A Novel Preparation Method of (Ti,Zr,Nb,Mo,W)B₂-SiC Composite Ceramic Based on Reactive Sintering of Pre-Alloyed Metals

Abstract

High-entropy diboride-based (MeB₂-based) ceramics are promising high-temperature structural materials because of their excellent mechanical properties, high-temperature stability, and oxidation resistance. In order to achieve low-temperature sintering of the high-entropy ceramics, a novel preparation method of high-entropy (Ti,Zr,Nb,Mo,W)B₂-SiC ceramics based on reactive sintering of pre-alloyed solid-solution metals and nonmetals of Si, C, B₄C was conducted in the current work. Mechanical alloying behavior of the mixed metal powders, as well as the phase composition, microstructure, mechanical properties, and oxidation behavior of the as-sintered MeB₂-SiC ceramic were investigated. The XRD, SEM, and EPMA results indicated that the primary MeB₂ solid-solution and SiC phases could be successfully formed during reactive sintering at a relatively low temperature of 1650 °C. The as-sintered MeB₂-SiC ceramics had a high relative density of 97.8% and high mechanical properties (hardness of 19.74 ± 0.8 GPa, flexure strength of 533 ± 38 MPa, and fracture toughness of 6.01 ± 0.77 MPa·m^1/2). Combining the oxidation behavior and microstructure evolution of the oxidation layer, a continuous and relatively dense MeO_x-SiO₂ oxidation layer was gradually formed and covered on the external surface, leading to decelerating oxidation behavior after an oxidation exposure time of 10 min.

1. Introduction

With the rapid development of aerospace technology, the demand for materials suitable for elevated temperature applications has generated significant interest in ultra-high temperature ceramics (UHTCs). In order to better meet their application under harsh environments, further improving mechanical, physical, and chemical performances is an important development direction for UHTCs [1,2,3,4]. Due to their ultra-high melting point, high mechanical properties, and excellent thermal and chemical stability, UHTC-based composites of diboride-silicon carbide have been considered as potential thermal-structural materials, such as ZrB₂-SiC, HfB₂-SiC, and TiB₂-SiC.

Since high-entropy alloys (HEAs) with superior mechanical and physical properties were reported by Yeh [5] and Cantor [6] in 2004, the design concept of high-entropy materials (HEM) has been considered as a new strategy to further improve the materials’ performance. In 2015, the HEM design concept was first applied to ceramic materials [7]. Since then, high-entropy ceramics, including high-entropy oxide [7,8,9], carbide [10,11], boride [12], nitride [13,14], fluoride [15], sulfides [16], and silicide [17,18], have attracted great interest as a new kind of ceramic material system with great potential. Among them, high-entropy diboride ceramic (MeB₂) has both advantages of HEM and UHTC, which are more suitable for high-temperature applications under harsh environments [2]. In general, MeB₂ is a hexagonal-structured solid-solution with multi-principal elements in the cation position. The reported research indicated that the hardness and oxidation resistance of some MeB₂ was higher than individual diboride ceramics [12]. Therefore, MeB₂-SiC composite ceramic is expected to be an ideal ultra-high-temperature thermal-structural material.

Based on the various starting materials, the preparation methods of MeB₂ followed four principal routes: (1) direct sintering combined with dissolving five individual metal diborides into solid-solution; (2) densification of pre-synthesized MeB₂ powder; (3) boro/carbothermal reduction of metal oxide, carbon, and B₄C powders; (4) borothermal reaction through different metals and B. For the route (1), Gild et al. prepared six different MeB₂ with a relative density of 92% via spark plasma sintering (SPS) at 2000 °C [12]. Li et al. prepared a (Zr_0.2Ta_0.2Nb_0.2Hf_0.2Mo_0.2)B₂ ceramic by oscillatory pressure sintering (OPS) at 1800–1900 °C; additionally, the single-phase solid-solution can be formed at 1900 °C [19]. For the route (2), the pre-synthesized MeB₂ powders can be densified at high temperatures (>1900 °C), while some residual oxides [20,21] or other secondary phases [22] appeared in the final ceramics. For the routes (3) and (4), in order to achieve a complete chemical reaction, the synthesis of MeB₂ always requires temperatures greater than 1950 °C [23,24]. In short, due to the limited diffusion coefficient of the diboride ceramics, the formation of MeB₂ solid-solution usually needs an ultra-high temperature (>1900 °C) and long holding time. In order to reduce the fabricating cost of MeB₂, a new preparation strategy that can prepare MeB₂ at a relatively lower temperature needs to be developed.

