High-temperature chemistry and oxidation of ZrB2 ceramics containing SiC, Si3N4, Ta5Si3, and TaSi2

Article abstract of DOI:10.1111/j.1551-2916.2008.02420.x

The effect of Si₃N₄, Ta₅Si₃, and TaSi₂ additions on the oxidation behavior of ZrB₂ was characterized at 1200°-1500°C and compared with both ZrB₂ and ZrB₂/SiC. Significantly improved oxidation resistance of all Si-containing compositions relative to ZrB₂ was a result of the formation of a protective layer of borosilicate glass during exposure to the oxidizing environment. Oxidation resistance of the Si₃N ₄-modified ceramics increased with increasing Si₃N ₄ content and was further improved by the addition of Cr and Ta diborides. Chromium and tantalum oxides induced phase separation in the borosilicate glass, which lead to an increase in liquidus temperature and viscosity and to a decrease in oxygen diffusivity and of boria evaporation from the glass. All tantalum silicide-containing compositions demonstrated phase separation in the borosilicate glass and higher oxidation resistance than pure ZrB₂, with the effect increasing with temperature. The most oxidation-resistant ceramics contained 15 vol% Ta₅Si₃, 30 vol% TaSi₂, 35 vol% Si₃N₄, or 20 vol% Si ₃N₄ with 10 mol% CrB₂. These materials exceeded the oxidation resistance of the ZrB₂/SiC ceramics below 1300°-1400°C. However, the ZrB₂/SiC ceramics showed slightly superior oxidation resistance at 1500°C.

Full text of DOI:10.1111/j.1551-2916.2008.02420.x

J. Am. Ceram. Soc., 91 [7] 2250–2257 (2008)
DOI: 10.1111/j.1551-2916.2008.02420.x
r 2008 The American Ceramic Society
No claims to original US government works
ournal
J
High-Temperature Chemistry and Oxidation of ZrB2 Ceramics
Containing SiC, Si3N4, Ta5Si3, and TaSi2
Inna G. Talmy,w James A. Zaykoski, and Mark M. Opeka
Naval Surface Warfare Center Carderock Division (NSWCCD), West Bethesda, Maryland 20817-5700
The effect of Si3N4, Ta5Si3, and TaSi2 additions on the oxida-
tion behavior of ZrB2 was characterized at 12001–15001C and
compared with both ZrB2 and ZrB2/SiC. Significantly improved
oxidation resistance of all Si-containing compositions relative to
ZrB2 was a result of the formation of a protective layer of
borosilicate glass during exposure to the oxidizing environment.
Oxidation resistance of the Si3N4-modified ceramics increased
with increasing Si3N4 content and was further improved by the
addition of Cr and Ta diborides. Chromium and tantalum oxides
induced phase separation in the borosilicate glass, which lead to
an increase in liquidus temperature and viscosity and to a
decrease in oxygen diffusivity and of boria evaporation from
the glass. All tantalum silicide-containing compositions demon-
strated phase separation in the borosilicate glass and higher
oxidation resistance than pure ZrB2, with the effect increasing
with temperature. The most oxidation-resistant ceramics con-
tained 15 vol% Ta5Si3, 30 vol% TaSi2, 35 vol% Si3N4, or 20
vol% Si3N4 with 10 mol% CrB2. These materials exceeded the
oxidation resistance of the ZrB2/SiC ceramics below 13001–
14001C. However, the ZrB2/SiC ceramics showed slightly
superior oxidation resistance at 15001C.
position of the oxide layer, leading to decreased inward diffusion
of oxygen, is an effective way of controlling oxidation resistance
of nonoxide ceramics. This modification can be accomplished by
changing the bulk composition of ceramics.
Numerous studies dedicated to the improvement of the oxi-
dation resistance of ZrB2 and HfB22,5,6,9–15 resulted in the selec-
tion of ceramics with 20–30 vol% SiC as optimal, preferred
compositions. The SiC-containing ZrB2 ceramics showed rela-
tively high oxidation resistance up to 15001C as a result of the
formation of the protective surface layer of borosilicate glass.
Compared with liquid B2O3, the borosilicate glass has higher
melting temperature, higher viscosity, and lower oxygen diff-
usivity, thus providing much more effective oxidation protec-
tion. Theoretical aspects of the oxidation behavior of the
zirconium boride-based materials have been reported earlier.1
The increase in the oxidation resistance of ZrB2/25 vol% SiC
ceramics was accomplished by the addition of CrB2, TiB2, TaB2,
NbB2, and VB2 (as substitution for ZrB2),16 despite the fact that
the oxidation resistance of all the modifying diborides (alone) is
much worse than that of the ZrB2 and HfB2.17,18 This improve-
ment in the oxidation resistance was related to the presence of
transition metal oxides (as a result of the oxidation of the cor-
responding diborides) in the borosilicate glass inducing its phase
separation (immiscibility). Both increased liquidus temperatures
and viscosities, which are characteristic features of immiscible
glasses, and are beneficial for decreasing oxygen diffusivity and
suppressing boria evaporation from the glass.
I. Introduction
HE unique combination of properties (high melting temper-
ature, high hardness and strength, good oxidation resis-
T
tance, and high thermal conductivity) makes ceramics based on
ZrB2 and HfB2 promising candidates for high-temperature
applications1–7 such as hypersonic vehicle aerosurfaces (leading
edges and nosecaps), which require oxidation resistance at tem-
peratures above 20001C.
The effectiveness of oxides in enhancing immiscibility increas-
es with increasing metallic element cation field strength, z/r,2
where z is the valence and r is the ionic radius.19–22 Because the
cation field strengths of Ti, Nb, Ta, Mo, Cr, and V are higher
than that of Zr, these elements are effective in promoting phase
separation of the ZrO2-containing borosilicate glass.
