There is no need to explain to metallurgists and technologists working with alloys the exceptional importance of the reliability of phase diagrams. The information in the phase diagrams is actually a guide for any specialist who has to deal with this or that alloy, and therefore, it has no right to be unreliable. However, those specialists who constantly use such diagrams in their work are already accustomed to the fact that many lines separating certain regions in phase diagrams should not be trusted. What these lines are, the researchers know well. These are lines separating regions with different the microstructure, in particular, solid solution regions.

Phase transitions in alloys, depicted on binary phase diagrams, are divided into two types: massive (diffusion-free), in which at a certain temperature all atoms of the alloy take part at once (melting, allotropy, magnetism) and diffusive, when individual atoms of one sort take part in a chemical interaction with individual atoms of either the same (phase separation), or different (ordering) sort. Such a chemical interaction occurs at when these two kinds of pairs of atoms become each other's nearest neighbors. It should be noted that this neighborhood does not appear by chance, but because of long-range forces of attraction between these two atoms.

They began to build phase diagrams at the beginning of the last century, when X-ray phase (XRD) analysis became widespread throughout the world, which was used to detect both massive and diffusion phase transitions. Couse is very simple:

If the lines of a massive transition could be determined with sufficient accuracy by this method, so for building the lines of diffusion transitions, it was not suitable. Highly dispersed particles of new phases, and even more so clusters, were simply impossible to detect by this method. This is because the X-ray method cannot identify particles of new phases whose dimensions are smaller than the regions of coherent X-ray scattering. Therefore, the absence of particles of highly dispersed phases in X-ray patterns was always interpreted by researchers unambiguously: the observed microstructure is a disordered (i.e., ideal) solid solution. However, such an interpretation come up against an uncompromising objection from thermodynamics: as is known, ideal solid solutions do not exist in nature (remember, for example, Raoult's law, which is never fulfilled for all pairs of elements).

In the 60-70s of the last century, when the transmission electron microscopy (TEM) method began widely used to study the microstructure of alloys, many authors were surprised to find that the microstructure of the vast majority of binary alloys quenched into water from the region of solid solutions was two-phase and contained highly dispersed particles of chemical compounds [1,2]. Practically, none of the metallurgists had any doubts that the experiment, and not the phase diagrams, was to blame for such a discrepancy. A version was even invented that the decomposition of alloys, proceeding according to the spinodal mechanism, occurs during the hardening process itself, i.e., in a very short time, comparable to the time of cooling the alloy in water. This behavior of the alloys was explained because during spinodal decomposition, the stage of formation of critical nuclei is absent, and therefore, the nucleation process is sharply sped up.

That is why a collision arose when, using X-ray diffraction data, phase diagrams were constructed containing, as stated, regions of disordered solid solutions at high temperatures, and using the TEM method, it was found that a completely different structure is formed these regions are two-phase[4] . Such areas of "solid solutions" could be found on almost the entire high-temperature part of each binary phase diagram. This means that we do not actually know what the microstructure is in these regions of the phase diagram. It is unlikely that such a situation in this field of knowledge can be called normal, when a TEM experiment reveals one type of structure, and another experiment (X-ray) - another type of structure. At the same time, the accepted theory unconditionally takes the side of one of them, while nothing explaining why there is such a discrepancy.

The “ordering-separation” phase transition was discovered for the first time by the authors [3] in alloys of the Fe-Cr system and later in other binary systems [4]. It implies because, at a well-defined temperature, a surprising phenomenon occurs in the alloy - the sign of the chemical interaction between dissimilar atoms changes. The tendency of the alloy to phase separation is replaced by a tendency to ordering, or vice versa. Because of such a change in the chemical's sign interaction, the type of microstructure also changes. The existence of such a transition in nature means that all previous ideas about the solid solution and its decomposition, which is believed to occur because of a decrease in solubility with decreasing temperature, are completely untrue [4].

