Ambipolar doping of a charge-transfer insulator in the Emery model

TL;DR

This study uses cellular dynamical mean-field theory to explore how doping affects a charge-transfer insulator in the Emery model, revealing asymmetries in orbital occupation but similar transition behaviors for electrons and holes.

Contribution

It provides new insights into the doping-driven transitions and orbital character asymmetries in the Emery model, challenging simple explanations of electron-hole asymmetry in cuprates.

Findings

Electrons mainly enter copper orbitals upon doping.

Holes mainly enter oxygen orbitals upon doping.

Both electron and hole doping induce a two-stage transition from insulator to pseudogap to metal.

Abstract

Understanding the similarities and differences between adding or removing electrons from a charge-transfer insulator may provide insights about the origin of the electron-hole asymmetry found in cuprates. Here we study with cellular dynamical mean-field theory the Emery model set in the charge-transfer insulator regime, and dope it with either electrons or holes. We consider the normal state only and focus on the doping evolution of the orbital character of the dopants and on the nature of the doping driven transition. Regarding the orbital character of the dopants, we found an electron-hole asymmetry: doped electrons mostly enter the copper orbitals, whereas doped holes mostly enter the oxygen orbitals. Regarding the nature of the doping driven transition, we found no qualitative electron-hole asymmetry: upon either electron or hole doping, there is a two-stage transition from a charge-transfer insulator to a strongly correlated pseudogap and then to a metal. This shows that a strongly correlated pseudogap is an emergent feature of doped correlated insulators in two dimensions, in qualitative agreement with recent experiments on ambipolar Sr$_{1-x}$La$_{x}$CuO$_{2+y}$ cuprate films. Our results indicate that merely doping with holes or electrons a charge-transfer insulator is not sufficient for explaining the electron-hole asymmetry observed in the normal state phase diagram of cuprates. Our work reinforces the view that actual hole-doped cuprates are more correlated than their electron-doped counterparts.