Int. Journal of Refractory Metals and Hard Materials 45 (2014) 23–30 Contents lists available at ScienceDirect Int. Journal of Refractory Metals and Hard Materials journal homepage: www.elsevier.com/locate/IJRMHM Synthesis of dense mullite/MoSi2 composite for high temperature applications Z.I. Zaki a,b,⁎, Nasser Y. Mostafa a,c, Y.M.Z. Ahmed b a b c Chemistry Department, Faculty of Science, Taif University, P.O. Box: 888, Al-Haweiah, Taif-Saudi Arabia Advanced Materials Division, Central Metallurgical R&D Institute (CMRDI), P.O. Box: 87, Helwan, Cairo, Egypt Chemistry Department, Faculty of Science, Suez Canal University, Ismailia 41522, Egypt a r t i c l e i n f o Article history: Received 12 December 2013 Accepted 6 March 2014 Available online 18 March 2014 Keywords: Synthesis Mullite MoSi2 Oxidation Electrical characters a b s t r a c t This work is aimed at producing a dense mullite/MoSi2 composite from a reactive blend of MoO3, SiO2 and Al by under load-combustion synthesis. Thermodynamic analyses proved the development of the entire combustion product in molten state with an adiabatic temperature of 2340 K. Diluting the reaction mixture with 25 vol.% of the combustion product was necessary to produce homogeneity between the obtained phases. The sample porosity was reduced to ~3.0 vol.% by controlling the pressing load, Al particle size and the delay time. The coefficient of thermal expansion of the obtained product was 5.96–8.93 × 10−6 K−1. The obtained dense composite exhibited good oxidation resistance at 1300 °C and a low electrical resistivity (1.80 Ω·cm). © 2014 Elsevier Ltd. All rights reserved. Introduction Due to its unique properties (thermal, electrical and chemical) in addition to its moderate density, MoSi2 has received the attention of many research groups. Despite its weak mechanical properties and pest oxidation problems, MoSi2 composites are becoming important elevated temperature structural materials for application in oxidizing and aggressive environments specially as heating elements in high temperature furnaces [1]. On the other hand, mullite (3Al2O3·2SiO2), a typical solid solution oxide, combines high melting point, low thermal expansion, low electrical conductivity, good mechanical strength and resilience at elevated temperatures. Therefore, it is considered as a potential candidate material for structural applications such as kiln furniture, protection tubes, and heat insulation parts [2]. MoSi2 is thermodynamically stable with mullite and a variety of other ceramic reinforcements, including TiC, ZrO2, Al2O3, TiB2, SiC and Si3N4 [3–11]. MoSi2 has also been employed as a conducting phase in Si3N4–MoSi2 particulate composites [10]. Alumina composites with 20 vol.% and higher MoSi2 are electroconductive due to the formation of a three-dimensional percolating network of the conductive MoSi2 phase [7,12]. Other composites of MoSi2 with Mo5Si3 [13], WSi2 [14], and ZrB2 [15,16] were also reported. A composite between MoSi2 and mullite is rarely mentioned in the literature. In a trial to improve the mechanical properties of MoSi2 by the powder metallurgy route, some ⁎ Corresponding author. http://dx.doi.org/10.1016/j.ijrmhm.2014.03.006 0263-4368/© 2014 Elsevier Ltd. All rights reserved. authors used mullite and mullite whisker as reinforcement [17]. Mullite containing 20 vol.% MoSi2 was synthesized by hot pressing a mixture of mullite and MoSi2 powders at 1650 °C with 69 MPa for 2 h. The product was found to have a Klc and strength values up to two times that of monolithic mullite [18]. Various starting materials and preparation methods have been used to prepare synthetic mullite and MoSi2 ceramics. These include reaction sintering of mechanically mixed powders [19–24], hydrothermal treatment of mixtures of sols [25–27], and chemical vapor deposition [28–31] for synthesis of mullite. On the other hand, mechanical alloying [4,32], hot pressing [16], reactive sintering [33], plasma-spray processing [34] and self-propagating high-temperature synthesis (SHS) are used for synthesis of MoSi2 [13,35,36]. Self-propagating