MoSi2 Oxidation in 670-1498 K Water Vapor

Article abstract of DOI:10.1111/jace.14120

Molybdenum disilicide (MoSi₂) has well documented oxidation resistance at high temperature (T > 1273 K) in dry O₂ containing atmospheres due to the formation of a passive SiO₂ surface layer. However, its behavior under atmospheres where water vapor is the dominant species has received far less attention. Oxidation testing of MoSi₂ was performed at temperatures ranging from 670-1498 K in both 75% water vapor and synthetic air (Ar-O₂, 80%-20%) containing atmospheres. Here the thermogravimetric and microscopy data describing these phenomena are presented. Over the temperature range investigated, MoSi₂ displays more mass gain in water vapor than in air. The oxidation kinetics observed in water vapor differ from that of the air samples. Two volatile oxides, MoO₂(OH)₂ and Si(OH)₄, are thought to be the species responsible for the varied kinetics, at 670-877 K and at 1498 K, respectively. Increased oxidation (140-300 mg/cm²) was observed from 980-1084 K in water vapor, where passivation is observed in air.

Full text of DOI:10.1111/jace.14120

J. Am. Ceram. Soc., 99 [4] 1412–1419 (2016)
DOI: 10.1111/jace.14120
© 2016 The American Ceramic Society
ournal
J
MoSi2 Oxidation in 670–1498 K Water Vapor
Elizabeth Sooby Wood,‡,† Stephen S. Parker,‡,§ Andrew T. Nelson,‡ and and Stuart A. Maloy‡
‡Materials Science and Technology Division, Los Alamos National Laboratory, Los Alamos, New Mexico
§Department of Nuclear Engineering, University of California, Berkeley, Berkeley, California
Molybdenum disilicide (MoSi2) has well documented oxidation
resistance at high temperature (T > 1273 K) in dry O2 contain-
ing atmospheres due to the formation of a passive SiO2 surface
layer. However, its behavior under atmospheres where water
vapor is the dominant species has received far less attention.
Oxidation testing of MoSi2 was performed at temperatures
ranging from 670–1498 K in both 75% water vapor and syn-
thetic air (Ar-O2, 80%–20%) containing atmospheres. Here
the thermogravimetric and microscopy data describing these
phenomena are presented. Over the temperature range investi-
gated, MoSi2 displays more mass gain in water vapor than in
air. The oxidation kinetics observed in water vapor differ from
that of the air samples. Two volatile oxides, MoO2(OH)2 and
Si(OH)4, are thought to be the species responsible for the var-
ied kinetics, at 670–877 K and at 1498 K, respectively.
Increased oxidation (140–300 mg/cm2) was observed from 980–
1084 K in water vapor, where passivation is observed in air.
The temperature regime where pesting is a concern is dic-
tated by the behavior of MoO3 when formed according to
Eq. (1) above.6 MoO3 behavior dominates the response of
Mo oxidation in air and water vapor, as recently shown by
Nelson and Sooby et al.7 At temperatures below roughly
873 K, MoO3 will remain
a solid and form whiskers
throughout the material. This causes expansion of the bulk
material and exposure of unoxidized MoSi2, preventing the
SiO2 from forming
a
protective layer.8 Above 873 K,
the MoO3 vapor pressure begins to play a significant role in
the oxidation. As the MoO3 is volatilized, the SiO2 is capable
of forming a continuous protective layer, inhibiting further
oxidation. It is reported that Mo5Si3 forms at the Si depleted
interface between the bulk MoSi2 and SiO2 due to the rapid
diffusion of Si as displayed in Eq. (2).6
Figure 1 displays an Ellingham-type diagram for several
possible reactions of MoSi2 and its resulting oxides in both
air and H2O containing atmospheres. It is noted that the
most energetic of these reactions in air containing atmo-
spheres are described in Eqs. 1 and 2 with the addition of
the reaction described in Eq. (3)
I. Introduction
OLYBDENUM disilicide (MoSi2) has been proposed as a
M
high temperature, corrosion-resistant material since its
discovery in 1907.1 It was first used for high-temperature fur-
nace elements in 1953 and since the 1990s used as a protec-
tive coating for high-temperature industrial applications,
such as turbine airfoils, combustion chambers, and missile
nozzles.2 The oxidation behavior of MoSi2 and MoSi2 com-
posites have been extensively studied in O2 containing atmo-
spheres up to 1973 K. However, its oxidation behavior in
water vapor is largely unknown. This behavior must be
understood if MoSi2 is to be considered for use in oxidizing
atmospheres where water vapor rather than oxygen is the
primary oxidizing reactant.
