**3. Results**

#### *3.1. Microstructure Evolution of the Alloys during HPT*

### 3.1.1. SEM Analyses

Figure 1a,b shows the microstructure of the as-cast Al-2Fe and Al-4Fe samples. For the Al-2Fe alloy (Figure 1a), the microstructure consists of the net/web of eutectic phase particles, surrounding the dendritic aluminum phase. This could also be described as the ellipsoidal-shape areas of the aluminum phase, divided by aluminum-intermetallic conglomerates. The volume fraction of the intermetallic phase in the Al-2Fe alloy is 4.6%.

The microstructure of the Al-4Fe alloy is different from that of the Al-2Fe alloy. The amount of the intermetallic phase is higher than 7.7%. It comprises very coarse plate-like particles with a thickness/width up to 10 microns, and a length up to 50 microns (Figure 1b). In addition to these coarse particles, the microstructure also consists of smaller fractions of particles distributed between them. Smaller particles in the Al-4Fe alloy form a structure, similar to Al-2Fe alloy, only these particles are larger and are located sparsely in comparison.

The intermetallic particles in the alloys are the mix of the Al6Fe and Al13Fe4 phases [21], formed during the crystallization [31–36].

**Figure 1.** Backscattered SEM images of the samples of the Al-2Fe and Al-4FE alloys: (**a**) Al-2Fe as-cast, (**b**) Al-4Fe as-cast, (**c**) Al-2Fe after HPT, (**d**) Al-4Fe after HPT.

It can be seen that the Al-2Fe alloy has a casting microstructure very similar to that in Al-RE (rare earth) alloys [12,13] and other Al-Fe alloys [20–23].

Figure 1c,d show the microstructure formed in alloys during HPT. While the intermetallic particles in the Al-2Fe alloy are pretty small and hardly distinguishable (Figure 1c), the size of these particles in Al-4Fe after HPT varies from a fraction to tens of microns.

Since the intermetallic particles in the Al-2Fe alloy exhibit a narrow size distribution, we can affirm that the fragmentation of the particles occurred homogeneously. On the contrary, the refinement of the coarse particles in the Al-4Fe alloy was incomplete, as fragments of large particles are still exhibited in the microstructure (Figure 1d) and the particle size distribution is large.

### 3.1.2. XRD Analyses

XRD profiles corresponding to the as-cast and HPT-processed states of Al-2Fe and Al-4Fe alloys are presented in Figure 2, where peaks of different phases are indexed. For the phase analysis we have used the reference crystal structures corresponding to Al (FCC, *a* = *b* = *c* = 4.049 Å), Al6Fe (orthorhombic, *a* = 6.46 Å, *b* = 7.44 Å, *c* = 8.78 Å), and Al13Fe4 (monoclinic, *a* = 15.49 Å, *b* = 8.08 Å, *c* = 12.47 Å, β=107.69◦), where *a, b, c* are the lattice parameters. Thanks to the analysis of the XRD profiles we have revealed the phases with the crystal structure close to stoichiometry of the reference phases. Interestingly, the ascast specimen of Al-2Fe alloy mainly contains a stable Al13Fe4 phase, while in the as-cast Al-4Fe alloy a metastable Al6Fe phase is dominant. This can be explained by the chemical composition of the as-cast materials: it is known that, depending on the alloy composition and cooling rate, different types of Al-Fe intermetallic phases can be formed [36].

**Figure 2.** XRD profiles of the Al-2Fe and Al-4Fe alloys in the as-cast state and after HPT.

After HPT the opposite trend is observed: Al6Fe and Al13Fe4 phases are formed in the Al-2Fe and Al-4Fe alloys, respectively, while the peaks of precursor particles disappear.

Revealing possible scenarios for phase transformations in Al-Fe alloys during HPT is beyond the scope of this study and will be reported later. The results of qualitative phase analysis and structural parameters evaluation are thus presented.

