[Paper Review] Membrane-less phonon trapping and resolution enhancement in optical microwave kinetic inductance detectors

Nicholas Zobrist, W. Hawkins Clay+5 more|arXiv (Cornell University)|Apr 28, 2022

Superconducting and THz Device Technology45 references27 citations

TL;DR

This paper demonstrates a membrane-free bilayer design in optical microwave kinetic inductance detectors (MKIDs) that doubles the energy resolution by trapping phonons via acoustic impedance mismatch, achieving a resolving power of ~20 at 1 µm without requiring complex membrane fabrication. The improvement arises because high-energy phonons cannot enter the second layer due to unavailable phonon states, reducing phonon escape losses and enhancing quasiparticle creation efficiency.

ABSTRACT

Microwave Kinetic Inductance Detectors (MKIDs) sensitive to light in the ultraviolet to near-infrared wavelengths are superconducting micro-resonators that are capable of measuring photon arrival times to microsecond precision and estimating each photon's energy. The resolving power of non-membrane MKIDs has remained stubbornly around 10 at 1 $\mu$m despite significant improvements in the system noise. Here we show that the resolving power can be roughly doubled with a simple bilayer design without needing to place the device on a membrane, avoiding a significant increase in fabrication complexity. Based on modeling of the phonon propagation, we find that the majority of the improvement comes from the inability of high energy phonons to enter the additional layer due to the lack of available phonon states.

Motivation & Objective

• To overcome the limited energy resolution (~10) in membrane-free optical MKIDs despite low system noise.
• To eliminate the need for complex membrane-based fabrication while maintaining or improving phonon confinement.
• To achieve high resolving power (R ~ 100) required for exoplanet atmospheric spectroscopy in space-based integral field spectrographs.
• To identify a scalable, high-quantum-efficiency detector architecture suitable for kilo- to mega-pixel arrays.

Proposed method

• Design a bilayer MKID structure with a hafnium superconductor layer on a sapphire substrate, with an additional low-impedance layer (e.g., indium or PMMA) to block phonon escape.
• Use acoustic impedance mismatch to reflect phonons back into the sensor layer, increasing their dwell time and chance of down-conversion into detectable quasiparticles.
• Model phonon propagation using a ray-tracing approach to estimate the phonon escape time τ_esc and the effective loss factor J = τ_pb / τ_esc.
• Introduce a modified Fano limit equation incorporating phonon loss: R_phonon = (1/2)√(2 ln 2) × √(η_pb E / [Δ(F + J)]), where J quantifies phonon escape losses.
• Measure resolving power experimentally on a 220 nm hafnium MKID on sapphire, with J ≈ 13, and compare to theoretical limits.
• Evaluate the impact of non-uniform current density and two-level systems, confirming they contribute negligibly for R < 40.

Experimental results

Research questions

• RQ1 Can phonon trapping be achieved in a membrane-free MKID design to improve energy resolution?
• RQ2 Does introducing a bilayer structure with acoustic impedance mismatch reduce phonon escape and increase quasiparticle creation?
• RQ3 What is the maximum resolving power achievable in a non-membrane MKID with a high-kinetic-inductance superconductor like hafnium?
• RQ4 How does the phonon loss factor J scale with material properties such as Debye temperature and acoustic impedance?
• RQ5 Can a bilayer design achieve R ~ 100 without the fabrication complexity of suspended membranes?

Key findings

• The bilayer MKID design achieves a resolving power of approximately 20 at 1 µm, roughly doubling the previous membrane-free limit of ~10.
• The improvement is attributed to phonon trapping via acoustic impedance mismatch, which reduces the effective phonon escape loss factor J from 13 to an effective value that enhances resolution.
• Modeling shows that high-energy phonons are blocked from entering the second layer due to unavailable phonon states, increasing their interaction time with the sensor.
• The resolving power is limited by phonon escape (J ≈ 13) rather than intrinsic noise, indicating that further reduction of J could push resolution toward the Fano limit.
• The design avoids the need for membranes, enabling scalable, high-fill-factor arrays with hafnium’s high kinetic inductance and low reflectivity.
• The results demonstrate that membrane-free MKIDs can achieve resolution suitable for detecting molecular absorption bands in exoplanet atmospheres, even at R ~ 25.