Oxidation behavior and outward diffusion of Al along oxide grain boundaries of FeCrAl alloys overdoped with Zr and Hf by means of Atom Probe Tomography (APT)

doi:10.21203/rs.3.rs-2084870/v1

Download PDF

Research Article

Oxidation behavior and outward diffusion of Al along oxide grain boundaries of FeCrAl alloys overdoped with Zr and Hf by means of Atom Probe Tomography (APT)

https://doi.org/10.21203/rs.3.rs-2084870/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

It is generally accepted that the formation of α-Al₂O₃ scale on FeCrAl alloys is governed by inward oxygen transport along grain boundaries. However, it metal ion outward diffusion plays a role in the development of the oxide scales and their microstructural characteristics. This paper studies oxidation behavior and outward diffusion for iron-chromium alloys containing ~ 10 at. % aluminum, ~ 22 at. % Chromium, occurred when doped with an over-critical concentration of reactive elements, i.e. Zr and Hf. All samples were investigated after thermal exposure at 1100°C by Scanning Electron Microscopy (SEM), Transmission Electron Microscopy (TEM), and Atom Probe Tomography (APT). As a result of the overdoping, a considerable increase in oxide growth, an increase in the depth of internal oxidation, and RE-oxide formation near and at oxide Grain Boundaries (GBs) were observed as a consequence of increased inward and outward diffusion. The effect of overdoping manifests itself differently depending on the RE type and amount due to different solubility, ionic size, and electronic structure. The overdoped sample with Zr is the retainment of the adhesion of alumina to the alloy after the first and second thermal exposure, contrary to Hf overdoping which leads to severe spallation after the second thermal exposure. And FeCrAl-Zr overdoped sample formed ZrO at oxide grain boundary and internal Al oxide.

Overdoping

Segregation

Grain boundary

Alumina

Iron-chromium-aluminum (FeCrAl) alloys are promising materials concerning corrosion and oxidation resistance for high-temperature applications, e.g., heating elements [1]. They form an alumina scale (α-Al₂O₃) at 1100 ^ͦC, which is responsible for the excellent oxidation resistance. Alumina has excellent thermodynamical stability and a low concentration of defects because the point defect formation energy of alumina is known to be large, especially for charged defects [2]. However, alumina is susceptible to breakaway corrosion due to consuming the aluminum reservoir in the alloy after long thermal exposure times.

The effects of Reactive Elements (REs) on the oxidation process and on the formed oxide layer have been studied extensively over the past decades [3–7], and it has been found that REs (i) improve scale adhesion, i.e. alumina scale adhesion to FeCrAl alloy by forming what is called oxide pegs [8], (ii) reduce the parabolic growth, (iii) hinder the outward diffusion of cation ions [9], and (iv) decrease the critical amount of Al and Cr necessary to develop the protective Cr₂O₃ or Al₂O₃ scales [10]. The advantages offered by REs are believed to stem from their large ion size and high oxygen affinity [11–13].

The effect of the RE depends on their distribution and concentration [6]. Most of the recent works suggest that the RE concentration has to be around a few hundred parts per million [14–16]. Overdoping, however, can generate disadvantageous effects, for instance, formation of a second phase such as yttrium aluminum oxide (Y₃Al₅O₁₂-YAG) [17, 18]. Also, high amounts of yttrium cause Y-rich oxide particles to form at alumina Grain Boundaries (GBs), which has a detrimental effect on scale adhesion by enabling micro-cracks and voids [6, 19]. This paper will discuss the effect of overdoping of Hf and Zr on the α-Al₂O₃ formation on FeCrAl alloys.

Alumina grows mainly by inward diffusion [20], which has been confirmed using ¹⁸O tracer [15, 21, 22]. The presence of two distinguished layered two-zone microstructures (columnar and equiaxed alumina layer) [23–26], and the formation of RE-rich oxide at the surface and along the GB is considered conclusive evidence of outward diffusion of cations [12, 27]. The oxygen chemical potential gradient across the scale is the driving force for the outward diffusion[28] and the concentration gradient is inversely proportional to the oxide thickness. So, the analysis of the new oxide formed along the GB ridge volume as a function of time and alumina scale thickness can give quantitative data on the magnitude of the GB flux and aluminum diffusivity. Nychka’s paper[29] and Torben’s paper[30] give a quantitive calculation of the Al GB flux based on GB ridge volume, where the presented data in both papers agree with the 1st Fick’s law.