The good oxidation resistance of pure Zr and Hf diborides is
a result of the formation of an oxidation product consisting of
ZrO2 (HfO2) and B2O3. Below 12001C liquid boria fills all of the
porosity in the fine-grained oxides, providing oxidation protec-
tion. However, above 12001C the diborides oxidize rapidly as a
result of the evaporation of the protective B2O3 from the oxide
scale. It should be noted that a fraction of boria (about 15 vol%)
remains within the oxide scale at temperatures up to 15001C
because of the high surface tension of the porous, fine-grained
ZrO2 layer.8 The presence of liquid boria explains the superior
oxidation resistance of these borides compared with the corre-
sponding carbides and nitrides, which form single-phase, porous
ZrO2 scales as a result of the formation of gaseous oxidation
products.
Studies on the effect of alternative sources of Si (instead of
SiC), such as Si3N4 and silicides, on the oxidation behavior of
ZrB2 ceramics are not systematic and are sometimes contradic-
tory. Bellosi and Monteverde23 reported an increase in the
oxidation resistance of ZrB2 ceramics at 13501 and 16001C
with the addition of 5 vol% Si3N4. During hot pressing at
17001C, Si3N4 decomposed with the formation of BN, ZrO2 and a
B–N–O–Zr–Si glassy phase. Unlike the previous data, Sato
et al.24 did not observe any chemical interaction between
Si3N4 fibers and the ZrB2 matrix after hot pressing at 14001C.
The effect of the disilicides of Ta, Nb, W, Mo, and Zr, as well
as of Zr5Si3 on the oxidation resistance of ZrB2 was studied by
Shaffer.25,26 In these studies, ZrB2 ceramics containing 5–15
mol% MoSi2 showed superior oxidation performance at both
14001 and 19501C. Shaffer also suggested the addition of ZrB2/
5–50 mol% TaSi2 mixture as oxidation-protective coatings for
Mo, borides, and carbides. An increase in the oxidation resis-
tance of ZrB2 up to 12001C with the addition of ZrSi2 was
reported by Lavrenko et al.27 The authors also found solid sol-
ubility between the components after pressureless sintering in
vacuum (no sintering temperature was given). Panasyuk
et al.28 reported an improvement in oxidation resistance of
ZrB2 with the addition of 15–50 wt% CrSi2. The experiments
were conducted at 6001–12001C and 740 torr (98.6 kPa) oxygen
In general, the oxidation behavior of non-oxide ceramics
largely depends on the chemical composition and properties of
the oxidation products and on the combination of physical and
chemical processes taking place on the surface exposed to oxy-
gen-containing atmosphere. Modification of the chemical com-
N. Jacobson—contributing editor
Manuscript No. 24005. Received November 19, 2007; approved February 8, 2008.
wAuthor to whom correspondence should be addressed. E-mail: inna.talmy@navy.mil
2250

July 2008
High-Temperature Chemistry and Oxidation of ZrB2-Based Ceramics
2251
pressure for 120 min. The improvement in oxidation resistance
was attributed to the formation of protective ZrSiO4.
X-ray and microscopic evaluations. XRD and SEM/EDS were
used to characterize the phase composition and microstructure
of the surface and cross-section of the samples after oxidation.
Pastor and Meyer29 evaluated the oxidation resistance of
ZrB2 with additions of MSi2 or M5Si3, where M is a transition
metal Zr, Ta, Cr, Mo, or W. The authors observed significant
interactions or solid solubility in the systems of ZrB2 with Zr,
Cr, and Mo silicides and very little or no interactions in the
systems with Ta and W silicides. After oxidation testing for up
to 100 h at 12001 and 14001C, the ZrB2/15 wt% CrSi2 compo-
sition was the most oxidation resistant. Opila et al.30 reported
that the addition of 20 vol% TaSi2 to ZrB2/20 vol% SiC
improved the oxidation resistance of the material at 16271C in
air. The improved behavior was attributed to the presence of Ta,
not to the increase in Si content as a result of the introduction of
the second Si-containing component. The processing, chemical
interactions, and properties of ceramics in the whole ZrB2/Ta5
Si3 system were reported by Talmy et al.31
III. Results and Discussion
(1) ZrB2–Si3N4 Ceramics
The ZrB2 ceramics containing 5–35 vol% Si3N4 were prepared
by hot pressing mixtures of the end-member powders at 18501C
and 20 MPa in He for 1 h. Unlike SiC, which is chemically
compatible with ZrB2,33 Si3N4 was decomposed during hot
pressing. The XRD analysis showed the presence of BN,
ZrSi2, and ZrN in addition to the residual ZrB2. The coexis-
tence of ZrSi2 and ZrB2 observed in the present research con-
firms the Brukl phase equilibria data,34 but contradicts the solid
solubility data reported by Lavrenko et al.27 and Kieffer and
Benesovsky.35
The purpose of the present work is to characterize the oxi-
dation behavior of ZrB2 modified with Si3N4, Ta5Si3, and TaSi2
as sources of Si and to compare the results with the oxidation
behavior of SiC-containing ZrB2. The introduction of Ta silic-
ides was of particular interest because it simultaneously modifies
the oxidation layer with SiO2 to form borosilicate glass and with
Ta2O5 to induce glass immiscibility for potential further
improvement in oxidation resistance.
The experimental compositions ZrB2/5, 20, and 35 vol%
Si3N4, ZrB2/8, 15, and 31 vol% Ta5Si3, and ZrB2/10, 20, and
30 vol% TaSi2 were selected to keep the B2O3/SiO2 ratio in the
surface borosilicate glass between 1.5 and 3, assuming that ZrB2
and Si-containing additives are totally oxidized. (This ratio in
the ZrB2/25 vol% SiC composition is 2.) According to the SiO2–
B2O3 phase diagram,32 the melting temperature of borosilicate
glass of these compositions would be about 6001–7001C. It
should be noted that SiO2 content in glass and glass melting
temperatures will gradually increase as B2O3 evaporates with
increasing oxidation temperature and/or time.
