Reactive Synthesis of B₄C–CrB₂, B₄C–TiB₂, AND B₄C–TiCrB₂ Heterophase Ceramics by Spark Plasma Sintering

The reactive synthesis of heterophase refractory ultrahard B₄C-based composites by spark plasma sintering (SPS) was examined. To produce heterophase B₄C + TiB₂ + CrB₂ ceramics, the chemical reaction between boron carbide and chromium oxide and between boron carbide and titanium carbide resulting in boron carbide–chromium diboride and boron carbide–titanium diboride composites was previously studied. The reactive sintering of B₄C + Cr₂O₃ + C and B₄C + TiC mixtures using boron carbide powders obtained from the Zaporizhzhya Abrasive Plant and Donetsk Chemical Reagent Plant (Ukraine) was compared. The boron carbide powders differed in the ratio of B₁₃C₂ and B₄C phases and particle sizes. The reactively synthesized TiB₂, CrB₂, and CrTiB₂ boride phases positively influenced the SPS consolidation and properties of the boron carbide composites. The B₄C–CrB₂ and B₄C–TiB₂ ceramics subjected to Vickers hardness testing under a load of 98 N showed HV levels of 23–29 GPa and 26–28 GPa. The ceramics demonstrated brittle fracture according to the Half-penny model, with a fracture toughness of 3 MPa∙m^1/2 for B₄C–CrB₂ and 4.4 MPa∙m^1/2 for B₄C–TiB₂. The 90 vol.% B₄C–5.5 vol.% TiCrB₂–4.5 vol.% C ceramics with ~33 GPa hardness and ~ 4 MPa∙m^1/2 fracture toughness were produced by reactive SPS from a mixture of B₄C (Zaporizhzhya Abrasive Plant), 6.6 wt.% TiC, and 11 wt.% Cr₂O₃. The high strength of TiCrB₂ ceramics was attributed to the stress–strain state, where the matrix phase of boron carbide was subjected to compressive stresses. The high hardness and fracture toughness allow the B₄C–TiCrB₂ composite to be classified as an ultrahard ceramic material.

Introduction

Hard and lightweight ceramics are gaining attention for their unique properties, such as wear resistance, heat resistance, and resistance to abrasive environments, and are thus attractive for the production of structural parts. The properties of boron carbide ceramics (boron carbide representing a combination of B₄C, B₁₀C, and B₁₃C₂ modifications), including rigidity, hardness, and wear resistance, make these ceramics suitable for nuclear energy applications, as part of control rods [1,2,3,4,5]. However, there are certain obstacles to the full implementation of their potential, primarily very high sintering temperatures (resulting from covalent bonds in boron carbide) and low fracture toughness. Moreover, sintering at temperatures above 2000°C leads to rapid grain growth and complicates the production of dense B₄C ceramics, even when external pressure is applied.

Liquid-phase sintering is an alternative method for synthesizing dense boron carbide ceramics. According to [6], the ceramics densify as the aluminum oxide melt wets boron carbide, surface diffusion energy reduces, and A1B₁₂C₂ forms through the reaction between B₄C and Al₂O₃. Similar results were reported in [7], where the addition of 2.5 vol.% Al₂O₃ significantly improved the consolidation of B₄C in hot pressing conditions at 2000°C, resulting in ceramics with 550 MPa bending strength. It is believed that TiB₂ and AlF₃ are effective additions for B₄C consolidation [8, 9]. Paper [10] reported that a relative density of more than 95% was achieved in the pressureless sintering of B₄C–TiB₂–1 wt.% Fe in the range 2150–2175°C and assumed that the iron-enriched liquid phase promoted the consolidation. The effect of SiC, TiC, WC, and BN additions was also addressed in [1, 11,12,13,14]. Skorokhod et al. [15, 16] consolidated dense B₄C–TiB₂ composites through the reaction sintering of B₄C with TiO₂ and C. The B₄C–15 vol.% TiB₂ composite (sintering temperature ranging from 1900 to 2050°C) demonstrated ~513 MPa bending strength and 3.71 MPa ∙ m^1/2 fracture toughness [16].

To achieve excellent properties, ceramic materials need to acquire a pore-free structure in the sintering process, with all phases being uniformly distributed throughout the volume. Papers [17, 18] demonstrated that the introduction of activating agents into the mixtures for producing boron carbide ceramics favorably influenced their hot pressing in a CO/CO₂ atmosphere. This increased the shrinkage rate, allowing the maximum density to be reached at low temperatures within short-term isothermal holding. Papers [19,20,21,22,23] examined the consolidation of B₄C, B₄C–B, and B₄C–BN under spark plasma sintering (SPS), where aggregate recrystallization and abnormal grain growth were slowed down. This contributed to the production of ceramic materials with a homogeneous, relatively fine-grained structure and high mechanical characteristics.

