Self-propagating high-temperature synthesis of a highly dispersed Si₃N₄–SiC ceramic powders composition using sodium azide and polytetrafluoroethylene

Introduction

Refractory ceramic composite materials of the silicon nitride–silicon carbide (Si₃N₄–SiC) system possess a wide range of valuable properties and are used in various industrial applications. Primarily, these materials are excellent refractories and can be used at high temperatures (up to 1500 °C) due to their outstanding oxidation and creep resistance, chemical stability, low thermal expansion coefficient, and thermal shock resistance [1]. Such refractories are typically produced by reaction bonding of SiC powder with Si₃N₄ particles synthesized via direct nitridation of silicon in a pressed powder mixture of SiC and Si under a nitrogen atmosphere at elevated temperatures (1300–1450 °C) for several hours [1; 2]. The resulting product is a two-phase mixture of α-Si₃N₄ and β-Si₃N₄. The density and mechanical strength of the composites increase with a higher content of the fibrous α-Si₃N₄ phase, which is promoted by reducing the particle sizes of SiC and Si, as well as by the presence of trace oxygen impurities in the nitrogen atmosphere. This oxygen leads to intensive oxidation of silicon, formation of SiO vapor, and gas-phase synthesis of α-Si₃N₄ [1; 2]. While commercially produced silicon carbide refractories (without Si₃N₄ in their composition) exhibit a compressive strength of up to 100 MPa at approximately 20 % porosity [3], a reaction-bonded composite with a composition of 46SiC–50Si₃N₄–4Y₂O₃ (wt. %) demonstrates a compressive strength of 319 MPa at 26 % porosity [1].

Si₃N₄–SiC composite refractories are also produced by other methods such as semi-dry pressing (with up to 8 % liquid binder), plastic forming (up to 24 % binder), and slip casting (with up to 80 % water), followed by drying and sintering at temperatures up to 1600 °C [4–6]. In these methods, the starting materials are typically mixtures of polydisperse Si₃N₄ and SiC micropowders, with particle sizes ranging from fine fractions (1–5 µm) to coarse ones (50–200 µm), to achieve optimal packing density in both the powder mixture and the final composite. However, the use of such relatively coarse powders, combined with a porosity of approximately 20 %, limits the ability of Si₃N₄–SiC refractories to reach their maximum mechanical strength, particularly at elevated temperatures.

To use Si₃N₄–SiC composites as structural heat-resistant materials under high mechanical loads at elevated temperatures (e.g., in gas turbine engines), or as ultra-hard tooling and wear-resistant materials, it is essential to achieve the highest possible values of hardness, strength, and fracture toughness. To achieve near-theoretical density and reduce porosity to below one percent, alternative fabrication methods must be employed. These include pressureless sintering of Si₃N₄ and SiC powder mixtures at temperatures up to 2000 °C (including liquid-phase sintering), hot pressing, hot isostatic pressing, spark plasma sintering, and additive manufacturing techniques [5; 7–10]. Numerous studies utilizing these methods have shown that the mechanical properties of Si₃N₄–SiC composite ceramics improve significantly with increased fineness of the constituent powders – that is, as particle sizes are reduced from the micron scale to highly dispersed (<1 µm), submicron (0.1–1.0 µm), and even nanoscale (1–100 nm) dimensions. Consequently, considerable research efforts are currently focused on the development of nanostructured Si₃N₄–SiC ceramic composites [11–15]. In addition to enhanced mechanical performance, Si₃N₄–SiC nanocomposites exhibit improved microwave absorption properties, making them promising candidates for use as lightweight functional materials for electromagnetic shielding in aerospace and aviation applications [16].

