Atomic-spin-based comagnetometers have long served as exceptionally sensitive instruments for measuring magnetic fields and testing some of the deepest assumptions in physics. Now, a team of researchers has demonstrated a new kind of comagnetometer built from two isotopes of ytterbium, jointly confined in an optical lattice and controlled through optical techniques. The experiment combines the nuclear spin of ^171Yb, which has spin one-half, with the spin-five-halves nucleus of ^173Yb. According to the study published in Nature Photonics, the system maintains spin coherence for 60 seconds, suppresses magnetic noise by more than a factor of 30,000 and measures the ratio of the two isotopes’ nuclear magnetic moments with four-parts-per-million precision. The result creates a cold-atom platform for precision sensing at distance scales that are difficult to access with conventional atomic systems, while also pointing toward future searches for subtle signals of physics beyond the Standard Model.
A comagnetometer works by allowing two different spin species to experience nearly the same magnetic environment at the same time. Each spin precesses around the applied magnetic field, much like a tiny gyroscope whose orientation rotates at a field-dependent frequency. If the field changes, both precession frequencies shift. But because the two species respond differently according to their magnetic moments, comparing their accumulated phases can reveal the magnetic field while also allowing common magnetic fluctuations to be rejected. In precision experiments, this common-mode rejection is essential: environmental magnetic noise can be millions of times larger than the signal associated with a new force, an exotic interaction or a tiny violation of a fundamental symmetry. Gas-cell comagnetometers have achieved remarkable sensitivity, but their atoms are distributed throughout a macroscopic vapor cell. Cold atoms confined in a lattice can instead be localized, optically manipulated and brought into controlled geometries, opening the door to measurements over much shorter distances.
The challenge is that cold atoms are not automatically ideal clocks or sensors. Optical trapping provides strong confinement, but the trapping light can also perturb the internal energy levels of the atoms. These perturbations are commonly described as light shifts, or differential shifts when they alter the relative energy between two spin states. Even a small, uncontrolled light shift can imitate a magnetic field or cause different atoms to accumulate different phases. As interrogation times become longer, the phase error grows, eventually destroying the interference signal. This problem is particularly important for diamagnetic atoms such as ytterbium, whose nuclear spins are attractive for precision sensing because their electronic structure offers reduced sensitivity to certain environmental effects. The researchers therefore had to design the optical trap and the spin states together, suppressing both vector and tensor light shifts rather than treating them as minor corrections after the experiment.
The first part of that strategy involves the polarization of the optical lattice. Light with an electric-field vector can interact with atomic angular momentum in a way that produces a vector light shift. In a spin measurement, this shift behaves in many respects like an artificial magnetic field: it depends on the polarization of the light and can vary if the lattice polarization drifts or becomes imperfect. The team suppressed this contribution by enforcing linear polarization of the lattice light. Under ideal linear polarization, the vector component is strongly reduced, preventing the trapping beam from introducing a spurious spin-dependent precession. This is a deceptively demanding requirement in a real optical system, where mirrors, lenses, vacuum windows and alignment imperfections can transform polarization. By making polarization control a central part of the apparatus, the researchers removed one of the key obstacles to long-lived nuclear-spin coherence.
The second obstacle arises from tensor light shifts, which are especially relevant for the spin-five-halves nucleus of ^173Yb. Unlike a spin-one-half system, a higher-spin manifold can respond to the orientation of the atomic state in a more complicated, direction-dependent way. The trapping light can therefore shift different magnetic sublevels by different amounts, even when vector effects have been eliminated. To overcome this problem, the researchers used a Schrödinger cat state: a coherent quantum superposition of two spin configurations chosen so that their tensor shifts cancel in the interferometric measurement. Rather than allowing the ^173Yb nucleus to occupy a single vulnerable sublevel, the experiment encodes the phase in a carefully engineered combination of states. The two components acquire the relevant tensor contributions with opposite signs, causing the unwanted shift to disappear from the measured phase while preserving sensitivity to the genuine magnetic precession. This use of a nonclassical state turns a normally damaging systematic effect into a quantity that can be rejected through symmetry.
