Thermally-Quenched Metastable Phases in Correlated Electron Systems

A large number of electrons that correlate with each other in crystals form ordered phases, such as magnetic orders and superconductivity, analogously to correlated atoms forming crystals. Whereas the most stable phase is uniquely determined once thermodynamic parameters are fixed, it has long been known that metastable crystals can be obtained when their transformation to the most stable phase is kinetically avoided by thermal quenching. It is intriguing whether the principle of quenching can also be applied to correlated electron systems. However, the cooling rates used in measurements of electronic properties are usually limited to a range between 10^-3-10^-1 Ks^-1, and thus the ubiquity of this principle had not been tested. We have been investigating metastability in correlated electron systems with rapid cooling techniques that produce a cooling rate up to 10⁸ Ks^-1.<br/><br/>Here, we present two cases, metastable magnetism in a chiral magnet MnSi [H. Oike et al., Nature Phys. 12, 62 (2016).] and metastable superconductivity in a transition-metal dichalcogenide IrTe₂ [H. Oike et al., Science Adv. 4, eaau3489 (2018).]. MnSi is a long-studied magnet without inversion symmetry, and, in the last decade, has been attracting a great deal of attention as a material hosting magnetic skyrmions [Y. Tokura and N. Kanazawa, Chem. Rev. 121, 2857 (2020).]. Two kinds of interactions are acting between electron spins in MnSi; ferromagnetic interaction that aligns neighboring spins in the same direction with each other and Dzyaloshinskii-Moriya interaction that twists neighboring spins. The competition between these interactions underlies a variety of spin structures in MnSi, such as spiral phase, where spins are aligned in a plane and modulated in a direction perpendicular to the plane, and skyrmion lattice phase (SkL), where vortex-like spin swirling objects form a lattice. We applied thermal quenching to the phase transition from SkL to spiral phase and found that SkL exists as a metastable state in a temperature range where a spiral phase is free-energetically most stable. Thus, the principle of quenching is demonstrated to work for magnetism.<br/><br/>Then, the principle is deployed to creation of metastable superconductivity. In correlated electron systems, a superconducting phase often has a slightly higher free-energy than other ordered phases of electrons. Therefore, if the formation of such competing orders is kinetically avoided by quenching, a superconducting state is expected to emerge. The target material IrTe₂ undergoes a complex ordering phenomena where the valence of the two-fifths Ir ions changes from trivalent to tetravalent and accordingly the network structure formed by Te ions deviating from divalent also changes [K.-T. Ko et al., Nature Commun. 6, 7342 (2015).]. We applied thermal quenching to the complex ordering and found that a phase without the ordering exists as a metastable state at low temperatures. As we expected, the metastable phase exhibits superconductivity highlighting that the principle of quenching opens up new frontiers in the exploration of quantum phenomena.<br/><br/>Thus, the ubiquity of the principle is being demonstrated in correlated electron systems. Because such quenching experiments have only been carried out on a limited number of materials, it is expected that new quantum phases will be pioneered when they are applied to a wide variety of materials. More importantly, there may be similarities between correlated atoms and correlated electrons in the mechanism of metastability, as well as similarities in the methods of creating metastable phases. I hope that general concepts would be found through this presentation and subsequent discussions.