MATERIALS TRANSACTIONS
Online ISSN : 1347-5320
Print ISSN : 1345-9678
ISSN-L : 1345-9678
Quantitative Evaluation of Microstructure in Mo-Si-B-TiC Alloy Produced by Melting and Tilt Casting Methods
Sojiro UemuraTakateru YamamuroJoung Wook KimYasuhiro MorizonoSadahiro TsurekawaKyosuke Yoshimi
Author information
JOURNALS FREE ACCESS FULL-TEXT HTML

2018 Volume 59 Issue 1 Pages 136-145

Details
Abstract

Mo-Si-B-TiC alloys are expected to be a candidate for an ultrahigh-temperature material beyond Ni-base superalloys. This work quantitatively investigated the microstructure of a Mo-Si-B-TiC alloy with the composition of Mo-5Si-10B-10TiC (65Mo alloy) (at%) produced via arc-melting and tilt-casting techniques. The alloy was composed of four constituent phases: Mo solid solution (Moss), Mo5SiB2(T2), (Ti, Mo)Cx, (Mo, Ti)2C, and their eutectic (or peritecteutectic) phases. The compositions of the constituent phases were determined by electron beam micro analyzer (EPMA). Scanning electron microscopy – backscattered electron diffraction (SEM-EBSD) measurements revealed that T2 and (Ti, Mo)Cx phases have orientation relationships with Mo phase: $\{1 \bar 10 \}_{\rm Mo}//(001)_{\rm T2}$, ${<}111{>}_{\rm Mo}//{<}001{>}_{\rm T2}$ and $\{1 \bar 10 \}_{\rm Mo}// \{1 \bar 11 \}_{\rm (Ti,Mo)Cx}$, ${<}111{>}_{\rm Mo}//{<}001{>}_{\rm (Ti,Mo)Cx}$. Furthermore, the three-dimensional SEM observation with the combination of the focused ion beam (FIB) serial sectioning technique demonstrated that the T2 phase had a thin plate shape with the orientation of (001) as plate surfaces and of {100} as side ones.

 

This Paper was Originally Published in Japanese in J. Japan Inst. Met. Mater. 80 (2016) 529–538.

1. Introduction

Modern society is facing worldwide severe environmental problems like a greenhouse effect. Thus, there are strong demands to decrease greenhouse gas emissions. In order to solve the problems, it is essential to enhance the energy efficiency of heat systems such as a jet engine. Ni-based superalloys are mainly used for the turbine blade of the plane. A higher energy efficiency can be achieved by increasing the turbine inlet temperature. Because the melting points of Ni-based superalloys are lower than the turbine entrance temperature at takeoff, a cooling system is indispensable. However, the cooling system limits the ability to improve operational efficiency, therefore the development of new ultrahigh-temperature materials is increasingly expected1).

Molybdenum-silicon-boron based alloys are some of the most promising materials. These alloys are composed of molybdenum solid solution (Moss), Mo3Si, Mo5Si3 (T1), Mo2B, MoB, Mo5SiB2 (T2)2,3), and have high melting point and excellent high-temperature strength. However, fracture toughness at room temperature of the alloys are low at approximately 10 MPa(m)1/2 and the density are high. To increase the fracture toughness at least up to a practical level of 15 MPa(m)1/2, the volume fraction of the Moss phase should be increased4). However, an increase in the volume fraction of the Moss phase leads to an increase in the alloy density and a decrease in high-temperature strength due to decreasing the volume fractions of reinforcing Mo3Si and T2 phases.

One of the authors (K. Yoshimi) recently developed TiC added Mo-Si-B alloys59). The alloys have excellent high-temperature strength and high fracture toughness of approximately 15 MPa(m)1/2 at room temperature9). In addition, the alloys have a lower density (≦ 9.0 Mg/m3) as same as Ni-based superalloys (8.6–9.2 Mg/m3). Therefore, the Mo-Si-B-TiC alloys have a great potential as an ultrahigh-temperature material. It is well known that mechanical properties of materials are significantly affected by their microstructures. Thus, in order to maximize the potential of Mo-S-B-TiC alloys, design and control of microstructure of the alloys is essential with understanding inherent properties of constituent phases. Yoshimi et al.6), who conducted chemical analysis using transmission electron microscope - energy dispersive X-ray spectroscopy (TEM-EDS) for two different Mo-Si-B-TiC alloys, found that Si and Ti were solid-solved in the Mo phase, and TiC and Mo formed a solid solution with the NaCl structure as (Ti, Mo)C, for example. However, they did not mention the chemical compositions of carbon and boron in the constituent phases because of the lower accuracy of quantitative analysis in TEM-EDS. Nevertheless, carbon and boron often exert a significant influence on the mechanical properties of materials. For example, TiC has nonstoichiometry over a wide-range of C/Ti atom ratio, which strongly affect the mechanical properties10,11). Moreover, boron also makes an influence on the mechanical property of TiC12). Therefore, quantitative analysis of carbon and boron is of great importance to understand properties of constituent phases in the Mo-Si-B-TiC alloy. In addition to chemical analysis, quantitative evaluation of microstructural feature, such as the configuration and orientation distribution (texture) of constituent phases and the orientation relationship between the phases, will give important clues for design and control of microstructure to achieve enhanced mechanical properties of the Mo-Si-B- TiC alloy.

