Doping dependence of the low temperature planar carrier density in overdoped YBa₂Cu₃O_7-δ

The rapid change in the low-T Hall number n_H from p (the number of doped holes) to 1 + p (the full LDA carrier density) observed in the first pulsed field measurements of YBa₂Cu₃O_7-δ (Y123) at T = 50 K suggests possible critical behaviour near the pseudogap end point (p*) [1]. In a more recent study of the single-layer cuprates Tl₂Ba₂CuO_6+δ (Tl2201) and La-doped Bi₂Sr₂CuO_6+δ (La-Bi2201) [2], the evolution from p to 1 + p was instead found to extend linearly across a far broader doping range -- out toward the edge of the superconducting dome. The inconsistency may arise from the presence of quasi-one-dimensional CuO chains in Y123, but not in Tl2201 or Bi2201. These chains contribute little to the Hall signal but generate a notable resistivity anisotropy, particularly in overdoped Y123 crystals with conductive chains [3]. As a result, n_H is no longer a direct measure of the carrier density in the CuO₂ planes and a correction factor is required to obtain the planar carrier density, n_pl = n_H(ρ_a/ρ_b)^-1 . While ρ_a/ρ_b is known to be strongly temperature and doping dependent above T_c [3] the low-T anisotropy ratio has not yet been rigorously studied.

We have used pulsed fields to examine the low-T behaviour of n_pl between optimal doping (p=0.16) and p* (p≈0.19) by measuring the resistivity anisotropy, ρ_a/ρ_b and Hall number n_H in individual detwinned Y123 single crystals. We reaffirm the previous finding of an abrupt rise in n_H(50 K) across this doping regime [1] and report two other key results: (i) at optimal doping, n_pl = p and (ii) the sharp rise in n_H(p) is softened by the resistivity anisotropy to such an extent that the full Luttinger count (n_pl = 1+p) is only partially recovered at p*. This latter result disfavours a conventional QCP scenario in which the pseudogap end point leads to a reconstructed Fermi surface.

[1] Badoux et al., Nature 531, 210 (2016)

[2] Putzke et al., Nature Physics, 17, 826 (2019)

[3] Segawa et al., PRB 69, 104521 (2004).