In this paper, for preparing high-entropy MeB₂-SiC ceramics at relatively low temperatures, a novel preparation method based on reactive sintering of pre-alloyed metals (Me) and nonmetals is proposed. According to the experimental results, the primary MeB₂ solid-solution and SiC can be formed at 1650 °C. Moreover, the alloying behavior of metals, as well as microstructure, phase composition, and elemental distribution of the as-sintered MeB₂-SiC were investigated. Finally, the oxidation behavior and microstructure evolution of MeB₂-SiC ceramics at 1400 °C was studied.

2. Materials and Experimental Procedures

2.1. Materials Preparation

Experiments were performed, starting with commercially-available Ti (Aladdin, Shanghai, China, particle size ≥ 48 μm, 99.99% purity), Zr (Aladdin, Shanghai, China, particle size ≥ 75 μm, 99.5% purity), Nb (Aladdin, Shanghai, China, particle size 25 μm, 99.95% purity), Mo (Aladdin, Shanghai, China, particle size 25 μm, 99.5% purity), W (Aladdin, Shanghai, China, particle size 45 μm, 99.99% purity), Si (Aladdin, Shanghai, China, particle size 75–380 μm, 99.9% purity), B₄C (Aladdin, Shanghai, China, particle size 2–3 μm, 99.8% purity), and C (Aladdin, Shanghai, China, particle size 40 nm, 99% purity) powders.

2.2. Fabrication of MeB₂-SiC

The detailed fabrication procedures of MeB₂-SiC are described as follows, and schematically represented in Figure 1. The raw metals of Ti, Zr, Nb, Mo, and W were mechanically alloyed by high-energy ball milling. Equimolar Ti, Zr, Nb, Mo, and W were milled at 350 rpm for 48 h. Tungsten carbide (WC) balls (diameter of Φ5 and Φ8 mm) and WC pots were selected as the ball-milling media. The milling process was performed in a protective argon atmosphere and the mass ratio of WC ball and powder was fixed at 10:1. Then, Si, B₄C, and C powders and 30 mL of anhydrous ethanol were added to the ball-mill pot and mixed with the pre-alloyed powders (Me) for 12 h. Anhydrous ethanol was used as the process control agent. The molar ratio of Me, Si, B₄C, and C was 2:4:1.1:3. The nominal chemical reaction was:

2Me_(s) + B₄C_(s) + 3C_(s) + 4Si_(s) = 2MeB_2(s) + 4SiC_(s)

(1)

Due to the inability to completely isolate the air during the experimental process, slight amounts of oxides are unavoidably present on the surface of the metal powders. The addition of a slight excess amount of B₄C (ten mole percent) is typically necessary [25,26]. The reaction between carbide and oxide according to this reaction was:

MeO_x(cr) + B₄C_(cr) → MeB_2(cr) + CO_(g) + B₂O_3(l)

(2)

Afterwards, the milled alloyed-Me, Si, B₄C, and C powders were dried and grinded. Ultimately, these powders were reactive hot-pressed (HP) at 1650 °C for 30 min, with the heating rate of 10 °C/min under a pressure of 30 MPa in vacuum, followed by furnace cooling to room temperature.