One of main motivation of this work is to determine chemical compositions of each constituent phase in a Mo-Si-B-TiC alloy more accurately using EPMA. In the present paper, an investigation was also made as to microstructural feature of Mo-Si-B-TiC alloy in connection with crystallographic information obtained from SEM-EBSD. Furthermore, three-dimensional analysis was performed to confirm the morphology of the constituent phases using SEM combined with the FIB serial sectioning. The results obtained from these analyses would open the possibility for modeling to be used as a tool to design microstructures for achieving enhanced properties of Mo-Si-B-TiC alloys.

2. Experimental Procedure

Two ingots of the TiC-added Mo-Si-B alloys, which had nominal compositions of 65Mo-5Si-10B-10TiC (at%), were prepared via the arc-melting and tilt-casting techniques13) in an Ar atmosphere. These alloys are referred to as 65Mo alloy throughout this article. Pure Mo rod (99.99 mass%), Si (99.9999 mass%), B (99.95 mass%), and cold-pressed TiC powder (99.95 mass%) were used as starting materials. TiC powder was compacted by applying a load of 20 kN using a hand press machine. After weighing these materials, rod-shaped ingots 100 mm long and 10 mm in diameter were produced by the arc-melting and tilt-casting technique in an Ar atmosphere (Fig. 1). Thereafter, one of the ingots were subjected to heat treatment at 2073 K for 24 h in an Ar atmosphere. For the heat treatment, an ultrahigh-temperature atmosphere controlled heat treatment furnace (Thermonic, THT-M03) was used. The interior of the furnace was evacuated to 1.0 × 10−3 Pa or less beforehand and set to Ar gas atmosphere (≥99.998%, 1 atm). Thereafter, the furnace was heated at 10 K/min up to 473 K and then at 16 K/min up to 2073 K with Ar gas flow at 1 L/min. After that, the furnace was kept at 2073 K for 24 h and subsequently cooled to room temperature. The as-cast and the heat-treated 65 Mo alloy ingots were cut into four equal parts in the longitudinal direction by wire electric discharge machining (EDM) to a size of about 10 mm as shown in Fig. 1(a)–(b), and then each cylindrical part was cut into semicylindrical shape (Fig. 1(c)), and lastly plate-shaped specimens with about 1.5 mm thick were cut in the vertical direction and the longitudinal direction (Fig. 1(d)) for evaluation of the microstructure. The cross sections in the vertical direction are referred to as C1 to C5 from the initial solidified part to the final solidified part, and cross sections in the longitudinal direction are referred to as L1 to L4 throughout this article. These samples were attached to a polishing jig and subjected to dry polishing in the order of #3M, #1M, #0, #2/0, #4/0, #6/0 using a precision parallel polishing machine. Furthermore, buffing was carried out using a diamond slurry having a particle diameter of 3 µm and 1 µm to obtain a mirror surface. Microstructures were observed by scanning electron microscopy (SEM; JEOL JEM-6510A) via back-scattered electron (BSE) images at an acceleration voltage of 20 kV and a working distance of 9–11 mm. The electron beam micro analyzer (Shimadzu EPMA-1720) was used to quantitatively evaluate the chemical composition of molybdenum solid solution (Moss), TiC, Mo5SiB2 (T2), Mo2C phase in the as-cast and the heat-treated 65Mo alloys. From the results of EPMA analysis, as described in section 3.3, it was revealed that the TiC phase made a solid solution with molybdenum and had nonstoichiometric composition, and the Mo2C phase did a solid solution with titanium. Therefore, the TiC and Mo2C phases are referred to as (Ti, Mo)Cx and (Mo, Ti)2C, respectively. Measurement was carried out under the conditions of an accelerating voltage of 15 kV and a beam current of 50 nA. As the spectroscopic crystal, PET was used for the Lα line of molybdenum and Kα line of silicon, and LiF was used for the Kα line of Ti, and LSA120 was used for the Kα lines of boron and carbon. Pure molybdenum, pure silicon, pure boron, and TiC were used as standard samples. However, since the B-Kα line used for the quantitative evaluation overlaps with the Mo-Mζ line, the interference correction method was applied as per the following procedure to improve the composition analysis accuracy for each constituent phase in the 65Mo alloy. First, the intensity ratio of Mo-Mζ (B-Kα) at the B-Kα peak position to Mo-Lα were measured from a molybdenum standard sample and the intensity ratio R = IMo-Mζ (B-Kα)/IMo-Lα was determined. Next, the intensity of Mo-Mζ (B-Kα) was estimated by multiplying Mo-Lα obtained from each constituent phase of the 65Mo alloy by the intensity ratio R, and then the corrected B-Kα line intensity was obtained by subtracting the Mo-Mζ (B-Kα) intensity mentioned above from the B-Kα line intensity obtained from the constituent phases in the 65Mo alloy. For the measurement of B-Kα peak wavelength, B standard sample was used. The EPMA measurements were performed at one point in five different grains in the respective constituent phases, and the compositions obtained at three points, where the total quantity of each element measured was close to unity, among five measurements was averaged to be the most probable composition of each phase. In addition, as the (Mo, Ti)2C phase and eutectic phases in the as-cast 65Mo alloy were finer than the spatial resolution of EPMA, it was hard to determine the composition of these phases. The crystal orientation relationship among the constituent phases in the 65Mo alloy were examined using scanning electron microscopy – backscattered electron diffraction (SEM-EBSD; HITACHI S-4200). The samples suitable for the SEM-EBSD measurement were mechanically polished and then buff-finished into the mirror surface using a vibration type automatic polishing machine (BUEHLER Vibromet 2) with a mixture of colloidal silica (OP-S) having a particle size of 0.04 µm and ethanol at 50:50 in volume ratio for 3 hours. SEM-EBSD observation was carried out using an EBSD-OIM system (TSL) attached to a field emission scanning electron microscope (HITACHI S-4200). Microstructures in the 65Mo alloy were multiphase analyzed with four phases of Moss, (Ti, Mo)Cx, Mo5SiB2 (T2), (Mo, Ti)2C phases and indexed. An area of 25 µm × 25 µm was scanned under conditions of an acceleration voltage of 20 kV, a beam current of 10 µA, a working distance of 10 mm, a direct observation magnification of 3000 times, and a step size of 0.1 µm. Moreover, to observe the three-dimensional form of the constituent phases, three-dimensional SEM examination was performed in combination with a serial sectioning using the focused ion beam (FIB; FEI Quanta 200 3D). Serially-sectioned images obtained by repeat sectioning were position-corrected on the computer and reconstructed as a three-dimensional image using the Image J software developed by National Institutes of Health (NIH). Serial sectioning with an ion beam was performed with an acceleration voltage of 30 kV and a beam current of 3 nA, and the SEM observation was carried out at an acceleration voltage of 15 kV, a beam current of 2 nA, a working distance of 15 mm and a direct magnification of 2000 times. The tilt angle between the Ga+ ion beam for FIB and the electron beam for SEM was 52°. The serially sectioned volume was 50 µm × 50 µm in area and 25 µm in depth, and each sectioning interval was 1 µm.

