Fe-Si Intermetallics/Al₂O₃ Composites Formed between Fe-20% Si and Fe-70.5% Si by SHS Metallurgy Method

Abstract

Fe–Si intermetallics–Al₂O₃ composites were fabricated by thermite-assisted combustion synthesis. Combustion reactions were conducted with powder compacts composed of Fe₂O₃, Al, Fe, and Si. The starting stoichiometry of powder mixtures had an atomic Fe/Si proportion ranging from Fe-20% to Fe-70.5% Si to explore the variation of silicide phases formed with Si percentage. Combustion in the mode of self-propagating high-temperature synthesis (SHS) was achieved and the activation energy of the SHS reaction was deduced. It was found that the increase of Si content decreased the combustion temperature and combustion wave velocity. Three silicide compounds, Fe₃Si, FeSi, and α-FeSi₂, along with Al₂O₃ were identified by XRD in the final products. Fe₃Si was formed as the single-phase silicide from the reactions with Si percentage from Fe-20% to Fe-30% Si. FeSi dominated the silicide compounds in the reactions with atomic Si content between Fe-45% and Fe-55% Si. As the Si percentage increased to Fe-66.7% Si and Fe-70.5% Si, α-FeSi₂ became the major phase. The microstructure of the composite product showed that dispersed granular or nearly spherical iron silicides were embedded in Al₂O₃, which was dense and continuous. Most of the silicide grains were around 3–5 μm and the atomic ratio of silicide particles from the EDS analysis confirmed the presence of Fe₃Si, FeSi, and FeSi₂.

1. Introduction

Transition metal silicides have been considered as the potential structural materials for high-temperature applications due to their excellent high-temperature mechanical properties, good oxidation resistance, creep resistance, and thermal stability [1,2,3]. Particular attention has been paid to Fe–Si intermetallic compounds, as they possess magnetic, electronic, optical, mechanical, and catalytic properties on account of different crystal structures and phase compositions [4,5,6,7]. Fe–Si silicides with various compositions have been synthesized and characterized, such as thermodynamically stable Fe₃Si, FeSi, and β-FeSi₂ and metastable Fe₂Si, Fe₅Si₃, and α-FeSi₂ [8,9,10,11,12]. According to the Fe–Si phase diagram [13], Fe₃Si exists in a wide stoichiometric range from 10 to 29.8 at.% Si but the homogeneity range of FeSi is very narrow. Both phases coexist between 25 and 50 at.% Si. The disilicide FeSi₂ of β phase is a line compound at 66.7 at.% Si and is stable at room temperature, while α-phase FeSi₂ forms within 70–73.5 at.% Si at temperatures above 937 °C [13].

Among the various fabrication methods for preparing transition metal silicides in a monolithic or composite form, self-propagating high-temperature synthesis (SHS) has the characteristics of simplicity, low energy requirement, short processing time, and high-purity products [14,15,16]. Furthermore, the high temperature gradient combined with rapid cooling makes the SHS reaction possible to produce metastable phases and unique structures. Unlike silicide compounds associated with molybdenum [17] and the transition metals of groups IVb and Vb [18,19], iron silicides cannot be produced by direct combustion of Fe and Si powders in the SHS mode due to weak reaction exothermicity. Consequently, mechanical and chemical activation on combustion synthesis has been considered as an alternative to resolve this issue. For example, Gras et al. [20] employed mechanically activated Fe and Si powders, which were treated by high-energy ball milling, to produce FeSi and β-FeSi₂ through self-sustaining combustion synthesis. KNO₃ was adopted as a reaction promoter to chemically stimulate the combustion between Fe and Si powders on the synthesis of FeSi and α-FeSi₂ [21]. On the other hand, mechanical alloying was utilized to fabricate FeSi– and Fe₃Si–Al₂O₃ composites from the SiO₂/Al/Fe and Fe₃O₄/Al/Si powder mixtures, respectively [22,23]. The addition of a hard secondary phase, such as Al₂O₃, was shown to improve mechanical properties of iron silicides [22]. Moreover, Fe₃Si–Al₂O₃ composites exhibited good ferromagnetic behavior with hysteresis loops [23]. With the use of the SHS metallurgy method, FeSi–Al₂O₃ composites with a molar ratio of FeSi/Al₂O₃ ranging from 1.2 to 4.5 were recently produced from combining the reduction of Fe₂O₃ and SiO₂ by Al with elemental Fe–Si reactions [24].

