Corrosion Science 221 (2023) 111348

Contents lists available at ScienceDirect

Corrosion Science
journal homepage: www.elsevier.com/locate/corsci

Revealing the elusive role of water vapor in the oxidation behavior of a Mn-Si
containing NiCr alloy at 950 ◦ C✩
M. Romedenne a ,∗, S. Lambeets b , M. Song b , C. Roach b , A. Devaraj b , R. Pillai a
a
Materials Science and Technology Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831-6156, USA
b
Physical and Computational Sciences Directorate, Pacific Northwest National Laboratory, Richland, WA, USA

ARTICLE INFO ABSTRACT

Keywords: Detailed nanoscale characterizations revealed the formation of a considerably higher fraction of SiO2 particles
Ni-based alloys in the external Cr2 O3 layer on a NiCr alloy. In addition, atom probe tomography revealed substantially
Transmission electron microscopy lower segregation of Mn at the Cr2 O3 grain boundaries in wet air compared to dry air exposures, where
Atom probe tomography
the intriguing formation of Mn-rich clusters was observed in the Cr2 O3 grains for the very first time. These
Chromia
findings provided additional evidence regarding the impact of distinct transport processes through the Cr2 O3
Grain boundaries
layer on the resulting oxidation rates and subsurface alloy compositional changes.

1. Introduction oxidized in Ar-4%H2 -7%H2 O compared to Ar-20%O2 at 1000 ◦ C were

suggested to be the result of greater inward oxygen transport through
The envisioned use of green hydrogen as fuel in the future energy oxide grain boundaries enabled by the much finer Cr 2 O3 grain size
landscape [1] for power generation turbines, and perhaps for hydrogen in the presence of water vapor [11]. The segregation of Si within the
combustion engines in freight vehicles, presents new challenges for Cr 2 O3 grains and grain boundaries was observed to affect the Cr 2 O3
selection and development of suitable high temperature materials. The growth for Fe-20Cr-0.5Si after exposure in Ar-20%O2 and Ar-20%CO2 -
burning of hydrogen can create higher operating temperatures and 20%H2 O at 650 and 818 ◦ C [7]. More recently, combined segregation
will result in higher water vapor content in exhaust gas than typically of Si, Mn, Nb and Ti within Cr 2 O3 grains and grain boundaries were
observed when burning natural gas (about 10 vol.% H2 O) [2]. The associated with changes in the oxide mass transport properties for
combined effects of water vapor contents in the combustion products various FeMnTiNbSi alloy variants after oxidation in Ar-4%H2 -4%H2 O
and alloy compositions on the long-term high temperature oxidation at 800 ◦ C [8,9]. Pillai et al. [10] investigated the role of typical alloying
behavior of existing candidate structural materials such as heat resis- additions such as Mn, Si, Al and Ti on the oxidation behavior of
tant steels (FeNiCr) and nickel based alloys (NiCr) alloys are not well model Ni–22Cr–X (X = Mn, Si, Al, Ti) alloys exposed in dry air and
understood. air + 10%H2 O between 800 ◦ C and 950 ◦ C. Exposures in water vapor
The attractive high-temperature oxidation behavior of these alloys reduced the Mn consumption from the Ni-22Cr-Mn-Si alloy suggesting
can mainly be attributed to their Cr content (>18 wt%) which ensures a slower growth of the Cr 2 O3 scale while also resulting in a more
the formation of external protective Cr 2 O3 scales [3,4]. The growth of porous Cr 2 O3 scale which subsequently led to increased nitridation in
Cr 2 O3 scales is governed by both diffusion in the oxide lattice and at the Ni-22Cr-Al-Ti alloy.
oxide grain boundaries [5,6]. Other alloying constituents such as Mn, The current paper is the continuation of a previously published
Si, Al, Ti, Nb in these alloys have been shown to influence these trans- work [10] with a focus on the Ni-22Cr-Mn-Si (NCMS) alloy system.
port processes in the Cr 2 O3 scales [7–10]. Furthermore, water vapor is Ni-22Cr-Mn-Si model alloys were exposed at 950 ◦ C in 1h-cycles
known to either disrupt the formation [11] or accelerate the breakdown in dry air and air + 10% H2 O. Detailed nanoscale characterization
of the Cr 2 O3 scales due to their susceptibility to volatilization in the including transmission electron microscopy (TEM) and atom probe
presence of water vapor [12]. Faster scaling rates observed for Ni-25Cr

✩ This manuscript has been authored by UT-Battelle, LLC under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy. The United States
Government retains and the publisher, by accepting the article for publication, acknowledges that the United States Government retains a non-exclusive, paid-
up, irrevocable, world-wide license to publish or reproduce the published form of this manuscript, or allow others to do so, for United States Government
purposes. The Department of Energy will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan
(http://energy.gov/downloads/doe-public-access-plan).
∗ Corresponding author.
E-mail address: romedennem@ornl.gov (M. Romedenne).

https://doi.org/10.1016/j.corsci.2023.111348
Received 28 February 2023; Received in revised form 8 June 2023; Accepted 21 June 2023
Available online 28 June 2023
0010-938X/© 2023 Elsevier Ltd. All rights reserved.
M. Romedenne et al. Corrosion Science 221 (2023) 111348