The content of particles in the secondary phase is such that it inhibits the B₄C grain growth during sintering and greatly improves the density of consolidated ceramics. The B₄C–TiB₂ ceramics are considered to be among the best composites that can retain the key B₄C properties to the maximum extent [24,25,26,27,28,29]. First, TiB₂ has a higher melting point (2980°C) than B₄C, ensuring that the B₄C–TiB₂ composite remains stable at high temperatures. Second, the B₄C–TiB₂ composite is strong and lightweight as TiB₂ has high hardness (25–32 GPa) and moderate density (4.52 g/cm³), allowing improvement in the bending strength. Moreover, TiB₂ exhibits good electrical conductivity, promoting easy consolidation of the B₄C–TiB₂ composites through SPS. Titanium diboride also significantly improves the mechanical properties of B₄C–TiB₂ composites if they are sintered in situ as follows:

$${\mathrm{B}}_{4}\mathrm{C}+2{\mathrm{TiO}}_{2}+3\mathrm{C}=2{\mathrm{TiB}}_{2}+4\mathrm{CO}\uparrow .$$

(1)

Papers [24, 26] reported that the B₄C–30 vol.% TiB₂ composite produced by SPS demonstrated 865 MPa bending strength and 39.3 GPa Vickers hardness, while the B₄C–43 vol.% TiB₂ composite produced by coprecipitation and consolidated in hot pressing conditions showed ~9.4 MPa ∙ m^1/2 fracture toughness.

According to [30], the B₄C–TiB₂ composite powders can be synthesized through the reaction between TiC_xN_1–x and B:

$${\mathrm{TiC}}_{x}{N}_{1-x}+\left(4x+2\right)\mathrm{B}={\mathrm{TiB}}_{2}+x{\mathrm{B}}_{4}\mathrm{C}+\left[\left(1-x\right)/2\right]{\mathrm{N}}_{2}.$$

(2)

The TiB₂ and B₄C phases form simultaneously in situ by reaction (2), resulting in the homogeneous distribution and finer sizes of TiB₂ and B₄C particles.

Chromium diboride is effective for improving the viscosity of B₄C ceramics because of a greater difference in thermal expansion between B₄C and CrB₂ compared to that between B₄C and TiB₂. This results in higher residual compressive stresses in the boron carbide matrix phase in the case of B4C and CrB₂. Furthermore, CrB₂ exhibits high hardness, a high melting point, and high chemical stability [31, 32]. The B₄C and CrB₂ phases may coexist according to the B₄C–CrB₂ phase diagram [33].

Makarenko [34] examined the interaction of B₄C with metal oxides in period IV according to the following reaction:

$${\mathrm{B}}_{4}\mathrm{C}-{{\mathrm{Me}}_{2}}^{\mathrm{IV}}{\mathrm{O}}_{3}+\mathrm{C}\to {\mathrm{B}}_{4}\mathrm{C}-{{\mathrm{Me}}_{2}}^{\mathrm{IV}}{\mathrm{B}}_{2}+\mathrm{CO}\uparrow .$$

(3)

Increased solubility of metals in this group (from calcium to chromium) in boron carbide was found.

The reduction of chromium oxide by boron carbide proceeds through the formation of intermediate boride phases with lower boron content: CrB–Cr₃B₄–CrB₂.

Our objective is to examine the reactive synthesis of heterophase mixtures of refractory ultrahard boron carbide composites by spark plasma sintering.

Experimental Procedure

Ukrainian boron carbide powders of two types were used to produce B₄C–TiB₂, B₄C–CrB₂, and B₄C–TiB₂–CrB₂ (B₄C–TiCrB₂) ceramic materials by consolidation under reactive SPS.

One boron carbide powder was produced at the Zaporizhzhya Abrasive Plant (B₄C (ZAP)) by electric arc carbon thermal reduction and had an average particle size of 20–350 μm (Fig. 1a). To unify the morphology of both starting boron carbide powders, the B₄C (ZAP) powder was ground and homogenized in a boron carbide-lined ball mill with boron carbide grinding balls at a speed of 60 rpm for 480 h in isopropyl alcohol at a material : balls : alcohol weight ratio of 1: 5: 2. After grinding, the average particle size was ~10 μm and the main size fraction was 3.5–9.0 μm, but there were both finer (0.5–2.5 μm) and coarser (12–35 μm) particles (Fig. 1b, c; Table 1).

Fig. 1.

Microstructure of the B₄C (ZAP) powder before (a) and after (b) grinding; particle-size distribution after grinding (c)

TABLE 1. Sedimentation Analysis Results

The other boron carbide powder was produced at the Donetsk Chemical Reagent Plant (B₄C (DCRP)) by direct vacuum synthesis from amorphous boron with carbon black. In the B₄C powder (DCRP), 0.6–2 μm particles constituted the main size fraction, but there were 4.5–9.0 μm agglomerates, which disintegrated when the powders were mixed (Fig. 2).