2.1 Sample preparation and exposure

The samples were prepared from high purity elements by arc melting technique under a protective Ar gas atmosphere. The compositions were determined by Spark-Optical Emission Spectrometry (Spark-OES/ for Fecralloy™ & FeCrAl-Zr) and Energy Dispersive Spectroscopy (EDS/ for Fecralloy™, FeCrAl-Zr & FeCrAl-Hf), the results are given in Table 1. Clearly, one notes Zr and Hf levels of around 0.5 at.% being much higher than above mentioned optimum doping level of several 100 ppm.

Table 1

Alloys composition as analyzed by ICP-OES/EDS
Alloy/Compositions (at. %)	Fe	Al	C	Cr	Mn	Si	Y	Zr	Hf
Fecralloy™ (OES)	Bal.	9.65	0.09	22.03	0.19	0.56	0.06	0.06	-
FeCrAl-Zr (OES)	Bal.	10.27	-	22.45	-	-	-	0.5	-
FeCrAl-Hf (EDS)	Bal.	9.64	-	22.74	-	-	-	-	0.52

The samples were sectioned using a diamond wire saw with a thickness of 2.5 mm, then annealed in air at 1200°C for 24 hours for homogenization. The samples were polished with abrasive SiC paper down to 4000-grit, then using oil-base diamond suspension down to 1 µm and ultrasonically cleaned in alcohol and distilled water for 10 min.

The first high-temperature oxidation experiments were carried out in a chamber furnace in air at 1100 °C for 200 hours to form an adherent alumina scale, followed by cooling to room temperature in air. The following procedures were carried out using experimental methods suggested in[31] to quantify the outward diffusion along GB in α-Al₂O₃ based on the formed ridge volume after second thermal exposure. Fig. 1 shows a schematic illustration of the used experimental techniques to investigate the outward diffusion of aluminum along the α-Al₂O₃ GBs. Starting with the previously mentioned steps and followed by a taper polishing to provide a variation of the oxide thickness, this is evident in the formation of various-sized ridges after the second oxidation. The taper polishing was done using oil-base diamond starting with 1 μm down to 0.04 μm particle size using colloidal silica. Also, the purpose of the taper polishing is to remove the outer fine-grained layer and expose the larger columnar grains underneath to provide an unobstructed view of the formed oxide after the second thermal exposure. Because a relatively long time will result in the growth of congested ridges zone the second thermal exposure must be adapted depending on ridge size evolution with timeOptimum conditions were found to be 60 minutes for Fecralloy^TM, 20 minutes for FeCrAl-Zr, and 40 minutes for FeCrAl-Hf. After the second exposure, a new oxide forms along the top of the GBs of the columnar oxide with time as aluminum segregation continues along the underlying alumina GBs.

As α-Al₂O₃ is an electrical insulator charging effect has been avoided by depositing a thin layer of silver (~ 200 nm) on the surface of the sample prior to SEM, and TEM sample investigation.

2.2 Analytical techniques

The oxide scale of Fecralloy™ and FeCrAl-Zr were investigated by means of scanning electron microscopy/focused ion beam (SEM/FIB-Zeiss Auriga 60), transmission electron microscope (FEI Thermo Fisher Themis Z), and laser-assisted atom probe tomography (APT-CAMECA LEAP 4000 XHR). The latter is equipped with a UV laser (wavelength 355 nm) and a spot size less than 1 µm. After the second exposure, the FeCrAl-Hf sample had significant spallation and cracks at most of the regions, due to the formation of Hf precipitates at oxide/alloy interface, hence TEM and APT were not employed in this condition.

