, d X-PEEM thermal image of cold rolled SDSS corresponding to e Cr L2.3 edge and f Fe L2 .3 edge side ( f). The XAS spectra are plotted at different spatial positions marked on the thermal images (a, d), the orange dotted lines in (b) and (e) represent the simulated XAS spectra of Cr3+ with a crystal field value of 2.0 eV. For X-PEEM images, use a thermal palette to improve image readability, where colors from blue to red are proportional to the intensity of X-ray absorption (from low to high).
Regardless of the chemical environment of these metallic elements, the chemical state of the additions of Ni and Ce alloying elements for both samples remained unchanged. Additional drawing. Figures 5-9 show X-PEEM images and corresponding XAS spectra for Ni and Ce at various positions on the surface of hot-worked and cold-rolled specimens. Ni XAS shows the oxidation states of Ni2+ over the entire measured surface of hot-worked and cold-rolled specimens (Supplementary Discussion). It should be noted that, in the case of hot-worked samples, the XAS signal of Ce was not observed, while in the case of cold-rolled samples, the spectrum of Ce3+ was observed. The observation of Ce spots in cold-rolled samples showed that Ce mainly appears in the form of precipitates.
In the thermally deformed SDSS, no local structural change in XAS at the Fe L2,3 edge was observed (Fig. 2c). However, the Fe matrix micro-regionally changes its chemical state at seven randomly selected points of the cold-rolled SDSS, as shown in Fig. 2f. In addition, in order to get an accurate idea of ​​the changes in the state of Fe at the selected locations in Fig. 2f, local surface studies were performed (Fig. 3 and Supplementary Fig. 10) in which smaller circular regions were selected. The XAS spectra of the Fe L2,3 edge of α-Fe2O3 systems and Fe2+ octahedral oxides were modeled by multiple crystal field calculations using crystal fields of 1.0 (Fe2+) and 1.0 (Fe3+)44. We note that α-Fe2O3 and γ-Fe2O3 have different local symmetries45,46, Fe3O4 has combination of both Fe2+ & Fe3+,47, and FeO45 as a formally divalent Fe2+ oxide (3d6). We note that α-Fe2O3 and γ-Fe2O3 have different local symmetries45,46, Fe3O4 has a combination of both Fe2+ & Fe3+,47, and FeO45 as a formally divalent Fe2+ oxide (3d6). Note that α-Fe2O3 and γ-Fe2O3 have different local symmetries45,46, Fe3O4 combines both Fe2+ and Fe3+,47 and FeO45 in the form of formally divalent oxide Fe2+ (3d6). Note that α-Fe2O3 and γ-Fe2O3 have different local symmetries45,46, Fe3O4 has a combination of Fe2+ and Fe3+,47 and FeO45 acts as a formal divalent Fe2+ oxide (3d6). All Fe3+ ions in α-Fe2O3 have only Oh positions, while γ-Fe2O3 is usually represented by Fe3+ t2g [Fe3+5/3V1/3]eg O4 spinel with vacancies in eg positions. Therefore, the Fe3+ ions in γ-Fe2O3 have both Td and Oh positions. As mentioned in a previous paper,45 although the intensity ratio of the two is different, their intensity ratio eg/t2g is ≈1, while in this case the observed intensity ratio eg/t2g is about 1. This rules out the possibility that in the current situation only Fe3+ is present. Considering the case of Fe3O4 with both Fe2+ and Fe3+, the first feature, which is known to have a weaker (stronger) L3 edge for Fe, indicates a smaller (larger) number of unoccupied t2g states. This applies to Fe2+ (Fe3+), which shows that the first feature of the increase indicates an increase in the content of Fe2+47. These results show that the coexistence of Fe2+ and γ-Fe2O3, α-Fe2O3 and/or Fe3O4 dominates on the cold-rolled surface of the composites.
Enlarged photoelectron thermal imaging images of the XAS spectra (a, c) and (b, d) crossing the Fe L2,3 edge at various spatial positions within selected regions 2 and E in Figs. 2d.
