Drude weight in hard core Boson systems: possibility of a finite temperature ideal conductor

TL;DR

This study investigates the Drude weight in hard core boson systems, revealing that in two dimensions, the normal phase can exhibit dissipationless transport with a finite Drude weight above the superfluid transition temperature.

Contribution

It demonstrates the persistence of a finite Drude weight above the superfluid transition in 2D systems, indicating non-dissipative transport in the normal phase, and explores anisotropic conduction in striped supersolid phases.

Findings

Drude weight remains finite above $T_{KT}$ in 2D superfluid phase.

In 3D, Drude weight vanishes at the superfluid transition.

Anisotropic conduction in striped supersolid phase above $T_{KT}$.

Abstract

We calculate the Drude weight in the superfluid (SF) and the supersolid (SS) phases of hard core boson (HCB) model using stochastic series expansion (SSE). We demonstrate from our numerical calculations that the normal phase of HCBs in two dimensions can be an ideal conductor with dissipationless transport. In two dimensions, when the ground state is a SF, the superfluid stiffness drops to zero with a Kosterlitz-Thouless type transition at $T_{KT}$. The Drude weight, though is equal to the stiffness below $T_{KT}$, surprisingly stays finite even for temperatures above $T_{KT}$ indicating the non-dissipative transport in the normal state of this system. In contrast to this in a three dimensional SF phase, where the superfluid stiffness goes to zero continuously with a second order phase transition at $T_c$, Drude weight also goes to zero at $T_c$ as expected. We also calculated the Drude weight in a 2D SS phase, where the charge density wave (CDW) order coexists with superfluidity. For the SS phase we studied, superfluidity is lost via Kosterlitz-Thouless transition at $T_{KT}$ and the transition temperature for the CDW order is larger than $T_{KT}$. In striped SS phase where the CDW order breaks the rotational symmetry of the lattice, for $T > T_{KT}$, the system behaves like an ideal conductor along one of the lattice direction while along the other direction it behaves like an insulator. In contrast to this, in star-SS phase, Drude weight along both the lattice directions goes to zero along with the superfluid stiffness and for $T > T_{KT}$ we have a finite temperature phase of a CDW insulator.