MoSi2 þ 3O2ðgÞ ! MoO2 þ 2SiO2
(3)
However, MoO2 will readily oxidize to MoO3 in O2 con-
taining atmospheres. Therefore, MoO3 is the resulting final
oxide in air exposures, until it becomes volatile.
(2) MoSi2 Oxidation in Water Vapor Containing
Atmospheres
The data available for water vapor containing atmospheres are
limited to a small range of temperatures. Hansson et al. found
that the oxidation rate of a MoSi2-composite increased with
the addition of 10% water vapor to an O2 containing atmo-
sphere, with the peak rate reported at 743, 40 K lower than
the peak oxidation temperature in pure O2.4 In addition, the
study reports increased mass loss at higher temperatures with
the addition of 10% water vapor. Referring to Fig. 1 in the
presence of both O2 and H2O a very energetic reaction can
occur resulting in the volatile Mo hydroxide species MoO2
(OH)2.
(1) MoSi2 Oxidation in Dry Atmospheres
MoSi2 displays “pesting” when subjected to air oxidation
experiments.3–5 The temperature range for pest formation is
reported at temperatures varying from 623–873 K. Full disin-
tegration of MoSi2 has been observed in O2 containing envi-
ronments after long duration (100 h) exposures in this
temperature range.3,5 Above 873 K, MoSi2 begins to display
excellent oxidation resistance, due to the formation of a pas-
sivating SiO2 layer. The primary MoSi2 oxidation reactions
in dry O2 atmospheres are as follows:
3:5O2ðgÞ þ MoSi2 þ H2O(g) ! MoO2ðOHÞ2 þ 2SiO2
(4)
However, exposure to Ar, Ar + 10% water vapor, and Ar
+ 40% water vapor at 723 K resulted in reportedly negligible
mass gain.4 Looking to Fig. 1, the reactions between MoSi2
and steam leading to volatiles are much less energetic. The
effect of water vapor on MoSi2 oxidation was subsequently
explored to 973 K in an O2 + 10% water vapor atmosphere.9
Increased mass loss, in agreement with Hansson et al., indi-
cated an increase in oxidation with the addition of water
vapor to O2 containing environments. No other studies
focused on the oxidation of MoSi2 in water vapor have been
conducted to the authors’ knowledge.
2MoSi2 þ 7O2ðgÞ ! 2MoO3 þ 4SiO2
5MoSi2 þ 7O2ðgÞ ! Mo5Si3 þ 7SiO2
(1)
(2)
D. Butt—contributing editor
Manuscript No. 37399. Received August 24, 2015; approved December 11, 2015.
†Author to whom correspondence should be addressed. e-mail: sooby@lanl.gov
1412

April 2016
MoSi2 Steam Oxidation
1413
Fig. 1. Ellingham-type diagram plotting the change in the Gibbs Free Energy of a number of relevant oxidation reactions in both O2 and H2O
containing atmospheres. Data obtained using HSC Chemistry 7 Software.
5.26 g/h to the STA. Ar was used as a carrier gas to limit
balance fluctuations while steam flowed into the system. The
transfer lines used to deliver the steam to the STA were held
at 473 K. In addition, 20 mL/min Ar was purged through
the balance as a protective gas. It was calculated that the
sample was exposed to 0.55 ATM water vapor during test-
ing. The steam was turned off at the end of the isotherm,
and the system cooled at 20 K/min under Ar.
II. Experimental Method
MoSi2 obtained from Goodfellow was segmented into
0.07 cm3 parallelepipeds and then ground and polished using
SiC paper to 1200 grit (U.S.) on all six sides. The prepared
samples measured to be 80% theoretical MoSi2 density
(6.26 g/cm3). The anticipated effect of decreased density on
the oxidation behavior observed is elaborated upon in the
Discussion section of this article. The resulting samples were
cleaned using acetone, rinsed with methanol, and allowed to
dry. The mass of each sample was measured using a balance
(XP205, Mettler-Toledo, Columbus, Ohio) calibrated to
0.1 mg. To calculate the total surface area of each block, the
length and width of each sample side were measured to a
0.1 mm accuracy, using digital calibers.
A simultaneous thermal analyzer (STA 449 F3, Netzsch
Instruments, Selb, Germany) equipped with a water vapor
furnace and water vapor generator (DV2ML, Astream, Ger-
many) was used to measure the mass change as a function of
both temperature and time. A thermocouple calibration was
performed using Ni, Au, Ag, Al, and Zn melt point standards
(6.223.5-91.3, Netzsch). The isothermal temperatures were
corrected to reflect this calibration factor. The sample was
placed on an Al2O3 platform in the STA. The furnace tube,
sample holder, and all internal fixturing were Al2O3. The
sample chamber was evacuated and backfilled three times to
minimize the effect of residual oxygen on the oxidation
behavior of the samples. In-line oxygen sensors (RapidOx
OEM447; Cambridge Sensotec, Saint Ives, UK) were used to
monitor the oxygen content in both the gas being delivered to
the system and the exhaust gas leaving the system.