tomography (APT) was employed to investigate the local changes in Table 1

the oxide constitution and chemistry, as well as diffusion through the Composition of studied model alloys (in wt.%) analyzed by inductively coupled plasma–
optical emission spectrometry (ICP–OES) and combustion analysis for carbon and
oxide scales supporting oxidation rates and impact on underlying alloy sulfur.
compositional changes.
Composition Ni Cr Mn Si C S
NCMS wt.% 76.86 21.17 1.43 0.50 174 wppm 4 wppm
Summary of previous SEM, EDS and EBSD analyses
NCMS at.% 74.32 23.11 1.48 1.01 823 appm 7 appm
NCM wt.% 76.30 21.75 1.93 0.01 <100 wppm 10 wppm
Measured oxidation kinetics and SEM/EDS analyses of correspond- NCS wt.% 76.23 21.85 <0.01 1.91 <100 wppm <10 wppm
ing NCMS specimens after exposures for 500 h in dry air and air +
10%H2 O at 800 ◦ C and 950 ◦ C were presented in the previous study
by the authors [10]. Results in water vapor were not as anticipated,
(EDS) and Electron Backscattered Diffraction (EBSD) using a TESCAN
suggesting decreased scaling and spallation kinetics for these NCMS
MIRA3 SEM equipped with an EDAX Velocity Plus EBSD camera and
model alloys compared to dry air after 500 1 h cycles at 950 ◦ C.
TEAM EBSD software package from EDAX. Specimens for transmission
Based on the results, it was proposed that the presence of water
vapor possibly reduced the free sites for Mn ions in the Cr 2 O3 lattice electron microscopy (TEM) were prepared using a FEI Helios Nanolab
and slowed the growth of the Cr 2 O3 and thereby the Cr depletion 600 focused ion beam (FIB)-SEM with 30 kV Ga ion milling. A TEM
in the alloy. It was hypothesized that outward Mn transport through (Titan ETEM 80–300) aberration-corrected environmental transmission
the Cr 2 O3 scale was slower than inward oxygen transport during the electron microscope (Thermo Fisher Scientific, USA) was employed at
wet air exposure and the Cr 2 O3 scale most likely continued to grow 300 kV for bright field (BF) imaging and selected area electron diffrac-
under the initially formed MnCr 2 O4 supported by mostly inward oxygen tion (SAED) pattern. A high-angle annular dark-field (HAADF) detector
transport. The deeper Mn and Cr depletion depths that were measured and an EDS system (X-Max N 100TLE detector, Oxford Instruments,
in the underlying alloy in dry air as compared to wet air, as well as the UK) were employed at 300 kV for bright field (BF), high resolution
presence of MnCr 2 O4 at a few locations underneath the Cr 2 O3 scale in (HR)TEM, HAADF-scanning TEM (STEM) and composition analysis by
wet air [10], seemed to support the proposed mechanism. Furthermore, EDS. The EDS detector exploits a 100 mm2 sensor, a windowless config-
Si additions were observed to result in the formation a SiO2 underlayer uration, and an innovative mechanical design that delivers ultra-high
at the Cr 2 O3 –alloy interface which might have additionally contributed solid angle. The detector delivers a high sensitivity for elements with
to the slower oxidation kinetics compared to Ni–Cr specimens [10]. low energy X-rays, including impurities and dopants at the nanoscale.
EDS elemental mapping was quantitatively analyzed using QuantMap
Present study of the AZtecTEM software. QuantMap not only can bring the benefits of
automatic qualitative analysis into two dimensions to identify elements
In the present study, to confirm the proposed mechanism in [10], and visualise their distributions, but also can recalculate mapping data
additional cyclic exposures of NCMS model alloys were performed to correct for X-ray background, peak overlaps, sample matrix effects
at 950 ◦ C for 80, 250, and 500 h. In addition, detailed TEM and and pile up artefacts and displays real quantitative chemical X-ray maps
APT characterizations of the specimens exposed for 500 h [10] were For atom probe tomography investigation (APT), specimens from
performed. The specific aims of the preset study were: various locations in the oxide scale were prepared with the standard
focus ion beam (FIB)-based lift out process [14] using FEI Helios
• Evaluate the temporal evolution of the oxidation behavior and
Nanolab DualBeam FIB-scanning electron beam (FIB-SEM). The APT
accompanying compositional changes in the alloy
measurements were carried out using a CAMECA local electrode atom
• Identify differences in segregation behaviors of Cr, Mn and Si in
probe (LEAP) 4000X HR system equipped with a 355 nm wavelength
the external oxide layers between dry and wet air exposures
ultraviolet laser. The data collection used the following user-selected
• Correlate the findings of nanoscale characterizations in the oxide
parameters: 125 kHz pulse frequency, 60 pJ/pulse laser energy, 45 K
scales with subsurface compositional changes of the alloy
specimen base temperature with 0.005 detected ions/pulse. Analyses
were performed using the Integrated Visualization and Analysis Soft-
2. Experimental procedures
ware (IVAS) version 3.8.8. A careful determination of reconstruction
parameters was performed with the standard tip profile approach using
Rectangular (20 × 10 × 1.5 mm) coupons of model alloys NCMS
SEM micrography of the specimens taken during their preparation in
(measured compositions in Table 1) were investigated. Before exposure,
the FIB-SEM system [15]. The following derivates CrO𝑥 H𝑛+ 𝑛+
𝑦 , SiO𝑥 H𝑦
the specimens were ground to a 600-grit finish and ultrasonically
and O𝑥 H𝑛+𝑦 , NiO 𝑥 H𝑛+ were considered to identify Cr, Si, Ni and O within
𝑦
cleaned in acetone and methanol. Seven samples (total surface area
the Cr 2 O3 grains. In addition, Cr𝑛+ , Si𝑛+ , Mn𝑛+ , Ni𝑛+ and O𝑛+ derivates
for 7 specimens 34.3 cm2 ) were exposed in automated cyclic rigs [13] 2
were used for Cr, Si, Mn, Ni and O to reveal the segregation of elements
in 1 h cycles in dry air and air + 10% H2 O (wet air) flowing at
at the Cr 2 O3 grain boundaries.
500 cm3 min−1 (velocities of 0.60–0.68 cm s−1 ) between 80 and 500 h.
To deconvolute the effects of Mn, Si and water vapor, additional Oxide thicknesses and grain sizes for Cr 2 O3 and spinel oxides were
exposures were conducted for Ni-22Cr-2Mn (NCM) and Ni-22Cr-2Si measured using the BSE images, EDS elemental distribution maps
(NCS) specimens (measured compositions in Table 1) at 950 ◦ C in dry and TEM bright field images and by integrating a customized Python
and wet air for 80 h. The specimens were heated to the test temperature code [16] with the image analysis software ImageJ [17].
in the selected test atmosphere and kept at that temperature for 1 h
and cooled to room temperature for 10 min. The mass changes of the 3. Results
specimens were measured using a Mettler-Toledo model XP208 balance
with an accuracy of ±0.04 mg. After oxidation tests, cross sections 3.1. Mass change, oxide morphology and depth profiling
were prepared, Cu-plated and mounted in conductive epoxy mounts
for metallographic analyses. The mounted samples were ground to a Fig. 1 shows the averaged mass change data (from a total of three to
1200 grit finish using SiC grinding papers and subsequently polished four specimens) and corresponding standard error of the mean values
with diamond pastes to 1 μm surface finish. The final polishing step acquired during 1 h cyclic oxidation at 950 ◦ C in dry air and air + 10%
was made using colloidal SiO2 slurry. H2 O (referred to as wet air hereafter). Variations between mass changes
The microstructure of the oxidized samples was analyzed using were measured between experiments performed in different cyclic rigs
scanning electron microscopy (SEM), energy-dispersive spectroscopy and between specimens in dry air in the mass loss regime (mass change

2
M. Romedenne et al. Corrosion Science 221 (2023) 111348

Fig. 1. Mass change per surface area as function of time for the NCMS alloy in 1 h cycle dry air and air + 10% H2 O at 950 ◦ C.

Fig. 2. BSE images of the NCMS alloy after (a, c) 80 and (b, d) 500 h in (a, b) 1 h cycle dry air and (c, d) air + 10% H2 O at 950 ◦ C.

3
M. Romedenne et al. Corrosion Science 221 (2023) 111348

Fig. 3. Thicknesses of the MnCr 2 O4 and Cr 2 O3 oxide scales of the NCMS alloy determined from EDS image analysis after exposure in 1 h cycle dry air and air + 10% H2 O at
950 ◦ C.