The empirical criteria parameters were used to predict whether the five metals would form a stable solid-solution [27,28,29]. The empirical criteria parameters were the mixing enthalpy ΔH_mix, related to the chemical affinity between the elements; the size parameter δ, linked to the atomic radii differences; the parameter Ω, which describes the relative contributions of the mixing enthalpy and the mixing entropy (ΔS_mix) in the Gibbs free energy; and Valence electron concentration VEC, a useful parameter for predicting the stability of solid-solution phases. The empirical parameters of this experimental composition perfectly match the requirements for a BCC solid-solution formation: −15 kJ/mol < ΔH_mix (−14.4) < 5 kJ/mol, Ω(2.492) < 6.6, δ(5.545) < 6%, and VEC(5) < 6.87. The parameters of the atomic radius and calculation process of empirical criterion parameters used in this paper are summarized in Supplement Tables S1 and S2, as well.

Thermodynamic calculations for the reactions between Me, Si, B₄C, and C were performed using HSC 9 software. Due to the lack of thermodynamic data for the solid-solution of alloyed Me, the individual metals of Ti, Zr, Nb, Mo, and W were used instead of Me. The thermodynamic analysis of each metal in Me with Si, B₄C, and C were performed in supplement (reaction S1–S5), and the corresponding results are shown in Figures S1–S5. The results indicated that the Gibbs free energy of the related reactions are negative at the sintering temperature; therefore, the reactions could occur from the perspective of thermodynamics.

2.3. Oxidation Tests

In order to study the oxidation processes of MeB₂-SiC, oxidation tests were performed using the samples of 4 mm × 5 mm × 2 mm, cut from the as-sintered MeB₂-SiC bulk. The MeB₂-SiC was oxidated in a tube furnace (GSL-1800X, Hefei Kejing Materials Technology Co., Ltd., Hefei, China). The sample containing crucible was pushed into the furnace and heated to 1400 °C (10 °C/min), in an air atmosphere. After being oxidized in air at 1400 °C for the prescribed time and then being allowed to cool, the samples were reweighed to determine their mass gain due to the uptake of oxygen. The oxidation prescribed times of the samples at 1400 °C were 0, 5, 10, and 20 min.

2.4. Characterizations

Phase composition of the individual metals, alloyed-Me powders, and as-sintered MeB₂-SiC bulk sample was characterized using an X-ray diffractometer (XRD, Shimadzu XRD-6000, Tokyo, Japan) with Cu-Kα radiation. Microstructures and elemental distribution of the as-received and oxidized samples were observed with a scanning electron microscope (SEM, NOVA NanoSEM 450, New York, NY, USA) and electron probe micro-analyzer (EPMA, JEOL JXA-8530F PLUS, Tokyo, Japan). Flexural strength was tested under four-point loading condition. Specimens with a size of 2 mm in width, 1.5 mm in thickness, and a length of at least 25 mm were used. The sample holder had a span between the two bearers of 20 mm. The distance between the two loading pistons was 10 mm. Supports and both loading pistons had steel knife edges, rounded to a radius of 1.5 mm. The flexural strength was calculated according to the equation:

$$\sigma =\frac{3Fd}{2b{h}^2}$$

(3)

where σ is the maximum center tensile stress (MPa), F is the load at fracture (N), d is the difference in the distance of the two supports and the distance of the two loading pistons (mm), b is the width of the specimen (mm), and h is the height of the specimen (mm). Fracture toughness was tested using the single edge notched beam (SENB) method under three-point loading condition. The specimens (2 mm × 4 mm × 25 mm) were used to measure fracture toughness. A blade was positioned in the center of the specimen to produce a 0.25 mm length notch. The fracture toughness was calculated according to the following equation:

$$K_{\mathrm{Ic} }=\left[\frac{PS}{B{W}^{1.5}}\right]f\left(\frac{a}{w}\right)$$

(4)

where P is the load of fracture (N), S is the spam between supports, B is the length, and W is the width (m). f(a/w) was calculated according to the standard test method [30]. Vickers microhardness (HVS-30, Shanghai Biman Instrument Co., Shanghai, China) was used to measure the hardness by indentation under a load of 9.8 N (Hv 1), with a dwelling time of 10 s. All indentations were carried out randomly during the hardness measurement. The distance between two neighboring indentations was kept at least four times the size of the indent, and an average of eight indentations were used as a hardness value.

