Effect of Ball to Powder Weight Ratio on the Mechanochemical Synthesis of MoSi 2-TiC Nanocomposite Powder

MoSi 2 -TiC nanocomposite powders were successfully synthesized with different ball to powder weight ratios (BPR) by ball milling of Mo, Si, Ti and graphite elemental powders. Formation of this composite was studied by X-ray diffraction (XRD). Morphology and microstructure of the milled powders were monitored by scanning and transmission electron microscopy (SEM and TEM), respectively. There was incomplete formation in BPR 5:1 after 30 hours of milling, however, the formation of this composite was completed after 10 hours in BPRs 15:1 and 20:1. Higher BPRs with longer milling time led to partially transformation of β to α MoSi 2 . Based on Rietveld refinement analysis, and subsequent verification by TEM image, nanostructure powders with the mean grain size less than 25 nm were obtained in all BPRs. Very fine submicron powders in agglomerated ones for BPRs 10:1 and 15:1were obtained at the end of milling.


Introduction
With the rapid development of aerospace technology, it is urgent to study the new material used in high temperature structural components.MoSi 2 intermetallic compound has become an attractive candidate for structural application at temperatures up to 1600 °C because of its interesting combination of thermo-physical and mechanical properties 1,2 such as high strength, ductility and oxidation resistance.Because of its relatively high conductivity, MoSi 2 and MoSi 2 -based materials can be electro-discharge machined.However, it exhibits ceramic-like brittleness at room and metal-like plasticity at elevated temperatures.
The major problem impeding the use of MoSi 2 as a high temperature structural material is its brittleness at low temperatures and low strength and creep resistance at temperatures above 1200 °C.An effective method to solve both of these problems is the addition of brittle or ductile reinforcements as second phase in MoSi 2 matrix.MoSi 2 is thermodynamically stable with a wide variety of potential ceramic reinforcements for composites, including SiC, Si 3 N 4 , ZrO 2 , Al 2 O 3 , TiB 2 , TiC etc.Its stability extends essentially to the full range of important structural ceramic materials [3][4][5][6][7][8][9][10] .TiC is another probable reinforcement for MoSi 2 with potential advantages: (1) it has a brittle-to-ductile transition above 600 °C, (2) the thermal expansion coefficient of TiC is virtually same as that of MoSi 2 , and (3) thermodynamic calculations indicate that TiC and MoSi 2 should not react at the temperatures used for densification of MoSi 2 [11]   .Preparation in nanostructure condition is another approach to improve mechanical properties 12 .
MoSi 2 -TiC composite can be produced easily by direct mixing of MoSi 2 and TiC powders.But the resulting heterogeneous microstructure and high cost of the starting materials are two important drawbacks of this method [13][14][15][16][17] .Alternatively, MoSi 2 -TiC nanocomposite powder can be synthesized in situ through high energy reactive milling of a mixture of Mo, Si, Ti and graphite powders.The formation of compounds via solid-state reaction that occurs during ball milling, called mechanical alloying (MA), is a promising way of producing such composites.During MA, both matrix and reinforcement are formed through in situ process, which will promote suitable bonding between matrix and reinforcement.Moreover, a homogeneous distribution of fine reinforcing particles can be obtained by the MA process [18][19][20] .There is no report on the synthesis of this composite except of our previous work.In that work MoSi 2 -30 wt.(%) TiC nanocomposite powder was successfully synthesized by ball milling and following heat treatment.Results showed that the synthesis of this composite begins after 10 hours of milling and progresses gradually up to 30 hours of milling.MoSi 2 -TiC composite was completely synthesized after annealing of 30 hours milled powder at 900 °C [21] .Effect of BPR on the mechanochemical synthesis was investigated in the other systems.For example FeAl-Al 2 O 3 composite was formed after 120, 270 and 360 minutes of milling at the BPRs of 5:1, 10:1 and 15:1 respectively.It means that in the higher BPR, the shorter milling time is required for the synthesis 22 .
The aim of this work is in situ synthesis of MoSi 2 -TiC nanocomposite powder by mechanical milling of the corresponding elemental powders with different BPRs.Formation of this composite has been studied by thermodynamic discussions and its microstructure has been investigated by Reitveld refinement analysis.XRD profiles were recorded on a Siemens diffractometer (30 kV and 25 mA) with Cu Kα 1 radiation (l.5404 A°).The Recorded XRD patterns were used for calculation of crystallite size and strain.Mean grain size and microstrain were calculated on the basis of Rietveld refinement method by using of X'Pert high score plus software (developed by PANalytical BV Company, Almelo, the Netherlands, and version 2.2b).In this method, peak profile fitting, size broadening and strain broadening were calculated based on the following equations 23 .

