Scaling properties of the optical conductivity of Bi-based cuprates

TL;DR

This work investigates whether the infrared optical conductivity in optimally doped Bi2223 obeys a universal $\omega/T$ scaling. It introduces a differential-integral transformation to extract the scaling function $g(\omega/T)$ from $\sigma(\omega,T)$ while separating a temperature-independent background $\sigma^{(0)}(\omega)$, and applies it to high-precision data. The extracted scaling function is Drude-like with a scattering rate that scales linearly with temperature and is independent of frequency, while a sizable background carries most of the spectral weight below 1 eV, leading to a broad collapse of spectra across temperatures after background subtraction. These results suggest the low-energy transport is governed by a single temperature scale, compatible with quantum-critical or incoherent transport pictures, but the interpretation is tempered by the non-negligible background and potential high-energy cutoff effects.

Abstract

We present novel infrared optical conductivity data on the three layer high Tc superconductor Bi2Sr2Ca2Cu3O10 at optimal doping. We extend the analysis of an earlier publication, providing a universal scaling function sigma(omega,T)=g(omega/T)/T for the optical conductivity. In the present manuscript we obtain a good scaling collapse of the experimental curves on the g(omega/T) over a wide range of values of omega/T (at least in range 0 to 10), if we assume that g(omega/T) is superimposed on a non-universal background which is temperature independent. We obtain the same result, if in our analysis we allow this background to have a T-squared temperature dependent correction. The most striking property of g(omega/T) is, that it corresponds to a scattering rate which varies linearly as a function of temperature, but which is independent of the frequency.

Figures (10)

• Figure 1: Temperature dependence of the ab-plane optical conductivity of Bi2223 for three selected frequencies.
• Figure 2: Optical conductivity spectra along the ab-plane of optimally doped Bi2223 at selected temperatures.
• Figure 3: In-plane frequency-dependent scattering rate of Bi2223.
• Figure 4: Phase of the in-plane optical conductivity of Bi2223.
• Figure 5: Real and imaginary parts of the scaling function $g(\omega,T)$ of Bi2223, calculated for 4 different temperatures from the experimental data using Eq. \ref{['g1']}. The grey open symbols are a simple Drude formula, with a frequency independent scattering which is proportional to the temperature.