The overdoping effect on oxide morphology was investigated using SEM/FIB-Zeiss Auriga 60 after the first and second thermal exposure. Also, TEM and APT samples were prepared using SEM/FIB with a maximum accelerating voltage of 30 kV for thinning steps and a minimum accelerating voltage of 2 kV to minimize Ga ions damage to the sample. The preparation of APT and TEM samples is very similar, starting from creating a lamella containing alumina GBs to attaching the lamella to Micro-post/TEM grid, but the next steps are different [32, 33]. The needle-shaped APT specimens with a shank angle of about 10° and an apex radius of less than 100 nm were prepared by the lift-out technique. And the samples containing oxide GB oxide were aligned approximately parallel to the field evaporation direction. The importance of attaining this particular APT sample shape is to acquire a high electric field of the order of 10¹⁰ V/m [34], to induce the removal of a conductive emitter surface atom as an ion. In case of poor electrical conductivity materials; like alumina, a combination of ultra-short laser pulses and high electrical voltage allows for studying this kind of material [35]. The tips were analyzed in the APT at a temperature of ~ 35 K, a UV-laser pulse frequency of 125–200 kHz with a laser pulse energy of 30–40 pJ, and a detection rate of 0.25%. Reconstruction of the atom probe data was performed using the AP suite 6 software based on the tip evolution profile. Thermo-Fisher Themis-Z has been operated in S/TEM mode, at an accelerated voltage of 300 keV. Both Bright-Field (BF) and High Angle Annular Dark-Field (HAADF) images mode were used.

The Gibbsian interfacial excess value Γ has been calculated, which indicates the solute atoms at internal interfaces. It is defined as [36]:

$${{\Gamma }}_{i}= {N}_{i}^{excess}/A=\left(1/A\right)\left({N}_{i}^{vol}-{N}_{i}^{\alpha }-{N}_{i}^{\beta }\right)=\left(1/A\right){N}^{vol}[{C}_{i}^{vol}-{C}_{i}^{\alpha }\xi -{C}_{i}^{\beta }\left(1-\xi \right)]$$

where ${N}_{i}^{excess}$is the excess number of atoms at an interface, A is the interfacial area over which ${{\Gamma }}_{i}$ is determined. The ${N}_{i}^{\alpha }$ and ${N}_{i}^{\beta }$ are the number of atoms in α-phase and β-phase adjacent to the interface, $\xi$ is the Gibbs dividing surface and C represents the concentration of an element $i$ in phase α or phase β. For more clarification [37].

2.3 Grain Boundary Outward Aluminum Flux

As previously stated, there were two oxidation treatments carried out, the first oxidation treatment to obtain a thick α-Al₂O₃ oxide scale, and a re-oxidation treatment after wedge polishing to observe ridges. The ridge volume measurements were obtained using SEM/FIB by making a cross-section, where the FIB acceleration voltage was 30 kV. For the cleaning cross-section, 5 kV was used with an emission current of 50 pA. After sectioning through the oxide at 0°, the SEM stage was tilted to 45° to view in order to measure the cross-sectional area of ridge ${A}_{ridge}$. The aluminum GB flux ${J}_{Al}$ is defined as the number of ions per length and time t. By assuming that the outward diffusion along an isolated GB without any lateral leakage of atoms from GBs to adjacent grains, it can be calculated using the relationship [38]:

$${J}_{GB}^{Al}= \frac{{N}_{GB}^{Al}}{{L}_{GB} . t}$$

where ${L}_{GB}$ is the lateral length of the GB, ${N}_{GB}^{Al}$ is the number of segregated Al atoms along GB, where ${N}_{GB}^{Al}$can be obtained from the measured grain boundary ridge volume ${V}_{m}={A}_{ridge} \bullet {L}_{GB}$. The unit cell volume of α-Al₂O₃ (${V}_{unitcell}$) is 0.254 nm³ and with the number of Al ions per unit cell of α-Al₂O₃ (${N}_{unitcell}^{Al}$) being 12, one yields:

inserting Eq (3) into Eq (2), the aluminum GB flux can be determined as follows:

3.1 Oxide scale microstructure and composition (SEM).

After 200 hours of oxidation at 1100 °C, the average oxide scale thickness of Fecralloy^TM, FeCrAl-Zr, and FeCrAl-Hf was found to be 3.1 μm, 5.6 μm, and 4.9 μm, respectively. Fig. 2 shows SEM images of the cross-sectional area of the oxidized samples. Both FeCrAl-Zr and FeCrAl-Hf samples formed internal oxides, while Fecralloy^TM did not form internal oxides. The average internal oxide depth for FeCrAl-Zr and FeCrAl-Hf is 79.2 μm and 18.2 μm, respectively.

