We study the physics of quantum phase transitions from the perspective of non-equilibrium thermodynamics. For first order quantum phase transitions, we find that the average work done per quench in crossing the critical point is discontinuous. This leads us to introduce the quantum latent work in analogy with the classical latent heat of first order classical phase transitions. For second order quantum phase transitions the irreversible work is closely related to the fidelity susceptibility for weak sudden quenches of the system Hamiltonian. We demonstrate our ideas with numerical simulations of first, second, and infinite order phase transitions in various spin chain models.

No abstract

The search for two-dimensional (2D) systems for applications in solar cells has continuously challenged our community. Here, we report screening of 2D monolayers from group IV to XI transition-metal dichalcogenides, MQ2, searching for candidates for high-performance photovoltaic devices, where M = 3d, 4d, and 5d transition metals and Q = S, Se, and Te, i.e., it yields 72 MQ2 monolayers. Through a robust theoretical framework that combines density functional theory, tight binding based on maximum localized Wannier functions, and the Bethe–Salpeter equation, we investigate the electronic, optical, and excitonic properties of the thermodynamic stable 2H-MQ2 monolayers. Furthermore, we employ a linear regression analysis of our data to identify nontrivial correlations between different optical and excitonic properties. From the 72 2H-MQ2 monolayers, we found from phonon calculations that only 22 are dynamically stable in the 2H form, of which 14 are semiconductors with a wide range of energy band gaps. Our results show that the presence of excitons affects the band alignment and the power conversion efficiency. In these compounds, the valley degeneracy breaking driven by spin-orbit coupling and the excitonic correction in the optical properties may be explored, through control of optical helicity, to tune the efficiency of the MQ2-based photovoltaic devices. We find high-efficiency van der Waals (vdW) heterostructures for solar cells and observe a strong linear correlation between exciton energy and the energy band gap for the stable semiconductors.

We study the relationship between the pseudogap and Fermi-surface topology in the two-dimensional Hubbard model by means of the cellular dynamical mean-field theory. We find two possible mean-field metallic solutions on a broad range of interaction, doping and frustration: a conventional renormalized metal and an unconventional pseudogap metal. At half-filling, the conventional metal is more stable and displays an interactiondriven Mott metal-insulator transition. However, for large interaction and small doping, region that is relevant for cuprates, the pseudogap phase becomes the ground state. By increasing doping, we show that a first-order transition from the pseudogap to the conventional metal is tight to a change of the Fermi surface from hole to electron like, unveiling a correlation-driven mechanism for a Lifshitz transition. This explains the puzzling link between pseudogap phase and Fermi surface topology which has been pointed out in recent experiments.In order to understand superconductivity [1-3] one must first understand the normal metallic state, appearing above a critical temperature (T c ), from which it takes its roots. The high-T c superconductivity in cuprates remains unsolved, mainly because its normal metallic state, the pseudogap (PG) phase, has not been well understood. It has been therefore a central issue to establish the origin of the pseudogap and its relation with the high-T c superconducting mechanism [4]. The PG has been revealed [5, 6] Here we give a rational explanation to all these observations within the framework of the two-dimensional Hubbard model solved with the cellular dynamical mean field theory (CDMFT) [15][16][17]. We first show that two metallic solutions exist: a rather regular correlated Fermi-liquid metal (CFM), and a PG metal (PGM), which violates Fermi liquid theory, by developing a pole-divergence in the self-energy. This result could account for contradicting reports about the existence of the Mott metal-insulator transition (MIT) at half-filling (zero doping) in two dimensions. The PGM is metastable at weak interactions, having higher energy than the CFM. However, by increasing interaction at low doping (region relevant for underdoped cuprates) the PGM emerges as the stable phase, up to the doping value p * . This is consistent with the CDMFT results of Sordi et al. [18]. Most importantly, we show that the PGM is bound to have always a h-FS. The CFM instead can undergo a Lifshitz transition at a doping p lt . However, for strong interaction the CFM is stable only for doping p > p lt , i.e. it has always an e-FS. Hence the transition from the PGM to the CFM at p * is accompanied by a corresponding change from a h-FS to an e-FS, unveiling a novel correlationdriven mechanism of the Lifshitz transition. These results explain why the PG must sharply end when a Lifshitz transition occurs[11], or is tuned by pressure [14], in cuprates.We consider the two-dimensional one-band Hubbard model:wheredestroys an electron with spin σ and momentum k, n iσ = c † iσ c iσ is th...