The “ordering-separation” phase transition in various binary alloys has certain characteristic features that manifest themselves at the level of microstructure changes: the sign of the energy of chemical interaction between dissimilar atoms changes for each system at a certain temperature (the temperature of the “ordered-separation” transition). The microstructure that existed before the transition, for example, because of the tendency of the alloy to ordering (solid solution with particles of chemical compounds), dissolves, and in its place a microstructure is formed, which occurs because of the tendency of the alloy to delaminate (solid solution with clusters or particles of the dissolved component). In different systems, such a change in sign (and microstructure) can occur at more than one temperature, and the morphology of each transition can change in different ways. When the energy of chemical interatomic interaction passes through zero during such a transition, it would be logical to assume that the microstructure of the alloy can be a homogeneous solid solution. Indeed, such a situation is observed in the phase diagram in alloys of the Fe-Cr system [4]. However, regions of a homogeneous solid solution between the regions of phase separation and regions of ordering were not found in all systems where such a transition occurred. For example, the authors [4], who studied alloys of the Ni-Co system by TEM, encountered a situation where, at the ordering-separation transition temperature, Co-clusters forming because of a tendency to phase separation, and particles of NiCo₂ chemical compounds forming because of the tendency towards ordering, were found in the same area of the foil.

The causes of attraction between dissimilar atoms, which lead to the formation of chemical compounds in alloys (tendency to ordering), and the causes of attraction between like atoms, which leads to the formation of clusters or particles of atoms of one of solutes (tendency to phase separation), have now found an adequate explanation [5] within the framework of the existing electronic theory. Assuming that a 100% metallic bond exists only in pure metals, while in alloys a certain part of the valence electrons takes part in the formation of other components of a strong chemical bond, the author of [5] analyzed what bonds exist in alloys between the atoms of the components. He concluded that ionic bonding occurs in alloys when one or more pairs of valence electrons, which should have taken part in the formation of an electron gas, are localized on two or more (depending on valence) nearest dissimilar atoms, which leads to formation of common orbitals. Here, the formation of chemical compounds in the alloy's structure occurs because of the attraction between the nearest dissimilar neighbors. A covalent bond forms when one or more pairs of valence electrons take part in the formation of hybridized orbitals between like atoms. Here, clusters or particles of atoms of the dissolved component form in the alloy's structure because of the attraction between the nearest neighbors of the same name. This means that the property of alloys to change the sign of a chemical bond with a change in the heating temperature follows from the very essence of the electronic structure of alloys [5].

We chose one binary alloy, Co₇₅V₂₅, and one ternary alloy, Ni₆₅Mo₂₀Cr₁₅, as objects for study. The first is to show what an ordering-phase separation transition is and how it will be displayed on modernized binary phase diagrams. The second is to explain what this new “diffusion micro-pair” structure is, which form in ternary alloys during their melting, and why its presence shows that us it is simply pointless to build ternary phase diagrams, since in alloys; there is only a pair interatomic chemical interaction. Alloys under study were water-quenched from the liquid state. Here, a small portion of liquid metal was poured from the crucible directly into water.

 After each low-temperature heat treatment, the alloys were also cooled in water in order to preserve this microstructure, which forms in the alloy at the interested by us temperature of tempering. The foils were made from the castings and the microstructure was studied on an EM-125 transmission electron microscope according to the standard method.

Binary Co₇₅V₂₅ Alloy

Our studies have shown that the quenching of the Co₇₅V₂₅ alloy [6] from the liquid state fixes particles of vanadium atoms, the bright-field image of which is shown in Figure 1. The electron diffraction pattern shows satellites appear near the fundamental reflections (Figure 1, inset). Obviously, these satellites arise from particles of a more refractory element - vanadium, which crystallizes in liquid solution because of the tendency to phase separation.

Figure 1: Alloy Co₇₅V₂₅. Quenching from the liquid state. Microstructure. Inset: electron diffraction pattern, zone axis [111]. V particles.

