Figure 1

Superconducting transition revealed by Ac susceptibility measurements. The onset temperature is $9.25\mathrm{K}$ with a transition width of about $0.25\mathrm{K}$. The inset is the schematic diagram of the junction configuration for point-contact spectra measurements.Reuse & Permissions

Figure 2

(Color online) Temperature dependence of the Andreev reflection spectra for $\mathrm{Nb}∕\mathrm{Au}$-foil point contact, which have been normalized by the high bias conductance. The corresponding temperatures from the top down are from $2\mathrm{K}\text{to}9.25\mathrm{K}$ with a step of $0.5\mathrm{K}$ between $2\mathrm{K}$ and $8.5\mathrm{K}$, and a step of $0.25\mathrm{K}$ between $8.75\mathrm{K}\text{to}9.25\mathrm{K}$. The solid lines represent the fits to the conventional BTK model.Reuse & Permissions

Figure 3

The temperature dependence of all the fitting parameters, namely, (a) barrier height $Z$, broadening factor $\Gamma$, and (b) superconducting gap $\Delta$. The solid line in (b) denotes the $\Delta \left(T\right)$ relation expected from BCS theory.Reuse & Permissions

Figure 4

Typical magnetic field dependence of the Andreev reflection spectra for $\mathrm{Nb}∕\mathrm{Au}$ foil point contact, which are normalized by the high bias data. The curves above $3.2\mathrm{T}$ is not presented here due to serious distortion and larger noise.Reuse & Permissions

Figure 5

(Color online) Fitting the spectra in various fields to the revised two-channel model with the free parameter $h_{\mathrm{eff}}$. The experimental data are denoted by solid circles and the calculations are indicated by solid lines. All the curves except for the lowest one are shifted upwards for clarity. The corresponding magnetic fields from the bottom up are 0, 0.05, 0.1, 0.2, 0.3, 0.4, 0.6, 0.8, 1.0, 1.2, 1.4, 1.6, 1.8, 2.0, 2.4, 2.8, 3.2 in units of T.Reuse & Permissions

Figure 6

Fitting parameters [(a) barrier height $Z$, (b) superconducting gap $\Delta$], and (c) broadening factor $\Gamma$ obtained from the experimental curves using the original two-channel model (solid diamonds) and revised two-channel model (open circles).Reuse & Permissions

Figure 7

(Color online) The dependence of the derived fitting parameter $Z$ and the junction resistance $R$ on the temperature and magnetic field, respectively. The $Z\left(T\right)$ relation was obtained by fitting the temperature dependent spectra to the generalized BTK model (open diamonds), while the $Z\left(H\right)$ relation was determined by fitting the field dependent spectra to the original two-channel model (solid diamonds). All the investigated spectra were measured at a fixed point-contact junction and the small variation of the junction resistance comes from the slight thermal relaxation of the device in the temperature controller system. Note that all the $Z$ values are almost the same except for that with the fields above $1.2\mathrm{T}$, indicating that the original two-channel model is not adequate above this crossover field.Reuse & Permissions

Figure 8

Field dependence of the fitting parameter $h_{\mathrm{eff}}$ obtained using the revised two-channel model. The thick straight line of $h=H∕H_{c2}$ is the parameter accepted in the original two-channel model based on the concept of rigid vortex core. The open circles indicate the best fitting values of $h_{\mathrm{eff}}$ determined by freely selecting $Z$ in the range of $\pm 2\%$ as discussed in the text. The solid circles represent $h_{\mathrm{eff}}$ obtained by restricting $Z$ to the extension line of the low-field $Z\left(H\right)$ relation. Inset: the derived normalized ${\xi }_{\mathrm{eff}}=\sqrt{h_{\mathrm{eff}}∕h}$, note the decrease starting at the crossover field $H^{*}\approx 1.2\mathrm{T}$.Reuse & Permissions

Figure 9

Schematic spatial variation of the normalized order parameter in the presence of Abrikosov vortices is shown by the solid lines (a) for low field and (b), (c) for high field (Ref. 5). The dotted lines denote the simplified description of the spatial variation. When the field is above $H^{*}$, the normal region defined by a continuously decaying ${\xi }_{\mathrm{eff}}$ as shown in (c) is more reasonable than that defined by the rigid core with the size of constant coherence length $\xi$ as shown in (b).Reuse & Permissions