Probe could ride to Earth's core in a mass of molten iron. (Going Down?).

Blast a crack in Earth's crust, pour in a few thousand tons of rock-busting molten iron, and then toss in a grapefruit-size instrument designed to ride the plunging elevator of liquid metal to the planet's core.

That scenario sounds like science fiction. Even its author, geophysicist David J. Stevenson of the California Institute of Technology in Pasadena, calls the proposal "highly speculative." However, in the May 15 Nature, he contends that such a mission to explore Earth's interior is technically feasible.

People have drilled into our planet's crust to investigate conditions there, but even the deepest borehole penetrates only about 10 kilometers, says Stevenson. Seismic analyses suggest that the continents, the thickest portions of Earth's outer layer, are between 200 and 250 km thick (SN: 5/3/03, p. 285). Below that hardened crust lies the viscous mantle, which surrounds a liquid outer core and a solid inner core that are both made predominantly of iron.

All that iron sank to Earth's core because it's so dense, and Stevenson's plan would exploit that property. If scientists pour molten iron into a narrow crack at least 300 meters deep, the pressure at the bottom of the fissure would be enough to fracture the rock there, Stephenson says. As the crack grows deeper, the molten iron would flow downward and maintain pressure at the crack tip. The self-propagating crack--which high pressure in deep rocks would seal after the iron passed by--would progress at about 5 m per second and reach Earth's outer core in about a week.

The energy required to blast the initial crack, which Stevenson estimates should also be at least 300 m long and about 10 centimeters wide, is equivalent to the explosive power of several million tons of TNT or a single modest hydrogen bomb. The molten iron needed to fill such a fissure--about 10,000 cubic meters--is the volume produced by all the world's foundries in an hour.

Getting the molten iron to Earth's core wouldn't be hard, says Norman H. Sleep, a geophysicist at Stanford University. Bigger challenges would arise in developing a suitable probe. Temperatures deep within Earth rise to 1,800[degrees]C, and pressures there are more than 1,000 times those found at the greatest ocean depths, he notes. In such conditions, most metals that probe-makers might use corrode in the presence of liquid iron or dissolve into it, and the electronics inside the device probably wouldn't fare well, either.

"This idea is workable but needs to be thought out carefully to be sure the probe measures something that's useful," says Sleep.

For example, on its plunge through the mantle, the instrument could provide information about a portion of Earth's interior that has been inaccessible to direct observation. Raymond Jeanloz, a geophysicist at the University of California, Berkeley says that such a probe could provide detailed data about the density and composition of rocks, properties that drive the movement of materials within the mantle and of tectonic plates at the planet's surface. Using an automotive metaphor, he asks: "Wouldn't it be nice to lift up Earth's hood and see what's going on inside?"