Stanford Physicists Observe Real-Time Quantum Jumps of Sound, Unlocking Acoustic Error Correction

To correct an error in a quantum computer, the system must first detect the exact moment a fragile quantum state collapses without destroying the data in the process. That condition has long eluded researchers working with acoustic quantum architectures, but it now holds. A team of physicists at Stanford University has directly observed a single quantum of sound—a phonon—abruptly dropping from one energy state to another in real time, demonstrating the measurement capability required for sound-based quantum error correction.[1][2]

The research, published in the journal Science on September 17, 2026, marks the first time scientists have tracked an individual phonon making a discrete quantum jump. While sudden transitions between energy states were demonstrated in trapped ions in 1986 and in photons in 2007, sound represents a far more complex mechanical system. A phonon is not a singular particle like a photon, but rather the coordinated movement of a large group of atoms.[1][2]

In the macroscopic world, the vibration of a struck bell fades gradually into silence. At the quantum scale, however, the energy changes in abrupt, discrete steps. To capture this century-old theoretical prediction, the Stanford team, led by applied physics professor Amir Safavi-Naeini, fabricated a microscopic mechanical resonator capable of vibrating for roughly two milliseconds. At that microscopic scale, two milliseconds is an exceptionally long "ringdown" time.[1]

The binding constraint in quantum mechanics is that observing a state typically destroys it. To bypass this hurdle, co-first authors Takuma Makihara and Erik Szakiel paired the mechanical resonator with a superconducting qubit. This qubit acted as an electrical detector, repeatedly checking whether the phonon's energy level was at state 1 or had dropped to state 0, without absorbing the acoustic energy itself.[1][2]

“We had to continually develop new processes to make this extremely long-lived, vibrating object and then integrate it with the qubit, which is our little electrical detector—without ruining either subsystem,” Makihara explained. The setup allowed the team to take hundreds of measurements within the two-millisecond window, pinpointing the exact moment the phonon vanished.[1]

While university press materials often highlight a broad range of future applications for such discoveries, the immediate utility of the Stanford breakthrough lies in error detection. In many quantum computing architectures, an unexpected quantum jump represents a computational error. If a system cannot recognize that a phonon has dropped to zero, the error propagates and ruins the calculation.[2]

By proving that a superconducting qubit can monitor a mechanical resonator and detect a missing phonon in real time, the team has laid the groundwork for a "dual-rail" mechanical qubit. In such a proposed design, two resonators would be coupled to a shared superconducting qubit. A lost phonon would register as a detectable erasure error, prompting the system to correct the fault before the stored quantum information degrades.[2]

However, the technology remains a proof-of-principle demonstration rather than a deployable commercial processor. Safavi-Naeini noted that while each individual measurement preserves the phonon number about 99 percent of the time, the sheer volume of measurements required means that the small disturbances eventually accumulate. The team has not yet demonstrated active error correction, only the detection mechanism that makes it possible.[1][2]

Beyond computing, the ability to monitor mechanical quantum states without disturbing them opens avenues for highly sensitive biological sensors. Safavi-Naeini’s group is currently collaborating with researchers at Caltech to adapt the resonator platform to detect and identify individual proteins inside cells, leveraging the extreme sensitivity of the acoustic quantum states.[2]

“What this study shows will allow us to move forward with developing new quantum technologies with sound,” said Safavi-Naeini. For the acoustic quantum computing sector, the milestone shifts the engineering challenge from proving that phonons jump to managing those jumps at scale. The next verifiable checkpoint will be integrating multiple mechanical resonators onto a single chip to demonstrate that real-time error detection can operate across an array.[1]