Experimental Procedures
( ) 1 180 4 2 ln 2 100 where G ik is Pseudo-Voigt function, C 0 = 2, C 1 = 4 ln 2, H k is full width at half maximum of the K th brag reflection, γ is shape parameter, X ik = (2θ i -2θ k )/H k .D i , η i , λ, U and W are grain size function, strain function, wavelength, strain parameter and size parameter of peak profile, respectively.
In the size and strain functions, i and std are referred to analyzed and standard samples, correspondingly.In this project, a pure MoSi 2 that annealed at 1500 °C for 10 hours was used as standard material for deconvolution of instrumental broadening.Structural observations of milled powders were carried out with a Philips EM208 TEM operating at 200 kV.The powders were ultrasonically dispersed in methanol.One drop of this suspension was placed on a copper grid for TEM observation.The morphology and particle size of samples were examined by using a Philips (XL30) SEM operating at 30 kV.Iron contamination of milled powders was measured by inductivity coupled plasma (ICP) method.

Synthesis
Effect of BPR was studied on the formation of MoSi 2 -TiC composite powders.As received powders including the mixture of Mo, Si, Ti and graphite elemental powders were blended to give the desired MoSi 2 -30 wt.(%) TiC. Figure 1 shows the XRD patterns of 10 hours milled powders with different BPRs.There is no reaction in BPR 5:1 and the as-received materials reflections can be seen after 10 hours of milling.Graphite reflections were disappeared as a result of large mass absorption coefficient of Mo and Ti in comparison with graphite (Table 1).In fact, the absorption of graphite reflections by Mo and Ti was promoted by the intimate mixing and size decreasing in the early stage of milling 24 .
With increasing milling energy at BPR 10:1, a partially reaction was performed between the starting materials.However, the strong Mo peak can still be seen in the pattern of this sample.Higher milling energy in BPR 15:1, led to the full reaction and formation of β-MoSi 2 and TiC.β-MoSi 2 is high temperature polymorph (HTP) and is unstable during room temperature.In fact, high exothermic reactions take place and increase the temperature of the powder particles.This condition favors the formation of the high temperature phase (β-MoSi 2 is stable in the temperatures above 1900 °C) 1 .Formation of HTP of MoSi 2 can be explained as follow; it is now well recognized that the structure and constitution of advanced materials can be better controlled by processing them under non-equilibrium (or far-from-equilibrium) conditions.Amongst many processes, which are in commercial use, MA has been receiving serious attention from researchers.The central underlying technique is to synthesize materials in a non-equilibrium state by energizing and quenching.The energization involves bringing the material into a highly non-equilibrium (meta-stable) state by some external dynamical forcing, such as mechanical energy 25 .On the basis of above discussion, non-equilibrium condition of milling leads to the formation of HTP of MoSi 2 that is unstable at room temperature.The synthesized β-MoSi 2 was partially transformed to α-MoSi 2 at higher BPR (20:1).
As discussed in Figure 1, formation of MoSi 2 -TiC composite was completed after 10 hours milling in BPRs 15:1 and 20:1.Two samples with shorter milling times (3.5 and 7 hours) was studied for the determining the exact formation time of this composite.Figure 2a shows there is no reaction in BPR 15:1 after 3.5 hours of milling.On the other hand, in BPR 20:1, some minor MoSi 2 was formed at this stage of milling.Increasing milling time to 7 hours in BPR 15, led to the formation of MoSi 2 in minor amount (Figure 2b).In the BPR 20:1, at this stage of milling, the intensity of Mo peak was significantly decreased due to the formation of β-MoSi 2 .
The BPRs 5:1 and 10:1 were compared at longer milling times (20 and 30 hours) due to their incomplete formation of MoSi 2 -TiC composite.Figure 3a shows some minor amount of starting materials was reacted to form MoSi 2 after   The beginning and completing time for the formation of MoSi 2 -TiC composite with the end product for all BPRs were summarized in Table 2.With increasing BPR, the beginning time was significantly decreased.Higher BPR led to the higher milling energy.It means that the activation energy for the formation of MoSi 2 -TiC composite is provided at the shorter milling time. .Higher milling energy also affect the completing time for the formation of this composite.Table 2 shows that the formation of this composite was not completed after 30 hours in BPR 5:1.However, in BPRs 15:1 and 20, this composite was formed after only 10 hours.It means that at higher BPRs, the higher energy per time is introduced to the milled powders.In the other words, the kinetic of MoSi 2 -TiC formation increases at higher BPRs.The long milling period for the formation of this composite at all BPRs indicates that it was performed in gradual mechanism.The longer milling time (10:1, 30 hours) or higher BPR (20:1, 10 hours) led to the partially transformation of β-MoSi 2 to α-MoSi 2 .It may be due to the activation of recovery mechanism or from the mechanical energy heating.Higher BPR with longer milling time led to the heating of milled powders from mechanical energy.Introducing of this energy to the milled powder promotes the transformation of meta-stable β-MoSi 2 to stable α-MoSi 2 .