Fig. 1

Arc-melted and solidified Mo-Si-B-TiC alloy: (a) whole view showing the designation of the samples for microstructure observations, (b)–(d) illustration for the preparation of the samples cut from the ingot.

3. Results and Discussion

3.1 Microstructure observation

3.1.1 As-cast 65Mo alloy

Figure 2 shows back-scattered electron (BSE) images of a vertical cross section of the as-cast 65Mo alloy. The left-hand and right-hand sides of the micrograph correspond to the directions of the casting wall and the center of a cylindrical ingot, respectively. Looking through from the bottom of the rod, which is the initial solidified part, to the top part of the rod, which is the final solidified part (Fig. 2(a)–(d)), we can see that the dark phase was almost homogeneously dispersed over the entire area of the rod. This dark phase was found to be (Ti, Mo)Cx primary phase from the EPMA analysis described below. However, it seems that the amount of the (Ti, Mo)Cx phase was small near the areas of the casting wall and the center of the rod. These areas were found to be dominated by the Moss + T2 + (Ti, Mo)Cx and Moss + T2 + (Mo, Ti)2C three-phase eutectic structure. In the 65Mo alloy composed of five elements, the liquid phase → Moss + T2 + (Ti, Mo)Cx, liquid phase → Moss + T2 + (Mo, Ti)2C reactions are not invariant reactions. Although these reactions could be peritecto-eutectic reactions, since the reaction route was not clear at the present moment, for convenience it is expressed as three-phase eutectic.

Fig. 2

BSE images showing the microstructures in the as-cast 65Mo alloy: (a) C1, (b) C3, (c) C4, (d) C5.

The microstructure of this alloy can be roughly divided into the following two kinds of regions.

Region A: (Ti, Mo)Cx primary phase is homogeneously dispersed

Region B: Moss + T2 + (Ti, Mo)Cx and Moss + T2 + (Mo, Ti)2C three-phase eutectic structures dominate.

The region B observed near the casting wall surface was the largest in the initial solidified part and tended to decrease as the region moved to the top of the rod. On the other hand, the cross-sections C4 and C5 (Fig. 2 (c), (d)), which are close to the final solidified part, largely possess homogeneous structures defined as the region A.

