The Effect of Pointlike Impurities on d_{x^2-y^2} Charge Density Waves in Cuprate Superconductors

TL;DR

This paper investigates how pointlike impurities affect the $d_{x^2-y^2}$ charge density wave order in cuprate superconductors, revealing a rapid suppression of bond order compared to superconductivity and questioning its relation to the pseudogap.

Contribution

The study introduces a model analyzing impurity effects on bond order in cuprates, highlighting the role of Fermi surface hotspots and contrasting impurity sensitivity with superconductivity.

Findings

Bond order is locally disrupted near impurities.

Transition temperature $T_{bo}$ decreases faster than $T_c$ with impurity concentration.

Charge order's insensitivity to Zn doping suggests it is distinct from the pseudogap.

Abstract

Many cuprate superconductors possess an unusual charge-ordered phase that is characterized by an approximate $d_{x^2-y^2}$ intra-unit cell form factor and a finite modulation wavevector $\bq^\ast$. We study the effects impurities on this charge ordered phase via a single-band model in which bond order is the analogue of charge order in the cuprates. Impurities are assumed to be pointlike and are treated within the self-consistent t-matrix approximation (SCTMA). We show that suppression of bond order by impurities occurs through the local disruption of the $d_{x^2-y^2}$ form factor near individual impurities. Unlike $d$-wave superconductors, where the sensitivity of $T_c$ to impurities can be traced to a vanishing average of the $d_{x^2-y^2}$ order parameter over the Fermi surface, the response of bond order to impurities is dictated by a few Fermi surface "hotspots". The bond order transition temperature $T_\mathrm{bo}$ thus follows a different universal dependence on impurity concentration $n_i$ than does the superconducting $T_c$. In particular, $T_\mathrm{bo}$ decreases more rapidly than $T_c$ with increasing $n_i$ when there is a nonzero Fermi surface curvature at the hotspots. Based on experimental evidence that the pseudogap is insensitive to Zn doping, we conclude that a direct connection between charge order and the pseudogap is unlikely. Furthermore, the enhancement of stripe correlations in the La-based cuprates by Zn doping is evidence that this charge order is also distinct from stripes.