Search

In this work, the wetting phase transition of grain boundaries (GBs) and their triple junctions (GB TJs) by the second solid phase in the magnesium-based alloy EZ33A is studied. The condition for complete wetting for the GB TJ ( $$\left({{\sigma _{{\rm{GB}}}}>\sqrt3{\sigma _{{\rm{SS}}}}}\right)$$ ) is weaker than that for GBs (σGB > 2σSS). Therefore, if the transition from partial to complete wetting occurs with increasing temperature, then all GB TJs should become completely wetted at a temperature TwTJ, which is lower than the temperature TwGB, at which all GBs become completely wetted. For the first time, it was experimentally found that GB TJs are completely wetted at TwTJ = 380±10 °C, which is approximately 70 °Clower than TwGB = 450 ± 10 °C. The wetting phase at the GBs is the intermetallic compound (Mg, Zn)12RE. A similar phenomenon was previously observed for GB TJ wetting with a liquid phase [B. B. Straumal, O. Kogtenkova, and P. Zieba, Acta Mater. 56, 925 (2008)].

In this work, we studied the behavior of the Nd–Dy–Fe–Co–Cu–B alloy for permanent magnets under high pressure torsion (HPT). In the initial state of the studied alloy, it mainly contained the crystalline phase τ1 (Nd, Dy)2(Fe, Co, Cu) 14B. After HPT at room temperature (THPT = 30°C), a mixture of an amorphous phase with nanocrystalline inclusions of the τ1 phase is observed in the alloy. In the equilibrium phase diagram, this state is equivalent to a mixture of the τ1 phase with the melt at the temperature Teff= ∼1100°C. The thus determined Teff value is called the effective temperature. When the THPT temperature of the HPT treatment increases to 300 and 400°C, the amorphous phase disappears, and the Fe2B and γ-Fe phases appear instead. In the equilibrium phase diagram, this state is equivalent to a mixture of phases τ1+ Fe2B + γ-Fe, which is observed in the temperature range from ∼950 to ∼1050°C. We explain this phenomenon by the fact that with an increase in the HPT temperature THPT, the rate of formation of defects during deformation remains constant, but the rate of their thermal relaxation (annihilation) increases. This is equivalent to decrease in the effective temperature Teff in the equilibrium phase diagram. The previously predicted decrease in Teff with an increase in THPT is observed for the first time.

The effect of the phase composition on the α/β-Ti(Fe)→ω-Ti(Fe) transformation in the Ti-4 wt % Fe alloy under shear strain with high-pressure torsion (HPT) has been studied. For shear deformation by means of HPT, two initial states of the alloy were used, which significantly differed in the morphology of the phases and the concentration of iron atoms in the β phase. During HPT, a stationary state occurred in both sample series, which is characterized by the presence of a single ω phase containing 4 wt % Fe and by a grain size of about 200 nm. Thus, the HPT state is equifinal and independent of the initial phase composition of the samples. It was found that under the influence of HPT in Ti-4 wt % Fe alloys not only martensitic (shear) transformation into the ω phase occurs, but also a significant mass transfer of atoms of the alloying element. An analysis of the change in the torsion torque directly in the HPT process made it possible to estimate the rate of deformation-induced mass transfer. It is 18–19 orders of magnitude higher than the rate of conventional thermal diffusion at the processing temperature THPT = 30°C, while it is close to the diffusivity values at 700–800°C. This is because HPT increases the concentration of lattice defects, which in turn is equivalent to an increase in temperature. A similar combination of accelerated mass transfer during HPT and martensitic (shear) transformation was previously observed in copper-based shape memory alloys, but for the first time studied for the formation of ω-phase in titanium alloys.

Spectra of grain boundaries existing in the polycrystalline copper-silver system (positive enthalpy of mixing) have been studied in comparison with those of the copper-indium system (negative enthalpy of mixing). Spectra of grain boundaries are formed spontaneously upon an increase in the temperature, occurrence of eutectic and peritectic reactions, and subsequent relaxation of structures in the two-phase solid/liquid region of phase diagrams. It has been shown that the rate of grain growth and the relation between different grain boundary types in the total spectrum depend on the enthalpy of mixing.

Phase transformations of some Hume-Rothery phases (electron compounds) to others in a copper-tin system subjected to high-pressure torsion were detected in our previous work [B.B. Straumal et al., JETP Lett. 100, 376 (2014)]. In particular, the torsion of the ζ + e phase mixture at high pressure led to the formation of the δ + e phase mixture, as after long-term annealing in the temperature interval Teff = 350–589°C. In this work, it has been shown that the high-pressure torsion of α-solid solutions of tin in copper results in the final stable solid solution whose composition is independent of the composition of the initial α phase before high-pressure torsion. The final composition is the same as after long-term annealing at the temperature Teff = (420 ± 10)°C. The rate of high-pressure torsion-induced mass transfer is several orders of magnitude higher than the rate of conventional thermal diffusion at the treatment temperature THPT and is close to values at Teff. This occurs because high-pressure torsion increases the concentration of lattice defects and this increase is in turn equivalent to an increase in the temperature.

The complete solubility of an impurity in a polycrystal increases with decreasing grain size, because the impurity dissolves not only in the crystallite bulk but also on the grain boundaries. This effect is especially strong when the adsorption layers (or the grain boundary phases) are multilayer. For example, the Mn solubility in the nanocrystalline films (where the size of grains is ∼20 nm) is more than three times greater than that in the ZnO single crystals. The thin nanocrystalline Mn-doped ZnO films in the Mn concentration range 0.1–47 at % have been obtained from organic precursors (butanoates) by the “liquid ceramic” method. They have ferromagnetic properties, because the specific area of the grain boundaries in them is greater than the critical value [B.B. Straumal et al., Phys. Rev. B 79, 205206 (2009)]. The high-resolution electron transmission microscopy studies show that the ZnO nanocrystalline grains with the wurtzite lattice are separated by amorphous layers whose thickness increases with the Mn concentration. The morphology of these layers differs greatly from the structure of the amorphous prewetting films on the grain boundaries in the ZnO:Bi2O3 system.