Experimental Investigation of the Phase Relations in the Fe-Zr-Y Ternary System

The phase relations of the Fe-Zr-Y system at 973 K and 1073 K were experimentally investigated by using the equilibrated alloys. New ternary compounds τ3-Fe3ZrY and τ4-Fe10Zr5Y2 were found in this ternary system. The solubility of Y in Fe2Zr was measured to be 3.5 at.% and the third component can hardly dissolve in the other binary intermetallic phases. Experiments have verified that Fe2.9Zr0.5Y0.5 has a solid solubility ranging from Fe73Zr12Y14 to Fe77Zr9Y13.


Introduction
Due to the use of nuclear fusion and the third-generation nuclear fission reactors, zirconium-based alloys have been widely studied as important nuclear cladding materials [1][2][3][4][5]. This relies on their specific characteristics: excellent corrosion resistance, good mechanical properties, high resistance to radiation damage and a low cross section of capture for thermal neutrons [6]. The Fe-Zr-Y system is an important member of the zirconium-based nuclear cladding materials, the high temperature resistance and radiation resistance of which can be greatly improved after oxidation [7][8][9][10][11]. Although the Fe-Zr-Y system has excellent prospects, there are still some problems that limit its applications in industry. For example, when the temperature of a nuclear reactor core continuously rises, the Fe-Zr-Y system may be at risk of being unstable [12].
An efficient and effective solution to this problem is to obtain a comprehensive and profound understanding on phases of different components and rule out the phases that are unstable at high temperatures in advance. which is of vital importance for the applications of the Fe-Zr-Y system [13]. The purpose of this work is to explore the phase diagram of the Fe-Zr-Y system at high temperature in order to intuitively express the relationship between phases under the thermodynamic equilibrium state so as to provide a basic theoretical guide for the research, development, and design of new materials.
The experimental investigations on the phase relations of the Fe-Zr-Y system [14][15][16][17][18][19][20] have been carried out, and several assessments of thermodynamic data have been obtained [21][22][23][24][25][26]. Although there are some controversies about the stability of hex-Fe 2 Zr and Fe 23 Zr 6 , it is widely accepted that both of the two phases exist [3]. According to the results of X-ray and magnetic measurements, Kai et al. [27] confirmed the presence of hexagonal Fe 2 Zr, which had the same structure of MgNi 2 . Subsequently, Stein et al. [28] found the Fe 2 Zr phase of C36-type structure through heat treatment at 1563 K. Meanwhile, Liu et al. [16] confirmed the existence of Fe 23 Zr 6 by using a transmission electron microscope (TEM) and a scanning transmission electron microscope (STEM). Recently, Yang et al. [29] and Lu et al. [30] applied first-principles calculations to evaluate the formation enthalpy of compounds in Fe-Zr system, in which Fe 23 Zr 6 was considered as a stable phase. After a thorough survey and thermodynamic evaluations based on a number of experiments, the binary phase diagrams reported by Lu et al. [30] were finally adopted in this work. As shown in Figure 1, among the four intermetallic phases, only Fe 2 Zr phase has an obvious homogeneity range. In other phases, FeZr 2 appears in the temperature range of 1215-1054 K, and FeZr 3 is formed through peritectoid reaction. Domagala et al. [31] reported the Fe-Y phase diagram for the first time in the whole composition range. Then, a thermodynamic assessment was carried out by Gschneider et al. [32] Recently, Zhang et al. studied the crystallography and thermodynamics of the compounds in the system, which provided a basis for further optimization of the Fe-Y system [33]. Figure 2 shows the Fe-Y phase diagram with four intermetallic compounds: Fe 17 Y 2 , Fe 23 Y 6 , Fe 3 Y and Fe 2 Y. Based on the previous research work mentioned above, Saenko et al. [34] reported the thermodynamic optimization of the Fe-Y binary system, as shown in Figure 2. The phase equilibrium relation of Zr-Y phase diagram was first studied by Wang [35], based on which Palenzona and Ciraflci [36] optimized the thermodynamic data. Recently, Flandorfer et al. [37] constructed the thermodynamic database of Zr-Y system. The calculated phase diagram used in this work adopts the latest thermodynamic parameters reported by Bu et al. [38], as shown in Figure 3. Research on the phase relationships and ternary compounds of Fe-Zr-Y ternary systems was relatively limited. In 1986, by combining three related binary phase diagrams, Harchenko et al. [39] measured the 1070K isothermal section of the Fe-Zr-Y ternary system for the first time, but no ternary compounds were discovered, as shown in Figure 4. In 1989, Jifan et al. reported Fe 9 Zr 2 Y phase with ThMn 12 structure [40]. At the same time, Itoh et al. also found Fe 2.9 Zr 0.5 Y 0.5 phase of Orthorhombic type [41]. The binary and ternary phases, the crystal structure and lattice parameters that were previously reported are listed in Table 1.