The obtained experimental data (Fig. 4a and Supplementary Fig. 11) are plotted and compared with the data for pure compounds 40, 41, 48. Three different types of experimentally observed Fe L-edge XAS spectra (XAS- 1, XAS-2 and XAS-3: Fig. 4a). In particular, spectrum 2-a (denoted as XAS-1) in Fig. 3b followed by spectrum 2-b (labeled XAS-2) was observed over the entire detection area, while spectra like E-3 were observed in figure 3d (labeled XAS-3) were observed at specific locations. As a rule, four parameters were used to identify the existing valence states in the sample under study: (1) spectral characteristics L3 and L2, (2) energy positions of the characteristics L3 and L2, (3) energy difference L3-L2. , ( 4) L2/L3 intensity ratio. According to visual observations (Fig. 4a), all three Fe components, namely, Fe0, Fe2+, and Fe3+, are present on the SDSS surface under study. The calculated intensity ratio L2/L3 also indicated the presence of all three components.
a Simulated XAS spectra of Fe with observed three different experimental data (solid lines XAS-1, XAS-2 and XAS-3 correspond to 2-a, 2-b and E-3 in Fig. 2 and 3) Comparison , Octahedrons Fe2+, Fe3+ with crystal field values ​​of 1.0 eV and 1.5 eV, respectively, the experimental data measured with bd (XAS-1, XAS-2, XAS-3) and the corresponding optimized LCF data (solid black line), and also in the form XAS-3 spectra with Fe3O4 (mixed state of Fe) and Fe2O3 (pure Fe3+) standards.
A linear combination fit (LCF) of the three standards 40, 41, 48 was used to quantify the iron oxide composition. LCF was implemented for three selected Fe L-edge XAS spectra showing the highest contrast, namely XAS-1, XAS-2 and XAS-3, as shown in Fig. 4b–d. For LCF fittings, 10% Fe0 was taken into account in all cases due to the fact that we observed a small ledge in all data, and also due to the fact that metallic iron is the main component of steel. Indeed, the probation depth of X-PEEM for Fe (~6 nm)49 is bigger than the estimated oxidation layer thickness (slightly > 4 nm), allowing detection of signal from the iron matrix (Fe0) beneath the passivation layer. Indeed, the probation depth of X-PEEM for Fe (~6 nm)49 is bigger than the estimated oxidation layer thickness (slightly > 4 nm), allowing detection of signal from the iron matrix (Fe0) beneath the passivation layer. Действительно, пробная глубина X-PEEM для Fe (~ 6 нм)49 больше, чем предполагаемая толщина слоя окисления (немного > 4 нм), что позволяет обнаружить сигнал от железной матрицы (Fe0) под пассивирующим слоем. Indeed, the probe X-PEEM depth for Fe (~6 nm)49 is greater than the assumed thickness of the oxidation layer (slightly >4 nm), which makes it possible to detect the signal from the iron matrix (Fe0) under the passivation layer.事实上，X-PEEM 对Fe（~6 nm）49 的检测深度大于估计的氧化层厚度（略> 4 nm），允许检测来自钝化层下方的铁基体（Fe0）的信号。事实上 ， X-PEEM 对 Fe （~ 6 nm） 49 的 检测 深度 大于 的 氧化层 厚度 略 略> 4 nm） 允许 检测 来自 钝化层 下方 铁基体 （fe0） 的。 信号 信号 信号 信号 信号 信号 信号 信号 信号 信号 信号 信号Фактически, глубина обнаружения Fe (~ 6 нм) 49 с помощью X-PEEM больше, чем предполагаемая толщина оксидного слоя (немного > 4 нм), что позволяет обнаруживать сигнал от железной матрицы (Fe0) ниже пассивирующего слоя. In fact, the depth of detection of Fe (~6 nm) 49 by X-PEEM is greater than the expected thickness of the oxide layer (slightly > 4 nm), which allows detection of the signal from the iron matrix (Fe0) below the passivation layer. . Various combinations of Fe2+ and Fe3+ were performed to find the best possible solution for the observed experimental data. On fig. 4b shows the XAS-1 spectrum for the combination of Fe2+ and Fe3+, where the proportions of Fe2+ and Fe3+ were similar by about 45%, indicating mixed oxidation states of Fe. While for the XAS-2 spectrum, the percentage of Fe2+ and Fe3+ becomes ~30% and 60%, respectively. Fe2+ ​​is less than Fe3+. The ratio of Fe2+ to Fe3, equal to 1:2, means that Fe3O4 can be formed at the same ratio between Fe ions. In addition, for the XAS-3 spectrum, the percentage of Fe2+ and Fe3+ becomes ~10% and 80%, which indicates a higher conversion of Fe2+ to Fe3+. As mentioned above, Fe3+ can come from α-Fe2O3, γ-Fe2O3 or Fe3O4. To understand the most likely source of Fe3+, the XAS-3 spectrum was plotted with different Fe3+ standards in Figure 4e, showing similarity to both standards when considering the B peak. However, the intensity of the shoulder peaks (A: from Fe2+) and the B/A intensity ratio indicate that that the spectrum of XAS-3 is close, but does not coincide with the spectrum of γ-Fe2O3. Compared to bulk γ-Fe2O3, the Fe 2p XAS peak of A SDSS has a slightly higher intensity (Fig. 4e), which indicates a higher intensity of Fe2+. Although the spectrum of XAS-3 is similar to that of γ-Fe2O3, where Fe3+ is present in the Oh and Td positions, identification of different valence states and coordination only along the L2,3 edge or the L2/L3 intensity ratio remains a problem. a topic of ongoing discussion due to the complexity of the various factors that affect the final spectrum41.