Two oxidizing atmospheres were used in this study. Syn-
thetic air produced by mixing ultra-high purity Ar and pure
O2 (Ar–O2, 80%–20%) was used to compare against litera-
ture data for MoSi2 oxidation at temperatures ranging from
670–1498 K. Water vapor exposures were performed from
670–1498 K at approximately 100 K increments. The sample
was ramped to the isothermal test temperature in Ar at
10 K/min. Negligible mass gain was observed during the
ramp to the isotherm (<0.1% mass change). The sample was
held at the test temperature for 15 min to allow the system
to stabilize under flowing Ar before water vapor was intro-
duced into the system. The oxygen content of the exhaust
was less than 10 parts per million (ppm) O2 for all experi-
ments, ensuring that water vapor was the dominant oxidizing
species for all testing. The water vapor generator supplied
Samples were analyzed by scanning electron microscopy
(SEM). Two orientations were observed. The as-oxidized sur-
face of each sample was imaged initially to analyze the
microstructure of the oxide formed during exposure. For
these as-oxidized surface samples, no surface preparation was
performed in order to preserve the surface oxide. Samples
were coated with Au prior to analysis using a sputter coater.
The samples were also cross-sectioned, potted in epoxy, pol-
ished to 0.25 lm diamond suspension and coated to image
the cross-section of the oxidized sample. The first form of
imaging allowed for visualization of the microstructure of
the as-formed oxide, while the second provided observations
of the depth of oxidation into the bulk of the sample. Ele-
mental analysis was performed by energy dispersive spec-
troscopy (EDS) using a combination of spot and line scans
over the cross-sectioned samples. Both the oxidation inter-
face (edge of the cross-sectioned samples) and the bulk of the
samples were analyzed. X-ray diffraction was performed on
the sample surface to identify the oxide phases that formed
during both air and water vapor exposure, using a Bruker
XRD (D2 Phaser; Bruker AXS, Madison, WI). A two theta
range of 10ꢀ–90ꢀ, 0.01ꢀ step size and 1 s dwell time were used
for all, but two samples. For the 1395 and 1498 K water
vapor samples, a 5 s dwell time was used to increase the sig-
nal-to-noise ratio of the scan.
III. Results
Table I provides a summary of results for MoSi2 exposed to
water vapor.
(1) Thermogravimetric Analysis Results
Figure 2 displays the MoSi2 mass gain normalized to surface
area (mg/cm2) for each 100 K test increasing from 670–
1498 K in water vapor. Four different oxidation kinetic
regimes are identified in the preliminary water vapor testing.

1414
Journal of the American Ceramic Society—Sooby Wood et al.
Vol. 99, No. 4
Table I. Summary of Results for H2O Isothermal Oxidation Testing of MoSi2
Parobolic
oxidation
H2O Exposure
Isothermal
Total mass
rate constant
Phases identified following
temperature (K) dwell time (h) gain (mg/cm2) [mg/(cm2-hð1=2Þ)]
Comments on kinetics
exposure XRD and EDS Analysis
670
773
877
980
1084
1188
1291
1395
1498
24
20
20
20
20
20
10
10
4
3.03
3.9
0.598
0.91
Parobolic
Parobolic
Nonpassivating, nonlinear oxidation
nonpassIvating, nonlinear, rapid oxidation MoSi2, MoO2
Nonpassivating, nonlinear, rapid oxidation MoSi2, MoO2
Passivates after 6 h
Passivates after 1.5 h
passivates after 0.5 h
Nonpassivating, nonlinear oxidation
MoSi2
MoSi2
†
18.95
298.21
141.62
17.01
11.77
9.25
N/A
N/A
N/A
N/A
N/A
N/A
N/A
MoSi2, Mo5Si3, SiO2
MoSi2, Mo5Si3, SiO2
MoSi2, Mo5Si3, SiO2
MoSi2, Mo5Si3
6.52
†Sample was rendered unsuable during sample preparation.
The samples exposed to 980 and 1084 K display rapid oxida-
tion with two separate kinetic regions, the second much more
rapid than the first. The 980 and 1084 K water vapor sam-
ples swelled and warped, as displayed in Fig. 3, during the
exposure with mass gains after 20 h of 300 and 140 mg/cm2,
respectively.
