*4.2. Thermal Stability of the HPT-Deformed Microstructure*

At ambient conditions, all of the HPT-deformed samples contained a mixture of three phases, i.e., α-Ti(Fe), TiFe and ω-Ti(Fe). The phase amounts were different, depending on the alloy composition. Upon heating, a relatively large exothermal DSC effect was registered at ~130 ◦C (onset temperature). However, in alloy Ti-4Fe, the beginning of the exothermal DSC effect was less abrupt than for the other alloys (compare Figure 4). The exothermal effect is related with the initiation of the decomposition process of the pressure-induced ω-Ti(Fe) phase and with the release of the deformation energy. Similar observations have already been made for pure Ti [51] and for Ti-1Fe [23]. At a temperature of ~320 ◦C, the phase amount of ω-Ti(Fe) decreases abruptly. Between 400 ◦C and 450 ◦C, this metastable phase completely disappeared (Figure 7). In a previous study [27] that was devoted to the investigation of the thermal stability of ω-Ti(Fe) in HPT deformed metastable β-(Ti,Fe) and α-Ti + β-(Ti,Fe) alloys, a cascade of exothermal DSC effects was observed between 150 ◦C and 450 ◦C. These effects originated from the gradual transformations of ω-Ti(Fe) to α-Ti(Fe) and from the defect annihilation and recrystallization processes [27]. Moreover, the strongest exothermal effect was observed at around 380 ◦C, whereas the released heat increases with increasing iron content within the investigated alloys. These results demonstrate that ω-Ti(Fe) possesses a lower thermal stability, if it is formed from α-Ti (present work) than if it is formed from β-(Ti,Fe). The differences in the phase composition of the HPT samples might cause this difference in the stability of the pressure-induced ω-Ti(Fe) phase, but also by the differences in the iron concentration in ω-Ti(Fe), because ω-Ti(Fe) originating from β-(Ti,Fe) possesses a higher Fe concentration than ω-Ti(Fe) stemming from α-Ti(Fe). The ω-Ti(Fe) phase, which was formed from the α-Ti + TiFe two-phase alloys contained approximately 1 wt.% Fe after HPT, whereas the ω-Ti(Fe) generated from metastable β-(Ti,Fe) alloys was found to contain ~4 wt.% Fe [26–28].

In the intermediate range between 400 ◦C and 562 ◦C, all of the alloys exhibit a two-phase α-Ti(Fe) + TiFe microstructure. At ~500 ◦C, which is still below the eutectoid reaction temperature, a shift of the line 0002 from α-Ti(Fe) towards lower diffraction angles was detected by HTXRD, together with the appearance of diffraction lines of β-(Ti,Fe). This phenomenon was already observed in former studies [27,29], where it was assigned to a variation of the iron solubility in the α-Ti(Fe) phase. The decomposition of the supersaturated α-Ti(Fe) phase to the equilibrium α-Ti phase is related to the shift of the diffraction line 0002α-Ti(Fe) towards smaller diffraction angles. In the present work, this effect is much less pronounced, because the phase amount of ω-Ti(Fe) only approaches approx. 50% after the HPT process. A pronounced endothermic heat effect corresponding to the eutectoid reaction β-(Ti,Fe) α-Ti + TiFe was registered at ~584 ◦C in the annealed, but not deformed, samples. The eutectoid reaction temperature is in good agreement with References [14,40–43], where this reaction was reported between 583 ◦C and 590 ◦C. After the HPT deformation, the measured temperature of the eutectoid reaction was found to be lowered to ~562 ◦C. The same behavior was observed for the β-(Ti,Fe)-transus temperature. Thus, strong deformations and/or small grain sizes of the α-Ti(Fe) and TiFe phases, which are characteristic microstructural features after the HPT process, lead to a shift of the equilibrium reaction temperatures towards lower values. Moreover, the development of the phase fractions upon heating (Figure 7) shows, in contrast to the phase diagram (Figure 6), that the phase amount of β-(Ti,Fe) does not continuously increase. At temperatures above 600 ◦C in the HTXRD measurements, the phase amount of β-(Ti,Fe) stagnated, and even decrease in Ti-2Fe and Ti-4Fe. The reason for that behavior is a slight oxidation of the samples during the HTXRD measurement. The CGHE measurements revealed increased oxygen content inside the samples after the HTXRD measurements (see Section 3.2), which confirms that the oxidation of the samples at high-temperatures, even under high vacuum, could not be prevented. Oxygen is an element stabilizing the α-Ti phase [52]. Therefore, α-Ti is stabilized at temperature above 600 ◦C and, thus, no single-phase β-(Ti,Fe) state was generated inside the samples upon further heating in the HTXRD device.