In addition to the spectral differences in the chemical state of the selected regions of interest described above, the global chemical heterogeneity of the key elements Cr and Fe was also assessed by classifying all XAS spectra obtained on the sample surface using the K-means clustering method. . Edge profiles Cr L are set to form two optimal clusters spatially distributed in the hot-worked and cold-rolled specimens shown in Figs. 5. It is clear that no local structural changes are perceived as similar, since the two centroids of the XAS Cr spectra are comparable. These spectral shapes of the two clusters are almost identical to those corresponding to Cr2O342, which means that the Cr2O3 layers are relatively evenly spaced on the SDSS.
Cr L K-means edge region clusters, and b is the corresponding XAS centroids. Results of K-means X-PEEM comparison of cold-rolled SDSS: c Cr L2.3 edge region of K-means clusters and d corresponding XAS centroids.
To illustrate more complex FeL edge maps, four and five optimized clusters and their associated centroids (spectral profiles) were used for hot-worked and cold-rolled specimens, respectively. Therefore, the percentage (%) of Fe2+ and Fe3+ can be obtained by fitting the LCF shown in Fig.4. The pseudoelectrode potential Epseudo as a function of Fe0 was used to reveal the microchemical inhomogeneity of the surface oxide film. Epseudo is roughly estimated by the mixing rule,
where \(\rm{E}_{\rm{Fe}/\rm{Fe}^{2 + (3 + )}}\) equals \(\rm{Fe} + 2e^ – \ to \rm { Fe}^{2 + (3 + )}\), 0.440 and 0.036 V, respectively. Regions with a lower potential have a higher content of the Fe3+ compound. The potential distribution in thermally deformed samples has a layered character with a maximum change of about 0.119 V (Fig. 6a, b). This potential distribution is closely related to the surface topography (Fig. 6a). No other position-dependent changes in the underlying laminar interior were observed (Fig. 6b). On the contrary, for the connection of dissimilar oxides with different contents of Fe2+ and Fe3+ in cold-rolled SDSS, one can observe a non-uniform nature of the pseudopotential (Fig. 6c, d). Fe3+ oxides and/or (oxy)hydroxides are the main constituents of rust in steel and are permeable to oxygen and water50. In this case, the islands rich in Fe3+ are considered to be locally distributed and can be considered as corroded areas. At the same time, the gradient in the potential field, rather than the absolute value of the potential, can be used as an indicator for the localization of active corrosion sites. This uneven distribution of Fe2+ and Fe3+ on the surface of cold rolled SDSS can change the local chemistry and provide a more practical active surface area in oxide film breakdown and corrosion reactions, thereby allowing continuous corrosion of the underlying metal matrix, resulting in internal corrosion. heterogeneity of properties and a decrease in the protective properties of the passivating layer.
K-means clusters and corresponding XAS centroids in the Fe L2.3 edge region of hot-deformed X-PEEM ac and df of cold-rolled SDSS. a, d K-means cluster plots overlaid on X-PEEM images. The calculated pseudoelectrode potential (Epseudo) is mentioned along with the K-means cluster plot. The brightness of the X-PEEM image, like the color in Fig. 2 is proportional to the X-ray absorption intensity.
