Nonlinear c-axis transport in Bi_2Sr_2CaCu_2O_(8+d) from two-barrier tunneling

TL;DR

This study extends a two-barrier tunneling model to explain the normal-state c-axis transport features in Bi-2212 superconductors, successfully reproducing experimental tunneling spectra and revealing that conventional tunneling accounts for many observed anomalies.

Contribution

The paper introduces an extended normal-state two-barrier tunneling model that accurately reproduces key experimental features of c-axis transport in Bi-2212, highlighting the role of conventional tunneling.

Findings

The model reproduces the parabolic V-dependence of dI/dV at high temperatures.

It explains the emergence and position of humps in the spectra.

The model matches experimental data without shifting the dI/dV curves vertically.

Abstract

Motivated by the peculiar features observed through intrinsic tunneling spectroscopy of Bi$_2$Sr$_2$CaCu$_2$O$_{8+\delta}$ mesas in the normal state, we have extended the normal state two-barrier model for the c-axis transport [M. Giura et al., Phys. Rev. B {\bf 68}, 134505 (2003)] to the analysis of $dI/dV$ curves. We have found that the purely normal-state model reproduces all the following experimental features: (a) the parabolic $V$-dependence of $dI/dV$ in the high-$T$ region (above the conventional pseudogap temperature), (b) the emergence and the nearly voltage-independent position of the "humps" from this parabolic behavior lowering the temperature, and (c) the crossing of the absolute $dI/dV$ curves at a characteristic voltage $V^\times$. Our findings indicate that conventional tunneling can be at the origin of most of the uncommon features of the c axis transport in Bi$_2$Sr$_2$CaCu$_2$O$_{8+\delta}$. We have compared our calculations to experimental data taken in severely underdoped and slightly underdoped Bi$_2$Sr$_2$CaCu$_2$O$_{8+\delta}$ small mesas. We have found good agreement between the data and the calculations, without any shift of the calculated dI/dV on the vertical scale. In particular, in the normal state (above $T^\ast$) simple tunneling reproduces the experimental dI/dV quantitatively. Below $T^\ast$ quantitative discrepancies are limited to a simple rescaling of the voltage in the theoretical curves by a factor $\sim$2. The need for such modifications remains an open question, that might be connected to a change of the charge of a fraction of the carriers across the pseudogap opening.