In addition to the SHS metallurgy, the sol-gel method and hot roll-bonding technique have been considered as the promising process for the fabrication of particles and composites. For example, Najafi et al. [25] synthesized zirconium carbide nanopowders by the sol-gel method in the alcoholic system, within which zirconium propoxide and phenolic resin were used as precursors of Zr and carbon, respectively. Functionally graded aluminum matrix Al6061/SiC_p composite laminates were fabricated by successive hot roll-bonding, which utilized two composite layers as outer strips and a layer of Al1050 as interlayer [26].

This study aims to investigate the different phases of Fe–Si intermetallics formed within a wide stoichiometric range from Fe-20 to Fe-70.5 at.% Si by thermite-assisted combustion synthesis in the SHS mode. By taking advantage of the highly exothermic thermite reaction of Fe₂O₃ with Al, the formation of in situ Fe–Si intermetallics/Al₂O₃ composites was conducted with reactant powder compacts consisting of Fe₂O₃, Al, Fe, and Si. Effects of the Si percentage on the reaction exothermicity, combustion temperature, flame-front propagation velocity, and phase composition and microstructure of the iron silicides/Al₂O₃ composites were explored. Moreover, the activation energy of the SHS reaction was determined from combustion wave kinetics.

2. Materials and Methods

The starting materials adopted by this study comprised Fe (Alfa Aesar Co., Ward Hill, MA, USA, <45 μm, 99.5%), Si (Strem Chemicals, Newburyport, MA, USA, <45 μm, 99.5%), Al (Showa Chemical, Tokyo, Japan, <45 μm, 99.9%), and Fe₂O₃ (Alfa Aesar Co., <45 μm, 99.5%). Based on the atomic proportions between Fe and Si in the reactant mixtures, two combustion systems expressed as Reactions (1) and (2) were considered.

$${\mathrm{Fe}}_2{\mathrm{O}}_3+2\mathrm{Al}+2.5\mathrm{Fe}+a\mathrm{Si}\to m{\mathrm{Fe}}_3\mathrm{Si}+n\mathrm{FeSi}+{\mathrm{Al}}_2{\mathrm{O}}_3$$

(1)

$${\mathrm{Fe}}_2{\mathrm{O}}_3+2\mathrm{Al}+2.5\mathrm{Fe}+b\mathrm{Si}\to p\mathrm{FeSi}+q{\mathrm{FeSi}}_2+{\mathrm{Al}}_2{\mathrm{O}}_3$$

(2)

Reactant mixtures were composed of thermite reagents of Fe₂O₃ + 2Al and metallic Fe and Si powders. Reaction (1) was formulated with the Si percentage from Fe-20% to Fe-45% Si, within which Fe₃Si and FeSi was considered as silicide compounds to be produced. Reaction (2) was designed for Fe/Si stoichiometry from Fe-50% to Fe-70.5% Si and the resulting silicides of Reaction (2) could be FeSi and FeSi₂.