Relatively uniform Cr but different chemical state of Fe leads to different oxide film damage and corrosion patterns in hot-worked and cold-rolled Ce-2507. This property of cold rolled Ce-2507 has been well studied. With regard to the formation of oxides and hydroxides of Fe in the ambient air in this almost neutral work, the reactions are as follows:
The above reactions occur in the following scenarios based on X-PEEM analysis. A small shoulder corresponding to Fe0 is associated with the underlying metallic iron. The reaction of metallic Fe with the environment results in the formation of an Fe(OH)2 layer (equation (5)), which enhances the Fe2+ signal in the Fe L-edge XAS. Prolonged exposure to air may result in the formation of Fe3O4 and/or Fe2O3 oxides after Fe(OH)252,53. Two stable forms of Fe, Fe3O4 and Fe2O3, can also form in the Cr3+ rich protective layer, of which Fe3O4 prefers a uniform and sticky structure. The presence of both results in mixed oxidation states (XAS-1 spectrum). The XAS-2 spectrum mainly corresponds to Fe3O4. While the observation of XAS-3 spectra in several places indicated complete conversion to γ-Fe2O3. Since the penetration depth of the unfolded X-rays is about 50 nm, the signal from the lower layer results in a higher intensity of the A peak.
The XPA spectrum shows that the Fe component in the oxide film has a layered structure combined with a Cr oxide layer. In contrast to the signs of passivation due to local inhomogeneity of Cr2O3 during corrosion, despite the uniform layer of Cr2O3 in this work, low corrosion resistance is observed in this case, especially for cold-rolled specimens. The observed behavior can be understood as the heterogeneity of the chemical oxidation state in the upper layer (Fe), which affects the corrosion performance. Due to the same stoichiometry of the upper layer (iron oxide) and the lower layer (chromium oxide)52,53 better interaction (adhesion) between them leads to slow transport of metal or oxygen ions in the lattice, which, in turn, leads to an increase in corrosion resistance. Therefore, a continuous stoichiometric ratio, i.e. one oxidation state of Fe, is preferable to abrupt stoichiometric changes. The heat-deformed SDSS has a more uniform surface, a denser protective layer, and better corrosion resistance. Whereas for cold-rolled SDSS, the presence of Fe3+-rich islands under the protective layer violates the integrity of the surface and causes galvanic corrosion with the nearby substrate, which leads to a sharp drop in Rp (Table 1). The EIS spectrum and its corrosion resistance are reduced. It can be seen that the local distribution of Fe3+ rich islands due to plastic deformation mainly affects the corrosion resistance, which is a breakthrough in this work. Thus, this study presents spectroscopic microscopic images of the reduction in corrosion resistance of SDSS samples studied by the plastic deformation method.
In addition, although alloying with rare earth elements in two-phase steels shows better performance, the interaction of this additive element with an individual steel matrix in terms of corrosion behavior according to spectroscopic microscopy data remains elusive. The appearance of Ce signals (via XAS M-edges) appears only in a few places during cold rolling, but disappears during hot deformation of the SDSS, indicating local precipitation of Ce in the steel matrix, rather than homogeneous alloying. While not significantly improving the mechanical properties of SDSS6,7, the presence of rare earth elements reduces the size of the inclusions and is thought to inhibit pitting in the initial region54.
In conclusion, this work discloses the effect of surface heterogeneity on the corrosion of 2507 SDSS modified with cerium by quantifying the chemical content of nanoscale components. We answer the question why stainless steel corrodes even under a protective oxide layer by quantifying its microstructure, surface chemistry, and signal processing using K-means clustering. It has been established that islands rich in Fe3+, including their octahedral and tetrahedral coordination along the whole feature of mixed Fe2+/Fe3+, are the source of damage and corrosion of the cold-rolled oxide film SDSS. Nanoislands dominated by Fe3+ lead to poor corrosion resistance even in the presence of a sufficient stoichiometric Cr2O3 passivating layer. In addition to methodological advances in determining the effect of nanoscale chemical heterogeneity on corrosion, ongoing work is expected to inspire engineering processes to improve the corrosion resistance of stainless steels during steelmaking.
To prepare the Ce-2507 SDSS ingot used in this study, a mixed composition including Fe-Ce master alloy sealed with a pure iron tube was melted in a 150 kg medium frequency induction furnace to produce molten steel and poured into a mold. The measured chemical compositions (wt%) are listed in Supplementary Table 2. Ingots are first hot forged into blocks. Then it was annealed at 1050°C for 60 min to obtain steel in the state of a solid solution, and then quenched in water to room temperature. The studied samples were studied in detail using TEM and DOE to study the phases, grain size and morphology. More detailed information about samples and production process can be found in other sources6,7.
