An investigation on the in situ synthesis–sintering and mechanical properties of MoSi2–xSiC composites prepared by spark plasma sintering

An investigation on the in situ synthesis–sintering and mechanical properties of MoSi2–xSiC composites prepared by spark plasma sintering

Int. Journal of Refractory Metals and Hard Materials 48 (2015) 263–271 Contents lists available at ScienceDirect Int. Journal of Refractory Metals a...

3MB Sizes 5 Downloads 167 Views

Int. Journal of Refractory Metals and Hard Materials 48 (2015) 263–271

Contents lists available at ScienceDirect

Int. Journal of Refractory Metals and Hard Materials journal homepage: www.elsevier.com/locate/IJRMHM

An investigation on the in situ synthesis–sintering and mechanical properties of MoSi2–xSiC composites prepared by spark plasma sintering Soheila Esmaeily ⁎, Milad Kermani, Mansour Razavi, Mohammad Reza Rahimipour, Mohammad Zakeri Materials and Energy Research Center (MERC), P.O. Box 14155-4777, Tehran, Iran

a r t i c l e

i n f o

Article history: Received 30 July 2014 Received in revised form 7 September 2014 Accepted 19 September 2014 Available online 20 September 2014 Keywords: MoSi2 SiC Spark plasma sintering Toughness

a b s t r a c t Composites of MoSi2–x wt.% SiC (x = 5, 10, 15, 20) prepared using spark plasma sintering. The effect of temperature on the in-situ synthesis–sintering was investigated between 1100 °C and 1500 °C. X-ray diffraction patterns showed that at 1100 °C the reactions were incomplete and elementary diffraction peaks of Mo, Si and C still exist. With an increase in temperature from 1100 to 1300 °C the reactions were performed completely. The study showed that the sintering ability at higher temperature at the presence of enough mechanical pressure was better because the heat released from the reaction between Mo, Si and C causes higher temperature than the melting point of Si (1410 °C). Consequently the silicon would melt during the heating process. The molten Si can strengthen the interconnections and it has higher diffusion rate. Therefore, due to the liquid phase sintering and at the presence of mechanical pressure, the sintering ability at higher temperature is better than lower temperature. Scanning electron microscopy showed that with the addition of carbon, there was no silica phase in the microstructure of the synthesized samples, due to the formation of SiC. Therefore, it can be noted that the addition of carbon leads to better mechanical properties due to elimination of silica phase. © 2014 Elsevier Ltd. All rights reserved.

1. Introduction The interest in MoSi2 is due to its high melting point, good oxidation and corrosion resistance and low density [1]. However, MoSi2 has its intrinsic limitation in properties such as low ductility at low temperature. Indeed the mechanical behavior of MoSi2 can be divided into three ranges [1]: strong and brittle (up to approximately 1000 °C), strong and ductile (1000–1250 °C) and weak and ductile (above 1250 °C) [2]. It is known that during the processing of MoSi2 using hot pressing, hot isostatic pressing and spark plasma sintering (SPS) or other processes, SiO2 forms at grain boundary or inside the grains. This SiO2 phase leads to the deterioration of the mechanical properties at both ambient and elevated temperatures. Although several investigators reported the observation of silica in MoSi2 based materials, but there is no agreement on the morphology and distribution of silica in MoSi2 and the effect of processing variables on these characteristics. A number of researchers claimed that SiO2 wets the grain boundary or it concentrates in triple junction [3,4], while others observed globular silica particles both at grain boundary and inside the grains [5–7]. The introduction of reinforcement improves mechanical properties of MoSi2. For instance, many reinforcements have been investigated over the past years, these reinforcements include metals (Nb, W, Ti), nitrides (Si3N4, AlN), oxides (Al2O3, ZrO2), carbides (TiC, SiC, ZrC) and borides (ZrB2, TB2) [3–5]. Attempts have been made to improve ⁎ Corresponding author. E-mail address: [email protected] (S. Esmaeily).

http://dx.doi.org/10.1016/j.ijrmhm.2014.09.020 0263-4368/© 2014 Elsevier Ltd. All rights reserved.

