Lanthanum cobaltite has been confounding solid-state physicists since at least 1958, when Goodenough proposed that cobalt ions in this material undergo spin-state transitions — switching between low-spin (all electrons paired, no magnetic moment) and high-spin (unpaired electrons, large moment) configurations as temperature increases. The transition drives an insulator-metal crossover around 500 K and produces anomalies in thermal expansion, resistivity, and magnetic susceptibility. Sixty-eight years later, the microscopic nature of the spin-state ordering — whether the high-spin and low-spin cobalt ions arrange themselves in a specific spatial pattern, and at what temperature — remains contested.
Ivashko, Manjo, Kauth, and collaborators (arXiv 2602.22725, February 2026) find the evidence in the lattice vibrations. Using inelastic neutron and X-ray scattering across the full temperature range from 2 K to 650 K, they identify an anomalous softening of a 10 meV oxygen phonon mode. The softening appears only between 100 K and 550 K — the temperature window where the spin-state crossover is active. And it occurs at a specific wave vector that matches the theoretically predicted ordering pattern for alternating high-spin and low-spin cobalt sites.
The phonon is acting as a reporter. The oxygen atoms sit between cobalt ions and mediate their magnetic coupling. When a cobalt ion switches spin state, its ionic radius changes — high-spin Co³⁺ is larger than low-spin Co³⁺ because the populated antibonding orbitals expand the ion. This size change modifies the local bonding environment, which shifts the phonon frequency. If the spin-state change is spatially ordered — high-spin and low-spin alternating on the cobalt sublattice — the phonon softening appears at the ordering wave vector. If the spin-state changes were random, the softening would be uniform across momentum space.
The momentum-resolved softening is the signature. It demonstrates that the spin-state fluctuations are spatially correlated at the predicted wave vector — not long-range ordered (no static Bragg peak appears), but dynamically correlated. The cobalt ions aren't switching independently; they're switching in a pattern, fluctuating between ordered and disordered configurations on a timescale that the phonons can track.
The lattice heard the spin transition. The phonons are the proof that the ordering theorized in 1958 is real — dynamic, fluctuating, never fully frozen, but spatially structured.