The precise quantitative phase analysis is complicated by texture effects especially in the coarse-grained states where statistically limited number of grains are oriented in the reflection position with respect to the scattering vector. Nevertheless, the XRD profile of Al-4Fe in the as-cast state demonstrates visibly the highest intensity of all major peaks related to the precipitate phase among all studied cases. It indicates that this state is characterized with the maximal volume fraction of precipitates.

HPT resulted in a noticeable shift in the positions of Al peaks to higher diffraction angles in both alloys. The calculated lattice parameter for Al matrix reduced by 0.078% for the Al-2Fe alloy and 0.035% for the Al-4Fe alloy (which is well consistent with the data reported in [36]). This indicates to a process of strain-induced dissolution of Fe in Al [36], which is more intensive in Al-2Fe. Besides, Al-2Fe has a smaller coherent domain size and higher dislocation density: about 90 nm and 2.2 × 10<sup>14</sup> m<sup>−</sup>2, respectively, versus 200 nm and 2.9 × 10<sup>13</sup> m<sup>−</sup><sup>2</sup> for the Al-4Fe alloy. From the presented data it follows that HPT-induced structure refinement, defect accumulation, and Fe dissolution are more pronounced in Al-2Fe compared to Al-4Fe despite the difference in the alloying element concentration.

#### 3.1.3. TEM and STEM Analyses

Figure 3a,b shows the bright-field TEM images, comparing the microstructures of the Al-2Fe and Al-4Fe alloys samples after HPT, respectively. To the right from BF images in Figure 3 the selected area diffraction (SAED) patterns obtained from a region of 1.5 μm size, are presented. The number of particles, contained within the aperture area of the SAED patterns, is quite small, thus the number of the reflections from particles is small as well. However, the number of reflections was just enough to index phases.

**Figure 3.** TEM bright-field (**left**) and SAED patterns (**right**) from the samples of the Al-2Fe (**a**) and Al-4Fe (**b**) after HPT. Indexed reflections are highlighted for the Al and Al13Fe4 phases with the corresponding color.

Analysis of the SAED patterns showed the presence of the Al and Al13Fe4 phase in both Al-2Fe and Al-4Fe alloys after HPT. Peaks that could be attributed to the Al6Fe phase could not be observed in TEM. This is probably due to the fact, that the amount of Al6Fe phase is quite small. Besides, overlap between the Al6Fe (131), Al6Fe (222) and Al6Fe (132) peaks with Al (111) and Al (200) peaks makes them hardly distinguishable.

The number of reflections on the Al-4Fe SAED patterns is lower due to the larger mean grain size.

The microstructure of the Al-2Fe alloy after HPT is composed of equiaxed, ultrafine grains with a mean grain size ~125 ± 10 nm. Figure 3b shows that the microstructure of the Al-4Fe alloy after HPT exhibits the equiaxed grains with a mean size ~340 ± 20 nm. The microstructures of both alloys exhibit grains with high-angles misorientation.

Bright-field TEM does not provide sufficient information for the size, distribution, and morphology of the intermetallic particles. Intermetallic particles in these alloys contains a high amount of Fe, which should provide a clear contrast between the matrix and the particles in the z-contrast mode (HAADF) of STEM [24].

Figure 4 shows STEM images of Al-2Fe (Figure 4a,b) and Al-4Fe (Figure 4c,d) alloys after HPT. According to these data, the mean size of particles in the Al-4Fe alloy is significantly larger than in the Al-2Fe alloy (277 ± 16 nm versus 78 ± 4 nm). The difference in grain size of these two alloys (Figure 3a,b) is very likely to be connected with the difference in particle size since the deformation conditions for both alloys were similar.

**Figure 4.** STEM HAADF images, obtained from the samples of the Al-2Fe alloy (**<sup>a</sup>**,**b**) and Al-4Fe alloy (**<sup>c</sup>**,**d**) HPT.

