Figure 5: Comparisons of experiment with 3D simulations that include disorder. | Nature Communications

Figure 5: Comparisons of experiment with 3D simulations that include disorder.

From: Symmetry protected Josephson supercurrents in three-dimensional topological insulators

Figure 5

(a) Dependence of critical current Ic on gate voltage Vg for a short channel junction (1:3 length to width ratio) (blue), overlayed with experimental results from device 1 (dashed black). (b) Surface (dashed blue) and bulk (solid blue) density of states (DOS) for the model topological insulator Hamiltonian, mapped to Vg for device 1. The bulk is gapped, with the first states appearing near −38 V (though the majority of bulk subbands appear above −20 V). The peak in the bulk DOS near Vg~ −10 V is not consistent with the monotonicity of the bulk DOS seen in the angle-resolved photoemission spectroscopy data; however, Fig. 4a shows that the bulk DOS does not have a significant role in transport within the junction, and thus the model can capture the salient transport physics. The effect of (c) bulk disorder and (d) surface and bulk disorder on Ic in a long-channel junction (2:1 length to width ratio), for δ-impurity strengths equal to 0, 0.2, 0.5 and 1.0 times the bulk bandgap energy. For this device, bulk states become dominant at approximately −20 V (the DOS is similar to that in 5b, but the gate voltage axis would range from −95 to 45 V due to the disparity in backgate capacitance between devices). In c, the supercurrent only decreases appreciably when the bulk disorder potential is large enough to begin hybridizing bulk and surface bands. Including surface disorder, as in d, causes the supercurrent to degrade immediately, signifying that supercurrent flows primarily through the surface band. Experimental results from device 2 are overlayed on both plots (dashed black).

Back to article page