mechanical properties of MoSi2 including: the introduction and control of second phase, tailoring interface properties, microstructural control and alloying [1]. The main criteria for the selection of second phase reinforcement have been reviewed in details elsewhere [5–9]. The most important issues are mechanical properties, coefficient of thermal expansion (CTE), density, chemical compatibility and interfacial characteristic. Among these various reinforcements, SiC is considered to be significantly effective due to its good compatibility with MoSi2. Many processing methods have been used to prepare MoSi2 composite with SiC particle. These methods include melting process [10], plasma spray deposition [11], mechanical alloying, self propagating high temperature synthesis, hot press, hot isostatic press, spark plasma sintering (SPS) of Mo, Si and C powders [12–15] or reaction of MoSi2 powder with C [16,17]. Except SPS method, the limitations of all these methods lie in that they are time consuming and/or the products are porous. In recent decades SPS has been widely employed for fabrication of many ceramics, metals, intermetallic compounds and different composites [18–20]. In this method the raw powders in a carbon die are pressed uniaxially and direct current (DC) is applied simultaneously. At early stage of the process, the powders are heated by spark discharge between particles and the carbon die is also heated by joule effect, so the powders are heated from the inside and outside [18,19]. Therefore this method leads to shortening the sintering time and good densification. In the present study we used spark plasma sintering (SPS) in order to synthesis and sinter MoSi2–x wt.% SiC (x = 5, 10, 15, 20) composites using Mo, Si and C powders in one step. The effect of temperature, mechanical pressure and composition on the synthesis, densification,

264

S. Esmaeily et al. / Int. Journal of Refractory Metals and Hard Materials 48 (2015) 263–271

a)

b)

5μm

50μm

c)

50μm

Fig. 1. SEM image of the raw materials used in this study, a) Mo, b) Si and c) C powder.

sintering, microstructure and mechanical properties of samples was also investigated. 2. Experimental 2.1. Mixing and drying Powders of molybdenum, silicon and graphite were used as starting materials with purity of 99.9%. Fig. 1 shows the scanning electron microscopy (SEM) images of the used raw materials. The atomic ratio of each element is based on the reaction (1): Mo þ ðx þ 2ÞSi þ xC ¼ MoSi2 þ xSiC:

ð1Þ

The powders dispersed in acetones according to stoichiometric ratio of MoSi2–(5, 10, 15, 20) wt.% SiC. The solids loading were approximately

50 wt.% and mixed using a high energy planetary ball mill for 5 h at 200 rpm. The ball to powder ratio was 5:1. After this step, in order to keep constituents from settling, drying was carried out using a hotplate, while stirring continuously. Then the powders loaded into the graphite die and the graphite die containing raw powders placed inside the SPS vacuum chamber and the powders compacted into a green body using 10 or 15 MPa uniaxial pressure depending on the maximum of pressure. 2.2. Synthesis and sintering of the compacted bodies The second step of the process, consisting of simultaneously reaction and consolidation of the mixed powder, performed using SPS apparatus. The synthesis–sintering process was performed under high pulsed direct current (between 1000 and 6000 A) in a vacuum atmosphere (8 Pa). Based on previous works by authors [18,19], half of the pressure

Fig. 2. Schematic of spark plasma sintering apparatus, a) vacuum chamber, b) the position of thermocouple and pyrometer.

S. Esmaeily et al. / Int. Journal of Refractory Metals and Hard Materials 48 (2015) 263–271

5

6 5

3 1373 K 2

1573 K 1873 K

Displacement (mm)

4 Displacement (mm)

265

1

4 3

20 MPa

2

30 MPa

1 0

0 -200

300

-1

800

1300

-1

1800

0

500

1000

1500

2000

Sintering time (S)

Sintering me (S)

Fig. 3. Variation of shrinkage displacement versus sintering time for MoSi2–20 wt.% SiC.