3. Results

3.1. Pre-Alloying Behavior of Multiple Metallic Elements

Figure 2 shows the XRD patterns of the Ti-Zr-Nb-W-Mo mixture at different milling times. As can be seen in Figure 2, the diffraction peaks of the five individual metals (Ti, Zr, Nb, Mo, and W) were clearly detected before ball milling. Then, with the milling time increased, the peak intensity of the individual metals gradually decreased. Especially after 24 h of ball-milling, some diffraction peaks of the individual metals, e.g., Ti and Zr, even become invisible, and meanwhile, the obvious broadening of the residual peaks can also be observed. Such phenomena indicated that the various metals were forming solid-solution. During the ball milling process, the metallic powders were collided and squeezed by the WC grinding balls in the milling tank, which resulted in severe plastic deformation, fracture, and cold welding in the metallic powders. During this process, the powders were easily formed to a BCC solid-solution, due to the shorter diffusion distance and larger contact area. As the milling time reached up to 48 h, only three major diffraction peaks of the BCC solid-solution were detected, those being 2θ = 40.14°, 2θ = 58.02°, and 2θ = 73.08°, which correspond to crystal planes {110}, {200}, and {211}. Due to the lack of symmetry in the peak of {110}, the software XPSPEAK4.1 was used to resolve the issue. The results indicate that this diffraction peak may be a superposition of two peaks (2θ = 38.5° and 2θ = 40.14°), as shown in the Supplementary Figure S6. According to the standard card, the diffraction peaks should be BCC solid-solution and a slight amount of Nb (PDF # 89-5291). Although a minor amount of unsolidified undissolved Nb is present, the pre-alloying of the five individual metallic elements was basically completed. Therefore, the pre-alloying metallic powders milled at 48 h were selected for further high-entropy diboride ceramic preparation.

3.2. Phase Composition and Microstructure of the As-Sintered MeB₂-SiC Ceramic

The pre-alloying metallic powders and the nonmetallic raw materials of Si, C, and B₄C were mixed, and then reactive sintered at 1650 °C by the hot-pressing method. The measured density and calculated theoretical density are 4.857 g/cm³ and 4.966 g/cm³, respectively. Thus, the relative density of the as-sintered MeB₂-SiC ceramic is 97.8%. Figure 3 shows the XRD pattern of the MeB₂-SiC high-entropy diboride-based ceramics prepared by HP sintering. As can be seen in Figure 3, the diffraction peaks of the primary MeB₂ solid-solution, SiC, and undesired WB₂ phases were detected.

Table 1. Crystal structures and parameters of the five individual diborides and the MeB₂ solid-solution.

Diborides	Pearson Symbol	Space Group	a = b (Å)	c (Å)	Standard Card (PDF #)
TiB2	hP3	P6/mmm (191)	3.036	3.238	65-8698
ZrB2	hP3	P6/mmm (191)	3.165	3.530	89-3930
NbB2	hP3	P6/mmm (191)	3.102	3.321	75-1048
MoB2	hP3	P6/mmm (191) *	3.005	3.173	86-1097
	hR6	R-3m (166)	3.015	20.971	73-0704
WB2	hP3	P6/mmm (191) *	3.020	3.050	89-3928
	hP12	P63/mmc (194)	2.983	13.879	43-1386
Calculation of MeB2 by average method	hP3	P6/mmm (191)	3.066	3.262	This work
Experimental results of MeB2 from XRD	hP3	P6/mmm (191)	3.078	3.384	This work

* Unstable phase at ambient temperature.

