⚛️ quantum physics

Noise Mitigation in Single Microwave Photon Counting by Cascaded Quantum Measurements

This paper presents a new superconducting single microwave photon detector that utilizes cascaded four-wave mixing processes and dual-qubit encoding to mitigate noise, achieving a record-low power sensitivity of $2.4 \cdot 10^{-23}\mathrm{W}/\sqrt{\mathrm{Hz}}$ over a 230 kHz bandwidth.

Original author: Alexandre S. May, Leo Sutevski, Jeanne Solard, Gil Cardoso, Léon Carde, Louis Pallegoix, Raphael Lescanne, Denis Vion, James O'Sullivan, Patrice Bertet, Emmanuel Flurin

Published 2026-11-04

📖 4 min read🧠 Deep dive

CC BY 4.0

Original author: Alexandre S. May, Leo Sutevski, Jeanne Solard, Gil Cardoso, Léon Carde, Louis Pallegoix, Raphael Lescanne, Denis Vion, James O'Sullivan, Patrice Bertet, Emmanuel Flurin

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the author. For technical accuracy, refer to the original paper. Read full disclaimer

In the invisible world of quantum physics, scientists are constantly trying to listen to the faintest whispers of nature. For decades, detecting single particles of light, known as photons, has been a cornerstone of optical science, enabling technologies from advanced medical imaging to the most secure forms of computing. However, light comes in many forms, and the photons used in modern quantum computers are not the bright, visible beams we see with our eyes. These are microwave photons, carrying vastly less energy than visible light. Because they are so weak, detecting a single one of them is incredibly difficult; at room temperature, the heat of the environment creates a roar of noise that completely drowns out the signal. To hear these faint whispers, researchers must cool their equipment to temperatures near absolute zero, colder than the depths of outer space. Even then, the challenge remains: distinguishing a single, precious microwave photon from the random thermal noise that naturally exists in the system.

A team of researchers has now developed a new way to listen to these whispers with unprecedented clarity. They have built a device that can count individual microwave photons with a sensitivity never before achieved. The key to their success lies in a strategy of redundancy. Instead of relying on a single sensor to make the decision, their device uses two separate quantum bits, which act as tiny, ultra-sensitive flags. When a photon enters the system, it triggers a chain reaction that must flip both flags to confirm the detection. If only one flag flips, the system assumes it was a random glitch and ignores it. This simple but powerful idea allows the device to filter out a vast amount of false alarms caused by random errors, leaving behind a clear signal of the actual photon.

The researchers, working with devices cooled to about ten millikelvin, demonstrated that this new architecture, which they call a cascaded single-microwave photon detector, can operate with a record-breaking level of precision. In their experiments, they measured the device's ability to detect power and found it could sense signals as weak as 2.4 times 10 to the power of negative 23 watts per square root of hertz. This number represents the lowest power sensitivity ever recorded for microwave photon counting. To put this in perspective, the device is so sensitive that it can detect the presence of a single photon even when the background noise is incredibly low, a feat that was previously impossible with single-sensor designs.

The path to this achievement involved overcoming significant hurdles related to noise. In previous attempts, the detectors were often fooled by thermal photons—random packets of energy from the heat of the environment—or by the sensors themselves making mistakes. The new design addresses these issues in two distinct ways. First, by using two sensors that must agree, the device effectively cancels out the errors that happen independently in each sensor. Second, the researchers tuned the device to operate at a higher frequency, around 11.7 gigahertz. Because higher frequency photons carry more energy, they are less likely to be mimicked by the random thermal noise that plagues lower frequencies. This combination of a dual-sensor check and a higher operating frequency allowed the team to suppress the background noise to levels where the true signal could finally be heard.

The team tested their device by sending a stream of photons into the detector and watching how often it correctly identified them versus how often it falsely claimed to see one. They found that while the device correctly identified photons about half the time, it made false alarms only about five times per second. This is a dramatic improvement over older designs, which would make false alarms hundreds of times per second. The researchers also showed that they could adjust the range of frequencies the device listens to by changing the strength of the control signals they applied, allowing them to tune the detector for different tasks.

Beyond the technical achievement, this work opens the door to new scientific possibilities. The ability to measure such low levels of microwave energy means scientists can now use these detectors as primary thermometers, measuring the temperature of the microwave environment itself with extreme accuracy. This is crucial for understanding the behavior of quantum computers, where even tiny amounts of heat can disrupt the delicate calculations being performed. The researchers also noted that their device could be useful for detecting dark matter candidates or for studying the magnetic properties of individual atoms. By proving that they can reliably count single microwave photons with such low noise, the team has provided a powerful new tool for exploring the quantum world, turning a once-noisy environment into a quiet space where the faintest signals can be studied in detail.