applied during the reaction and increased to a maximum amount after achieving to the expected temperature and maintained during holding time. The synthesis–sintering process was performed at three different temperatures (1100, 1300 and 1500 °C) in order to investigate the effect of temperature on the density, microstructure and mechanical properties of the samples. On the other hand, in order to investigate the effect of mechanical pressure on the density and mechanical properties, we synthesized and sintered another sample at 1500 °C using 30 MPa pressure. The sintering temperature was measured by a K type thermocouple between 25 and 1000 °C, and after 1000 °C, it was measured by a pyrometer. Fig. 2 shows the schematic of SPS apparatus and the position of thermocouple and pyrometer. The SPS holding time was 5 min for all of the samples. After holding time the pressure decreased to half of its previous amount and maintained during cooling. The products were typically disks of 50 mm in diameter and about 6 mm in height. The samples were polished with 80-grit silicon carbide and up to 4 μm with alumina–SiC slurry and finally cleaned in an ultrasonic ethanol bath in order to remove surface contamination from graphite foil. 2.3. Characterization The density of the products was evaluated using Archimedean's method. X-ray powder diffraction (XRD) measurements performed using a Philips PW 3710 diffractometer (40 kV and 30 mA) with Cu Kα radiation (k = 0.154051 nm). The microstructure of the powders and sintered samples was studied by scanning electron microscopy (SEM) with energy dispersive X-ray spectrometry (EDXS) using a Cambridge scanning electron microscope operating at 25 kV. Three-point bending tests were performed using an Instron 4411 universal testing machine, at a loading rate of 0.5 mm/min, at room

Fig. 5. Variation of shrinkage displacement versus the sintering time at 1500 °C using different mechanical pressure.

temperature, to determine the transverse rupture strength (TRS) of the segments. At least five tests were repeated for each specimen, and the results were averaged. The size of the sintered specimens for the three-point bending test was 20 × 5 × 5 mm. The hardness of the samples determined from the Vickers indentations was obtained with a load of 2 kg for 10 s. Fracture toughness values (KIc) of the product were determined from hardness experiments, by measuring crack lengths and using equation reported by Anstis et al. [21].

3. Results and discussion 3.1. The effect of temperature on the synthesis–sintering behavior Fig. 3 shows the variation of shrinkage displacement versus sintering time during reaction process for MoSi2–20 wt.% SiC sintered at different temperatures using 20 MPa pressure. It is worth noting that sample displacement represents an important indication of the occurrence of the reaction. The displacement toward negative or positive directions means expansion or shrinkage, respectively. As it is seen in this figure, with the exception of 1100 °C, the synthesis–sintering behavior at 1300 and 1500 °C is characterized by three different stages, i. e. initial expansion, abrupt shrinkage and slight shrinkage. In the first stage, due to the thermal expansion of gas and particles by heating, slight expansion took place. What was followed is that the reaction between Mo, Si and C took place, the displacement increased abruptly due to density enhancement resulting from molar volume change associated with the formation of MoSi2 and SiC. In the next stage, due to the increase in pressure from 10 to 20 MPa, a slight shrinkage occurred and displacement increased slightly.

100

Relave density (%)

96 92

5 wt% SiC 10 wt% SiC

88

15 wt% SiC 84

20 wt% SiC

80 76 15

20

25

30

35

Mechanical pressure (MPa) Fig. 4. Relative density of the sintered samples at different temperatures using 20 MPa.

Fig. 6. The effect of mechanical pressure on relative density at 1500 °C.

266

S. Esmaeily et al. / Int. Journal of Refractory Metals and Hard Materials 48 (2015) 263–271

2500

MoSi2

C Si Mo

c)

increase in relative density of the samples. For instance, the relative density of the MoSi2–20 wt.% SiC at 1100 and 1500 °C is 63.46% and 79.13%, respectively. In order to improve the relative density of samples, we increased the pressure from 20 to 30 MPa at 1500 °C. Fig. 5 shows the variation of shrinkage displacement versus sintering time for MoSi2–20 wt.% SiC sintered using 20 and 30 MPa at 1500 °C. As it is seen, the increment of pressure from 20 to 30 MPa, resulted in more shrinkage in this sample, consequently it was expected that the relative density of the sample sintered using 30 MPa pressure at 1500 °C was more than the sample sintered using 20 MPa pressure at the same temperature. Fig. 6 shows the relative density of the samples sintered using the 20 and 30 MPa at 1500 °C. As it is seen, the increment of pressure from 20 to 30 MPa at 1500 °C resulted in a significant increase in the relative density of the samples. For instance the relative density of the MoSi2– 20 wt.% SiC sintered using 20 MPa at 1500 °C was 79.13% while the relative density of this sample after sintering using 30 MPa and the same temperature was 96.8%.

Mo5Si3 (Mo4.8Si3C0.6) SiC

Intensity (a. u.)

2000

b)

1500

1000

a) 500

0 20

30

40

50

60

70

80

2θ (Degree)

Fig. 7. XRD pattern of synthesized products at, a) 1100 °C, b) 1300 °C and c) 1500 °C.