With respect to different Si percentages denoted by x of Fe-x% Si,

Table 1. Stoichiometric coefficients (a, m, and n) and adiabatic temperatures (T_ad) of Reactions (1) under different Si percentages (x) in terms of Fe-x%Si.

x, Fe-x% Si	a	m	n	Tad (K)
20	1.13	1.13	0	2870
25	1.5	1.5	0	3206
30	1.93	1.29	0.64	3158
35	2.42	1.04	1.38	3110
40	3.0	0.75	2.25	3059
45	3.68	0.41	3.27	3008

and

Table 2. Stoichiometric coefficients (b, p, and q) and adiabatic temperatures (T_ad) of Reactions (2) under different Si percentages (x) in terms of Fe-x%Si.

x, Fe-x% Si	b	p	q	Tad (K)
50	4.5	4.5	0	2960
55	5.5	3.5	1.0	2820
60	6.75	2.25	2.25	2670
65	8.36	0.64	3.86	2504
66.7	9.0	0	4.5	2444
70.5	10.75	0	4.5	2292

list the values of stoichiometric coefficients, including a, m, and n, of Reaction (1) and b, p, and q of Reaction (2). These values were calculated based on the conservation of atomic species, except for two off-stoichiometric conditions. One is Reaction (1) of Fe-20% Si in Table 1, which represents an under-stoichiometric condition (i.e., a Si-lean composition) to produce Fe₃Si because Fe₃Si exists in a wide homogeneity range from 10 to 29.8 at.% Si. The other is Reaction (2) of Fe-70.5% Si in Table 2 for the production of FeSi₂. This case was intended for FeSi₂ of the α phase since α-FeSi₂ is an over-stoichiometric (i.e., Si-rich) disilicide and is present within a region of 70–73.5 at.% Si.

The reaction exothermicity of Reactions (1) and (2) was studied by calculating their adiabatic combustion temperature (T_ad) as a function of x in Fe-x% Si, based on the following energy balance equation [27] with thermochemical data taken from [28,29].

$$∆H_r+{{\displaystyle \int }}_{298}^{T_{ad}}{\displaystyle \sum }{n}_j{c}_p\left(P_j\right)dT+{\displaystyle \sum }_{298-T_{ad}}{n}_{j}L\left(P_j\right)=0$$

(3)

where ΔH_r is the reaction enthalpy at 298 K, n_j is the stoichiometric coefficient, c_p and L are the heat capacity and latent heat, and P_j refers to the product. ΔH_r was calculated from the heats of formation of the reactants and products. Heats of formation for Fe₂O₃, Al₂O₃, Fe₃Si, FeSi, and FeSi₂ are −823.4, −1675.7, −73.1, −79.4, and −81.2 kJ/mol, respectively. Specific heats, c_p (in J/mol⋅K), of the product species as a function of temperature are given below [28,29].

$$c_p\left(A{l}_2{O}_3\right)=117.49+10.38·{10}^{-3}·T-3.71·{10}^6·T^{-2}$$

(4)

$$c_p\left(F{e}_{3}Si\right)=80.2+25.7·{10}^{-3}·T-0.2·{10}^6·T^{-2}$$

(5)

$$c_p\left(FeSi\right)=44.6+14.72·{10}^{-3}·T-0.11·{10}^6·T^{-2}$$

(6)

$$c_p\left(FeS{i}_2\right)=64.2+20.6·{10}^{-3}·T-0.15·{10}^6·T^{-2}$$

(7)

In this study, the reactant powders were dry mixed in a tumbler ball mill (Quancin Tech. Co., Tainan, Taiwan), which consisted of a cylinder partially filled with raw materials and alumina grinding balls rotating about its longitudinal axis. The diameter of the grinding ball was 2.5 mm. The tumbler mill machine operated at 60 rpm and the milling time was 4 h. The purpose of this ball milling process was to disperse the agglomerated particles and to achieve the good mixing of the reactant powders. The action of ball milling in this study did not affect the size and shape of the particles. Well-mixed powders were uniaxially compressed to form cylindrical test specimens with 7 mm in diameter, 8.5 mm in height, and a relative density of 55%.