At first glance, the ex situ fabrication of Si₃N₄–SiC nanocomposites – based on mechanical mixing of pre-synthesized Si₃N₄ and SiC nanopowders followed by compaction and sintering – appears to be the most straightforward approach. However, such nanopowders are extremely costly, as they cannot be produced by simple mechanical milling of inexpensive coarse powders, but instead require synthesis via various complex and expensive chemical routes [5]. Furthermore, nanoparticles exhibit pronounced interparticle adhesion, forming dense agglomerates that are difficult to disperse through mechanical mixing, thereby impeding the formation of a homogeneous nanopowder blend. This lack of homogeneity often results in structural defects during densification and sintering, including chemical inhomogeneity, grain size non-uniformity, porosity, and crack formation [14]. From this standpoint, when dealing with highly dispersed powders, in situ chemical synthesis methods are preferable. These approaches enable the direct formation of the desired ceramic powder composition from precursor reagents, eliminating the drawbacks associated with ex situ mechanical mixing of pre-synthesized ceramic powders [1; 14].

Several chemical synthesis methods have been developed for producing submicron and nanopowder compositions of Si₃N₄–SiC. These include pyrolysis of silicon-containing polymers, methane pyrolysis followed by coating of Si₃N₄ particles with highly dispersed carbon, carbothermal reduction of silicon dioxide in the presence of gaseous nitrogen, gas-phase reactions, and plasma-chemical synthesis [14–21]. However, these methods are characterized by high energy consumption due to the need for elevated temperatures and prolonged heat treatment, as well as the use of expensive equipment and reagents. As a result, such techniques remain largely confined to laboratory-scale production of composite powders and have not yet been transferred to industrial-scale manufacturing. Currently, commercial suppliers of ceramic powders continue to produce primarily single-phase powders, and most nanocomposite ceramic powders are still obtained by conventional methods involving the mixing and milling of their constituent single-phase powders [14; 15]. Nevertheless, chemical synthesis methods for composite powders are regarded as advanced and promising, warranting further development and industrial implementation. Once this goal is achieved, high-quality nanocomposite powders will become commercially available, which is expected to have a positive impact on the performance characteristics of the resulting composite ceramics [14; 15].

Among the chemical methods used to synthesize submicron and nanoscale Si₃N₄–SiC powder compositions, self-propagating high-temperature synthesis (SHS) stands out for its simplicity and energy efficiency. This method is based on the combustion of inexpensive precursor reagents [22–24]. In the simplest case, combustion of a powder mixture of silicon (particle size d < 15 µm) and technical carbon (d < 1 µm) in gaseous nitrogen at a pressure of 1–7 MPa – that is, in the Si–nC–N₂ system – can be initiated when the carbon black content does not exceed n = 0.8. The combustion product is a Si₃N₄–SiC powder composition containing 5–60 wt. % SiC, with the remainder comprising silicon nitride predominantly in the β-phase, along with a minor amount of unreacted silicon. Most of the resulting particles are 2–3 µm in size, though larger particles up to 15 µm are also present, typically in the form of columnar crystals or agglomerates of finer particles.

The addition of powdered polytetrafluoroethylene (PTFE) (C₂F₄)_n in small amounts (5–15 wt. %) acts as an activating agent, extending the combustion limits of the Si–C–N₂ system and enabling the synthesis of composites with a wide range of Si₃N₄: SiC ratios (from 0 to 100 %). The resulting powders exhibit micron-scale particle sizes and a low fraction of the α-Si₃N₄ phase [22]. A more advanced approach involves azide self-propagating high-temperature synthesis (azide SHS), applied to systems based on Si–C–NaN₃–halide salts. In this method, silicon (Si), technical carbon (C), sodium azide (NaN₃) as the nitriding agent, and halide salts (NH₄F, Na₂SiF₆, and (NH₄)₂SiF₆) as activating, gasifying additives are used. Combustion of these bulk powder mixtures in a 4.5 L reactor under 4 MPa of nitrogen produces highly dispersed Si₃N₄–SiC powder compositions containing 1.6–41.8 wt. % SiC. Compared to earlier combustion-based SHS products, these compositions feature finer particle sizes (mainly in the range of 100–500 nm) and a markedly higher content of the α-Si₃N₄ phase – up to ten times greater than that of β-Si₃N₄ [23–25]. However, the experimental compositions deviated significantly from the theoretical ones, with an excess of silicon nitride, a reduced fraction of silicon carbide, and notable levels of free silicon (up to 5.7 wt. %) and free carbon (up to 5.1 wt. %) as impurities.