Fig. 2.

Microstructure of the B₄C (DCRP) powder according to scanning electron microscopy (a) and particle-size distribution (b)

Table 1 summarizes the sizes and specific surface areas of the boron carbide powders according to sedimentation analysis. The sedimentation data provide averaged values for particles and agglomerates. The DCRP TiC and Cr₂O₃ powders had a wide size distribution (2–25 μm), probably resulting from weak agglomerates.

The mixtures were subjected to SPS using an FCT SPS–HP D25 machine (FCT, Germany) in graphite dies of increased strength (MPG-7, Ukraine) with punches 20 or 30 mm in diameter in an argon atmosphere (PCh class, Ukraine). The heating rate was 100°C/min and compaction pressure varied from 15 to 60 MPa. The holding time at a consolidation temperature of 1950–2050°C was 5–10 min. The samples were first cooled down at a rate of 50°C/min to 900°C and then under standard SPS conditions for no more than 10 min. The kinetic dependences for compaction and shrinkage rate take into account the thermal expansion of the graphite die and SPS equipment.

The density of the samples was measured hydrostatically. The microstructure of the consolidated samples with a relative density of 96.0–98.6% was examined by scanning electron microscopy (SEM) employing JEOL JEM-2100F, EVO Zeiss 50, JEOL JSM 7001F, and Tescan Mira 3 microscopes. All microscopes were equipped with energy-dispersive X-ray analyzers (EDX) for elemental analysis.

X-ray diffraction studies of the powders and ceramics were conducted using a DRON-3M diffractometer in Cu-K_α radiation. The X-ray diffraction patterns were photographed point by point in 0.05° steps with 10 sec exposure time at one point in the range 15–70°. The Analyze software package and PDF-2 database were used for qualitative phase analysis. Quantitative X-ray phase analyses were carried out with the PowderCell software package using the Rietveld refinement. At the first stage, theoretical crystal lattices were constructed for each of the phases contained in the ceramics. At least three reflections of each ceramic phase had to be present in the diffraction pattern. At the second stage, the superpositions of the constructed theoretical lattices were compared with the experimental diffraction patterns to determine the phase composition of the ceramics.

The approach elaborated in [35] was used to determine the amount of bound carbon in the B₄C (ZAP) and B₄C (DCRP) powders and to analyze their structural state. The approach included precision X-ray photography of several lines of the boron carbide powder with a silicon reference sample. The dependence (Fig. 3) of the bound carbon content on the lattice parameter c in the starting boron carbide samples was plotted using the data provided in [36]. The hexagonal lattice parameters a and c were determined considering this dependence and the 2θ values for the diffraction peaks for several boron carbide reflection planes calculated from the X-ray photography results.

Fig. 3.

Content of bound carbon in boron carbide versus lattice parameter c [36]

For precision analyses of lattice parameters of the boron carbide powders, the photography step was reduced to 0.02° and exposure time was increased to 25 sec. The NewProfile Version 3.5 software based on the Nadler–Mead optimization method was employed to separate the superimposed diffraction lines and analyze their parameters for the ceramics [37]. The measurements were performed using a standard silicon powder sample (a = 0.5430825 nm and d₅₀ = 7 μm), which was uniformly mixed with the starting samples of boron carbide powders.

The particle-size distribution was determined using an SA-CP3 centrifugal particle analyzer (Shimadzu, Japan). The analyzer operates by detecting changes in the optical density of suspensions, resulting from gravitational and/or centrifugal sedimentation of particles, through photometric measurements. The analyzer enables the determination of particle-size distribution in the range 0.02–150 μm. To prepare the suspension, ultrasonic dispersion was used with the addition of sodium hexametaphosphate.

The Vickers hardness (HV) was measured with an MVK-E hardness tester (Akashi Co., Japan) by holding the sample under a load of 98 N (10 kg) for 15 sec, followed by the standard procedure per ASTM C 1327-03. The value for each sample was averaged over 10 measurements. To avoid the mutual penetration of cracks and the influence of defects during Vickers indentation, all 10 tests were performed at a distance of 300 μm from each other. The fracture toughness was calculated with the Evans equation [38].

The X-ray diffraction patterns for the starting B₄C powders (ZAP and DCRP) and the results from qualitative phase analysis are summarized in Fig. 4. The B₄C (ZAP) powder significantly differs by a much higher amount of graphite and carbon black, reaching almost 5.0 vol.%. This amount of graphite and carbon black in the B₄C (ZAP) powder corresponds to the amount needed to complete the transition of B₁₃C₂ to B₄C.

Fig. 4.

X-ray diffraction patterns for the starting powders: a) B₄C (ZAP) and b) B₄C (DCRP)

It should also be noted that the B₄C (ZAP) powder contains up to 1.5 vol.% SiC. The results of quantitative phase analysis are summarized in Table 2.