Figure 5 compares the water vapor and air TG data for
1188–1498 K. MoSi2 appears to passivate in both air and
water vapor from 1188 to 1395 K, though the rate of mass
15
The samples exposed to 670 and 773 K water vapor for
24 and 20 h, respectively, displayed less than 4 mg/cm2
mass gain, equating to less that 1% overall mass change
for each sample. Additionally, the 670 and 773 K samples
display very little change upon visual inspection. The ther-
mogravimetric (TG) results for 670–877 K exposures to
both the air and water vapor samples are plotted in
Fig. 4.
10
877 K H2O
773 K H2O
670 K H2O
877 K Air
773 KAir
670 K Air
5
0
200
400
Time (min)
600
Fig. 4. Mass gain (mg/cm2) vs time (min) data for MoSi2 exposed
to 670–877 K isotherms in 0.55 ATM water vapor (hollow symbols)
and synthetic air (solid symbols). The legend is ordered top to
bottom by decreasing mass gain.
20
15
1188 K H2O
1291 K H2O
1395 K H2O
1498 K H2O
1188 K Air
1291 K Air
1395 K Air
1498 K Air
Fig. 2. Mass gain (mg/cm2) vs time (min) data for MoSi2 exposed
to 670–1498 K isotherms in 0.55 ATM water vapor. The legend is
ordered top to bottom by decreasing mass gain.
10
5
0
200
400
600
800
1000
1200
Time (min)
Fig. 5. Mass gain (mg/cm2) vs time (min) data for MoSi2 exposed
to 1188–1498 K isotherms in 0.55 ATM water vapor (hollow
symbols) and synthetic air (solid symbols). The legend is ordered top
to bottom by decreasing mass gain in the water vapor samples.
Fig. 3. Photographs of MoSi2 before (left) and after (right) 1084 K
steam exposure for 20 h.

April 2016
MoSi2 Steam Oxidation
1415
gain decreases, indicating diffusion limited oxidation, much
sooner in air than in water vapor. For both 1498 K sam-
ples, the oxidation rate decreases to a comparatively ele-
vated linear mass gain, indicating that the samples are not
passivating. These samples are further investigated using
SEM.
(2) X-Ray Diffraction
X-Ray diffraction (XRD) patterns of the 773, 980, 1395, and
1498 K water vapor sample surfaces are displayed in Fig. 6.
These four diffraction patterns help to summarize the varied
effects of water vapor on MoSi2 in the regions identified in
the Discussion section of this paper. Diffraction patterns
were also collected for the air samples, although those results
are less informative since the oxide layer formed is much
thinner than the interaction depth of the X-rays with the
sample surface.
(3) Microscopy Results
Both surface and cross-sectional SEM were used to analyze
the resulting microstructure of each sample. The most rapid
oxidation is observed at 980 and 1084 K. Figure 7 displays
the as-oxidized surface of a MoSi2 sample after a 20 h,
1084 K water vapor exposure in increasing magnification
from left to right. The 1084 K sample displays large oxide
grains. In comparison, Fig. 8 displays the top-down perspec-
tive of a MoSi2 sample after a 20 h, 1188 K water vapor
exposure in increasing magnification from left to right. The
1188 K sample has formed a finer-grained oxide than the
sample exposed to 1084 K water vapor. In addition, the layer
is less cracked and uniformly covers the sample surface.
Micrographs of the cross-sections of the 670, 1084, and
1498 K water vapor samples are displayed in Figs. 9–11 to
illustrate three of the four distinct behaviors observed in the
water vapor TG data displayed in Section IV(1). The 670 K
sample displayed in Fig. 9 exhibits a similar microstructure
to the as-received material displayed in Fig. 12, with uni-
formly distributed pores. There is no observable oxide layer
from the cross-sectional view of the sample. The 773 K sam-
ple displayed the same microstructure as the 670 K sample,
with no visible oxide layer in the micrographs and the bulk
of the material having similar structure to the as-received
material.
The 1084 K sample displayed in Fig. 10, which exhibits
dramatic mass gain in Fig. 2, appears to be completely
reacted, with grain boundaries exposed throughout the sam-
ple. Likewise, the 980 K exposure resulted in the same full
sample deterioration as the 1084 K sample shown in Fig. 10.
Finally, the 1498 K sample shown in Fig. 11 has formed a
region of increased porosity approximately 80 lm in depth,
expanding from the edge of the cross-section into the bulk of
the sample. Both the air and the water vapor 1498 K samples
exhibit similar trends of continued oxidation while the 1188–
1395 K samples passivated in both atmospheres. Figure 13
compares the 1498 K synthetic air and water vapor exposed
samples. It should be noted that the synthetic air sample
experienced a 10 h isothermal hold while water vapor iso-
therm was only 4 h.