The SHS experiment was conducted in a windowed combustion chamber filled with high-purity (99.99%) argon at 0.2 MPa. The ignition was accomplished by a heated tungsten coil with a voltage of 60 V and a current of 1.5 A. The propagation rate of combustion wave (V_f) was measured by recording the whole combustion event with a color CCD camera at 30 frames per second. The exposure time of each recorded image was set at 0.1 ms. To facilitate the accurate measurement of instantaneous locations of the combustion front, a beam splitter (Rolyn Optics; Covina, CA, USA), with a mirror characteristic of 75% transmission and 25% reflection, was used to optically superimpose a scale onto the image of the test sample. The combustion temperature was measured by an R-type (Pt/Pt-13%Rh) fine-wire thermocouple (Omega Engineering Inc.; Norwalk, CT, USA) with a bead size of 125 μm. The details of the experimental setup were reported elsewhere [30].

The phase constituents of the synthesized products were analyzed by an X-ray diffractometer (Bruker D2 Phaser; Billerica, MA, USA) with CuK_α radiation. Scanning electron microscopy (SEM) analysis was performed (Hitachi S3000H; Tokyo, Japan) with energy dispersive spectroscopy (EDS) to examine the microstructure and elemental composition on the fracture surface of the final product. The open porosity of the product was determined by the Archimedes method. Vickers hardness (H_v) measurements were conducted with 1.0 kg load for selected samples.

The activation energy (E_a) of the SHS reaction was deduced from the temperature dependence of combustion wave velocity. The combustion wave velocity derived from an energy equation with a heat source can be expressed as [31,32]

$$V_f{}^2=\frac{2\lambda }{\rho Q}\frac{R{T}_c{}^2}{E_a}{k}_o\exp (-E_a/R{T}_c)$$

(8)

where λ is the thermal conductivity, ρ the density, R the gas constant, Q the heat of reaction, T_c the combustion front temperature, and k_o the Arrhenius rate constant. According to Equation (8), E_a can be obtained from the slope of a best-fitted linear line correlating ln(V_f/T_c)² with 1/T_c.

3. Results and Discussion

3.1. Combustion Exothermicity and Combustion Wave Kinetics

The calculated values of T_ad are presented in Table 1 and indicate that T_ad of Reaction (1) decreases with increasing Si percentage except for the off-stoichiometric case of Fe-20% Si. For Reaction (1), the decrease of T_ad was mainly attributed to the increase of the numbers of mole of iron silicides (m + n) formed in the products. Since the formation of iron silicides is less exothermic than aluminothermic reduction of Fe₂O₃, the overall exothermicity of combustion synthesis tends to decrease as the amount of silicide compounds increases. The exception was due to a smaller amount of Fe₃Si (m = 1.13) produced from the Si-lean sample of Fe-20% Si when compared to that (m = 1.5) associated with the stoichiometric condition of Fe-25% Si.

On the other hand, the numbers of mole of iron silicides (p + q = 4.5) synthesized from Reaction (2) are constant. As shown in Table 2, the increase of Si content leads to an increase in FeSi₂ but a decrease in FeSi. Although both FeSi and FeSi₂ have comparable enthalpies of formation, the heat capacity of FeSi₂ is much larger than that of FeSi. This suggested lower formation exothermicity for FeSi₂ and explained the decrease of T_ad with increasing Si percentage for Reaction (2).

Figure 1a,b illustrates two combustion sequences recorded from Reaction (1) of Fe-25% Si and Fe-40% Si, and Figure 1c,d is associated with Reaction (2) of Fe-50% Si and Fe-66.7% Si. It is evident that for all reactions a distinct combustion front formed upon ignition and propagated along the powder compact in a self-sustaining manner. A comparison unfolded that the sample of Figure 1a exhibits stronger combustion luminosity, the flame spreading time is shorter, and the final product is partially melted. The burned sample of Figure 1b essentially retained its original shape.

In Figure 1c,d, however, the combustion glow gradually fades away as the flame front progresses, the propagation time becomes longer, and the products are stretched out. These observations reflect vigorous combustion for Reaction (1) of Fe-25% Si. The partial melting of the burned sample of Reaction (1) with Fe-25% Si could be mainly attributable to the combustion temperature exceeding the melting point of Fe₃Si of about 1210 °C [13]. The elongated products obtained from Reaction (2) of Fe-50% Si and Fe-66.7% Si were due partly to the combustion temperature below the melting points of FeSi and FeSi₂ at 1410 and 1220 °C, respectively [13] and in part were interpreted as the combined result of the green compact porosity, small pores evolved from Kirkendall effect, and density change between the reactants and products [15].

