A 2026 arXiv preprint (not yet peer-reviewed) by Goncharov and colleagues delivers the first simultaneous high-pressure, high-temperature measurements of thermal conductivity in ferropericlase, the second-most abundant mineral in Earth's lower mantle. Using laser flash heating and X-ray free-electron laser techniques inside diamond-anvil cells, the team tested single-crystal samples (iron content 9–13%) up to 130 GPa and 2200 K. The experiments revealed a sharp drop in conductivity between 60 and 100 GPa at roughly 1700 K—precisely where iron atoms undergo spin crossover, switching from high-spin to low-spin electron configurations.

This quantum transition scatters phonons more effectively, turning the mineral into a poorer heat conductor. When combined with the authors' prior bridgmanite data, the resulting lower-mantle conductivity profile rises overall with depth yet shows a clear dip caused by spin crossover, reaching only about 10 W m⁻¹ K⁻¹ near the core-mantle boundary.

Previous coverage and standard geodynamic models often assumed conductivity increases monotonically with pressure, implying efficient core-heat extraction and vigorous convection. That picture is incomplete. The observed reduction acts like a partial thermal blanket, trapping heat in the deep mantle. This alters plume buoyancy, reduces overall mantle heat flux, and slows long-term planetary cooling—potentially extending the lifetime of plate tectonics and delaying the shutdown of the geodynamo.

Synthesizing this work with Badro et al. (Science, 2013, doi:10.1126/science.1244828), which mapped spin transitions via X-ray emission spectroscopy but lacked direct conductivity data, and with Li et al. (Geophysical Research Letters, 2020) modeling variable-conductivity convection, reveals a consistent pattern: mid-mantle seismic anomalies line up with the 60–100 GPa spin-crossover window. Earlier models missed how this dip stabilizes large low-shear-velocity provinces and changes superplume frequency.

Limitations are clear. Diamond-anvil cell samples are micrometers in size, the iron concentrations tested may not perfectly match every mantle region, and minor elements such as aluminum were not fully explored. Extrapolation to the real, heterogeneous mantle therefore carries uncertainty. Still, the experimental advance forces a recalibration of Earth's thermal history: the mantle may lose heat more slowly than most simulations predict, keeping the core hotter longer and subtly reshaping surface geology across hundreds of millions of years.