The relative densities of the ceramics were not determined
because of the unknown quantitative phase composition of the
final materials. However, the samples did not have any open
porosity. The microstructure of materials containing up to 20
vol% Si3N4 in the starting mixtures consists of coarse crystals
surrounded by a fine-grained phase (Fig. 1). The EDS analysis
showed only Zr in the coarse crystals (presumably ZrB2) and Zr,
Si, and N in the fine-grained phase (ZrB2, BN, ZrSi2, and ZrN
reaction products). The sizes of both ZrB2 and fine-phase crys-
tals decrease with increasing Si3N4 content, and ceramics con-
taining 35 vol% Si3N4 had a uniform fine-grained structure.
The oxidation resistance of the Si3N4-containing ceramics
during furnace heating increased (butmass gain decreased) with
increasing Si3N4 (Fig. 2). The lower mass gain for pure ZrB2
compared with the sample containing 5 vol% Si3N4 is due to the
significant evaporation of B2O3 from pure ZrB2 above 12001C,
decreasing the total weight gain from the formation of ZrO2 and
B2O3 during oxidation. The significantly larger mass gain mea-
sured for the 5 vol% Si3N4-containing sample in comparison with
that for pure ZrB2 resulted from the lower evaporation of B2O3
from borosilicate glass formed on the surface of this sample.
The smaller thickness of the oxidation layer for the 5 vol%
Si3N4-containing sample (140 mm) compared with that for pure
ZrB2 (160 mm) after 2 h oxidation at 13001C affirms improved
oxidation resistance of ZrB2 with the addition of even small
amounts of Si3N4. The data on the formation of the surface
glass and improvement in the oxidation resistance of ZrB2 by
the addition of 5 vol% Si3N4 support the results reported by
Bellosi and Monteverde.23 The thickness of the oxidized layer
for the ZrB2/20 vol% Si3N4 ceramics tested at 13001C was 60
mm. Only ZrO2 was identified on the surface of all oxidized
Si3N4-containing samples.
II. Experimental Procedure
The starting borides and silicides (ꢀ325 mesh and 99.5% purity)
were purchased from Cerac Inc. (Milwaukee, WI). Silicon nit-
ride (LC12, particle size 0.6 mm) was purchased from HC Starck,
Inc. (Newton, MA). The powders were mixed with a mortar and
pestle in acetone. The mixing procedure was repeated three
times with intermediate drying. The materials were densified by
hot pressing (Model 912G, Thermal Technology Inc., Santa
Rosa, CA) in a graphite mold in He atmosphere. The hot-press-
ing conditions will be given in the sections corresponding to the
specific systems.
The samples were heated to the maximum temperature with a
ramp rate of 401C/min under a pressure of 5 MPa, and the full
pressure was applied when the desired densification temperature
was reached. The hot-pressed samples were characterized by
density and open porosity (Archimedes principle), phase com-
position (X-ray diffraction (XRD), Siemens Theta/Theta, Mod-
el D 500, Bruker AXS, Madison, WI), and microstructure
(SEM, Model ISI ABT SR-50A, Withington, Manchester,
UK). Energy-dispersive spectroscopy (EDS) was also used to
identify the elemental composition of phases. The boron content
data were inconclusive and are not reported in the paper.
The oxidation behavior was characterized by measuring mass
changes of the bars (average size about 3 mm ꢁ 4 mm ꢁ 5 mm
and surface area about 1.0 cm2) during furnace heating in air at
10001–15001C for 2 h. The bars were supported by their ends on
alumina semirings for maximum contact with the atmosphere.
The mass with and without the semi-rings was recorded to allow
posttest mass measurements in the event that the glassy oxida-
tion product reacted with the alumina. The mass change results
were an average from three to five samples. Air quenching of the
samples after furnace heating was conducted to retain the high-
temperature condition of the surface layer for room temperature
By analogy with ZrB2/SiC materials, CrB2 and TaB2 were
added to the ZrB2/Si3N4 ceramics to induce phase separation in
Fig. 1. SEM of ZrB2/20 vol% Si3N4 ceramics hot-pressed at 18501C
and 20 MPa for 1 h.

2252
Journal of the American Ceramic Society—Talmy et al.
Vol. 91, No. 7
the surface glass during the oxidation test. The additives were
introduced as a 10 mol% substitution for ZrB2 with constant 20
vol% Si3N4. No significant changes in the microstructure were
observed for the modified ceramics compared with the baseline
material. As previously discussed, Si3N4 reacted with ZrB2 dur-
ing hot pressing with the formation of BN, ZrSi2, and ZrN,
which were localized in the fine-grained areas. By EDS analysis,
the modifying elements were detected only in the fine-grained
phase. The dissolution of CrB2 and TaB2 in ZrB2 was deter-
mined by XRD. The ZrB2 peaks were shifted to higher angles,
indicating the contraction of the ZrB2 lattice as a result of the
incorporation of CrB2 and TaB2, which have smaller lattice
constants than ZrB2.
The oxidation behavior of the modified compositions togeth-
er with that of ZrB2 with 20 vol% Si3N4 and ZrB2 with 25 vol%
SiC ceramics is shown in Fig. 3. The modifying diborides
increased the oxidation resistance of the baseline ceramics.
