**2. Materials and Methods**

In the frame of the present work, three alloy compositions (Ti-2Fe, Ti-4Fe, and Ti-10Fe) were prepared and investigated. The numerical values give the iron concentrations in wt.%. The alloys were produced by induction melting of pure materials (Ti: 99.9% and Fe: 99.97%) in vacuum. The cast rod-shaped samples (10 mm in diameter) were cut into the disks of 0.7 mm thickness, which were polished, etched, and annealed for 4000 h at 470 ◦C in vacuum and subsequently quenched in water. The annealing was carried out in sealed fused silica ampoules, which were evacuated up to the pressure of 4 <sup>×</sup> <sup>10</sup>−<sup>4</sup> Pa. The heat-treated samples were deformed by high-pressure torsion (HPT, five rotations) in a Bridgeman anvil-type press (Klement, Lang, Austria). The HPT process was carried out at an ambient temperature, at a pressure of 7 GPa and with the deformation speed of 1 rpm while using a computer-controlled HPT device.

The oxygen content in the samples was determined using carrier gas hot extraction (CGHE) after the HPT process and after the HTXRD measurements. For the CGHE analyses, the GALILEO G8 device (Bruker AXS, Karlsruhe, Germany) was used. The measurements were performed in a graphite crucible while using He as the carrier gas. Before each measurement, the samples were dipped in diluted HCl acid to remove oxide layers from the sample surface. The spatial distribution of the phases was obtained from the scanning electron micrographs for both heat-treated and deformed states. The SEM images were recorded while using back-scattered electrons (SEM/BSE) with the JSM-7800 F (JEOL, Tokyo, Japan) microscope operating at the accelerating voltage of 20 kV. The grain-size distributions of the phases were determined using electron backscatter diffraction (EBSD) that was performed with an EDAX/EBSD system. The analysis of the data was done while utilizing the OIM Analysis software (version 8, AMETEK/EDAX TSL, Mahwah, NJ, USA). The phase identification in the heat-treated and HPT deformed samples was performed by means of X-ray diffraction (XRD) in a D8 Advanced (Bruker AXS, Germany) diffractometer. The diffractometer worked in the Bragg–Brentano geometry, and it was equipped with a sealed X-ray tube with Co anode (wavelength of CoKα<sup>1</sup> = 0.178897 nm), with a Johannson-type monochromator in the primary beam that suppressed the spectral line CoKα<sup>2</sup> and with a LynxEye one-dimensional detector. The ex situ XRD patterns were analyzed using the Rietveld method [30,31] (whole pattern refinements), as implemented in the TOPAS [32] software package (version 5, Bruker AXS, Karlsruhe, Germany).

The HTXRD experiments were performed at a constant heating rate of 10 K/min in the temperature range between room temperature and 750 ◦C using a θ-θ Bruker D8 Advance diffractometer (Bruker AXS, Germany) that utilized the Bragg–Brentano geometry, CuKα radiation (wavelengths: 0.154056 nm and 0.154437 nm), a LynxEye XE one-dimensional detector (Bruker AXS, Germany) and the high-temperature chamber MTC HighTemp+ (Bruker AXS, Germany). During the HTXRD measurements, the samples were placed on a resistively heated tantalum strip that was surrounded by an additional Ta radiation heater to reduce the temperature gradients in the sample. A type D thermocouple (W-Re3/W-Re25), which was contacted on the heating strip below the sample position, was used to measure the sample temperature. The sample was placed on a (100)-oriented sapphire plate in order to avoid reactions between the sample and the Ta heater. The chamber was evacuated to a pressure of ~7 <sup>×</sup> 10−<sup>3</sup> Pa to reduce the oxidation of the sample. The HTXRD patterns were recorded in a 2θ range between 33◦

and 45◦ with a step size of 0.02◦ in order to speed up the measurement and reduce the difference in the annealing time and in the temperature during the acquisition of individual diffraction patterns. The temperature difference between the beginning and the end of each measurement was 20 K at the acquisition time of 120 s per diffraction pattern.

Differential scanning calorimetry (DSC) was utilized to determine phase transformation temperatures in the heat-treated samples before and after the HPT process. For DSC, the samples were placed in a platinum crucible with a thin Al2O3 inlet to avoid reactions of the samples with the Pt crucible. The measurements were performed in inert Ar atmosphere (99.999% + Varian cleaning system) while using the differential scanning calorimeter DSC Pegasus 404C (Netzsch, Selb, Germany). Before each experiment, the calorimeter was evacuated and back-filled with Ar several times to remove the remaining oxygen from the DSC chamber. The measurements were done in the temperature range between 35 ◦C and 1100 ◦C using the same heating rate of 10 K/min as used for the HTXRD measurements. The temperatures that were recorded in the DSC device were calibrated using melting points of pure metals (In, Sn, Al and Au).