Experimental Procedures
The phase relationship of the Fe-Zr-Y system was studied through the equilibrium alloy method of static measurements. Iron rod (99.99 wt.%), zirconium rod (99.99 wt.%) and yttrium block (99.99 wt.%) were selected as raw materials. The weight of the samples was controlled around 12 g, with an error of ±0.005 g. The samples were placed in a non-expendable arc melting furnace with high purity argon atmosphere. At the same time, a sponge titanium button was added as the oxygen absorber to prevent oxidation. Each alloy button was melted for at least 4 times to ensure its uniformity. The alloy buttons were divided into two parts and sealed in the quartz tubes filled with argon. According to the reported heat treatment processes of similar systems [3], 60 and 90 days were set as the annealing times for 700 • C and 800 • C, respectively. After annealing, the samples were quenched in cold water to preserve the microstructure at high temperature.
The morphology and phase compositions of the alloy were analyzed by electron probe microanalysis (EPMA, JAXA-8800R, JEOL, 15 kV, 1 × 10 −8 A, Tokyo, Japan). X-ray diffraction was employed (XRD, Rigaku d-max/2550 VB, Cu K, 40 kV, 250 mA, Tokyo, Japan) to analyze the crystal structure of typical alloys, with the scanning range of 10-90 • and a speed of 0.133 • /s. Backscattering electron (BSE) images of the alloy samples were acquired using a scanning electron microscope (SEM, TESCAN MIRA3 LMH, 15 kV, working distance of 15 mm, Brno, The Czech Republic).

Experimental Results
Based on the phase equilibrium data of 24 alloy samples, the isothermal section of the Fe-Zr-Y ternary system at 973 K was determined for the first time, as shown in Figure 5. A total of 12 three-phase regions and 12 two-phase regions were measured. In this isothermal section, there are four ternary compounds, among which, τ1, τ3 and τ4 have solid solubility intervals, and seven binary compounds, three of which have solid solubility, which are Fe 2 Zr, FeZr 3 , Fe 2 Y. The isothermal section of Fe-Zr-Y ternary system at 1073 K is similar to the system at 973 K except that the maximum solid solubility of compounds τ1 and τ3 are slightly higher at 1073 K. Additionally, FeZr 2 as a high temperature phase appeared at 1073 K. This is consistent with the binary optimized phase diagram, and therefore its appearance is reasonable and in accordance with expectation [30]. The 1073 K isothermal section obtained in this work is shown in Figure 6.

Discussion
The experimental data obtained from SEM, EPMA, XRD and EBSD examination were analyzed to determine the isothermal section and the phase relationship of the Fe-Zr-Y ternary system at 973 K and 1073 K. The phase relations in several key alloys were discussed in detail as follows.