lower than 0.95 mg cm−2 after 200 h in Fig. 1). Lower oxidation (mass in an apparent uneven interface for the selected TEM specimen in
gain) rates and mass losses in wet air than in dry air are seen in Fig. 1, Fig. 5.
in agreement with previous observations [10]. TEM bright-field (BF) images are shown in Fig. 7(a) and (b). In dry
Fig. 2 shows SEM images of metallographic cross sections of the air, a columnar grain microstructure was observed below the MnCr 2 O4
NCMS after 80 and 500 h (1 h cycles) in dry and wet air at 950 ◦ C. Cross layer and equiaxed grains closer to the alloy-oxide interface in Fig. 7(a).
sections reveal duplex oxide scales consisting of an outer Mn- and Cr- In wet air, only equiaxed grains were formed and were slightly finer
rich oxide and inner Cr 2 O3 layer (phases identified using EDS elemental than in dry air Fig. 7(b). White regions along the oxide-alloy interface
maps and EBSD as reported in [10]). The average thicknesses of the were identified as amorphous SiO2 in both atmospheres as reported in
different oxide scales derived from EDS image analysis are presented Figs. 5 and 6. An increase in Cr 2 O3 grain size from the alloy-oxide
in Fig. 3. It can be seen that the Cr 2 O3 thickness increased between 80 towards the gas/oxide interface was measured, with a slightly larger
and 500 h, as expected from the mass change data. Furthermore, the increase in dry air than in wet air, Fig. 7(c). The median grain sizes
thickness of the Cr 2 O3 scale was smaller in wet air compared to dry were 420 ± 222 μm and 301 ± 125 μm in dry and wet air respectively.
air while the thickness of the MnCr 2 O4 did not increase significantly In addition, on top of the multi-layered Cr 2 O3 scale, a single-layered
(Fig. 3). Fig. 4 displays EDS elemental concentration profiles measured MnCr 2 O4 scale was observed and measured to be about 600 and 900 nm
within the alloy after oxidation starting from the alloy/oxide interface. thick in wet and dry air respectively, Fig. 7. EDX line scans were
The depletion depths were defined from the alloy-oxide interface until collected across grain boundaries in the Cr 2 O3 scale (yellow lines in
the measured concentrations were within ±5% of their initial values in Fig. 7(a) and (b)). In dry air, all line scans measured 70 ± 4 at.%,
the alloy (before high temperature exposures). In dry air, both Cr and 30 ± 2 at.%, 0.2 ± 0.2 at.% and 0.1 ± 0.1 at.% and for O, Cr, Mn
Mn depletion depths increased from 20 to 80 μm between 80 and 500 h and Ni respectively and no particular change was observed between
(Figs. 4(a) and 4(c)). However, in wet air, Cr depletion depth increased the Cr 2 O3 grains or grain boundaries. For Si however, most line scans
slightly from 20 to 40 μm (Fig. 4(b)), and no visible increase in the Mn measured 0.2 ± 0.2 at.% except for line scan number 8 which measured
depletion depth was observed between 80 and 500 h (Fig. 4(d)). These Si enrichment across a grain boundary of about 10 at.% (Fig. 8(a)). In
observations suggest lower oxidation-induced consumption of Cr and wet air, 4 line scans were collected (Fig. 8(b)). Line scan number 1 was
Mn in wet air than in dry air, in agreement with previous results [10]. collected over a SiO2 and measured about 25 ± 5 at.% of Si and 70 ± 4
at.% of O. However, the line scans 2, 3 and 4 measured within the
3.2. Oxide scale analysis by TEM Cr 2 O3 across grain boundaries revealed local Si enrichment up to 20
at.% while the O and Cr concentrations were 70 ± 4 at.% and 30 ± 2
Figs. 5 and 6 show high angle annular dark-field (HAADF) images at.% (Fig. 8(b)).
labeled with the location of the transmission diffraction analyses, cor-
responding quantified EDS mappings for O, Cr, Mn, Si and Ni elements 3.3. Oxide scale analysis by APT
and selected area electron diffraction patterns (SAED) for MnCr 2 O4 ,
Cr 2 O3 and SiO2 phases for specimens exposed for 500 h in dry and wet For gaining insights into the local composition of the oxide scale,
air at 950 ◦ C. The EDS mappings and transmission diffraction analyses specimens oxidized at 950 ◦ C for 500 1 h cycle in dry and wet air were
confirm the duplex microstructure of the surface oxide scale, with an selected for APT investigation. Fig. 9 shows a schematic cross section
outer layer as MnCr 2 O4 spinel and inner layer as Cr 2 O3 . Beneath the of the characterized specimens indicating the APT-analyzed regions,
Cr 2 O3 scale, discontinuous and amorphous SiO2 precipitates were also marked with numbers 1 to 3 in dry air and letters A to C in wet air.
identified (Fig. 6). As seen from the BSE images in Fig. 2, the Cr 2 O3 - Three specimens were successfully analyzed by APT in both dry and
alloy interface was convoluted in both dry and wet air and resulted wet air conditions. Among the three APT samples prepared from the dry

4
M. Romedenne et al. Corrosion Science 221 (2023) 111348

Fig. 4. EDS (a, b) Cr and (c, d) Mn elemental depletion below the oxide scale in the NCMS alloy after 80, 250 and 500 h in (a, c) 1 h cycle dry air and (b, d) air + 10% H2 O
at 950 ◦ C.

atmosphere, two were probed directly from the SiO2 /Cr 2 O3 interface, at.% Mn. Concentration errors bars were automatically calculated using
and one sample was prepared within the Cr 2 O3 scale (Fig. 9(a)). In method from [18]. In Fig. 11(a), the oxygen concentration remained
wet air, two APT samples were prepared from the Cr 2 O3 while one relatively constant along the profile (60 at.%) while increasing Cr and
was successfully made from the alloy/oxide interface (Fig. 9(b)). In wet decreasing Si and Mn concentrations were measured within the Cr 2 O3
air, many specimens were not mechanically stable under high voltage scale. In the SiO2 enriched region, up to 20 at.% Si, 30 at.% Cr and
and developed multiple fractures during APT analyses and could not be 1 at.% Mn were measured. In the Cr 2 O3 scale the average Mn and Si
analyzed. concentrations were about 0.1 and 1 at.% respectively. Si-rich clusters
(C1 and C2 in Fig. 11(b)) were identified within the Cr 2 O3 scale with Si
concentration up to 33 at.% and 60 at.% O showing SiO2 stoichiometry
3.3.1. Alloy/oxide and SiO2 /oxide interfaces
with minor Cr impurities.
Fig. 10 displays the reconstructions of the three dimensional atomic
Grain boundaries with local Mn enrichments were also observed
maps from the APT analyses performed at the SiO2 /Cr 2 O3 interface in
within the Cr 2 O3 scale in dry air (four grain boundaries GB1a, b and
dry air (regions 1 and 2 in Fig. 9(a)) and at the alloy/Cr 2 O3 interface in GB2a, b from regions 1 and 2 in Fig. 12). On average, 0.8 at.% Mn was
wet air (region A in Fig. 9(b)). For each reconstruction, the positions of measured at the grain boundaries while about 0.1 at.% was measured
the relevant metal and metal oxide ions are displayed. Figs. 10(a) and within the Cr 2 O3 grains (Fig. 12) in agreement with measurements
10(b), show the presence of Si𝑛+ , Mn𝑛+ and O𝑛+ 2
at the Cr 2 O3 grain in Fig. 11(b). For grain boundary GB2b, much larger Mn enrichment
boundaries in dry air. No oxide grain boundary or SiO2 precipitate was measured (up to 5 at.%) compared to other grain boundaries
could be isolated in the wet air specimen, as seen with the presence of (Fig. 12). For Si, grain boundaries enrichment was less noticeable with
Ni derivates (Ni𝑛+ , NiO𝑥 H𝑛+
𝑦 ) from the underlying alloy in Fig. 10(c). a measured average of 1.6 at.% Si, except for the GB2b where almost
In dry air, a composition profile from the SiO2 /Cr 2 O3 interface 5 at.% Si was measured. Si enrichment of 1 at.% was measured within
towards the surface was measured and displayed in Fig. 11(a) from the Cr 2 O3 grains. The O and Cr concentrations did not vary between
region 2 in Fig. 9(a). The reconstruction displays Si (blue) and Mn the four characterized grain boundaries and were measured to be
(purple) iso-volumes containing at least 4 at.% of Si and at least 0.8 56 ± 1 at.% and 40 ± 1 at.% respectively.

5
M. Romedenne et al. Corrosion Science 221 (2023) 111348

Fig. 5. TEM HAADF micrographs labeled with selected area electron diffraction (SAED) locations and EDS maps showing distribution of elements in the oxide scale of NCMS
specimens oxidized for 500 h in (a) 1 h cycle dry air and (b) 1 h air + 10% H2 O at 950 ◦ C.

