Scientists have found something strange at the bottom of Earth’s mantle. About 2,900 kilometers beneath the surface, right where the rocky mantle meets the liquid metal core, small patches of material act like thermal insulators. These structures trap heat and may influence everything from volcanic activity to the stability of Earth’s magnetic field.
Researchers led by Professor Wen-Pin Hsieh at Academia Sinica in Taiwan measured the thermal properties of materials that likely form these mysterious zones. Their findings, published in Nature Communications, show these patches conduct heat far worse than surrounding rock. The discovery changes how scientists think about heat flow in Earth’s deepest layers.
Seismologists have known about ultralow velocity zones, or ULVZs, for decades. When earthquake waves pass through these regions, they slow down dramatically. Compressional waves drop by up to 20 percent. Shear waves slow by as much as 50 percent. The material in these zones is also denser than the mantle around them, up to 20 percent heavier.
These patches are small by geological standards. They rise only tens of kilometers high and stretch hundreds of kilometers wide. Think of them as irregular bumps on the core-mantle boundary. Recent seismic studies suggest ULVZs exist around the globe, not just under the massive low-velocity provinces beneath Africa and the Pacific Ocean. Some appear in high-velocity regions associated with ancient subducted slabs.
Despite their small size, ULVZs sit in a critical location. The core-mantle boundary separates Earth’s hottest region from the overlying mantle. Heat flowing from the core drives mantle convection, powers volcanic activity, and maintains the geodynamo that generates our magnetic field. Any structure that interferes with this heat transfer could have significant effects.
Scientists have proposed several explanations for what makes up ULVZs. Some researchers think they contain iron-rich minerals like magnesiowüstite, a mixture of magnesium oxide and iron oxide. Others suggest they might be patches of partial melt, remnants of an ancient magma ocean that once covered the core. Still others point to subducted oceanic crust that has sunk to the base of the mantle over billions of years.
Professor Hsieh’s team focused on magnesiowüstite as a likely ULVZ candidate. They synthesized two compositions in the laboratory: one with 25 percent iron content and another with 75 percent iron. Then came the hard part. Conditions at the core-mantle boundary are extreme. Pressure reaches 136 gigapascals, more than a million times atmospheric pressure. Temperature hits 4,000 Kelvin. No conventional instrument can recreate these conditions while also measuring thermal properties.
The researchers used diamond anvil cells, devices that squeeze tiny samples between two diamond tips. They compressed the magnesiowüstite samples to pressures exceeding 100 gigapascals. For temperature control, they built an externally heated diamond cell that could reach 973 Kelvin while maintaining pressure. To measure thermal conductivity, they employed time-domain thermoreflectance. This technique uses ultrafast laser pulses to heat a thin aluminum film deposited on the sample. A second laser probes how quickly heat dissipates from the aluminum into the sample below. The entire measurement happens in picoseconds, allowing the team to determine thermal conductivity with precision better than 15 percent.
The experiments revealed surprising behavior. Both iron-rich compositions showed low thermal conductivity compared to other mantle minerals. The 25 percent iron sample reached about 10 watts per meter per Kelvin at deep mantle conditions. The 75 percent iron sample conducted even less heat, dropping to approximately 3.4 watts per meter per Kelvin near the core-mantle boundary. For comparison, the mantle minerals bridgmanite and davemaoite conduct about 8 watts per meter per Kelvin at similar depths. The iron-rich material forming ULVZs transfers heat three times slower than surrounding rock.
The low thermal conductivity relates to an electronic transformation in iron atoms under pressure. At low pressures, iron electrons occupy high-energy orbital states, a configuration called high-spin. As pressure increases past about 40 to 80 gigapascals, depending on temperature and iron content, electrons drop into lower-energy orbitals. This spin transition fundamentally changes how the material conducts heat.
The researchers found that thermal conductivity decreases across the spin transition. For the 25 percent iron sample, conductivity dropped by about 30 percent as the material shifted from high-spin to low-spin states at room temperature. The 75 percent iron composition showed a larger decrease of 45 percent, though this happened at higher pressures. At elevated temperatures representative of the deep mantle, the transition became more gradual. Instead of a sharp drop, thermal conductivity decreased slowly over a wider pressure range.
The team also discovered that temperature affects thermal conductivity differently in low-spin versus high-spin states. For high-spin material, conductivity typically decreases with temperature following a power law with an exponent of about negative 0.5. The low-spin 25 percent iron sample showed a weaker temperature dependence with an exponent near negative 0.39. The 75 percent iron material displayed an even weaker dependence at negative 0.23. This weaker temperature sensitivity means low-spin iron-rich materials maintain relatively low thermal conductivity even at the extreme temperatures near the core-mantle boundary.
