Niobium diselenide, a layered superconductor, stayed superconducting at a slightly higher temperature when it was placed inside a resonator built to amplify the quantum fluctuations of empty space. In a six-layer device the critical temperature rose by up to 5.4 percent. The team at the University of Science and Technology of China, led by Changgan Zeng and Guanghui Cheng, reported it in Nature with MIT’s Frank Wilczek among the co-authors.
The figure is an upper bound for one device, and the team describes it as the first experimental observation of vacuum-fluctuation-enhanced superconductivity.
Terahertz dark cavity and the 5.4% rise
The experiment is described in the Chinese Academy of Sciences’ English-language release, dated August 19, 2026. The USTC group embedded a few-layer flake of NbSe2 in a terahertz split-ring resonator, the device the team calls a terahertz dark cavity. Fluctuations of the electromagnetic vacuum are normally far too weak to matter in a solid. The cavity concentrates them until they couple to the electrons.
The reported result is a critical temperature that rose “up to 5.4%” in a six-layer device, according to the USTC Low-dimensional Physics lab page, which carries the paper’s title, “Evidence for vacuum-enhanced superconductivity in NbSe2.” Cheng added that the critical current and critical magnetic field were also significantly enhanced near the superconducting transition. The release gives no kelvin values behind the percentage, so the size of the shift in absolute degrees is not stated there.
Confinement, virtual photons and resonance
“Squeezing” is shorthand for confinement. The cavity confines and reshapes the electromagnetic modes available in its small volume, and the superconductor interacts with that reshaped vacuum through virtual photons, which, as press coverage of the paper describes it, lowers the energy of the superconducting state. Shaping the surroundings in this way changes the material without driving it with a laser or an electrical current.
The enhancement depends on tuning. Qingdong Jiang of Shanghai Jiao Tong University said that when the cavity mode’s characteristic energy matched the low-energy superconducting fluctuations, the device showed a resonant enhancement. The lab page adds that control experiments ruled out strain, degradation and inhomogeneity as the cause, and that the peak-like dependence on frequency pointed to resonant coupling between the superconducting state and the cavity modes.
Wilczek, the Nobel-winning physicist, framed the idea in a line reported by Science Times: the background itself can become an actor, engineered to strengthen superconductivity.
Critical temperature is the point below which a material loses all electrical resistance, so even a modest rise widens the range in which the superconducting state survives. NbSe2 is a convenient test bed because it remains superconducting in very thin layers, and the number of layers matters: the USTC result is for six layers in a single device. The Chinese Academy of Sciences release stresses that the effect was seen in the critical current and critical magnetic field as well, the other two limits that decide how much current and field a superconductor can tolerate before it reverts to ordinary metal.
Scope of the 5.4% and what remains untested
The number carries three limits. It applies to a six-layer NbSe2 device, it is described as “up to,” and it comes from carefully engineered cavity conditions. The CAS release calls the work a proof of concept and says practical use would call for further optimization of cavity structures and material systems. A 5.4 percent shift in a critical temperature that is already low does not bring superconductivity near room temperature.
A separate preprint on arXiv, from a group that includes Singapore’s Nanyang Technological University and Centre for Quantum Technologies, reports cavity-enhanced superconductivity in few-layer NbSe2 too, with a roughly 10 percent rise, from 3.02 K to 3.41 K, in a trilayer coupled to a 2.04 THz resonator. That is a different team, device and cavity, so its 10 percent should not be merged with the USTC 5.4 percent. It does show that the effect has drawn independent attention in the same material, and that thinner samples in that work responded more strongly: the ten-layer sample gained only 0.10 K.
The paper itself, in Nature, is the record to check for the full dataset. Its six-layer 5.4 percent is the figure the USTC release and lab page both give. The authors named in the CAS release are Zeng and Cheng of USTC, Jiang of Shanghai Jiao Tong University and Wilczek of MIT, and the paper’s DOI is 10.1038/s41586-026-11037-x. Any claim of a larger shift, in another thickness, another cavity or another material, would need its own measurement, because the 5.4 percent in this device is the only value the institutional release reports.
This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.
More from Morning Overview
- NTSB report details how a family of four died when their plane’s nose suddenly pitched up
- Doctors warn a silent liver disease now affects one in three American adults
- Common allergy, bladder and sleep pills tied to sharply higher dementia odds
- The second-largest U.S. reservoir just fell to its lowest level ever recorded
