Fine-grained dual-phase high-entropy ceramics derived from boro/carbothermal reduction

In the current work ne-grained dual-phase, high-entropy ceramics (Ti 0.2 Zr 0.2 Hf 0.2 Nb 0.2 Ta 0.2 )B 2 -(Ti 0.2 Zr 0.2 Hf 0.2 Nb 0.2 Ta 0.2 )C with different phase ratios are prepared from powders synthesized via a boro/carbothermal reduction approach, by adjusting the content of B 4 C and C in the precursor powders. Phase compositions, densication, microstructure, and mechanical properties have been investigated. Due to the combination of pinning effect and the boro/carbothermal reduction approach, the average grain sizes (0.5–1.5 µm) of the dual-phase high-entropy ceramics, were much smaller as compared with previously reported values. The dual-phase high-entropy ceramics with 15 mol% boride phase exhibit the highest Vickers hardness (24.21 GPa) and fracture toughness (3.2 MPa•m 1/2 ).


Introduction
In 2015, an entropy stabilized oxide ceramic was reported for the rst time, which exhibited a singlephase rock salt structure [1]. Since then, the concept of high-entropy materials has applied to the eld of ceramics more widely. In the last two to three years, high-entropy ceramics have become a major global research emphasis. Different series of high-entropy ceramics, such as silicides [2,3], oxides [1,4], nitrides [5], carbides [6][7][8][9], and borides [10][11][12][13] have been widely studied. The subject of high-entropy boride (HEB) and high-entropy carbide (HEC) ceramics are, so far, the most intensively researched.
Equiatomic ve (or four)-metal HEB and HEC ceramics of transition metal (group IVB, B, and B) have the AlB 2 structure (space group P6/mmm #191) and rock salt structure (space group Fm-3m #225), respectively. HEB ceramics are known to be di cult to achieve full density, due to the strong, primarily covalent chemical bonds that lead to slow diffusion rates [10]. Six different high-entropy borides were fabricated via high-energy ball milling and spark plasma sintering, whereas the relative density reported was only about 92% [11]. Many researchers have shown that HEC ceramics have lower thermal conductivity, and higher elastic modulus and hardness values as compared to monocarbides. Yan et al. [6] have synthesized (Hf 0.2 Zr 0.2 Ta 0.2 Nb 0.2 Ti 0.2 )C by spark plasma sintering (SPS), which showed much lower thermal diffusivity and conductivity than the binary carbides like HfC, ZrC, TaC, and TiC. In prior work, most of the HEC and HEB ceramics were fabricated through the use of blends of commercially available single boride or carbide powders, and subsequent in-situ HEB or HEC formation during densi cation [6,[8][9]. As an alternative, high-entropy ceramics can be prepared and sintered using highentropy powders derived by the boro/carbothermal reduction approach, using metal oxides as powder precursors. For example, a single-phase rock salt structure (Hf,Zr,Ti,Ta,Nb)C powder was prepared by carbothermal reduction with an average particle size of about 550 nm and an associated oxygen content of 0.2 wt% [14]. Wei et al. [15] [16]. A series of studies have demonstrated that highentropy ceramic powders prepared by using the boro/carbothermal reduction approach exhibited better sintering behavior than the commercial powders [17][18][19]. Huo et al. showed that a (Ti,Zr)B 2 -(Zr,Ti)C composite fabricated by reactive hot pressing can re ne the grain size and improve the mechanical properties [20,21]. Recently, dual-phase high-entropy ultrahigh temperature ceramics (DPHE-UHTCs) have been fabricated starting from a combination of n binary borides and (5-n) binary carbides powders [22]. The relative density of this dual-phase high-entropy ceramic is up to 99%. However, the mean grain sizes of the boride and carbide formed in these dual-phase high-entropy ceramics were larger than 4.2 and 4.9 µm, respectively.
In the present work, (Ti 0.2 Zr 0.2 Hf 0.2 Nb 0.2 Ta 0.2 )B 2 -(Ti 0.2 Zr 0.2 Hf 0.2 Nb 0.2 Ta 0.2 )C dual-phase high-entropy ceramics were prepared using dual-phase high-entropy powders, which were in-situ synthesized by boro/carbothermal reduction approach using oxide precursors. The phase compositions, densi cation behavior, mechanical properties, and microstructures of the dual-phase high-entropy ceramics were studied. The objective of this study is to provide insight into the feasibility of fabrication of ne-grained dual-phase high-entropy ceramics in order to enhance its mechanical properties based upon applying a low-cost boro/carbothermal reduction approach. However, it should be noted that, a large number of researches have shown that when borides are prepared with the stoichiometric amounts of raw powders, according to Reactions (1), (2) and (3), a small amount of residual oxides would invariably remain in the product [10,16,18,23]. This is due to the evaporation of intermediate gas products (i.e., B 2 O 3 and boron-rich oxides), resulting in the loss of B form the initial mixture. In order to obtain two-phase high-entropy ceramics with different high-entropy phase fractions, the amounts of B 4 C and C were thus adjusted in the present work. For the current study, samples with nominally 100 mol% HEB, 80 mol% HEB + 20 mol% HEC, 60 mol% HEB + 40 mol% HEC, 40 mol% HEB + 60 mol% HEC, 20 mol% HEB + 80 mol% HEC, 100 mol% HEC sintered samples are referred to as HEB, B8C2, B6C4, B4C6, B2C8, and HEC, respectively. For boro/carbothermal synthesis, the ratio details of these mixes of powders are presented in Table 1.