Fig. 6. Diffraction patterns for the 1498, 1395, 980, and 773 K
sample surfaces following oxidation in water vapor.
Fig. 7. Top-down SEM micrographs of MoSi2 samples after a 20 h, 1084 K water vapor exposure.

1416
Journal of the American Ceramic Society—Sooby Wood et al.
Vol. 99, No. 4
Fig. 8. Top-down SEM micrographs of MoSi2 samples after a 20 h, 1188 K water vapor exposure.
Fig. 9. Cross-sectional SEM micrographs of the MoSi2 after 24 h,
670 K water vapor exposure.
Fig. 11. Cross-sectional SEM micrographs of the MoSi2 after 4 h,
1498 K water vapor exposure.
Fig. 12. Backscatter, cross-sectional SEM micrographs of the as-
received MoSi2.
Fig. 10. Cross-sectional SEM micrographs of the MoSi2 after
20 h, 1084 K water vapor exposure.
(4) Elemental Analysis
Energy dispersive spectroscopy analysis was performed on
several water vapor and air exposed samples to qualitatively
assess the elemental composition of the various phases identi-
fied in BSED. Elemental standards were not used to calibrate
The air sample displayed on the left of Fig. 13 shows uni-
form porosity, similar to the seemingly unaffected 670 K
sample in Fig. 9.

April 2016
MoSi2 Steam Oxidation
1417
Fig. 13. Backscatter, cross-sectional SEM micrographs of MoSi2 samples after 10 h, 1498 K synthetic air exposure (left) and 4 h, 1498 K water
vapor exposure (right).
the detector prior to analysis, therefore the presented analysis
is considered qualitative. In addition, EDS is not the ideal
method to quantify small levels of oxygen in materials; it is
used here to identify relative changes in the oxygen content
and Mo/Si ratio across the samples’ surfaces.
The spot analysis results for the regions identified in
Figs. 9, 10, 11, and 13, are reported in Table II. Line scans
of the cross-sectioned samples were performed, analyzing
from the oxidation interface of each sample to 50 lm into
the bulk of each sample. Due to the porosity of these sam-
ples, the data acquired displays a significant amount of fluc-
tuation as the beam traverses pores in the material.
However, the ratio of Mo/Si as compared to the as-received
material provides insight to the compositional variation
through the cross-section of the exposed samples. Figure 14
displays the line scans for the 670, 1084, and 1498 K water
vapor exposures. The 1498 K water vapor exposure is also
compared to the 1498 K air exposed sample.
Energy dispersive spectroscopy did not reveal any elevated
oxygen levels in either the bulk or along the edges of the 670
and 773 K samples. There are elevated oxygen levels across
the cross-sectioned sample of the 1084 K sample identified
by EDS, indicating oxide formation deep into the bulk of the
sample. EDS indicates that oxygen is only present on the
very edge (top of the 1498 K air sample), in the lighter
region of the material. Each pore in the 80 lm affected
region of the water vapor sample exhibits a similar lighter
outlining area which also contained oxygen, as seen in
Fig. 13.
IV. Discussion
The material obtained from GoodFellow was measured to be
80% theoretical MoSi2 density (6.26 g/cm3). A cross-section
micrograph of the as-received material is displayed in
Fig. 12. The effects of porosity on the oxidation behavior of
MoSi2 was studied using 70% dense materials in a previous
investigation by Kurokawa et al.10 The authors noted a
change in the oxidation kinetics and overall increase in mass
gain of the porous MoSi2 in 773 K air; however, the oxida-
tion behavior observed in the present study agrees with
accepted oxidation kinetics at both high and low tempera-
ture.11 In addition, passivation of MoSi2 is observed in air
above 877 K, in agreement with past experiments, although
at increased total oxidation. Therefore the authors believe
Table II. EDS Spot Analysis for Figs. 9–11
Sample
O K
Si K
Mo L
670 K H2O
Fig. 9
Spot 1
Spot 2
Fig. 10
Spot 1
Spot 2
Spot 3
Fig. 11
Spot 1
Spot 2
Fig. 12
Spot 1
Spot 2
Fig. 12
Spot 1
Spot 2
4.23
6.35
65.45
64.6
30.32
29.05
1084 K H2O
4.21
18.26
13.87
65.57
56.10
58.58
30.22
25.64
27.56
1498 K H2O
1498 K H2O
1498 K air
5.38
30.06
64.87
39.08
29.75
30.86
—
41.9
64.87
29.07
29.75
29.44
Fig. 14.