3.3.2. Chromia scale boundary found in the dry specimen and show a local increase of Mn
Figs. 13 and 14 show APT atom maps of the Cr 2 O3 and composition in the grain boundary up to 1 at.% (Fig. 14(a)). Differences can be
profile built across the grain boundaries from region 3 in dry air observed, however, between the dry and wet conditions. In wet air, the
and regions B and C in wet air in Fig. 9. In Fig. 13, grain boundary local increase of Mn composition across the grain boundary is much
segregation of Mn can be neatly observed from Mn𝑛+ and O𝑛+ maps in lower than in dry air with only 0.2 at.%. In wet air, Mn enrichment
Fig. 13 in dry air. Composition profiles are measured across the grain was accompanied with a local Si enrichment visible from the atom

6
M. Romedenne et al. Corrosion Science 221 (2023) 111348

Fig. 6. Transmission diffraction performed at different location on the NCMS specimens oxidized for 500 h in (a) 1 h cycle dry air and (b) 1 h air + 10% H2 O at 950 ◦ C.

maps in Figs. 13(b) and 13(c) and Si enrichment detected in the grain air than in dry air (Fig. 15) along with the formation of Mn clusters
boundary in Fig. 14(b). However in dry air, no particular Si enrichment in wet air (Fig. 14(b)). Mn segregation at the Cr 2 O3 grain boundaries
was measured in the Cr 2 O3 grain boundaries (Fig. 14(a)), where the has been observed and reported in the literature and is a characteristic
Si concentration was about 1 at.% both in the Cr 2 O3 grains and grain outcome of the diffusion of Mn through the Cr 2 O3 grain boundaries
boundaries. Interestingly, Mn clusters were only identified in wet air in in order to form the observed topmost MnCr 2 O4 scale (Figs. 5(a) and
the Cr 2 O3 scale in both regions B and C. Mn clusters, were identified 5(b)) [8,9]. However, the presence of Mn dispersed as nanoparticles in
using Mn iso-volumes of concentration equal and above 0.5 at.% and the Cr 2 O3 grains is surely intriguing and has never been reported before
gathered as much as 4 at.% in certain cases as reported in Fig. 14(b) for Ni-base alloys. Galiullin [19] observed a similar clustering of Mn in
from region B in Fig. 9(b). the oxide scale formed on Co-based alloys Haynes 188 and 25 in the
finer grained regions of the Cr 2 O3 scale after exposure in dry air at
1000 ◦ C for 300 h. The question of interest concerns the mechanism of
4. Discussion
the formation of these clusters in the present study, observed solely in
wet air, and their impact on alloy oxidation rate and oxidation-induced
4.1. Impact of water vapor on segregation behavior of elements
elemental depletion in the underlying alloy. The solubility of Mn in
Cr 2 O3 is about 1.6 at.% 1000 ◦ C in air [20] but about 4 at.% of Mn
4.1.1. Segregation behavior of Mn was measured to be clustered in the Cr 2 O3 grains after exposures in
Average values of the calculated Gibbsian interfacial excess (G𝑖 ) for wet air (Fig. 14(b)).
Mn at the Cr 2 O3 grain boundaries are reported in Fig. 15 for a total
of five grain boundaries in dry air and two grain boundaries in wet 4.1.2. Segregation behavior of Si
air. Mn segregation was identified in all the studied grain boundaries The presence of 0.5 wt% Si in the NCMS specimens led to the
and analyses revealed substantially lower segregation activity in wet formation of a discontinuous layer of amorphous SiO2 at the alloy-oxide

7
M. Romedenne et al. Corrosion Science 221 (2023) 111348

Fig. 6. (continued).

interface in both dry and wet air (Figs. 5 and 6). APT characterizations in the scale grown in wet air was not located near a SiO2 precipitate
at the SiO2 -Cr 2 O3 interface demonstrated the diffusion of Si from the due to the non-uniform distribution of SiO2 particles (Fig. 9). Therefore,
amorphous SiO2 through the Cr 2 O3 scale (presence of Si diffusion Si diffusion through the Cr 2 O3 scale and the formation of SiO2 clusters
profile in Fig. 11(a)) and the formation of Si-rich clusters (Figs. 10(a) could not be confirmed (location A in Fig. 10(c)) with APT. However,
and 10(b)) of stoichiometry closely matching that of SiO2 (Fig. 11(b)) up to 0.4 at.% Si and 58 ± 1 at.% O and 40 ± 1 at.% Cr were
in dry air. In the Cr 2 O3 scale probed region (locations B and C in measured in the Cr 2 O3 scale formed in wet air (Fig. 14(b)). Calculations
Figs. 13(b) and 13(c)), grain boundary segregation of Si was also of the Gibbsian interfacial excess of Si (number of segregating Si atoms
measured (between 0.2 and 1.6 at.% Si in Fig. 14(a)) in dry air. per grain boundary unit area) [22] did not reveal any particular Si
Similar findings have been reported in the literature where amorphous enrichment in wet air compared to dry air (Fig. 15). However, the
SiO2 particles embedded in the Cr 2 O3 scale along with the segregation very small size of the tip analyzed in the atom probe could lead to
of Si at the Cr 2 O3 grain boundaries was observed for Fe-20Cr-0.5Si non-representative results compared to TEM characterizations.
after 240 h in Ar-20CO2 and Ar-20O2 -20H2 O at 650 and 818 ◦ C [7, Nguyen and co-authors [7] revealed a greater rate of SiO2 incor-
21] and in the Cr 2 O3 scale formed on various alloys based on the poration in the form of glassy SiO2 nanoparticles within the Cr 2 O3
Crofer 22 H composition with Nb, Ti and Si additions after oxidation scale formed on the Fe-20Cr-0.5Si (wt.%) after 240 h at 818 ◦ C in
in Ar-4%H2 -4%H2 O [8,9]. Ar-20O2 -20H2 O compared to Ar-20CO2 . They combined EDX TEM line
TEM EDX linescans across the Cr 2 O3 grain boundaries and cor- scans through the Cr 2 O3 and estimated values of SiO2 volume fractions
responding diffraction patterns indicated the presence of amorphous to suggest that volatilization of SiO2 in the water vapor environment
(CrSi)O𝑥 in the Cr 2 O3 scales formed on NCMS in both dry and wet was responsible for the qualitative higher Si enrichment in the Cr 2 O3
air. However, a significantly higher fraction of grain boundaries were scale. Si enrichment (0.2 at.%) was measured in the Cr 2 O3 scale formed
enriched with these oxides for scales formed in the presence of water on Fe-30Cr-0.002Si (wt.%) after 240 h in Ar-10%H2 O at 650 ◦ C and
vapor than in dry air (Fig. 8). Unfortunately, the APT probe extracted was correlated to the transport of Si impurities (0.167 mg/L) from the

8
M. Romedenne et al. Corrosion Science 221 (2023) 111348

Fig. 7. TEM bright-field (BF) images and locations of EDX line scans (yellow) and SAED (red) of cross-sectioned NCMS specimens oxidized for 500 h in (a) 1 h cycle dry air and
(b) 1 h air + 10% H2 O at 950 ◦ C and (c) Average Cr 2 O3 grain size from the alloy-oxide interface towards the surface in both environment. The blue lines mark the alloy/Cr 2 O3
and Cr 2 O3 /MnCr 2 O4 interfaces. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

9
M. Romedenne et al. Corrosion Science 221 (2023) 111348

Fig. 7. (continued).

Fig. 8. Measured Si EDX profiles across grain boundaries in Cr 2 O3 of cross-sectioned NCMS specimens oxidized for 500 h in (a) 1 h cycle dry air and (b) 1 h air + 10% H2 O at
950 ◦ C.