Experimental Procedure
The raw materials were mixed in anhydrous ethanol with Si 3 N 4 milling media for 24 hours, using a roll jar mill, and then dried within a rotary evaporator. The dried mixtures were subsequently passed through a 100-mesh sieve to remove any large agglomerates. The powder mixture was then compacted into 30 mm diameter and ~ 5 mm thickness monoliths, and loaded in a graphite crucible. The preparation of the insitu synthesized, boro/carbothermal reduction powder is then carried out in a vacuum furnace. The mixtures were heat treated at 1650℃ for 1 h under vacuum ( 10 Pa), with a heating rate of 10 ℃/min, to complete the boro/carbothermal reaction(s) process. The as-prepared high-entropy powders were subsequently ground with an agate mortar and pestle, and then through a 100-mesh sieve. For powder densi cation, Spark Plasma Sintering (SPS, HPD-10-FL, FCT Systeme GmbH, Germany) consolidation was conducted, using a graphite die with an inner diameter of 30 mm. All samples were sintered at 2000 ℃ for 10 min, with a heating rate of 100 ℃/min, in an Ar atmosphere. A uniaxial pressure of 30 MPa was applied above 1650℃, and then maintained for the remainder of the sintering cycle.
Bulk densities of the sintered compacts were measured using the Archimedes method in distilled water.
The Vickers hardness (ISO 14705: 2008, MOD) was measured by the indentation method, with an applied load of 0.2 kg, held for 10 s. The fracture toughness, K IC , was measured by the indentation method with an applied load of 2 kg, held for 10 s [24]. Crystalline phases within the sintered specimens were determined by X-ray diffractometry (XRD; model D8 ADVANCE,Bruker Corp. Germany). The Rietveld re nement approach was used to calculated the lattice parameters and phase fraction of the highentropy ceramics from the recorded XRD pattern, using the Rietveld re nement EXPGUI software. The microstructures of SPS processed samples were examined using scanning electron microscopy (SEM; model SU-8220, Hitachi High-Technologies, Japan), which is equipped with an energy dispersive spectroscopy (EDS) Si-drift detector (X-Max N 50, Oxford, UK) for chemical analysis and mapping.