high-temperature synthesis (SHS) is one of the rapidly emerging cost-effective technologies used to synthesize monolithic and composite in situ ceramics [35,37]. The principle of this technique is that the initial reagents, when ignited, spontaneously transform into products due to the exothermic heat of the reaction. The major problem to overcome in the SHS processing of refractory ceramics is the high-retained porosity of the synthesized products. However, applying sudden mechanical pressure during the reaction has resulted in highly dense materials [38,39]. The densification behavior in that case is only due to the compaction and is favored by the formation of liquid phases during the reaction [39]. The very short processing time of typical SHS reactions limits the diffusion phenomena responsible for the material shrinkage in other processes [40]. Therefore SHS under the load process represents a one-step manufacturing route capable of 24 Z.I. Zaki et al. / Int. Journal of Refractory Metals and Hard Materials 45 (2014) 23–30 producing highly dense composites with a homogeneous distribution of the different phases. Another serious drawback of SHS is it's limitation to a highly exothermic reaction. Recently, successful management of the heat evolved from the common SHS reactions have enabled carrying out side endothermic reactions [36,41,42]. In a previous work, we have successfully prepared a MoSi2/mullite composite in a self sustaining manner without applying external mechanical pressing. The product unfortunately was highly porous, cracked and was non-homogeneous [42]. The aim of the current work is to in-situ develop a high density and homogeneous composite of MoSi2 and mullite starting from relatively low cost raw materials (MoO3, SiO2 and Al) by applying sudden external mechanical pressing. Designing a composite of MoSi2 and mullite will result in the fabrication of a new low cost ceramic conductor that can be easily shaped using an electric discharge machine and can be used as a heating element. Both MoSi2 and mullite resist oxidation at high temperature and are thermodynamically stable with each other. The mullite phase will be the matrix while the MoSi2 phase will be responsible for conducting the electricity through that matrix. SHS under load methodology is chosen to perform simultaneous synthesis and densification of MoSi2 and mullite where the technical parameters controlling the process are studied. These parameters include: diluent vol.%, pressing load, Al particle size and delaying time. To the best of our knowledge this composite did not synthesize before as a dense object by SHS methods especially from these starting materials. Experimental The materials used in this study were molybdenum trioxide of 99.5% purity and 15 μm average particle size (Atlantic Equipment Engineers, NJ, USA), aluminum metal powder of 99.5% purity and different average particle sizes of 52.3, 37, 18.5 and 5.6 μm (GFS Chemicals, Inc., USA) and silica of 11 μm average particle size with 99% purity (Riedel-de Haen, Germany). Particle size distribution of the raw materials was investigated using Microtrac Bluewave particle size analyzer operated with FLEX software (Microtrac S3500). The wet samples were inserted as a suspension with double distilled water. Ultrasonic waves were used for making the suspension more homogenous. The powders were mixed in the required molar ratio to form MoSi2-7/9 mullite composite and dry blended in a slow rotating mill with alumina ball for 5 h. The powder mixture was uniaxially pressed without a binder at 78 MPa into a cylindrical compact of 15 mm height and 20 mm diameter with an approximately 55% relative density. The compact was transferred into a combustion reactor and placed into a steel die filled with sand where the mechanical pressure was applied on the movable punch. The pressure was transferred to the green compact through a sand medium. Ignition of samples is carried out in thermal explosive mode using a resistance coil placed around the compact. All the work was conducted under vacuum. A detailed procedure of