It is well known that the synthesis reaction of silicon carbide from elemental silicon and technical carbon powders (Si + C = SiC) is weakly exothermic and cannot proceed in combustion mode under standard conditions [26]. However, the reactivity of the Si + C mixture can be enhanced through preheating, application of an electric field, mechanical activation, chemical activation, or by using a gaseous nitrogen or air environment, enabling the synthesis of SiC via combustion. One of the simplest and most effective approaches involves the use of polytetrafluoroethylene (C₂F₄)_n as a chemical activator. This method enables complete reaction of the mixture Si + 0.9C + 0.05C₂F₄ → SiC + 0.1F₂ in a nitrogen atmosphere at 3 MPa, resulting in the formation of SiC particles with an average size of ~200 nm [27]. The reaction stages and structure formation in the combustion wave of Si–C–C₂F₄ mixtures with relatively high PTFE content (C₂F₄: C = 0.5÷3.0) were studied in [28] by burning pressed powder mixtures in an argon atmosphere at 0.5 MPa. The combustion products, in addition to silicon carbide, contained 16–33 % residual silicon (necessary for subsequent Si₃N₄ formation in Si₃N₄–SiC composite synthesis) and consisted of porous agglomerates 10–20 µm in size, incorporating crystallized silicon droplets and SiC grains measuring 0.3–0.7 µm. Furthermore, PTFE can be used not only as an activating additive but also as a carbiding reagent in place of technical carbon during the combustion of a bulk Si + C₂F₄ mixture in gaseous argon at 0.5 MPa and the synthesis of silicon carbide in the form of fibers (100–500 nm in diameter) and equiaxed particles (0.5–3.0 µm), which aggregate into porous agglomerates [29]. However, the overall SiC yield in this case is quite low – only about 10 % of the charge reacts to form cotton-like (fluffy) SiC, while the remainder consists of a black powder of unreacted carbon and silicon. This result is attributed to the following sequence of reactions involved in the formation of silicon carbide with PTFE (C₂F₄) [28; 29]:

At the first stage (1), PTFE undergoes exothermic decomposition in the preheating zone, producing gaseous fluorides and solid carbon particles. During the intermediate stages (2) and (3), the gaseous fluorides react with each other and with molten silicon. In the final stage (4), silicon and carbon particles – originating both from the initial soot and from PTFE decomposition – react to form the target silicon carbide. Completion of all these stages is essential to obtain highly dispersed SiC; if only the first stage occurs, the result is merely the formation of gaseous fluorides and soot particles [29]. Combustion involving PTFE proceeds at a high rate and is accompanied by intense gas evolution, resulting in the scattering of the charge components. Under these conditions, silicon particles are unlikely to react with carbon particles, as such a reaction is improbable in the gas phase [29]. Therefore, when using PTFE, the charge should be pressed into briquettes – rather than used in bulk form – preferably with a diameter of at least 30 mm, and the combustion should be conducted under elevated gas pressure in an SHS reactor to prevent the escape of PTFE decomposition products from the reaction zone [27–30].

As demonstrated in our previous study [31], this approach proved effective in the azide SHS of another highly dispersed powder composition, AlN–SiC, using PTFE. Partial replacement of carbon in the carbiding mixture with 0.9C + 0.05C₂F₄ eliminated, in most cases and across various target AlN/SiC phase ratios, the drawbacks associated with conventional azide SHS using halide salts such as (NH₄)₂SiF₆, AlF₃, and NH₄F. While maintaining the high dispersity of the synthesized AlN–SiC powder compositions, their phase composition – especially when pressed charges were used – became significantly closer to the intended theoretical composition. The SiC phase content increased substantially, while undesirable secondary phases such as silicon nitride and the water-insoluble salt cryolite (Na₃AlF₆) were eliminated.