Figure 3(a) shows a magnified BSE image of region A in C3. The EPMA analysis, to be described later, identified that the areas observed in a bright- and a dark- contrast corresponded to molybdenum solid solution (Moss) and titanium carbide containing molybdenum (Ti, Mo)Cx, respectively, and that the light- and dark- gray areas corresponded to Mo5SiB2 containing titanium and carbon, and (Mo, Ti)2C, respectively. From these results, the microstructure of as-cast 65Mo alloy was found to be composed of Moss, T2, (Ti, Mo)Cx, (Mo, Ti)2C and their eutectics like Moss + (Ti, Mo)Cx, Moss + T2 + (Ti, Mo)Cx and Moss + T2 + (Mo, Ti)2C. Figure 3(b) is a magnified image of the area surrounded by a square in Fig. 3(a). It is found that the the Moss + T2 + (Ti, Mo)Cx eutectic forms a colony structure surrounded by the boundaries consisting of the Moss + T2 + (Mo, Ti)2C eutectic phase.

Fig. 3

BSE images showing the microstructures in the as-cast 65Mo alloy: (a) C3, (b) the enlargement of the area shown in Fig. 3 (a), (c) the eutectic-phases dominated area (designated as area B) in section C3, (d) the area near the center in section L3.

According to the microstructure observed, it is most likely that the Moss + T2 + (Mo, Ti)2C three-phase eutectic reaction is the final coagulation reaction. Therefore, the colony structure would be formed in the following solidification route. Moss + T2 + (Ti, Mo)Cx three-phase eutectic (or peritecto-eutectic) reaction occurs first in the stage closest to the final stage of solidification. In a quinary system, this reaction is not an invariant reaction but a bivariant reaction, so that the liquid phase can vary its composition and temperature while coexisting with the Moss + T2 + (Ti, Mo)Cx three-phase eutectic during the solidification process. In that process, titanium and carbon atoms are likely evacuated from the liquid phase, which results in change in the reaction into “liquid phase → Moss + T2 + (Mo, Ti)2C”. Therefore, the colony structure is formed as the Moss + T2 + (Ti, Mo)Cx three-phase eutectic is surrounded by the Moss + T2 + (Mo, Ti)2C three-phase eutectic.

Figure 3(c) shows a magnified image of region B in C3. Region B is composed of Moss, (Ti, Mo)Cx, T2, Moss + T2 + (Mo, Ti)2C, Moss + T2 + (Ti, Mo)Cx, and Moss + (Ti, Mo)Cx. The microstructure of region B is mostly occupied by the colony structure as same as that mentioned above, and the (Ti, Mo)Cx in dendritic form is slightly distributed. As shown in Fig. 2, three-phase eutectics probably having a low melting point like Moss + T2 + (Mo, Ti)2C and Moss + T2 + (Ti, Mo)Cx existed unexpectedly in the area near the cast wall surface as well as in the area in the center of the ingot. This could be due to the fact that the liquid phase was rapidly cooled below the three-phase eutectic temperature without crystallization of the primary (Ti, Mo)Cx during tilt-casting. Figure 3(d) shows a BSE image of L3. The bottom side and the top side of the image corresponds to the directions of the initial and the final solidified portion, respectively, and the left-hand side and right-hand side of the image correspond to the directions of the casting wall and the center of the ingot, respectively. It is found that the T2 phase is prone to grow in an anisotropic manner along two distinct directions: one is from the surface of the casting wall toward the center direction and the other is from the initial solidified portion along the direction of the final solidified portion.