Cylindrical samples (φ10 mm×15 mm) for hot compression were processed so that the axis of the cylinder was parallel to the deformation direction of the block. High-temperature compression was carried out at various temperatures in the range of 1000-1150°C using a Gleeble-3800 thermal simulator at a constant strain rate in the range of 0.01-10 s-1. Before deformation, the samples were heated at a rate of 10 °C s-1 for 2 min at a selected temperature to eliminate the temperature gradient. After achieving temperature uniformity, the sample was deformed to a true strain value of 0.7. After deformation, the samples were immediately quenched with water to preserve the deformed structure. The hardened specimen is then cut parallel to the compression direction. For this particular study, we chose a specimen with a hot strain condition of 1050°C, 10 s-1 because the observed microhardness was higher than other specimens7.
Massive (80 × 10 × 17 mm3) samples of the Ce-2507 solid solution were used in an LG-300 three-phase asynchronous two-roll mill with the best mechanical properties among all other deformation levels6. The strain rate and thickness reduction for each path are 0.2 m·s-1 and 5%, respectively.
An Autolab PGSTAT128N electrochemical workstation was used for SDSS electrochemical measurements after cold rolling to a 90% reduction in thickness (1.0 equivalent true strain) and after hot pressing at 1050°C for 10 s-1 to a true strain of 0.7. The workstation has a three-electrode cell with a saturated calomel electrode as the reference electrode, a graphite counter electrode, and an SDSS sample as the working electrode. The samples were cut into cylinders with a diameter of 11.3 mm, to the sides of which copper wires were soldered. The samples were then fixed with epoxy, leaving a working open area of ​​1 cm2 as the working electrode (bottom side of the cylindrical sample). Be careful during curing of the epoxy and subsequent sanding and polishing to avoid cracking. The working surfaces were ground and polished with a diamond polishing suspension with a particle size of 1 μm, washed with distilled water and ethanol, and dried in cold air. Before electrochemical measurements, the polished samples were exposed to air for several days to form a natural oxide film. An aqueous solution of FeCl3 (6.0 wt%), stabilized to pH = 1.0 ± 0.01 with HCl according to ASTM recommendations, is used to accelerate the corrosion of stainless steel55 because it is corrosive in the presence of chloride ions with a strong oxidizing capacity and low pH Environmental standards G48 and A923. Immerse the sample in the test solution for 1 hour to reach near steady state before making any measurements. For solid-solution, hot-formed, and cold-rolled samples, impedance measurements were carried out at open circuit potentials (OPC) of 0.39, 0.33, and 0.25 V, respectively, in the frequency range from 1 105 to 0.1 Hz with an amplitude of 5 mV. All chemical tests were repeated at least 3 times under the same conditions to ensure data reproducibility.
For HE-SXRD measurements, rectangular duplex steel blocks measuring 1 × 1 × 1.5 mm3 were measured to quantify the beam phase composition of a Brockhouse high-energy wiggler at CLS, Canada56. Data collection was carried out in Debye-Scherrer geometry or transmission geometry at room temperature. The X-ray wavelength calibrated with the LaB6 calibrator is 0.212561 Å, which corresponds to 58 keV, which is much higher than that of Cu Kα (8 keV) commonly used as a laboratory X-ray source. The sample was located at a distance of 740 mm from the detector. The detection volume of each sample is 0.2 × 0.3 × 1.5 mm3, which is determined by the beam size and sample thickness. All data were collected using a Perkin Elmer area detector, flat panel X-ray detector, 200 µm pixels, 40×40 cm2 using an exposure time of 0.3 s and 120 frames.
