In 2020, Shabir Barzanjeh and colleagues at TU Wien carried out the first experimental demonstration of quantum illumination in the microwave band. Using a superconducting device called a Josephson parametric converter (JPC), they generated entangled pairs of signal and idler photons to detect a room-temperature object one meter away. The results only marginally outperformed classical noise radar—but the heart of the setup sat inside a dilution refrigerator at 7 millikelvin above absolute zero, or roughly -273.14°C.
To put that temperature in perspective: it's an extreme environment reached by evaporating liquid helium even further, produced by a refrigerator the size of a desk that costs on the order of tens of millions of yen to operate. The theoretical advantages of quantum radar and quantum communication have been known since the 1990s, but the biggest obstacle to practical deployment was never the underlying physics—it was this engineering constraint that the system simply wouldn't work without deep cooling.
The basic principle behind generating correlated microwave photon pairs is straightforward. When a single input photon is split into two photons through a parametric process, conservation of energy dictates that the sum of the two output frequencies equals the input frequency, and their phases become synchronized with each other. Send one of these "twin photons" out as a signal while keeping the other on hand, then perform a correlation measurement at the receiving end, and you can extract a weak signal buried in noise. Quantum radar, jamming-resistant communication, and quantum simulators are all designed around this principle of correlated photon pairs.
The catch was that the device responsible for this splitting had to be a superconductor. Maintaining a superconducting state requires cooling below a critical temperature, and for niobium- or aluminum-based JPCs, that standard falls in the millikelvin range. In other words, the "seed" of correlated photon pairs—the foundation of quantum technology—could only sprout in the "special soil" of extreme cold.
A Different Path Opened by Magnetic Nonlinearity
While pursuing a separate line of research, the group led by Associate Professor Luqiao Liu in MIT's Department of Electrical Engineering and Computer Science stumbled onto an unexpected fact: superconducting circuits aren't strictly necessary to generate correlated microwave signals. A magnetic material—an ordinary magnet—can do the same job.
Inside a magnetic material, electron spins undergo collective precession. The quantized quasiparticle of this collective excitation is called a magnon. When a microwave photon is injected into the magnetic material, a parametric process occurs in which one photon splits into two magnons. This phenomenon itself has been known since the 1960s from studies of ferromagnetic resonance.
However, conventional parametric magnon generation has a fatal limitation: the two resulting magnons share the same frequency (they are degenerate), so there's no way to distinguish and extract them by frequency. Even if you wanted to use one as a signal and the other as a key, you can't separate two overlapping waves of identical frequency.
Liu's team brought the physics of "cavity magnonics" to bear on this problem. When a magnetic thin film is placed inside a microwave resonator and the magnon mode is strongly coupled to the resonator's photon mode, level repulsion occurs—a phenomenon in which the frequencies of two modes, as they approach each other, instead push apart, as if avoiding one another.
The research team designed a microstrip resonator fabricated on a printed circuit board. This resonator supports two resonances—a half-wavelength mode (around 3.354 GHz) and a full-wavelength mode (around 6.709 GHz)—with dimensions precisely tuned so that the two frequencies are exactly related by a factor of two. A YIG thin film was placed in a narrow section of the resonator (40 micrometers wide), and a pump microwave at 6.708 GHz, corresponding to the full-wavelength mode, was injected at an output power of 12 dBm.
Level repulsion then split the frequencies of the two magnon polaritons (hybrid quasiparticles of magnons and photons), yielding a signal mode at 3.3532 GHz and an idler mode at 3.3548 GHz. The frequency difference is a mere 1.6 MHz, but that small gap is enough to distinguish the "twins." And crucially, this entire process happens at room temperature.
A 475-Pixel Image That Survived the Noise
Using the generated correlated signal pair, Liu's team conducted a communication experiment. Information was encoded onto the signal channel using QPSK (quadrature phase-shift keying) modulation and transmitted from an antenna at a modulation frequency of 500 Hz. On the receiving end, a correlation operation with the idler channel was used to decode the information.
Each channel's phase is completely random on its own. An eavesdropper intercepting only the signal channel would obtain nothing but random phase noise. Recovering the information requires the corresponding idler channel—the "key."