The samples were re-polished at an angle of 0.08° and then re-oxidized at 1100 °C in air. Fe-based oxide layer spalled off completely, revealing the partially spalled alumina oxide scale containing ridges, see Fig. 3 (c). In some areas, all the formed oxide is spalled off, and the alloy surface is exposed, see Fig. 3 (a). Contrary to that, at the center of the sample, it seems that the alumina oxide is embedding other oxides, notably hafnia, see Fig. 3 (d).

3.2 Oxide scale and composition (TEM).

After the second exposure, high-resolution electron imaging was performed using the FEI Thermo Fisher Themis Z for all samples, except for FeCrAl-Hf due to the severe spallation that occurred after the second exposure. A high-angle annular dark field (HAADF) was used to record images in scanning mode (STEM) with an accelerating voltage of 200 kV. Also, energy-dispersive X-ray spectroscopy (EDX) mapping was acquired along with alumina GB.

Fig. 4 (b,c,d) shows the elemental maps of Fe, Cr, and Zr for Fecralloy^TM revealing no segregation of Fe and Cr to the oxide layer. While Zr has segregated to the oxide layer. The cross-sectional HAADF-STEM image of the oxide scale grown on the FeCrAl overdoped with Zr alloy (FeCrAl-Zr) and EDX elemental mapping images are shown in Fig. 5.

There is no clear evidence from the elemental maps in Fig. 5 (b,c,d) of Fe segregation to the alumina oxide layer, but Fig. 5 (c) shows a Cr enrichment region formed at the alumina oxide layer. A 1-D concentration profile was used to identify and scrutinize Cr and Fe segregation to the alumina scale.

The white arrow in Fig. 6 (a) indicates the direction of the concentration profile analysis. The Cr segregated along alumina GB, as indicated in the concentration profile in Fig. 6 (c), where the atomic concentration of Cr suddenly increased in the GB region. It is also evident from the same figure that the Fe concentration at alumina grains is higher than it is along the alumina GB. The Cr and Fe concentration increased to 0.14, 0.05 at. % (left Cr-precipitate region on the concentration profile line) and 0.13, 0.04 at. % (right Cr-precipitate region on the concentration profile line), respectively, as shown in Fig. 6 (b). The presence of Zr in another region of the alumina containing GB was also investigated, see Fig. 7. Based on the concentration profile in Fig. 7 (b), the Zr concentration at the alumina GB shows a sudden increase to 7.27 at. %.

3.3 Atom probe tomography (APT)

APT was utilized to examine the microstructure in greater detail and gain a better understanding of the distribution of the atoms in the oxide scale. An APT analysis of a volume containing alumina GB for Fecralloy^TM and FeCrAl-Zr, as well as the oxide-alloy interface of FeCrAl-Zr, are given in the next figure.

Fig. 8 depicts the elemental map distribution of Fecralloy^TM’s oxide scale after the second oxidation. Elemental enrichment at these characteristics often allows APT to monitor the GBs. Figure 9 (c) shows the location of the GB on the APT sample, which indicates a sudden increase in the density of carbon atoms at GB, and there was no hint of Fe or Cr segregation at the GB. Further investigations were conducted by creating a 1-D concentration map across the GB, which shows a decrease of O around the GB and an increase in the concentration of O₂, see Fig. 9 (c).

In case of FeCrAl-Zr, two samples of oxide scale containing GB and a sample containing oxide-alloy interface were run successfully. As shown in Fig. 10 (a,b,c,d), the 3-dimensional atom maps exhibit a triple junction point at the oxide-alloy interface. The triple junction point is confined between alloy and two oxide grains and it has been identified by its hump-like location and shape (e.g. Fig. 5a and 8a show the same hump-like at the triple junction point). According to these figures, segregation of Zr and Fe along oxide GB occurred which coincides with the TEM results. Also, Fig. 10 (a) reveals the formation of the Fe-rich band, and Cr-rich band (Fig. 10 (b)) near the alumina GB and above each other. According to the proxigram profile of the triple junction, it seems that Cr is the major constituent compared to Fe and Zr. Further investigation was conducted by performing a 1D concentration profile on the GB at two positions, one close to the triple junction and the other further away, see Fig. 12. The Zr concentration relatively far away from the triple junction is nearly equal to its concentration close to the triple junction, and Fig. 12 (b). The Gibssian excess of Zr for the two ROIs in Fig. 12) a,c) were calculated, and it was

Fig. 13 shows the elemental map distribution of Zr atoms (right) and ZrO (left), and it shows zirconium oxide (ZrO) forming at the alumina GB. The segregation of Zr at GB is expected and has already been reported in several papers [39–42]. But preferentially RE oxide formed at oxide GB due to the RE amalgamated with a relatively high ratio into the scale. Additionally, the presence of Zr precipitate close to the GB may facilitate Zr segregation along the oxide GB.