3.3. The effect of temperature on the microstructure of samples 3.2. The effect of temperature and pressure on density Fig. 7 shows the XRD pattern of synthesized products for MoSi2– 20 wt.% SiC at different temperatures. As it is seen, at 1100 °C, MoSi2 and Mo5Si3 diffraction peaks are found. On the other hand, a variety of elementary diffraction peaks of Si, C and Mo still exist. This means that at 1100 °C the reactions did not perform completely. With the

The relative density of the sintered samples using 20 MPa pressure at various temperatures is shown in Fig. 4. As it is seen in this figure, the relative density increased with increment of temperature. It should be noted that the increase in temperature did not result in a significant

a)

b) Si

porosity 20μm

20μm

c) 3

d)

1 2

3

Porosity

1

2

20μm

20μm

e) 1

2 3

5μm

Fig. 8. Scanning electron microscopy of the synthesized MoSi2–20 wt.% SiC at, a) 1100 °C and 20 MPa, b) 1300 °C and 20 MPa, c) 1500 °C and 20 MPa, d) 1500 °C and 30 MPa and e) at 1500 °C and 30 MPa in higher magnification.

S. Esmaeily et al. / Int. Journal of Refractory Metals and Hard Materials 48 (2015) 263–271

267

Fig. 9. EDXS patterns of the synthesized MoSi2–x wt.% SiC, a) MoSi2, b) Mo4.8Si3C0.6, c) SiC.

increment of temperature to 1300 °C the elementary diffraction peaks of Mo, Si and C became weaker and the diffraction peaks of SiC are observed and at the same time the diffraction peaks of MoSi2 and Mo5Si3 became stronger. It should be noted that the type of the Mo5Si3 phase is much close to the carbon stabilized form which its approximate composition is Mo4.8Si3C0.6. With further increase in temperature up to 1500 °C, the diffraction peaks of SiC, MoSi2 and Mo5Si3 get stronger. Fig. 8 shows the microstructure of the synthesized MoSi2–20 wt.% SiC at various temperatures using different mechanical pressures. As it is seen in Fig. 8a, three phases are observed at 1100 °C, because the reactions did not perform completely at this temperature, it was hard to determine the different phases which were identified by XRD. However it seems that the dark phase is carbon, the gray phase contains MoSi2 and the white phase contains Mo, and Mo5Si3. The increment of temperature from 1100 to 1300 °C resulted in the formation of a porous microstructure which did not contain unreacted Mo, Si, and C (Fig. 8b). Porosities form due to the carbon reduction of oxides. The carbon reduction of oxides resulted in the release of gaseous species. Fig. 8c shows the microstructure of the sample sintered using 20 MPa pressure at 1500 °C. Fig. 8d and e shows the microstructure of the sample sintered using 30 MPa pressure at 1500 °C. As it is seen in Fig. 8c and d, it is obvious that the sample sintered using lower mechanical pressure contained more porosities and had lower relative density (Fig. 8c) in comparison with the samples sintered using higher mechanical pressure (Fig. 8d). Three phases are observed in Fig. 8c and d. EDXS patterns of synthesized phases are shown in Fig. 9. Semi-quantitative analysis showed that the gray phase (point 1 in Fig. 8c–e) consists Mo and Si

which its atomic ratio is about 1:1.978 and the light phase (point 2 in Fig. 8c–e) consists of Mo, Si and C, while the dark one (point 3 in Fig. 8c–e) contains Si and C, which its atomic ratio is about 1:1.011. Therefore it is reasonable to think that the gray phase is MoSi2, the light phase is likely to be Mo4.8Si3C0.6, and the dark one is SiC. It should be noted that as it is seen in Fig. 8d and e, larger amount of Mo4.8Si3C0.6 is visibly present in higher sintering temperature. It seems that the SiC particles located at MoSi2 grain boundaries have a relatively larger size in comparison with those within the matrix. This means that the grain size of SiC particles may vary in different regions of the sample. It is necessary to note that the enough high sintering temperature or intensity of the exothermal reactions in Mo, Si and C system causes higher temperature than the melting point of Si (1410 °C), and it would be melted during the continuous reaction. Molten Si has relatively higher mobility and diffusion rate; consequently it may strengthen the interconnections between particles and prompts diffusions of reactants [15]. Hu et al. [15], reported that when the liquid phase penetrates between the particle boundaries, a tremendous capillary force can be developed which leads to particle rearrangement and pore filling for closer packing in the presence of a liquid phase, consequently liquid phase sintering, grain growth and densification take place simultaneously and a higher relative density can be obtained under mechanical pressure [15]. It should be noted that the mechanical pressure has a significant effect on the relative density of samples. Because, as it is shown in Fig. 6, for example, MoSi2–5 wt.% SiC sample sintered at 1500 °C using 20 MPa pressure has relative density about 89.21% while with increment of pressure from 20 to 30 MPa, the relative density of this sample