shows the space group and crystal parameter of the five individual MeB₂ phases (ZrB₂, NbB₂, TiB₂, MoB₂, and WB₂) reported in the standard PDF cards and the primary MeB₂ solid-solution phase detected in this study. Compared to the individual MeB₂ phases, the primary MeB₂ solid-solution phase detected in this study has the same space group (P6/mmm (191)) and a similar crystal parameter (a = b = 3.078 Å, c = 3.384 Å), which are close to the average crystal parameter (a = b = 3.066 Å, c = 3.262 Å). The average crystal parameter of MeB₂ is averaged from the individual diboride of the P6/mmm (191) space group. It should be noted that besides primary MeB₂ solid-solution and SiC phases, a minor amount of WB₂ phase with a P6₃/mmc (194) crystal structure was also detected in the XRD pattern (Figure 3). As the previous studies reported [31,32,33], for WB₂ ceramics, the P6₃/mmc (194) crystal structure is more stable than the P6/mmm (191) crystal structure at an ambient condition. Generally, a WB₂ ceramic with a P6/mmm (191) crystal structure is thermodynamically unstable and difficult to synthesize [34,35]. Therefore, in the current study, massive WB₂ is difficult to dissolve completely into the MeB₂ solid-solution phase with a P6/mmm (191) crystal structure, and consequently, minor WB₂ with a P6₃/mmc (194) crystal structure was precipitated in the as-sintered MeB₂-SiC ceramic.

In order to further understanding of the microstructure of the as-sintered MeB₂-SiC ceramic, the typical morphology and elemental distribution on the polished surface are shown in Figure 4 and Figure 5, respectively. The ceramic consists of three distinctly different phases, named gray, black, and bright phases. As can be seen in the low-magnification morphology (Figure 4a), gray and black phases are the main phases, which were well dispersed with each other. Moreover, the bright phase has an extremely low volume fraction, indicating that it is not the major product after the reactive sintering. The elemental composition of each phase was determined by EPMA mapping (Figure 5) and spot (

Table 2. Chemical composition (at.%) of the three phases measured by EPMA spot analysis in the as-sintered MeB₂-SiC ceramic.

Phases	Ti	Zr	Nb	Mo	W	B	C	Si
MeB2 (Gray)	6.39	6.43	6.65	5.61	5.08	69.84	-	-
SiC (Black)	-	-	-	-	-	-	53.89	46.11
WB2 (White)	-	-	-	12.66	36.77	50.57	-	-

) analysis. Figure 5a,b show the BSE and SE2 images, while Figure 5c–j are the individual elemental distribution maps of Ti, Zr, Mo, Nb, W, B, C, and Si, respectively. The gray phase is enriched with Ti (6.39%), Zr (6.43%), Nb (6.65%), Mo (5.61%), W (5.08%), and B (69.84%). Combined with XRD, SEM, and EPMA results, the gray phase is the primary MeB₂ solid-solution phase. For the black phase, it is enriched with Si (46.11%) and C (53.89%). The atomic ratio of Si:C is approximately equal to 1:1, indicating that the black phase is identified as the primary SiC phase. For the minor amount of bright phase, it is enriched of W (36.77%), B (50.57%), and Mo (12.66%). Combined with XRD, SEM, and EPMA results, it is probably a WB₂ phase with P6₃/mmc (194) crystal structure, which contains some Mo solute.

3.3. Mechanical and Oxidation Properties of the MeB₂-SiC Ceramic

The mechanical properties of the as-sintered MeB₂-SiC ceramic are shown in

Table 3. Summary of mechanical properties of the as-sintered high-entropy MeB₂-SiC ceramics and other similarly reported ceramics.