APT tip-sample containing oxide-scale interface was studied, Fig. 14 shows the 3D atom map distribution of Fe, Cr, Zr, Al, and O. It can be inferred from the images (Fig. 14 (c)) below, that Zr segregated to the oxide layer. Fe and Cr segregation through the oxide-alloy interface was identified, and it was confirmed using the 1D concentration profile, see Fig. 15. 1D concentration profile was generated at four segments on the oxide-alloy interface, it can be seen that the value of the atomic concentration of Fe at the alloy region (ID6 figure) is at its maximum. And based on the curve of the interface near the bottom of the 3D elemental map (Fig. 14), it seems that it’s near a triple junction region, Also, Fig. 15 depicts the concentration variation across the oxide-alloy interface. On average, 29.22 at. % / 1.24 at. % of Fe is present in the oxide-alloy interface/oxide grain, respectively. Likewise, 6.12 at. % / 0.39 at. % of Cr and 0.097 at. % / 0.11 at. % of Zr are present in the oxide/alloy and oxide grain, respectively.

3.4 Outward flux of Al along alumina GB.

The data acquired on the GB aluminum flux is shown in Fig. 16, which is plotted using a double-logarithmic of the oxide thickness against the Al flux along GB J. Both samples (Fecralloy^TM and FeCrAl-Zr) followed 1^st Fick law.

The methodology of sample preparation and quantifying the Al flux along oxide GB is based on the second polishing step and tuning of the second oxidation time to have a specific microstructure on the oxide surface (ridge), where its growth is time and oxide thickness dependent. The second polishing is considered a critical stage, but it is also a contentious step since polishing could introduce additional dislocation, and subsequently, yield a higher flux. To reduce this effect, the second polishing should be done using as lower as possible polishing particle size and for at least 11 minutes. Using some other milling techniques, for instance, Xe gas miller has a disadvantageous point, where it has been reported [30] that it affects Al flux, where it can impede in oxide GB, consequently hindering Al segregation along GB. They found that the mechanical polished Hf-doped sample exhibited a higher flux for the same specie compared to the Xe-milled of the same sample.

The straightforward formula to evaluate the flux along oxide GB has disregarded several factors that could affect the results, for instance, low and high angle GB is not addressed in the equation, since these two factors have an impact on segregation. The leakage of atoms from GBs to adjacent grains was also ignored. All of these variables influence the result. But the equation can serve its purpose by giving an estimated value of Al flux along GB and it follows 1st Fick law. Also, it seems that our results correlate with Nychka's results in Fig. 16 (a) with (b) and it also complies the 1st Fick law. Based on observing the SEM images, concentration profiles, and Al flux values, it is evident that the alloy constituents (especially Al ions) flux along GB has been enhanced as an effect of overdoping contrary to REs beneficial effect on the outward diffusion of Al in the Fecralloy™ sample.

The Fecralloy™ formed the minimum oxide thickness with no internal oxidation after the first and second oxidation at 1100°C, demonstrating the feasibility of RE addition when doped under defined conditions [6, 12]. Overdoped sample with Hf has drastic consequences, such as reduced oxide scale adhesion, formation of a second phase oxide such as HfO₂ or intermetallic compound with Al and it can promote internal oxidation by enhancing inward diffusion [4]; occurring counter-diffusion of minority and majority point defects, and increasing the growth rate [43]. That’s what we observe in Fig. 3, and 5, where FeCrAl-Hf and FeCrAl-Zr oxidized internally, resulting in an alumina scale of inhomogeneous thickness. After the oxidation process, FeCrAl-Zr formed internal oxide of alumina, and according to APT reconstruction result in Fig. 13, ZrO formed near and at alumina oxide GB, also Fe and Cr segregated and formed FeO and CrO. According to the simple theory of equilibrium segregation [44], adding an extra amount of RE would raise the concentration of segregant at GB until it reaches the saturation concentration leading to the formation of RE oxides [45]. Other papers proposed that Increasing the amount of [46]REs increases the disorder in the alumina lattice allowing improved diffusion rate through the oxide [46].