An attempt to synthesize the target stoichiometric composition of the highly dispersed Si₃N₄–SiC nitride–carbide system by fully replacing carbon with PTFE in the initial azide SHS reagent mixture was unsuccessful [32]. In bulk mixtures, only small amounts of the desired Si₃N₄ and SiC phases were formed, with free carbon being the predominant combustion product. This result aligns with the findings reported in [29] and is attributed to the combustion proceeding only through the initial stages – specifically, reactions (1) and (2), which involve PTFE decomposition and the generation of gaseous silicon fluorides and free carbon. In pressed mixtures, a substantial fraction of the silicon particles was able to react with carbon (reaction 4), leading to increased formation of the target Si₃N₄ and SiC phases. However, due to continued scattering of the charge components during combustion, a significant portion of the silicon remained unreacted. As a result, the amount of synthesized SiC was considerably lower than the theoretical value predicted by the stoichiometric equations.

In light of these findings and building upon previous studies [23–25; 32], the present work aimed to increase the SiC content and bring the composition of the highly dispersed Si₃N₄–SiC material closer to its theoretical target. To this end, we investigated the azide SHS process involving partial substitution of carbon with PTFE in the initial reagent mixture – specifically, the combustion process and resulting products of the Si–NaN₃–C–C₂F₄ system.

Research methodology

To investigate the azide SHS process for synthesizing highly dispersed Si₃N₄–SiC compositions with partial substitution of carbon by PTFE in the initial charge (reagent mixture), the following starting materials were used (wt. % unless otherwise noted): silicon powder, grade Kr00 (≥99.9 %, d = 40 µm); sodium azide powder, analytical grade (≥98.71 %, d = 100 µm); polytetrafluoroethylene (PTFE), grade PN-40 (≥99.0 %, d = 40 µm); and technical carbon black, grade P701 (≥88.0 %, d = 70 nm, in the form of agglomerates up to 1 µm).

According to [22], achieving a high SiC content in silicon nitride-based composites requires partial replacement of technical carbon with an activating carbiding additive – polytetrafluoroethylene (PTFE) – in amounts of 5, 10, and 15 %. This yields carbiding mixtures of technical carbon and PTFE with the following compositions, each equivalent to 1 mol of carbiding carbon:

In this azide SHS approach, the strongly nitriding reagent sodium azide (NaN₃) is added to the charge in quantities sufficient to neutralize the fluorine released during the complete decomposition of PTFE and bind it as water-soluble sodium fluoride (NaF), which can subsequently be removed from the SHS product by washing with water. As a result, the stoichiometric equations for the azide SHS of Si₃N₄–SiC powder compositions for five molar ratios of the target phases (Si₃N₄: SiC = 4:1, 2:1, 1:1, 1:2, 1:4) using carbiding mixtures (A)–(C) containing PTFE and combusted in a nitrogen atmosphere are as follows:

– for carbiding mixture (A):

– for carbiding mixture (B):

– for carbiding mixture (C):

