Correlated quantum-chemical methods for condensed matter systems, such as the random phase approximation (RPA), hold the promise of reaching a level of accuracy much higher than that of conventional density functional theory approaches. However, the high computational cost of such methods hinders their broad applicability, in particular for finite-temperature molecular dynamics simulations. We propose a method that couples machine learning techniques with thermodynamic perturbation theory to estimate finite-temperature properties using correlated approximations. We apply this approach to compute the enthalpies of adsorption in zeolites and show that reliable estimates can be obtained by training a machine learning model with as few as 10 RPA energies. This approach paves the way to the broader use of computationally expensive quantum-chemical methods to predict the finite-temperature properties of condensed matter systems.

The electronic structural properties in the presence of constrained magnetization and a charged background are studied for a monolayer of FeSe in non-magnetic, checkerboard-, and stripedantiferromagnetic (AFM) spin configurations. First principles techniques based on the pseudopotential density functional approach and the local spin density approximation are utilized. Our findings show that the experimentally observed shape of the Fermi surface is best described by the checkerboard AFM spin pattern. To explore the underlying pairing mechanism, we study the evolution of the non-magnetic to the AFM-ordered structures under constrained magnetization. We estimate the strength of electronic coupling to magnetic excitations involving an increase in local moment and, separately, a partial moment transfer from one Fe atom to another. We also show that the charge doping in the FeSe can lead to an increase in the density of states at the Fermi level and possibly produce higher superconducting transition temperatures.Recent experimental advances in molecular beam epitaxy and scanning tunneling microscopy have made it possible to study superconducting monolayer systems, such as FeSe, which is the simplest iron-based superconductor. Studies of FeSe monolayer systems on different substrates show significant sensitivity to interface effects and give signs of the presence of superconductivity above 77 K [1,2]. The latter fact is especially interesting because bulk samples only show superconducting transition temperatures T c of about 8 K, or 37 K with the application of pressure [3,4]. When double-layer graphene is used as a substrate, the superconducting gap seems to increase with the FeSe film thickness, and a monolayerthick film was found to be non-superconductive [2]. At the same time, when FeSe films are studied on SrTiO 3 , scanning tunneling microscopy reveals the presence of a 20 meV gap for single unit cell thick films, whereas thicker films show no presence of superconductivity [1].Another difference between the bulk and monolayer FeSe is the Fermi surface topology: ARPES measurements show distinctive shape where only pockets at the Brillouin zone (BZ) boundaries are present, and no pockets are seen at the center point Γ [5]. The same study shows the presence of large and nearly isotropic superconducting gaps without nodes. Another very recent report by the same research group [6] describes the presence of two phases with different Fermi surface topology, where a superconducting phase can be obtained from a nonsuperconducting one through annealing. Both reports give evidences for substantial interface-induced changes in electronic structure. Suppression of superconductivity by twin boundaries interconnected with the Se-atom height with respect to the Fe layer was also found [7].Several theoretical efforts were made to study thin films of FeSe [8,9]. First-principles study of atomic and electronic structures of one-and two-monolayer thick films on SrTiO 3 found semiconductor-like behavior and collinear antiferromagneti...

Using a first-principles pseudopotential approach we study the origin of superconductivity in lithium under pressure. A recently developed Wannier interpolation based technique that allows for ultradense sampling of electron-phonon parameters throughout the Brillouin zone was employed. The electron-phonon coupling strength as a function of pressure was calculated, precisely resolving many of the fine features of its distribution. The contributions to coupling arising from the Fermi surface topology, phonon dispersions, and electronphonon matrix elements were separately analyzed. It is found that of the constituent components, the electronphonon matrix elements are the most sensitive to pressure changes, and a particular phonon is responsible for high values of coupling. Additionally, the distribution of matrix elements over the Fermi surface is seen to be non-uniform and possesses a two-peak structure. Analysis of the Eliashberg spectral function ␣ 2 F͑͒ shows a considerable increase in spectral weight in the low-frequency region with the application of pressure. We estimate the superconducting transition temperature and find that the obtained values are in good accord with experiment.

We present first-principles calculations for quasiparticle excitations in sodium and lithium including the effects of charge and spin fluctuations. We employ the Overhauser-Kukkonen form for the electron self energy arising from spin fluctuations and demonstrate that the coupling of electrons to spin fluctuations gives an important contribution to the quasiparticle lifetime, but does not significantly reduce the occupied bandwidth. Including correlation effects beyond the random-phase approximation in the screening from charge fluctuations yields good agreement with experiment. PACS numbers: 71.20.Dg, 71.10.Ca, 71.15.Qe, 71.45.Gm Introduction.-The coupling of electrons to spin fluctuations causes many fascinating phenomena: for example, it has been proposed that spin fluctuations can "glue" electrons together to form Cooper pairs giving rise to unconventional high-temperature superconductivity [1][2][3][4]. In particular, spin fluctuations were invoked to explain superconductivity in the cuprates [1][2][3] and recently also in the iron pnictide and chalcogenide materials [5][6][7]. In addition, it is well known that the coupling of spin fluctuations to electrons can affect the electronic effective mass and consequently transport properties and the specific heat.Theoretically, the effect of spin fluctuations on quasiparticle excitations is usually calculated using model Hamiltonians. Early studies [8-10] constructed empirical theories including spin fluctuations based on the homogeneous electron gas and simple tight-binding models. More recently, many empirical theories involving spin fluctuations were constructed to investigate superconductivity in the cuprates and pnictides. In these theories the spin susceptibility is either parametrized using experimental neutron scattering and nuclear magnetic resonance data [1,3,11] or estimated by combining density-functional theory (DFT) with interaction parameters (such as the Hubbard U ) adjusted to reproduce experimental findings [6,7].While the aforementioned theories have been very instructive, their applications have been limited by the availability of concrete experimental data needed to determine their input parameters, supporting the need for a fully first-principles theory without empirical parameters. There have been several attempts to compute the spin fluctuation-electron coupling from first principles. Notably, Winter and coworkers [12,13] calculated the spin susceptibility and the spin fluctuation-electron self energy from DFT and evaluated the correction to the specific heat for palladium and vanadium. Later studies [14-16] employed a first-principles T-matrix approach to calculate satellites in the photoemission spectrum of nickel