Calculations of Spin Fluctuation Spectral Functions $\alpha^{2}F$ in High-Temperature Superconducting Cuprates

TL;DR

This paper develops a first-principles method combining density functional theory and FLEX-RPA to compute spin fluctuation spectral functions in cuprates, elucidating their role in high-temperature superconductivity.

Contribution

It introduces a novel computational approach to derive material-specific spin fluctuation spectral functions and predicts the superconducting gap symmetry in cuprates.

Findings

Spectral functions show a peak near 40-60 meV and decay at higher frequencies.

Superconducting gap symmetry is $d_{x^{2}-y^{2}}$ in all studied cuprates.

Reproduces $T_c$ range by varying Coulomb repulsion $U$, sensitive to spin density wave proximity.

Abstract

Spin fluctuations have been proposed as a key mechanism for mediating superconductivity, particularly in high-temperature superconducting cuprates, where conventional electron-phonon interactions alone cannot account for the observed critical temperatures. Traditionally, their role has been analyzed through tight-binding based model Hamiltonians. In this work we present a method that combines density functional theory with a momentum- and frequency-dependent pairing interaction derived from the Fluctuation Exchange (FLEX) type Random Phase Approximation (FLEX-RPA) to compute Eliashberg spectral functions $\alpha ^{2}F(\omega )$ which are central to spin fluctuation theory of superconductivity. We apply our numerical procedure to study a series of cuprates where our extracted material specific $\alpha ^{2}F(\omega )$ are found to exhibit remarkable similarities characterized by a sharp peak in the vicinity of 40-60 meV and their rapid decay at higher frequencies. Our exact diagonalization of a linearized BCS gap equation extracts superconducting energy gap functions for realistic Fermi surfaces of the cuprates and predicts their symmetry to be $d_{x^{2}-y^{2}}$ in all studied systems. Via a variation of on-site Coulomb repulsion $U$ for the copper $d$-electrons we show that that the range of the experimental values of $T_{c}$ can be reproduced in this approach but is extremely sensitive to the proximity of the spin density wave instability. These data highlight challenges in building first-principle theories of high temperature superconductivity but offer new insights beyond previous treatments, such as the confirmation of the usability of approximate BCS-like $T_{c}$ equations, together with the evaluations of the material specific coupling constant $\lambda $ without reliance on tight-binding approximations of their electronic structures.