The reagent mixtures corresponding to equations (5)–(19), each with an average mass of 22 g, were combusted in a 4.5 L azide SHS reactor under an initial nitrogen pressure of P₀ = 3 MPa. Combustion was performed in two forms: as a bulk charge, placed in a tracing-paper crucible (30 mm in diameter and 45 mm in height), and as briquetted charges, compacted under 7 MPa into cylindrical pellets measuring 30 mm in diameter and approximately 22 mm in height. Combustion was initiated using a tungsten spiral heater. The maximum gas pressure (P_max), generated during combustion was recorded via a manometer. After cooling, the combustion products were removed from the reactor, ground into free-flowing powder using a porcelain mortar, and washed with water to remove the byproduct – sodium fluoride (NaF). The dried and washed combustion product was weighed, and the mass loss (∆m, %) was determined as the difference between the initial charge mass (m₀) and the final product mass (m_f). This mass loss was interpreted as deviation of reactants and products caused by intense combustion. The phase composition of the synthesized products was analyzed using an ARL X’TRA powder X-ray diffractometer (Thermo Fisher Scientific, Switzerland) equipped with a copper anode X-ray tube. Diffraction patterns were processed and the quantitative phase composition determined by the Rietveld refinement method using HighScore Plus software and the COD-2024 crystallographic database. The morphology and particle size of the synthesized powders were examined using a JSM-6390A scanning electron microscope (JEOL, Japan).

Results and discussion

In chemical equations (5)–(19), the composition of the reaction products is expressed in moles, while in the experiments, it is given in weight percent. When converting the molar ratios of silicon nitride to silicon carbide to weight percentages, the following theoretical ratios of the target Si₃N₄–SiC compositions are obtained, assuming complete removal of the water-soluble byproduct sodium fluoride (NaF) from the products of reactions (5)–(19):

The experimentally determined combustion parameters – including the maximum pressure in the reactor (P_max) and mass loss (Δm) – for the initial bulk and pressed powder charges corresponding to reactions (5)–(19), along with the phase compositions of the washed solid reaction products, are summarized in the Table.

The Table shows that the bulk charges corresponding to reactions (5)–(8) did not combust, while combustion of the charge from equation (9) did not yield the target phases Si₃N₄ and SiC. Instead, the combustion products were merely a mixture of free silicon and carbon. These results can be explained by the fact that reactions (5)–(9) used carbiding mixture (A), which contains the lowest amount of the activating additive – polytetrafluoroethylene (0.05C₂F₄ per 1 mole of carbiding carbon). At the same time, these charges included a relatively large amount of poorly reactive silicon powder – ranging from 4 to 13 moles. Only the charge in equation (9), which had the highest PTFE content (0.2C₂F₄), was able to sustain combustion. However, under bulk conditions, this reaction yielded only a mixture of Si and C particles, consistent with the findings reported in [29]. This outcome is attributed to the progression of only reaction (1) – the exothermic decomposition of PTFE – as the first stage of the process. The table also shows that the pressed charges from equations (5)–(7), which also used 0.05C₂F₄ as the activating additive, did not combust. Meanwhile, the charges from equations (8) and (9), containing 0.1C₂F₄ and 0.2C₂F₄ respectively, did combust, but the resulting phase composition deviated significantly from the theoretical one. The products contained considerable amounts of free silicon and less SiC than expected. The recorded maximum gas pressures in the reactor – P_max = 3.00 and 3.29 MPa – were either equal to or only slightly higher than the initial pressure P₀ = 3.00 MPa. The relatively small mass losses during combustion (Δm = 15.0 and 9.9 %) also point to the low combustion intensity of charges using carbiding mixture (A). (The observed match between the maximum gas pressure and the initial pressure – P_max = P₀ = 3.00 MPa – can be by the concurrent increase in nitrogen gas pressure in the reactor due to the combustion-induced temperature rise and the simultaneous decrease in the amount of gaseous nitrogen caused by its significant uptake during the formation of silicon nitride).

It is worth noting that in a similar case of synthesizing a different composition – AlN–SiC – via azide-assisted SHS with the use of PTFE, all bulk and pressed charges based on carbiding mixture (A) underwent combustion, and did so intensively, reaching maximum reactor pressures ranging from 3.51 to 4.38 MPa. These conditions yielded good results in the formation of the target phases AlN and SiC [31]. This outcome is explained by the fact that, in the charges intended for AlN–SiC synthesis, the 0.05C₂F₄ PTFE additive was combined with only 1 mol of poorly reactive Si powder and 1 to 4 mol of highly reactive Al powder.

