Reflection-Mode RF-SET

• Reflection-mode RF-SET is a high-sensitivity charge detector that leverages RF reflectometry with a spiral inductor architecture to transduce quantum dot state changes.
• Its multimode design, modeled as a one-dimensional ladder of LC cells, enables simultaneous, frequency-multiplexed measurements with optimized impedance matching.
• Experimental results demonstrate up to 98% single-shot spin-qubit readout fidelity at cryogenic temperatures, highlighting its potential for scalable quantum technologies.

A reflection-mode radio-frequency single-electron transistor (RF-SET) is a high-sensitivity charge detector that leverages radio-frequency (RF) reflectometry to transduce minute changes in quantum-dot device states into measurable voltage signals. In the multimode superconducting inductor architecture, distinct resonance modes arising from distributed inter-turn capacitance and a one-dimensional ladder structure enable simultaneous, frequency-multiplexed reflectometry. Embedding an SET at the termination of this circuit allows for fast and high-fidelity detection of both charge and spin states, with demonstrated single-shot spin-qubit readout fidelities reaching 98% within microsecond-scale integration times (Rivard et al., 4 Dec 2025).

1. Circuit Architecture and Theoretical Foundation

The core of the multimode reflection-mode RF-SET system is a spiral inductor formed from 100 nm NbN on sapphire, utilizing 150 tightly wound turns (1 μm trace width and spacing). Owing to the substantial length and continuity of the superconducting trace, the device cannot be approximated as a simple lumped-element LC resonator. Instead, it functions as a distributed-element system, modeled as a one-dimensional transmission line or a ladder of approximately 30 discrete Lₙ–Cₙ cells. Each cell comprises a series inductance (Lₙ) and a shunt capacitance (Cₙ, combining capacitance to substrate and inter-turn contributions).

The entire spiral supports standing wave modes. For each mode $n$, the effective modal capacitance $C_{n,\mathrm{eff}}$ is set by the standing current antinode at the drive point and nodes elsewhere, leading to mode-dependent impedance matching conditions. Resonant angular frequencies are given by:

$$\omega_n = \frac{1}{\sqrt{L_{\text{tot}} C_{n,\mathrm{eff}}}}$$

where $L_{\text{tot}} = \sum L_n$.

A 50 Ω RF source is coupled to the spiral, which is terminated by the sample impedance $Z_s$ (the SET in parallel with stray capacitance $C_s$). The input impedance $Z_\text{in}(\omega)$ emerges from cascading the ladder and load. The voltage reflection coefficient is:

$$\Gamma(\omega) = \frac{Z_\text{in}(\omega) - Z_0}{Z_\text{in}(\omega) + Z_0}$$

Distinct resonant modes yield minima in $|\Gamma(\omega_n)|$. Critical coupling (maximal power transfer, $\Gamma \simeq 0$) is achieved when $C_{n,\mathrm{eff}}$0, and each mode has a unique matching condition due to differences in $C_{n,\mathrm{eff}}$1.

2. Reflection-Mode RF-SET Operation

The SET, fabricated on a multigate CMOS quantum-dot device, is attached at the far end of the spiral. Its differential resistance ($C_{n,\mathrm{eff}}$2) and parasitic capacitance ($C_{n,\mathrm{eff}}$3) load the tank circuit. When the state of the quantum dot changes—either via charge tunneling events or spin transitions—both $C_{n,\mathrm{eff}}$4 and $C_{n,\mathrm{eff}}$5 experience discrete shifts ($C_{n,\mathrm{eff}}$6, $C_{n,\mathrm{eff}}$7), directly altering $C_{n,\mathrm{eff}}$8 and thus modulating the reflection coefficient $C_{n,\mathrm{eff}}$9.

Charge sensitivity is realized as small resonance frequency shifts (via $$\omega_n = \frac{1}{\sqrt{L_{\text{tot}} C_{n,\mathrm{eff}}}}$$0) and modulation of the depth or width of the reflection minimum (via $$\omega_n = \frac{1}{\sqrt{L_{\text{tot}} C_{n,\mathrm{eff}}}}$$1). The output voltage variation for a change in dot charge $$\omega_n = \frac{1}{\sqrt{L_{\text{tot}} C_{n,\mathrm{eff}}}}$$2 is given by $$\omega_n = \frac{1}{\sqrt{L_{\text{tot}} C_{n,\mathrm{eff}}}}$$3.

