Synthesis of B₄C–TiB₂ composition powder mixtures by carbidobor reduction using nanofibrous carbon for ceramic fabrication

Introduction

Over the past 20 years, the study of the production and properties of B₄C–TiB₂ composite ceramics has attracted a great interest, driven by the combination of the unique properties of their components – boron carbide and titanium diboride.

Boron carbide exhibit a high melting temperature (2447 °C) and possesses a unique combination of high hardness (up to 43 GPa) and low density (2.52 g/cm³)[1; 2]. However, ceramics based on it are characterized by poor sintering and low fracture toughness. The use of modifying additives, such as titanium diboride, can significantly increase these indices by 10–40 % [3; 4].

Titanium diboride, as well as boron carbide, is a refractory compound, its melting temperature is ~3225 °C [5]. The microhardness of TiB₂ ceramics is rather high and amounts to 25–35 GPa. Titanium diboride exhibit a relatively high thermal conductivity coefficient (66.4 W/(m∙K)) and low specific electrical toughness (~10\(^{-}\)⁷ Ohm·m) [6; 7]. Furthermore, TiB₂ is quite stable when heated in air, and it is not oxidized at a temperature of up to 800 °C. At a temperature of 900 °C, it is slightly oxidized with the formation of a vitreous protective film on the surface of the material, which prevents its further oxidation [8].

Many researchers note an increase in the fracture toughness and sinterability of the ceramics based on B4C–TiB2 system in comparison with the ceramics containing only B₄C [9; 10]. The presence of TiB₂ prevents the growth of В4С grains, reduces the sintering temperature and improves the mechanical properties of the resulting composite [11; 12]. Besides, there are data [13; 14] being indicative of the fact that the presence of TiB₂ increases the electrical conductivity of B₄C–TiB₂ material. B₄C–TiB₂ composite of eutectic composition can itself act as a modifying additive for refractory ceramics, increasing its mechanical properties [15].

In most cases, ready-made B₄C and TiB₂ powders are used as batch components for the production of B₄C–TiB₂ composite ceramic material [14; 16]. The preparation of such a batch before compaction involves mixing using a planetary ball mill.

The literary sources present the data on the production of B₄C–TiB₂ composites by in situ methods in accordance with reactions (1) [17], (2) [3] and (3) [18]:

The advantage of running the processes in accordance with reactions (1) and (2) is the absence of gaseous products, which is particularly important for the simultaneous synthesis and compaction of the material. On the other hand, the use of expensive boron and long-term mixing of the batch are required. Reaction (3) is prospective for the preliminary production of B₄C–TiB₂ batch due to the use of cheaper reagents. For instance, in terms of producing a mixture with a molar ratio of B₄C:TiB₂ = 1:1, the cost of reagents for reaction (3) is almost 5 times lower than for reactions (1) and (2). In addition, the gaseous product release during the heat treatment can contribute to additional mixing of the batch and more uniform heating.

Acetylene black is most often used as a carbon source in carbide synthesis. This material exhibit a rather high surface area of ~50 m²/g. However, nanofibrous carbon (NFC) with a developed surface area (~150 m²/g) can serve as a more efficient carbon material [19]. A highly dispersed carbon agent can accelerate the formation of titanium diboride due to more intensive diffusion of carbon into titanium dioxide particles. It should be noted that there are some technological challenges in using NFC. This highly dispersed material is prone to caking and requires thorough homogenization of the reaction mixture before the heat treatment. Besides, NFC is a more expensive reagent, its price is about 5 times higher than that of acetylene black. However, in light of the fact that the mass fraction of carbon agent in the reaction mixture for reaction (3) is relatively low, the cost of final product increases insignificantly.

The purpose of this paper is to research the synthesis and study the properties of B₄C + TiB₂ composite powder materials obtained by the carbide reduction of titanium dioxide (reaction (3)) using NFC.

Research methods

To obtain B₄C + TiB₂ powder composites, the following reagents were used:

– highly dispersed boron carbide В₄С (assay 98.5 wt. %, average particle size d = 2.1 μm) synthesized from simple substances according to method [20];

– titanium dioxide (TU 6-09-3811-79, assay 99.0 wt. %, d = 1.0 μm);

– nanofibrous carbon (carbon content 99 wt. %) [21; 22].

The used NFC contained catalyst residues: ~0.1 wt. % Al₂O₃ and 0.9 wt. % Ni. In its initial form, the carbon material consisted of pellets being 0.4–8.0 mm in size formed by densely intertwined fibers with an average diameter of 73 nm. The NFC pellets were pre-ground in the AGO-2S planetary ball mill for 5 min at an accele­ration of 15g and a NFC to ball mass ratio of 1:15. The average particle size of NFC after grinding was 3.9 µm.