TABLE 2. Phase Composition (vol.%) of the Starting B₄C (ZAP) and B₄C (DCRP) Powders According to XRD

Since the total amount of carbon black and graphite does not exceed 5 vol.%, their presence can be ascertained from lines with a relative intensity of 100% in the diffraction pattern for the B₄C (ZAP) powder (Fig. 4). Hence, the position of the peak for carbon black in copper radiation is determined by a 2θ diffraction angle of ~25.73° and for graphite by a 2θ diffraction angle of ~26.56°.

Figure 5 shows precision diffraction patterns for the boron carbide powders at 2θ diffraction angles of 68.3–72.9°. The diffraction curves for the B₄C (DCRP) powder can be quite accurately described by one Pearson VII dome function [37], which indicates that the starting powder contains a small amount of B₁₃C₂. However, to describe the diffraction curve for B₄C (ZAP) (Fig. 5), at least two appropriate Pearson VII functions are to be used, the second function describing quite a significant amount of B₁₃C₂. Therefore, the structural state of the ZAP and DCRP boron carbide powders differs significantly.

Fig. 5.

Precision diffraction pattern for the starting powders at diffraction angles 2θ = 68.3–72.9 deg: a) B₄C (ZAP) and b) B₄C (DCRP)

According to the analysis that used one Pearson VII function, the concentration of bound carbon in the B₄C (DCRP) powder is 19.7 at.%. The powder can be considered virtually stoichiometric.

In the case of the B₄C (ZAP) powder, the above approach for determining the lattice parameters has a substantial error because the diffraction curve significantly differs from the Pearson VII function. Therefore, when the structural state of this powder was analyzed, it was considered a mixture of B₁₃C₂ and B₄C (Fig. 4). The calculation results were used to construct theoretical crystal lattices for B₁₃C₂ and B₄C with the PowderCell software, and the superposition of the theoretical lattices was compared with the experimental diffraction pattern.

According to the structural state analysis for the starting powders from different manufacturers, should they be considered a mixture of B₁₃C₂ and B₄C, we have B₁₃C₂: B₄C = 28: 72 for B₄C (ZAP) and B₁₃C₂: B₄C = 7: 93 for B₄C (DCRP) (which is accounted for in Table 2). Hence, the B₁₃C₂ content is approximately 28 vol.% in the B₄C (ZAP) powder, which is almost four times higher than its content in the B₄C (DCRP) powder, reaching approximately 7 vol.%. These powders are expected to behave differently when sintered in the same conditions with the same sintering activators.

Results and Discussion

Calculation of Starting Mixtures. To produce heterophase B₄C–TiB₂–CrB₂ ceramics, the formation of TiB₂ particles in SPS conditions through the chemical reaction between boron carbide and titanium carbide at 1200–1700°C was studied:

$$2{\mathrm{B}}_{4}\mathrm{C}+2\mathrm{TiC}={\mathrm{B}}_{4}\mathrm{C}+2{\mathrm{TiB}}_{2}+3\mathrm{C}.$$

(4)

This reaction is exothermic, and additional heat generation in the synthesis of the composites promotes their rapid compaction. The final composition of the ceramics was determined taking into account Eq. (3); the initial composition included ~15 wt.% TiB₂ and ~4 wt.% free carbon and final density ρ was 2.68 g/cm³ [39]. The ratio between the thermal expansion coefficients (TECs) for B₄C and TiB₂ leads to compressive stresses in the matrix phase (B₄C) [17, 18]. However, reaction (4) proceeds with the release of free carbon, affecting the strength of the composites. To neutralize the negative influence, components that chemically react with carbon to form strong refractory compounds at high temperatures should be added to the starting mixture.

By analogy with TiB₂, CrB₂ is expected to be an effective addition for the sintering of B₄C-based ceramics through the formation of a eutectic at 2150°C, according to the B₄C–CrB₂ phase diagram [33]. Moreover, the high TEC of chromium diboride additionally increases residual compressive stresses in B₄C, and a small amount of free carbon in the composites can lead to increase in the fracture toughness of the ceramics [18]. Chromium oxide in the range 1400–1500°C loses oxygen and binds free carbon that forms by reaction (3).

Chromium diboride CrB₂ in the B₄C–10 vol.% CrB₂ composite with density ρ = 2.89 g/cm³ forms by reaction [39]:

$$2{\mathrm{B}}_{4}\mathrm{C}+{\mathrm{Cr}}_{2}{\mathrm{O}}_{3}+3\mathrm{C}={\mathrm{B}}_{4}\mathrm{C}+2{\mathrm{CrB}}_{2}+3\mathrm{CO}+\mathrm{C}.$$

(5)

Considering that the reaction between B₄C and TiC leads to free carbon in the amount required for the reaction between B₄C and Cr₂O₃ to proceed, we believe that almost all free carbon is combined with oxygen. As a result, B₄C–TiB₂–CrB₂ ceramic material is produced [39].

