Interbilayer charge dynamics inYBa2Cu3

Cooper Sl

¹

,

P. Nyhus

²

,

D. Reznik

³

et al.

Abstract

We present optical reflectivity and Raman scattering studies of YBaiCujOe-i-x as a function of doping using <7-axis polarized light. The c-axis optical response of YBa2Cu307 is characterized by a Drude-like conductivity and an electronic Raman continuum, indicating a weak coupling between adjacent Cu02 bilayers which is mediated by the CuO chains. With reduced doping, the c-axis Drude and continuum responses rapidly decrease, revealing a strong sensitivity of the interbilayer coupling strength to the removal of oxygen from the chain sites. We suggest that interbilayer charge transport in YBa2Cu306+x may be governed by carrier scattering from in-plane fluctuations.PACS numbers: 74.72. Bk, 78.30.Er While the in-plane properties of the layered high-7"c cuprates are generally well known, the nature of c-axis charge transport remains unresolved. Band structure calculations of the cuprates indicate that the antibonding Cu(3rf)-0(2/7)-plane bands at the Fermi surface are quasi-two-dimensional, exhibiting weak dispersion in the c direction and a c-axis Fermi velocity on the order of r/r~0.7xl0^ cm/sec [1]. By contrast, the Luttinger liquid model proposes that the charge carriers are confined to the Cu02 planes in the cuprates, predicting that transport perpendicular to the Cu02 planes should exhibit "semiconducting" behavior in the normal state with a 1/r temperature-dependent resistivity [2]. Experimentally, the nature of c-axis transport in the layered cuprates is also ambiguous: c-axis transport measurements on Bi2Sr2CaCu208 [3] and YBa2Cu306.6 (^^---60 K) [4] show a semiconductorlike \/T temperature dependence above Tc, while c-axis transport [5] and optical [6l measurements on YBa2Cu307 {Tc^9Q K) indicate "metallic" behavior in the normal state.

Deconfined fermions but confined coherence?

“…YBa 2 Cu 3 O 6+x and Bi 2 Sr 2 CaCu 2 O 8 exhibit activated c axis conduction at low temperatures, with ab transport again remaining metallic [2]. Meanwhile, the frequency dependent conductivity is perhaps even more anomalous with no signs of anything remotely like a Drude peak with any appreciable weight in the c-axis conductivity in either underdoped YBa 2 Cu 3 O 6+x [3] or La 2−x Sr x CuO 4 [4]. In fact, c-axis conduction is rather generically incoherent in these materials and has anomalous temperature dependence in a range near to T c .…”

One-particle interchain hopping in coupled Hubbard chains

“…Such an unusual behavior could, for example, explain the unconventional c-axis conductivity found in two-layered high-T c compounds [1]. In addition, the high transition temperatures of these materials may also be connected to incoherent hopping [2]: as a consequence of incoherence, the coupled planes can lower the total energy by coherent pair-hopping processes, leading to an amplification of the tendency towards superconducting pairing.…”

Collective Charge Density Fluctuations in Superconducting Layered Systems with Bilayer Unit Cells

“…The out-of-phase plasma edge, ω c− ≡ ω − (k = 0), will obviously lie much higher in energy than ω c+ ≡ ω c because the intracell interlayer hopping is much stronger than the intercell interlayer hopping. In particular, even though the ω c+ mode may lie in the superconducting gap 16,12 , we expect ω c− to lie much above the superconducting gap energy in Y BCO. A crude qualitative estimate can be made by assuming that the intra-and intercell hopping amplitudes scale as inverse squares of lattice parameters: t intra /t inter ≈ (d/c) 2 = 16.…”

“…To make things really complicated one of these modes (ω c1 ) could be below the gap and the other (ω c2 ) above the gap, (or, both could be below or above the gap). While each of these scenarios is possible, c-axis optical response experimental results on inter-bilayer charge dynamics in Y BCO have been interpreted 16 to exhibit only one c-axis plasma edge in the superconducting state with the frequency ω c between 60 cm −1 and 200 cm −1 , depending on the oxygen content. There are three possibilities: (1) The two plasma modes (ω c1 ≈ ω c2 ≈ ω c ) are almost degenerate because the corresponding intercell hopping amplitudes are close in magnitudes; (2) ω c2 is much lager than ω c1 (≪ ω c2 ) because the two intercell hopping amplitude are very different in magnitudes (we consider this to be an unlikely scenario); (3) one of the two modes carries very little optical spectral weight and is not showing up in c-axis reflectivity measurements, leaving only the other one as the observed c-axis plasma edge.…”

“…This then leads to the approximate formula ω c− ≈ 16 2 ω c+ = 256 ω c , which, for Y BCO, implies that the long wavelength out-of-phase mode should lie between 2 eV and 6 eV, depeding on the oxygen content (assuming that the c-axis plasma edge varies between 60 cm −1 and 200 cm −1 , as reported in ref. 16, depending on the oxygen content). While there is some minor observable structure in optical experiments at high energies, we cannot find any compelling evidence in favor of the existence of a high energy out-of-phase mode in the currently available experimental data.…”