As the temperature of the alloy decreases to 1150°C, the absolute value of the chemical interaction energy increases. This can be judged from the size and number of particles of vanadium atoms formed at 1550 and 1150°C. Figure 2 shows that the size and number of such particles is greater than after quenching from 1550 °C. In the electron diffraction pattern obtained from such a colony, satellites from particles of vanadium atoms are also observed (Figure 2, inset).

When the heat treatment temperature is lowered to 800°C, a dissolving microstructure of phase separation is observed (dark particles of vanadium atoms) located between light grains of particles of the chemical compound Co3V (L12). Satellites are no longer observed in the electron diffraction pattern (Figures 1 and 2), and diffuse scattering is observed instead (inset to Figure 3). These facts indicate that at 800 C a phase transition, ordering-separation, occurs in the alloy.

Figure 3: Alloy Co₇₅V₂₅. Quenching from 800°C. Phase separation microstructure: bright field image; Inset: Micro-electron diffraction pattern shows the formation of the L1₂ phase takes place.

Exactly the same picture we observed earlier on alloys of the Ni-Co [7] system. At the same time, in alloys of a number of other systems (for example, Fe-Cr), this transition passes through the stage of formation of a disordered solid solution between the region in which separation takes place and in which ordering occurs.

A further decrease in the heat treatment temperature to 500°? leads to the fact that reflections from the L1₂ phase become more distinct and intense (Figure 4). The bright-field image shows highly dispersed particles of the chemical compound L1₂. If the heat treatment of the alloy is reduced to 350°C, then the type of the alloy structure does not change, but the morphology changes (compare Figures 4 and 5). It is not yet clear why this change in structure is connected. But this is not an ordering-separation phase transition, since the tendency to ordering at 350? C is preserved.

Figure 4: Alloy Co₇₅V₂₅. Aging at 550°C. Bright-field image of the microstructure and micro-electron diffraction pattern (inset).

Figure 5: Alloy Co₇₅V₂₅. Water quenching from 350°C. (a) Bright-field image of the phase-separation microstructure; (b) electron diffraction pattern, [001] zone axis.

Figure 6: Co-V phase diagram presented by the National Physical Laboratory (USA). Dashed horizontal lines at 800? C show temperature of ‘ordering – separation’ phase transition, at 450? C – temperature of morphology change.

Figure 6 shows a part of the phase diagram of Co-V, on which a line is plotted, corresponding in composition to the Co75V25 alloy. We can see it from the graph that the coincidence of the experimental results with what they showed in the diagram takes place only in 2 points – 1 and 2.

Ternary Ni₆₅Mo₂₀Cr₁₅ Alloy

In this article, we would like to touch upon one more point that has not yet found an answer from researchers - this is the question of multi-component phase diagrams.

It became possible to answer this question recently, when such microstructural elements as diffusion micro-pairs were discovered in ternary alloys. Previously, ternary alloys were considered as ordinary alloys, in which interatomic interactions proceed in the same way as they did in binary ones. However, our discovery of the ordering-separation phase transition, together with the existence of the concept of pair-wise chemical interatomic interaction, showed that this is not at all the case. Therefore, in the presented article, we will show the reader in a very brief form and on one alloy how the process of formation of new phases in binary and ternary alloys differs and why the formation of new phases in these alloys cannot be approached from the common positions.

At temperatures of the liquid state in the Ni₆₅Mo₂₀Cr₁₅ ternary alloy [8], there is a tendency for all three pairs of its components to separation. Investigating the alloy by TEM after quenching from the liquid state (1550°C), we found different phases can form in the foil under study in different areas. This means that the alloy already in the liquid state is divided into sections that differ from each other in composition. Since the solvent occupies almost 2/3 of the volume of the alloy under study, it can only decompose into micro-pairs of two types - enriched in Mo and enriched in Cr, while the Mo/Cr micro-pair doesn’t form. After quenching from 1550°C, clusters of chromium atoms are observed in the form of concentration waves emanating from some light-colored centers in a thin foil (Figure 5a).