Mo/Si ratio measured using EDS line scans from the
—
46.71
66.25
27.99
33.75
25.30
oxidized surface to 50 lm into the cross-section sample for 670,
1084, and 1498 K as compared to the as-received material.

1418
Journal of the American Ceramic Society—Sooby Wood et al.
Vol. 99, No. 4
Fig. 15. 670–877 K water vapor (left) and air (right) TGA data plotted in mass gain (mg/cm2) vs h1=2. Note the different scales for both the
mass gain and time axis in the two plots.
that the kinetic behavior and microstructural changes to this
material are not affected by porosity, although the total oxi-
dation of each sample measured by thermogravimetric analy-
sis is expected to be higher in the 80% dense material than
in a fully dense sample.
that MoO2 will form via the reaction displayed in Eq. (6)
and plotted in Fig. 1 but will remain on the surface unless
the kinetics are sufficiently favorable to convert it into MoO3
or MoO2(OH)2.
MoSi2 þ 6H2O(g) ! MoO2 þ 2SiO2 þ 6H2
(6)
(1) 670–773 K Oxidation Behavior
MoO2 formation has been observed in Mo steam oxida-
tion testing, and while it could provide a diffusion barrier to
further oxidation in steam, it is rapidly oxidized to MoO3.16
At lower temperatures (below 1223 K), some MoO2 remains
on the surface. It is observed in the XRD data for the 980 K
water vapor sample, though not enough MoO2 was formed
at 670–773 K to be detected via XRD.
The data for 670–877 K water vapor (left) and synthetic air
(right) exposures is replotted in Fig. 15. The x-axis units are
h1=2, allowing for a clear view of the parabolic nature of the
670–773 K water vapor mass gain curves. Longer duration
tests are necessary to determine if the resulting material is
passivating in water vapor at these lower temperature expo-
sures to water vapor, where in O2 containing atmospheres
the materials is known to “pest”. The air curves do not dis-
play signs of parabolic oxidation, which is to be expected,
given the documented pest formation in this temperature
range. The oxidation rate in the water vapor of MoSi2 is
expected to decrease as a function of time, due to the diffu-
sion limited, parabolic oxidation kinetics displayed at 670–
773 K.12,13
It is expected that the volatile MoO2(OH)2, formed via the
reaction displayed in Eq. (5), contributes to the change in
oxidation dynamics seen at 670–773 K in water vapor.
Though the thermodynamics are not favorable in this tem-
perature range according to the data plotted in Fig. 1, The
reaction has been stated in the literature to occur even at
these lower temperatures.14,15
(2) 877–1084 K Oxidation Behavior
At 877 K, the oxidation kinetics is no longer parabolic in
water vapor, reference Fig. 15. In addition, the mass gain
displayed after 20 h has increased by nearly a factor of five
when compared to the 773 K water-vapor exposure. This
supports the conclusion that the oxidation of MoSi2 in water
vapor is no longer diffusion limited. In this temperature
regime, both MoO3 and MoO2(OH)2 are volatile species.17
The XRD data of the 980 K water-vapor sample displays
MoSi2 and broad MoO2 peaks, where as only Mo5Si3 and
SiO2 phases were found in the air sample. This suggests that
MoSi2 oxidized in this temperature regime is incapable of
forming a uniform SiO2 required for passivation. Though
there are signs (low angle, broad peak in XRD data) that
some amorphous oxide could have formed.
MoO3 þ H2O(g) ! MoO2ðOHÞ2ðgÞ
(5)
The 980–1084 K samples were observed to swell and warp
following testing, see Fig. 3. Oxidation penetrating through-
out the bulk of MoSi2 at 973–1073 K water vapor is attribu-
ted here to the lack of uniform SiO2 formation in water
vapor at these temperatures, likely due to the formation of
MoO2 observed in XRD analysis (reference Fig. 6) through-
out the bulk of the reacted sample inhibiting the formation
of a uniform SiO2 layer.
In dry O2 containing atmospheres, MoO3 volatility aids in
homogeneous SiO2 scale formation with Mo5Si3 intermetallic
formed between the native MoSi2 and oxide scale via the
reaction presented in Eq. (2), thus leading to passivation
above 873 K.6,8,5 Similarly, it is argued here, at 670–773 K,
the MoO2(OH)2 volatility allows for a more passive oxide
formation, primarily SiO2. In a previous study, the hydroxide
volatility was displayed as linear mass loss of Mo exposed to
steam as low as 823 K where in air MoO2 remained a stable
oxide on the surface of the Mo sample.7 In addition to SiO2
formation, MoO2 is also formed at this temperature range
during steam oxidation. In high temperature steam, the DG
of the reaction between H2O and MoO2 resulting in MoO3
and H2 is positive across this temperature range as is the
reaction forming the hydroxide. Therefore, it is anticipated
(3) 1188–1395 K Oxidation Behavior
It is in the 1188–1395 K temperature range that the water
vapor and air data display similar oxidation trends with two
linear oxidation regions, the first more rapid than the second.