Production of B₄C–CrB₂ Ceramics in SPS Conditions. To support the theoretical calculations for selecting the ceramic composition, mixtures of boron carbide from different manufacturers with Cr₂O₃ and with TiC were subjected to preliminary reactive sintering.

At the first stage, SPS of the B₄C–CrB₂ ceramics from B₄C (ZAP) or B₄C (DCRP) mixtures with 10 vol.% Cr₂O₃ and 2 vol.% C to complete the transformation of chromium oxide to chromium diboride was examined. The powder mixture was heated at a rate of 100°C/min before the reaction began and at 20°C/min during the reaction up to the holding temperature. The initial pressure was 15 MPa until the transformation reactions between the components finished and then pressure was raised to 50 MPa. However, the final temperature of 1900°C turned out to be insufficient, since the ceramics had approximately 20% porosity. A 50°C increase in the final holding temperature allowed ceramics with residual porosity ≤5% to be produced, but this did not significantly influence the boron carbide grain growth: the grain growth factor was 2.

The microstructures of B₄C–CrB₂ are presented in Fig. 6a, b. The electric current and external pressure in SPS conditions accelerate the transformation of Cr₂O₃ to CrB₂ and the formation of the liquid phase, which can be seen in SEM images of the fracture surfaces resulting from three-point bending at room temperature (Fig. 6c). An enlarged image of the selected area is shown in Fig. 6d.

Fig. 6.

Microstructure of the B₄C (DCRP)–CrB₂ composite produced by reactive SPS: a) 1900°C and b) 1950°C holding temperature; c, d) fracture surfaces of the composite

A vapor/gas phase is known [34] to coexist with the B₄C + Cr₂O₃ mixture over a wide temperature range of 750–1500°C, which includes CO, B₂O₃, B₂O₂, and atomic chromium. Figure 7 shows the dependence of reactive SPS parameters for the B₄C (DCRP) + 10 vol.% Cr₂O₃ + 2 vol.% C mixture. Three key stages of consolidation under the action of electric current and pressure can be singled out. The first significant peak denoting increase in the shrinkage rate is observed at 650–680°C. This may be associated with the decomposition of a partially hydrated B₂O₃ layer on the B₄C powder surface (Fig. 7, region I).

Fig. 7.

Reactive SPS of the B₄C (DCRP) + 10 vol.% Cr₂O₃ + 2 vol.% C mixture: 1) shrinkage; 2) shrinkage rate; 3) temperature; regions I–III—consolidation stages under the action of electric current and pressure

A further increase in temperature to 1150°C activates the evaporation and dissociation of Cr₂O₃, and Cr₂₃C₆ and CrB begin to form when temperature reaches 1300°C (Fig. 7, region II). In reactive SPS conditions, two-component B4C (DCRP)–CrB₂ ceramics start forming in the presence of excessive boron carbide in the mixture at 1300°C, and this process continues up to the holding temperature (Fig. 7, region III). The insignificant shrinkage observed during isothermal holding probably results from completion of the consolidation process for the B4C (DCRP)–CrB₂ ceramics.

Carbon black introduced into the mixture as a source of atomic carbon is highly reactive and thus enabled the completion of all synthesis reactions to produce a single CrB₂ phase at 1900°C (Table 3).

TABLE 3. Phase Composition of the Two-Component Ceramics

The same sintering behavior was shown by the B₄C (ZAP) + Cr₂O₃ mixture, but X-ray diffraction indicated that 0.5 vol.% Cr₅B₃ also formed besides the CrB₂ phase (Table 3), which can influence the mechanical properties of the ceramics.

X-Ray Diffraction of B₄C–CrB₂ Ceramics. Figure 8 shows X-ray diffraction patterns for the twocomponent ceramics produced from the B₄C (ZAP) + 10 vol.% Cr₂O₃ + 2 vol.% C and B₄C (DCRP) + 10 vol.% Cr₂O₃ + 2 vol.% C mixtures. The phase composition of the ceramics is summarized in Table 3. Note that the B₄C (ZAP) powder samples contain the Cr₅B₃ phase, while this phase is not present in the B4C (DCRP) ceramics. The B₄C (DCRP) samples have a higher CrB₂ content (~3.5 vol.%) than the B₄C (ZAP) ceramics (~1.0 vol.%). Both samples contain an insignificant amount of graphite (Table 3). Traces of α-SiC are also present in the B₄C (ZAP) ceramics.

Fig. 8.

X-ray diffraction patterns for the B₄C–CrB₂ ceramics produced from the B₄C (ZAP)+ 10 vol.% Cr₂O₃ + 2 vol.% C (a) and B₄C (DCRP) + 10 vol.% Cr₂O₃ + 2 vol.% C (b) powder mixtures

According to XRD, the Cr₅B₃ phase formed besides the CrB₂ phase (Table 3), and its impact on the mechanical properties of the ceramics needs to be additionally analyzed.