The variations of flame-front propagation velocity (V_f) and combustion front temperature (T_c) with Si content of Fe-x% Si for Reactions (1) and (2) are shown in Figure 2. As the Si proportion increased from Fe-25% to Fe-70.5% Si, it was noticed that V_f decreased from 3.1 to 0.8 mm/s and T_c declined from 1950 to 1367 K. The combustion wave propagation rate of the SHS process is mostly governed by the layer-by-layer heat transfer from the reaction zone to unburned region and therefore is subject to the combustion front temperature. This clarifies the consistency between combustion wave velocity and reaction front temperature. Moreover, the trend variation of T_c agrees reasonably with that of T_ad. In particular, Figure 2 indicates that T_c of Reaction (1) with Fe-20% Si of about 1890 K is lower than that of 1950 K reached by Reaction (1) with Fe-25% Si, confirming the lower reaction exothermicity of this under-stoichiometric sample.

According to combustion wave kinetics, the activation energies (E_a) of Reactions (1) and (2) were determined on the basis of the correlation between ln(V_f/T_c)² and 1/T_c. Two sets of experimental data with their best-fitted linear lines are plotted in Figure 3. From the slopes of straight lines, the values of E_a equal to 171.3 and 106.4 kJ/mol were deduced for Reactions (1) and (2), respectively. The activation energy represents a kinetic barrier that depends on the reaction mechanism and transport phenomena. The possible reaction pathways of Reactions (1) and (2) were initiated by the aluminothermic reduction of Fe₂O₃. In addition to the kinetic aspect, the reduction of Fe₂O₃ by Al was thermodynamically favorable and acted as the major heat-releasing step to sustain the SHS process. Different silicide phases formed from SHS reactions could be responsible for a larger E_a for Reaction (1) than Reaction (2).

3.2. Phase Composition and Microstructure of Synthesized Products

Figure 4 presents XRD patterns of the products of Reaction (1) with Fe-20%, Fe-30%, and Fe-40% Si. As revealed in Figure 4, Fe₃Si is the only silicide formed from Reaction (1) with Fe-20% and Fe-30% Si. This agrees with the homogeneity range of Fe₃Si in the phase diagram. In the product of Reaction (1) with Fe-40% Si, silicide compounds comprising the three different phases of Fe₃Si, FeSi, and FeSi₂ were identified with Fe₃Si and FeSi being two primary phases. This reflects the fact that Fe₃Si and FeSi are co-existent under stoichiometry from Fe-30% to Fe-50% Si. In addition to silicide compounds, Figure 4 points out the formation of Al₂O₃ from the aluminothermic reduction of Fe₂O₃ and the presence of a small quantity of aluminum silicate that was formed through the dissolution of Si into Al₂O₃ during the SHS process [33].

Three XRD spectra depicted in Figure 5 show the phase compositions of the products of Reaction (2) with Fe-50%, Fe-60%, and Fe-66.7% Si. In addition to Al₂O₃, two silicide compounds, FeSi and FeSi₂, were detected in Figure 5. FeSi was the major silicide and FeSi₂ the minor under stoichiometry of Fe-50% Si. It was evident that FeSi₂ substantially increased in the case of Fe-60% Si and became predominant over FeSi for a further increase in Si to Fe-66.7% Si. It should be noted that the FeSi₂ produced in this study is α-phase disilicide. The formation of metastable α-FeSi₂ was ascribed to the SHS process characterized by a unique temperature gradient starting with a sudden temperature rise over 1360 K and followed immediately by a rapid cooling process. Because of the presence of FeSi, a trivial amount of remnant Si was found in the final product of the reaction with Fe-66.7% Si.