The mass gain of the ceramics containing TaB2 and CrB2 was
26 and 34 g/m2, respectively, compared with 49 g/m2 for the
baseline composition after 2 h oxidation at 13001C. Higher ox-
idation resistance of these ceramics at temperatures below
14001C compared with that of ZrB2/SiC ceramics can be attrib-
uted to the oxidation rate difference between ZrSi2 (the Si source
in these ceramics) and SiC. It is known that SiC does not oxidize
appreciably at these lower temperatures to participate in the
formation of borosilicate glass.16 However, ZrSi2 oxidizes more
readily and enhances the low-temperature formation of
the protective borosilicate glass. Lavrenko’s observation27 of
the formation of a vitreous borosilicate film after oxidation
of the ZrB2/50 wt% ZrSi2 material at 7001–8501C is consistent
with the present results.
300
200
100
0
Scale Thickness
5 v% Si N
3
4
1
2
3
60 µm
140 µm
160 µm
ZrB
2
20 v% Si N
3
4
2
35 v% Si N
3
1
3
4
ZrB /25 v% SiC
2
1200
1300
1400
1500
Temperature (°C)
Fig. 2. Mass changes during furnace oxidation (2 h) of ZrB2/Si3N4
ceramics at 12001–15001C.
200
Baseline (ZrB / 20 v%Si N )
2
3 4
150
100
50
10 mole% CrB
2
10 mole% TaB
2
The presence of Ta and Cr oxides, which induce phase sep-
aration in the glass, further improves glass-protective capabili-
ties. The effect of the additives was very similar at temperatures
below 14001C. Above 14001C the best oxidation resistance was
shown by the CrB2-containing material.
ZrB / 25 v% SiC
2
Extensive phase separation (mostly ‘‘cabbage’’-like configu-
rations) was observed on the surface of all the modified samples
after oxidation tests. Figure 4 shows the SEM micrograph of the
surface of the CrB2-containing sample after oxidation at 13001C
for 2 h. The overall XRD of the surface of this sample identified
ZrO2, Cr2O3, and CrO2. The presence of Cr in the Cr41 oxida-
tion state having a high cation field strength (1322 compared
with 793 nmꢀ2 for Cr31) contributed significantly to the phase
0
1200
1300
Temperature, °C
1400
1500
Fig. 3. Mass changes during furnace oxidation (2 h) of ZrB2/20 vol%
Si3N4 ceramics modified with 10 mol% CrB2 and TaB2 at 12001–15001C.
Zr – 6.6; Si – 56.8; Cr – 0 at. %
Zr – 1.9; Si – 63.9; Cr – 5.9 at. %,
Zr – 17.5; Si – 8.9; Cr – 0 at. %
100 µm
Fig. 4. SEM of the surface of ZrB2/20vol% Si3N4/10 mol% CrB2 ceramics after oxidation at 13001C for 2 h.

July 2008
High-Temperature Chemistry and Oxidation of ZrB2-Based Ceramics
2253
Table I. Phase Composition of ZrB2/Ta5Si3 Ceramics After
Hot Pressing at 19001C for 30 min
The samples containing 10 mol% Ta5Si3 also showed the pres-
ence of ZrB2 and ZrB2-based solid solution. However, the solid-
solution unit cell volume only slightly changed with an increase in
Ta5Si3 content from 4 to 10 mol%. This indicated that the sol-
ubility limit of Ta in the ZrB2 lattice, with the present reactant
ratios, was achieved in the material containing about 4 mol%
Ta5Si3. The simultaneous presence of ZrB2 and ZrB2-based solid
solutions is probably the result of insufficient time (30 min) dur-
ing hot pressing to completely homogenize the composition.
Additionally, the samples with 10 mol% Ta5Si3 contained
substantial amounts of two new phases: TaB and a Ta5Si3-based
hexagonal (D88) solid solution with considerable expansion of
Ta5Si3 (tetragonal) content in
ZrB2/Ta5Si3 starting mixtures
Phases determined by XRD
Ta5Si3 (D88
Mol%
Vol%
ZrB2
ZrB2 (ss)
TaB
hexagonal) (ss)
0
2
4
10
0
8
15
31
x
x
x
x
x
x
x
3
˚
x
x
the silicide cell volume from 252.77 A (for pure hexagonal
3
Ta5Si3) to 273.83 A for the solid solution formed in the mate-
˚
XRD, X-ray diffraction.
rial, as determined by XRD. The possible reaction between the
components could be presented as a hypothetical equation:
separation in the glass and, subsequently, to the improvement in
its oxidation protection capability, which is especially pro-
nounced at temperatures below 14001C. The EDS analysis of
this sample showed a significant concentration of Zr in the top
area of the ‘‘cabbage’’ (17.5 at.% compared with 1.9 at.% in the
glassy area) and the presence of Cr only in the glass.
Ta5Si3 þ ZrB2 ! ðZr; TaÞB2 þ TaB þ ðTa; ZrÞ5Si3B
This reaction involved the decomposition of some fraction of
the ZrB2, which provided B for the formation of TaB, as well as
Zr and B for the formation of the D88-based solid solution. Both
the interstitial incorporation of B and the substitution of Ta with
Zr resulted in significant D88 lattice expansion. It is also note-
worthy that the TaB lattice parameters corresponded to the sto-
ichiometric composition, indicating no dissolution of Zr in TaB.
The observed chemical interaction and solid solubility be-
tween Ta5Si3 and ZrB2 contradict the only existing literature
data,29 which reports that very little or no interaction is observed
even after heating at 22001C and 10ꢀ2 torr (1.33 Pa) for 1 h. This
dramatic discrepancy in these two studies is not easy to resolve.
The bulk densities of the samples were 6.61, 7.18, and 8.27
g/cm3 for ceramics containing 8, 15, and 31 vol% Ta5Si3,
respectively, with close-to-zero open porosity. (The relative den-
sities of the ceramics were not determined because of the un-
known quantitative phase composition of the final ceramics.)
Evidently, a decrease in the hot pressing temperature of the
Ta5Si3-containing materials compared with pure ZrB2 (22001C)
can be related to chemical interactions and solid solubility be-
tween the components.