Figure 9 presents kinetic dependences for shrinkage and shrinkage rate to compare the effects of the starting boron carbide powder during consolidation in the same sintering conditions in an argon atmosphere: at a heating rate of 100°C/min under 1900°C and at a heating rate of 20°C/min at 1900–2050°C. An external pressure of 50 MPa was applied after the reactions proceeded in the system, the holding time being 5 min. The density of the starting samples differed by no more than 5% at the beginning of sintering. In these consolidation conditions, B₄C (ZAP)–CrB₂ samples with a density of 2.55 ± 0.05 g/cm³ and B₄C (DCRP)–CrB₂ samples with a density of 2.61 ± 0.05 g/cm³ were produced. The lower density values compared to the calculated ones are associated with ~0.5% porosity and 0.5 vol.% residual graphite in the case of the B₄C (DCRP)–CrB₂ ceramics. The B₄C (ZAP)–CrB₂ ceramics contain also 0.5 vol.% Cr₅B₃ and have ~2% residual porosity. The structures of both ceramics are presented in Fig. 10. Chromium diboride particles appear as high-contrast areas in the darker B₄C matrix. The 3–40 μm CrB₂ particles are dispersed in the boron carbide matrix.

Fig. 9.

Temperature dependence of shrinkage, shrinkage rate, and density under SPS of the B₄C (ZAP) + 10 vol.% Cr₂O₃ + C (1) and B₄C (DCRP) + 10 vol.% Cr₂O₃ + 2 vol.% C (2) mixtures

Fig. 10.

Structures of the B₄C (ZAP)–CrB₂ (a) and B₄C (DCRP)–CrB₂ (b) ceramics after SPS in the conditions illustrated in Fig. 9

At the initial sintering stages, the B₄C (ZAP) and B₄C (DCRP) powders with Cr₂O₃ show similar behavior. When temperature approaches 1900°C, the shrinkage of the B₄C (DCRP) mixture intensifies and its shrinkage rate increases compared to the B₄C (ZAP) powder mixture, which results from the particle size and ratio of the B₁₃C₂ and B₄C phases in the starting boron carbide powders (Table 2).

The B₄C (ZAP)–CrB₂ and B₄C (DCRP)–CrB₂ ceramics with hardness HV = 23.29 ± 1.78 GPa and 27.43 ± 2.12 GPa were produced at a load of 98 N. Note that the ceramics demonstrated brittle fracture. Hence, a microscopic image of the indenter impression area was used to determine fracture toughness, considering the criteria for calculating fracture toughness depending on the type of fracture, such as the c/a ratio, where c is the length from the indenter impression center to the crack end and a is half the diagonal of the indenter impression. In the case of c/a < 2.5, the crack develops according to the Palmqvist model [38] and fracture toughness is calculated with the Niihara equation. If c/a ≥ 2.5 is observed, the crack is classified according to the Half-penny model and fracture toughness K_Ic is calculated with the Evans equation. In our case, fracture occurs through the Half-penny model and the second c/a ratio is obeyed (Fig. 11). Fracture toughness K_Ic was 3.03 ± 0.47 MPa∙m^1/2 and 2.91 ± 0.35 for B₄C (ZAP)–CrB₂ and B₄C (DCRP)–CrB₂. These values agree well with the published data [40], where the B4C –20 mol.% CrB₂ (~13 vol.% CrB₂) ceramics demonstrated K_Ic = 2.6–3.7 MPa∙m^1/2.

Fig. 11.

Indenter impression on the B₄C (ZAP)–CrB₂ ceramic surface

Production of B4C–TiB₂ Ceramics in SPS Conditions. To examine the interaction between B₄C (ZAP) or B₄C (DCRP) and TiC to form TiB₂ under reactive SPS, the B₄C (ZAP) + 10 vol.% TiC and B₄C (DCRP) + 10 vol.% TiC mixtures were sintered. It was established that TiB₂ began to form at the same temperature in both mixtures: 1500°C (Fig. 12, region I). The process actively continued up to 1700°C, and then the formation of TiB₂ slowed down (Fig. 12, region II).

Fig. 12.

Reactive SPS of the B₄C (DCRP) + 10 vol.% TiC mixture: 1) shrinkage; 2) shrinkage rate; 3) temperature; regions I and II—consolidation stages under the action of electric current and pressure

The kinetic dependences of shrinkage and shrinkage rate for comparing the effects of the starting boron carbide powder consolidated in the same sintering conditions are presented in Fig. 13. The B₄C (ZAP) + 10 vol.% Cr₂O₃ + 2 vol.% C and B₄C (DCRP) + 10 vol.% Cr₂O₃ + 2 vol.% C mixtures were sintered in the same conditions. Analysis of the sintering kinetics shows that the B₄C (DCRP) + TiC mixture exhibits more intensive shrinkage under reactive SPS after pressure is applied. This results from the fine particles and reactivity of the starting B₄C (DCRP) powder, according to XRD studies.

