Nature Visual. Artwork Credit: Valentin Heinrich
Single Molecule Shows Promise as Tiny Memory for Quantum Computers
In A Nutshell
- One carbene molecule, chilled to 4.5 Kelvin inside a crystal, kept quantum information alive for about 2.2 thousandths of a second with protective microwave pulses.
- Researchers read the molecule’s spin with a laser, and its light color stayed steady for more than an hour.
- By the authors’ account, the spin results beat earlier molecular qubit results from large groups by more than tenfold.
- Molecules could eventually be built into light-based chips without the etching diamond defects need, but no chip or network has been built yet.
A single molecule, chilled inside a crystal to about minus 452 degrees Fahrenheit, kept a fragile piece of quantum information alive for roughly 2.2 thousandths of a second, with help from protective microwave pulses. Fleeting as that sounds (a blink lasts dozens of times longer), researchers could read the information back just by shining a laser on the molecule and watching whether it glowed. Chemists made the molecule in a lab, which hints at a different way to build quantum computers and communication networks.
Quantum networks, which could one day link quantum computers the way the internet links laptops, need a part that does two jobs: store a quantum bit, or qubit, long enough to be useful, and exchange information with the outside world through particles of light called photons. Defects in diamond have done both. Putting those defects on chips that run on light, though, usually means etching the host crystal, which can strain it and make its light less stable, according to a new paper in Nature.
Molecules made through ordinary chemistry might avoid that problem. This one is a carbene, a carbon-based compound. The researchers say no single molecule had combined bright light, steady color, and a long-lived spin (a tiny internal magnet that can point in different directions) before this experiment.
Crystal Trap Makes an Unstable Molecule a Stable Qubit
Carbenes are built around a carbon atom missing some of its usual bonds, so they grab onto neighbors fast. To keep one around, a team based at NVision Quantum Technologies and Ulm University in Germany started with a calmer precursor molecule and grew it inside a crystal made of a similarly shaped compound. The precursor slotted in like a puzzle piece dropped into a matching puzzle, holding it in place.
After cooling the crystal to 4.5 Kelvin, colder than any naturally occurring spot on Earth, the team pointed a focused laser at it, knocking a nitrogen molecule off the precursor and turning it into the carbene. The beam can be steered with pinpoint accuracy, so the researchers controlled where the qubits formed and how crowded they were, thinning them out until single molecules could be studied.
Light Tests Confirm Single Molecules Stayed Steady for Over an Hour
To prove the glow came from one molecule and not a small cluster, the researchers sent the light into a detector that tracks whether photons arrive alone or in bunches. A cluster would produce frequent pairs, but this molecule produced a clear shortage of them, the signature of a single emitter.
Steady color matters because quantum networks need light from different sources to match, like instruments tuned to the same note. In semiconductor crystals like diamond, stray electrical charges can make an emitter’s color wander. Molecular crystals lack those charge traps, the authors say. Over more than an hour of repeated scans, the color of two tracked molecules barely budged, drifting by less than a millionth of a percent. The authors describe the result as “remarkable spectral stability.”
Microwave Pulses Keep Quantum Memory Alive for 2.2 Milliseconds
Qubit information lives in the molecule’s spin, and pulses of microwaves, cousins of the energy inside a kitchen microwave oven, flipped that spin between states. Each spin state responds to a slightly different color of laser light, so shining one color and watching for a glow showed which way the spin pointed.
Left unprotected, the spin signal faded in a few millionths of a second. A repeating sequence of microwave pulses designed to shield the spin from its surroundings stretched that to roughly 2.2 thousandths of a second, hundreds of times longer. That is the molecule’s memory time, the point where the signal has faded to about a third of its strength. For near-perfect storage, 99% accuracy, the authors calculate about 40 millionths of a second. A separate measure, how long the spin takes to drift back to rest, came in at about 21 thousandths of a second. By the authors’ account, those single-molecule numbers top earlier results from large groups of molecular qubits by more than tenfold, and they call them “a substantial leap for molecular qubits.”
Molecular Qubits Could Sidestep Tricky Chip Etching, Researchers Say
Measured against the checklist for quantum network parts, which includes long-lasting memory, usable light, and steady color, the molecule “already combines many of the required properties,” the paper reports.
Chemistry could also ease manufacturing. Molecules can be spread into thin films and placed directly onto prebuilt light-based chips, the authors write, which would avoid that etching step. None of that has been built yet, and linking photons to a single molecule’s spin, the move a working quantum network requires, remains a future goal. If molecules can be tweaked the way chemists adjust dyes or drugs, quantum parts might someday be designed to order.
A compound type known for being too reactive to last has become an individually controlled memory, which makes molecular design a credible second route to quantum hardware alongside diamond defects. A molecule-powered quantum internet is not around the corner. Still, quantum parts designed in a flask rather than carved from exotic crystals now have a first proof of concept.
Paper Notes
Limitations
Experiments ran in a laboratory at 4.5 Kelvin, so the setup is far from a room-temperature device. Light output was modest: typical photon counts were about 2,000 per second, limited largely by a 10.8% quantum yield (the share of absorbed light re-emitted as a glow). The net emission into the sharp color line was about 2.1%, which the authors describe as comparable to diamond-based defects. The molecules also degrade under continuous laser and microwave exposure, with a lifetime of about 119 hours in the tests. The sample was small: the single-photon test used one isolated molecule, the stability tracking followed two, and the main text reports the spin measurements on individual molecules without a larger count, so how typical the results are remains to be shown. The paper also does not demonstrate entanglement between a photon and the molecule’s spin, a working network node, or integration onto a photonic chip. The authors describe those as future routes.
Funding and Disclosures
Support came from the state of Baden-Württemberg (computing resources through bwHPC), the German Research Foundation, and the DLR Quantum Computing Initiative with Germany’s Federal Ministry for Economic Affairs and Climate Action (COMIQC project). Individual authors also acknowledged the Clore Israel Foundation Scholars Programme, the Israeli Council for Higher Education, the Milner Foundation, an ERC grant (QRES), a Quantera grant (MfQDS), the Israeli Science Foundation, the Schwartzmann University Chair, and the Israeli Innovation Authority. Competing interests: 12 of the 15 authors, including both corresponding authors, are employees of NVision Quantum Technologies, a company commercializing products in quantum information technology. Four authors (Martin B. Plenio, Alex Retzker, Fedor Jelezko, and Ilai Schwartz) are also shareholders in the company.
Publication Details
Title: “A single-molecule spin–photon interface.” Authors: Simon Roggors, Thomas Unden, Anna Aubele, Paul Mentzel, Gregor Bayer, Alon Salhov, Jochen Scharpf, Martin B. Plenio, Alex Retzker, Fedor Jelezko, Tim R. Eichhorn, Tobias A. Schaub, Matthias Pfender, Philipp Neumann, and Ilai Schwartz. Corresponding authors: Matthias Pfender and Ilai Schwartz. Affiliations: NVision Quantum Technologies (Ulm, Germany); Institute for Quantum Optics, Ulm University; Racah Institute of Physics, The Hebrew University of Jerusalem; Institute of Theoretical Physics, Ulm University; and Center for Integrated Quantum Science and Technology, Ulm University. Journal: Nature, Volume 658, pages 1153 to 1159 (October 29, 2026 issue). Received April 8, 2026; accepted September 10, 2026; published online October 28, 2026. DOI: 10.1038/s41586-026-11147-6. The paper is open access under a Creative Commons Attribution 4.0 International License.