The yield percentage of different silicide phases (Fe₃Si, FeSi, and FeSi₂) in the product was estimated by a semi-quantitative analysis based on the integrated intensity values of their highest XRD peaks, including Fe₃Si (220) at 2θ = 45.337°, FeSi (210) at 2θ = 45.062°, and FeSi₂ (102) at 2θ = 49.098°. The yield percentage of silicide phases as a function of Si percentage is plotted in Figure 6. For Reaction (1), Fe₃Si was the only silicide produced from the samples with Si content from Fe-20% to Fe-30% Si. Iron silicides generated from the sample of Fe-35% Si contained 83% Fe₃Si and 17% FeSi. Three silicide compounds were present in the products of samples with Fe-40% and Fe-45% Si and the proportions of Fe₃Si/FeSi/FeSi₂ were near to 60.0/45.3/4.7 and 26.7/67.3/6.0, respectively. It is evident for Reaction (1) that the dominant silicide shifted from Fe₃Si to FeSi as the Si percentage increased from Fe-20% to Fe-45% Si.

As also indicated in Figure 6, the resulting silicides of Reaction (2) with Fe-50% and Fe-55% Si were dominated by FeSi with a yield percentage up to 92.9% and 84.5%, respectively, and FeSi₂ was the minor phase. With the increase of Si, comparable amounts of FeSi and FeSi₂ were produced from the reactions of Fe-60% and Fe-65% Si. Finally, FeSi₂ became the governing phase with high yield percentages of 79.1% and 87.4% in the products of samples with Fe-66.7% and Fe-70.5% Si, respectively. For Reaction (2) with Si from Fe-50% and Fe-70.5% Si, the change in dominant silicide from FeSi to FeSi₂ was observed.

Figure 7a–c illustrates the SEM images and EDS spectra of the fracture surfaces of products synthesized from the samples with Si percentages of Fe-25% Si, Fe-50% Si, and Fe-70.5% Si, respectively. As can be seen in the SEM image of Figure 7a, the morphology exhibits granular particles distributing over a dense and continuous substrate and the particle size ranges approximately from 2 to 10 μm. According to the EDS analysis, those particles were identified as iron silicide and the dense substrate was made up of aluminum oxide. The atomic ratio of Fe:Si based on the EDS spectrum of Figure 7a is 74.3:25.7, which is very close to 3:1 and proves the formation of Fe₃Si from the sample of Fe-25% Si. The dense matrix of Figure 7a is composed of Al and O elements with an atomic ratio of Al:O = 42.5:57.5, which matches reasonably with the stoichiometry of Al₂O₃.

Nearly spherical particles with a broad size range from submicron to about 10 μm were observed in the SEM image of Figure 7b. The ratio of Fe:Si obtained from the EDS spectrum was about 47.7:52.3. This suggested the phase of FeSi and agreed reasonably with the starting Fe/Si stoichiometry of Fe-50% Si. The EDS spectrum associated with the dense substrate of Figure 7b indicates that constituent elements are Al and O. The composition of the matrix was confirmed to be Al₂O₃ by the atomic ratio of Al:O = 37.9:62.1 deduced from the EDS analysis.

In Figure 7c, the SEM photo shows that many granular particles partially embedded in the dense substrate can be seen, most of them have a size of around 3 μm, and some pores are formed in the Al₂O₃ substrate. The EDS spectrum provided the atomic ratio of Fe:Si equal to 31.3:68.7, which is slightly Si-rich in comparison to the atomic composition of FeSi₂. This could be due to the fact that the silicide formed from the sample of Fe-70.5% Si is α-FeSi₂, which is a Si-rich disilicide. Similarly, the EDS spectrum of Figure 7c shows that Al and O are two major constituent elements for the substrate and the atomic proportion of Al:O = 42.1:57.9 is in good agreement with that of Al₂O₃.

