No Arabic abstract
We study the orbital diamagnetic susceptibility in excitonic condensation phase using the meanfield approximation for a two-band model defined on a square lattice. We find that, in semiconductors, the excitonic condensation acquires a finite diamagnetic susceptibility due to spontaneous hybridization between the valence and the conduction bands, whereas in semimetals, the diamagnetic susceptibility in the normal phase is suppressed by the excitonic condensation. We also study the orbital diamagnetic and Pauli paramagnetic susceptibilities of Ta2NiSe5 using a two-dimensional three-band model and find that the calculated temperature dependence of the magnetic susceptibility is in qualitative agreement with experiment.
The idea of exciton condensation in solids was introduced in 1960s with the analogy to superconductivity in mind. While exciton supercurrents have been realized only in artificial quantum-well structures so far, the application of the concept of excitonic condensation to bulk solids leads to a rich spectrum of thermodynamic phases with diverse physical properties. In this review we discuss recent developments in the theory of exciton condensation in systems described by Hubbard-type models. In particular, we focus on the connections to their various strong-coupling limits that have been studied in other contexts, e.g., cold atoms physics. One of our goals is to provide a dictionary which would allow the reader to efficiently combine results obtained in these different fields.
We show that finite temperature variational cluster approximation (VCA) calculations on an extended Falicov-Kimball model can reproduce angle-resolved photoemission spectroscopy (ARPES) results on Ta2NiSe5 across a semiconductor-to-semiconductor structural phase transition at 325 K. We demonstrate that the characteristic temperature dependence of the flat-top valence band observed by ARPES is reproduced by the VCA calculation on the realistic model for an excitonic insulator only when the strong excitonic fluctuation is taken into account. The present calculations indicate that Ta2NiSe5 falls in the Bose-Einstein condensation regime of the excitonic insulator state.
We consider a two-orbital Hubbard model with Hund coupling and crystal-field splitting and show that in the vicinity of the high-spin/low-spin transition, crystal-field quenches can induce an excitonic condensation at initial temperatures above the highest ordering temperature in equilibrium. This condensation is the effect of an increase in the spin entropy and an associated cooling of the effective electronic temperature. We identify a dynamical phase transition and show that such quenches can result in long-lived nonthermal excitonic condensates which have no analogue in the equilibrium phase diagram. The results are interpreted by means of an effective pseudo-spin model.
We study the excitonic insulating (EI) phase in the two-band Hubbard models on the Penrose tiling. Performing the real-space mean-field calculations systematically, we obtain the ground state phase diagrams for the vertex and center models. We find that, in some regimes, the stable EI phase is induced by small interband interactions. We argue that this originates from the electron-hole pairing for the completely or nearly degenerate states, which are characteristic of the Penrose tiling. We also study spatial distribution of the order parameter, mapping it to the perpendicular space.
The microscopic quantum interference associated with excitonic condensation in Ta$_2$NiSe$_5$ is studied in the BCS-type mean-field approximation. We show that in ultrasonic attenuation the coherence peak appears just below the transition temperature $T_{rm c}$ whereas in NMR spin-lattice relaxation the rate rapidly decreases below $T_{rm c}$; these observations can offer a crucial experimental test for the validity of the excitonic condensation scenario in Ta$_2$NiSe$_5$. We also show that the excitonic condensation manifests itself in a jump of the heat capacity at $T_{rm c}$ as well as in softening of the elastic shear constant, in accordance with the second-order phase transition observed in Ta$_2$NiSe$_5$.