carrying out combustion reactions was provided elsewhere [39]. The electricity was supplied from an AC power source (30–40 A). Linear thermal expansion (LTE) and the coefficient (CTE) of its dense product were measured using dilatometery (Linseis Inc., Germany, Model L76/ 1550) in the temperature range of 200 up to 1250 °C with a heating rate of 10 °C/min. The electrical resistivity investigation was carried out in air under one atmospheric pressure from a room temperature of up to 900 °C with a heating rate of 2 °C/min. The data (temperature vs resistivity) were recorded every 2 min. Oxidation resistance of the combustion product was investigated in open atmosphere at 1300 °C for 3 h in order to determine the thermodynamic stability (compatibility) between the obtained phases of MoSi2 and mullite. Weight changes during oxidation were measured by means of an automatic weight recording balance. Different phases of the combustion products were identified by X-ray diffraction analysis using an X-ray diffractometer (D8 Advanced Bruker AXS, GMbH, Karlsruhe, Germany). The microstructure of specimens was investigated using a scanning electron microscope (SEM, Model JSM-5410, JEOL, Tokyo, Japan) equipped with electron dispersive spectroscopy (EDX). SEM investigation is conducted on polished surfaces coated with gold to ensure good electrical conductivity of the entire components of the sample. Sample porosity was measured via Archimedes' method (submersion in water). Results and discussions Adiabatic temperatures and molten factions The combustion process of this work is based on the following reactions: MoO3 þ 2SiO2 þ 14=3Al ¼ MoSi2 þ 7=3Al2 O3 7=3Al2 O3 þ 14=9SiO2 ¼ 7=9ð3Al2 O3
2SiO2 Þ ΔH ¼ −1470 kJ ΔH ¼ 52:6 kJ MoO3 þ 2SiO2 þ 14=3Al þ 14=9SiO2 ¼ MoSi2 þ 7=9ð3Al2 O3
2SiO2 Þ ð1Þ ð2Þ ΔH ¼ −1417:4 kJ : ð3Þ It could be easily noticed that reaction (1) is highly exothermic while reaction (2) is endothermic in its nature. However, according to the endothermic nature of the reaction (2) it could not proceed alone in a self-sustaining manner. Meanwhile, the addition of SiO2 to the reactant of reaction (1) in a stoichiometric amount (14/9 mol) equivalent to the quantity of liberated alumina (7/3 mol) generates the overall reaction (3). This reaction (3) was found to be highly exothermic in which the heat released by the progress in reaction (1) will support reaction (2). On the other hand reaction (3) seems to be a complicated one which can be composed of many reactions. The possible reactions that can compete to yield reaction (3) and their thermodynamic properties are organized in a consecutive manner according to its expected order and given in Table 1. Regarding reaction (3), the adiabatic combustion temperature and the fractions of molten phases of the products can be calculated from the following equation: −ΔHðr;298Þ þ ¼ Z 2303 298 Z To 298 Cp ðMoO3 þ 2SiO2 þ 14=3Al þ 14=9SiO2 Þ dT Z Cp ðMoSi2 Þs dT þ 7=9 2123 298 ð4Þ Cp ð3Al2 O3
2SiO2 Þs dT þαΔHð f;MoSi2Þ þ 7=9γΔHð f;3Al2O3
2SiO2Þ where, ∆H(r,298) is the enthalpy change of reaction (3), Cp is the specific heat capacity, ∆H(f,MoSi2) and ∆H(f,3Al2O3·2SiO2) are the enthalpies of fusion [42], and α and γ are the molten fractions of MoSi2 and mullite respectively. Fig. 1 shows the adiabatic temperatures and the molten fractions of mullite and MoSi2 phases calculated from Eq. (4) as a function of initial temperatures. It is found that the adiabatic temperature at 298 K is equal to 2340 K. This temperature exceeds the melting points of mullite (2123 K) and MoSi2 (2303 K). This means that both phases will be produced in the molten states. This could be a good point in order to produce a highly dense mullite/MoSi2 structure via compaction during the reaction progress. Table 1 Expected sequence of the reactions involved in the combustion process. Reaction 1 2 3 4 MoO3 + 2Al = Mo + Al2O3 3SiO2 + 4Al = 3Si + 2Al2O3 Mo + 2Si = MoSi2 3Al2O3 + 2SiO2 = (3Al2O3·2SiO2) ΔH, kJ ΔG, kJ −927.5 −614.1 −131.3 67.6 −911 −589 −115 56.0 Z.I. Zaki et al. / Int. Journal of Refractory Metals and Hard Materials 45 (2014) 23–30 25 