Fig. 4: Interference at ν = 3.
a QPC conductance versus gate voltage at ν = 3. \(G=3\frac{{e}^{2}}{h}\) and \(G=2\frac{{e}^{2}}{h}\) plateaus are observed. The \(G=\frac{{e}^{2}}{h}\) plateau is not well-formed, likely due to the small N = 0 Landau level spin gap at B = 0.9 T, which prevents formation of a wide incompressible region between the spin-down and spin-up N = 0 Landau level edge modes. However, an inflection near \(G=\frac{{e}^{2}}{h}\) is observed, indicating a transition from backscattering the spin-down N = 0 edge mode to backscattering the spin-up N = 0 edge mode. Red circles indicate the transmission used to interfere each mode at ν = 3. b Interference generated by partially backscattering the innermost edge mode (spin-down N = 1 Landau level). (c) Interference produced when backscattering the middle edge mode (spin-up N = 0 Landau level). Black solid lines are drawn along lines of nearly constant phase in (b) and (c). d Interference when partially backscattering the outermost edge mode (spin-down N = 0 Landau level). e 2D FFT for interference of the outermost edge mode at ν = 3 identifying the most significant frequencies in the interference. f Plot of the magnitude of discontinuous phase slips observed in Fig. 4d with \(\overline{\Delta \theta }/2\pi \approx 0.44\). g Differential conductance measurement for the outermost mode at ν = 3. h Plot of FFT amplitude vs. δVSD1 used to determine the node spacing for the outermost mode.