40000 20000

Temperature (K)

Gibbs free energy (Kj)

0 300

500

700

900

1100

1300

1500

1700

1900

-20000 Mo2C -40000

Si + Mo2C

-60000

SiC MoSi2

-80000

MoSi2 + SiC -100000 -120000 -140000 -160000 Fig. 10. Ellingham diagram of various materials [22].

268

S. Esmaeily et al. / Int. Journal of Refractory Metals and Hard Materials 48 (2015) 263–271

0.71

10

Log of shrinkage

1/2

0.69 0.68 0.67

20 wt% SiC

0.66

15 wt% SiC

0.65

10 wt% SiC

= 3.31

0.64 5 wt% SiC

= 3.25.

0.63

= 3.22

0.62

= 3.38

0.61 0.6 2.47

2.57

2.67

2.77

2.87

7 6 5

0



Mo þ 2Si →MoSi2 ΔGT ¼ −131800 þ 65:1 T J:

ð2Þ

ð3Þ

J:

These reactions have the Gibbs free energy of − 16378 and − 41200 J at 1500 °C, respectively [22]. The negative ΔG shows that these reactions can be performed during heating process at 1500 °C. Chemical affinity (A) definition is the negative partial derivation of Gibbs energy (G) with respect to the extent of reaction (ε) at constant pressure and temperature [23]:   δG P; T : δε

5

10

20

25

ð4Þ

Fig. 13. Variation of fracture toughness versus the amount of SiC addition.

A question may be raised in mind: Why Mo2C was not formed instead of SiC during heating process according to the reaction (5) [22]? 

2Mo þ C → Mo2 C ΔGT ¼ −28032 þ 0T J:

On the basis of this equation, both reactions (2) and (3) have positive chemical affinity, but it should be noted that in the same extent of reaction (ε), reaction (2) has higher chemical affinity because of its bigger ΔG. However, the heat released from the first reaction leads to the initiation of second reaction. Consequently, it seems that these reactions take place simultaneously.

ð5Þ

We can answer the question using chemical affinities of reactions (3) and (5). As discussed earlier, chemical affinity of reaction (3) is more than reaction (5) due to its bigger ΔG. Consequently, reaction (3) took place before reaction (5) and all of the graphite powders consumed by reaction (3). It should be noted that the amount of graphite powder in starting composition for the formation of 5, 10, 15 and 20 wt.%– SiC is based on reaction (3). On the other hand, we can answer this question in another way according to the Ellingham–Richardson diagram of these carbides (Fig. 10). Variation of Gibbs free energy versus temperature is shown in Ellingham–Richardson diagram for the formation of SiC and Mo2C (reactions 3 and 5). According to this diagram, each line which is lower than the other line, the product of its reaction is more stable than the other line. In other words, the products of lower line are more stable than the higher line. As it is seen in Fig. 10, the line of SiC formation is lower than the line of Mo2C formation, therefore SiC is more stable than Mo2C. This theory can be explained as reaction (5)–reaction (3): 

Vicker’s Hardness (GPa)

15

Amount of SiC (wt %)

at the same temperature and similar holding time increased to 99.4%. This is due to the effect of mechanical pressure on the diffusion rate and mobility of the liquid phase. Consequently it seems that the increase in mechanical pressure leads to promotion and increase in diffusion rate during liquid phase sintering. Formation of MoSi2 and SiC are on the base of two following reactions:



8

4

Fig. 11. Log–log plots of shrinkage for MoSi2–x wt.% SiC at 1500 °C and 30 MPa.