Detector performance is characterized by the charge sensitivity ($$\omega_n = \frac{1}{\sqrt{L_{\text{tot}} C_{n,\mathrm{eff}}}}$$4), defined as the rms charge fluctuation for unity signal-to-noise ratio (SNR = 1) in 1 Hz bandwidth:

$$\omega_n = \frac{1}{\sqrt{L_{\text{tot}} C_{n,\mathrm{eff}}}}$$5

where $$\omega_n = \frac{1}{\sqrt{L_{\text{tot}} C_{n,\mathrm{eff}}}}$$6 is the spectral density of output-referred voltage noise. The measurement bandwidth for a selected mode is $$\omega_n = \frac{1}{\sqrt{L_{\text{tot}} C_{n,\mathrm{eff}}}}$$7, set by the loaded quality factor. SNR for pulsed measurements with integration time $$\omega_n = \frac{1}{\sqrt{L_{\text{tot}} C_{n,\mathrm{eff}}}}$$8 is

$$\omega_n = \frac{1}{\sqrt{L_{\text{tot}} C_{n,\mathrm{eff}}}}$$9

with $L_{\text{tot}} = \sum L_n$0 the signal amplitude between states (e.g., singlet vs triplet) and $L_{\text{tot}} = \sum L_n$1 the noise standard deviation over $L_{\text{tot}} = \sum L_n$2.

3. Resonant Modes and System Performance

The spiral inductor naturally produces multiple resonant modes due to inter-turn capacitances, each supporting unique impedance matching and measurement bandwidths. Empirically, four principal modes were characterized (Rivard et al., 4 Dec 2025):

Mode	Frequency (MHz)	Loaded $L_{\text{tot}} = \sum L_n$3
1	150	500
2	222.5	600
3	360.2	720
4	523.8	850–900

The system operates over RF frequencies up to 2 GHz. Charge detection is simultaneously performed on multiple modes by exploiting these distinct resonances. When a barrier gate voltage $L_{\text{tot}} = \sum L_n$4 sweeps the tunnel rate $L_{\text{tot}} = \sum L_n$5 of the quantum dot, each mode selectively probes the same transition at different $L_{\text{tot}} = \sum L_n$6. The response amplitude as a function of $L_{\text{tot}} = \sum L_n$7 follows the dispersive relations:

$L_{\text{tot}} = \sum L_n$8

$L_{\text{tot}} = \sum L_n$9

4. Experimental Procedures and Single-Shot Readout

The cryogenic measurement apparatus incorporates a dilution refrigerator (20 mK base, $Z_s$0 mK), with the signal path including room-temperature RF sources, attenuators, a circulator, the spiral+SET circuit, a cryogenic low-noise amplifier (gain ≈ 40 dB, noise temperature ≃2 K), and subsequent signal processing.

Single-shot singlet–triplet spin readout is demonstrated using mode 2 (245.1 MHz, $Z_s$1) at the (4,0)–(3,1) interdot transition under $Z_s$2 mT. The pulse protocol loads a singlet, permits spin mixing, and measures at the Pauli spin-blockade (PSB) point. The reflected amplitude $Z_s$3 displays two distinct plateaus, corresponding to singlet and triplet outcomes. With integration time $Z_s$4, the separation between signals yields $Z_s$5 ($Z_s$6). Histograms of $Z_s$7 over $Z_s$8 single shots, analyzed via two-Gaussian fits and optimal thresholding, establish a readout fidelity $Z_s$9, where:

$C_s$0

with $C_s$1 the error of assigning "triplet" to a true singlet, and vice versa.

5. Design Principles and Optimization Guidelines

Selection of operational mode is governed by the trade-off between sensitivity and bandwidth: lower modes ($C_s$2–200 MHz) offer greater $C_s$3-driven sensitivity but reduced bandwidth, while higher modes ($C_s$4 MHz) support higher bandwidth with diminished signal amplitude per single-electron event. Optimal charge sensitivity is achieved when the resonant frequency matches the relevant tunnel rate ($C_s$5).

Impedance matching for multiple modes is optimized by tuning the coupling capacitance ($C_s$6) and, if necessary, introducing a series capacitor to flatten $C_s$7. Electromagnetic simulations must incorporate parasitic elements such as on-chip bond-wire inductance and mounting-board stray capacitance.

Noise minimization is essential: implementation of near-quantum-limited cryogenic amplifiers (e.g., JPA, TWPA), rigorous line filtering above a few kilohertz, thermalized attenuators, and infrared-blocking filters are standard measures.

The geometric configuration of the spiral (spacing, trace width) allows for tuning of mode spacing and resonance characteristics. Parasitic inter-turn capacitance, traditionally viewed as a limitation, is repurposed as a functional contributor, driving the presence of higher-order modes and facilitating frequency multiplexing for scalability.

6. Applications and Scalability

This multimode spiral inductor architecture with reflection-mode RF-SET enables flexible impedance matching across a wide frequency (and tunnel-rate) range, rapidly switching between different quantum dots or qubits in a single device. The same spiral can simultaneously support the readout of multiple quantum devices by assigning each to a well-separated resonance below the bandwidth ceiling of the cryogenic low-noise amplifier. The system architecture provides a viable platform for fast spin-qubit readout, high-fidelity quantum nondemolition measurements, and in situ defect spectroscopy, with demonstrated broad tunability and quantum-limited performance (Rivard et al., 4 Dec 2025).