(2) ZrB2–Ta5Si3 Ceramics
ZrB2 ceramics modified with 8, 15, and 31 vol% (2, 4, and 10
mol%) Ta5Si3 were prepared by hot pressing starting mixtures
at 1900 1C and 20 MPa for 30 min in a graphite mold and He
atmosphere. The XRD analysis revealed significant interaction
between ZrB2 and Ta5Si3 during hot pressing. The polymor-
phism of Ta5Si3, as well as the chemical interactions and solid
solubility in the entire ZrB2–Ta5Si3 system, is described else-
where.31 The phase compositions of the ceramics containing 2,
4, and 10 mol% are summarized in Table I.
The ceramics containing 2 and 4 mol% Ta5Si3 consisted of
stoichiometric ZrB2 and ZrB2-based solid solution with a con-
traction of the ZrB2 unit cell. The unit cell volume of ZrB2
3
˚
decreased linearly from 30.70 to 29.89 A as Ta5Si3 content in-
creased from 0 to 4 mol%. Based on the insolubility of Si and
ZrSix in ZrB2 reported in the literature34,36 and the contraction
observed in the ZrB2 lattice, it was deduced that Ta entered the
ZrB2 cell, causing the contraction (the atomic radii of Zr and Ta
The microstructure of the hot-pressed ZrB2/15 vol% Ta5Si3
sample (Fig. 5) is composed of large crystals (average size about
5–10 mm) containing predominantly Zr (by EDS) surrounded by
˚
are 1.60 and 1.46 A, respectively).
10 µm
Fig. 5. SEM of ZrB2/15 vol%/Ta5Si3 ceramics HP at 19001C.

2254
Journal of the American Ceramic Society—Talmy et al.
Vol. 91, No. 7
practically no Ta (from the size of the ‘‘far’’, ꢂ8 keV, Ta peak),
while the crystallized area contains Ta and a significant amount
of Zr. The strong peak belonging to both Si and Ta in the EDS
scan of the glassy area can be therefore associated predomi-
nantly with Si, and the peak from the crystallized area of the
scan is from a combination of both elements.
Thus, the improved oxidation resistance of the Ta5Si3-con-
taining compositions compared with pure ZrB2 is due to the
formation of an immiscible borosilicate glass during oxidation.
The test results of this study did not match the results of Pastor
and Meyer,29 who did not observe any improvement in the
oxidation resistance of ZrB2 by the addition of 15 wt% Ta5Si3
(the composition corresponds to 8 vol% Ta5Si3 in the present
study) after oxidation in air at 12001C for 100 h.
200
Scale Thickness
115 µm
1
2
ZrB
2
2
300 µm
160
120
80
8 v.% Ta Si
5
3
15 v.% Ta Si
5
3
31 v.% Ta Si
5
3
25 v.% SiC
1
40
In spite of the improvement in oxidation resistance of ZrB2
with the addition of up to 31 vol% Ta5Si3, ZrB2 ceramics con-
taining 25 vol% SiC have superior oxidation resistance at tem-
peratures above 13001C, as shown in Fig. 6. The enhanced
oxidation resistance of Ta5Si3-containing ceramics below
13001C can be attributed to the insufficient oxidation of SiC
at these temperatures to form a continuous, protective borosil-
icate glass, while significant oxidation of Ta5Si3 occurred below
11001C.31
0
1100
1200
Temperature (°C)
1300
1400
Fig. 6. Mass changes during furnace oxidation (2 h) of ZrB2 ceramics
containing 0, 8, 15, and 31 vol% Ta5Si3 at 11001–14001C.
fine-grained crystals (about 1 mm) consisting of both Zr and Ta,
which is an indication that fine ZrB2 particles preferentially par-
ticipated in the chemical interactions between the components.
The location of Si was difficult to identify because of the over-
lapping of Ta and Si peaks, but it can be assumed that it was
localized in the fine-grained phase.
Figure 6 shows the mass gain for ZrB2–Ta5Si3 ceramics dur-
ing 2 h oxidation at 11001–14001C. Materials containing Ta5Si3
exhibited increased oxidation resistance, which is further sup-
ported by measurements of the thickness of the oxidized layer
after oxidation at 14001C. The thickness of the oxidized layer
was 115 mm for ceramics containing 15 vol% Ta5Si3 compared
with 300 mm for pure ZrB2. The presence of ZrO2 and
TaZr2.75O8 was determined by XRD on the surface of the
oxidized samples. No crystalline silicon- or boron-containing
compounds were identified with certainty.
(3) ZrB2–TaSi2 Ceramics
ZrB2 ceramics containing 10, 20, and 30 vol% (approximately 7,
15, and 23 mol%, respectively) TaSi2 were prepared by hot
pressing at 20001C and 20 MPa for 20 min in a He atmosphere.
The bulk densities of the samples were 5.63, 6.44, and 6.73
g/cm3, respectively. The XRD patterns of the ceramics showed
solid solubility in the system (Fig. 8). The ceramics containing 10
vol% TaSi2 did not show the diffraction peaks of TaSi2. As the
amount of TaSi2 increased to 20 and 30 vol%, the TaSi2 peaks
became evident and increased in intensity with increasing TaSi2
concentration. The peaks were shifted to the left from the
position of the stoichiometric TaSi2, indicating lattice expan-
sion. The diffraction peaks of ZrB2 were shifted to the right of
the expected positions, indicating lattice contraction. The pres-
ence of the shoulder corresponding to the position of the stoic-
hiometric ZrB2 at 2y 5 41.61, on the left of the shifted ZrB2 peak
on the pattern of the material containing 20 and 30 vol% TaSi2,
probably indicates insufficient processing time to homogenize
the phase composition. The calculated unit cell volume
A periodic pattern of glassy and crystalline areas on the
oxidized surface of the ceramics containing 8 vol% Ta5Si3
(Fig. 7) implies glass phase separation, which was induced by
the presence of Ta51. In spite of the overlapping of major Ta
and Si peaks, EDS analysis of the surface leads to the conclusion
that the glassy area contains a very small amount of Zr and
3
˚
decreased for ZrB2 from 30.70 to 30.02 A and increased for
Fig. 7. SEM/EDS of the surface of ZrB2/8 vol% Ta5Si3 ceramics after oxidation at 13001C for 2 h.