Technical Summary: Noise Mitigation in Microwave Photon Counting by Cascaded Quantum Measurements

Problem Statement
Single-photon detection in the microwave domain is a critical capability for quantum sensing, dark matter searches, and quantum computing protocols. However, achieving high sensitivity is challenging due to the low energy of microwave photons, which necessitates millikelvin operating temperatures. Existing Single Microwave Photon Detectors (SMPDs), particularly those based on superconducting qubits and four-wave mixing (4WM), face a trade-off between operational efficiency and noise. The sensitivity of these devices is limited by two primary factors: the dark count rate (false positives) and thermal noise at the detector input. Previous implementations, while achieving high efficiency, were often constrained by residual qubit excitation errors and thermal photons, limiting their power sensitivity to approximately $10^{-21}$ W/ $\sqrt{\text{Hz}}$ .

Methodology
The authors propose and demonstrate a new architecture called the Cascaded Single Microwave Photon Detector (cSMPD). This design introduces redundancy into the detection encoding to mitigate noise.

Architecture: The cSMPD utilizes two spatially separated transmon qubits coupled via a "memory" resonator mode. The system employs two cascaded 4WM processes. An incoming photon in the "buffer" mode is parametrically converted to the "memory" mode by the first qubit (Q1), and subsequently from the memory to the "waste" mode by the second qubit (Q2).
Detection Logic: A successful detection event is defined only when both qubits are found in their excited states ( $|e_1e_2\rangle$ ) following the interaction. All other outcomes ( $|g_1g_2\rangle$ , $|e_1g_2\rangle$ , $|g_1e_2\rangle$ ) are discarded.
Noise Mitigation Strategy: The core hypothesis is that spurious qubit excitations (false positives) are largely uncorrelated between the two qubits. By requiring a coincidence of excitation, the cSMPD suppresses the false positive rate quadratically ( $\alpha_{err} \propto \alpha_{err,1} \times \alpha_{err,2}$ ), whereas thermal photons, which trigger the actual 4WM process, remain a dominant noise source.
Device Implementation: Two devices (cSMPD1 and cSMPD2) were fabricated using tantalum and aluminum Josephson junctions. cSMPD2 operates at a higher buffer frequency (11.722 GHz) compared to cSMPD1 (8.798 GHz) to reduce thermal photon occupation. The devices operate in a cyclic mode involving detection, readout, and conditional reset.

Key Contributions and Results

Record-Low Power Sensitivity: The cSMPD2 device achieved a record power sensitivity of $2.4 \times 10^{-23}$ W/ $\sqrt{\text{Hz}}$ over a 230 kHz bandwidth. This represents a significant improvement over previous state-of-the-art devices.
Dark Count Suppression: The correlated detection scheme successfully reduced the dark count rate to $5.2 \pm 0.2$ s $^{-1}$ for cSMPD2, compared to hundreds of counts per second for individual qubits. The analysis confirms that the majority of these residual dark counts are due to thermal photons rather than uncorrelated qubit errors, validating the noise mitigation strategy.
Operational Efficiency: The device maintains a high operational efficiency of $\eta \approx 0.49 \pm 0.01$ (up to 0.5 in optimal conditions), comparable to single-qubit SMPDs, despite the added complexity of the cascaded process.
Bandwidth Tunability: The detection bandwidth ( $\kappa_d$ ) is tunable via the pump amplitude, ranging from approximately 100 kHz to 1 MHz, allowing optimization for specific applications.
Thermal Photon Measurement: The high sensitivity of the cSMPD enabled the measurement of thermal photon occupation numbers as low as $n_{th} \approx 5 \times 10^{-6}$ in the input line, corresponding to an effective temperature of 46 mK. This is approximately two orders of magnitude lower than previous measurements in circuit quantum electrodynamics.
Frequency Dependence: The study highlights that increasing the detection frequency (from ~7.5 GHz to 11.7 GHz) significantly reduces the thermal photon contribution, suggesting a path toward further sensitivity improvements.

Significance
The paper claims that the cSMPD architecture represents a major advancement in microwave photon counting by effectively decoupling the detection efficiency from the noise floor through redundant encoding. By suppressing uncorrelated qubit errors, the device shifts the noise limitation to thermal photons, which can be further mitigated by increasing the operating frequency or improving thermalization. The authors assert that this technology enables applications requiring extreme sensitivity, such as single-spin magnetic resonance, dark matter detection, and primary microwave thermometry. They conclude that with further improvements in reducing correlated errors and thermalization, power sensitivities better than $1 \times 10^{-23}$ W/ $\sqrt{\text{Hz}}$ are achievable in the near future.

More like this