Si þ C →SiC ΔGT ¼ −53429:7 þ 6:9 T

9

2.97

Log of me (min)



Fracture Toughness (MPa. m )

0.7

 ð5Þ

 ð3Þ

SiC þ 2Mo →Si þ Mo2 C ΔGT ¼ ΔGT −ΔGT

¼ 25397:7−6:9T J: ð6Þ

16

The positive ΔG of this reaction (13164) at 1500 °C indicates that it cannot be performed during heating process [22]. It should be noted that this means that Mo cannot attract the carbon of SiC on the base of the above reaction. On the other hand, Si can attract carbon of Mo2C to form SiC on the base of opposite route of above reaction.

14

3.4. Isothermal sintering

12

It seems that the sintering processes that go along with temperature rise can generally be envisaged as [19]: the formation of MoSi2, SiC and Mo5Si3 due to the reaction between Mo, Si and C; homogenization and limited particle sliding and rearrangement when the liquid phase (liquid Si) starts to form and wet the grain boundaries; dissolution of the solid phases in the liquid phase; mass transport through the grain boundary liquid phase by solution, diffusion and precipitation of MoSi2, SiC and a little Mo5Si3. Most of these processes lead to densification, and it is difficult to separate one process from the other since they overlap and are interwoven together. The sintering of MoSi2–SiC composite is through a liquid phase mechanism, and the main factors that

10 8 6 0

5

10

15

20

Amount of SiC (wt %) Fig. 12. Variation of hardness versus amount of SiC addition.

25

S. Esmaeily et al. / Int. Journal of Refractory Metals and Hard Materials 48 (2015) 263–271

a)

b)

c)

d)

269

Fig. 14. The effects of SiC addition on the fracture toughness of MoSi2–x wt.% SiC, a) 5 wt.% SiC, b) 20 wt.% SiC, c) higher magnification of selected zone in part b, and d) crack bridging in MoSi2–20 wt.% SiC.

affect sintering behavior are the liquid phase volume and its properties such as viscosity. As it was noted earlier, during the sintering process, the specimens were rapidly heated up to 1500 °C and the sintering was monitored with time at a constant temperature. According to Kingery's model for liquid phase sintering [24–26], the whole densification process can be divided into the following steps: (1) formation of liquid phase leading to initial particle rearrangement; (2) solution of the solid phase in the liquid, followed by diffusion and precipitation, and (3) the development of a solid framework. Solution–diffusion–reprecipitation is considered as the main step in liquid phase sintering. The relative shrinkage during this stage can be described by the following expression: 1 ΔL ¼ kt n L

ð7Þ

where k is a value involving parameters such as temperature, liquid film thickness, grain size and diffusion coefficient in the liquid phase, and t is the time. The value n symbolizes the rate controlling step in the 800

Bend strength (MPa)

750 700 650 600 550 500 450 400 0

5

10

15

20

25

Amount of SiC (wt %) Fig. 15. The variation of transverse rupture strength (TRS) versus amount of SiC reinforcement.

dissolution–diffusion–precipitation processes. It should be noted that it is also a grain shape dependent factor. For prismatic and acicular particles, a n value of 3 indicates that phase–boundary reaction is the rate controlling step, and n = 5 suggests that the rate controlling step is diffusion, while for equiaxed grains, n values of 2 and 3 stands for the same rate controlling steps. Fig. 11 shows the isothermal shrinkage change with time as plotted on a log–log scale. The plots indicate relatively clearly two or three different stages. After the first couple of minutes, the curves follow a straight line up to 10–15 min with n values mostly close to 3 which is considered as the first stage. It may be argued that the dissolution of solid phase particles into the already formed liquid phase is the major on-going process. Consequently phase boundary reaction is the rate controlling step. After this period of time, the curves are flattened into straight line with n values close to 5. Thus it may be considered that the rate controlling step turns from reaction to diffusion. This second stage carries through 15–20 min. 3.5. Mechanical properties of the samples Fig. 12 shows the variation of hardness versus the increase in the amount of SiC addition for samples sintered at 1500 °C using 30 MPa. As it is seen in this figure, the hardness of the samples increased with an increment of amount of SiC addition. This is due to the higher hardness of SiC in comparison with MoSi2. Fig. 13 shows variation of fracture toughness versus weight percent of SiC for samples sintered at 1500 °C using 30 MPa. As it is seen in this figure, the fracture toughness of samples increased with an increase in the amount of SiC addition. It should be noted that the room temperature fracture toughness of all the MoSi2 matrix composites containing SiC reinforcements is higher than that of monolithic MoSi2 produced by authors in the previous researches [18,19]. Fig. 14 shows the effect of SiC additions on the type of cracking observed at room temperature hardness indent. A transition from intergranular cracking in the MoSi2–5 wt.% SiC specimen (Fig. 14a) to transgranular cracking (arrowed in Fig. 14b) in MoSi2–20 wt.% SiC sample (Fig. 14b and c) is