July 2008
High-Temperature Chemistry and Oxidation of ZrB2-Based Ceramics
2255
The oxidation resistance of ZrB2–TaSi2 ceramics increased
with increasing TaSi2 content (Fig. 10). Samples containing 10
vol% TaSi2 showed higher mass gain than pure ZrB2 at all tem-
peratures. This behavior was also observed for the ZrB2/5 vol%
Si3N4 ceramics and was due to the significant evaporation of
B2O3 from pure ZrB2 above 12001C, decreasing the total weight
gain from the formation of ZrO2 and B2O3 during oxidation.
The ceramics containing 20 and 30 vol% TaSi2 exhibit signifi-
cantly less mass gain than ZrB2 up to 14001C. The matching
values of the mass gain for ZrB2 and ceramics with 30 vol%
TaSi2 at 15001C are again the result of extensive evaporation of
B2O3 from ZrB2 at this temperature. This is supported by the
thickness of the oxidized layer for ceramics containing 30 vol%
TaSi2 after oxidation at 14001 and 15001C (50 and 60 mm com-
pared with 300 and 600 mm for pure ZrB2, respectively). A small
increase in the thickness of the oxidized layer for the TaSi2-con-
taining material between 14001 and 15001C together with 5–10
times higher thickness for pure ZrB2 at these temperatures are
additional and conclusive indications of the protective capabil-
ities of Ta2O5-containing borosilicate glass. The SEM/EDS of
the oxidized surfaces of all the ZrB2–TaSi2 samples showed glass
immiscibility (similar to that for the ZrB2–Ta5Si3 ceramics) with
a Zr-containing glassy phase and the crystalline phase contain-
ing mostly Zr with a small amount of Ta. The XRD of the
oxidized surface of the samples showed the presence of ZrO2
and TaZr2.75O8. No Si- and B-containing crystalline phases were
identified.
Figure 11 shows the behavior of the most oxidation-resistant
compositions from each series of Si-containing additives char-
acterized in this research together with oxidation data previously
reported for ZrB2/25 vol% SiC modified with 10 mol% TaB2.16
An inset presents a magnified picture corresponding to the mass
gains at 12001–14001C. All the compositions exhibited improved
oxidation resistance at 12001 and 13001C compared with the
ZrB2-SiC ceramics. The improved oxidation resistance persists
with some of the compositions approaching 14001C, with the
best performance shown by the ZrB2/SiC/TaB2 material. As
discussed above, the lower oxidation resistance of the SiC-con-
taining ceramics below 13001–14001C compared with all other
materials can be attributed to the insufficient oxidation of SiC at
these temperatures to form a continuous protective borosilicate
glass, while the significant oxidation of ZrSi2, Ta5Si3, and TaSi2
occurred at much lower temperatures. The presence of transi-
tion-metal oxides in the borosilicate glass, inducing immiscibil-
ity, additionally increased glass-protective capability. The
diminishing effect of these oxides at temperatures above
Fig. 8. X-ray diffraction of ZrB2/10, 20, and 30 vol% TaSi2 ceramics
HP at 20001C for 20 min.
3
TaSi2 from 130.19 to 132.67 A in the ceramics containing 20 vol%
˚
(15 mol%) TaSi2, confirming the formation of solid solutions.
The solid solubility in the system was not expected and was
not previously reported in the literature. The crystal structures
of both ZrB2 and TaSi2 are hexagonal but belong to different
space groups (ZrB2 is AlB2-type (P6/mmm) and TaSi2 is CrSi2-
type (P6222)). Based on the results of Brukl’s research,34 which
excluded solubility of silicon or silicides in ZrB2, one way to
interpret the formation of solid solutions between ZrB2 and
TaSi2 is to consider that only the transition metals mutually dis-
solve, leaving the boride and silicide lattices intact. With atomic
radii of Zr and Ta being 0.160 and 0.146 nm, respectively, the
incorporation of Ta into the ZrB2 crystal lattice and Zr into the
TaSi2 crystal lattice had to result in the observed contraction of
the ZrB2 and expansion of the TaSi2 crystal lattices.
A mutual solubility of ZrB2 and TaSi2 is supported by the
results of SEM/EDS studies, which showed that the majority of
the crystals had a core–shell structure with the core being ZrB2
and the shell containing both Ta and Zr (illustrated on Fig. 9)
for ceramics with 30 vol% TaSi2. It can be assumed that the
simultaneous presence (Fig. 8) of the shifted ZrB2 peaks and the
stoichiometric ZrB2 peak reflects the core/shell structure of the
material. It is expected that an increase in holding time during
hot pressing would complete the formation of the solid solutions
and eliminate the core–shell structure of the crystals.
Fig. 9. SEM/EDS of ZrB2/30 vol% TaSi2 ceramics HP at 20001C for 20 min, BS image.

2256
Journal of the American Ceramic Society—Talmy et al.
Vol. 91, No. 7
performed best, but was only slightly better than the
ZrB2/SiC/TaB2 and ZrB2/Si3N4/CrB2 ceramics.
500
400
300
200
100
0
Scale Thickness
ZrB2
1
2
3
4
50 µm
300 µm
60 µm
600 µm
10 v% TaSi2
20 v% TaSi2
30 v% TaSi2
ZrB2 / 25 v% SiC
Acknowledgments
The authors would like to acknowledge ONR (Drs. S. Fishman and D. Shifler)
for providing support for this project and Dr. B. Varshal for valuable discussions
on phase separation in glasses.