Phase Equilibria at 973 K
A total of 24 alloy samples were prepared for the study of the phase equilibria of the Fe-Zr-Y ternary system at 973 K. The constituent phases of each alloy sample were summarized in Table 2. BSE images and XRD patterns of alloy samples A7, A13 and A24 are shown in Figure 7a,c,e). EPMA-WDS results show that the dark gray phases in A7, A13 and A24 have the same composition, close to Fe60Zr20Y20, so they are confirmed as the same phase. However, this composition is different from the existing ternary compounds in the Fe-Zr-Y system. For the further identification of this phase, XRD analysis was performed on all three samples. In the XRD patterns of the three alloys, there are unrecorded characteristic peaks, the positions of which are basically fixed. Only a few peaks have slight displacement, which may be due to the change in lattice parameters. According to the above analysis, the ternary compound was identified as a new phase with the chemical formula Fe 3 ZrY and named τ3. Since no single phase of τ3 was obtained in this work, follow-up experiments are needed to explore the crystal structure of τ3. There are three phases in the BSE images of alloy samples A2 and A9 (Figure 8a,c). According to the results of XRD and EPMA-WDS tests, all the phases are already identified except for the black phase. The compositions of the black phase in Figure 8a,c are Fe59.7Zr29.6Y10.5 and Fe56.2Zr27.9Y15.9, respectively. The stoichiometric ratios represent some differences between the two compositions, so the black phase was initially considered to be two different compounds. However, the XRD patterns (Figure 8b,d) show that the characteristic peaks of the two matched up, indicating that the crystal structures of the two phases were consistent, thus the two phases were confirmed to be the same compound. No ternary compound with similar stoichiometric ratio and diffraction peak matching was found in PDF cards. According to the results described above, the black phase was defined as the new ternary compound Fe 10 Zr 5 Y 2 and named as τ4. The BSE images of alloy samples A5, A10 and A17 are shown in Figure 9a,c,e. The black phase in A10 and the dark gray phase in A5 and A17 have the same atomic ratio of Fe 2.9 Zr 0.5 Y 0.5 , which was denoted as τ1 phase. The XRD (as shown in Figure 9b,d,f) characteristic peaks of compound τ1 have fixed positions, which are different from all the existing PDF cards. According to the reported literature [41], Fe 2.9 Zr 0.5 Y 0.5 was considered to be transformed from Fe 3 Y, as Zr atoms replaced Y atoms in Fe 3 Y. In the study of Itoh et al. [41], Fe 3 Y did not exist in the Fe-Zr-Y system, as it was considered to be present in the form of Fe 2.9 Zr x Y 1−x (0 ≤ x ≤ 1), meaning that Y would be replaced by Zr until Y disappears completely. This conclusion was not adopted in this work due to the following reasons: (1) Fe 2.9 Zr 0.5 Y 0.5 would be adjacent to Fe 3 Y phase according to the reported theory [41], which is not consistent with what is shown in Figure 9a,c,e. Although there is Fe 3 Y shown in Figure 9c, it is separated by Fe 23 Y 6 , and τ1 is completely wrapped by Fe 23 Y 6 instead of Fe 3 Y. (2) If Fe 2.9 Zr x Y 1−x (0 ≤ x ≤1) phase exists, the solubility range of Zr in Fe 3 Y would pass through most of the phase equilibrium region, which violates the law of phase equilibrium relation. In conclusion, the phase Fe 2.9 Zr x Y 1−x (0 ≤ x ≤ 1) is not adopted in this work.  Figure 10a illustrates the phase composition in alloy A4, which is composed of three phases. The results of EPMA-WDS show the continuously distributed matrix phase and the gray phase had the composition of Fe68.1Zr29.3Y2.6 and Fe88.2Y11.8, and were, respectively, determined as Fe 2 Zr and Fe 17 Y 2 according to the XRD results of Figure 10b. The black phase labeled τ2 was presumed to be an unknown ternary compound at first, because the diffraction peaks cannot be matched by any PDF card corresponding to the crystal structure of the stable solid phase in the ternary system. However, they are indexed by a ThMn 12 type crystal structure and obtain the characteristic peaks of Fe 9 Zr 2 Y, which coincide with τ2 in the literature [40]. Therefore, τ2 phase was finally identified as the ternary compound Fe 9 Zr 2 Y.  Figure 11 shows the BSE images and XRD patterns of A16 and A18. According to the data from EPMA-WDS, the maximum atomic percentage of Zr in Fe 2 Zr is 37.6 at 973 K in A16. Combining Figure 11c with Figure 5, it can be seen that Y tends to precipitate in the form of metal simple substance most of the time, which is in accord with the results obtained by Nouri et al. [38] Meanwhile, Y phase is dispersed, indented and porous, which can be mainly ascribed to the result of rapid galvanic corrosion of the (Y), which acts as anode with Continuous-distributed matrix as cathode during polishing (or exposure to the moisture) [38].

Phase Equilibria at 1073 K
The phase equilibrium of the Fe-Zr-Y ternary system at 1073 K was also investigated in this work. A total of 25 alloy samples were annealed at 1073 K for 60 days. The phases and compositions of each alloy are listed in Table 3. With the increase in temperature, the solid solubility region of τ1 and Fe 2 Zr becomes larger, and a new high temperature phase FeZr 2 appears. This result is consistent with the binary phase diagram of Fe-Zr, but the exact formation temperature of FeZr 2 had not been reported to date. The XRD pattern of B23 in Figure 12b shows the same characteristic peaks of FeZr 2 as standard PDF cards. Combined with EPMA-WDS data, it can be seen that the dark phase in BSE image Figure 12a is the binary compound FeZr 2 .

Conclusions
In this work, the phase relationship of the Fe-Zr-Y ternary system at 973 K and 1073 K combined with EPMA-EDS, XRD and SEM results was studied systematically. There are twelve and thirteen three-phase regions measured at 973 K and 1073 K, respectively. At 1073 K, the solution ranges of τ1-Fe 2.9 Zr 0.5 Y 0.5 are from Fe73Zr12Y14 to Fe77Zr9Y13. The maximum solid solubility of Y in Fe 2 Zr is 4 at.% and atomic ratio of Fe 2 Zr is from Fe62Zr38 to Fe71Zr29 at 1073 K. New ternary compounds τ3-Fe 3 ZrY and τ4-Fe 10 Zr 5 Y 2 were investigated. Those isothermal sections at 1073 K and 973 K provide the possibility of providing a thermodynamic description of the system through the CALPHAD method, allowing more reliable extrapolations of the Fe-Zr-Y system to be used in the nuclear industry.