With both isotopes trapped in the same optical lattice, the researchers carried out simultaneous Ramsey interferometry. In a Ramsey measurement, a first coherent control pulse prepares a superposition of spin states. The atoms then evolve freely for a chosen interrogation time, during which the relative phase between the components records the influence of magnetic fields and other energy shifts. A second control pulse converts that phase into a measurable population difference. Repeating this sequence for both ytterbium isotopes allows their phases to be compared directly. The shared lattice and nearly simultaneous interrogation ensure that the two nuclear spins experience closely related environmental conditions. The experiment’s 60-second spin coherence time is especially significant because the longer the coherent evolution, the more precisely a small frequency difference can be resolved. It also demonstrates that careful optical control can preserve quantum phase information in a trapped cold-atom system for a duration far beyond the timescales commonly associated with delicate spin manipulations.
The reported magnetic noise suppression factor, exceeding 3 × 10^4, shows the practical power of the isotope comparison. Environmental field fluctuations affect both nuclear spins, but the comagnetometer combines their phase records according to the known relationship between their magnetic moments. Common fluctuations are then rejected, leaving a differential signal that is far less sensitive to ordinary magnetic noise. This does not mean that the instrument becomes completely immune to magnetic fields. Rather, it measures the two responses together and uses their correlation to distinguish field noise from effects that would act differently on the two isotopes. Such rejection is crucial for experiments searching for tiny spin-dependent interactions, anomalous forces or signatures associated with violations of established symmetries. The ability to place both species in the same lattice also reduces uncertainties associated with spatially separated samples, an important advantage when the sought-after interaction changes over short distances.
As a demonstration of the platform’s precision, the researchers determined the ratio of the ^171Yb and ^173Yb nuclear magnetic moments to be –0.726076(3), with a precision of approximately four parts per million. The negative sign indicates that the two nuclear magnetic moments have opposite orientations relative to the chosen spin conventions. Measuring this ratio through a common optical-control sequence provides a stringent test of the experiment’s ability to distinguish genuine nuclear-spin dynamics from light-induced shifts and magnetic noise. The result is not merely a number characterizing ytterbium. It is also a calibration of the comagnetometer itself: an accurate magnetic-moment ratio allows the instrument to cancel ordinary magnetic fields more effectively and to identify residual phase shifts that cannot be explained by them. In future measurements, this capability could help separate technical imperfections from possible signals generated by new interactions between spins and matter or between different forms of quantum fields.
The significance of the work extends beyond ytterbium spectroscopy. Comagnetometers have been used in searches for electric dipole moments, exotic spin couplings, short-range forces and other phenomena that could reveal physics beyond the Standard Model. Many of these experiments require long interrogation times, exceptionally stable phase references and an accurate understanding of every field that can mimic the desired signal. The combination demonstrated here—diamagnetic atoms, optical-lattice confinement, linear-polarization control and a tensor-insensitive cat state—addresses several of those requirements simultaneously. Cold atoms also offer a flexible route to changing the geometry and separation of the sensing particles, potentially enabling measurements of interactions that disappear over longer distances. The platform may be integrated with advanced optical quantum-control methods, squeezed or entangled states and tailored spatial arrangements to improve sensitivity still further. The researchers’ result therefore represents more than an improved measurement of ytterbium’s magnetic properties: it establishes a versatile architecture in which nuclear spins can remain coherent, can be compared with high precision and can be deployed in the continuing search for faint clues that the known laws of physics are incomplete.
Subject of Research: Cold-atom comagnetometry using the nuclear spins of ^171Yb and ^173Yb, optical control of spin states, long spin coherence and precision tests of magnetic interactions.
Article Title: Cold-atom comagnetometry via optical control of spin states
Article References: Zhang, JL., Luo, WT., Yang, Y.A. et al. Cold-atom comagnetometry via optical control of spin states. Nature Photonics (2026). https://doi.org/10.1038/s41566-026-01982-4
Image Credits: AI Generated
DOI: https://doi.org/10.1038/s41566-026-01982-4
Keywords: Cold atoms; comagnetometry; ytterbium; nuclear spins; optical lattice; Ramsey interferometry; quantum coherence; light shifts; magnetic noise suppression; physics beyond the Standard Model
