Science1 publisher2 min readPublished
Stanford catches single phonons jumping to zero inside a two-millisecond ringdown
Stanford physicists report in Science the first real-time observation of quantum jumps of sound, using a superconducting qubit to check a mechanical resonator hundreds of times during its two milliseconds of vibration.
The Scientist · Science desk

What happened
- A Stanford team led by the applied physicist Amir Safavi-Naeini reports in Science the first direct observation of quantum jumps of sound in a mechanical resonator.
- The resonator rings long enough that a regular-sized tuning fork with the same capability would sound for several hours, by the group's own comparison.
- Co-first authors Takuma Makihara and Erik Szakiel built the readout by pairing the microscopic resonator with a superconducting qubit that stores quantum information and doubles as the detector.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- capability In many quantum computing architectures a jump is an error and knowing when one happened has been difficult, so a detector that flags the transition while the mode is still ringing gives error-correction schemes something to act on.
- constraint The public account gives no readout fidelity and no jump count, so phonon hardware cannot yet be scored against photonic or ion systems. A mechanical-qubit budget today buys a demonstrated measurement and no demonstrated error rate.
- precedent The resonator is patterned with chipmaking techniques. Many resonators on one chip is the demonstration the field will now expect next.
A phonon, the quantum unit of sound, is the coordinated movement of a large group of atoms [7]. Measuring one runs into a standing problem in quantum engineering: how to get a signal out of a quantum system without disturbing its fragile state [12]. Here the superconducting qubit checks repeatedly during the two milliseconds of vibration whether the phonon is in energy state 1 or 0 [14].
Hundreds of those checks fit inside the ringdown, and they are what locate the moment the vibrational energy fell from 1 to 0 [10]. Divide two milliseconds by 200 readings and each one has 10 microseconds; divide by 1,000 and each has 2 [11]. Earlier experiments had found evidence that phonons jump, and this is the first record of individual phonons making the transition as it happened [6].
Building a very long-lived vibrating object and a working electrical detector on the same device was the fabrication problem. Makihara, a recent Stanford doctoral graduate, said the team "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," and had to do that "without ruining either subsystem" [15].
Safavi-Naeini framed the result in terms of what the platform now permits. "We have seen that vibrating objects can exhibit quantum behavior, which is the prerequisite for many of the operations needed by quantum computing and sensing," said Safavi-Naeini, associate professor of applied physics in the Stanford School of Humanities and Sciences [16]. He also said the work "will allow us to move forward with developing new quantum technologies with sound" [17].
The phys.org account calls the study a foundational first step [22], and it does not report a readout fidelity, the number of jumps recorded, or any comparison with photonic or trapped-ion hardware [23]. Those comparison classes have long head starts. Trapped ions showed quantum jumps in 1986 and photons in 2007, 21 years later [3][4][5], against a theory that dates to the early 1900s [2].
The nearer application is sensing, where the small size and high sensitivity of the resonator and qubit together are the draw [20]. Safavi-Naeini's group is already working with the physicist Michael Roukes and his team at Caltech to use the platform to detect and identify proteins within cells [21].
What to watch
- Whether the Science paper and its supplement give a readout fidelity and jump statistics that allow a comparison with photonic and trapped-ion hardware.
- Whether the protein-identification collaboration reports a measured signal from inside a cell. That is the first real test of the sensing claim.
- Whether the group shows jump detection on more than one resonator sharing a chip.