Self-propagating high-temperature synthesis of highly dispersed Si₃N₄–TiC composition using sodium azide and various carbon sources

Introduction

Silicon nitride ceramics exhibit excellent mechanical and thermal properties, making them an ideal material for high-temperature applications such as aerospace structural components and turbine engines [1]. Their microstructure after sintering resembles that of composites reinforced with β-Si₃N₄ whiskers, which act as reinforcing elements [2–4]. Adequate fracture toughness, high hardness, and good wear resistance are critical characteristics for advanced ceramics, which find applications in cutting tools and automotive components such as cam rollers and ball bearings in diesel engines [5–7]. Recently, silicon nitride ceramics have attracted significant interest due to their high mechanical properties, wear resistance, and corrosion resistance [8–12]. Thermal shock resistance is a key property for their high-temperature applications. However, most studies on silicon nitride nanocomposites focus on optimizing fracture toughness and flexural strength. To expand the application areas of silicon nitride ceramics, improvements in shock resistance and high-temperature creep resistance are essential [13].

Considerable efforts have been made to improve the mechanical properties of Si₃N₄ by controlling the microstructure or creating various types of composites. During the densification process, β-Si₃N₄ fibers significantly increase fracture toughness as cracks interact with these large grains [14].

Tensile strength is widely recognized as one of the most important factors for enhancing the thermal stability of ceramics [15; 16]. Incorporating second-phase particles into a ceramic matrix can enhance the mechanical properties of ceramics. Studies have also shown that the addition of a secondary phase can improve crack initiation and propagation resistance in various ways [17–20]. For example, adding TiC particles to a silicon nitride matrix enhances mechanical performance, thermal shock behavior, and fatigue resistance, with an optimal TiC content of 10 wt. % [13]. Other researchers [21–23] have also reported the influence of TiC particles on the Si₃N₄ matrix.

Moreover, silicon nitride is extremely hard and non-conductive, making machining with conventional diamond tools challenging and costly, significantly increasing the final cost of ceramic components. Consequently, new electrically conductive composites based on silicon nitride have been developed for more cost-effective electrical discharge machining by incorporating certain amounts of electrically conductive particles such as TiC, TiN, or TiCN into the ceramic matrix [24; 25]. For instance, a Si₃N₄ + TiN composite with critical TiN content can be machined using inexpensive electrical discharge machining [26].

A Si₃N₄–TiC nanocomposite with high mechanical properties was obtained by hot pressing with the addition of 10 wt. % nanosized Si₃N₄ particles and 15 wt. % TiC to a submicron Si₃N₄ matrix, using Al₂O₃ and Y₂O₃ as sintering aids. Layered composites demonstrated high strength, fracture toughness, and wear resistance due to the presence of compressive surface stresses in the layers. A ceramic nanocomposite Si₃N₄–TiC was fabricated using a Si₃N₄ micro-matrix with nanosized Si₃N₄ and TiC particles. Cutting tools made from this ceramic exhibited better wear resistance than those made from sialon. Wear occurred mainly through abrasion and adhesion, whereas sialon cutting tools predominantly experienced abrasion, adhesion, thermal cracking, and delamination [27; 28].

The addition of secondary phases, namely the production of composites with ceramic matrices, offers many significant advantages, such as improved fracture toughness compared to unreinforced ceramics [29; 30]. Moreover, recent studies have shown that in-situ phase formation provides additional opportunities compared to composites produced using traditional ex-situ methods. The main advantages of in-situ manufacturing methods include enhanced mechanical properties, the ability to achieve unique microstructures, process simplicity, and inexpensive raw materials [27].

One promising in-situ technology is the self-propagating high-temperature synthesis (SHS) process, which enables the production of a wide range of refractory compounds, including nitrides and carbides, by utilizing the heat released during combustion in simple, compact equipment with short processing times [30].

To produce highly dispersed (d < 1 μm) Si₃N₄–TiC powder compositions, the authors of this article investigated the use of azide synthesis, a variation of SHS where sodium azide (NaN₃) powder serves as the nitriding agent, and various activating halide salts are used alongside elemental reactants. This approach results in relatively low combustion temperatures, the formation of numerous intermediate vapor-gas reaction products, and final by-products consisting of condensed and gaseous phases. These by-products separate the target powder particles and prevent their agglomeration into larger particles.

The study summarizes the results of azide SHS synthesis of Si₃N₄–TiC ceramic compositions with various nitride-to-carbide phase ratios of Si₃N₄:TiC = 4:1, 2:1, 1:1, 1:2, and 1:4, according to the following stoichiometric equations, using halide salts such as Na₂SiF₆ and (NH₄)₂SiF₆ [33; 34].

