Low energy phenomenology of the overdoped cuprates: Viability of the Landau-BCS paradigm

N R Lee-Hone,H U Özdemir,D M Broun,V Mishra,P J Hirschfeld

doi:10.1103/physrevresearch.2.013228

Abstract

Recently, we have argued that experimental data on superfluid density and terahertz conductivity of overdoped LSCO are compatible with a Landau Fermi liquid/Bardeen-Cooper-Schrieffer description of these samples, provided dopants are treated within "dirty $d$-wave theory" as weak scatterers. Here we test these ideas by comparing to specific heat and thermal conductivity data on LSCO, showing that the theory works extremely well across the overdoped region for similar disorder parameters. We then study the same properties in another overdoped cuprate, Tl-2201, thought to be quite "clean" since it exhibits quantum oscillations, low residual resistivities and small superconducting state Sommerfeld coefficients. Our results are consistent with the Tl-2201 system being $\approx 3$ times cleaner due in part to the dopant atoms' being located further from the CuO$_2$ plane. We conclude that cuprates can be described semiquantitatively in the overdoped regime by "dirty $d$-wave" theory, subject to significant Fermi liquid renormalizations, without introducing physics beyond the Landau-BCS paradigm.

Highlights

Of the various phases observed in the hole-doped cuprate phase diagram [1], the one which seems most conventional in many ways is the d-wave superconductor between hole concentrations of roughly pc1 5% and pc2 30%
Starting from the same dirty d-wave framework and disorder parameters used previously to study the relative temperature dependence of superfluid density [26] and terahertz (THz) conductivity [27] in LSCO, we show that the absolute superfluid density, residual specific heat, Volovik effect, and thermal conductivity can be understood in LSCO
General statements about any part of the cuprate phase diagram are a priori dangerous because cuprates consist of differing numbers of CuO2 planes per unit cell, as well as differing charge reservoir layers

Summary

INTRODUCTION

Of the various phases observed in the hole-doped cuprate phase diagram [1], the one which seems most conventional in many ways is the d-wave superconductor between hole concentrations of roughly pc1 5% and pc2 30%. A few authors have discussed the role of disorder in suppressing Tc [17], and this scenario was emphasized by Rullier-Albenque et al [18], who pressed the analogy between systematic irradiation disorder, which suppresses pc in YBa2Cu3O6+y, and dopant disorder Another longstanding puzzle is the apparent contradiction between the “universal” physics of the CuO2 planes [1,19] and the wide variation of maximum Tc’s across cuprate families. The remarkable agreement found between theory and experiment depends on disorder effects being properly treated and on a realistic parametrization of the electronic structure, including substantial Fermi-liquid renormalizations Our conclusions contradict those of recent experimental studies of superfluid density and optical conductivity on high quality LSCO films [24,25]. We show that despite being significantly cleaner than LSCO, overdoped Tl-2201 is describable within the same dirty d-wave framework and is sufficiently dirty to display the same unusual non-BCSlike proportionality of superfluid density to Tc, as recently measured in overdoped LSCO [24]

DIRTY d-WAVE THEORY

COMPARISON OF LSCO AND Tl-2201

CONCLUSION