270

S. Esmaeily et al. / Int. Journal of Refractory Metals and Hard Materials 48 (2015) 263–271

a)

b)

Fig. 16. SEM micrograph of fractured surface of a) 5 wt.% SiC, b) 20 wt.% SiC.

apparent. It should be noted that with an increase in the amount of SiC addition, the crack segments run both along the MoSi2–SiC phase boundaries and through the SiC phase itself (Fig. 14b and c). On the other hand crack bridging of the MoSi2–20 wt.% SiC is evident in Fig. 14d. The variation of transverse rupture strength (TRS) or bending strength versus the weight percent of SiC addition for specimens sintered at 1500 °C using 30 MPa is shown in Fig. 15. The mechanism of improving transverse rupture strength can be observed in Fig. 16. As the amount of SiC reinforcement increases, a transition from intergranular fracture to transgranular is observed in MoSi2–5 wt.% Si and MoSi2–20 wt.% SiC, respectively. It seems that the removal of SiO2 promotes transgranular fracture at an ambient temperature. It should be noted that as it is seen in Fig. 15, with an increase in the amount of SiC reinforcement the transverse rupture strength of the specimens increased from 520 MPa in MoSi2–5 wt.% SiC to 721 MPa in MoSi2– 20 wt.% SiC. 4. Conclusions MoSi2 composites can be in situ synthesized–sintered in one step using SPS apparatus. XRD results showed that at 1100 °C the reactions did not perform completely but with the increase in temperature from 1100 to 1300 °C the reactions performed completely. It should be noted that the heat released from the reaction between Mo, Si and C causes higher temperature than the melting point of Si (1410 °C) and the silicon would melt during the heating process. The molten Si can strengthen the interconnections and it has higher diffusion rate. Consequently, due to the liquid phase sintering and at the presence of enough mechanical pressure, the sintering ability at higher temperature is better than lower temperature. The formation of MoSi2 and SiC is according to the reactions (2) and (3). Based on the Gibbs free energies of these reactions and chemical affinity definition (Eq. (4)), it seems that reactions (2) and (3) take place simultaneously. On the other hand, it is reasonable to think that why Mo2C did not form during the heating process. As it was discussed earlier, due to bigger Gibbs free energy and chemical affinity for the formation of SiC in comparison with Mo2C, the reaction (3) takes place sooner than the reaction (5). Therefore, it can be noted that all of the graphite powders consumed by reaction (3). On the other hand, according to Ellingham–Richardson diagram (Fig. 10), the line of SiC formation is lower than the line of Mo2C formation, which means that the thermodynamic condition is appropriate for the formation of SiC. The densification process is achieved by a liquid phase mechanism. However, in the solution–diffusion–reprecipitation process, two substages have been identified. It is proposed that the transition from the first sub-stage to the second one is characterized by the shift of the rate controlling mechanism from solution–precipitation to diffusion. Or it may be argued that the rate controlling mechanism in liquid

phase sintering of the MoSi2–SiC materials is transient while the Kingery Model appears to be applicable. With an increment in the amount of SiC, the bending strength and fracture toughness of the samples increased, because a transition from intergranular to transgranular fracture mode was seen as the amount of SiC reinforcement increased. On the other hand the hardness had a similar manner to that for bending strength and fracture toughness. The reason of the increment in hardness value of the composites with the increase in amount of SiC addition is the higher hardness of SiC in comparison with MoSi2.