Si–Ti–NaN₃–Na₂SiF₆–C system

Si–Ti–NaN₃–(NH₄)₂SiF₆–C system

In these stoichiometric reactions, the composition of the reaction products is expressed in moles. When converted to weight percent (wt. %), the following ratios are obtained for the expected theoretical composition of the target Si₃N₄–TiC compositions after the removal of the water-soluble byproduct salt, NaF, from the condensed reaction products:

The results of the phase composition analysis of the washed solid combustion products from reactions (1)–(10), determined experimentally, are presented in Table 1. In most cases, the products consist of a highly dispersed powder with a complex composition, appearing as submicron equiaxed particles ranging in size from 100 nm to 1 μm (Fig. 1).

From the data in Table 1, it is evident that the experimental phase composition of the products from azide SHS differs significantly from the expected theoretical composition of Si₃N₄–TiC powder compositions. The content of the target carbide phase TiC is significantly lower than its theoretical value (ranging from 2.7 to 19.9 %), the amount of Si₃N₄ is excessive, and an undesirable secondary phase, titanium nitride, is present (ranging from 1.9 to 19.2 %). These results can be attributed to the fact that very fine and lightweight particles of technical carbon may be partially or completely removed from the burning, highly porous bulk charge sample by gases released during the decomposition of sodium azide and halide salts, preventing their participation in the formation of titanium carbide. As a result, silicon nitride forms in larger quantities, and titanium nitride forms due to an excess of nitrogen (since combustion in a nitrogen atmosphere is essential for in-situ nitride formation in SHS compositions), while titanium carbide forms in smaller amounts than predicted by the initial stoichiometric reaction equations and thermodynamic calculations. Additionally, the synthesized compositions may contain impurities of unreacted free silicon (up to 1.9 %) or carbon (up to 1.5 %).

To address these shortcomings, several directions for further research on applying the SHS process to produce highly dispersed Si₃N₄–TiC compositions can be pursued. The simplest approach is to use polytetrafluoroethylene (PTFE (C₂F₄)_n) as an activating and carbon-containing additive in the charge, promoting the formation of TiC, as successfully demonstrated in previous studies [35; 36].

In this context, the aim of the present study was to maximize the convergence of the theoretical and experimental compositions of the Si₃N₄–TiC powder composition by modifying the initial reactant composition with full or partial replacement of carbon with PTFE and optimizing the conditions of the azide SHS process.

Research methodology

To synthesize the target Si₃N₄–TiC composition with phase molar ratios ranging from 2:1 to 1:4, chemical reaction equations involving full (11)–(14) and partial (15), (16) substitution of carbon with polytetrafluoroethylene (PTFE) were used:

To achieve the Si₃N₄–TiC composition with the maximum titanium carbide phase content (Si₃N₄:TiC = 1:4), carbiding mixtures with increased PTFE content were also used:

Thermodynamic calculations were conducted using the Thermo software [37] to predict the feasibility of combustion reactions by determining thermal effects (enthalpies), adiabatic temperatures, and the compositions of synthesis products.

The following raw materials were used in the experiments:

– silicon powder, grade Kr0 (main substance content ≥98.8 wt. %, average particle size d = 5 μm;

– titanium powder, grade PTOM-1 (98.0 wt. %, d = 30 μm);

– sodium azide powder, analytical grade (≥98.71 wt. %, d = 100 μm);

– carbon black, grade P701 (≥99.7 wt. %, d = 70 nm, agglomerates up to 1 μm);

– polytetrafluoroethylene, grade PN-40 (≥99.0 %, d = 40 μm).

Combustion of the starting reactant mixtures (charge) with a bulk relative density of 0.4 was conducted in a paper crucible with a diameter of 30 mm and a height of 45 mm. The experiments were performed in a laboratory SHS-Az reactor with a volume of 4.5 L, equipped with two thermocouples, under a nitrogen pressure of 4 MPa. The thermocouples were used to measure combustion temperatures and calculate the combustion rate. The pressure variation in the reactor during the combustion process was monitored with a pressure gauge.

The synthesized product was weighed and compared to the theoretical yield calculated from reactions (11)–(19). The combustion product was washed with water to remove water-soluble impurities, and the pH of the wash water was measured to assess the presence of free sodium in the combustion product and the completeness of the chemical reaction. The phase composition of the synthesized combustion products was determined using an automated ARL X’trA X-ray diffractometer (Thermo Scientific, Switzerland). CuK_α radiation was employed with continuous scanning over the angular range of 2θ = 20–80° at a scan rate of 2°/min. The resulting spectra were processed using the WinXRD software package. The surface topography and powder particle morphology were examined using a JSM-6390A scanning electron microscope (Jeol, Japan) equipped with a Jeol JED-2200 energy-dispersive spectroscopy attachment.

