*3.1. Microstructure and Texture of the Extruded ZK60 Alloys with and without RE Addition*

The SEM micrographs in Figure 1a–c revealed that the addition of Y and Ce-rich elements to the ZK60 alloy increased the volume fraction of precipitates from 2% to 14% and 15% in the ZK60–2Y and ZK60–2RE alloys, respectively (see also Table 1). Spherical and network-like precipitates were observed in both the Y- and RE-containing alloys; however, the ZK60–2RE alloy contained more network-like secondary phase particles than the ZK60– 2Y alloy did. Figure 1d,e shows color-coded EDS elemental maps for the ZK60–2Y and

ZK60–2RE alloys (Mg: orange, Zn: red, Y and Ce: turquoise). The EDS analysis suggests that the precipitates were enriched in Y and Zn in the ZK60–2Y alloy, and in Ce and Zn in the ZK60–2RE sample. Namely, the secondary phase particles indicated by the red arrows in Figure 1a–c had the following composition, as obtained by EDS (in at.%): 66.1% Mg and 33.9% Zn for the ZK60 alloy, 29.6% Mg, 33.1% Zn and 27.3% Y for sample ZK60–2Y, and 89.2% Mg and 10.8% Ce for the ZK60–2RE alloy. The corresponding EDS spectra are shown in Figure 2.

**Figure 1.** SEM images for (**a**) ZK60, (**b**) ZK60–2Y, and (**c**) ZK60–2RE alloys. The bright areas in the SEM images correspond to the secondary phase particles. The red arrows and the letters A, B and C show the locations where the EDS analysis was performed. The corresponding EDS spectra are shown in Figure 2. Color-coded EDS elemental maps for (**d**) ZK60–2Y and (**e**) ZK60–2RE alloys (Mg: orange, Zn: red, Y and Ce: turquoise).

**Table 1.** Microstructural characteristics and mechanical properties of the three studied alloys (*f* P: fraction of secondary phase particles as obtained by SEM; *f* DRX: fraction of recrystallized volumes determined by EBSD; *f* HAGB: HAGB fraction obtained by EBSD; *ρ*: dislocation density determined by XLPA; UTS: ultimate tensile strength).


**Figure 2.** EDS spectra corresponding to (**a**) point A in Figure 1a, (**b**) point B in Figure 1b, and (**c**) point C in Figure 1c.

The structure of the secondary phase particles was studied by XRD. A part of the XRD patterns containing peaks of the precipitates is shown in Figure 3. Using the ICDD PDF-2018 database, the particle structures were identified. It was found that the particles in the ZK60 alloy were Mg4Zn7 (PDF card no.: 01-071-9625) with a monoclinic structure. The magnitude of the edge vectors of the cell (i.e., the lattice constants) are a = 2.596 nm, b = 0.524 nm, and c = 1.428 nm. The angles between these edge vectors are α = 90◦, β = 102.5◦, and γ = 90◦. In the ZK60 alloy, the precipitates were identified as Mg3Y2Zn3 (PDF card no.: 00-036-1275) with an fcc structure (the lattice constant is 0.6833 nm). In the ZK60–2RE alloy, the secondary phase particles were identified as Mg41Ce5 (PDF card no.: 01-071-7012) with a tetragonal crystal structure. The lattice constants of this structure are a = 1.478 nm and c = 1.043 nm. It is worth noting that there is reasonable agreement between the chemical compositions determined by EDS (see the former paragraph) and the nominal compositions of the precipitate structures identified by XRD. The slight difference could have been caused by the fact that it was not possible to exclude a fraction of the EDS signal that came from the material beneath the particles. In addition, the actual compositions of the precipitates may have slightly differed from the nominal chemical compositions given in the XRD cards without changing the structure; therefore, these small differences could not be observed by XRD. Similarly to SEM, XRD also suggested a higher amount of precipitates in the ZK60–2Y and ZK60–2RE alloys compared to the ZK60 alloy (see Figure 3).

