Detection mechanisms in superconducting point contacts at frequencies of 44 times the energy gap

Abstract

We have performed mixing experiments with CO2 lasers in the 9 to 11 μm wavelength band using superconducting point contacts. The experiments involve difference frequencies ranging from 19 MHz to 106 GHz. The mixing mechanism definitely appears to require the superconducting state since the beat signal amplitude decreases by as much as 60 dB as the bath temperature is raised through the superconducting transition temperature. At least three conceptually distinguishable mechanisms may be involved in the observed response. These include : (1) a conventional Josephson response, (2) a thermally modulated Josephson response, and (3) a single particle tunneling response. The usual signature of a Josephson interaction is one or more current steps in the I-V curve. Steps, the first of which should be at 67 mV, are not observed in these experiments. However, we do not feel that is a definitive test for the mixing mechanism between two external sources at such high frequencies. In fact, Werthamer's theory of tunnel junctions predicts a large response for these experiments, a response as large as that from two X-band sources, for example. It should be noted, however, that the theory assumes the junction is maintained at a fixed low temperature and whether or not that is the case depends directly on the current density at the contact. Very little is known about the current density at the present time. These experiments emphasize a fact that is usually overlooked in Josephson junction mixing experiments with an external local oscillator. If the contact or tunneling region is sufficiently small, it may be thermally modulated at difference frequencies extending into the microwave range. This effect becomes more likely with increasing applied frequencies, particularly in the infrared. Theoretical modeling for the purpose of calculating ultimate sensitivities of these devices requires a detailed consideration of thermal effects. We fully expect large mixing signals from Josephson junctions even at optical frequencies if thermal effects do not prove to be a limitation.