The Ni₆₅Mo₂₀Cr₁₅ [8] alloy was chosen for this study. At liquid state temperatures, there is a tendency for all three of its components to phase separation. Investigating the alloy by TEM, we encountered the fact: in some areas of the foil, we find some phases, in other areas of the same foil, others. This immediately suggests that the alloy, already in the liquid state, there is a separation into certain sections that differ from each other in composition. Since the solvent in the alloy under study Ni occupies almost 2/3 of the volume of the alloy, such a composition can break down into only two types of micro-pairs - enriched in Mo and enriched in Cr while the micro-pair Mo/Cr is not formed. After quenching from 1550°C, clusters of chromium atoms are observed as concentration waves of the absorption contrast, which emanate from a certain light colored centers in a thin foil (Figure 7a).

Figure 7: Ni₆₅Mo₂₀Cr₁₅ alloy. Ni/Cr diffusion pair. Microstructure. Quenching from 1550°C (a), 1500°C (b). Absorption contrast.

When the quenching temperature is lowered to 1500°C (Figure 7b), the entire surface area of the foil in the Ni/Cr diffusion pair turns out to be dissected by such contrast waves as in Figure 7b. Such concentration inhomogeneities can form only because of the tendency to phase separation. The size of the region occupied by the Ni/Cr diffusion pair can be estimated from the number of such contrast waves that fit into such a region. This value approximately reaches several tens of microns.

Figure 8: Ni₆₅Mo₂₀Cr₁₅ alloy. Quenching in water from 1300 ° C. Absorption contrast. Diffusion micro-pair Ni/Cr.

 Figure 8 shows the microstructure of the clusters, which are found inside the Ni/Cr diffusion micro-pair [8] after the alloy has been quenched from 1300°C. It can be seen that both in the liquid (Figure 7b) and in the solid state (Figure 8), the microstructure has almost the same morphology. With a further decrease in the quenching temperature of the alloy, the “ordering-phase separation” phase transition occurs in diffusion micro-pairs. In other areas of the same foil of the Ni₆₅Mo₂₀Cr₁₅ alloy, accumulation of rounded particles are observed (Figure 9a). The electron diffraction pattern in Figure 9b shows that these are particles of Mo atoms, i.e., it is a microstructure of the diffusion micro-pair Ni/Mo. At 1200°C, when the “ordering-phase separation” transition occurs in Ni/Mo diffusion micro-pairs, Mo atom particles dissolve and take part in the formation of Ni₃Mo(D0₂₂) and Ni₄Mo(D1_a) phases.

Figure 9: Ni₆₅Mo₂₀Cr₁₅ alloy. Ni/Mo diffusion micro-pair. Quenching in water from 1500°C: (a) Microstructure (Mo-particles); (b) micro-electron diffraction pattern.

The presented data allows us to conclude:

Because chemical bonds in alloys can only exist between two neighboring atoms and nothing else, any compounds in ternary (or more) alloys can be formed only in two stages. The first stage is the formation of clusters of atoms of one or another component by "up-hill" diffusion. We called these clusters “diffusion micro-pairs”. The second stage is the formation of particles of a new phase inside the diffusion micro-pair. Sometimes (liquid state, higher melting point of some dissolved component), these two stages proceed at a time.

The views existing in metal science that with an increase in the heating temperature of alloys, the diffusion of atoms speeds up and becomes chaotic, and in the process chemical interatomic interactions disappear, is a profound delusion. Chemical bonds exist in alloys as long as the alloys themselves with their electronic structure exist, i.e. during the entire time of their condensed state. These bonds cannot disappear, just as electronic structure cannot disappear in alloys.