APT reconstruction analysis coincides with STEM/EDS results, where both showed no segregation of Fe or Cr at alumina oxide GBs in the Fecralloy™ sample, see Figs. 4 and 9. But in FeCrAl-Zr alloy, Cr (see Fig. 6 (c)) and Zr (see Fig. 7 (b) and Figs. 10 and 13) have segregated along oxide GBs. The formation of a Cr-precipitate near the GB could contribute to the segregation of Cr along GB. Also, Since RE has a high affinity for oxygen, ZrO has been detected in the oxide GB of FeCrAl-Zr using APT reconstructed elemental maps (see Fig. 13).

Regarding the surface morphology of overdoped samples, it was expected to form Hf-oxide on the surface after the first oxidation, since it was doped with a high amount of Hf, as seen in Fig. 3 (b,d). The major undesirable effect is the oxide scale spalled out due to the formation of HfO₂ on the surface and at the oxide/scale interface after the first exposure which has a different thermal expansion factor compared to other formed oxides [47, 48]. But it was not expected with the overdoped FeCrAl-Zr sample to not form a RE-oxide at the surface and stand against the isothermal exposure.

The oxidation behavior of Fecralloy™ and FeCrAl overdoped samples with Hf/Zr were investigated after the first thermal exposure at 1100°C followed by an isothermal exposure. The obtained results can be summarized as follows:

The inward diffusion was enhanced in the overdoped samples where the inward diffusion is more enhanced by overdoping with Zr than overdoping with Hf based on the depth of internal oxidation.

The outward diffusion of Al along oxide GBs was enhanced in the overdoped sample with Hf/Zr according to the thickness of the oxide after the first oxidation and according to the Al GB flux J values (see Fig. 16).

In contrast to the Fecralloy^™ sample, Fe and Cr segregations were seen in oxide grains in the overdoped samples as a result of the overdoping. Zr (from STEM-EDS) and ZrO (from APT analysis) were identified in the oxide GB of the overdoped sample with Zr.

The beneficial effects including the hindered the outward diffusion of Al and other constituents along the alumina GBs, reducing the growth rate, and enhanced alumina scale adhesion to the alloy were observed in the Fecralloy^™ sample. While the overdoped sample with Hf severed strong spallation after the second exposure which is relatively much higher than the overdoped sample with Zr.

The susceptibility to overdoping effects depends on the type and the amount of the RE and the Al concentration in the alloy.

Acknowledgments The authors would like to thank the German Academic Exchange Service (DAAD) scholarship. This work was carried out with the support of the Karlsruhe Nano Micro Facility (KNMFi, www.knmf.kit.edu), a Helmholtz Research Infrastructure at Karlsruhe Institute of Technology (KIT, www.kit.edu) (project no. ha 2020-024-028730). Also, special thanks to Prof. Dr.-Ing. Martin Heilmaier for the scientific discussion.

