First-hand information for everyone
Researchers used diamond quantum sensors to study water squeezed into channels just 5.6 nanometers wide. By watching the tiny magnetic signals from the water’s hydrogen nuclei, they found that the confined liquid stays mobile but shows extremely slow motion. On some runs the proton signal briefly split into two lines separated by tens of kilohertz, suggesting an unusual, long-lived magnetic state in the confined water.
The team probed the water with nitrogen‑vacancy (NV) centers, a well-known defect in diamond that acts like a tiny magnetic sensor. They built channels from layers of hexagonal boron nitride placed on a diamond chip with near‑surface NV centers. A controlled magnet set a field of about 35–50 millitesla, and a Peltier cooler let them change temperature from room temperature down to near freezing. A pulsed sensing protocol (an XY8 dynamical‑decoupling sequence) made the NV centers sensitive to the magnetic noise of nearby hydrogen nuclei and let the researchers record a nanoscale nuclear magnetic resonance (NMR) spectrum from the confined water.
Across the temperature range the water remained liquid and the proton resonance stayed detectable. The resonance linewidth was roughly constant at about 40 kilohertz. From that linewidth the authors estimate an effective molecular diffusivity on the order of 10−12 square meters per second or lower. That is many orders of magnitude slower than ordinary bulk water at room temperature (about 2×10−9 m2/s) and slower than supercooled bulk water. The slow motion is what makes it possible to record a proton NMR signal with these shallow NV sensors, because rapid diffusion would wash out the signal.
Occasionally — most often near 4–5 °C but sometimes up to 14 °C — the single proton line changed into a resolved doublet with a splitting of several tens of kilohertz. The split appears and then disappears over timescales of hours while other experimental conditions stay the same. The authors tentatively interpret this intermittent splitting as arising from hyperfine interactions (magnetic coupling between electron and nuclear spins) mediated by long‑lived paramagnetic charge‑hydration complexes. In plain terms, they suggest that optical excitation during the measurement can create solvated electrons (free electrons stabilized in water), which seed a metastable, correlated population of charged, magnetic complexes that change the local magnetic field felt by the protons.
The result matters because it shows how nanoconfinement can slow molecular motion enough to reveal magnetic effects that are invisible in bulk liquids. It also demonstrates the capability of NV‑based NMR to probe both transport and local electromagnetic environments on nanometer scales. Important caveats apply: the interpretation involving long‑lived paramagnetic complexes is tentative and based on the intermittent appearance of the doublet. The feature is not always present, and the authors stress that further work will be needed to confirm the mechanism and to rule out other explanations.