Figure 5 shows the mass gain curves in both atmospheres to
be very similar. The XRD data displays both Mo5Si3 and

April 2016
MoSi2 Steam Oxidation
1419
SiO2 in these water vapor samples, similar to the passivated
air samples. However, MoO2 is seen in some low intensity,
broad peaks. The higher mass gain observed in the water
vapor samples is likely caused by the continued formation of
MoO2. Likewise, while not displayed here, the SEM data
showed very little porosity after the 1188 K water vapor
exposure, and even less porosity in the 1291–1395 K samples.
In this temperature regime, while there is an increase in mass
gain, MoSi2 passivates in water vapor.
to parabolic oxidation kinetics, slowing the oxidation reac-
tion at 670–773 K. The second removes the protective SiO2
layer, rendering the material susceptible to further oxidation
at temperatures above 1473 K.
Acknowledgments
This work was supported by the U.S. Department of Energy, Office of Nuclear
Energy Fuel Cycle Research and Development program. The authors thank
Ming Tang for his help in imaging the 1498 K samples.
(4) 1498 K Oxidation Behavior
References
While the TG data for the 1498 K air and water vapor expo-
sures display similar oxidation trends, the effect of the water
vapor exposure on the resulting microstructure is vastly dif-
ferent than the air exposure as exhibited in Fig. 13. The
1498 K water vapor sample displays increased porosity and
oxygen rich regions outlining the pores. This indicates that
MoSi2 was not passivating at 1498 K in water vapor, where
it formed only a small surface oxide layer in synthetic air.
The behavior in water vapor is attributed here to a second
volatile hydroxide. SiO2 forms a volatile hydroxide Si(OH)4
via the reaction described in Eq. (7).18–20 Though the ener-
getics of this reaction and the other volatile hydroxides
known to form from SiO2 reactions with steam were not
available to the authors.
1O. Hnigschmid, “ber das Molybdnsilicid MoSi2, das Wolframsilicid WSi2
und das Tantalsilicid TaSi22,” Monatshefte fr Chemie und verwandte Teile
anderer Wissenschaften, 28 [8] 1017–28 (1907).
2Z. Yao, J. Stiglich, and T. Sudarshan, “Molybdenum Silicide Based Materi-
als and Their Properties,” J. Mater. Eng. Performance, 8 [3] 291–304 (1999).
3T. Chou and T. Nieh, “Pesting of the High-Temperature Intermetallic
MoSi2,” J. Miner. Metals Mater. Soc. (TMS), 45 [12] 15–21 (1993).
4K. Hansson, M. Halvarsson, J. Tang, R. Pompe, M. Sundberg, and J.-E.
Svensson, “Oxidation Behaviour of a MoSi2-Based Composite in Different
Atmospheres in the Low Temperature Range (400–550ꢀC),” J. Eur. Ceram.
Soc., 24 [13] 3559–73 (2004).
Available at: http://www.sciencedirect.
com/science/article/pii/S0955221903008975 [accessed on 19 October 2015].
5T. Chou and T. Nieh, “New Observations of MoSi2 Pest at 500ꢀC,” Scr.
Metall. Mater., 26 [10] 1637–42 (1992). Available at: http://www.sciencedirect.
com/science/article/pii/0956716X9290270O [accessed on 19 October 2015].
6C. Wirkus and D. Wilder, “High-Temperature Oxidation of Molybdenum
Disilicide,” J. Am. Ceram. Soc., 49 [4] 173–7 (1966).
7A. Nelson, E. Sooby, Y.-J. Kim, B. Cheng, and S. Maloy, “High Tempera-
ture Oxidation of Molybdenum in Water Vapor Environments,” J. Nucl.
Mater., 448 [13] 441–7 (2014). Available at: http://www.sciencedirect.
com/science/article/pii/S0022311513011938 [accessed on 27 November 2015].
8T. Chou and T. Nieh, “Kinetics of MoSi2 Pest During Low-temperature
Oxidation,” J. Mater. Res., 8 [07] 1605–10 (1993).
