Low-Frequency Noise Performance of Microstrip-Coupled Lumped-Element Aluminum KIDs using Hydrogenated Amorphous Silicon Parallel-Plate Capacitors for NEW-MUSIC

TL;DR

The paper addresses whether Al/a-Si:H MS-PPC-LEKIDs can operate at sub-Hz modulation with photon-noise-limited performance for NEW-MUSIC. It introduces a two-tone IQ measurement approach to suppress low-frequency electronics noise and characterizes the detectors' low-frequency PSDs, applying TLS-noise scaling from prior GHz measurements to predict low-frequency behavior. The results show GR noise dominates dark spectra down to $0.1$ Hz, while TLS noise remains below extrapolated predictions, and under optical loading the detectors are expected to be GR+photon-noise limited down to tenths of a Hz. Overall, the work demonstrates the viability of a-Si:H PPC-based KIDs for low-modulation-rate astronomy and supports their use in NEW-MUSIC as photon-noise-limited detectors across observing conditions.

Abstract

We present measurements of the low-frequency noise of microstrip-coupled, lumped-element aluminum kinetic inductance detectors that use hydrogenated amorphous silicon parallel-plate capacitors (Al/a-Si:H MS-PPC-LEKIDs), which are under development for the Next-generation Extended Wavelength Multiband Submillimeter Inductance Camera (NEW-MUSIC). We show that, under dark conditions, these devices are generation recombination (GR) noise dominated down to 0.1 Hz and, under optical load, they are likely dominated by GR and photon noise down to tenths of a Hz and possibly lower, both in spite of the use of a-Si:H PPCs. Our measurements set limits on the low-frequency two-level-system (TLS) noise of the a-Si:H material that are consistent with higher frequency measurements in the 0.1-10 kHz regime. These results establish that our MS-PPC-LEKID design for NEW-MUSIC will be photon-noise-limited under a range of observing conditions and, more generally, that a-Si:H PPC-KIDs are a viable new detector technology for even low modulation-rate applications such as astronomy.

Figures (4)

• Figure 1: Two-tone IQ measurement setup. I and Q signals at tens to hundreds of kHz produced by the arbitrary waveform generator (AWG) are up-mixed with a local oscillator (LO) signal from a RF synthesizer. The two-tone signal is sent into the cryostat to probe the KID on and off resonance. The output signal is down-mixed with a copy of the original LO, amplified and anti-alias filtered at baseband, and digitized. Final demodulation of the two tones is done in software. The scheme enables simultaneous acquisition of on- and off-resonance tones to enable removal of correlated multiplicative electronics noise. It also mitigates baseband amplifier additive low-frequency noise.
• Figure 2: TLS noise performance of a reference Nb superconducting resonator incorporating a-Si:H dielectric PPCs defrance. Noise PSDs are shown for an 847 MHz resonator. Data were fit with $a f^{-0.5}+b$ to capture both TLS and white noise.
• Figure 3: Dark noise data and TLS noise models. Blue: Noise PSD under dark conditions for an Al MS-PPC-LEKID ($f_r = 253.83$ MHz). Red and green: scaling of TLS noise measured for a-Si:H PPCs in defrance to this resonator's PPC area and electric field for the feedline readout power used. Since defrance only measured noise above 100 Hz, below 100 Hz we show both the Eq. 2 broken power law and a simple $f^{-1}$ extrapolation (red and green, respectively). Mustard and yellow-green: approximate fits of these two models to the measured dark noise PSDs, from which we derive approximate upper limits on TLS noise. It is not certain the low-frequency rise is TLS noise, and there is good reason to think it is not (see text).
• Figure 4: Noise PSD under optical load, compared to dark noise PSD and TLS noise estimates. Blue solid: GR+photon noise measured above 100 Hz under $T_{Load} = 180$ K. Blue dashed: extrapolation of that noise level to low frequency. Grey solid: dark noise PSD. Grey dashed: TLS noise model curves from Fig. \ref{['FIG:dark_load']}, unmodified. Colored dashed: the same TLS noise model curves extrapolated to the operating conditions under optical load (i.e., scaling by electric field magnitude): they increase because the stored power decreases. Discussion is provided in the text.