Conflict of Interest The authors declare that they have no known competition for financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Sundberg M, Malmqvist G, Magnusson A, El-Raghy T. Alumina forming high temperature silicides and carbides. In: Ceramics International. Elsevier; 2004.30:1899–1904.
Kitayama M. The Wulff Shape of Doped and Undoped Alumina. California, USA: University of California, Berkeley; 1996.
Chevalier S. What did we learn on the reactive element effect in chromia scale since Pfeil’s patent? Materials and Corrosion. 2014;65:109–115.
Hou PY. The reactive element effect - Past, present and future. Materials Science Forum. 2011;696:39–44.
Pint BA, Martin JR, Hobbs LW. The oxidation mechanism of θ-Al2O3 scales. Solid State Ion. 1995;78:99–107.
Quadakkers J, Singheiser L. Practical aspects of the reactive element effect. Materials Science Forum. 2001;369–372:77–92.
Stringer J. The reactive element effect in high-temperature corrosion. Materials Science and Engineering A. 1989;120–121:129–137.
Eklund J, Jönsson B, Persdotter A, Liske J, Svensson JE, Jonsson T. The influence of silicon on the corrosion properties of FeCrAl model alloys in oxidizing environments at 600 °C. Corros Sci. 2018;144.
Swadźba R. Influence of hafnium on high temperature oxidation of NiAl and Ni3Al alloys. Prace Instytutu Metalurgii Żelaza. 2018;70.
Hou PY, Stringer J. The influence of ion-implanted yttrium on the selective oxidation of chromium in Co-25 wt.% Cr. Oxidation of Metals. 1988;29:45–73.
Mennicke C, Schumann E, Rühle M, Hussey RJ, Sproule GI, Graham MJ. The effect of yttrium on the growth process and microstructure of α-Al2O3 on FeCrAl. Oxidation of Metals. 1998; doi:10.1023/A:1018803113093.
Pint BA. Progress in understanding the reactive element effect since the Whittle and Stringer literature review. In: Proc. John Stringer Symposium on High Temperature Corrosion. Citeseer; 2003.9–19.
Wang X, Peng X, Tan X, Wang F. The reactive element effect of ceria particle dispersion on alumina growth: A model based on microstructural observations. Sci Rep. 2016;6:1–10.
Pedraza F, Troncy R, Pasquet A, Delautre J, Hamadi S. Critical Hafnium Content for Extended Lifetime of AM1 Single Crystal Superalloy. In: Superalloys 2020. Springer; 2020:781–788.
Pint BA. Optimization of Reactive-Element Additions to Improve Oxidation Performance of Alumina-Forming Alloys. Journal of the American Ceramic Society. 2003;86:686–95.
Smialek JL. Invited Review Paper in Commemoration of Over 50 Years of Oxidation of Metals: Alumina Scale Adhesion Mechanisms: A Retrospective Assessment. Oxidation of Metals. 2022;1–50.
Cueff R, Buscail H, Caudron E, Riffard F, Issartel C, el Messki S. Effect of reactive element oxide coating on the high temperature oxidation behaviour of FeCrAl alloys. Appl Surf Sci. 2004;229:233–241.
Mortazavi N, Geers C, Esmaily M, et al. Interplay of water and reactive elements in oxidation of alumina-forming alloys. Nat Mater. 2018;17:610–617.
Gottstein G. Physical Foundation of Material Science. Nature. 2004.282.
Engkvist J, Bexell U, Grehk M, Olsson M. High temperature oxidation of FeCrAl‐alloys–influence of Al‐concentration on oxide layer characteristics. Materials and corrosion. 2009;60:876–881.
Quadakkers WJ, Holzbrecher H, Briefs KG, Beske H. Differences in growth mechanisms of oxide scales formed on ODS and conventional wrought alloys. Oxidation of Metals. 1989;32:67–88.
Reddy KPR, Smialek JL, Cooper AR. 18O Tracer studies of Al2O3 scale formation on NiCrAl alloys. Oxidation of Metals. 1982;17:429–449.
Golightly FA, Stott FH, Wood GC. The influence of yttrium additions on the oxide-scale adhesion to an iron-chromium-aluminum alloy. Oxidation of Metals. 1976;10:163–187.
Pint BA. Experimental observations in support of the dynamic-segregation theory to explain the reactive-element effect. Oxidation of metals. 1996;45:1–37.
Wessel E, Kochubey V, Naumenko D, Niewolak L, Singheiser L, Quadakkers WJ. Effect of Zr addition on the microstructure of the alumina scales on FeCrAlY-alloys. Scr Mater. 2004;51:987–992.
Huang T, Naumenko D, Song P, Lu J, Quadakkers WJ. Effect of titanium addition on alumina growth mechanism on yttria-containing FeCrAl-base alloy. Oxidation of metals. 2018;90:671–690.