Using their experimental data, the researchers built models of thermal conductivity throughout the lower mantle. They started with a representative temperature profile that begins at 1900 Kelvin at 660 kilometers depth and rises to 4000 Kelvin at the core-mantle boundary. For the 75 percent iron magnesiowüstite that might form ULVZs, thermal conductivity stays remarkably low throughout the deep mantle, around 3.6 to 3.8 watts per meter per Kelvin from 1600 to 2800 kilometers depth. At the core-mantle boundary, where temperature spikes upward, conductivity drops further to 3.4 watts per meter per Kelvin.
If ULVZs contain partial melt instead of solid iron-rich minerals, their thermal insulation becomes even more pronounced. The team calculated that basaltic melt at core-mantle boundary conditions would have a thermal conductivity near 1.9 watts per meter per Kelvin. This represents an upper limit since iron-enriched melts would conduct even less. Either way, ULVZs act as powerful thermal barriers between the core and mantle.
When ULVZs block heat flow from the core, temperature builds up at their base. This creates a positive feedback loop. Higher temperatures reduce thermal conductivity further, which traps more heat, which raises temperatures even more. The result is thermal runaway that produces hot, buoyant material. This mechanism could explain why thermal plumes, the upwelling currents that feed volcanic hotspots like Hawaii and Iceland, often seem rooted in the deepest mantle.
The thermal insulation also creates steep temperature gradients across the thin ULVZ layers. Above these zones, temperatures might actually be cooler than in surrounding areas. This local cooling could stabilize minerals like post-perovskite, which forms only under specific high-pressure, moderate-temperature conditions. The combination of hot material below and cooler material above sets up small-scale instabilities that affect regional mantle flow patterns.
Right at the core-mantle boundary, something striking happens. Earth’s outer core, made of liquid iron and nickel alloy, has a thermal conductivity somewhere between 25 and 250 watts per meter per Kelvin, depending on composition and measurement method. The uncertainty is large, but even the lowest estimates put core conductivity an order of magnitude higher than ULVZ materials. Heat flowing from core to mantle encounters massive thermal resistance at ULVZ locations.
This resistance sharply reduces local heat flux from the core. In extreme cases, if a ULVZ becomes hotter than the core-mantle boundary temperature, heat might actually flow backward, from mantle into core. Such patches of negative heat flux would create complex patterns in core fluid flow. The outer core’s motion generates Earth’s magnetic field through the geodynamo process. Any irregularities in heat extraction affect convection patterns in the core, which in turn influence the magnetic field’s strength and stability.
The researchers calculated something called the Q-star parameter, which measures how much heat flux varies across the core-mantle boundary. Strong variations in this parameter correlate with geomagnetic instability, including the reversal of magnetic poles. Earth’s magnetic field has flipped hundreds of times over geological history, with the north and south magnetic poles swapping places. The timing between reversals varies wildly, from tens of thousands to millions of years. Scientists still don’t fully understand what triggers these flips, but heterogeneous heat flux at the core-mantle boundary likely plays a role.
ULVZs appear scattered across the planet’s base. Their global distribution means they create a patchwork of high and low heat flux regions rather than a uniform boundary condition. This heterogeneous pattern makes the geodynamo more complex and possibly less stable than if heat extraction were smooth and even. The evolution of ULVZ locations over geological time could help explain the irregular timing of magnetic reversals.
Professor Hsieh noted that these findings represent significant progress in understanding heat transport at Earth’s deepest levels. The work connects laboratory measurements at extreme conditions with planetary-scale processes that shape our world. The team emphasized that much remains unknown. Future experiments need to measure thermal conductivity of other candidate ULVZ materials, including various melt compositions and mixed mineralogies. Computer simulations that incorporate realistic thermal properties will help test whether the proposed mechanisms actually operate in Earth’s interior.
The identification of thermal insulation at the core-mantle boundary reveals new information about how our planet works. These tiny patches, barely visible in seismic data, may exert outsized influence on processes ranging from volcanic hotspots to magnetic field generation. Understanding them better could reveal fundamental truths about Earth’s thermal evolution over billions of years.
Source:
Research published in Nature Communications by Wen-Pin Hsieh and team at Academia Sinica, Taiwan.
Source Link: https://doi.org/10.1038/s41467-025-65430-7