3.1.2 65Mo alloy after heat treatment

Figure 4 shows BSE images of 65Mo alloy after heat treatment at 2073 K for 24 h. In C4 (Fig. 4(a)), the left-hand side corresponds to the direction of the surface of the casting wall and the right-hand side corresponds to the direction of the ingot center. Microstructure of the 65Mo alloy after heat treatment is composed of Moss, (Ti, Mo)Cx, T2, (Mo, Ti)2C and Moss + (Ti, Mo)Cx two-phase eutectic. Each constituent phase is coarsened and spheroidized by the heat treatment, and the colony structure observed in the as-cast 65Mo alloy can no longer be observed. The Moss + T2 + (Mo, Ti)2C eutectic, which had formed the colony boundary before the heat treatment, changed to a microstructure in which the (Mo, Ti)2C and (Ti, Mo)Cx is finely dispersed in the Moss and T2 matrix as shown by ① in Fig. 4(a). The Moss + T2 + (Ti, Mo)Cx eutectic, which was surrounded by the Moss + T2 + (Mo, Ti)2C eutectic, changed to a microstructure composed of Moss + (Ti, Mo)Cx eutectic and spheroidized T2 as shown by ② in Fig. 4(a). In addition, it is found that (Ti, Mo)Cx is precipitated in the Moss (Fig. 4(b)), and that the Moss, (Ti, Mo)Cx, and (Mo, Ti)2C are precipitated in T2. In particular, (Ti, Mo)Cx in the T2 existed in two distinct manners: one is independently present in the form of platelet and the other in the eutectoid-like with the Moss as shown in Fig. 4(c). Furthermore, careful observation of (Mo, Ti)2C revealed that some of (Mo, Ti)2C transformed into Moss and (Ti, Mo)Cx in the form of lamella structure during heat treatment at 2073 K for 24 h as shown in Fig. 4(d). Such phase transformation of (Mo, Ti)2C into Moss and (Ti, Mo)Cx was frequently observed where (Mo, Ti)2C was present in contact with the T2 phase. According to the Mo-Ti-C ternary isothermal phase diagram at 2023 K14), which is close to the heat treatment temperature of 2073 K, there is a Moss + TiC + Mo2C three-phase region at the Ti rich side of the Mo2C single phase region, and Moss + TiC two phase region is also present at further Ti rich side. As described below, the composition of (Mo, Ti)2C phase in the 65Mo alloy heat-treated is Mo 51.7 at%, Ti 14.0 at%, C 34.4 at%, which is just at Moss + TiC + Mo2C three phase region in the Mo-Ti-C ternary isothermal phase diagram at 2023 K14). Accordingly, it is considered that the phase transformation of (Mo, Ti)2C to the Moss and (Ti, Mo)Cx may occur due to diffusion of Ti that is supersaturated in the T2 through contact with (Mo, Ti)2C. Yoshimi et al. reported6) that the deformation-induced transformation of (Mo, Ti)2C to Moss and (Ti, Mo)Cx could occur during high-temperature compressive test in the alloy of Mo-5Si-10B-7.5TiC(at%), which is slightly different in TiC composition from that of the 65Mo alloy. However, in the previous report5) using alloys of the same composition as the 65Mo alloy, phase transformation of (Mo, Ti)2C was not found in heat treatment at 2073 K for 24 h.

Fig. 4

BSE images showing the 65Mo alloy after heat treatment at 2073 K for 24 h: (a) C3, (b)–(d) showing precipitations in (b) Moss, (c) T2 and (d) (Mo, Ti)2C.

3.2 Area fraction

The area fractions of the constituent phases in the region A of the sections C1 to C5 in as-cast and heat-treated 65Mo alloys were measured from the BSE images (Fig. 5). The area fraction data of T2 and (Mo, Ti)2C are combined in this figure because it was difficult to separate these two phases in the SEM-BSE images owing to their weak contrast difference. In the as-cast 65Mo alloy, the area fraction of each constituent phase shows a nearly uniform irrespective of the position along the direction of solidification, and the average value of all cross sections was approximately 16.3% for Moss, 11.4% for T2, 10.2% for (Ti, Mo)Cx, 20.4% for Moss + T2 + (Ti, Mo)Cx eutectic, 38.1% for Moss + T2 + (Mo, Ti)2C (Fig. 5(a)). On the other hand, in the heat-treated 65Mo alloy, the average of all cross sections was approximately 45% for Moss, 33.9% for T2 + (Mo, Ti)2C, 22.1% for (Ti, Mo)Cx. Therefore, it was confirmed that the microstructure in region A, which is particular microstructure in the 65Mo alloy, was substantially uniform along the casting direction before and after heat treatment.

Fig. 5

Area fraction of the individual phases measured in sections C1–C5: (a) as-cast, (b) post heat-treatment at 2073 K for 24 h.

3.3 EPMA analysis of constituent phases

In the previous report6), the composition of each constituent phase of Mo-Si-B-TiC alloys were analyzed. However, the analysis for light elements like boron and carbon was not conducted. Therefore, quantitative analysis of the constituent phases, including the light elements, was performed using EPMA. Table 1 shows the chemical compositions of the constituent phases in the as-cast and heat-treated 65Mo alloy. Since the difference between the compositions of individual cross sections of the ingot was small within the experimental error, the average values are shown in the Table 1. The composition of constituent phases in the as-cast 65Mo alloy is as follows. The Moss had a 1.7 at% for Si, 4.8 at% for Ti, and 1.4 at% for C. The (Ti, Mo)Cx had a 32.7 at% for Ti, 43.4 at% for C, and 23.8 at% for Mo. The atomic ratio of Ti to Mo is approximately 11 : 8 and that of (Ti + Mo) to C is approximately 1 : 0.77. Accordingly, TiC phase observed in the 65Mo alloy was confirmed to be a nonstoichiometric compound with NaCl structure, in which some of Ti sites was replaced by Mo, being described as (Ti, Mo0.73)C0.77. The T2 phase had a 56.7 at% for Mo, 10.8 at% for Si, 21.7 at% for B, 9.2 at% for Ti, and 1.6 at% for C. Assuming that Ti substitute for the Mo position, the concentrations of Si and B could be shifted to a lower concentration side than the stoichiometric ones of T2.

