Long-range magnetic forces govern ferrimagnet phase transitions

In a laboratory in Kyoto, researchers have solved a puzzle that has lingered at the edge of magnetic physics for years. When materials undergo phase transitions—sudden shifts in their magnetic properties—physicists have long known that wildly different systems often follow the same mathematical rules. This principle, called universality, has held up reliably in ferromagnets and antiferromagnets when short-range forces dominate. But ferrimagnets, a third class of magnetic materials with their own peculiar structure, have remained an open question. Now, an international team has shown that ferrimagnets do follow these universal rules—but only when long-range forces take the lead.

Ferrimagnets are materials with two magnetic sublattices pointing in opposite directions, like dancers facing away from each other. The catch is that their magnetic moments have different strengths, so they don't cancel out completely. Instead of pure opposition, you get a net magnetization—a residual pull in one direction. This makes them fundamentally different from ferromagnets, which align uniformly, and antiferromagnets, which cancel perfectly. The question researchers faced was whether the same physics that governs ferromagnets near their critical point would also govern ferrimagnets. The answer, it turned out, depended on which forces you paid attention to.

Yusuke Nambu, the corresponding author on the study, explains the counterintuitive finding: the strongest force at the microscopic level is not always the one that matters most when a material approaches its transition. In ferrimagnets, exchange interactions—the short-range forces that bind atoms to their neighbors—build the magnetic structure itself. But dipole-dipole interactions, which reach much farther across the material, turn out to be the ones that set the rules for how the transition unfolds. It's a distinction that matters profoundly for understanding magnetic behavior.

The team, drawn from Kyoto University, Tohoku University, and Australia's Nuclear Science and Technology Organisation, chose to test their theory on a compound called Eu2MnSi2O7. This melilite-type material is an ideal laboratory: the large magnetic moments of its europium and manganese ions make it a clean system for observing how long-range interactions shape critical behavior. The researchers synthesized the compound in polycrystalline form and subjected it to a battery of measurements. They used magnetization studies to track how the material's magnetic properties changed with temperature, and they deployed neutron powder diffraction at two specialized instruments—Echidna and Wombat—to map the atomic-scale magnetic structure.

What they found was striking. Three independent magnetization analyses all yielded transition temperatures and critical exponents that matched mean-field theory predictions—the mathematical framework that assumes long-range interactions dominate. A separate analysis of temperature-dependent neutron reflections confirmed the same conclusion. The neutron diffraction data also revealed that the europium and manganese ions order simultaneously in a tilted ferrimagnetic structure, a pattern consistent with the crystal's asymmetry. Every measurement pointed in the same direction: long-range dipolar forces were governing the transition.

This result fills a gap that has existed in magnetic physics. Dipolar-driven mean-field criticality had been firmly established in ferromagnets but never convincingly demonstrated in ferrimagnets. The new work extends the principle to a more complex magnetic class, broadening the scope of universality in magnetic phase transitions. Nambu notes that ferrimagnets occupy a unique position in the magnetic landscape—they combine a net magnetization with internal antiferromagnetic correlations, making them a bridge between ferromagnets and antiferromagnets. By showing that ferrimagnets follow mean-field rules, the team has closed one important gap. The antiferromagnetic case, however, remains unresolved, leaving room for future investigation.