*3.1. Characterization of the Initial State of the Samples*

After annealing at 470 ◦C, i.e., below the temperature of the eutectoid reaction β α-Ti + TiFe, the alloys Ti-2Fe, Ti-4Fe, and Ti-10Fe show a two-phase α-Ti + TiFe microstructure (see Figure 1a–c). Some TiFe particles were ordered in chains along the boundaries of the α-Ti grains (see, for example, Figure 1a). This phenomenon is called incomplete or complete grain boundary wetting by a second solid phase and it has been previously observed in various Ti-based alloys [23,33–37]. In the corresponding SEM/BSE micrographs, the TiFe grains appear to be bright due to a higher Fe content and, thus, a higher mean atomic number of TiFe as compared to α-Ti. Moreover, no coarse-grain microstructure of TiFe was observed, even though the samples were annealed for 4000 h at 470 ◦C. The averaged grain sizes of α-Ti, determined using EBSD were (3.1 ± 2.0) μm, (3.0 ± 2.0) μm, and (1.4 ± 1.0) μm in the alloys Ti-2Fe, Ti-4Fe, and Ti-10Fe, respectively. The sizes of the TiFe grains were (0.6 ± 0.2) μm, (0.7 ± 0.3) μm, and (2.1 ± 1.0) μm in the same samples. The limited growth of the α-Ti and TiFe grains can be explained by either very slow recrystallization and grain-growth kinetics or by very slow diffusion velocities at the heat-treatment temperature of 470 ◦C. The phase fractions of α-Ti and TiFe were concurrently determined from the whole XRD patterns using the Rietveld method and from the chemical composition of the respective alloy using the lever rule for comparison (see Table 1) [14].

**Figure 1.** Initial microstructures of samples Ti-2Fe (**a**), Ti-4Fe (**b**), and Ti-10Fe (**c**) annealed at 470 ◦C for 4000 h. The white grains belong to TiFe, the gray areas to α-Ti.

Upon the Rietveld refinement, the lattice parameters of α-Ti and TiFe were calculated in addition to the phase composition that was fit together with the degree of the preferred orientation {0001} of the α-Ti crystallites. The March–Dollase model described this texture, which was probably caused by the foregoing sample preparation and annealing [38]. Applying this model, the phase fractions obtained from XRD for Ti-2Fe and Ti-4Fe agree very well with the phase fractions calculated from the chemical composition (Table 1). In alloy Ti-10Fe; however, the apparent texture was much more complex, because bad grain statistics caused the differences in diffracted intensities. Consequently, it was not possible to describe the texture satisfactorily using the March–Dollase model, which led to a larger uncertainty in the calculated phase fractions of α-Ti and TiFe (see Table 1). Assuming that the equilibrium state was achieved in all samples, the chemical compositions of α-Ti and TiFe are the same for all of investigated alloys (independent of the iron content). The lattice parameters of α-Ti and TiFe were *a*α-Ti = 0.2951(2) nm, *c*α-Ti = 0.4691(2) nm and *a*TiFe = 0.2978(2) nm, respectively. The obtained lattice parameters *a*α-Ti and *c*α-Ti correspond to the maximum Fe content in α-Ti at 470 ◦C. The lattice parameter *a*TiFe agrees with the reference value for TiFe [39].

determined from the X-ray diffraction (XRD) measurements and calculated using the thermodynamic description of the binary Ti–Fe system by application of the lever rule. All of the phase amounts are given in wt.%. The uncertainty of the XRD phase analysis utilizing Rietveld refinement falls generally within a range of 1% to 3%.

**Table 1.** Comparison of the phase fractions in the samples annealed at 470 ◦C for 4000 h, which were


The phase transition temperatures of the annealed alloys were determined by means of DSC upon heating at the heating rate of 10 K/min and they are shown in Figure 2. The endothermic heat effect registered at ~584 ◦C corresponds to the eutectoid reaction β α-Ti + TiFe that was detected in all of the samples, but with different extents. The increase of the heat amount with increasing Fe content is related to a higher amount of TiFe taking part in the eutectoid reaction. The β-transus temperatures (marked by crosses in Figure 2) decrease with an increasing Fe content, because the eutectoid point is located at higher iron contents (between 11 wt.% and 14 wt.% Fe) [14,40–43].

**Figure 2.** Differential scanning calorimetry (DSC) heating curves of alloys Ti-2Fe, Ti-4Fe, and Ti-10Fe measured with the heating rate of 10 K/min. The DSC curves were shifted vertically for better visibility. The dashed line indicates the temperature of the eutectoid reaction β α-Ti + TiFe. The crosses indicate the β-transus temperatures (solvus temperatures of the β phase), which were determined as inflection points of the respective DSC curve.
