POLONAISE: Searching both ends of the dark matter mass axis with levitated magnets

by Christopher Tunnell (Rice University)

Thursday Oct 15, 2026, 2:00 PM → 4:00 PM Europe/Madrid

C7b/058 - Seminar (IFAE Main Building C7b)

 

How heavy is dark matter? We know how much surrounds us, and its flux falls as one over the unknown mass. This seminar starts from that flux and the state of WIMP searches, then walks the mass axis with one platform called POLONAISE: sub-milligram magnets levitated in a superconducting trap in Leiden.

At the light end, dark matter is a classical wave. A vector field coupled to B−L drives the magnet as a forced oscillator, and the signal is one narrow line at the Compton frequency. We show the signal model at equation level. Our first search, a four-hour run, constrains the B−L coupling near 10⁻¹³ eV. Torsion balances and MICROSCOPE are still stronger there.

At the heavy end, dark matter is rare, and each particle gives the magnet one kick. In 220 hours of night data we set limits on dark matter coupled to neutrons through a light mediator. We reach seven orders of magnitude higher in mass than optical levitation. For one composite benchmark, where xenon experiments lose coherence, our bounds lead over part of the mass range. You will leave with a napkin test for your model. Data and code are public.

The last minutes cover how we used AI coding agents in the analysis. I will name the successes and failures, and how we could tell the difference.