Morphology and Microstructure
Figure 4 shows the morphology and particles size of 30 hours milled powders for all BPRs.In BPR 5:1, very small submicron particles with spherical morphology were obtained after 30 hours of milling (Figure 4a).A few small agglomerated particles can be seen in this sample.Higher BPR (10:1) led to the excessive agglomeration because of  higher milling energy.Cold welding and particles fracturing are in competition during milling process.For preventing excess agglomeration, Stearic acid was used as process control agent (PCA).In the lower BPR (5:1), the used PCA acts as surfactant and decreases the cold welding and agglomeration process.At higher BPR (10:1), the added PCA is decomposed due to the formation of MoSi 2 -TiC composite and its heat releasing.Therefore, there is no inhibitor for cold welding.All of primary particles were agglomerated at this BPR (Figure 4b).With increasing milling energy at higher BPR (15:1), cold working resulted in work hardening and fracturing of milled powders.Figure 4c shows that the amount and size of agglomerates were significantly decreased.These phenomena continue at higher BPR (20:1), so few agglomerates can be seen in this sample (Figure 4d).Compositions of the milled powders were analyzed by energy dispersive spectroscopy (EDS) that was performed on the marked particles in Figure 4.The results of these analyses were presented in Figure 5.In all BPRs, there are Mo, Si and Ti peaks that confirm the composition of MoSi 2 and TiC.As expected, carbon cannot be detected by this method.There is another peak that attributed to Fe in all BPRs.This peak intensity, increases at higher BPRs.Fore confirming the existence of Fe peak, the Fe amount of 30 hours milled powders was measured by ICP method.The results of this measurement were presented in top right corner of each BPR in Figure 5.As seen, the Fe content increases from the minimum value of 0.23 wt.(%) in BPR 5:1 to 8 wt.(%) in BPR 20:1.These results indicate that the higher milling energy led to the higher wearing rate of steel cup and balls.In lower BPRs, this Fe impurity can be neglected due to its minor amount while in higher BPR; it can be removed by conventional leaching methods.
The mean grain size and microstrain of the milled powders were calculated on the basis of Reitveld refinement method (Figure 6).Longer milling time as well as higher BPRs led to the smaller mean grain size (Figure 6a).On the other hands, the milled powders with larger micro strain were obtained at higher BPRs and longer milling times (Figure 6b).As seen in both parameters, a steady state condition was obtained at longer milling time.It means that there is equilibrium between the grain size lessening and recovery mechanisms at longer milling times.In all BPRs and milling times, a nanostructure powders was obtained.The minimum mean grain size acquired for 30 hours-milled powder with 20:1 BPR.The bright field TEM image of 30 hours-milled sample with BPR of 10:1 was shown in Figure 7 due to the complete formation of MoSi 2 -TiC with the lower Fe impurity.A mean grain size of 12 nm was obtained for this sample by Rietveld method.Figure 7 shows that it is approximately in consistent with the Rietveld result.

Conclusions
In this research, the effect of the BPR was studied on the formation of MoSi 2 -TiC composite powder.Beginning and completing time for the formation of this composite were decreased with increasing BPR.High temperature polymorph of MoSi 2 (β) was obtained at higher BPR or longer milling time.The minimum and maximum Fe impurity of 0.23 and 8 wt.(%) were introduced to the 30 hours milled powders with the BPRs of 5:1 and 20:1, respectively.Nanostructure powders with the mean grain size less than 25 nm were obtained in all BPRs that confirmed by TEM image.
MA experiments were performed in a planetary ball mill at approximately room temperature and cup of speed 750 rounds per minute (RPM).The four cup planetary ball mill of Retch Company was used for MA experiments.Pure Merck Mo (99.7 wt.(%), 50 µm), Si (99.8 wt.(%), 25 µm), Ti (99.0 wt.(%), 30 µm) and graphite (99.3 wt.(%), 10 µm) were mixed to give the desired MoSi 2 -30 wt.(%) TiC composition.Four BPRs of 5:1, 10:1, 15:1 and 20:1 were used.A distribution of 20, 15 and 10 mm stainless steel balls was used in MA experiments.In order to prevent excess agglomeration 1 wt.(%) Stearic acid was added.The powders and balls were charged into a stainless steel cup (250 mL) in an Ar atmosphere.Samples were removed for analysis in a glove box under an Ar atmosphere by interrupting milling at various intervals.

Figure 1 .
Figure 1.XRD patterns of the 10 hours milled powders with different BPRs.

Figure 2 .
Figure 2. Effect of BPR on the formation of MoSi 2 -TiC nanocomposite powder after a) 3.5 hours and b) 7 hours of milling time.

Figure 3 .
Figure 3.Effect of BPR on the formation of MoSi 2 -TiC nanocomposite powder after a) 20 hours and b) 30 hours of milling time.

Figure 6 .
Figure 6.Mean grain size and microstrain of milled powders for all BPRs that were calculated on the basis of Rietveld refinement method.

Figure 7 .
Figure 7. Bright field TEM image of 30 hours milled powder With the BPR of 10:1.

Table 1 .
22ss absorption coefficients of starting materials and products22.

Table 2 .
effect of BPR on the beginning and completing time of MoSi 2 -TiC formation.