Figure 1

Sketch of experimental setup. A chain of small-capacitance SQUIDs is measured using a voltage-bias scheme. The sample leads are filtered at several stages of the dilution refrigerator.Reuse & Permissions

Figure 2

(a) Current-voltage characteristic of the 255-islands array at a temperature well below the thermal activation of transport and at an external flux of ${\Phi }_0/2$ per SQUID loop. Within the setup's resolution, no current can be detected below 1 mV bias voltage due to the Coulomb blockade. Left inset: With increasing temperature, a finite conductance ($G_T$) develops close to zero voltage bias. Right inset: Above a threshold voltage (1.2 mV for $N=255$ near full suppression of $E_J$), a constant differential conductance ($G_V$) can be observed. (b) Thermally activated conductance $G_T$ in the small-voltage bias range for the array with $N=255$ islands. The dependence on the external flux has been fitted according to Eqs. (6) and (7). The value of $G_0$ is determined by the offset of the conductance at ${\Phi }_0/2$, and $\gamma$ captures the magnitude of the $E_J^2$-dependent contribution. (c) Data points showing the differential conductance $G_V$ at temperature 20 mK as a function of flux, in the case of $N=255$. The solid line is a fit according to $G_V\propto {\cos }^2(\pi \Phi /{\Phi }_0)$.Reuse & Permissions

Figure 3

Temperature dependence of the zero-bias conductance $G_T$ for both arrays. (a) ${E_J}^2$-dependent part [$\gamma \left(T\right)$], $\alpha =3.5$. (b) Flux-independent part. The lines represent fits according to Eq. (4), $\alpha =1$. The conductance of the array with $N=59$ islands shows a temperature-independent plateau.Reuse & Permissions