Advances in light sources and time resolved spectroscopy have made it possible to excite specific atomic vibrations in solids and to observe the resulting changes in electronic properties but the mechanism by which phonon excitation causes qualitative changes in electronic properties, has remained unclear. Here we show that the dominant symmetry-allowed coupling between electron density and dipole active modes implies an electron density-dependent squeezing of the phonon state which provides an attractive contribution to the electron-electron interaction, independent of the sign of the bare electron-phonon coupling and with a magnitude proportional to the degree of laser-induced phonon excitation. Reasonable excitation amplitudes lead to non-negligible attractive interactions that may cause significant transient changes in electronic properties including superconductivity. The mechanism is generically applicable to a wide range of systems, offering a promising route to manipulating and controlling electronic phase behavior in novel materials. * These two authors contributed equally 1 arXiv:1609.03802v2 [cond-mat.supr-con] May 2017Strong mode-specific excitation of specific atomic vibrations (phonon modes) in solids [1] has been shown [2-6] to drive drastic changes in collective electronic properties. Of the many effects observed, perhaps the most dramatic is the superconducting-like behavior observed at temperatures far above the equilibrium transition temperature in strongly irradiated YBa 2 Cu 3 O 6+x [7] and K 3 C 60 [2]. Because optically addressable phonon modes are dipole active (odd parity) zone center vibrations which typically do not couple linearly to local electronic quantities such as the density or orbital occupancy, the mechanism by which phonon excitation changes electronic properties has not been clear, although many interesting proposals have been made [9][10][11][12][13][14]. Here we present a new and general mechanism that naturally explains how optical phonon excitation can lead to changes in electronic properties. The key idea is that the dominant symmetry-allowed coupling between electron density and dipole active modes is quadratic, implying an electron density-dependent squeezing of the phonon state. We show that this electron density dependence of the phonon squeezing provides a sizable attractive contribution to the electron-electron interaction. The attractive contribution is independent of the sign of the bare electron-phonon coupling, has a magnitude proportional to the degree of laser-induced phonon excitation, and is large enough that reasonable values of the excitation amplitude lead to non-negligible attractive interactions.The mechanism is generically applicable to a wide range of systems, offering a promising route to manipulating and controlling electronic phase behavior.To demonstrate the mechanism we consider a minimal tight-binding model of electrons that can hop between sites of a lattice, are subject to electronic interactions, and are coupled quadratically to ...

The existence of more than one steady-state in a many-body quantum system driven out-ofequilibrium has been a matter of debate, both in the context of simple impurity models and in the case of inelastic tunneling channels. In this paper, we combine a reduced density matrix formalism with the multilayer multiconfiguration time-dependent Hartree method to address this problem. This allows us to obtain a converged numerical solution of the nonequilibrium dynamics. Considering a generic model for quantum transport through a quantum dot with electron-phonon interaction, we prove that a unique steady-state exists regardless of the initial electronic preparation of the quantum dot, consistent with the converged numerical results. However, a bistability can be observed for different initial phonon preparations. The effects of the phonon frequency and strength of the electron-phonon couplings on the nonequilibrium dynamics and on the emergence of bistability is discussed.

The nonequilibrium dynamics of a quantum dot with electron-phonon interactions described by a generalized Holstein model is presented. A combination of methodologies including the reduced density matrix formalism, the multilayer multiconfiguration time-dependent Hartree method, and a time-dependent nonequilibrium Green function approach, is used to explore the transient behavior on multiple timescales as the system approaches steady-state. The dot population dynamics on short to intermediate times is governed by the dot-lead hybridization parameter (Γ) and by the typical phonon frequency (ωc) and depends on the location of the energy level of the dot relative to the bias window. At longer times, the dynamics show a distinct behavior depending on whether the system is in the adiabatic or non-adiabatic regime, with a quantum dot occupation that may depend on the initial preparation of the phonons degrees of freedom. A "phase" diagram of this localization effect as a function of the polaron shift (λ) for various phonon frequencies is derived, suggesting the existence of bistability on experimentally observable timescales.

The reduced dynamics formalism has recently emerged as a powerful tool to study the dynamics of non-equilibrium quantum impurity models in strongly correlated regimes. Examples include the non-equilibrium Anderson impurity model near the Kondo crossover temperature and the nonequilibrium Holstein model, for which the formalism provides an accurate description of the reduced density matrix of the system for a wide range of timescales. In this work, we generalize the formalism to allow for non-system observables such as the current between the impurity and leads. We show that the equation of motion for the reduced observable of interest can be closed with the equation of motion for the reduced density matrix and demonstrate the new formalism for a generic resonant level model. Gesellschaft is currently under investigation [4], and basic questions regarding hysteresis and bi-stability in systems governed by strong electron-phonon couplings remain under debate [5][6][7][8][9][10]. Notably, many successful approaches to problems of this kind are based on master equation treatments and cumulant expansions [11][12][13][14][15], or on diagrammatic partial summations [16], all of which are approximate in general. A major theoretical challenge lies in the need to provide an accurate account of time propagation of open quantum systems, starting from some known initial state and proceeding all the way to an unknown steady-state.Numerically exact methods play a particularly important role in the quest to obtain a reliable, unbiased description of non-equilibrium phenomena. Several different types of bruteforce approaches developed in recent years have been applied to open non-equilibrium quantum systems. These include the time-dependent numerical renormalization group [17] and functional renormalization group [18][19][20], time-dependent density matrix renormalization group [21][22][23][24], iterative [25][26][27][28] and stochastic [29][30][31][32][33] diagrammatic methods and wavefunction based approaches [34,35]. While the application of these approaches to the non-equilibrium Holstein, the Anderson impurity and the spin-fermion models has been very fruitful, they are still restricted to a relatively small range of parameters, typically characterized by a rapid decay to steady-state. Situations or observables exhibiting slow dynamics are inaccessible by these brute-force methods.An alternative approach recently proposed by Cohen and Rabani [36] is based on a combination of a brute-force impurity solver (one of the above) with a generalized quantum master equation (GQME). The Nakajima-Zwanzig-Mori [37-39] formalism was used to derive an exact equation of motion for the reduced density matrix of the system, which includes a memory kernel giving rise to non-Markovian effects. This kernel, along with some information regarding the initial conditions, determines the dynamics of the system and contains all New Journal of Physics 15 (2013) 073018 (http://www.njp.org/)

We develop a new fermionic path-integral formalism to analyze the phase diagram of open nonequilibrium systems. The formalism is applied to analyze an ensemble of two-level atoms interacting with a single-mode optical cavity, described by the Dicke model. While this model is often used as the paradigmatic example of a phase transition in driven-dissipative systems, earlier theoretical studies were limited to the special case when the total spin of the atomic ensemble is conserved. This assumption is not justified in most experimental realizations. Our new approach allows us to analyze the problem in a more general case, including the experimentally relevant case of dissipative processes that act on each atom individually and do not conserve the total spin. We obtain a general expression for the position of the transition, which contains as special cases the two previously known regimes: i) non-equilibrium systems with losses and conserved spin and ii) closed systems in thermal equilibrium and with the Gibbs ensemble averaging over the values of the total spin. We perform a detailed study of different types of baths and point out the possibility of a surprising non-monotonous dependence of the transition on the baths' parameters.Comment: 6 pages. Comments welcom