Fig. 1. Adiabatic combustion temperatures and molten fractions of MoSi2 & mullite as a function of initial temperature and dilutions. Synthesis of mullite/MoSi2 composite The work is started using a sample having aluminum with an average particle size of 37 μm and ignited by a thermal explosion mode. A compaction pressure of 82 MPa is suddenly applied during the reaction. The combustion product was characterized by phase identification using XRD analysis, shown in Fig. 2. The pattern revealed that MoSi2 and mullite appear as the main phases while alumina appears as a minor one. On the other hand no peaks characteristic for the starting materials are detected. This indicates that the reaction between the starting precursors have been going to completion according to reaction (3) using thermal explosion mode under applied pressure. However the microstructure homogeneity of such a composite is regarded as one of the highly important parameters that have a direct impact on its physic–mechanical properties. Accordingly the microstructure of the produced composite has been investigated via SEM analysis. Fig. 3a presents the SEM micrograph of the produced composite. It could be noticed that there is almost a complete separation between the different phases such that one big MoSi2 ball is developed at the center of the specimen and other small ones are found embedded into the mullite matrix. This could be attributed to the fact that the combustion temperature of reaction (3) is very high (Tad = 2340 K) as shown in Fig. 1. This temperature exceeds the melting points of MoSi2 (2303 K) and that of mullite (2123 K) causing a high agglomeration of the melting MoSi2 phase especially during compaction. This inhomogeneous microstructure is responsible for deteriorating all the physical–mechanical properties of the produced composite. Fig. 2. XRD of combustion product prepared by using samples having 37 μm Al particle size pressed during the reaction at 82 MPa. 26 Z.I. Zaki et al. / Int. Journal of Refractory Metals and Hard Materials 45 (2014) 23–30 Fig. 3. SEM images of combustion products containing: a) no diluents b) 20 vol.% dilution c) 25 vol.% dilution and BES images of d) 30 vol.% dilution pressed during the reaction at 82 MPa. According to the previous investigation, the vigor of the reaction (3) should be inhibited to some extent for preventing the agglomeration of the molten MoSi2 during compaction. One solution is the addition of diluents which could compensate some of the heat evolved from reaction (3) and decreasing the probabilities for agglomeration of the molten MoSi2 phase. In this regard and in order to keep the sample Fig. 4. Effect of dilution amount, pressing load and Al average particle size on the combustion product porosity. Z.I. Zaki et al. / Int. Journal of Refractory Metals and Hard Materials 45 (2014) 23–30 composition without changes the combustion product of reaction (3) is crushed, pulverized to pass a 20 μm sieve and used as a diluent. In this case the combustion reactions will be proceeding as follows: MoO3 þ 2SiO2 þ 14=3Al þ 14=9SiO2 þ x ½MoSi2 þ 7=9ð3Al2 O3
2SiO2 ފ ¼ ð1 þ xÞ½MoSi2 þ 7=9ð3Al2 O3
2SiO2 ފΔH ¼ −1417:4 kJ ð5Þ where x, is the weight percentage of the diluents. Regarding reaction (5), the adiabatic combustion temperature and the fractions of molten phases of the products can be calculated from the following equation: −ΔHðr;298Þ þ Z To 298  Cp MoO3 þ 2SiO2 þ 14=3Al þ 14=9SiO2 Z   þ x MoSi2 þ 7=9ð3Al2 O3
2SiO2 ފ dT ¼ ð1 þ xÞ Z þ Tad Z þ Tad 2303 2123 Z CpðMoSi2 Þ1 dTÞ þ 7=9ð1 þ xÞ  2123 298 2303 298 CpðMoSi2 Þs dT 27 and the results are shown in Fig. 4. It is found that the pressing load has a strong effect on the sample porosity such that a sharp decrease in the sample porosity is achieved. The sample porosity is reduced to ~ 3.65 vol.% upon applying a mechanical pressure of 117 MPa and falls to 2.09 vol.