According to the diagram of B₄C–TiB₂ system state at the eutectic point, the content of TiB₂ is ~26 mol. % [23]. The composite powder materials, the composition of which corresponds to the eutectic point and beyond it, were selected for research. The ratio of reagents was selected so that the composite powder materials containing 10, 20, 25, and 30 mol. % of TiB₂ were obtained in accordance with reaction (3). In calculating the batch composition, the presence of impurities in the composition of reagents was taken into account. The samples were designated as T10, T20, T25 and T30, respectively. The initial powders were mixed in a planetary ball mill for 5 min at an acceleration of 20g, and then they were sifted through a sieve with a mesh size of 100 µm.

The synthesis was performed in the VCh-25AV induction crucible furnace (Russia). Argon was chosen as an inert atmosphere preventing nitriding of boron carbide and titanium diboride. During carbide reduction of titanium dioxide, gaseous products (CO and CO₂) are released and the pressure in the system increases. To ensure the safety of the process, the synthesis should be performed in a flow reactor, ensuring continuous removal of the resulting gases by an argon flow. The temperature was controlled using the Kelvin Compact 2300 optical pyrometer (PC EUROMIX, Russia). The pressure in the reactor was almost atmospheric. The temperatures of the beginning of titanium dioxide reduction were determined by performing the thermodynamic calculations in accordance with the procedure [24]. The temperatures were calculated for different CO pressures, since it is difficult to estimate the partial pressure of CO in Ar + CO gas mixture.

The value of the isobaric-isothermal potential of the reaction of carbide reduction of titanium dioxide turns negative at the temperatures of 745, 849 and 994 °C for CO pressures of 0.001, 0.01 and 0.1 MPa, respectively.

In this research, the heat treatment of the batch was initially performed to prepare a mixture of B₄C–25 mol. % TiB₂ (T25) at t = 1200, 1400, 1650 and 1900 °C for 20 min in accordance with reaction (3). The completeness of the process was evaluated by weighing the batch and the reaction products, as well as by comparing experimental data with the theoretical ones. The estimated weight loss upon the completeness of this reaction was 19.05 wt. %. In practice, this value may slightly differ from the estimated value. This is due to the presence of impurities in the reagents used, as well as due to the possibility of the formation of aluminum borocarbide Al₃B₄₈C₂, being prone to oxidation and hydration, during the synthesis. Since the results of the conducted studies revealed that the optimal synthesis temperature is 1650 °C, further experiments with a batch of a different composition were performed at the same temperature for 20 min.

The X-ray phase analysis (XPA) of the obtained powders was performed by means of DRON-3 diffractometer using CuK_α-radiation. The diffraction patterns were interpreted using Power Diffraction File (PDF-2) database. The ratio of B₄C and TiB₂ phases was estimated using the corundum number method.

The total carbon content was determined by infrared absorption spectrometry using the CS 844 sulfur and carbon analyzer (LECO Corporation, the USA) as per GOST 12344-2003.

The microstructure of the powders and the morphology of the particles were studied using S-3400N scanning electron microscope (Hitachi, Japan) equipped with an energy-dispersive analysis attachment (Oxford Instruments Analytical, the United Kingdom). The particle size distribution was evaluated by means of laser particle size analyzer MicroSizer 201 VA Instrument (VA Instalt LLC, Russia). The surface area was determined by the method of low-temperature nitrogen adsorption using NOVA 2200e device (Quantachrome Instruments, the USA).

The thermal-oxidative stability of the samples was determined using STA 449 C Jupiter synchronous thermal analysis instrument (Netzsch, Germany). During the analysis, the sample was oxidized in an atmosphere of synthetic air when heated up to a temperature of 1000 °C at a rate of 15 °C/min.

Experiments on the production of B₄C–TiB₂ composite ceramics were performed on a hot pressing unit designed by the Institute of Automation and Electrometry, the Siberian Branch of the Russian Academy of Sciences (Novosibirsk), using a synthesized powder containing 30 mol. % of TiB₂. In this case, the batch was pre-ground in a planetary ball mill at an acceleration of 20g for 5 min at a mass ratio of the batch to balls of 1:30. The process was carried out in argon atmosphere at a pressing pressure of 25 MPa and a temperature of 2100 °C.

The relative density and open porosity of ceramics were evaluated in accordance with GOST 2409-2014 using AD-1653 hydrostatic weighing set installed on GR-300 analytical balance (AND, Japan).

