Skip to content

Atomic-scale Imaging of Time Reversal Breaking Supercurrents in Bi2Sr2CaCu2O8

Bin Hu, Catherine Dawson, Shuqiu Wang, Siyuan Wan, Huiyu Zhao, Kuanysh Zhussubekov, S. Ishida, Hiroshi Eisaki, Peter J. Hirschfeld, Jonathan Ward, J. C. Séamus Davis

cond-mat.supr-conarXiv:2611.23060

Abstract

A persistent mystery of CuO2-based high temperature superconductivity has been the mechanism and consequences of time reversal (T)-symmetry breaking. Although theoretical hypotheses focus upon persistent electronic currents as a potential cause, these have been invisible to direct experimental probes. That impasse has recently been removed by development of techniques for atomic-scale supercurrent visualization. Here, by imaging the energy-resolved density of electronic states N(r,E) in Bi2Sr2CaCu2O8, we search for the specific signatures of flowing supercurrents in a disordered d-wave superconductor. Employing the strategy of artificial neural networks (ANN) for complex quantum matter visualization, we designed a dual-ANN (VMAP) to efficiently implement d-wave supercurrent visualization, achieving imaging of supercurrent speeds V(r) at magnetic fields |B| ≤ 50 μT in Bi2Sr2CaCu2O8. Correspondence between specific characteristics of N(r,E) around each supercurrent loop and a quantum &#34;Rosetta Stone&#34; which translates N(r,E) to V(r), emphasizes the validity of V(r) detection and visualization. Then, for hole-densities 0.17 ≤ p ≤ 0.22 we observe statistically repeatable networks of supercurrent flows. Crucially, these supercurrents satisfy Kirchhoff's law at junctions within each network. Further, theoretical predictions for supercurrent speeds V in disordered d-wave superconductors are consistent with measured 1250 m/s < V < 4000 m/s, and with reported magnetic field amplitudes |B| ≈ 0.1 mT. Thus, detection and visualization of these supercurrent loop networks provides unique perspectives on T-symmetry breaking in cuprates. It also emphasizes the accelerating utility of ANN techniques for physics discovery from high data volume quantum matter research.

Create a lesson