Fig. 13.

Temperature dependence of shrinkage, shrinkage rate, and density under SPS of the B₄C (ZAP) + 10 vol.% TiC (1) and B₄C (DCRP) + 10 vol.% TiC (2) mixtures

Note that density of the starting samples differed by 20% at the beginning and then equalized when the heating rate decreased to 20°C/min at 2000°C: 2.72 ± 0.05 g/cm³ for B₄C (ZAP)–TiB₂ and 2.65 ± 0.05 g/cm³ for B₄C (DCRP)–TiB₂. Open porosity in both ceramics was no more than 0.3%. As was the case for the boron carbide ceramics with CrB₂, this ceramic material had lower density than the calculated one. This is associated with 12–15 vol.% residual graphite depending on the starting mixture (Table 3). The microstructures of the ceramics are presented in Fig. 14.

Fig. 14.

Structures of the B₄C (ZAP)– TiB₂ (a) and B4C (DCRP)–TiB₂ (b) ceramics after SPS in the conditions illustrated in Fig. 13

Vickers hardness measurements determined HV = 26.55 ± 1.97 GPa for B₄C (DCRP)–TiC and 28.29 ± 1.36 GPa for B4C (ZAP)–TiC at a load of 98 N. Their fracture toughness K_Ic was 4.77 ± 0.17 MPa∙m^1/2 and 4.44 ± 0.23 MPa∙m^1/2.

X-Ray Diffraction of B₄C–TiB₂ Ceramics. X-ray diffraction patterns for the ceramics produced from the B₄C (ZAP) + 10 vol.% TiC and B₄C (DCRP) + 10 vol.% TiC mixtures are provided in Fig. 15, and their phase composition is summarized in Table 3. Note that TiB₂ forms in both samples, but its amount is greater in the B₄C (DCRP) ceramics: ~11.5 vol.% against ~9.5 vol.% in the B₄C (ZAP) sample. There are also traces of α-SiC in the samples, as it is present in the starting powder. The residual graphite content in the B₄C (ZAP)–TiB₂ ceramics is somewhat lower (12 vol.%) than in the B₄C (DCRP)–TiB₂ ceramics (15 vol.%). Recall (Table 2) that carbon content was higher in the starting B₄C (ZAP) powder. To identify mechanisms of structural transformations between the boron carbide modifications in the ceramics, we plan to conduct additional studies.

Fig. 15.

X-ray diffraction patterns for the ceramics consolidated from the B₄C (ZAP) + 10 vol.% TiC (a) and B₄C (DCRP) + 10 vol.% TiC (b) powders

Production of B₄C–CrB₂–TiB₂ Ceramics. After analyzing the reactive consolidation of the ceramics including boron carbide from two manufacturers and additions of 10 vol.% Cr₂O₃ or 10 vol.% TiC, B₄C (ZAP) was chosen for developing heterophase B₄C–CrB₂–TiB₂ ceramics [39]. This choice was made because the kinetic dependences for B₄C (ZAP) homogenized by grinding to an average size of approximately 10 μm did not substantially differ from those for B₄C (DCRP). The B₄C (ZAP) samples had somewhat higher final density and lower content of residual carbon as graphite and carbon black. The final phase composition of the ceramics (Table 3) was used to determine the optimal amounts of chromium oxide and titanium carbide in the mixture for producing heterophase B₄C–CrB–TiB₂ ceramics, specifically (wt.%): B₄C (ZAP) + 6.6 TiC + 11 Cr₂O₃ (mixture 1) and B₄C (ZAP) + 3.3 TiC + 5.5 Cr₂O₃ (mixture 2) [39].

Both mixtures (for B₄C (ZAP)–CrB₂ and B₄C (ZAP)–TiB₂ ceramics) were sintered in the same SPS conditions in the range 1900–2050°C. Figure 16 shows temperature dependences of shrinkage and shrinkage rate for mixture 1. Reactive sintering stages can be singled out. At temperatures ranging from 550 to 700°C, B₂O₃ melts and evaporates from the boron carbide surface. The reaction between TiC and B₄C to release CO and atomic carbon begins in the range 1100–1300°C, and the reaction of Cr₂O₃ simultaneously with B₄C and free carbon to form TiB₂ and CrB₂ begins at 1200°C. According to XRD (Table 4), TiB₂ and CrB₂ are not formed, but double diboride, TiCrB₂, shows up under reactive SPS. The microstructure and elemental composition of the ceramics produced from mixture 1 are presented in Fig. 17.

Fig. 16.