Vickers microhardness measurements were performed on 402MVD unit (Wolpert Group, Great Britain). The indentation load was 500 g. At least 5 punctures were applied to the samples in such a way that the distance between the center of one indent and the edge of the next one was at least 2.5 lengths of the diagonal of the indent.

The fracture toughness was determined by indentation on a hardness tester of TP model No. 3534 (Russia) with an indenter in the form of a 4-sided diamond Vickers pyramid with a load of 5 kg. Its values were calculated according to equation [25]

\[{K_{1c}} = 0.048{\left( {\frac{l}{a}} \right)^{ - 0,5}}{\left( {\frac{{{H_v}}}{{E\Phi }}} \right)^{ - 0,4}}\frac{{{H_v}{a^{0,5}}}}{\Phi },\]

where l is the fracture length, µm; a is the half-diagonal of impression, µm; H_v is the microhardness, GPa; Е is the Young modulus, GPa; Ф = 3 is the constant.

Results and discussion

Fig. 1 shows the X-ray diffraction patterns of the synthesized samples of B₄C–25 mol. % TiB₂ mixture. It can be seen that at the synthesis temperatures t = 1200 and 1400 °C the peaks of both the target B₄C and TiB₂ phases and the unreacted carbon are observed for condensed products. At t = 1650 and 1900 °С, B₄C and TiB₂ phases are formed in the reaction products, and the X-ray diffraction patterns show the impurity reflections of Al₃B₄₈C₂ phase. Its presence is caused by the fact that NFC has aluminum oxide impurity Al₂O₃, which reacts with the components of the batch [20].

Fig. 2 presents the electron micrographs of mixture samples synthesized at the temperatures of 1400, 1650 and 1900 °C. The micrographs were taken in the mode of registration of secondary electrons. The scanning electron microscopy (SEM) images of a sample obtained at t = 1400 °C clearly show heterogeneous particles, some of which are fragmented. To clarify their nature, an elemental mapping was performed, which indicated that the particles constituted remains of unreacted NFC (Fig. 3). Besides, the energy- dispersive analysis data revealed the presence of oxygen in the amount of 5 wt. %.

The samples obtained at t = 1650 and 1900 °C have aggregated particles with smooth edges, the size of which does not exceed several micrometers. According to the energy-dispersive analysis, these samples contain titanium, boron, carbon, as well as nickel and aluminium (~1 wt. % in total).

The theoretical weight loss of the batch as a result of reaction (3) is 19.05 % at a ratio of reagents corresponding to 25 mol. % of TiB₂ in the resulting powder. The experimental weight loss was 0.9, 1.7, 19.5 and 19.4 % at the processing temperatures of 1200, 1400, 1650 and 1900 °C, respectively. It follows from the obtained results that the reaction of boride formation is fully completed at t = 1650 °C.

The results of granulometric analysis of the samples of B₄C –25 mol. % TiB₂ composition synthesized at t = 1650 and 1900 °C showed that the average particle size of the obtained powders increases from 8.4 to 9.8 μm upon an increase in the synthesis temperature. Since an increase in the particle size of the powder can lead to adeterioration in its sintering properties, the further experiments at t = 1650 °C were conducted.

To evaluate the effect of the mixture composition on the properties of the resulting powder, the batch with the composition corresponding to 10, 20, 25 and 30 mol. % of TiB₂ was heat-treated. The experimental weight loss of the batch during the synthesis was close to the theoretical value in all cases (the relative deviation did not exceed 3 %), which indicates the completeness of the synthesis process at t = 1650 °C, regardless of the composition of the sample. This was also confirmed by X-ray phase analysis data (Fig. 4). The diffraction patterns of the condensed reaction products contain TiB₂ and B₄C phases for all samples. The TiB₂ phase content estimated by the corundum number method was 9, 18, 24, and 29 mol. % for T10, T20, T25, and T30 samples, respectively. These data turned out to be close to the estimated values.

From the results of determining the total carbon content, presented in Table 1, it can be seen that the obtained experimental data slightly exceed the values corresponding to the given composition of the synthesized mixtures. This also indicates a complete synthesis process. It should be noted that with an increase in TiB₂ phase content in the powders, the excess of carbon decreases.

Fig. 5 shows the micrographs of the samples of composite powder materials with different TiB₂ contents. All SEM images contain aggregated particles of several micrometers in size, and the absence of fragmented particles indirectly bespeaks of the absence of unreacted particles of the initial components of the reaction mixture.