The origin of differences in the two alloys, subjected to the same deformation treatment, probably lies in the particle fragmentation process. In the Al-2Fe alloy, linear arrays of fine particles can be observed (Figure 4b). Such arrays are former intermetallic particles, fragmented and elongated during the deformation.

The particle fragmentation in the Al-4Fe alloy after HPT is different. During the observation, no similar arrays could be observed. Instead, "clouds" of finer particles surrounding coarse ones were exhibited (Figure 4d). This means that the amount of deformation induced by HPT was probably not large enough to cause cracks throughout the whole particle. Instead, small particles are fragmented from the surface region, while the core of the coarse particle remains intact. Thus, even after HPT, the coarse particles are still present, and fine particles are distributed non-uniformly. All this contributes to a reduced impact on the structure evolution during the HPT in comparison to the Al-2Fe alloy, as it will be discussed later

#### *3.2. Physical and Mechanical Properties*

Table 1 presents the results of the study of electrical and mechanical properties of Al-2Fe and Al-4Fe alloys in as-cast condition and after HPT.


**Table 1.** Mechanical and electrical properties of Al-Fe alloys in the as-cast state and after HPT.

Despite the difference in the iron content and the intermetallic particle morphology in the initial state, both alloys have a comparable ultimate tensile strength (~90 MPa). The presence of coarse intermetallic particles in the Al-4Fe alloy embrittles it Al-4Fe alloy elongation to failure is 8% lower than that of the Al-2Fe alloy in the initial state (Figure 5).

**Figure 5.** Engineering stress-strain curves obtained by tensile tests of Al-Fe alloys in the as-cast state (1: Al-2Fe, 2: Al-4Fe) and after HPT (3: Al-2Fe, 4: Al-4Fe). For the sake of clarity, only one curve is provided for each state.

The level of the electrical conductivity is higher for the Al-2Fe alloy in the as-cast state. This is due to the lower total amount of iron in the alloy, and due to the absence of coarse intermetallic particles, contrary to the Al-4Fe alloy.

As a result of microstructure refinement during HPT, the level of the ultimate tensile strength increases to ~650 MPa for the Al-2Fe alloy and to ~340 MPa for the Al-4Fe alloy. The increase in tensile strength is not as big for the Al-4Fe alloy as for the Al-2Fe alloy due to the larger intermetallic particle size and a nonuniform distribution of nanoscaled particles (Section 3.1.3). HPT also leads to a decrease in the ductility of the Al-2Fe and the Al-4Fe alloys. It should be noted, however, that after HPT both alloys still maintain a relatively high level of elongation to failure: ~5% for the Al-2Fe alloy and ~8% for the Al-4Fe alloy. Because the fragmentation of the intermetallic particles in the Al-4Fe alloy is not as complete as in the Al-2Fe alloy, the Al-4Fe alloy samples still contain areas free of the finely dispersed second phase particles. These areas improve the elongation to failure for the Al-4Fe alloy in comparison to the Al-2Fe alloy.

The electrical conductivity after HPT decreases for both Al-2Fe and Al-4Fe alloys. The decrease in the electrical conductivity is 15.4 %IACS for the Al-2Fe alloy and 6.2 %IACS for the Al-4Fe alloy (Table 1). The electrical conductivity is affected, among other things, by the presence of the finely dispersed second phase particles and the size of the grains. It was observed in Section 3.1.3 that Al-4Fe alloy after HPT has a higher mean grain size and also contains areas, free of the fine intermetallic particles, while Al-2Fe alloy after HPT is characterized by lower mean grain size and uniformly distributed fine second-phase particles. Such difference in microstructure may explain the difference in the level of electrical conductivity.

The electrical conductivity is known to be particularly sensitive to the presence of the solid solution, so the sharp decrease of it in the Al-2Fe alloy can be the indirect sign of the solid solution formation. This will be discussed in the detail in the Discussion section.