SEM images presented in Figure 8a–c are associated with the final composites produced from samples of Fe-20%, Fe-55%, and Fe-66.7% Si, respectively. It is evident that granular particles revealed in Figure 8a,b are more spherical than those in Figure 8c. The silicide grains formed were Fe₃Si in Figure 8a and mostly FeSi in Figure 8b. Their particle size ranged from submicron to about 10 μm. The matrix made up of Al₂O₃ in both Figure 8a,b is dense and continuous. Figure 8c shows that FeSi₂ granules with a size of 2–5 μm are formed over the Al₂O₃ substrate. Some pores were observed in the matrix due to volume expansion of the product. A comparison between Figure 7 and Figure 8 pointed out that similar microscopic structures of the final composites were observed as their initial compositions had close Fe/Si proportions.

For the reactions with Si percentage from Fe-20% to Fe-30% Si, their resulting Fe₃Si–Al₂O₃ composites were produced in a partially melted form so that the porosity of Fe₃Si–Al₂O₃ composites was relatively low and varied between 17% and 22%. Due to the slight volume expansion of the final products, the porosity of FeSi–Al₂O₃ composites synthesized from samples of Fe-50% and Fe-55% Si was around 30%. Higher porosity of about 37–42% was attained for the FeSi₂–Al₂O₃ composites produced from the reaction of Fe-66.7% and Fe-70.5% Si, because the products were prolonged after combustion. For selected composites obtained from the reactions with Fe-20%, Fe-50%, and Fe-66.7% Si, Vickers hardness measurements were conducted. The values of H_v were found to be 12.5, 10.4, and 7.6 GPa for the Fe₃Si–, FeSi–, FeSi₂–Al₂O₃ composites, respectively.

4. Conclusions

The formation of Fe–Si intermetallics–Al₂O₃ composites was investigated using thermite-assisted combustion synthesis with an emphasis on the relationship between iron silicide phases formed and initial Fe/Si proportions. Combustion reactions in the SHS mode were achieved with the powder compacts of the Fe/Si stoichiometry covering Si percentage from Fe-20% to Fe-70.5% Si. Results showed that the increase of Si content lowered the reaction exothermicity and decreased the combustion temperature from 1950 to 1367 K and flame-front propagation velocity from 3.1 to 0.8 mm/s. This was mainly due to the fact that the formation of iron silicide compounds was less exothermic than the aluminothermic reduction of Fe₂O₃. Depending on different iron silicides formed from the SHS reactions, two activation energies of 171.3 and 106.4 kJ/mol were determined. By means of XRD, three iron silicides, Fe₃Si, FeSi, and α-FeSi₂, were identified along with Al₂O₃ in the final products. Fe₃Si was the only or major silicide for the reactions containing Si between Fe-20% and Fe-40% Si. Specifically, Fe₃Si was the only silicide in the products of Fe-20%, 25%, and 30% Si. The ratio of Fe₃Si/FeSi in the product generated from the sample of Fe-35% Si was 83/17. Three silicide compounds with a proportion of Fe₃Si/FeSi/FeSi₂ was equal to 60.0/45.3/4.7, when produced from the sample of Fe-40% Si. The reactions of Fe-45%, Fe-50% and Fe-55% Si produced FeSi as the dominated silicide and the corresponding silicide compositions were Fe₃Si/FeSi/FeSi₂ = 26.7/67.3/6.0, FeSi/FeSi₂ = 92.9/7.1, and FeSi/FeSi₂ = 84.5/15.5. For the reactions of Fe-65%, Fe-66.7%, and Fe-70.5% Si, their resulting silicides comprised FeSi and α-FeSi₂ and were predominated by α-FeSi₂ which exhibited an increasing yield percentage from 56.7% to 79.1% and to 87.4%. The SEM images of the product fracture surface revealed that granular or nearly spherical silicides were distributed over or embedded partially in the dense substrate made up of Al₂O₃. The particle size of iron silicide grains varied from submicron to about 10 μm and most of them were around 3–5 μm. The phases of silicide grains, including Fe₃Si, FeSi, and FeSi₂, were also confirmed by the EDS analysis.