A quantum charge sensor gets a strong signal from a faint microwave probe

Reading charge without adding much heat

Detecting individual electrons matters for studying electrical transport and reading quantum devices. But a detector is not a passive observer: its measurement signal can heat the sample or alter the electron movements it is supposed to record. This disturbance is called back-action.

In an arXiv preprint, Samuel Andersson and colleagues demonstrate a charge detector that uses coherent interaction between a double quantum dot and a microwave resonator. They report distinguishing two charge states with a signal-to-noise ratio of 11.7 using 16 microseconds of averaging per measurement, at an input power of −129 dBm.

The result is not a demonstration of a quantum-limited detector or nanosecond readout. Its central achievement is a strong charge-sensitive response at unusually low probe power.

Turning one electron into a resonance shift

A quantum dot confines electrons in a small region. Here, two coupled dots form the detector, while a nearby single dot provides the charge being measured. The dots are defined in an indium arsenide nanowire, and the detector is coupled to a microwave resonator built using Josephson junctions.

When an extra electron enters the nearby single dot, electrostatic repulsion shifts the relative energies of the detector’s two dots. That changes how their quantum states mix with the resonator’s microwave states. This mixing is called hybridisation.

The researchers first mapped the microwave response while tuning the double dot. They observed avoided crossings: regions where the coupled states separate rather than crossing as their uncoupled energies would. These features establish the interaction underlying the sensing scheme.

At the chosen operating point, adding one electron shifts the detector resonance by about 22 megahertz, larger than its roughly 11-megahertz linewidth. A fixed-frequency microwave probe therefore encounters a resonant detector for one charge state and an off-resonant detector for the other. The reflected signal changes substantially, allowing the two states to be distinguished.

Watching electrons arrive and leave

The team then recorded the signal as electrons tunnelled into and out of the nearby dot. The traces showed transitions between two levels, corresponding to charge states differing by one electron.

For the 16-microsecond measurements, the signal separation was 260 nanovolts and the root-mean-square noise was 22 nanovolts. Their ratio gives the reported signal-to-noise value of 11.7. This is a measure of signal separation relative to noise, not a percentage readout accuracy.

The authors translate that sensitivity into an estimated averaging time of 120 nanoseconds for a signal-to-noise ratio of one. That estimate should not be confused with the clearly separated charge states demonstrated in the longer measurements.

Compared with specific earlier high-sensitivity sensors discussed in the paper, the device used more than a thousand times less input power while reaching comparable estimated sensitivity. This is a comparison with selected published results, not a claim that it outperforms every detector on every measure.

Testing disturbance—and finding limits

The measured charge occupancy was consistent with an electron temperature of 69 millikelvin. Increasing the probe power fourfold left the results essentially unchanged over the central energy range examined, supporting the authors’ conclusion that readout caused no significant heating under those conditions.

They also found no significant evidence of photon-assisted tunnelling, in which measurement photons help electrons cross an energy barrier. Neither observation establishes that all back-action vanishes.

The operating scheme offers two forms of protection. In one charge state, the resonator and double dot are far apart in frequency, limiting excitation of the dot. In the other, the shifted detector resonance causes most probe photons to reflect before entering. Calculations support substantial suppression of probe-induced charge fluctuations, but those estimates are not a direct measurement of every disturbance channel.

Faster operation remains a proposal

The present resonator becomes nonlinear as power increases: its resonance shifts and its response weakens. That prevents the researchers from simply increasing probe strength to accelerate readout. Amplifier noise is another limitation.

The authors propose a more linear resonator and a lower-noise amplifier. Their model projects sub-nanosecond sensitivity under improved conditions, rather than demonstrating it experimentally. Very fast events can also be missed in the current recordings, while finite recording duration limits estimates of rare events.

The demonstrated advance is narrower but useful: coherent quantum-dot–resonator coupling can turn a faint microwave probe into a strong single-charge signal, with little observed heating in the tested regime.