X-PEEM measurements of two selected model systems were carried out at the Beamline MAXPEEM PEEM end station in the MAX IV laboratory (Lund, Sweden). Samples were prepared in the same way as for electrochemical measurements. The prepared samples were kept in air for several days and degassed in an ultrahigh vacuum chamber before being irradiated with synchrotron photons. The energy resolution of the beam line was obtained by measuring the ion yield spectrum in the excitation region from N 1 s to 1\(\pi _g^ \ast\) near hv = 401 eV in N2 with the dependence of the photon energy on E3/2 , 57. Approximation spectra gave ΔE (width of the spectral line) of about 0.3 eV in the measured energy range. Therefore, the beamline energy resolution was estimated to be E/∆E = 700 eV/0.3 eV > 2000 and flux ≈1012 ph/s by utilizing a modified SX-700 monochromator with a Si 1200-line mm−1 grating for the Fe 2p L2,3 edge, Cr 2p L2,3 edge, Ni 2p L2,3 edge, and Ce M4,5 edge. Therefore, the beamline energy resolution was estimated to be E/∆E = 700 eV/0.3 eV > 2000 and flux ≈1012 ph/s by utilizing a modified SX-700 monochromator with a Si 1200-line mm−1 grating for the Fe 2p L2.3 edge, Cr 2p L2.3 edge, Ni 2p L2.3 edge, and Ce M4.5 edge. Таким образом, энергетическое разрешение канала пучка было оценено как E/∆E = 700 эВ/0,3 эВ > 2000 и поток ≈1012 ф/с при использовании модифицированного монохроматора SX-700 с решеткой Si 1200 штрихов/мм для Fe кромка 2p L2,3, кромка Cr 2p L2,3, кромка Ni 2p L2,3 и кромка Ce M4,5. Thus, the energy resolution of the beam channel was estimated as E/∆E = 700 eV/0.3 eV > 2000 and flux ≈1012 f/s using a modified SX-700 monochromator with a Si grating of 1200 lines/mm for Fe edge 2p L2 ,3, Cr edge 2p L2.3, Ni edge 2p L2.3, and Ce edge M4.5.因此，光束线能量分辨率估计为E/ΔE = 700 eV/0.3 eV > 2000 和通量≈1012 ph/s，通过使用带有Si 1200 线mm-1 光栅的改进的SX-700 单色器用于Fe 2p L2,3 边缘、Cr 2p L2,3 边缘、Ni 2p L2,3 边缘和Ce M4,5 边缘。因此 ， 光束线 能量 分辨率 为 为 为 为 δe = 700 EV/0.3 EV> 2000 和 ≈1012 PH/S ， 使用 带有 带有 1200 线 mm-1 光栅 改进 的 SX-700 单色器 于 于 于 用 用 用Fe 2p L2.3 边缘、Cr 2p L2.3 边缘、Ni 2p L2.3 边缘和Ce M4.5 边缘。 Thus, when using a modified SX-700 monochromator with a 1200 line Si grating. 3, Cr edge 2p L2.3, Ni edge 2p L2.3 and Ce edge M4.5. Scan photon energy in 0.2 eV steps. At each energy, PEEM images were recorded using a fiber-coupled TVIPS F-216 CMOS detector with 2 x 2 bins, which provides a resolution of 1024 x 1024 pixels in a 20 µm field of view. The exposure time of the images was 0.2 s, averaging 16 frames. The photoelectron image energy is chosen in such a way as to provide the maximum secondary electron signal. All measurements were carried out at normal incidence using a linearly polarized photon beam. More information about measurements can be found in a previous study. After studying the total electron yield (TEY) detection mode and its application in X-PEEM49, the trial depth of this method is estimated to be about 4-5 nm for the Cr signal and about 6 nm for Fe. The Cr depth is very close to the thickness of the oxide film (~4 nm)60,61 while the Fe depth is larger than the thickness. XRD collected at the edge of Fe L is a mixture of XRD of iron oxides and Fe0 from the matrix. In the first case, the intensity of the emitted electrons comes from all possible types of electrons that contribute to TEY. However, a pure iron signal requires higher kinetic energy for the electrons to pass through the oxide layer to the surface and be collected by the analyzer. In this case, the Fe0 signal is mainly due to LVV Auger electrons, as well as secondary electrons emitted by them. In addition, the TEY intensity contributed by these electrons decays during the electron escape path, further reducing the Fe0 spectral response in the iron XAS map.
