Full text loading...
(a) Schematic of the SCPT and junction. The regions of two different gaps are indicated. Note that the lower gap region is isolated from the normal metal bond pads (by the higher gap region) to avoid quasiparticle trapping. (b) Energy diagram of the SCPT, showing the potential minima for quasiparticles on the device island. (c) When the junction is biased to quasiparticle extraction occurs from the SCPT reservoir.
(a) Unaveraged gate sweep for the SCPT. (b) A single time trace at odd-integer charge. The points are separated by a time interval of . (c) Histogram of the time trace—solid lines are fits to a Gaussian distribution. (d) Example histogram showing distribution of even times (no quasiparticle on island). The solid line indicates a fit to an exponential giving a time constant of . (e) Example of an odd-time (quasiparticle on island) histogram showing dual-Poissonian distribution. The time constants are and .
(a) The extracted even time constant , normalized to , as a function of junction bias. There is a factor of 2 increase when indicating cooling of one of the reservoirs. A simulation, based on balancing cooling power and heat transfer from the phonon system, is also shown. (b) The behavior of the odd-time constant , normalized to under the same bias conditions.
(a) Simulated cooling power of a junction, at constant temperature, vs bias. Calculation parameters are similar to the experimental case: , , and . (b) The cooling power of the same junction biased close to as the temperature of the material, corresponding to the lead of the SCPT, is reduced.
Article metrics loading...