Table 1 Chemical compositions of the constituent phases (at%).
Element As-cast   Heat-treatment
Phase Mo Si B Ti C   Mo Si B Ti C
Moss 92.1 +0.7 1.7 +0.0 4.8 +0.1 1.4 +0.7   93.7 +1.0 1.9 +0.1 2.5 +0.3 1.9 +1.7
−0.5 −0.0 −0.2 −0.7   −1.5 −0.1 −0.1 −0.8
(Ti, Mo)Cx 23.8 +1.0 32.7 +0.7 43.4 +0.8   23.5 +1.3 31.0 +0.5 45.5 +0.9
−1.0 −0.5 −0.6   −1.3 −1.1 −1.7
Mo5SiB2(T2) 56.7 +1.6 10.8 +0.2 21.7 +1.9 9.2 +0.4 1.6 +0.4   58.2 +1.8 12.3 +0.3 20.0 +0.9 7.4 +0.3 2.0 +0.8
−1.8 −0.5 −1.4 −0.5 −0.4   −0.8 −0.2 −1.5 −0.1 −0.9
(Mo, Ti)2C   51.7 +0.9 14.0 +1.5 34.4 +1.0
  −0.8 −0.4 −1.4

For the 65Mo alloy heat treated at 2073 K for 24 h, the Si, Ti and C dissolved in the Moss were 1.9 at%, 2.5 at%, 1.9 at% respectively. The (Ti, Mo)Cx was composed of Ti : 31.0 at%, Mo : 23.5 at% and C : 45.5 at%. The ratio of Ti to Mo is approximately 5 : 4, and the ratio of (Ti + Mo) to C is approximately 1 : 0.83, so that (Ti, Mo)Cx is expressed as (Ti, Mo0.77)C0.83. It is of interest that the atomic ratio of (Ti, Mo) to C in (Ti, Mo0.73)C0.77 and (Ti, Mo0.77)C0.83 is close to the carbon concentration at which TiC possesses the congruent melting point in the Ti–C binary system15). Furthermore, the composition of the T2 is Mo 58.2 at%, Si 12.3 at%, B 20.0 at%, Ti 7.4 at%, C 2.0 at%. The concentration of the Si and B were also shifted to the lower concentration side than the stoichiometric ones of T2. In addition, the composition of (Mo, Ti)2C, which was coarsened by heat treatment to a size that can be analyzed by EPMA, was Mo 51.7 at%, Ti 14.0 at%, C 34.4 at%. This composition is in the range of Mo2C + Mo + TiC phase coexisting region in the Mo–Ti–C ternary isothermal state diagram at 2023 K14).

3.4 SEM-EBSD observation

SEM-EBSD observations were carried out on as-cast and heat-treated 65Mo alloys. Figure 6 shows a phase map and IPF maps of the Moss, T2, (Ti, Mo)Cx, and (Mo, Ti)2C for the as-cast 65Mo alloy. In the phase map (Fig. 6(a)), the green, red, blue and yellow colors correspond to Moss, (Ti, Mo)Cx, (Mo, Ti)2C, T2 respectively. The IPF maps are displayed as the crystal orientation in the longitudinal direction of the ingot (the Z direction shown in the figure). It is found from Fig. 6(b) that the <111> orientation in the Moss phase was localized in the longitudinal direction of the ingot. It is interesting to see that Moss consisting of eutectics possessed the same orientation as the Moss single phase. Moreover, the IPF map revealed that there is a relatively large orientation dispersion in the Moss grain, suggesting that the Moss may accumulates some residual strain16). For (Ti, Mo)Cx and (Mo, Ti)2C, there are no definite sign of any texture formation (Fig. 6 (c), (d)). On the other hand, either <100> or [001] orientation in the T2 with the tetragonal structure is found to be oriented in the longitudinal direction of the ingot (Fig. 6(e), (f)). The elongated T2 appears to have <100> direction, but the plate-like T2 does the [001] direction. Figure 7 shows a phase map and IPF maps on the cross section (C plane) of the 65Mo alloy after heat treatment. As described above, each constituent phase is coarsened, but no significant changes were observed in the crystal orientation distribution.

Fig. 6

SEM-EBSD micrographs taken from section C4 (except Fig. 6(f)) in the as-cast 65 Mo alloy: (a) phase map, (b)–(f) IPF maps for (b) Moss, (c) (Ti, Mo)Cx, (d) (Mo, Ti)2C, (e) T2 and (f) T2 (C5 section). The IPF maps are displayed as the orientation parallel to the growing direction.

Fig. 7

SEM-EBSD micrographs taken from section C4 (except Fig. 7(e)) in the 65Mo alloy heat-treated at 2073 K for 24 h: (a) phase map, (b)–(f) IPF maps for (b) Moss, (c) (Ti, Mo)Cx, (d) (Mo, Ti)2C, (e) T2 (C5 section) and (f) T2 (C4 section). The IPF maps are displayed as the orientation parallel to the longitudinal direction of the ingot.

