Slave Boson Approach to Neutron Scattering inYBa2Cu3O6+y

Abstract

The evolution of the so-called "41 meV resonance" in the magnetic response of YBCO cuprates is studied with slave-boson theory for the t-t ′ -J-model. The resonance appears as a collective spin fluctuation in the d-wave superconducting (SC) state. It is undamped at optimal doping due to a threshold in the excitation energies of particle-hole pairs with relative wave vector (π, π) . When hole filling is reduced, the resonance moves to lower energies and broadens. Below the resonance energy we find a crossover to an incommensurate response in agreement with a recent experiment on YBa2Cu3O6.6 . We show that dynamic nesting in the d-wave SC state causes this effect.PACS numbers: 71.10. Fd, 74.25.Ha, 74.72.Bk, 75.20.Hr One of the puzzling phenomena in copper-oxide superconductors is the so-called "41 meV resonance" observed in inelastic neutron-scattering (INS) experiments on optimally doped Y Ba 2 Cu 3 O 6+y (YBCO) compounds in the superconducting (SC) phase [1][2][3][4][5] . The magnetic susceptibility χ ′′ (q, ω) shows a peak at the antiferromagnetic (AF) wave vector (π, π) and an energy ω 0 =41 meV . The peak appears to be resolution limited, i.e., almost undamped, and vanishes above T c . In underdoped YBCO the peak shifts to lower energies ω 0 < 41 meV [6][7][8] and develops some damping. It is also visible in the spingap regime above T c . Several theoretical approaches to the magnetic excitations have been proposed [9][10][11][12][13][14][15] . In most cases [9][10][11][12][13][14] the resonance at optimal doping is identified as a collective spin fluctuation, which is stabilized through the suppression of quasi-particle damping in the SC state, and vanishes in the normal phase. The question arises how this mode evolves with decreasing doping. In particular it is unclear how the finite damping at intermediate doping levels can be obtained, since the peak is expected to narrow and become a spin-wave (Bragg) peak at zero energy when the Néel state is reached. There is also experimental evidence [2,16] that the magnetic response in YBCO at energies somewhat below ω 0 is incommensurate. It is dominated by four peaks at q = (π, π±δ) and q = (π ± δ, π) , a pattern previously known only for the La 2−x Sr x CuO 4+y family of compounds. In view of the new experimental data, we re-investigate the spin response in slave-boson mean-field theory [9-11] , focusing on the evolution of the resonance with hole filling (doping) and the crossover commensurate-incommensurate with variation of energy.We start from the t-t'-J-model for a single CuO 2 -layer. Results specific to the bi-layer structure of YBCO will be presented elsewhere. The model readsSums include nearest neighbor (i, j) or next n.n. (i, j) ′ Cu-sites on a 2D square lattice. Physical operators c iσ = The magnetic susceptibility is calculated in ladder approximation,with J(q) = J[cos(q x )+cos(q y )] and α = 1 . Fig. 1 shows χ 0 (q, ω) , which is built of fermion bubbles with a vertex function to include all ladder diagrams. Boson excitations do not appear in l...

Phase separation models for cuprate stripe arrays

“…An EPS crossover from stripes to domains near x ∼ 1/8 would provide a natural explanation of these phenomena. Interestingly, the only calculation 82 to approximately describe the doping dependence is based on a slave boson model, which as noted above tends to mimic EPS behavior as x → 0.…”

“…For both materials, this is the doping at which the pseudogap closes. While in YBCO, these peaks have been interpreted as a Fermi surface nesting effect 82,83 , this cannot explain the charge order fluctuations, and in LSCO the Tranquada data 2 is associated with elastic magnetic peaks. Furthermore, a study of the neutron diffraction pair distribution function 84 for LSCO finds evidence for charge fluctuations, presumably associated with stripes.…”

Spin anisotropy of the magnetic excitations in the normal and superconducting states of optimally doped YBa2Cu3O6.9Headings1, Hayden2, Kulda3 et al. 2011Phys. Rev. B312261View full textAdd to dashboardCite

“…[42][43][44][45][46][47][48][49][50][51][52][53] Many features are explained by a magnetic exciton scenario 42,43,46,47,52 in which the resonance is a bound state in the particle-hole channel, which appears in a region of q − ω space where there are no damping processes due to electron-hole pair creation. This is illustrated schematically in Fig.…”

Magnetic fluctuations and resonant peak in cuprates: Towards a microscopic theory

“…Relevant microscopic models, such as the Hubbard model and the t-J model have been so far studied in the weak coupling or random-phase approximation [6], neglecting strong correlations. The latter have been considered using a Hubbard-operator technique [10], and more recently within the self-consistent slave-boson approach [11], self-consistent spin-fluctuation method [12], as well as within the phenomenological spin-fermion model [8,9].…”