Phase compositions and densi cation
XRD patterns of all of the sintered samples are shown in Fig. 1. Only single phases of either a hexagonal lattice (nominally corresponding to HfB 2 ) or face-centred cubic lattice (nominally corresponding to HfC) were detected in the SPS densi ed HEB and HEC ceramics, respectively. In contrast, the B8C2, B6C4, B4C6, and B2C8 high-entropy ceramic composites are composed of a combination of boride and carbide phases as anticipated. No secondary phases were detected in all any of the SPS densi ed samples. The XRD results show that using a suitable combination metal oxides, B 4 C, and graphite as precursors, synthesis of both single-phase and dual-phase high-entropy ceramics is feasible via boro/carbothermal approach. For the dual-phase B8C2 to B2C8 sequence of samples, the changes in XRD peak intensities are consistent with the changes in the boride (or carbide) phase fraction. This indicates that the highentropy boride (or carbide) phase fraction in these dual-phase high-entropy ceramics can be carefully tailored through adjustment of the content of B 4 C and C.
Considering that the actual phase molar fraction in the sintered dual-phase high-entropy ceramics may not be consistent with the nominal molar fraction, it is necessary to calculate the action fractions of the high-entropy boride phase in the dual-phase ceramics utilizing Rietveld XRD re nement. The molar fraction of the (Ti 0.2 Zr 0.2 Hf 0.2 Nb 0.2 Ta 0.2 )B 2 phase in dual-phase high-entropy ceramic was calculated from the recorded XRD pattern, as shown in Table 2. The nominal molar fractions for the (Ti0.2Zr0.2Hf0.2Nb0.2Ta0.2)B2 phase in compositions B8C2, B6C4, B4C6, and B2C8 are 80 mol%, 60 mol%, 40 mol%, and 20 mol%, respectively. The actual molar fractions within B8C2, B6C4, B4C6, and B2C8 are 90 mol%, 59 mol%, 35 mol%, and 15 mol%, respectively. The reason for this discrepancy is that it is di cult to accurately quantify the evaporation of gas (i.e., B 2 O 3 and rich-boron oxides) during the high-entropy powder preparation step, when using boro/carbothermal reduction. Therefore, as a consequence, the targeted molar ratios of raw powders (speci cally, the contents of B 4 C and C) needs an increased theoretical and experimental veri cation to obtain accurate nal phase fractions in dual-phase high-entropy ceramics.  Fig. 3. Results show that there is no indication of notable clustering or segregation is visible within the individual element maps of B4C6, suggesting that these metal elements formed a solid solution in each HEB and HEC phase.
Microstructures of the fracture surfaces of the sintered specimens are shown in Fig. 4. Some small pores were observed within the grains and at the grain boundaries for the HEB, B8C2, HEC samples (e.g., Fig. 3a,   3b, 3f, 4a, 4b, and 4f). Conversely, no obvious evidence of pores was found in the B6C4, B4C6, or B2C8 samples. These results are consistent with the relative density values of the sintered specimens (Shown previously in Table 2). Based on SEM images of both the polished and fracture surfaces of the HEB, B8C2, and HEC samples (Figs. 3 and 4), an exaggerated grain growth could be observed relative to the other samples. This rapid grain growth could led to the formation of intragranular porosity, which becomes di cult to remove, and then inhibited full densi cation. However, due to the mutual grainboundary pinning effect, the dual-phase ceramics with a more even ratio of boride to carbide have ner grain size and signi cantly less residual porosity. The fracture morphology of the single-phase highentropy boride (HEB) samples was transgranular in nature (Fig. 4a), indicating a strong interface is formed between the individual high-entropy boride grains. However, in contrast, the fracture morphology of the single-phase high-entropy carbide (HEC) was intergranular in character (Fig. 5f). In further comparison, the dual-phase high-entropy ceramics containing both the high-entropy boride and carbide phases, showed both transgranular and intergranular fracture features (i.e., a mixed fracture model).
The variations in the grain sizes of the carbide and boride phases, as a function of the overall composition in the sintered samples, and hence the respective phase fraction, are presented in Fig. 5. The average grain sizes of the HEB and HEC samples are 2.58 µm and 2.31 µm, respectively. Due to the grain pinning effects, the grain size of the dual-phase high-entropy ceramics showed obvious mutual growth, inhibition, and each phase was smaller than that of the respective single-phase carbide and boride highentropy ceramics. The grain size decreased with the increase in counterpart phase fraction, similar to results previously reported in the literature [22]. The B8C2 and B2C8 samples have the smallest carbide and boride grain sizes of 0.53 µm and 0.64 µm, respectively. However, Huo et al. [22] used blended mixtures of commercial borides (TiB 2 , ZrB 2 , NbB 2 , HfB 2 , and TaB 2 ) and carbides (TiC, ZrC, NbC, HfC, and TaC) as raw materials to prepare high-entropy dual-phase ceramics; this work achieved minimum carbide and boride average grain sizes of 4.9 µm (for the equivalent of the 8B2C formulation) and 4.2 µm (for the equivalent of the 2B8C formulation), respectively. In the present work, the grain sizes of the dual-phase were signi cantly decreased for high-entropy ceramics composites SPS densi ed using 'pre-alloyed' powders, in-situ synthesized using a simple and low-cost boro/carbothermal reduction approach.

Mechanical properties
The hardness and toughness of dual-phase high-entropy ceramics with different phase fractions are presented in Fig. 6. Compared to the other samples, HEB shows the lowest hardness of 21.27 GPa, similar to Zhang and colleague's reported value (i.e., 21.7 GPa) [18]. With an increase of the carbide phase content in the dual-phase high-entropy ceramics, the hardness increases, and the B2C8 sample reaches a maximum hardness of 24.21 GPa. Due to the larger grain size, the hardness of the HEC samples (23.13 GPa) is slightly lower than that of B2C8.
The change in trend of indentation fracture toughness with carbide content is similar to the hardness.

Conclusions
Using a mixture of metal oxides, B 4 C and graphite powders as precursors, single-phase high-entropy boride, and carbide ceramic powders, together with dual-phase high-entropy powders, were synthesized via a simple one-step boro/carbothermal reduction approach. A series of high density, dual-phase highentropy ceramics, with different boride to carbide phase ratios, were then obtained by SPS processing at 2000℃. Compared to previously reported mean grain size values for similar dual-phase high-entropy materials processed from mixtures of single element commercial boride and carbide powders, the averaged grain sizes of high-entropy dual-phase ceramics in the present work are signi cantly reduced, by roughly one order of magnitude. For the SPS processed samples in the current work, dual-phase highentropy ceramics with 15 mol% of the boride phase exhibit the highest Vickers hardness (24.21 GPa) and fracture toughness (3.2 MPa•m 1/2 ). Figure 1 Representative XRD patterns of samples after SPS process.  Backscatter electron image and the corresponding EDS elemental mapping for the B4C6 sample. B2C8, and (f) HEC.

Figure 5
Averaged grain size of the SPS processed high-entropy ceramics.

Figure 6
Vickers hardness and fracture toughness of the SPS processed high-entropy ceramics.