Tolpygo VK, Clarke DR. Microstructural evidence for counter-diffusion of aluminum and oxygen during the growth of alumina scales. Materials at High Temperatures. 2003;20:261–271.
Kitaoka S, Matsudaira T, Wada M. Mass-transfer mechanism of alumina ceramics under oxygen potential gradients at high temperatures. Mater Trans. 2009;50:1023–1031.
Nychka JA, Clarke DR. Quantification of aluminum outward diffusion during oxidation of FeCrAl alloys. Oxidation of Metals. 2005;63:325–352.
Boll T, Unocic KA, Pint BA, Mårtensson A, Stiller K. Grain Boundary Chemistry and Transport Through Alumina Scales on NiAl Alloys. Oxidation of Metals. 2017;88:469–479.
Tolpygo VK, Clarke DR. Evidence of outward aluminum grain-boundary diffusion in Y-and Zr-doped alumina scales. High Temperature Corrosion and Materials Chemistry. New Jersey, USA: The Electrochemical Society; 1999.99–38.
Cerezo A, Clifton PH, Galtrey MJ, et al. Atom probe tomography today. Materials Today. 2007;10:36–42.
Rao SD, Muraleedharan K, Humphreys CJ. TEM specimen preparation techniques. Microscopy: Science, Technology, Applications and Education. 2010;2:1232.
Chen YM, Ohkubo T, Hono K. Laser assisted field evaporation of oxides in atom probe analysis. Ultramicroscopy. 2011;111:562–566.
Miller MK, Forbes RG. Atom-probe tomography: The local electrode atom probe. Atom-Probe Tomography: The Local Electrode Atom Probe. 2014.9781489974.
Krakauer BW, Seidman DN. Absolute atomic-scale measurements of the Gibbsian interfacial excess of solute at internal interfaces. Phys Rev B. 1993;48:6724–6727.
Miller MK, Smith GDW. Atom probe analysis of interfacial segregation. Appl Surf Sci. 1995;87–88:243–250.
Boll T, Unocic KA, Pint BA, Stiller K. Interfaces in Oxides Formed on NiAlCr Doped with Y, Hf, Ti, and B. In: Microscopy and Microanalysis. 2017.23.
BUBAN JP, MIZOGUCHI T, SHIBATA N, ABE E, YAMAMOTO T, IKUHARA Y. Zr segregation and associated Al vacancies in alumina grain boundaries. Journal of the Ceramic Society of Japan. 2011;119:840–844.
Behera SK, Cantwell PR, Harmer MP. A grain boundary mobility discontinuity in reactive element Zr-doped Al2O3. Scr Mater. 2014;90–91:33–36.
Wang CM, Reddy JN, Lee KH. Relationships for Inhomogeneous Plates. Shear Deformable Beams and Plates. 2000;243–277. doi:10.1016/B978-008043784-2/50013-7.
Lallemant L, Roussel N, Fantozzi G, et al. Effect of amount of doping agent on sintering, microstructure and optical properties of Zr- and La-doped alumina sintered by SPS. J Eur Ceram Soc. 2014;34:1279–1288.
Hayashi S, Maeda T. Effect of Zr on Initial Oxidation Behavior of FeCrAl Alloys. Oxidation of Metals. 2020;93:573–586.
HONDROS ED, SEAH MR, HOFMANN S, LEJČEK P. INTERFACIAL AND SURFACE MICROCHEMISTRY. In: Physical Metallurgy. Elsevier; 1996:1201–1289. doi:10.1016/B978-044489875-3/50018-1.
Wakai F, Nagano T, Iga T. Hardening in Creep of Alumina by Zirconium Segregation at the Grain Boundary. Journal of the American Ceramic Society. 1997;80:2361–2366.
Klöwer J. Factors affecting the oxidation behaviour of thin Fe-Cr-Al foils. Part II: The effect of alloying elements: Overdoping. Materials and Corrosion. 2000;51:373–385.
Stott FH. The oxidation of alumina-forming alloys. Materials Science Forum. 1997;251–254:19–32.
Leyens C, Pint BA, Wright IG. Effect of composition on the oxidation and hot corrosion resistance of NiAl doped with precious metals. Surf Coat Technol. 2000;133–134:15–22.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Oxidation behavior and outward diffusion of Al along oxide grain boundaries of FeCrAl alloys overdoped with Zr and Hf by means of Atom Probe Tomography (APT)

Status:

Version 1

Abstract

Figures

1. Introduction

2. Materials And Methods

2.1 Sample preparation and exposure

2.2 Analytical techniques

2.3 Grain Boundary Outward Aluminum Flux

3. Results

3.1 Oxide scale microstructure and composition (SEM).

3.2 Oxide scale and composition (TEM).

3.3 Atom probe tomography (APT)

3.4 Outward flux of Al along alumina GB.

4. Discussion

5. Conclusion

Declarations

References

Additional Declarations

Status:

Version 1