Abstract
The melting temperatures of $K_2CO_3$ were experimentally determined to be $1220 \pm 20$ °C (4.0 $\pm$ 0.5 GPa), $1290 \pm 10$ °C (9.0 $\pm$ 0.5 GPa), and $1313 \pm 10$ °C (11.5 $\pm$ 0.5 GPa) in a 2000 ton split-sphere apparatus and $1195 \pm 15$ °C (5.0 $\pm$ 0.5 GPa) in a 1000 ton uniaxial split-cylinder apparatus. The fusion curves of $K_2CO_3$ were calculated up to ~12.0 GPa for various $K_0'$ values of the liquid, according to the thermodynamic properties for crystalline and liquid $K_2CO_3$. On the basis of these experimental results and fusion curves, the $K_0'$ for liquid $K_2CO_3$ is constrained to be ~14.4 $\pm$ 1.1 at pressures lower than 5.0 GPa. However, the results at pressures above 9.0 GPa deviate from this trend, suggesting a possible phase transformation in either the crystalline or liquid phase between 5.0 and 9.0 GPa. $K_2CO_3$ is shown to have the lowest melting temperature among common carbonates.
Introduction
Potassium is a highly mobile and geochemically significant element in Earth's interior. As tectonic plates subduct, they transport potassium-rich minerals into the mantle. These minerals can melt, triggering profound geological changes, such as lowering the solidus of surrounding peridotite or creating high-conductivity "Low Velocity Zones" (LVZs) observed in seismic data.
Understanding these processes requires a precise mapping of the melting behavior—the fusion curve—of potassium carbonate ($K_2CO_3$) under deep-mantle pressures. Previous studies were limited to pressures below 3.2 GPa or provided inconsistent data regarding liquid density. This research aims to extend the experimental fusion curve to 12 GPa to better constrain the thermodynamic properties of $K_2CO_3$ and its role in mantle metasomatism and diamond formation.
Methods
The study utilized high-pressure experimental techniques paired with thermodynamic modeling. Experiments were conducted using 1000-ton uniaxial split-cylinder and 2000-ton split-sphere multi-anvil apparatuses to simulate the extreme conditions of the upper mantle.
To determine the melting point, the researchers employed the Pt bead "bracket" technique. A small platinum (Pt) bead was placed atop a sample of synthetic $K_2CO_3$ powder within a capsule. After being subjected to target pressure and temperature, the sample was quenched and sectioned. The final position of the Pt bead served as a diagnostic tool: if the bead remained at the top, the sample had stayed solid; if it sank through the sample, the material had melted into a low-viscosity liquid.
Theoretical fusion curves were then calculated using a third-order Birch-Murnaghan equation of state (EOS). By fitting these curves to the experimental data, the team constrained the $K_0'$ value, which represents the pressure dependence of the liquid's bulk modulus.
Key Results and Discussion
The experimental campaign yielded several critical melting benchmarks:
- $1220 \pm 20$ °C at 4.0 GPa
- $1195 \pm 15$ °C at 5.0 GPa
- $1290 \pm 10$ °C at 9.0 GPa
- $1313 \pm 10$ °C at 11.5 GPa
At pressures below 5 GPa, the data is well-described by a $K_0'$ of $14.4 \pm 1.1$. However, a significant deviation occurs at pressures above 9 GPa. This divergence from the predicted EOS trend strongly indicates a phase transition occurring between 5 and 9 GPa. This structural change could involve a transition in the crystalline solid or a coordination change in the liquid phase.
Furthermore, $K_2CO_3$ was found to have a lower melting temperature than other common mantle carbonates like $CaCO_3$ or $MgCO_3$. This makes $K_2CO_3$-rich melts the "first responders" in mantle melting scenarios. The study also confirms that $K_2CO_3$ liquid is significantly less dense than typical silicate melts in the upper mantle, suggesting these carbonate melts are highly buoyant and likely to migrate upward, chemically altering the surrounding mantle rock (metasomatism).
Conclusion and Future Directions
This work provides a robust experimental foundation for the $K_2CO_3$ fusion curve up to 12 GPa. The identification of a phase transition between 5 and 9 GPa highlights a gap in our current understanding of carbonate structures at depth.
Future research should focus on in-situ X-ray diffraction at synchrotron facilities to directly observe structural changes in $K_2CO_3$ under pressure. Additionally, investigating more complex systems, such as the $K_2CO_3–MgCO_3$ binary, will be essential for developing realistic models of deep-Earth carbon and alkali cycling.
Summary updated