As is known, all fundamental properties of alloys are determined by the chemical bond between the nearest neighbors. It is believed that such a model is sufficiently correct for binary alloys only. The “ordering-separation” transition that we show in this article using the example of the Co₇₅V₂₅ alloy testifies to that all diffusion structural and phase transformations that occur in binary alloys with a change in temperature owe their origin to chemical interatomic interactions that exist in alloys at all temperatures of the condensed state. What does this mean for existing phase diagrams built using experimental data got by X-ray diffraction analysis? This means that such sections of the phase diagrams must be reconstructed using the TEM method. One person cannot do the work that is happening now. This work needs to be done by the entire community of metallurgists. It is necessary that all "white spots" (as we now call the areas of solid solutions in existing phase diagrams) will disappear, because only chemical interatomic interactions are the major cause of all changes in the structure of binary and ternary alloys.

The “ordering-separation” phase transition, found in alloys of 17 binary systems, led to the conclusion that all structural changes in metal alloys occur because of chemical interatomic interactions. The discovery of such chemical bonds in the liquid state of alloys led to the conclusion that a chemical bond exists in alloys at temperatures of the condensed state. These two discoveries mean that the binary phase diagrams, which were built based on the notion that at high temperatures, the chemical interaction in alloys disappears and the structure is a disordered solid solution, are incorrect. The last two ideas arose because X-ray phase analysis was previously used in Metals Science to study the microstructure, which does not allow one to detect highly dispersed particles of new phases after high-temperature quenching.

In the current situation in Metal Science, there are two ways out:

• Leave binary phase diagrams with no changes, informing users that only lines of massive phase transformations are correct in these diagrams. The areas of solid solutions and the areas in which the microstructure formed thanks to up-hill diffusion of the solutes should be considered as not inspiring confidence.
• To put on the phase diagrams the lines of the phase transition "ordering - separation", using the method of transmission electron microscopy. Before this, it is necessary to remove all regions of the solid solution, as well as those regions in which the microstructure formed because of the diffusion of atoms.

The article raises the question of what to do with the main achievement of Metal S?ience - binary phase diagrams. This question arose because our inconsistency in one of the main areas of Materials Science - the science of metal alloys – was revealed. In these areas, phase diagrams play a key role and their reliability must be complete. However, the discovery of the “ordering-separation” phase transition, which shows that chemical interatomic interactions are realized in the temperature range of the condensed state of alloys, forces us to reconsider all ideas about metal alloys. The author does not give an unequivocal answer to this question, believing that the entire community of metallurgists should do this. However, he calls on the community to analyze the current situation and decide.

1. Higgins J, Nicholson RB, Wilkes P. Precipitation in iron-beryllium system. Acta metallurgica. 1974; 22: 201-217.
2. Laughlin DE, Cahn JW. Spinodal decomposition in age-hardening Cu-Ti alloys. Acta metallurgica. 1975; 23: 329-339.
3. Ustinovshikov Y, Shirobokova M, Pushkarev B. A structural study of the Fe-Cr system alloys. Acta Materialia. 1996; 44: 5021-5032.
4. Ustinovshikov Y. The “Ordering – Phase Separation” Transition in Alloys, Monograph 2019. Cambridge Scholar Publishing. Great Britain.
5. Ustinovshikov Y. Changes in electronic structure of metallic alloys at the Transition ‘Ordering-Phase Separation’. J Alloys and Compounds. 2017; 714: 476-483.
6. Ustinovshikov Y. Phase transition ‘Ordering-Phase Separation’ in the Co₃V alloy. J Alloys Compounds. 2015; 639: 669-674.
7. Ustinovshikov Y, Shabanova I, Lomova N. TEM study of the 'ordering-phase separation' transition in Ni-Co alloys. J Advanced Microscopy Researches. 2013; 8: 276-282.
8. Ustinovshikov Y. Phase and structural transformations in the Ni₆₅Mo₂₀Cr₁₅ alloy at changing the temperature of heat treatment. J Alloys Compounds. 2014; 588: 470-473.