**Figure 3.** A part of the XRD patterns for the studied alloys.

The grain orientation maps and misorientation angle distribution histograms obtained by EBSD for the three alloys are shown in Figure 4. The analysis of the EBSD images revealed that the grain size of the ZK60 alloy was reduced from 6.5 μm to 2.1 μm and 2.8 μm when Y and Ce-rich RE elements were added to the base alloy, respectively (Figure 4a–c). According to the misorientation angle distribution histograms (see Figure 4d–f), the fraction of HAGBs (*θ* > 15°) was more than 90% for all three alloys. In addition, the presence of a sharp peak at 30°indicated the occurrence of continuous DRX (CDRX) [25]. Thus, it is suggested that the fine grains in the extruded samples formed due to recrystallization. On the other hand, some large elongated grains were present in the ZK60–2Y and ZK60–2RE alloys (see Figure 4g,h). These grains can be attributed to the segregation of large atoms, such as Ce, La, Nd, and Pr or Y, due to dislocations and grain boundaries which hindered recrystallization during extrusion and increased the required strain to complete the DRX process [18]. On the other hand, coarse secondary phase particles (usually larger than 1 μm) could promote PSN as a DRX mechanism in the ZK60–2Y and ZK60–2RE alloys [26]. It is obvious from Figure 5a that the predominant texture component for all alloys was fiber-like, in which the c-axis of the HCP crystals were aligned perpendicular to ED, and the (1010) and (2110) planes were parallel to the surface analyzed by EBSD. This type of texture is common in extruded Mg alloys and limits their formability due to the limited activation of easy glide systems [2]. It is evident in Figure 5a–c that the texture intensity increased with the addition of Y and Ce-rich RE elements. This type of texture change can be explained by the higher volume fractions of un-DRXed grains (see Figure 4a–c).

**Figure 4.** (**a**–**c**) EBSD grain orientation maps (the extrusion direction is horizontal), and (**d**–**f**) misorientation angle distribution histograms for ZK60, ZK60–2Y, and ZK60–2RE alloys, and higher magnification EBSD maps for (**g**) ZK60–2Y, and (**h**) ZK60–2RE alloys.

The average crystallite size and the dislocation density were determined by the XLPA method. As an example, Figure 5d demonstrates the CMWP fitting on the X-ray diffraction pattern obtained for the ZK60–2Y alloy. Only a part of the fitted pattern is shown in the figure. Table 1 shows that the crystallite size decreased with the addition of RE alloying elements. This trend is in line with the grain size reduction observed by EBSD. On the other hand, the crystallite size obtained by XLPA was much smaller than the grain size determined by EBSD. This phenomenon is well-known in plastically deformed metallic materials and is caused by the hierarchical nature of the deformed microstructures [22]. Indeed, during deformation, dislocations are formed and arranged into low-angle grain boundaries and/or dipolar walls in order to reduce the energy of the dislocation structure. The volumes separated by these dislocation boundaries exhibit low misorientations which break the coherency of the scattered X-rays; therefore, XLPA detected these volumes as individual crystallites which were considerably smaller than the grain size.

**Figure 5.** Inverse pole figures for (**a**) ZK60, (**b**) ZK60–2Y, and (**c**) ZK60–2RE alloys, and (**d**) CMWP fitting for the extruded ZK60–2Y alloy. The open circle and the solid line represent the measured data and the fitted XRD pattern, respectively. The intensity is in a logarithmic scale. (**e**) The dislocation density for the three studied alloys determined by CMWP fitting.

XLPA revealed that in the extruded ZK60 alloy the dislocation density was about 1.2 × 1014 <sup>m</sup><sup>−</sup>2, which did not change remarkably due to the addition of 2% Y (see Figure 5e). On the other hand, the dislocation density increased to 2.3 × <sup>10</sup><sup>14</sup> <sup>m</sup>−<sup>2</sup> in the ZK60– 2RE alloy. The higher dislocation density for sample ZK60–2RE can be attributed to the

alloying effect since the large RE atoms could hinder the dislocation annihilation during the extrusion either by dissolving into the Mg matrix or by forming precipitates. It is surprising that the addition of 2% Y did not yield an increased dislocation density in the extruded sample. This effect can be understood if we assume that most yttrium was concentrated in the precipitates. Since a major fraction of these particles formed at the grain boundaries, as suggested by SEM (see Figure 1), they had no significant effect on the dislocation multiplication and annihilation occurring in the grain interiors. Therefore, the yttrium addition did not increase the dislocation density compared to the base ZK60 alloy. For sample ZK60–2RE, the relative fraction of RE atoms in the precipitates was much lower than that for specimen ZK60–2Y. Namely, the fractions of RE atoms in Mg41Ce5 and Mg3Y2Zn3 particles were 11% and 25%, respectively. On the other hand, the volume fractions of precipitates in the two alloys were similar (see Table 1). Therefore, most probably, the ZK60–2RE alloy contained more solute RE atoms in the grain interiors than the ZK60-2Y alloy did, which possibly contributed to the higher dislocation density after extrusion. In addition, the average size of the Ce, La, Nd, and Pr atoms present in the ZK60-2RE sample was about 5% larger than that of the Y added to the ZK60-2Y alloy. Therefore, the pinning effect of the former elements should have been higher on dislocations compared to that of the latter one. This effect coud have also contributed to the higher dislocation density in sample ZK60-2RE.