Integrating data mining into a data cube (X-PEEM data) is a key step in extracting relevant information (chemical or physical properties) in a multidimensional approach. K-means clustering is widely used in several fields, including machine vision, image processing, unsupervised pattern recognition, artificial intelligence, and classificatory analysis. For example, K-means clustering has performed well in clustering hyperspectral image data. In principle, for multi-feature data, the K-means algorithm can easily group them based on information about their attributes (photon energy properties). K-means clustering is an iterative algorithm for dividing data into K non-overlapping groups (clusters), where each pixel belongs to a certain cluster depending on the spatial distribution of chemical inhomogeneity in the microstructural composition of steel. The K-means algorithm includes two stages: in the first stage, K centroids are calculated, and in the second stage, each point is assigned a cluster with neighboring centroids. The center of gravity of a cluster is defined as the arithmetic mean of the data points (XAS spectrum) for that cluster. There are various distances to define neighboring centroids as Euclidean distance. For an input image of px,y (where x and y are the resolution in pixels), CK is the center of gravity of the cluster; this image can then be segmented (clustered) into K clusters using K-means63. The final steps of the K-means clustering algorithm are:
Step 2. Calculate the membership of all pixels according to the current centroid. For example, it is calculated from the Euclidean distance d between the center and each pixel:
Step 3 Assign each pixel to the nearest centroid. Then recalculate the K centroid positions as follows:
Step 4. Repeat the process (equations (7) and (8)) until the centroids converge. The final clustering quality results are strongly correlated with the best choice of initial centroids. For the PEEM data structure of steel images, typically X (x × y × λ) is a cube of 3D array data, while the x and y axes represent spatial information (pixel resolution) and the λ axis corresponds to a photon. energy spectral picture. The K-means algorithm is used to explore regions of interest in X-PEEM data by separating pixels (clusters or sub-blocks) according to their spectral features and extracting the best centroids (XAS spectral profiles) for each analyte. cluster). It is used to study spatial distribution, local spectral changes, oxidation behavior, and chemical states. For example, the K-means clustering algorithm was used for Fe L-edge and Cr L-edge regions in hot-worked and cold-rolled X-PEEM. Various numbers of K clusters (regions of microstructure) were tested to find the optimal clusters and centroids. When these numbers are displayed, the pixels are reassigned to the corresponding cluster centroids. Each color distribution corresponds to the center of the cluster, showing the spatial arrangement of chemical or physical objects. The extracted centroids are linear combinations of pure spectra.
Data supporting the results of this study are available upon reasonable request from the respective WC author.
Sieurin, H. & Sandström, R. Fracture toughness of a welded duplex stainless steel. Sieurin, H. & Sandström, R. Fracture toughness of a welded duplex stainless steel. Sieurin, H. & Sandström, R. Вязкость разрушения сварной дуплексной нержавеющей стали. Sieurin, H. & Sandström, R. Fracture toughness of welded duplex stainless steel. Sieurin, H. & Sandström, R. 焊接双相不锈钢的断裂韧性。 Sieurin, H. & Sandstrom, R. 焊接双相不锈钢的断裂韧性。 Sieurin, H. & Sandström, R. Вязкость разрушения сварных дуплексных нержавеющих сталей. Sieurin, H. & Sandström, R. Fracture toughness of welded duplex stainless steels. Britannia. Fractional part. fur. 73, 377–390 (2006).
Adams, FV, Olubambi, PA, Potgieter, JH & Van Der Merwe, J. Corrosion resistance of duplex stainless steels in selected organic acids and organic acid/chloride environments. Adams, FV, Olubambi, PA, Potgieter, JH & Van Der Merwe, J. Corrosion resistance of duplex stainless steels in selected organic acids and organic acid/chloride environments. Adams, F.W., Olubambi, P.A., Potgieter, J. Kh. and Van Der Merwe, J. Corrosion resistance of duplex stainless steels in environments with some organic acids and organic acids/chlorides. Adams, FV, Olubambi, PA, Potgieter, JH & Van Der Merwe, J. 双相不锈钢在选定的有机酸和有机酸/氯化物环境中的耐腐蚀性。 Adams, FV, Olubambi, PA, Potgieter, JH & Van Der Merwe, J. 双相stainless steel在选定的organic酸和organic酸/chlorinated environment的耐而性性。 Adams, F.W., Olubambi, P.A., Potgieter, J. Kh. and Van Der Merwe, J. Corrosion resistance of duplex stainless steels in selected environments of organic acids and organic acids/chlorides. preservative. Materials Methods 57, 107–117 (2010).
Barrera, S. et al. Corrosion-oxidative behavior of Fe-Al-Mn-C duplex alloys. Materials 12, 2572 (2019).