4
References
1M. M. Opeka, I. G. Talmy, and J. A. Zaykoski, ‘‘Oxidation-based Materials
Selection for 20001C1 Hypersonic Aerosurfaces: Theoretical Considerations and
Historical Experience: Special Section: Ultra-High Temperature Ceramics (Guest
Editors: Joan Fuller and Michael D. Sacks),’’ J. Mater. Sci., 39 [19] 5887–904
(2004).
2
3
2W. G. Fahrenholtz, G. E. Hilmas, I. G. Talmy, and J. A. Zaykoski, ‘‘Refrac-
tory Diborides of Zirconium and Hafnium,’’ J. Am. Ceram. Soc., 90 [5] 1347–64
(2007).
1
1200
1300
1400
1500
3R. E. Loehman, ‘‘Ultrahigh-Temperature Ceramics For Hypersonic Vehicle
Applications,’’ Ind. Heating, 71 [1] 36–8 (2004).
Temperature (°C)
4S. R. Levine, E. J. Opila, M. C. Halbig, J. D. Kiser, M. Singh, and J. A. Salem,
‘‘Evaluation of Ultra-High Temperature Ceramics for Aeropropulsion Use,’’
J. Eur. Ceram. Soc., 22 [14/15] 2757–67 (2002).
Fig. 10. Mass changes during furnace oxidation (2 h) of ZrB2 ceramics
containing 10, 20, and 30 vol% TaSi2 at 12001–15001C.
5J. R. Fenter, ‘‘Refractory Diborides as Engineering Materials,’’ SAMPE
Quart., 2 [3] 1–15 (1971).
6E. V. Clougherty, D. Kalish, and E. T. Peters, ‘‘Research and Development
Oxidation Resistant Diborides’’; Technical Report AFML-TR-68-190, Manlabs
Inc., Cambridge, MA 1968
60
40
20
ZrB
300
200
100
0
25 v % SiC
7A. L. Chamberlain, W. G. Fahrenholtz, G. E. Hilmas, and D. T. Ellerby,
‘‘Characterization of Zirconium Diboride for Thermal Protection Systems,’’ Key
Eng. Mater., 264–268, 493–6 (2004).
15 v % Ta Si
35 v % Si N
1200
1300
1400
8I. G. Talmy and J. A. Zaykoski, ‘‘Properties of Ceramics in the ZrB2/ZrC/SiC
System Prepared by Reactive Processing,’’ Ceram. Eng. Sci. Proc., 19 [3] 105–12
(1998).
9W. C. Tripp, H. H. Davis, and H. C. Graham, ‘‘Effect of SiC Additions on the
Oxidation of ZrB2,’’ Ceram. Bull., 52 [8] 612–6 (1973).
30 v % TaSi
20 v % Si N
10J. W. Hinze, W. C. Tripp, and H. C. Graham, ‘‘High-Temperature Oxidation
of a HfB2 Composite,’’ J. Electrochem. Soc., 122 [9] 1249–54 (1975).
11E. V. Clougherty, R. L. Pober, and L. Kaufman, ‘‘Synthesis of Oxidation
Resistant Metal Diboride Composites,’’ Trans. Metals Soc. AIME, 242, 1077–82
(1968).
/
10 mole % CrB
25 v % SiC/
10 mole % TaB
1200
1300
1400
1500
12R. F. Voitovich, E. A. Pugach, and L. A. Menshikova, ‘‘High Temperature
Oxidation of Zirconium Diboride,’’ Sov. Powd. Met., 7 [6] 44–8 (1967).
13F. Monteverde and A. Bellosi, ‘‘The Resistance to Oxidation of an HfB2–SiC
Composite,’’ J. Eur. Ceram. Soc., 25 [7] 1025–31 (2005).
Temperature (°C)
Fig. 11. Mass changes during furnace oxidation (2 h) of ZrB2 ceramics
containing SiC, Si3N4, TaSi2, and Ta5Si3 at 12001–15001C.
14F. Monteverde and A. Bellosi, ‘‘Production and Characterization of Ultra
Refractory HfB2–SiC Composites,’’ Ceram. Eng. Sci. Proc., 25 [4] 13–8 (2004).
15W. G. Fahrenholtz, G. E. Hilmas, A. L. Chamberlain, and J. W. Zimmerm-
ann, ‘‘Processing and Characterization of ZrB2-Based Ultra-High Temperature
Monolithic and Fibrous Monolithic Ceramics,’’ J. Mater. Sci., 39 [19] 5951–7
(2004).
14001C and the increase in oxidation rates could be related to a
decrease in glass viscosity, which leads to the reduction of
immiscibility effects and an increase in oxygen diffusivity
through the glass layer.
16I. Talmy, J. Zaykoski, M. Opeka, and S. Dallek, ‘‘Oxidation Behavior of ZrB2
Ceramics Modified with SiC and Group IV–VI Transition Metal Diborides’’; pp.
144–158, Proceedings of the International Symposium on High Temperature Cor-
rosion and Materials Chemistry III, 2001-12, Edited by M. McNallan, and E.
Opila. The Electrochemical Society Inc., Pennington, NJ, 2001.
17L. Kaufman and E. Clougherty, ‘‘Investigation of Boride Compounds for
Very High Temperature Applications’’; Report RTD-TDR-63-4096, Part 1,
AFML, WPAFB, OH, Dec. 1963.