deionized water source to the water vapor-containing gases [23]. The molar mass of Cr 2 O3 .
deionized water used in the present study contained 0.3 mg/L of Si Cr O3 𝑁Si Cr 2 O3
(measured by inductively coupled plasma atomic emission spectroscopy 𝑛Si 2 = 𝑛 (1)
𝑁Cr Cr
ICP-AES) while the NCMS alloys contained 0.5 wt% of Si (Table 1). 𝛿Cr O 𝜌Cr O
Cr O3
The liquid water volume consumed during the 500 h reaction in 𝑛Cr 2 =2 2 3 2 3 (2)
𝑀Cr 2 O3
air + 10%H2 O flowing at 500 cm3 min−1 was 𝑉H2 O = 1.35 L and
N𝑆𝑖 and N𝐶𝑟 are the average concentrations (at.%) of Si and Cr in the
therefore a total amount of Si of 0.41 mg or 4.3 10−7 mol cm−2 (Si𝑐𝑎𝑙𝑐
Cr 2 O3 scale. The experimental amount of Si was between 0.2 and 2 at.%
from air + 10% H2 O) was available to be incorporated into the 7
from the APT measurements in Figs. 12 and 14 and 0.2 ± 0.2 at.% from
specimens present in the cyclic rig (total surface area of 4.9 cm2 × 7 the TEM measurements. Using Eqs. (1) and (2) the total amount of Si in
(specimens) = 34.3 cm2 ) (Fig. 16). The total amount of Si (n𝑆𝑖 Cr 2 O3 in the Cr 2 O3 scale (Si𝑒𝑥𝑝 ) was between 2.1 × 10−7 and 3.0 × 10−6 mol cm−2
mol cm−2 ) in the Cr 2 O3 scale was calculated using Eqs. (1) and (2), with and reported in Fig. 16. Clearly, the Si impurity in the water cannot
𝛿Cr 2 O3 the Cr 2 O3 thickness (8.7 and 6.2 μm respectively after 500 h in account for the measured Si concentrations in the Cr 2 O3 scale (Fig. 16).
dry and wet air in Fig. 3), 𝜌Cr 2 O3 equal to 5.22 g cm−3 and MCr 2 O3 the Indeed, with an initial alloy content of Si of 0.5 wt% in the alloy, the

10
M. Romedenne et al. Corrosion Science 221 (2023) 111348

(air + 10% H2 O) decreased the initial mass gains (Fig. 1), and mass
losses (spallation behavior) (Fig. 1) which were further confirmed by
less depletion of Cr and Mn from the alloy (Fig. 4(d)) compared to dry
air (Figs. 1 and 4(c)). The better adherence of Cr 2 O3 scales grown in the
presence of water vapor has been attributed in the literature to suppres-
sion of void nucleation and scale detachment due to oxide formation at
the scale–alloy interface supported by mostly inward oxygen transport
in wet air [11,30]. However, the decreased scaling rates in the present
study are in contradiction to the established understanding that water
vapor accelerates scaling rates of Cr 2 O3 scales by resulting in a finer
grain size and enhancing inward grain boundary transport [11,26].
The exposures in water vapor clearly resulted in a finer grained Cr 2 O3
scale (Fig. 7(c)) but did not increase the Cr 2 O3 growth rate. This can
most likely be attributed to phenomena that counter-acted the grain
size reduction, likely including the distinct segregation behaviors of Mn
and Si observed between the Cr 2 O3 scales formed in dry and wet air and
its impact on altering the grain boundary transport properties.
Zurek et al. showed that Mn additions to Ni-25Cr led to lower
oxidation rates in Ar-4%H2 -7%H2 O compared to Ar+20% O2 [11] at
1050 ◦ C. However, this effect was mainly attributed to the impact of
different partial pressures of oxygen on the solubility of Mn in Cr 2 O3
suggesting the reduction in Cr 2 O3 scale growth rate in the low-pO2
Ar-4%H2 -7%H2 O gas. The authors [11] speculated that in the low-pO2
gas, Mn hinders inward oxygen transport by segregating at oxide grain
Fig. 9. Schematics of cross-sectioned NCMS specimens oxidized for 500 h in (a) 1 h boundaries. In the present study, the presence of 10% H2 O does not
cycle dry air and (b) 1 h air + 10% H2 O at 950 ◦ C. (For interpretation of the references
significantly impact pO2 at the gas/oxide interface which is similar in
to color in this figure legend, the reader is referred to the web version of this article.)
dry and wet air. Hence, the findings reported by Zurek et al. [11]
cannot explain the differences in the oxidation rates and segregation of
amount of Si that can be transported from a 1.5 mm thick specimen Mn in the Cr 2 O3 scale observed in the present study. Perez et al. [31]
to the specimen’s surface at which the alloy Si concentration is set at showed with photoelectrochemical measurements that the Cr 2 O3 scale
zero can be calculated with the solution proposed by Crank [24] (for beneath a MnCr-spinel formed on Mn-containing alloys demonstrates
the case of a plate with initial uniform concentration and fixed surface
an n-type contribution. This supports the fact that a cation of lower
concentration) and using the diffusion coefficient of Si in gamma FCC
valence state, in this case Mn2+ , is expected to increase the oxidation
from [25] to be DSi = 1.3 × 10−11 cm2 s−1 . This calculation results in a
rate by increasing the concentration of oxygen vacancies, i.e. Cr and
value of 4 × 10−6 mol cm−2 after 500 h at 950 ◦ C for the amount of Si
Mn interstitials [5]. Indeed, they observed faster scaling rates for the
from the alloy (Si𝑐𝑎𝑙𝑐 from alloy) available to be incorporated into the
Ni-25Cr-xMn (x = 0.5,1.0,1.5 wt,%) compared to the binary Ni-25Cr
Cr 2 O3 scale (Fig. 16).
alloy. Vayyala et al. [9] elucidated that substitution of host Cr3+ ions
at grain boundaries with Mn2+ will induce the formation of oxygen
4.2. Effect of water vapor and segregation of Mn and Si on Cr 2 O3 growth
vacancies promoting the inward oxygen transport and counteract the
and microstructure
impact of additions of higher valence Nb4+∕5+ and Ti4+ ions which
would typically generate cation vacancies.
NCMS specimens formed duplex oxide scales consisting of outer
The question of interest here concerns the reasons why even lower
MnCr 2 O4 spinel and inner Cr 2 O3 in both dry and wet air. The Cr 2 O3
oxide layer was made of equiaxed and elongated Cr 2 O3 grains in dry oxidation rates and therefore also lower Mn depletion depths were
air, while smaller and equiaxed grains were formed in wet air (Fig. 7). observed for the studied NCMS alloys in wet air. Although, the nature
Similar impact of water vapor on the Cr 2 O3 growth and microstructure of the defects could not be ascertained, it is most likely that inward
was observed [11] where columnar grains were formed on the surface transport of oxygen-bearing species is governing growth of the Cr 2 O3
of Ni-25Cr-Mn after 1000 h in Ar+20% O2 at 1000 ◦ C while smaller scale while Mn transport along the Cr 2 O3 grain boundaries supports
and equiaxed grains were formed in Ar-4%H2 -7%H2 O [11]. The Cr 2 O3 the formation of the MnCr 2 O4 at the gas/oxide interface in wet air.
scale formed on Fe-20Cr and Fe-20Cr-20Ni presented relatively coarse, The continued depletion of Mn from the alloy (up to 500 h) in dry
columnar grains after 1000 h in Ar-20CO2 at 650 ◦ C while equiaxed air exposures supports this hypothesis. The observed lower segregation
Cr 2 O3 grains were formed in Ar-20CO2 -20H2 O [21]. Additions of water activity of Mn (Fig. 15) in wet air suggests that water vapor reduced
vapor during oxidation usually provides accelerated mass transport the amount of available free sites at grain boundaries for Mn2+ ions to
within voids whilst simultaneously promoting the formation of an in- diffuse outward, in agreement with the proposed mechanism in [10].
ward growing fine-grained Cr 2 O3 oxide scale [11,26]. The finer grained This consequently reduced the generation of oxygen vacancies and
Cr 2 O3 scale provides more diffusion paths in wet air than dry air which contributed to the observed retardation of the overall growth kinetics.
will potentially enhance the transport of defects across the Cr 2 O3 scale. The limited outward diffusion of Mn in wet air also resulted in the
Enhanced Cr loss due to combined oxidation and volatilization of the formation of Mn-rich clusters in the Cr 2 O3 scale (Figs. 13 and 14(b)).
Cr 2 O3 scale formed on 304L stainless steel and 709 alloy was also The formation of SiO2 at the Cr 2 O3 /alloy interface has been known
shown in presence of water vapor in the atmosphere between 600 and to typically slow down the kinetics of Cr 2 O3 scale growth in Si con-
800 ◦ C [27,28]. In that study, the impact of Cr 2 O3 volatilization was taining Cr 2 O3 forming alloys by acting as a diffusion barrier [32,33].
minimal and corresponded to an increase in the Cr loss from 4.54 mg Underlying SiO2 particles were observed at the Cr 2 O3 –alloy interface
cm−2 in dry air to 4.57 mg cm−2 in air + 10% H2 O after 500 h at 950 ◦ C, in the NCMS specimens in both dry and wet air and are most likely
using the method reported in [12] and data by Gindorf et al. [29] and contributing to the reduction in oxidation kinetics of NCMS specimens.
oxidation rate constant, k𝑝 , equal to 8.8 10−3 mg2 cm−4 h−1 [10]. Furthermore, Si additions to a NiCr alloy have been reported to in-
The obtained results in the present study were counter to expected crease the diffusion coefficient of Cr which would have contributed to
trends and showed that the presence of water vapor in the atmosphere the increased flux of Cr to the oxide–alloy interface [10,34]. Nguyen