% at 234 MPa. While with further increase in the pressing load (at 292 MPa) there is not much of an enhancement in sample porosities noticed. On the other hand SEM investigation of the microstructure of the sample pressed at 234 shows homogeneous distribution of the different phases such that no collection of phases is formed due to pressing (Fig. 5a). This is a confirmation that increasing the pressing load enhances the densification parameter of such composite without any changes in both phase formation as well as the microstructure homogeneity. However, it is well known that the Al particle size usually plays a very important role during the combustion synthesis process. Accordingly, the effect of reducing Al particle size (52.3, 37, 18.5 and 5.6 μm) on the porosity and microstructure of the product is investigated under Cpð3Al2 O3
2SiO2 Þs dT ð6Þ  Cpð3Al2 O3
2SiO2 Þ1 dT þ αð1 þ xÞΔHð f;MoSi2Þ þ7=9γð1 þ xÞΔHð f;3Al2O3
2SiO2Þ where, x is the number of moles of the product which is added to the reaction media as a diluent. For comparison, the adiabatic temperatures and the molten fractions of mullite and MoSi2 phases as a function of initial temperatures and dilution volume percentage are assembled in Fig. 1. It is clear that increasing the amount of diluents in the reaction medium at any value of the initial temperature leads to a parallel decrease in adiabatic temperature. This could be highly useful in controlling the microstructure of the produced product. The XRD of all samples produced with different dilution amounts varied from 10 to 30 wt.% (not shown) and revealed that the target composite of mullite/MoSi2 is the main phase without detection of any peaks for the reactant precursors. The addition of dilution did not affect the phase composition of the produced sample according to the fact that it is one of the products during the reaction between the starting precursors. The microstructure of the sample produced with different dilution percentages (20, 25, and 30 wt.%) are investigated and compared with that of the sample without dilution, Fig. 3. It reveals a regular decrease in the particle size of MoSi2 (from several mm to few μm) with a significant improvement in the homogeneity among sample phases with an increasing diluent volume percentage. At 20 vol.% diluents (Fig. 3b), although a large agglomerate of MoSi2 phase is still present in the combustion product, its amount as well as particle size are largely reduced compared with one without a dilution (Fig. 3a). However, a complete disappearance of agglomerated phase is achieved at 25 vol.% diluents (Fig. 3c). This could be attributed to the continuous decrease in the combustion temperatures and the molten fractions of mullite and MoSi2 phases by increasing the amount of diluents. With a further increase in the dilution addition to 30 vol.% (Fig. 3d), although a homogeneous microstructure is also produced, a higher degree of porosity could be clearly noticeable. The behavior of increasing sample porosity with dilution addition was clearly observed in all samples. The porosities of these samples were measured and given in Fig. 4. These measured porosities confirmed the observation from SEM images that the porosity of the sample increased with an increase in the amount of diluents. Although the sample having a 25 vol.% diluent has a higher porosity than the 0, 10 and 20 vol.% diluents' samples, it has the highest degree of homogeneity among its phases. Accordingly, a trial was made for enhancing the densification parameter of such sample via varying the pressing load during the reaction. The effect of different pressing loads (82–292 MPa) on the sample open porosity is investigated using the sample having a 25 vol.% dilution Fig. 5. BES images of combustion product containing 25 vol.% diluent pressed at 234 MPa, a) using 37 μm particle size Al, b) 5.6 μm particle size Al, and c) 37 μm particle size Al after 20 s of reaction. 28 Z.I. Zaki et al. / Int. Journal of Refractory Metals and Hard Materials 45 (2014) 23–30 Fig. 6. Effect of delay time on the sample porosity. the mechanical pressing of 234 MPa with the sample containing 25 vol.