Figure 1

Inset: layout of the experiment: a $10\mathrm{\text{−}}\mu \mathrm{m}$-long, 120-nm-wide, and 40-nm-thick superconducting (aluminum) wire can be current biased at $I_S$ or exposed to a magnetic field $B$. A normal probe electrode forms a tunnel junction (dashed area) with the wire. Main panel: measured $dI/dV(V)$ for different combinations of the bias current and magnetic field: dashed line: $I_S=0$ and $B=0$; solid lines: $I_S=70\mathrm{\mu }\mathrm{A}$ and $B=0$, and $I_S=0$ and $B=23\mathrm{m}\mathrm{T}$. To a good approximation (see text), the differential conductance of the junction $dI/dV(V)$ is proportional to the DOS in the superconductor.Reuse & Permissions

Figure 2

Normalized differential conductance $dI/dV(V)$ of the probe tunnel junction: Top: at $B=0$, as a function of the supercurrent $I_S$ (from right to left: $I_S=17$, 51, and $85\mathrm{\mu }\mathrm{A}$). Bottom: at $I_S=0$, as a function of the magnetic field $B$ (from 11.5 to 69 mT by steps of 11.5 mT). Solid lines are best fits with $dI/dV(V)$ calculated with an electronic temperature dependent on $V$ (see text); dashed lines are the best fits with $dI/dV(V)$ calculated with a constant electronic temperature. Insets: depairing energy $\Gamma$ (in units of the gap ${\Delta }_0$ at $B=0$ and $I_S=0$) for different currents and magnetic fields, deduced from the fits of $dI/dV(V)$. In the top inset, square symbols correspond to the data in the main panel ($B=0$), whereas triangles and disks were obtained from data taken at $B=10.2\mathrm{m}\mathrm{T}$ and $B=27\mathrm{m}\mathrm{T}$, respectively. Solid lines: fits with theory, leading to depairing current and magnetic field $I_{\Gamma }=240\mu \mathrm{A}$ and $B_{\Gamma }=105\mathrm{m}\mathrm{T}$.Reuse & Permissions