11
M. Romedenne et al. Corrosion Science 221 (2023) 111348

et al. [7] suggested that SiO2 particles incorporated in the Cr 2 O3 into H+ and OH− ions ruptures the Si-O-Si bridge [35]. Moulson and
decreased oxidation rates by blocking grain boundary transport of Cr3+ Roberts [35] explained that although the silicates are highly stable,
and O2− ions. The presence of amorphous SiO2 at the Cr 2 O3 grain there is a continuous range of inter-ionic bond strengths which most
boundaries could be confirmed in the present study and possibly had probably ruptures the weaker bonds.
a similar effect by additionally slowing Mn2+ ions outward diffusion Cr 2 O3 and SiO2 have been known to be immiscible [36] which has
and reducing scaling rates in wet air. The considerably higher fraction been supported with thermodynamic calculations by Wang et al. [37]
of SiO2 in wet air than dry air can be explained with the modifying at 950 ◦ C. However, a minimal concentration of Si ions (0.1–0.5%)
effect of OH− ions in SiO2 which accelerates growth of SiO2 on silicon in in Cr 2 O3 was reported by Wood et al. [38]. In the current work, Si
water vapor containing environments [5,35]. Glassy SiO2 is considered concentrations of up to 1.0 at.% were measured in the Cr 2 O3 grain
a network former oxide [5] and the dissociation of the water molecule boundaries. Potential doping of the Cr 2 O3 with the higher valent Si4+

Fig. 10. APT maps of the (a, b) oxide at the SiO2 and Cr 2 O3 interface (regions 1 and 2 in Fig. 9(a)) formed on the NCMS specimen after 500 h in 1 h cycle dry air at 950 ◦ C
and (c) of the oxide at the alloy and Cr 2 O3 interface (region A in Fig. 9(b)) 1 h air + 10% H2 O at 950 ◦ C.

Fig. 11. (a) O, Cr, Mn and Si atomic concentrations along profile A and (b) along clusters C1 and C2 of NCMS specimen (region 2 in Fig. 9(a)) after 500 h in 1 h cycle dry air
at 950 ◦ C. Respective color codes and scale bar for the APT needle are reported in Fig. 10(b). (For interpretation of the references to color in this figure legend, the reader is
referred to the web version of this article.)

12
M. Romedenne et al. Corrosion Science 221 (2023) 111348

Fig. 12. APT composition profiles built across the grain boundaries GB1 and GB2 of the SiO2 and Cr 2 O3 interface (regions 1 and 2 in Fig. 9(a)) for Mn and Si segregation formed
on the NCMS specimen after 500 h in 1 h cycle dry air at 950 ◦ C.

ions will result in generation of Cr vacancies and O interstitials [39]. of SiO2 at the Cr 2 O3 grain boundaries which played a larger role in
Consequently, the effects of Mn and Si doping will tend to balance reducing inward oxygen transport and outward transport of Mn and Cr
each other, which might explain the findings of the study conducted ions. This combined with SiO2 at the Cr 2 O3 /alloy interface which acted
by Perez et al. [31], which concluded that simultaneous additions of as a diffusion barrier slowed the overall oxidation kinetics in wet air.
Mn (0.5 wt%) and Si (0.5, 1.0 and 1.5 wt%) had no influence on the Fig. 17 shows the measured Cr depletion profiles (average of 3 EDS
oxidation kinetics compared to the Ni-25Cr binary alloy after exposures linescans) for the Ni-22Cr-2Mn and Ni-22Cr-2Si (in wt.%) specimens
in dry air at 1050 ◦ C for 100 h. at 950 ◦ C in dry and wet air for 80 h. The increased Cr consumption
Based on the experimental observations in the present study, it can after exposures in dry air compared to wet air is evident for both model
be concluded that the exposures in water vapor resulted in formation alloys. Furthermore, the measurements suggest a stronger impact of Si

13
M. Romedenne et al. Corrosion Science 221 (2023) 111348

Fig. 13. APT maps of the (a) Cr 2 O3 (region 3 in Fig. 9(b)) formed on the NCMS specimen after 500 h in 1 h cycle dry air at 950 ◦ C and (b) of the Cr 2 O3 (regions B and C in
Fig. 9(b)) 1 h air + 10% H2 O at 950 ◦ C.

than Mn additions on reducing the Cr depletion in wet air. A detailed enrichment of Si at the Cr 2 O3 grain boundaries in wet air exposures.
characterization of the oxide scales on these specimens could not be A detailed APT analysis revealed the reduced segregation of Mn at
undertaken to confirm the formation of Mn-rich clusters in the Cr 2 O3 the Cr 2 O3 grain boundaries and formation of Mn-rich clusters within
scale formed in wet air and the higher enrichment of Si in the Cr 2 O3 the Cr 2 O3 scale after exposure in wet air compared to dry air which
scale formed in wet air than in dry air. However, the Cr profiles in indicated a substantially lower outward diffusion of Mn. This combined
Fig. 17 provide partial evidence that ternary additions of either Mn or with the enhanced formation of Si-rich oxide clusters in the Cr 2 O3 scale
Si can decrease oxidation and Cr depletion rates during exposures in reduced the outward diffusion of Cr (thereby depletion from the alloy)
water vapor. which could explain the observed lower oxidation rates in wet air and
decreased Mn consumption in the underlying NCMS alloy compared to
dry air.
5. Summary and conclusion
The findings of the current study demonstrate that water vapor
effects on alloy oxidation behavior can vary significantly, illustrating
The present study further confirmed the combined role and inter- the need to improve the mechanistic understanding of the impact of
actions between alloying additions Mn and Si and water vapor in the water vapor on the oxidation behavior of high temperature alloys.
oxidation behavior of Ni-22Cr model alloys containing 1.5 wt% Mn and This is critical to a successful development of materials for H2 -fueled
0.5 wt% Si oxidized in flowing dry air and air +10% H2 O at 950 ◦ C. technologies aiming to balance alloy chemistry (role of key alloying
The studied alloys showed the formation of an outer MnCr 2 O4 scale and constituents such as Mn, Si, Al, Ti) for desired mechanical properties
inner Cr 2 O3 scale with underlying amorphous SiO2 precipitates in both while simultaneously setting constraints on costs. Future research ef-
environments. TEM analyses showed significant formation of SiO2 and forts are focusing on integrating density-functional theory modeling

14
M. Romedenne et al. Corrosion Science 221 (2023) 111348

Fig. 14. APT composition profiles built across the grain boundaries GB1 and GB2 and clusters (C1, C2 and C3) in the Cr 2 O3 (regions 3 and B in Figs. 9(a) and 9(b)) for Mn and
Si segregation in the NCMS specimen after 500 h in (a) 1 h cycle dry air and (b) 1 h air + 10% H2 O at 950 ◦ C. Respective color codes and scale bar for the APT needle are
reported in Figs. 13(b) and 13(c). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 15. Average and standard deviation of the Gibbsian interfacial excess values (G𝑖 ) Fig. 16. Minimum and maximum Si measured in the Cr 2 O3 scale from APT analyses
for the grain boundary segregation of Si and Mn in each of the identified Cr 2 O3 scale in dry and wet air (Siexp ) and calculated Si available from the NCMS alloy (Sicalc -alloy)
grain boundaries in Fig. 9 after 500 h in 1 h cycle dry air and air + 10% H2 O at and deionized water (Sicalc -Air + 10% H2 O) after 500 h in 1 h cycle dry air and
950 ◦ C. air + 10% H2 O at 950 ◦ C.