% diluent. This trial was made in order to reduce the sample porosity of the produced sample as much as possible. Fig. 4 illustrates the effect of Al particle size on the porosity of the samples. This figure shows that reducing the Al particle size to 5.6 μm leads to an increase in the porosity of the sample. This is because finer Al particle size usually leads to vigorous reactions. On the other hand, increasing Al particle size from 37 to 52.3 μm does not affect the porosity of the sample. Although the sample produced using Al 5.6 μm still has a low porosity (~4.6 vol.%), it suffers from non-homogeneity between phases such that agglomerations of MoSi2 are noticed as it can be seen from the SEM image (Fig. 5b). These results revealed that although the particle size of Al could play an important role during the reaction between the starting precursors, it did not affect the densification parameter of the sample using the modified SHS technique. Another trial for enhancing the sample porosity is to control the delay time at which the pressing load is applied. Delaying the time of pressing is important in some cases in order to give enough time for the gases evolved during combustion to release. This may help in reducing the sample porosity. Therefore the pressing load is investigated before the beginning of the reaction, during the reaction, and 5, 10, 15 and 20 s after the reaction has finished. The effect of delay time on the porosity of the samples is given in Fig. 6. It is concluded that pressing the sample at any time after starting the reaction and up to 10 s after the reaction has finished leads to almost the same sample porosity. Delaying the pressing time for more than 10 s leads to an increase in the sample porosity due to a dramatic reduction of the combustion temperature with time. This behavior is also noticed during SEM investigation (Fig. 5c). Composite properties Coefficient of thermal expansion Although optimizing the controlling parameters lead to a high density composite, it suffers from the existence of microcracks. The coefficient of thermal expansion (CTE) as a function of temperature (200–1250 °C) is Fig. 7. Coefficient of thermal expansion and electrical resistivity behavior of the dense product at different temperatures. Z.I. Zaki et al. / Int. Journal of Refractory Metals and Hard Materials 45 (2014) 23–30 29 Fig. 8. XRD of the combustion product treated in air at 1300 °C for 3 h. given in Fig. 7. It can be seen that the presence of MoSi2 (~18.96 vol.%) as a second phase increases the CTE of the mullite matrix (5.96–8.93×10−6 K−1) compared with CTE of pure mullite in the same temperature range (4.4–6.5×10−6 K−1). This means that during cooling, the mullite matrix will suffer from thermal stress that causes the development of such cracks. On the other hand, a simple calculation of volume changes shows that reaction (3) is accompanied by a decrease in volume equal to 18% due to the conversion of the reactants to the product. The fast reaction and sudden cooling behavior of combustion synthesis and the mismatch in the thermal expansion coefficient between MoSi2 and mullite may contribute to the formation of cracks. inhomogeneous. The addition of a 25 vol.% combustion product as a diluent was necessary to produce a homogeneous product. The lowest sample porosity was achieved upon applying 234 MPa mechanical pressure during the combustion reaction. Using Al with an average particle size less than 37 μm was found to increase sample porosity. The obtained high density composite has good oxidation resistance at 1300 °C and low electrical resistivity (1.8 Ω·cm) at room temperature. The measured coefficient of the thermal expansion of this product was 5.96– 8.93 × 10−6 K−1 in the temperature range of 200–1250 °C. References Electrical resistivity The electrical resistivity behavior at different temperatures is given in Fig. 7. The measured electrical resistivity of this composite is 1.8 Ω·cm at room temperature. This reflects its good electrical conductivity at room temperature which will facilitate its machining into complex shapes by using the economical electrical discharge machining (EDM) technique. This technique essentially needs the component to be electrically conductive (resistivity b 100 Ω·cm) [33]. By increasing the temperature, the electrical resistivity slightly increases as a common behavior for resistance heating elements. Oxidation resistance investigation Oxidation resistance of the dense combustion product was carried out in open atmosphere at 1300 °C for 3 h. No weight gains or weight losses were noticed. This ensures that this composite has good resistance to oxidation within the investigated temperature range. XRD pattern of the oxidized specimen at 1300 °C, Fig. 8, shows the existence of a SiO2 cristobalite phase and a weak diffraction pattern of alumina. There is no evidence for the decomposition of MoSi2 to the Mo-rich phase (Mo5Si3) as commonly reported for samples containing MoSi2 [31,32]. This could be attributed to the embedding of MoSi2 grains in the mullite matrix which prevents direct contact between MoSi2 grains and atmospheric oxygen. Conclusion Combustion synthesis under load methodology has been successfully used to produce a high density mullite/MoSi2 composite having less than 3.0 vol.% porosity. The sample treated without dilution was totally [1] Yao Z, Stiglich J, Sudarshan TS. Molybdenum silicide based materials and their properties. J Mater Eng Perform 1999;8:291–304. [2] Schneider H, Schreuer J, Hildmann B. Structure and properties of mullite—a review. J Eur Ceram Soc 2008;28:329–44. [3] Sun L, Pan J. Fabrication and characterization of TiC-particle-reinforced MoSi2 composites. J Eur Ceram Soc 2002;22:791–6. [4] Zakeri M, Ramezani M. Synthesis of MoSi2–TiC nanocomposite powder via mechanical alloying and subsequent annealing. Ceram Int 2012;38:1353–7. [5] Meng J, Lu J, Wang J, Yang S. Preparation and properties of MoSi2 composites reinforced by TiC, TiCN, and TiB2. Mater Sci Eng A 2005;396:277–84. [6] Ignat S, Sallamand P, Nichici A, Vannes B, Grevey D, Cicală E. MoSi2 laser cladding—a new experimental procedure: double-sided injection of MoSi2 and ZrO2. Surf Coat Technol 2003;172:233–41. [7] Köbel S, Plüschke J, Vogt U, Graule TJ. MoSi2–Al2O3 electroconductive ceramic composites. Ceram Int 2004;30:2105–10. [8] Raju GB, Basu B, Suri AK. Thermal and electrical properties of TiB2–MoSi2. Int J Refract Met Hard Mater 2010;28:174–9. [9] Micele L, Palombarini G, Guicciardi S, Silvestroni L. Tribological behaviour and wear resistance of a SiC–MoSi2 composite dry sliding against Al2O3. Wear 2010;269:368–75. [10] Manukyan KV, Kharatyan SL, Blugan G, Kocher P, Kuebler J. MoSi2–Si3N4 composites: influence of starting materials and fabrication route on electrical and mechanical properties. J Eur Ceram Soc 2009;29:2053–60. [11] Shon I-J, Lee D-M, Doh J-M, Yoon J-K, Ko I-Y. Consolidation and mechanical properties of nanostructured MoSi2–SiC–Si3N4 from mechanically activated powder by high frequency induction heated sintering. Mater Sci Eng A 2011;528:1212–5. [12] Lin GY, Costil V, Jorand Y, Fantozzi G. Experiments on packing and sintering of composite powder mixtures of MoSi2 + Al2O3 platelets. Ceram Int 1999;25:367–73. [13] Yeh CL, Chen WH. Combustion synthesis of MoSi2 and MoSi2–Mo5Si3 composites. J Alloy Compd 2007;438:165–70. [14] Ingemarsson L, Halvarsson M, Hellström K, Jonsson T, Sundberg M, Johansson LG, et al. Oxidation behavior at 300–1000 °C of a (Mo, W) Si2-based composite containing boride. Intermetallics 2010;18:77–86. [15] Guicciardi S, Swarnakar AK, Van der Biest O, Sciti D. Temperature dependence of the dynamic Young's modulus of ZrB2–MoSi2 ultra-refractory ceramic composites. Scripta Mater 2010;62:831–4. [16] Liu H-T, Wu W-W, Zou J, Ni D-W, Kan Y-M, Zhang G-J. In situ synthesis of ZrB2–MoSi2 platelet composites: reactive hot pressing process, microstructure and mechanical properties. Ceram Int 2012;38:4751–60. [17] McFayden AA, Thomas G, Petrovic JJ. Mullite whisker–molybdenum disilicide composites. Ceram Eng Sci Proc 1995;16:147–54. 