Experimental Identification of the Second-Order Non-Hermitian Skin Effect with Physics-Graph-Informed Machine...

Area of Science:

• Condensed Matter Physics
• Topological Matter
• Non-Hermitian Systems

Background:

• Topological phases conventionally follow bulk-boundary correspondence.
• Higher-order topological insulators exhibit bulk-edge-corner correspondence.
• Non-Hermitian systems introduce the non-Hermitian skin effect (NHSE).

Purpose of the Study:

• To experimentally demonstrate the higher-order NHSE.
• To explore the second-order NHSE in a 2D non-Hermitian system.
• To investigate the role of machine learning in discovering topological phenomena.

Main Methods:

• Utilized physics-graph-informed machine learning (PGIML) for data analysis.
• Experimentally realized a 2D non-Hermitian topoelectrical circuit.
• Analyzed admittance spectra under open boundary conditions.

Main Results:

• Successfully demonstrated the second-order NHSE.
• Observed an extensive number of corner skin modes.
• Showcased extreme sensitivity of spectral flow to boundary conditions.
• Violated conventional bulk-boundary correspondence.

Conclusions:

• Higher-order NHSE is experimentally achievable.
• The conventional bulk-boundary correspondence needs modification for higher-dimensional non-Hermitian systems.
• PGIML is a powerful tool for exploring complex physical phenomena.