Material	Relative Density (%)	Vickers Hardness (GPa)	Flexure Strength (MPa)	Fracture Toughness (MPa·m½)	Ref.
(Ti,Zr,Nb,Mo,W)B2-SiC	97.8	19.74 ± 0.8	533 ± 38	6.01 ± 0.77	This work
(Ti0.2Zr0.2Hf0.2Nb0.2Ta0.2)B2	99.8	22.0 ± 0.9	339 ± 17	3.81 ± 0.40	[37]
(Hf0.2Zr0.2Ta0.2Nb0.2Ti0.2)B2	96.3	21.7  ±  1.1	-	4.06  ±  0.35	[38]
(Ti0.2Zr0.2Hf0.2Nb0.2Ta0.2)B2-SiC	99	22.4 ± 0.7	447 ± 45	4.85 ± 0.33	[37]
(Hf0.2Zr0.2Mo0.2Nb0.2Ti0.2)B2	98.1	26.3  ±  1.8	-	3.64  ±  0.36	[38]
(Hf0.2Mo0.2Ta0.2Nb0.2Ti0.2)B2	98.5	27.0  ±  0.4	-	4.47  ±  0.40	[38]
(Ti0.2Zr0.2Hf0.2Ta0.2Mo0.2)B2	92.4	19.1 ± 1.8	-	-	[12]
(Ti0.2Zr0.2Hf0.2Ta0.2Cr0.2)B2	92.2	19.9 ± 2.6	-	-	[12]
TiB2-SiC	98.1–98.7	23.7–23.8	349–601	4.8–5.3	[39]
ZrB2–SiC	99.1	20–21	-	3.8  ±  0.1	[40]
NbB2–SiC	97.5	17.9–18.9	-	4.16–4.62	[41]
MoB2–SiC *	-	-	-	-	-
WB2–SiC *	-	-	-	-	-

* Unstable phase at ambient temperature.

. The hardness, flexure strength, and fracture toughness were 19.74 ± 0.8 GPa, 533 ± 38 MPa, and 6.01 ± 0.77 MPa·m^1/2, respectively.

Table 4. Vickers hardness of metal diborides, SiC and MeB₂-SiC.

Materials	Vickers Hardness (GPa)	Ref.
TiB2 (191)	23.6	[42]
ZrB2 (191)	22.0	[43]
NbB2 (191)	20.25	[44]
MoB2 (191)	15.2	[45]
WB2 (191)	11.5	[36]
Calculated value of MeB2 (191) *	18.6	/
SiC	23.0	[46]
WB2 (194)	26.8	[33]
Calculated value of MeB2-SiC *	21.26	/
Measured value of MeB2-SiC	19.74 ± 0.8	This work

* Estimated by the rule-of-mixture law.

presents the Vickers hardness of the five individual diborides (TiB₂, ZrB₂, NbB₂, MoB₂, and WB₂) and β-SiC reported in the previous studies. According to the rule-of-mixture law, the theoretical hardness of the MeB₂-SiC ceramics was estimated as 21.26 GPa. Generally, due to solid-solution strengthening, the hardness of the high-entropy ceramic is usually slightly higher than the estimated one calculated by rule-of-mixture law. Moreover, a minor amount of P6₃/mmc (194)-WB₂ grains with high-hardness (26.8 GPa) [36] were precipitated in the as-sintered MeB₂-SiC ceramic, which may also increase the hardness of the ceramic. However, in the current study, the measured hardness (19.74 ± 0.8 GPa) is slightly lower than the estimated one (21.26 GPa). This may be caused by the as-sintered MeB₂-SiC ceramic not being dense enough. Some mechanical properties of the reported diborides are also listed in Table 3. The hardness value of MeB₂-SiC in this work is in the same range of individual metal diborides and some high-entropy diborides reported in the previous literature (17.9–23.8 GPa). However, there are also similar materials with higher hardness (26.3–27 GPa). The difference in hardness is not only due to the difference in density, but also to the elemental composition of the material. Therefore, in order to further improve the mechanical properties, the optimization of elemental composition and processing parameters need to be investigated in further works.