SiO2 þ 2H2O(g) ! Si(OH)4ðgÞ
(7)
SiO2 is identified via XRD in the water vapor samples
exposed to temperatures exceeding 1084 K. SiO2 has been
found to exhibit porosity when exposed to T >1473 K water
vapor.18,21 It is expected that starting at temperatures greater
than 1473 K MoSi2 will not passivate during oxidation under
water vapor environments. SiO2 is not identified in the
1498 K sample surface, though Mo5Si3 is detected, as shown
in Fig. 6. This further indicates that SiO2 was likely formed
via the reaction described in Eq. (2) but was volatilized at
this temperature.
9K. Hansson, J. Tang, M. Halvarsson, R. Pompe, M. Sundberg, and J.-E.
Svensson, “The Beneficial Effect of Water Vapour on the Oxidation at 600
and 700ꢀC of a MoSi2-Based Composite,” J. Eur. Ceram. Soc., 25 [1] 1–11
(2005). Available at: http://www.sciencedirect.com/science/article/pii/S095522
1904000299 [accessed on 19 October 2015].
10K. Kurokawa, H. Houzumi, I. Saeki, and H. Takahashi, “Low Tempera-
ture Oxidation of Fully Dense and Porous MoSi2,” Mater. Sci. Eng.: A, 261
[1] 292–9 (1999).
11Y. Liu, G. Shao, and P. Tsakiropoulos, “On the Oxidation Behaviour of
MoSi2,” Intermetallics, 9 [2] 125–36 (2001).
12C. Wagner, “Theoretical Analysis of the Diffusion Processes Determining
the Oxidation Rate of Alloys,” J. Electrochem. Soc., 99 [10] 369–80 (1952).
13N. Cabrera and N. Mott, “Theory of the Oxidation of Metals,” Rep.
Prog. Phys., 12 [1] 163–84 (1949).
V. Conclusions
14G. Belton and A. Jordan, “The Volatilization of Molybdenum in the Pres-
ence of Water Vapor,” J. Phys. Chem., 69 [6] 2065–71 (1965).
MoSi2 displays different oxidation behavior in water vapor
than in dry air. In the 670–1498 K temperature range, four
distinct behaviors are observed. Parabolic oxidation is exhib-
ited in only 670–773 K water vapor, a temperature range in
which the material pests in dry O2 environments. From 877–
1084 K in water vapor, MoSi2 undergoes rapid mass gain
resulting in oxidation throughout the bulk of the sample at
980 and 1084 K. The resulting material displays swelling and
warping after the 980–1084 K exposures. MoSi2 exhibits the
greatest resistance to water vapor oxidation in the 1188–
1395 K temperature range, passivating and displaying very
little visual or microstructural signs of oxidation. At 1498 K,
the passivating SiO2 layer is volatilized as a hydroxide in
water vapor environments, degrading its performance.
15T. Millner and J. Neugebauer, “Volatility of the Oxides of Tungsten and
Molybdenum in the Presence of Water Vapour,” Nature, 163, 601–2 (1949).
16M. Kilpatrick and S. K. Lott, “Reaction of Flowing Steam with Refrac-
tory Metals. 1, 2 i. Molybdenum (1100–1700),” J. Phys. Chem., 69 [5] 1638–40
(1965).
17P. E. Blackburn, M. Hoch, and H. L. Johnston, “The Vaporization of
Molybdenum and Tungsten Oxides,” J. Phys. Chem., 62 [7] 769–73 (1958).
18T. Cheng and P. F. Tortorelli, “Silicon Carbide 389 Oxidation in High-
Pressure Steam,” J. Am. Ceram. Soc., 96 [7] 2330–7 (2013).
19E. J. Opila, N. S. Jacobson, D. L. Myers, and E. H. Copland, “Predicting
Oxide Stability in High-Temperature Water Vapor,” J. Miner., Metals Mater.
Soc. (TMS), 58 [1] 22–28 (2006).
20E. J. Opila, D. S. Fox, and N. S. Jacobson, “Mass Spectrometric Identifi-
cation of Si-O-H(g) Species from the Reaction of Silica with Water Vapor at
Atmospheric Pressure,” J. Am. Ceram. Soc., 80 [4] 1009–12 (1997).
21K. A. Terrani, B. A. Pint, C. M. Parish, C. M. Silva, L. L. Snead, and Y.
Katoh, “Silicon Carbide Oxidation in Steam up to 2MPa,” J. Am. Ceram.
The two volatile hydroxides MoO2(OH)2(g) and Si(OH)4
(g) play significant roles in the behavior of MoSi2 in water
vapor. The first facilitates uniform SiO2 formation and leads
Soc., 97 [8] 2331–52 (2014).
h