In the course of particle size analysis, the samples of B₄C–TiB₂ powders were subjected to ultrasonic dispersion at a power of 200 W for 30 s. Two peaks were found in the particle size distribution histograms of T10 and T30 samples (Fig. 6), with the second peak increasing upon an increase in TiB₂ phase content. Since the ratio of the heights of the first and the second maxima on the bimodal curve changes with an increase in the concentration of titanium diboride in the synthesized mixture, it can be assumed that the part of the histogram with a smaller particle size mainly characterizes the B₄C phase; consequently, its other part with a larger particle size refers to the TiB₂ phase. Based on this assumption, the average size of particles and aggregates was calculated for each phase (Table 2), and the values of standard deviations and asymmetry indices were determined using method [26].

Table 2 shows that the average 50 % particle size increases with an increase in the TiB₂ content of the powders under research. There is also an increase in the particle size of B₄C phase compared to pure B₄C (2.4 µm). The standard deviation values indicate a wide range in particle size distribution, i.e. the powder is polydisperse. The low value of asymmetry degree proves the symmetry of the distribution curves for each phase. The largest value of the average particle size of B₄C and TiB₂ phases is typical for the sample containing 30 mol. % of TiB₂.

The surface area values were 5, 4, 3, and 3 m²/g for T10, T20, T25, and T30 samples, respectively, whereas the said value was 4 m²/g for the initial boron carbide sample without modifying additives.

In order to determine the thermal-oxidative stability of the obtained B₄C–TiB₂ powders, they were oxidized in a synthetic air atmosphere. Similar thermogravimetric curves were obtained for all samples of different composition. The derivatogram of T10 sample is presented in Fig. 7 as an example.

X-ray phase analysis was conducted to identify the products of oxidation of the mixture with oxygen. The diffraction pattern of the sample of composite powder material after heating up to 1000 °C in an oxidizing atmosphere is shown in Fig. 8.

The results of thermogravimetric analysis show that the weight gain is caused by the oxidation process starting at t ~500 °C. Upon the temperature reaching 1000 °C, there are unoxidized B₄C and TiB₂ phases, as well as TiBO₃, TiO₂ and B₂O₃ oxidation products present in the samples. It can be assumed that when this temperature is reached, the process proceeds in accordance with the following reactions

Upon that, the oxidation of minimum 80 wt. % of composite powder material occurs. The mass fraction of the oxidized substances at t = 1000 °C is 80, 75, 69 and 73 wt. % for T10, T20, T25 and T30 samples, respectively, and 83 wt. % for the initial boron carbide. The incomplete oxidation of the samples can be explained by the formation of a liquid protective film of B₂O₃, the melting temperature of which is ~450 °C, on the surface of B₄C and TiB₂ particles [27].

The synthesized powder containing 30 mol. % of TiB₂ was selected for the preparation of composite ceramics. The relative density of the obtained material was 99.0±1.1 %, and the relative density of B₄C ceramics produced in a similar way without the use of modifying additives was 97.7±0.5 %.

Thus, the use of a batch with В₄С–30 mol. % TiB₂ composition obtained by carbide reduction allows to produce ceramics with a high relative density. Its structure consists of a boron carbide matrix (gray area) and light inclusions of titanium diboride of various sizes (Fig. 9).

The microhardness of the composite ceramics was 33.0±3.4 GPa, and the fracture toughness was 5.0±0.2 MPa∙m\(^{0.5}\); for ceramics without TiB₂ additives, these indices were 45.5±5.2 GPa and 3.6±0.11 MPa∙m\(^{0.5}\), respectively. Thus, the presence of a modifying additive in the composition of ceramics naturally led to a decrease in microhardness and an increase in the material fracture toughness.

Conclusion

B₄C–TiB₂ composite powder materials have been obtained by the carbide reduction of titanium dioxide using an excess of boron carbide and nanofibrous carbon. It has been established that the process of formation of the TiB₂ phase starts at t = 1200 °C, but it is fully completed at 1650 °C. a further increase in a temperature leads to an increase in the particle size of B₄C–TiB₂ powder. The average size of 50 % particles of the composite powder material containing 10–30 mol. % of TiB₂ is 15 µm maximum, and the surface area value does not exceed 5 m²/g. The average particle size of the B₄C phase is in the range of 5.3–5.5 µm, and that of the TiB₂ phase is 33.6÷41.9 µm.

The oxidation of the obtained mixtures with atmospheric oxygen starts at t ~500 °C. Upon that, maximum 80 wt. % of the powders under study are oxidized when the temperature reaches 1000 °C.

The presence of 30 mol. % of TiB₂ in the composite powder material allows to perform the hot pressing production of the ceramics with a higher relative density (99.0±1.1 %) and fracture toughness (5.0±0.2 MPa∙m\(^{0.5}\)) as compared to the ceramics obtained in a similar way only from B₄C.