Results and discussion

The results of thermodynamic calculations for reactions (11)–(19) performed using the Thermo software are presented in Tables 2–4.

From the presented data, it is evident that all reactions exhibit high adiabatic temperatures sufficient for the realization of the SHS process in a combustion mode. The reaction products contain all the phases corresponding to the right-hand sides of equations (11)–(19), including the target phases of silicon nitride (Si₃N₄) and titanium carbide (TiC), the water-soluble byproduct salt NaF, and minor impurities of free silicon (Si) and titanium (Ti).

The results of the microstructural analysis of the combustion products of the initial powder mixtures (charges) from reactions (11)–(19) after washing with water to remove the byproduct salt NaF are shown in Fig. 2.

As shown in Fig. 2, the combustion products of charges from reactions (11)–(19) consist of highly dispersed equiaxed particles, comprising a mixture of nanosized (less than 100 nm) and submicron (100–500 nm) particles of titanium carbide and titanium nitride, as well as silicon nitride fibers with diameters of 50–200 nm and lengths of up to 5 μm.

The results of X-ray phase analysis for the washed combustion products of systems with the maximum titanium carbide phase content (Si3N4 :TiC = 1:4) are presented in Fig. 3.

The results of quantitative processing of the XRD spectra presented in Fig. 3 are summarized in Table 5. These results show the phase content in the washed combustion products of charges with the maximum titanium carbide phase fraction (Si₃N₄:TiC = 1:4) under various conditions: full replacement of carbon black with PTFE (reaction 14), carbiding mixtures with minimal PTFE content (reaction 16), and carbiding mixtures based on reactions (17)–(19) to synthesize the Si₃N₄–TiC composition. The experimental data are compared with theoretical phase compositions of target products Si₃N₄ and TiC based on the stoichiometric equations (11)–(19).

From the data in Table 5, it is evident that using PTFE as a carbon source for synthesizing nitride-carbide compositions via azide SHS is a promising research direction. A comparison of the experimental phase compositions of combustion products from charges (11)–(14) with theoretical values indicates that full replacement of carbon black with PTFE is not advisable, although the titanium carbide phase content increases compared to azide SHS without PTFE. For example, the titanium carbide content in the combustion products of charge (5): 2Si + 4Ti + 4NaN₃ + Na₂SiF₆ + 4C with the maximum TiC content (Si₃N₄:TiC = 1:4), is 19.9 wt. %. With full replacement of carbon black by PTFE in charge: 3Si + 4Ti + 8NaN₃ + 2C₂F₄) the titanium carbide content increases to 31.0 wt. %. However, partial replacement of carbon black and its combined use with PTFE as a carbon source allows the titanium carbide content to reach 52.3 wt. % in the combustion products of charge (16): 3Si + 4Ti + 0.64NaN₃ + 3.68C + 0.16C₂F₄ + 1.04N₂.

The best results were obtained using carbiding mixtures with increased PTFE content according to equations (17)–(19), where the TiC content in the experimental products ranged from 58.6 to 61.7 wt. %. Furthermore, the use of PTFE reduced the content of the secondary phase, titanium nitride, to 2.0–4.0 wt. % in the products of carbiding mixtures (17)–(19).

Conclusion

The presented results demonstrate that the SHS technology can make a significant contribution to the development of methods for producing highly dispersed Si₃N₄–TiC nitride-carbide compositions. The SHS process is attractive for its simplicity and cost-effectiveness and is one of the promising in-situ chemical methods for the direct synthesis of ceramic powders within the desired composition from a mixture of inexpensive starting reagents.

Traditional azide SHS using NaN₃ and gasifying halide fluorides, such as Na₂SiF₆ and (NH₄)₂SiF₆, is characterized by comparatively low combustion temperatures, the formation of large amounts of intermediate vapor and gaseous reaction products, as well as final byproduct condensed and gaseous products that separate the target powder particles. This enabled the synthesis of a highly dispersed (<1 μm) Si₃N₄–TiC powder composition, with Si₃N₄ predominantly in the α-modification phase during the combustion of all studied mixtures.

However, in all cases of traditional azide SHS application, the amount of TiC phase synthesized in the experiments was significantly lower than expected. Additionally, all synthesized compositions contained the TiN phase, with its content exceeding that of titanium carbide in mixtures without PTFE additives. Furthermore, the synthesized compositions may include impurities of unreacted free silicon (up to 3.0 wt. %).

The use of PTFE as an activating and carbiding additive with partial replacement of carbon in the mixtures (15)–(19) in azide SHS eliminated, in most cases, the shortcomings of the traditional approach for various ratios of target phases Si₃N₄ and TiC. This allowed for the synthesis of highly dispersed Si₃N₄–TiC powder compositions with a phase composition significantly closer to the calculated theoretical composition.