Figure 1

STM topography image of (a) the SIC phase of Pb-Si surface, (b) a globally uniform 5-ML Pb film, and (c) 2-d Pb islands with various lateral sizes and thicknesses. The numbers labeled on the image (c) indicate thickness of islands in ML. (d) Two 3-ML Pb islands with effective lateral size ($d_{\mathrm{eff}}$) of 74 nm and 15 nm (inset) are shown in the same length scale (sample bias $V_{\mathrm{s}}$ = (a,d) 0.3 V, (b,c) 2 V, tunneling current $I$ = (a) 20 pA, (b,c,d) 10 pA).Reuse & Permissions

Figure 2

Lateral size dependence of differential conductance spectra (dI/dV) measured at 4.3 K is shown for (a) 5-ML islands and (b) 3-ML islands. All differential conductance spectra were taken with the same tunneling parameter with the junction stabilized at $V_{\mathrm{s}}$ = 20 mV and $I_{\mathrm{t}}$ = 30 pA tunneling current. (c) Normalized conductance spectra (blue) measured as a function of temperature from a 3-ML Pb island with a 62-nm-lateral size were fitted using the BCS DOS for the tunneling conductance (red). Each normalized spectrum is offset successively by 0.5 for clarity. (d) The superconducting energy gaps (Δ) for each temperature were obtained from (c) and plotted as red circles. The blue curve is a fitting of these energy gap data using a BCS gap equation to obtain a $T_{\mathrm{C}}$ of ∼5.6 ± 0.1 K for a 3-ML island with 62-nm-lateral size. All $T_{\mathrm{C}}$ values determined from such fitting are estimated to have an error bar of ±0.1 K.Reuse & Permissions

Figure 3

(a) $T_{\mathrm{C}}$ as a function of island lateral size for 3-ML, 4-ML, and 5-ML islands. A transition region from slow to dramatic reduction of superconductivity is shaded in green. (b) $T_{\mathrm{C}}$ of islands was normalized to the globally flat film $T_{\mathrm{C}}$ for each thickness and plotted as a function of cube root of island volume. For the $T_{\mathrm{C}}$ normalization, 5-ML film $T_{\mathrm{C}}$ was measured to be 6.1 K, and estimated values of 6.9 K and 6.3 K from the $T_{\mathrm{C}}$ trend in (a) were used for 3-ML and 4-ML film, respectively.Reuse & Permissions

Figure 4

(a) dI/dV tunneling spectra obtained from 9-ML Pb films at 4.3 K and 8.5 K. (b) The result of normalization with the spectrum acquired at 8.5 K (blue line) and the 5th-order polynomial fitting (red line).Reuse & Permissions

Figure 5

Temperature dependence of superconducting gap spectra taken from 9-ML Pb films (a) without RF filter and (b) with RF filter. 0.8-meV and 0.2-meV broadening were applied in the BCS DOS (red lines) of (a) and (b), respectively. Each normalized spectrum is offset successively by 0.3 for clarity.Reuse & Permissions

Figure 6

(a) Direct comparison between experimental spectra measured from 9-ML Pb films with our best RF shielding configuration and theoretical BCS DOS without any broadening. Each spectrum is offset successively by 0.4 for clarity. (b) The superconducting energy gaps (Δ) obtained from Figs. 5a and 5b were plotted as empty and solid blue squares for each temperature, respectively (error bar for each Δ value is smaller than the size of square mark). The red curve is a fitting of these energy gap data using a BCS gap equation. Those gap values from two shielding configurations align very well with a single BCS curve within a $T_{\mathrm{C}}$ of ∼6.6 ± 0.1 K. (c) Temperature dependent gap spectra measured from 5-ML Pb films.Reuse & Permissions

Figure 7

(a,b) Comparison of dI/dV spectrum measured at 6 K (above $T_{\mathrm{C}}$) for 3-ML islands and theoretical tunneling DOS based on electron-electron interaction and disorder [Eq. (B1)]. Because $T$ = 0 K is assumed in Eq. (B1), 1.2-meV broadening effect is applied to the fitting curves to take finite temperature and instrument noise into account. (c,d) 2-d conductivity (σ) values for Pb islands obtained from the fitting result in (a,b).Reuse & Permissions

Figure 8

Shown is a comparison of the pseudogap feature at temperatures below and above the superconducting transition temperature. The lack of any clear temperature dependence for energies above the gap [clear from fit to Eq. (B1)] indicates the high-energy part of the pseudogap is not related to superconductivity but is rather a normal-state property of the Pb nano-islands.Reuse & Permissions