Upper bounds on superconducting and excitonic phase-stiffness for interacting isolated narrow bands

Abstract

Inspired by the discovery of superconductivity in moir\'e materials with isolated narrow bandwidth electronic bands, here we analyze critically the question of what is the maximum attainable $T_c$ in interacting flat-band systems. We focus specifically on the low-energy effective theory, where the density-density interactions are projected to the set of partially-filled flat bands. The resulting problem is inherently non-perturbative, where the standard mean-field approximation is not applicable. Here we develop further our recent Schrieffer-Wolff transformation based approach (PNAS, 120 (11), e2217816120 (2023)) to compute the effective electromagnetic response and the superconducting phase-stiffness in terms of "projected" gauge-transformations, and extend the formalism to compute the stiffness for excitonic superfluids. Importantly, our method requires neither any "wannierization" for the narrow bands of interest, regardless of their (non-)topological character, nor any knowledge of an underlying pairing-symmetry, and can be setup directly in momentum-space. We use this formalism to derive upper bounds on the phase-stiffness for sign-problem-free models, where their values are known independently from numerically exact quantum Monte-Carlo computations. We also illustrate the analytical structure of these bounds for the superconducting and excitonic phase-stiffness for perfectly flat-bands that have Landau-level-like wavefunctions.