(a) Scanning electron micrograph of the structure used to measure tunnel junction characteristics. It is an array of seven parallel lines with 64 tunnel junctions in series in each vertical line. Cooling fins are added to the islands between junctions to facilitate low temperature measurements with minimal decoupling of the electrons to the phonon system. (b) Two tunnel junctions of the array with higher magnification. The junction area is nominally .
Resistance of three arrays with tunnel junctions manufactured with double oxidation. All the three samples were fabricated in the same vacuum cycle in batch 3 (see Table I). The resistance was monitored over an extended time while the junctions were stored in ambient air at room temperature. A small decrease of resistances is seen in the first days but then the junctions stabilize. The overall change in resistance is over this duration.
Measured data (○) of dynamic conductance plotted vs voltage over one junction. Fitting a parabola (solid line) gives, using the Simmons model, an effective barrier width of and a barrier height of . This measurement was performed at . The measurement at gave almost identical results apart from the Coulomb blockade feature around zero bias.
Measured (○) normalized conductance vs voltage across junctions. Fitting with expression for conductance (solid line) gives together with a temperature of (Ref. 8). The measurement was carried out in a magnetic field of approximately oriented perpendicular to the metal film.
Samples fabricated and characterized. Note that the two first batches were fabricated with the standard single oxidation of aluminum. In the table, is the characteristic resistance at room temperature and at . is defined as .
Article metrics loading...
Full text loading...