Dual-Resonator Kinetic-Inductance Detector for Distinction between Signal and 1/f Frequency Noise

Abstract

Astronomical Kinetic Inductance Detectors (KIDs), similar to quantum information devices, experience performance limiting noise from materials. In particular, 1/f (frequency) noise can be a dominant noise mechanism, which arises from Two-Level System defects (TLSs) in the circuit dielectrics and material interfaces. Here we present a Dual-Resonator KID (DuRKID), which is designed for improved signal to noise (or noise equivalent power) relative to 1/f-noise limited KIDs. We first show the DuRKID schematic, fabricated circuit, and we follow with a description of the intended operation, first measurements, theory, and discussion. The circuit consists of two superconducting resonators sharing an electrical capacitance bridge of 4 capacitors, each of which hosts TLSs. The device is intended to operate using hybridization of the modes, which causes TLSs to either couple to one mode or the other, depending upon which capacitor they reside in. In contrast, the intended KID signal is directed to an inductor, and due to hybridization this causes correlated frequency changes in both (hybridized) modes. Therefore, one can distinguish photon signal from TLS frequency noise. To achieve hybridization, a TiN inductor is current biased to allow tuning of one bare resonator mode into degeneracy with the other and measurements show that the intended resonator modes frequency tune and hybridize as expected. The interresonator coupling and unintentional coupling of the 2 resonators to transmission lines are also characterized in measurements. In the theory, based on a quantum-information-science modes, we calculate the 4-port S parameters and simulate the 1/f frequency noise of the device. The study reveals that the DuRKID can exhibit a large and fundamental performance advantage over 1/f-noise-limited KID detectors.

Similar papers