15
M. Romedenne et al. Corrosion Science 221 (2023) 111348

Fig. 17. Cr EDS line scans (average of three lines) measured below the oxide scale formed on the surface (a) Ni-22Cr-2Mn and (b) Ni-22Cr-2Si after 80 h in 1 h cycle dry air
and air + 10% H2 O at 950 ◦ C.

methods with the ultimate goal of describing the interactions between References
atmosphere and alloying constituents.
[1] M. Noussan, P.P. Raimondi, R. Scita, M. Hafner, The role of green and blue
hydrogen in the energy transition-A technological and geopolitical perspective,
Sustainability 13 (1) (2021) 1–26, doi: ARTN 298 10.3390/su13010298.
CRediT authorship contribution statement
[2] M.H. du Toit, A.V. Avdeenkov, D. Bessarabov, Reviewing 𝐻2 combustion: A
case study for non-fuel-cell power systems and safety in passive autocatalytic
recombiners, Energy Fuels (ISSN: 0887-0624) 32 (6) (2018) 6401–6422, http:
M. Romedenne: Conceptualization, Writing – original draft, Writ-
//dx.doi.org/10.1021/acs.energyfuels.8b00724.
ing – review & editing, Data curation, Oxidation experiments, Vi- [3] F.H. Stott, Influence of alloy additions on oxidation, Mater. Sci. Technol. 5
sualization, Investigation, Analysis, Interpretation. S. Lambeets: APT (8) (1989) 734–740, http://dx.doi.org/10.1179/mst.1989.5.8.734, arXiv:https:
characterization, Analysis, Review and editing. M. Song: TEM charac- //doi.org/10.1179/mst.1989.5.8.734.
[4] C.S. Giggins, F.S. Pettit, Oxidation of Ni-Cr alloys between 800 degrees and 1200
terization. C. Roach: APT preparation. A. Devaraj: Funding, Charac- degrees c, Trans. Metall Soc. Aime (ISSN: 0543-5722) 245 (12) (1969) 2495, URL
terization, Supervision, Review and editing. R. Pillai: Funding, Con- https://link.springer.com/article/10.1007/BF02811822.
ceptualization, Oxidation experiments, Supervision, Writing – review [5] P. Kofstad, High Temperature Corrosion, Elsevier Applied Science, 1988, URL
https://books.google.com/books?id=dsdlQgAACAAJ.
& editing, Analysis, Interpretation. [6] R.E. Lobnig, H.P. Schmidt, K. Hennesen, H.J. Grabke, Diffusion of cations in
chromia layers grown on iron-base alloys, Oxid. Met. 37 (1) (1992) 81–93,
http://dx.doi.org/10.1007/BF00665632.
Declaration of competing interest [7] T.D. Nguyen, A. La Fontaine, L. Yang, J.M. Cairney, J. Zhang, D.J. Young,
Atom probe study of impurity segregation at grain boundaries in chromia scales
grown in CO2 gas, Corros. Sci. 132 (2018) 125–135, http://dx.doi.org/10.
The authors declare that they have no known competing finan- 1016/j.corsci.2017.12.024, URL https://www.sciencedirect.com/science/article/
cial interests or personal relationships that could have appeared to pii/S0010938X17311423.
[8] A. Vayyala, I. Povstugar, T. Galiullin, D. Naumenko, W.J. Quadakkers, H.
influence the work reported in this paper. Hattendorf, J. Mayer, Effect of Nb addition on oxidation mechanisms of high Cr
ferritic steel in Ar–H2–H2O, Oxid. Met. 92 (2019) 471–491, http://dx.doi.org/
10.1007/s11085-019-09933-x, URL https://link.springer.com/article/10.1007%
Data availability 2Fs11085-019-09933-x.
[9] A. Vayyala, I. Povstugar, D. Naumenko, W.J. Quadakkers, H. Hattendorf, J.
Mayer, A nanoscale study of thermally grown chromia on high-Cr ferritic steels
Data will be made available on request. and associated oxidation mechanisms, J. Electrochem. Soc. 167 (6) (2020)
061502, http://dx.doi.org/10.1149/1945-7111/ab7d2e.
[10] R. Pillai, M. Romedenne, J.A. Haynes, B.A. Pint, Oxidation behavior of candidate
Acknowledgments NiCr alloys for engine exhaust valves: Part I—Effect of minor alloying elements,
Oxid. Met. 95 (2021) 157–187, http://dx.doi.org/10.1007/s11085-020-10017-4.
[11] J. Zurek, D. Young, E. Essuman, M. Hänsel, H. J. Penkalla, L. Niewolak, W.
J. Quadakkers, Growth and adherence of chromia based surface scales on Ni-
G. Garner and J. Wade are thanked for the oxidation exposures.
base alloys in high- and low-p𝑂2 gases, Mater. Sci. Eng. A 477 (2008) 259–270,
T. Lowe and D. Newberry are acknowledged for the metallographic http://dx.doi.org/10.1016/j.msea.2007.05.035.
preparations. A. Haynes and J. Poplawsky are thanked for their valu- [12] M. Romedenne, R. Pillai, S. Dryepondt, B.A. Pint, Effect of water vapor on
able comments on the manuscript. FIB based APT and TEM sample lifetime of 625 and 120 foils during oxidation between 650 and 800 ◦ C, Oxid.
Met. 96 (2021) 589–612, http://dx.doi.org/10.1007/s11085-021-10069-0.
preparation and TEM, APT analysis was conducted using facilities at the [13] B.A. Pint, P.F. Tortorelli, I.G. Wright, Effect of cycle frequency on high-
Environmental Molecular Sciences Laboratory (EMSL), a DOE national temperature oxidation behaviour of alumina- and chromia-forming alloys, Cycl.
user facility funded by the Biological and Environmental Research Oxid. High Temp. Mater. (27) (1999) 111–132, URL https://link.springer.com/
article/10.1023/A:1016064524521.
Program located at Pacific Northwest National Laboratory, USA. PNNL
[14] K. Thompson, D. Lawrence, D.J. Larson, J.D. Olson, T.F. Kelly, B. Gorman, In situ
is operated by Battelle for DOE under contract DE-AC05-76RL01830. J. site-specific specimen preparation for atom probe tomography, Ultramicroscopy
Haynes and J. Poplawsky are thanked for their valuable comments on 107 (2–3) (2007) 131–139, http://dx.doi.org/10.1016/j.ultramic.2006.06.008.
[15] A. Devaraj, D.E. Perea, J. Liu, L.M. Gordon, T.J. Prosa, P. Parikh, D.R. Diercks,
the manuscript. This research was sponsored by the U.S. Department
S. Meher, R.P. Kolli, Y.S. Meng, S. Thevuthasan, Three-dimensional nanoscale
of Energy Office of Vehicle Technologies, Powertrain Material Core characterisation of materials by atom probe tomography, Int. Mater. Rev. 63 (2)
Program. (2018) 68–101, http://dx.doi.org/10.1080/09506608.2016.1270728.