From these results, it is most likely that the Moss has some orientation relationships with T2. Figure 8(a) shows the relationship between the $\{1 \bar 10 \}_{\rm Mo}$ and $(001)_{\rm T2}$, ${<}111{>}_{\rm Mo}$ and ${<}100{>}_{\rm T2}$ directions. Because Moss and the T2 possess considerably sharp texture, the orientations with the highest intensity were plotted in Fig. 8(a). The stereographic analyses revealed that Moss had such an orientation relation to T2 as $\{1 \bar 10 \}_{\rm Mo}//(001)_{\rm T2}$ and ${<}111{>}_{\rm Mo}//{<}100{>}_{\rm T2}$. Takata et al. also reported the same orientation relation between Mo and T2 in Mo-Nb-Si-B alloy17). Furthermore, stereographic analysis for the orientation relationship between the Moss and (Ti, Mo)Cx is shown in Fig. 8(b). Here, the pole of (Ti, Mo)Cx shown in Fig. 8 (b) is obtained from the crystal grain shown in ① in Fig. 6. Kurishita et al.18) reported that TiC precipitated at the grain boundary of Mo had an orientation relationship with the Mo matrix of $\{1 \bar 10 \}_{\rm Mo}// \{1 \bar 11 \}_{\rm TiC}$, ${<}111{>}_{\rm Mo}//{<}110{>}_{\rm TiC}$ and ${<}110{>}_{\rm Mo}//{<}110{>}_{\rm TiC}$. We also found the same orientation relation between Moss and (Ti, Mo)Cx as the previous report: $\{1 \bar 10 \}_{\rm Mo}// \{1 \bar 11 \}_{\rm (Ti,Mo)Cx}$, ${<}111{>}_{\rm Mo}//{<}110{>}_{\rm (Ti,Mo)Cx}$ (Kurdjumov-Sachs (K-S) relation). However, (Ti, Mo)Cx, did not necessarily possess the K-S relationship.

Fig. 8

Stereographic projections showing the orientation relationships (a) between Mo and T2 phases, and (b) between Mo and (Ti, Mo)Cx phases.

Figure 9 shows IPF maps of Moss and (Ti, Mo)Cx after heat treatment. The Moss surrounded by (Ti, Mo)Cx has a larger orientation dispersion in the crystal grain compared with the Moss shown in Fig. 7(a). This finding suggests that (Ti, Mo)Cx may exert an influence on the large orientation dispersion in Moss phase. The thermal expansion coefficient of TiC0.8 is about 60% larger than that of pure Mo19,20). Thus, it is considered that compression and tensile stresses would be applied to the Moss phase and the (Ti, Mo)Cx phase, respectively, during cooling from the heat treatment temperature to room temperature due to the restriction from the interface between them. The compressive residual stress can generally provide a beneficial effect for improving the fracture resistance. According to Kruzic et al.4), the Mo ductility plays an important role in determining the mechanical properties like fracture toughness in Mo-Si-B alloys. It is therefore considered that the compressive stress probably introduced in the Moss phase may cause the high fracture toughness of Mo-Si-B-TiC alloys compared to Mo-Si-B alloys4,9).

Fig. 9

IPF maps for (a) Moss and (b) (Ti, Mo)Cx including a coherent twin boundary in the 65Mo alloy heat treated at 2073 K for 24 h. The IPF maps are displayed as the orientation parallel to the longitudinal direction of the ingot. The Moss phase surrounded by (Ti, Mo)Cx is likely to have residual strain in grains interior to a certain extent.

Furthermore, it is surprising to see that the {111}∑3 coherent twin boundary was frequently observed in (Ti, Mo)Cx not only in the heat-treated 65Mo alloy but also as-cast 65Mo alloy, as shown in Fig. 9 (b). Because TiCx has a very high stacking fault energy of 130~300 mJ/m2 21), to authors' knowledge, the formation of twin boundary in TiCx or (Ti, Mo)Cx has not been reported. Since elements other than Ti, Mo and C were not detected in EPMA, the reason for the formation of the twin boundary in the (Ti, Mo)Cx in the 65Mo alloy is unclear at present moment.

3.5 Three-dimensional microstructure observation

It is important to understand precise form of the constituent phases and their arrangement in microstructure, so that three-dimensional analysis was carried out using SEM observations combined with the FIB serial sectioning. For the three-dimensional analysis, the cross-section of L1 in as-cast 65Mo alloy was used. Figure 10 presents three-dimensional images, in which (Ti, Mo)Cx is displayed in red and T2 in gray. Figure 10 (b) and (c) show the three-dimensional form of (Ti, Mo)Cx, and T2 respectively. Figure 10 (b) reveals that the (Ti, Mo)Cx has dendrite-like form with stretching branches in all directions around the spherical core. The lengths of the branches were different, and small branches extended from the nucleus to fill in between the long ones. As shown in Fig. 10 (c), we can see that the T2 phase has a thin plate form with a thickness of approximately 3 µm, rather than rod-like shape. From the results of the SEM-EBSD analysis, the T2 phase probably had a plate shape of orientation (001) for the plate surface and {100} for side ones, because the surface orientations of the elongated (rod-like) T2 and the plate-like T2 observed in EBSD-IPF maps tends to localized near {100}, and (001), respectively. This finding suggests that the surface energy of T2 phase has high anisotropy. Furthermore, as shown in Fig. 10 (a), the growth of the T2 phase was arrested by the (Ti, Mo)Cx phase. Therefore, as reported in the previous work5), (Ti, Mo)Cx should be primarily crystal rather than T2 for the 65Mo alloy.