Levkov, L., Shurygin, D., Dub, V., Kosyrev, K. & Balikoev, A. New generation of super duplex steels for equipment gas and oil production. Levkov, L., Shurygin, D., Dub, V., Kosyrev, K. & Balikoev, A. New generation of super duplex steels for equipment gas and oil production. Levkov L., Shurygin D., Dub V., Kosyrev K., Balikoev A. New generation of super duplex steels for oil and gas production equipment. Levkov L., Shurygin D., Dub V., Kosyrev K., Balikoev A. New generation of super duplex steels for gas and oil production equipment. Webinar E3S 121, 04007 (2019).
Kingklang, S. & Uthaisangsuk, V. Investigation of hot deformation behavior of duplex stainless steel grade 2507. Metall. Kingklang, S. & Uthaisangsuk, V. Investigation of hot deformation behavior of duplex stainless steel grade 2507. Metall. Kingklang, S. & Uthaisangsuk, V. Исследование поведения горячей деформации дуплексной нержавеющей стали марки 2507. Metall. Kingklang, S. & Uthaisangsuk, V. A Study of Hot Deformation Behavior of Type 2507 Duplex Stainless Steel. Metall. Kingklang, S. & Uthaisangsuk, V. 双相不锈钢2507 级热变形行为的研究。 Kingklang, S. & Uthaisangsuk, V. 2507 级热变形行为的研究。 Kingklang, S. and Utaisansuk, V. Investigation of the Hot Deformation Behavior of Type 2507 Duplex Stainless Steel. Metal. alma mater. trance. 48, 95–108 (2017).
Zhou, T. et al. Effect of controlled cold rolling on the microstructure and mechanical properties of cerium-modified super-duplex SAF 2507 stainless steel. alma mater. the science. Britannia. A 766, 138352 (2019).
Zhou, T. et al. Structural and mechanical properties induced by thermal deformation of cerium-modified super-duplex SAF 2507 stainless steel. J. Alma mater. storage tank. technology. 9, 8379–8390 (2020).
Zheng, Z., Wang, S., Long, J., Wang, J. & Zheng, K. Effect of rare earth elements on high temperature oxidation behaviour of austenitic steel. Zheng, Z., Wang, S., Long, J., Wang, J. & Zheng, K. Effect of rare earth elements on high temperature oxidation behavior of austenitic steel. Zheng Z., Wang S., Long J., Wang J. and Zheng K. Influence of rare earth elements on the behavior of austenitic steel under high temperature oxidation. Zheng, Z., Wang, S., Long, J., Wang, J. & Zheng, K. 稀土元素对奥氏体钢高温氧化行为的影响。 Zheng, Z., Wang, S., Long, J., Wang, J. & Zheng, K. Zheng Z., Wang S., Long J., Wang J. and Zheng K. Influence of rare earth elements on the behavior of austenitic steels at high temperature oxidation. koros. the science. 164, 108359 (2020).
Li, Y., Yang, G., Jiang, Z., Chen, C. & Sun, S. Effects of Ce on the microstructure and properties of 27Cr-3.8Mo-2Ni super-ferritic stainless steels. Li, Y., Yang, G., Jiang, Z., Chen, C. & Sun, S. Effects of Ce on the microstructure and properties of 27Cr-3.8Mo-2Ni super-ferritic stainless steels. Li Y., Yang G., Jiang Z., Chen K. and Sun S. Influence of Se on the microstructure and properties of superferritic stainless steels 27Cr-3,8Mo-2Ni. Li, Y., Yang, G., Jiang, Z., Chen, C. & Sun, S. Ce 对27Cr-3.8Mo-2Ni 超铁素体不锈钢的显微组织和性能的影响。 Li, Y., Yang, G., Jiang, Z., Chen, C. & Sun, S. Effects of Ce on the microstructure and properties of 27Cr-3.8Mo-2Ni super-steel stainless steel. Li, Y., Yang, G., Jiang, Z., Chen, C. & Sun, S. Влияние Ce на микроструктуру и свойства суперферритной нержавеющей стали 27Cr-3,8Mo-2Ni. Li, Y., Yang, G., Jiang, Z., Chen, C. & Sun, S. Effect of Ce on microstructure and properties of superferritic stainless steel 27Cr-3,8Mo-2Ni. Iron sign. Steelmak 47, 67–76 (2020).