The team transmitted a 475-pixel binary image using this scheme. Although noise was superimposed on both channels, the correlation operation eliminated the noise, and the image was reconstructed with zero bit errors.
As Liu explained in an MIT press release: "Magnon systems display remarkably rich nonlinear dynamics, but these nonlinearities haven't been put to practical use as extensively as those in nonlinear optics or other dynamical systems. In this work, we tackled a key challenge—the spectral overlap of the 'twin' magnons generated from the same pump photon. By exploiting the level repulsion arising from the coupling between magnons and microwave photons, we were able to separate the two magnons in frequency."
Professor Can-Ming Hu of the University of Manitoba, who was not involved in the research, called the achievement "an important milestone for cavity magnonics," adding that it "extends the field beyond the stage of generating incoherent microwaves and toward the production of multi-channel correlated microwave photons."
Structural Differences From the Superconducting Approach
To clarify where this research stands relative to existing sources of correlated microwaves, here is a comparison:
| Item | JPC (Superconducting Approach) | This Study (Cavity Magnon Approach) |
|---|---|---|
| Operating temperature | ~7 mK (near absolute zero) | Room temperature (~300 K) |
| Cooling equipment | Dilution refrigerator (large, expensive) | Not required |
| Device substrate | Superconducting circuit (niobium, etc.) | Printed circuit board + YIG thin film |
| Output signal | Entangled photon pairs (quantum regime) | Correlated polariton pairs (currently classical regime) |
| Frequency separation | Can be made non-degenerate by design | Made non-degenerate via level repulsion (new finding of this study) |
| Demonstration | Quantum illumination (Barzanjeh 2020, 1 m detection) | QPSK communication + image transmission (475 pixels, zero errors) |
| Scalability | Roughly one chip per refrigerator | Highly compatible with PCB manufacturing processes |
The difference in operating temperature spans roughly a factor of 43,000 (7 mK versus 300 K). That gap translates directly into differences in device size, cost, and power consumption.
There is, however, an important caveat. The current work demonstrates operation in the "classical regime." At room temperature, the thermal photon number at 3.35 GHz reaches roughly 1,000 photons per mode, and any quantum entanglement would be buried in thermal noise. The research team states in their paper that "lowering the operating temperature to suppress the thermal contribution would make it possible to establish and detect quantum entanglement between the two modes, enabling unconditional security"—explicitly identifying the leap into the quantum regime as future work.
What a Serendipitous Discovery Reveals
This research originated from a chance observation made while Liu's group was pursuing an entirely different topic. While investigating the nonlinear response of a magnetic material, they noticed that correlated signals were being generated at room temperature. As Liu put it in the press release, "We want to open a path toward room-temperature quantum simulators." Quantum simulators are devices that mimic the complex interactions of subatomic particles—interactions that classical computers cannot handle—and are used in the search for new drugs and materials. Correlated microwave signals are a building block of such simulators, and being able to generate them at room temperature offers major advantages in both cost and scale.
The team's next goals are to extend the architecture toward scalability and to further explore the fundamental physics of correlated microwave signals. Wang and Liu have also filed a patent disclosure with the MIT Technology Licensing Office, and MIT is pursuing patent protection.
From Room Temperature to Quantum: The Distance That Remains
For this device to reach true quantum technology—quantum cryptography with unconditional security, or radar with genuine quantum advantage—at least two more steps are needed. First, the device must be cooled to suppress the thermal photon count and demonstrate a two-mode squeezed (entangled) state. Second, detection and control at the single-magnon level must be achieved.
The field of cavity magnonics itself has developed rapidly since strong coupling between YIG spheres and microwave cavities was first reported around 2014. But practical applications in the nonlinear regime are only now taking their first step with this research. Professor Hu's remark that the work "may be remembered as the starting point for quantum-inspired microwave sensing and communication technologies based on nonlinear cavity magnonics" reflects exactly this position in the field's timeline.
The fact that correlated signals can be obtained at room temperature points to one concrete pathway for quantum technology to move from "specialized laboratory apparatus" to "real-world infrastructure." What lies at the end of that path will be revealed by the results of the next low-temperature experiments.