IV. Summary
Silicon carbide is conventionally used to improve the mechanical
properties and oxidation resistance of ZrB2 ceramics. The effect
of SiC on the oxidation resistance of ZrB2 is due to the forma-
tion of protective surface borosilicate glass, which enhances
oxidation protection by acting as a barrier to the inward diffu-
sion of oxygen. The effect of Si3N4, Ta5Si3, and TaSi2 (as sourc-
es of Si instead of SiC) on the phase composition,
microstructure, and oxidation behavior of ZrB2 ceramics was
characterized. Silicon nitride decomposed during the processing
of ZrB2–Si3N4 compositions, and the final ceramics contained
ZrB2, BN, ZrSi2, and ZrN. Oxidation resistance of Si3N4-mod-
ified ceramics increased with increasing Si3N4 content and was
further improved by the addition of Cr and Ta diborides due to
phase separation of the surface glass. The ZrB2/Ta5Si3 materials
exhibited complex chemical interactions and solid solubility
between the components during hot pressing. Limited solid
solubility was also observed between ZrB2 and TaSi2. Both Ta
silicides significantly improved oxidation resistance of pure ZrB2
ceramics, with the best performance exhibited by ZrB2 contain-
ing 15 vol% Ta5Si3 and 30 vol% TaSi2. The effect is attributed
to the pronounced immiscibility in the surface glass formed
during the oxidation tests. All studied compositions exhibited
improved oxidation resistance below 14001C compared with
ZrB2/SiC. At 15001C, the SiC-modified ZrB2 composition
18I. Talmy, J. Zaykoski, and M. Opeka Unpublished data
19W. Vogel, Glass Chemistry, 2nd edition, Springer-Verlag, New York, (1994).
20F. P. Glasser, I. Warshaw, and R. Roy, ‘‘Liquid Immiscibility in Silicate Sys-
tems,’’ Phys. Chem. Glasses, l, 39–45 (1960).
21B. G. Varshal, ‘‘A Structural Model for Immiscibility in Silicate Glass-Form-
ing Melts,’’ Glass Phys. Chem., 19 [2] 107–10 (1993).
22B. G. Varshal (ed.), Two-Phase Glasses: Structure, Properties, and Applica-
tions. AN SSSR, Nauka, Leningrad, 11–33, (1991).
23A. Bellosi and F. Monteverde, ‘‘Ultra-Refractory Ceramics: The Use of Sinte-
ring Aids to Obtain Microstructure Control and Properties Improvement,’’ Key
Eng. Mater., 264–268 [Part 2] 787–92 (2004).
24K. Sato, H. Kakisawa, Y. Kagawa, H. Kaya, O. Funayama, and T. Isoda,
‘‘Reaction between Newly-Developed Si3N4 Fiber and Various Ceramics,’’ J.
Mater. Sci. Lett., 19 [13] 1179–83 (2000).
25P. Shaffer, ‘‘An Oxidation Resistant Boride Composition,’’ Ceram. Bull., 41
[2] 96–9 (1962).
26P. Shaffer, ‘‘Oxidation-Resistant Ceramics and Method of Manufacturing
Same’’; U.S. Patent 3, 189, 477 (1965)
27V. A. Lavrenko, A. D. Panasyuk, T. G. Protsenko, V. P. Dyatel, E. S.
Lugovskaya, and E. I. Egorova, ‘‘High-Temperature Reactions of Materials in the
ZrB2-ZrSi2 System with Oxygen,’’ Sov. Powder Metal Metal Ceram., 21 [6] 471–3
(1982).
28A. D. Panasyuk, L. A. Glebov, T. G. Protsenko, K. M. Timofeeva, and V. P.
Dyatel, ‘‘High-Temperature Oxidation of Materials in the System ZrB2–CrSi2’’;
pp. 127–30 in Vysokotemp. Boridy Silitsidy, Edited by T. Ya. Kosolapova. Nayk-
ova Dumka, Kiev, USSR, 1978.
29H. Pastor and R. Meyer, ‘‘An Investigation of the Effect of Additions of Metal
Silicides on Titanium and Zirconium Borides from the Point of View of their

July 2008
High-Temperature Chemistry and Oxidation of ZrB2-Based Ceramics
2257
Sintering Behavior and their Resistance to Oxidation at High Temperatures,’’ Rev.
Int. Hautes Temp. Refract., 11 [1] 41–54 (1974).
34C. E. Brukl, ‘‘Ternary Phase Equilibrium in Transition Metal–Boron–Car-
bon–Silicon Systems. Part II. Ternary Systems, Volume X’’; AFML-TR-65-2, Part
II, Vol. X, Air Force Materials Laboratory, Wright Patterson Air Force Base,
Dayton, OH, November 1966.
30E. Opila, S. Levine, and J. Lorincz, ‘‘Oxidation of ZrB2- and HfB2-Based Ul-
tra-High Temperature Ceramics: Effect of Ta Additions,’’ J. Mater. Sci., 39 [19]
5969–77 (2004).
35R. Kieffer and F. Benesovsky, ‘‘Recent Developments in the Field of Silicides
and Borides of High-Melting-Point Transition Metals,’’ Powder Metal., [1/2] 145–
71 (1958).
31I. Talmy, J. Zaykoski, M. Opeka, and A. Smith, ‘‘Properties of Ceramics in
the System ZrB2–Ta5Si3,’’ J. Mater. Res., 21 [10] 2593–9 (2006).
32T. J. Rocket and W. R. Foster, ‘‘Phase Relations in the System Boron Oxide–
Silica,’’ J. Am. Ceram. Soc., 48 [2] 78 (1965).
36E. Parthe and J. T. Norton, ‘‘Phase Equilibria in M–X–X0 and M–Al–X Ter-
nary Systems (M5 Transition Metal; X, X0 5 B, C, N, Si) and the Crystal Chem-
33S. S. Ordan’yan, A. I. Dmitriev, and E. S. Moroshkina, ‘‘Reaction of SiC with
ZrB2,’’ Inorg. Mater., 25 [10] 1487–9 (1989).
istry of Ternary Compounds,’’ Monatsh. Chem., 91, 1127 (1960).
&