Isothermal oxidation tests at 1400 °C with different exposure times were conducted in order to investigate the oxidation property of as-sintered MeB₂-SiC ceramics. The mass change per unit surface area (Δm/S) with respect to oxidation exposure time is presented in Figure 6. The results depicted in Figure 6 show that the Δm/S increased rapidly before 10 min, and then increased slowly, indicating an oxide passivation stage after 10 min. In order to study the oxidation process, the oxidation behavior was discussed combining with the microstructure evolution of the oxidation layer. Figure 7 shows the surface (Figure 7a,c,e,g) and cross-sectional (Figure 7b,d,f,h) morphologies of the MeB₂-SiC ceramic after oxidation with different exposure times. As can be seen in Figure 7, with the increase of exposure time, the oxidation layer gradually thickened, meanwhile the oxidated surface morphologies changed significantly. When the oxidation time is 0 min, the oxidated surface is rough and porous, as shown in Figure 7a,b. During the heating process, the oxidation of MeB₂ preceded that of SiC; therefore, since the temperature had just risen to 1400 °C, the oxidation products of B₂O₃ and MeO_x were more than the oxidation products of SiO₂. Due to the evaporation temperature of B₂O₃ being lower than 1400 °C, a massive amount of micron pores appeared on the oxidated surface because of the volatilization of B₂O₃. With the oxidation exposure time increasing to 5 min, the oxidation product of B₂O₃ continued to volatilize, leading to lots of large-size pores appearing on the oxidated surface (Figure 7c,d). As the exposure time continued to increase, the SiO₂ oxidation products gradually accumulated on the oxidated surface. Due to the viscous flow characteristic of SiO₂ at 1400 °C, massive accumulated SiO₂ filled the pores. This made the oxidation products become a continuous and relatively dense SiO₂-MeO_x oxidation layer to cover the external surface of the MeB₂-SiC ceramic, as shown in Figure 7e,f (10 min) and Figure 7g,h (20 min). Such a continuous and relatively dense oxidation layer could prevent a further oxidation process, which leads to decelerating oxidation behavior after the exposure time of 10 min.

Figure 8 shows the elemental distribution mapping results on the cross-sectional of the MeB₂-SiC ceramic after oxidation at 1400 °C for 20 min. As can be seen in Figure 8, the external region II was enriched with O, Si, and five individual metallic elements, implying that the oxidation layer is mainly composed of SiO₂ and MeO_x. In this region, the dense SiO₂-MeO_x oxidation layer played a dominant role in the oxidation resistance of MeB₂-SiC ceramic. The oxygen element could only enter into the interior of the ceramic by anionic oxygen diffusion through the oxidation layer. In other words, after the formation of the dense and continuous SiO₂-MeO_x oxidation layer, the oxidation reactions are controlled by diffusion of oxygen (atomic or anionic oxygen) from the air-ceramic interface to the oxide-ceramic interface. Therefore, only trace amounts of oxygenelements was detected in the inner region, indicating that the dense SiO₂-MeO_x oxidation layer can effectively slow down the entry of oxygen.

4. Conclusions

In summary, a high-entropy diboride-based ultra-high temperature ceramic (Ti,Zr,Nb,Mo,W)B₂-SiC (MeB₂-SiC), has been successfully prepared at a relatively low sintering temperature of 1650 °C by using reactive hot-press sintering of pre-alloyed Me solid-solution metals and nonmetals of Si, C, and B₄C. The as-sintered MeB₂-SiC ceramic showed a high relative density of 97.8% and high mechanical properties (hardness of 19.74 ± 0.8 GPa, flexure strength of 533 ± 38 MPa, and fracture toughness of 6.01 ± 0.77 MPa·m^1/2). SEM and EPMA results showed that the as-sintered ceramic had a fine and homogeneous microstructure without significant elemental aggregation and grain coarsening. Aside from the primary high-entropy P6/mmm (191)-MeB₂ and SiC phases, a minor amount of P6₃/mmc (194)-WB₂ was also detected in the MeB₂-SiC ceramics. In addition, the oxidation tests at 1400 °C with different exposure times were conducted to investigate the oxidation behavior and microstructure evolution of the oxidation layer. With the oxidation exposure time increasing to 10 min, a continuous and relatively dense SiO₂-MeO_x oxidation layer was gradually formed. After that, the oxygen (atomic or anionic oxygen) diffusion mechanism was changed from air-ceramic interface diffusion-controlled to the oxide-ceramic interface diffusion-controlled, leading to decelerating oxidation behavior.