Oxygen vacancy modulated superconductivity in monolayer FeSe on $\mathrm{SrTi}{\mathrm{O}}_{3-\delta }$

The monolayer FeSe on $\mathrm{SrTi}{\mathrm{O}}_{3-\delta }$ exhibits a twofold enlarged pairing gap compared with other electron-doped FeSe, which is postulated to originate from cooperative effects of interface charge transfer and electron-boson interaction that relate closely to surface oxygen vacancies in $\mathrm{SrTi}{\mathrm{O}}_3$. We create gradient oxygen vacancies on the surface of $\mathrm{SrTi}{\mathrm{O}}_{3-\delta }$ substrates using direct current heating. Combining spatially resolved spectroscopy characterizations, we disclose gradient pairing gaps but negligible doping variation in monolayer FeSe on such substrates. As oxygen vacancy concentration gradually increases from the anode to cathode region, the pairing gap increases from 12 to 17 meV, right beyond the values of electron-doped FeSe. This modulation provides a platform for spatially resolved investigations to unveil the interplay between electronic correlation and electron-phonon interaction and their cooperation to drive superconductivity.

Figure 1

Schematic and STM characterization of $\mathrm{SrTi}{\mathrm{O}}_{3-\delta }$(001) substrate. (a) A photo of a pristine $\mathrm{SrTi}{\mathrm{O}}_{3-\delta }$ showing gradient color change from the anode to cathode region (top panel) and schematic of gradient oxygen vacancies (bottom panel). (b) and (c) The typical STM topographic images ($V_{\mathrm{s}}=500\mathrm{mV}, I=50\mathrm{pA}$) of $\mathrm{SrTi}{\mathrm{O}}_{3-\delta }$(001) substrates in the L and H regions with local preheating temperatures of 1173 °C and 1143 °C, respectively. The insets are the corresponding FFT images with the red circles marking the 2×2 Bragg points. (d) The histogram distribution of the work function in the L and H regions.

Figure 2

STEM characterization of monolayer FeSe/$\mathrm{SrTi}{\mathrm{O}}_{3-\delta }$ heterostructures. (a) Atomically resolved cross-sectional HAADF images viewed along the [010] direction at the L, M, and H regions. The blue circles mark the columns of Ti in the topmost $\mathrm{Ti}{\mathrm{O}}_{2-\delta }$ layer. (b) Local electron energy loss spectroscopy of $\mathrm{Ti}-{\mathrm{L}}_{2,3}$-edge taken at the top three $\mathrm{Ti}{\mathrm{O}}_{2-\delta }$ layers in the three regions shown in (a). Experimental data are shown as colored dots, fitted (black dashed lines) by the addition of the reference spectra shown in the green line ($\mathrm{T}{\mathrm{i}}^{3+}$) and black line ($\mathrm{T}{\mathrm{i}}^{4+}$) at the bottom. The residual to the fit is given by the purple dotted line.

Figure 3

STM characterization of spatial-dependent superconductivity. (a)-(c) Atomically resolved topographic images of monolayer FeSe at the L, M, and H regions ($V_{\mathrm{s}}=200\mathrm{mV}, I=500\mathrm{pA}$). (d)–(f) Tunneling spectra taken along the white lines in (a)–(c) ($V_{\mathrm{s}}=50\mathrm{mV}, I=500\mathrm{pA}$), respectively. (g) The histogram distribution of superconducting gaps in the M and H regions showing the gap variation. (h) Tunneling spectra in a large bias range taken on the L, M, and H regions ($V_{\mathrm{s}}=500\mathrm{mV}, I=250\mathrm{pA}$), in comparison with the tunneling spectra evolution during postannealing of monolayer FeSe (bottom group of three curves). In (d)–(f) and (h), the spectra are shifted along the vertical axis. The horizontal bars in (d)–(f) indicate the zero-conductance position of each curve.

Figure 4

Spatial-dependent distribution of coverage, superconducting gaps and doping level. (a) and (b) The topographic images of monolayer FeSe on $\mathrm{SrTi}{\mathrm{O}}_3$(001) taken on the 0 and 2.5 mm positions marked in (c) ($V_{\mathrm{s}}=1\mathrm{V}, I=10\mathrm{pA}$), respectively. (c) The magnitude of superconducting gap (pink dots) and doping level (purple dots) as a function of positions, with the preheating temperature for $\mathrm{SrTi}{\mathrm{O}}_3$ (black dots) and postannealing temperature for monolayer FeSe (red dots) inserted. The horizontal error bar indicates a position uncertainty of ±0.25 mm in ARPES measurements. (d) The topographic images of monolayer FeSe on $\mathrm{SrTi}{\mathrm{O}}_{3-\delta }$(001) preheated at a lower temperature of 1200 °C ($V_{\mathrm{s}}=1\mathrm{V}, I=10\mathrm{pA}$). (e) Tunneling spectra taken on the sample shown in (d) ($V_{\mathrm{s}}=50\mathrm{mV}, I=500\mathrm{pA}$). The spectra are shifted along the vertical axis, and the horizontal bars indicate the zero-conductance position of each curve.