The Table shows that when using carbiding mixture (B), which contains an increased amount of the activating PTFE additive – 0.1C₂F₄ per 1 mol of carbiding carbon – in the charges for reactions (10)–(14), only the charges for reaction (10) fail to combust, both in bulk and pressed form. This is due to the presence of the highest amount (13 mol) of poorly reactive silicon powder in these mixtures. In reactions (11)–(14), the silicon content decreases to 4–7 mol, and the corresponding charges undergo combustion. The combustion of bulk mixtures is more intense, characterized by a somewhat higher pressure rise and greater mass loss. In some Si₃N₄: SiC molar ratios, the phase composition of the target products from bulk charges more closely approximates the theoretical composition than that of the pressed charges, although up to 3 % of free silicon impurity is still present.

Finally, the Table shows that when using carbiding mixture (C), which contains the highest amount (0.15C₂F₄) of the activating PTFE additive per 1 mol of carbiding carbon for reactions (15)–(19), all the corresponding charges undergo intense combustion in both bulk and pressed form, reaching similar maximum reactor pressures ranging from 3.50 to 4.36 MPa. The mass loss strongly depends on the Si₃N₄: SiC molar ratio. At a 1:4 ratio, the mass loss reaches very high values (up to 80 %) for both bulk and pressed charges. For other Si₃N₄: SiC ratios, the mass loss remains high for bulk charges (from 57.0 to 80.1 %) but is relatively low for pressed charges (from 12.6 to 35.3 %). The composition of the washed combustion products obtained from the pressed charges is significantly closer to the theoretical values compared to that of the bulk charges, both in terms of Si₃N₄ and SiC phase content at all ratios, and in terms of lower levels of free silicon impurity: up to 2.0 % for pressed charges versus up to 4.0 % for bulk charges. The mass loss during combustion and the SiC phase content in the washed products obtained using carbiding mixture (C) are presented graphically in Figs. 1 and 2.

As shown in Fig. 1, the mass loss during combustion of the pressed mixtures is significantly lower than that of the bulk mixtures and becomes nearly identical – and very high – only at the Si₃N₄: SiC molar ratio of 1:4, reaching 80.4 and 81.4 %, respectively. However, an experiment involving the combustion of this pressed mixture at a higher initial nitrogen pressure in the reactor (increased from 3 to 4 MPa) demonstrated a substantial reduction in mass loss – by a factor of two – to 41.9 %, while maintaining a similar phase composition in the washed combustion product: α-Si₃N₄ = 42.0 wt. %, β-Si₃N₄ = 51.0 wt. %, SiC = 6.0 wt. %, Si = 1.0 wt. %.

As evident from the Table and Fig. 2, the phase composition of the washed combustion products of the pressed charges containing PTFE in carbiding mixture (C) shows the best agreement with the theoretical phase composition in terms of Si₃N₄ and SiC content across all investigated Si₃N₄: SiC molar ratios. This experimentally obtained phase composition is much closer to the theoretical composition of Si₃N₄–SiC composites than the results previously reported for conventional azide SHS (without PTFE), which yielded a significantly lower SiC content in the combustion products compared to the theoretical values [23–25].

The X-ray diffraction (XRD) patterns obtained from the washed combustion products of the pressed charges synthesized using carbiding mixture (C) are shown in Fig. 3.

The XRD patterns display strong reflections from the target phases Si₃N₄ and SiC, as well as either weak reflections from residual free Si or none at all. Silicon nitride is present in two polymorphic forms: α-Si₃N₄ and β-Si₃N₄. As shown in Fig. 3 and the table, the contents of these phases are approximately equal at Si₃N₄: SiC molar ratios of 4:1 and 1:4. At other ratios, however, the α-phase content is roughly half that of the β-phase. Overall, the α-phase constitutes a significant fraction – ranging from 30 to 50 % – of the silicon nitride phase in the synthesized Si₃N₄–SiC composites.