16
M. Romedenne et al. Corrosion Science 221 (2023) 111348

[16] Y.F. Su, P. Stack, C. Stephens, K. Kane, S.N. Dryepondt, R. Pillai, B. Pint, [27] H. Asteman, J.E. Svensson, L.G. Johansson, M. Norell, Indication of chromium
B.M. Tossey, Quantifying High Temperature Corrosion, Research Org.: Oak oxide hydroxide evaporation during oxidation of 304L at 873 K in the presence
Ridge National Lab. (ORNL), Oak Ridge, TN (United States), 2023, doi: Con- of 10% water vapor, Oxid. Met. 52 (1–2) (1999) 95–111, http://dx.doi.org/10.
ference: CORROSION 2021 - Salt Lake City, Utah, United States of America 1023/a:1018875024306.
- 4/19/2021 8:00:00 AM-4/30/2021 4:00:00 AM, URL https://www.osti.gov/ [28] D.J. Young, B.A. Pint, Chromium volatilization rates from 𝐶𝑟2 𝑂3 scales into
biblio/1783299https://www.osti.gov/servlets/purl/1783299. flowing gases containing water vapor, Oxid. Met. 66 (3–4) (2006) 137–153,
[17] C.T. Rueden, J. Schindelin, M.C. Hiner, B.E. DeZonia, A.E. Walter, E.T. Arena, http://dx.doi.org/10.1007/s11085-006-9030-1.
K.W. Eliceiri, ImageJ2: Imagej for the next generation of scientific image data, [29] C. Gindorf, L. Singheiser, K. Hilpert, Vaporisation of chromia in humid air, J.
BMC Bioinformatics 18 (1) (2017) 529, http://dx.doi.org/10.1186/s12859-017- Phys. Chem. Solids 66 (2–4) (2005) 384–387, http://dx.doi.org/10.1016/j.jpcs.
1934-z. 2004.06.092.
[18] B. Gault, M.P. Moody, J.M. Cairney, S.P. Ringer, Atom Probe Microscopy, Vol. [30] M. Michalik, M. Hansel, J. Zurek, L. Singheiser, W.J. Quadakkers, Effect of water
160, Springer Science & Business Media, 2012. vapour on growth and adherence of chromia scales formed on Cr in high and
[19] K. Wollgarten, T. Galiullin, W.J. Nowak, W.J. Quadakkers, D. Naumenko, Effect low p𝑂2 -environments at 1000 and 1050◦ C, Mater. High Temp. 22 (3–4) (2005)
of alloying additions and presence of water vapour on short-term air oxidation 213–221.
behaviour of cast Ni-base superalloys, Corros. Sci. 173 (2020) 108774, http:// [31] T. Perez, S. Mathieu, Y. Parsa, L. Latu-Romain, Y. Wouters, M. Vilasi, About
dx.doi.org/10.1016/j.corsci.2020.108774, URL https://www.sciencedirect.com/ the synergetic influence of manganese and silicon on the oxidation rate of
science/article/pii/S0010938X1932606X. chromia forming nickel-based model alloys at 1050 ◦ C, Oxid. Met. 94 (3) (2020)
[20] A. Naoumidis, H.A. Schulze, C. Garciarosales, Phase studies in the system 235–249, http://dx.doi.org/10.1007/s11085-020-09988-1.
chromium-manganese-titanium oxide at different oxygen partial pressures, Epdic [32] D.L. Douglass, J.S. Armijo, The effect of silicon and manganese on the oxidation
1 : Eur. Powder Diffr. Pts 1 2 79 (1991) 691–695, URL https://www. mechanism of Ni-20 Cr, Oxid. Met. 2 (2) (1970) 207–231, http://dx.doi.org/10.
sciencedirect.com/science/article/abs/pii/0955221991900544. 1007/BF00603657.
[21] T.D. Nguyen, J. Zhang, D.J. Young, Effects of silicon and water vapour on [33] H.E. Evans, D.A. Hilton, R.A. Holm, S.J. Webster, Influence of silicon additions
corrosion of Fe–20Cr and Fe–20Cr–20Ni alloys in CO2 at 650 ◦ C, Oxid. Met. on the oxidation resistance of a stainless steel, Oxid. Met. 19 (1) (1983) 1–18,
87 (3) (2017) 541–573, http://dx.doi.org/10.1007/s11085-016-9681-5. http://dx.doi.org/10.1007/BF00656225.
[22] B.W. Krakauer, D.N. Seidman, Absolute atomic-scale measurements of the Gibb- [34] B. Li, B. Gleeson, Effects of silicon on the oxidation behavior of Ni-base chromia-
sian interfacial excess of solute at internal interfaces, Phys. Rev. B 48 (1993) forming alloys, Oxid. Met. 65 (1) (2006) 101–122, http://dx.doi.org/10.1007/
6724–6727, http://dx.doi.org/10.1103/PhysRevB.48.6724, URL https://link.aps. s11085-006-9003-4.
org/doi/10.1103/PhysRevB.48.6724. [35] A.J. Moulson, J.P. Roberts, Water in silica glass, Trans. Faraday Soc. 57
[23] T.D. Nguyen, J. Zhang, D.J. Young, Silicon contamination during alloy oxidation (8) (1961) 1208–1216, http://dx.doi.org/10.1039/tf9615701208, URL <Go
in water vapour at 650 ◦ c, Oxid. Met. 97 (5) (2022) 559–574, http://dx.doi. to ISI>://WOS:A19613304C00033 https://pubs.rsc.org/en/content/articlepdf/
org/10.1007/s11085-022-10108-4. 1961/tf/tf9615701208.
[24] J. Crank, The Mathematics of Diffusion, Oxford University Press, Oxford, 1956, [36] E. Bunting, Phase equilibria in the system 𝐶𝑟2 𝑂3 -𝑆𝑖𝑂2 , J. Res. Natl. Bur. Stand.
p. 402, URL http://books.google.de/books?id=eHANhZwVouYC&dq=J.+Crank, 5 (1930).
+The+Mathematics+of+Diffusion&printsec=frontcover&source=bn&hl=de&ei= [37] C.R. Wang, W.Q. Zhang, R.Z. Zhu, Simultaneous sulfidation and oxidation of
eHykTMyJKcG88ga_naX9AQ&sa=X&oi=book_result&ct=result&resnum=4&ved= Fe-25wt-percent Cr and silicon-containing fe-25wt-percent Cr alloys in 𝐻2 -𝐻2 𝑂-
0CDgQ6AEwAw#v=onepage&q=J.%20Crank%2C%20The%20Mathematics% 𝐻2 𝑆 gas-mixtures at high-temperatures, Mater. Sci. Eng. A-Struct. Mater. Prop.
20of%20Diffusion&f=false. Microstruct. Process. 125 (2) (1990) 223–233, http://dx.doi.org/10.1016/0921-
[25] R.A. Swalin, A. Martin, R. Olson, Diffusion of magnesium, silicon, and molyb- 5093(90)90172-Y, URL <Go to ISI>://WOS:A1990DH63900008.
denum in nickel, JOM 9 (7) (1957) 936–939, http://dx.doi.org/10.1007/ [38] G.C. Wood, J. Richardson, M.G. Hobby, J. Boustead, Identification of thin
BF03397944. healing layers at base of oxide scales on 𝐹 𝑒𝐶𝑟 base alloys, Corros. Sci. 9
[26] D.J. Young, Chapter 11 - Effects of water vapour on oxidation, in: High (9) (1969) 659–671, URL https://www.sciencedirect.com/science/article/abs/
Temperature Oxidation and Corrosion of Metals, second ed., Elsevier, 2016, pp. pii/S0010938X69800971.
549–601, http://dx.doi.org/10.1016/B978-0-08-100101-1.00011-X. [39] R.J. Tilley, Defects in solids, in: Defects in Solids, John Wiley and Sons, In.,
Hoboken, 2008, http://dx.doi.org/10.1002/9780470380758.

17