Reactive SPS of the B₄C (ZAP) + 6.6 wt.% TiC + 11 wt.% Cr₂O₃ mixture: 1) shrinkage; 2) shrinkage rate; 3) temperature; I and II—consolidation stages under the action of electric current and pressure

TABLE 4. Phase Composition of the Ceramics According to XRD

Fig. 17.

Microstructure of the ceramics produced from the B₄C (ZAP) + 6.6 TiC + 11 Cr₂O₃ mixture and EDX results after reactive SPS

X-Ray Diffraction of _B4C–CrB₂–TiB₂ Ceramics. According to XRD, TiCrB₂ with different ratios of Ti and Cr formed in the samples produced from mixtures 1 and 2. This is confirmed by the TiCrB₂ lattice parameters. Hence, the lattice parameters are a = 0.3015 nm and c = 0.3178 nm for the ceramics produced from mixture 1 and a = 0.3021 mm and c = 0.3197 nm for the ceramics from mixture 2 (Table 4 and Fig. 18). The concentration of TiCrB₂ in the samples substantially differs as well: ~5.5 vol.% in the ceramics with a greater activator content (from mixture 1) and ~2.0 vol.% in the ceramics from mixture 2. Note also a significant graphite amount (~4.5 vol.%) in the ceramics from mixture 1, while graphite is hardly present in the ceramics from mixture 2.

Fig. 18.

X-ray diffraction pattern for the ceramic samples produced from mixtures 1 (a) and 2 (b)

The B₄C–TiCrB₂ composite produced under reactive SPS demonstrated promising Vickers hardness HV at a load of 98 N: ~33 GPa for mixture 1 and >25 GPa for mixture 2. These hardness values are due to ratios between the TECs of B₄C, TiB₂ CrB₂, and TiCrB₂ [41, 42]. In terms of the TEC ratio for B₄C (5.6 K^–1) and CrB₂ (10.5 K^–1), residual compressive stresses in the boron carbide matrix after reactive SPS in the system with CrB₂ are somewhat higher than for the ceramics with TiB₂ (TEC = 8.1 K^–1). It can be assumed that residual stresses in the ceramics with TiCrB₂ will be lower than those in the ceramics with CrB₂ and higher than in the samples with TiB₂.

This stress–strain state imparts high mechanical properties to the ceramics with TiCrB₂ (Table 5), where the matrix B₄C phase is actually subjected to compressive stresses. Additional studies are needed to examine the effect of internal stresses.

TABLE 5. Hardness and Fracture Toughness of the Ceramics Produced by Reactive SPS

The hardness and fracture toughness of the B₄C–TiCrB₂ composite allow it to be classified as an ultrahard ceramic material.

Therefore, the best characteristics were exhibited by the boron carbide composite with additions of 6.6 wt.% TiC and 11 wt.% Cr₂O₃ (mixture 1), produced from double titanium–chromium boride (B₄C–TiCrB₂) by reactive SPS.

Conclusions

The development of boron carbide–chromium diboride and boron carbide–titanium diboride composites under SPS through the chemical reaction between boron carbide and chromium oxide and between boron carbide and titanium carbide was previously studied to produce heterophase B₄C–TiB₂–CrB₂ ceramics.

The reactive consolidation of the B₄C + 10 vol.% Cr₂O₃ + 2 vol.% C and B₄C–TiC mixtures based on two boron carbide powders from the Donetsk Chemical Reagent Plant and Zaporizhzhia Abrasive Plant was compared to support theoretical calculations for choosing the ceramic composition.

Precision X-ray diffraction of intermediate and final products from the reactive consolidation of boron carbide with chromium oxide revealed that the B₄C (ZAP) ceramics contained the Cr₅B₃ phase, while this phase was not found in the B₄C (DCRP) ceramics. The chromium diboride content was 1.0 and 3.5 vol.%, respectively.

The positive effect of the TiB₂, CrB₂, and CrTiB₂ phases on the sintering under SPS conditions and on the mechanical properties of the B₄C composites was ascertained. The composites demonstrated 23–29 GPa hardness and ~3 MPa∙m^1/2 fracture toughness at a load of 98 N.

The two-component B₄C–TiB₂ composites contained 9.5 vol.% TiB₂ in the B₄C (ZAP) ceramics and 11 vol.% TiB₂ in the B₄C (DCRP) ceramics and demonstrated 28 and 26 GPa hardness and ~4.4 MPa∙m^1/2 fracture toughness.

Reactive SPS of the B₄C (ZAP) + 6.6 wt.% TiC + 11 wt.% Cr₂O₃ mixture produced high-temperature ultrahard 90 vol.% B₄C–5.5 vol.% TiCrB₂–4.5 vol.% C ceramics, which showed a Vickers hardness of ~33 GPa at a load of 98 N and a fracture toughness of ~4 MPa∙m^1/2. These mechanical properties result from stress–strain state of the matrix boron carbide phase that occurs under reactive SPS and requires additional studies.