The microstructure of the synthesized composites is presented in Fig. 4.

As shown in Fig. 4, the combustion products of all pressed charges are predominantly composed of highly dispersed particles smaller than 1–2 μm. The combustion product of the charge from reaction (15) with a Si₃N₄: SiC molar ratio of 4:1 contains a significant fraction of columnar crystals with a transverse size of approximately 500 nm and a length of up to 2 μm, along with a small amount of finer equiaxed particles and nanofibers with diameters below 100 nm. The combustion product of the charge from reaction (16) with a 2:1 ratio includes a large fraction of coarser columnar crystals, approximately 1 μm in diameter and up to 5 μm in length, as well as equiaxed particles up to 2 μm and a small fraction of nanofibers. The product from reaction (17) with a 1:1 ratio stands out for its large proportion of nanofibers in a mixture with equiaxed particles up to 1 μm in size. The combustion product of reaction (18) with a 1:2 ratio primarily consists of agglomerates of relatively fine equiaxed particles ranging from 100 nm to 0.5 μm. Finally, the combustion product of reaction (19) with a 1:4 ratio mainly contains agglomerates of larger equiaxed particles, ranging from 200 nm to 1 μm.

Conclusion

The conventional approach of azide-assisted SHS using halide salts such as NH₄F, Na₂SiF₆, and (NH₄)₂SiF₆ as activating gasifying additives previously enabled the synthesis of highly dispersed (˂1 μm) Si₃N₄–SiC powder compositions through combustion in a nitrogen atmosphere from mixtures of silicon powder, technical carbon, sodium azide, and a halide salt. These compositions exhibited a high content of the α-phase of silicon nitride. However, their phase compositions significantly deviated from the theoretical values, showing a much higher content of silicon nitride and a markedly lower content of silicon carbide, along with notable impurities of free silicon (up to 5.7 wt. %) and free carbon (up to 5.1 wt. %) [23–25].

In the present study, polytetrafluoroethylene (PTFE) was used instead of halide salts as both an activating and carbiding additive – partially replacing technical carbon – while sodium azide served as the nitriding additive in an amount sufficient to neutralize the fluorine released during complete PTFE decomposition. This approach facilitated the combustion of silicon–carbon powder mixtures in a nitrogen atmosphere and significantly increased the silicon carbide content in the synthesized Si₃N₄–SiC product, while reducing the amounts of free silicon and carbon impurities. The most favorable results were obtained using carbiding mixture (C), which contained the highest PTFE amount (0.15C₂F₄ + 0.7C). This can be attributed to the higher content of poorly reactive silicon powder (4 to 13 mol) in the charges for synthesizing Si₃N₄–SiC compositions compared to, for example, a similar PTFE-based azide SHS process for AlN–SiC compositions. In the latter case, the best results were achieved using carbiding mixture (A) with the lowest PTFE content (0.05C₂F₄ + 0.9C) in charges containing only 1 mol of poorly reactive Si powder and 1 to 4 mol of highly reactive Al powder.

Based on the obtained data on mass loss during combustion and on the phase composition of the combustion products, it can be recommended that Si₃N₄–SiC powder compositions be synthesized via combustion of pressed charges in azide SHS with partial replacement of 0.3 mol of carbon by 0.15 mol of PTFE, at an initial nitrogen pressure of 3 MPa (or 4 MPa to reduce mass loss when the Si₃N₄: SiC phase ratio is 1:4). The resulting combustion products are predominantly mixtures of highly dispersed particles smaller than 1–2 μm and contain a substantial fraction (30–50 %) of the α-phase in the silicon nitride component of the synthesized Si₃N₄–SiC compositions. The phase composition of these products is significantly closer to the calculated theoretical composition of the target Si₃N₄–SiC system than that achieved using conventional azide SHS without PTFE.