30 Z.I. Zaki et al. / Int. Journal of Refractory Metals and Hard Materials 45 (2014) 23–30 [18] Brynsvold R. Processing of a mullite matrix, molybdenum disilicide reinforced composite. University of Washington; 1991. [19] S-k Zhao, Y Huang, C-a Wang, X-x Huang, J-k Guo. Mullite formation from reaction sintering of ZrSiO4/α-Al2O3 mixtures. Mater Lett 2003;57:1716–22. [20] Mazzei AC, Rodrigues JA. Alumina–mullite–zirconia composites obtained by reaction sintering: Part I. Microstructure and mechanical behaviour. J Mater Sci 2000;35:2807–14. [21] Sarydikmen H, Kushkonmaz N. Microstructure and phase analysis of ethyl silicatebonded alumina-zircon refractories. Refract Ind Ceram 2005;46:212–6. [22] Chen CY, Lan GS, Tuan WH. Preparation of mullite by the reaction sintering of kaolinite and alumina. J Eur Ceram Soc 2000;20:2519–25. [23] Chen G, Qi H, Xing W, Xu N. Direct preparation of macroporous mullite supports for membranes by in situ reaction sintering. J Membr Sci 2008;318:38–44. [24] Kong LB, Ma J, Huang H. Mullite whiskers derived from an oxide mixture activated by a mechanochemical process. Adv Eng Mater 2002;4:490–4. [25] Ribero D, Restrepo R, Paucar C, García C. Highly refractory mullite obtained through the route of hydroxyhydrogels. J Mater Proc Technol 2009;209:986–90. [26] Kaya C, He JY, Gu X, Butler EG. Nanostructured ceramic powders by hydrothermal synthesis and their applications. Microporous Mesoporous Mater 2002;54:37–49. [27] Somiya S, Yoshimura M, Suzuki M, Yanaguchi T. Mullite powder from hydrothermal processing. In: Somiya S, Davies RF, Pask JA, editors. Ceramic transactions mullite and mullite matrix composites. Westerville, OH: American Ceramic Society; 1990. p. 287–317. [28] Basu SN, Kulkarni T, Wang HZ, Sarin VK. Functionally graded chemical vapor deposited mullite environmental barrier coatings for Si-based ceramics. J Eur Ceram Soc 2008;28:437–45. [29] Botero CA, Jimenez-Piqué E, Kulkarni T, Fargas G, Sarin VK, Llanes L. Cross-sectional nanoindentation and nanoscratch of compositionally graded mullite films. Surf Coat Technol 2011;206:1927–31. [30] Hou P, Basu SN, Sarin VK. Structure and high-temperature stability of compositionally graded CVD mullite coatings. Int J Refract Met Hard Mater 2001;19:467–77. [31] Armas B, Sibieude F, Mazel A, Fourmeaux R, de Icaza Herrera M. Low-pressure chemical vapour deposition of mullite layers using a cold-wall reactor. Surf Coat Technol 2001;141:88–95. [32] Zamani S, Bakhsheshi-Rad HR, Shokuhfar A, Vaezi MR, Kadir MRA, Shafiee MRM. Synthesis and characterization of MoSi2–Mo5Si3 nanocomposite by mechanical alloying and heat treatment. Int J Refract Met Hard Mater 2012;31:236–41. [33] Cabouro G, Le Gallet S, Chevalier S, Gaffet E, Grin Y, Bernard F. Dense Mosi2 produced by reactive flash sintering: control of Mo/Si agglomerates prepared by high-energy ball milling. Powder Technol 2011;208:526–31. [34] Wu H, Li H-j, Lei Q, Fu Q-g, Ma C, Yao D-j, et al. Effect of spraying power on microstructure and bonding strength of MoSi2-based coatings prepared by supersonic plasma spraying. Appl Surf Sci 2011;257:5566–70. [35] Liu G, Li J, Chen K. Combustion synthesis of refractory and hard materials: a review. Int J Refract Met Hard Mater 2013;39:90–102. [36] Zaki ZI, Mostafa NY, Rashad MM. High pressure synthesis of magnesium aluminate composites with MoSi2 and Mo5Si3 in a self-sustaining manner. Ceram Int 2012;38:5231–7. [37] Feng P, Qu X, Akhtar F, Islam SH. Self-propagating high temperature synthesis of MoSi2 matrix composites. Rare Metals 2006;25:225–30. [38] Mishra SK, Das SK, Sherbacov V. Fabrication of Al2O3–ZrB2 in situ composite by SHS dynamic compaction: a novel approach. Compos Sci Technol 2007;67:2447–53. [39] Vallauri D, Shcherbakov VA, Khitev AV, Deorsola FA. Study of structure formation in TiC– TiB2–MexOy ceramics fabricated by SHS and densification. Acta Mater 2008;56:1380–9. [40] Khina BB, Formanek B, Solpan I. Limits of applicability of the “diffusion-controlled product growth” kinetic approach to modeling SHS. Phys B Condens Matter 2005;355:14–31. [41] Zaki ZI. Combustion synthesis of mullite–titanium boride composite. Ceram Int 2009;35:673–8. [42] Zaki ZI, Hessien MM, Ahmed YMZ. Synthesis of molybdenum silicide/mullite composites for high-temperature applications. Int J Self Propag High Temp Synth 2011;20:72–81.