Proximity-induced superconductivity in a topological crystalline insulator

Superconducting topological crystalline insulators (TCIs) are predicted to host new topological phases protected by crystalline symmetries, but available materials are insufficiently suitable for surface studies. To induce superconductivity at the surface of a prototypical TCI SnTe, we use molecular beam epitaxy to grow a heterostructure of SnTe and a high-$T_c$ superconductor Fe(Te,Se), utilizing a “buffer” layer to bridge the large lattice mismatch between SnTe and Fe(Te,Se). Using low-temperature scanning tunneling microscopy and spectroscopy, we measure a prominent spectral gap on the surface of SnTe, and demonstrate its superconducting origin by its dependence on temperature and magnetic field. Our work provides a platform for atomic-scale investigations of emergent topological phenomena in superconducting TCIs.

Figure 1

(a) Schematic of SnTe/${\mathrm{Bi}}_2{\mathrm{Te}}_3$/Fe(Te,Se) heterostructure. STM topograph showing (b) exposed Fe(Te,Se) substrate, (c) 1 QL ${\mathrm{Bi}}_2{\mathrm{Te}}_3$ buffer layer, and (d) the topmost SnTe layer. Insets in (b)–(d) are the respective Fourier transforms showing their lattice symmetries. (e) Height profile taken along the red dashed line in the embedded STM topograph. The step height is consistent with consecutive bilayers of SnTe. STM setup conditions: (b) $I_{\mathrm{set}}=15$ pA, $V_{\mathrm{sample}}=10$ mV; (c) $I_{\mathrm{set}}=30$ pA, $V_{\mathrm{sample}}=6$ mV; (d) $I_{\mathrm{set}}=40$ pA, $V_{\mathrm{sample}}=40$ mV; (e) $I_{\mathrm{set}}=10$ pA, $V_{\mathrm{sample}}=-500$ mV.

Figure 2

(a) Long-range average dI/dV spectra on sample A (blue), sample B (red), and sample C (green). VBT denotes the approximate position of the top of the valence band. (b)–(d) Average dI/dV spectra on the surface of (b) sample A, (c) sample B, and (d) sample C, with a schematic of their respective heterostructures (inset in the lower right corner). (e) A series of dI/dV spectra, offset for clarity, along a 25-nm line taken on the surface of sample A. (f) Average dI/dV spectra, offset for clarity, as a function of temperature, showing the gap closing at $\sim 12$ K. (g) Spectral gap ${\mathrm{\Delta }}_{\mathrm{SnTe}}$ as a function of temperature extracted from spectra in (f) (black squares), and the overlaid BCS trend (gray line). STM setup condition: (a) Sample A: $I_{\mathrm{set}}=100$ pA, $V_{\mathrm{sample}}=300$ mV, $V_{\mathrm{exc}}=10$ mV (zero-to-peak); sample B: $I_{\mathrm{set}}=200$ pA, $V_{\mathrm{sample}}=400$ mV, $V_{\mathrm{exc}}=10$ mV; sample C: $I_{\mathrm{set}}=100$ pA, $V_{\mathrm{sample}}=300$ mV, $V_{\mathrm{exc}}=10$ mV; (b) $I_{\mathrm{set}}=40$ pA, $V_{\mathrm{sample}}=10$ mV, $V_{\mathrm{exc}}=0$.2 mV; (c) 4 BL: $I_{\mathrm{set}}=100$ pA, $V_{\mathrm{sample}}=10$ mV, $V_{\mathrm{exc}}=0$.2 mV; 5 BL: $I_{\mathrm{set}}=60$ pA, $V_{\mathrm{sample}}=10$ mV, $V_{\mathrm{exc}}=0$.2 mV; 6 BL: $I_{\mathrm{set}}=60$ pA, $V_{\mathrm{sample}}=10$ mV, $V_{\mathrm{exc}}=0$.2 mV; (d) $I_{\mathrm{set}}=150$ pA, $V_{\mathrm{sample}}=10$ mV, $V_{\mathrm{exc}}=0$.3 mV; (e) $I_{\mathrm{set}}=40$ pA, $V_{\mathrm{sample}}=10$ mV, $V_{\mathrm{exc}}=0$.2 mV, (f) $I_{\mathrm{set}}=30$ pA, $V_{\mathrm{sample}}=5$ mV, $V_{\mathrm{exc}}=0$.2 mV.

Figure 3

Quasiparticle interference (QPI) imaging of SnTe(111) surface of sample C. (a) The schematic of the SnTe(111) constant energy contour, where $q_1$ (purple), $q_2$ (blue), and $q_3$ (green) denote dominant scattering vectors. The dashed hexagon in (a) denotes the first Brillouin zone. (b)–(d) Fourier transforms of dI/dV maps acquired at 100, 8, and −8 mV bias, respectively. The peaks in the Fourier transform in (b)–(d) circled in purple, blue, and green correspond to the scattering channels denoted in panel (a). Yellow dashed circles in (b)–(d) denote the atomic Bragg peaks associated with the Te lattice. STM setup conditions: (b) $I_{\mathrm{set}}=200$ pA, $V_{\mathrm{sample}}=100$ mV, $V_{\mathrm{exc}}=2$ mV; (c) $I_{\mathrm{set}}=200$ pA, $V_{\mathrm{sample}}=-8$ mV, $V_{\mathrm{exc}}=2$ mV; (d) $I_{\mathrm{set}}=200$ pA, $V_{\mathrm{sample}}=8$ mV, $V_{\mathrm{exc}}=2$ mV.

Figure 4

(a)–(c) dI/dV maps over the same region of the sample at 1, 2, and 4 T magnetic fields, respectively. All fields are applied perpendicular to the surface of the sample. (d) Radially averaged dI/dV spectra as a function of distance away from the vortex core in (b). Zero-bias peak is visible in the vortex core (darkest blue). STM setup condition: (a) $I_{\mathrm{set}}=4$ pA, $V_{\mathrm{sample}}=2$ mV, $V_{\mathrm{exc}}=0.4$ mV; (b), (d) $I_{\mathrm{set}}=8$ pA, $V_{\mathrm{sample}}=2$ mV, $V_{\mathrm{exc}}=0.4$ mV; (c) $I_{\mathrm{set}}=6$ pA, $V_{\mathrm{sample}}=2$ mV, $V_{\mathrm{exc}}=0.4$ mV.