Science 343 522–525 (2014)

At the boundary between Earth's mantle and core, solid silicate rock meets liquid iron metal. High-pressure melting experiments on natural samples of mantle rocks now suggest that temperatures at the core–mantle boundary are lower than previously thought.

Ryuichi Nomura and Kei Hirose at the Tokyo Institute of Technology, Japan, and colleagues performed laboratory experiments to determine the maximum temperature at which primitive mantle rock remains completely solid, under the extreme pressures of the core–mantle boundary. Mantle rocks were thought to remain solid up to temperatures of about 4,000 K, but the researchers found that the rocks began to melt at just 3,570 K. Since the lowermost mantle is largely solid, this suggests temperatures cannot exceed 3,570 K. Such modest temperatures in the lowermost mantle also imply that the outer core is cooler than assumed, by up to 700 K. Pure iron would not be molten at these lower temperatures, but an iron-hydrogen mixture would.

The researchers suggest that hydrogen may have entered the core from the hydrous magma ocean that is thought to have existed on Earth during core formation.