Fig. 10

Three-dimensional phase images reconstructed from serially-sectioned images: (a) (Ti, Mo)Cx and T2 (Mo5SiB2), (b) (Ti, Mo)Cx, (c) T2.

4. Conclusions

In this study, the microstructure of the 65Mo-5Si-10B-10TiC (at%) alloy produced via arc-melting and tilt-casting techniques was quantitatively evaluated. The main results obtained are as follows.

(1) As-cast 65Mo alloy was composed of Moss, (Ti, Mo)Cx, T2 (Mo5SiB2), and their eutectic phases such as Moss + (Ti, Mo)Cx, Moss + T2 + (Ti, Mo)Cx, and Moss + T2 + (Mo, Ti)2C. The three-phase eutectic formed a characteristic colony structure such that the Moss + T2 + (Ti, Mo)Cx eutectic was surrounded by boundaries consisting of the Moss + T2 + (Mo, Ti)2C eutectic phase.

(2) There was no change in the phases constituting the microstructure after heat treatment at 2073 K for 24 h. However, the colony structure disappeared due to coarsening and spheronization of the constituent phases by heat treatment. In addition, it was found that some (Mo, Ti)2C transformed into Moss and (Ti, Mo)Cx, in the form of lamella structure.

(3) The area fraction of constituent phases in both the as-cast and heat-treated 65Mo alloy along the casting direction was almost uniform regardless of the location of the ingot.

(4) EPMA measurements determined the compositions of individual constituent phases as follows.

As-cast 65Mo alloy

  • ●   Moss: 92 at% Mo – 2 at% Si – 5 at% Ti – 1 at% C
  • ●   (Ti, Mo)Cx: 33 at% Ti – 24 at% Mo – 43 at% C
  • ●   T2(Mo5SiB2): 57 at% Mo – 11 at% Si – 22 at% B – 9 at% Ti – 2 at% C

65Mo alloy after heat treatment at 2073 K for 24 h

  • ●   Moss: 94 at% Mo – 2 at% Si – 3 at% Ti – 2 at% C
  • ●   (Ti, Mo)Cx: 31 at% Ti – 24 at% Mo – 45 at% C
  • ●   T2(Mo5SiB2): 58 at%Mo – 12 at% Si – 20 at% B – 7 at% Ti – 2 at% C
  • ●   (Mo, Ti)2C: 52 at% Mo – 14 at% Ti – 34 at% C

(5) SEM/EBSD analysis revealed that the <111> orientation of Moss phase was localized in the longitudinal direction of the rod-shaped ingot. In addition, the Moss phase consisting of eutectics also possessed the same orientation as the Moss single phase. As for T2 phase, the <100> orientation was likely to be localized in the longitudinal direction of the ingot in the area of the center of the rod, while [001] direction was likely to be oriented in the longitudinal direction of the ingot in area near the casting wall. The three-dimensional analysis via SEM observations combined with the FIB serial sectioning revealed that T2 phase possessed a thin plate shape with the thickness of approximately 3 µm. Taking into account of the SEM-EBSD observation, it was likely that the T2 phase had a thin plate shape of orientation (001) for the plate surface and {100} for side ones.

(6) SEM-EBSD measurements revealed that the Moss phase possessed such orientation relations to T2 phase as $\{1 \bar 10 \}_{\rm Mo}//(001)_{\rm T2}$, ${<}111{>}_{\rm Mo}//{<}100{>}_{\rm T2}$, and to (Ti, Mo)Cx as $\{1 \bar 10 \}_{\rm Mo}// \{1 \bar 11 \}_{\rm (Ti,Mo)Cx}$, ${<}111{>}_{\rm Mo}//{<}110{>}_{\rm (Ti,Mo)Cx}$.

Acknowledgement

The authors wish to express their sincere thanks to Prof. J. Otani and Mr. T. Sato for their help with three-dimensional analysis, and to Assoc. Prof. Y. Kimura for useful discussion. This work was supported by the Advanced Low Carbon Technology Research and Development Program (ALCA) of the Japan Science and Technology Agency (JST), and by Japan Society for the Promotion of Science (JSPS) KAKENHI Grant Number 16H06366.

REFERENCES
 
© 2017 The Japan Institute of Metals and Materials
feedback
Top