References [1] A. Costa e Silva, M.J. Kaufman, Applications of in situ reactions to MoSi2-based materials, Mater. Sci. Eng. A 195 (1995) 75–88. [2] D.A. Hardwick, P.L. Martin, S.N. Patankar, J.J. Lewandowski, Effects of processing on properties of MoSi2, 1993. [3] J.D. Cotton, Y.S. Kinm, M.J. Kaufman, Intrinsic second-phase particles in powderprocessed MoSi2, Mater. Sci. Eng. A 144 (1991) 287–291. [4] T.E. Mitchell, R.G. Castro, J.J. Petrovic, Dislocations, twins, grain boundaries and precipitates in MoSi2, Mater. Sci. Eng. A 155 (1992) 241–249. [5] S. Malloy, A.H. Heur, J.J. Lewandowski, J.J. Petrovic, Carbon additions to molybdenum disilicide: improved high-temperature mechanical properties, J. Am. Ceram. Soc. 74 (1991) 2704–2706. [6] R.M. Aikin Jr., On the ductile-to-brittle transition temperature in MoSi2, Scr. Metall. Mater 26 (1992) 1025–1030. [7] R.K. Wade, J.J. Petrovic, Fracture modes in MoSi2, J. Am. Ceram. Soc. 75 (1992) 1682–1684. [8] K.K. Chawla, Composite materials science and engineering, Springer, New York, 1987. [9] P.F. Bechter, Microstructural design of toughened ceramics, J. Am. Ceram. Soc. 74 (1991) 255–269. [10] D.J. Tilly, J.P.A. Lofvander, C.G. Levi, In-situ growth of SiC in MoSi2 by melt processing, Mater. Res. Soc. 273 (1992) 273–295. [11] R.G. Castro, R.W. Smith, A.D. Rollett, P.W. Stanek, Ductile phase toughening of molybdenum disilicide by low pressure plasma spraying, Mater. Sci. Eng. A 155 (1992) 101–107. [12] C.H. Henager, J.L. Brimhall, J.P. Hirth, Synthesis of a MoSi2–SiC composite in situ using a solid state displacement reaction, Mater. Sci. Eng. A 155 (1992) 109–114. [13] L. Xiao, Y.S. Kim, R. Abbaschian, Ductile phase toughening of MoSi2—chemical compatibility and fracture toughness, Mater. Res. Soc. Symp. Proc. 194 (1990) 399–702. [14] J. Subrahmanyam, R.M. Rao, Combustion synthesis of MoSi2–SiC composites, J. Am. Ceram. Soc. 78 (1995) 487–490. [15] Q. Hu, P. Luo, Y. Yan, Microstructures and densification of MoSi2–SiC composite by field-activated and pressure-assisted combustion synthesis, J. Alloy Compd. 468 (2009) 136–142. [16] S.A. Maloy, J.J. Leandowski, A.H. Heuer, J.J. Petrovic, Effects of carbon additions on the high temperature mechanical properties of molybdenum disilicide, Mater. Sci. Eng. A 155 (1992) 159–163. [17] S.A. Maloy, A.H. Heuer, J. Lewandowski, J. Petrovic, Carbon additions to molybdenum disilicide: improved high-temperature mechanical properties, J. Am. Ceram. Soc. 74 (1991) 2704–2706. [18] M. Kermani, M. Razavi, M.R. Rahimipour, M. Zakeri, The effect of temperature on the in situ synthesis-sintering and mechanical properties of MoSi2 prepared by spark plasma sintering, J. Alloys Compd. 585 (2014) 229–233. [19] M. Kermani, M. Razavi, M.R. Rahimipour, M. Zakeri, The effect of mechanical alloying on microstructure and mechanical properties of MoSi2 prepared by spark plasma sintering, J. Am. Ceram. Soc. 593 (2014) 242–249. [20] D. Sciti, M. Nygren, Spark plasma sintering of ultra refractory compounds, J. Mater. Sci. 43 (2008) 6414–6421.

S. Esmaeily et al. / Int. Journal of Refractory Metals and Hard Materials 48 (2015) 263–271 [21] G.R. Anstis, P. Chantikul, B.R. Lawn, D.P. Marshall, A critical evaluation of indentation techniques for measuring fracture toughness: I, direct crack measurements, J. Am. Ceram. Soc. 64 (1981) 533–539. [22] O. Kubaschewski, Materials thermochemistry, Pergamon Press, Oxford, 1993. [23] B.E. Smith, Basic chemical thermodynamics, University Press, Oxford, 2004. [24] J.D. Bolton, A.J. Gant, Microstructural development and sintering kinetics in ceramic reinforced high speed steel, Powder Metall. 40 (1997) 143–151.

271

[25] W.D. Kingery, Densification during sintering in the presence of a liquid phase. I. Theory, J. Appl. Phys. 30 (1959) 301–307. [26] W.D. Kingery, M.D. Narasimhan, Densification during sintering in the presence of a liquid phase. II. Experimental, J. Appl. Phys. 30 (1959) 307–310.