Messengers from the Mantle

David Pacchioli
December 01, 2004

The rocks on Chuma Mbalu-Keswa's table, she admits, are not exactly eye-catching. "They look uninteresting, I know," she writes in a note penciled neatly on the back of a departmental announcement.

"But," the note continues, "look at the photographs!"

Lined up at the table's edge are five rugged chunks of kimberlite. They are blue-gray—sooty, which makes sense since they were lately rescued from a coal mine in southwestern Pennsylvania. Flecks of mica shine through the grime of their surfaces, glittering like nightclub chic.

The photographs, arrayed on a board behind these clunky specimens, are—true to Mbalu-Keswa's promise—something else again. They represent thin slices (shavings, really) taken at cross-section, lit from below, and magnified 40 to 80 times. It's a little tough to fathom that these are the same rocks.

Splashes of vivid hue — buttercup, fuschia, azure—shimmer as though submerged, in all their igneous magnificence. Spidery cracks zig-zag from imbedded islands of color. The luminous elegance of these images recalls the poet Gerard Manley Hopkins:

  • The world is charged with the grandeur of God.
  • It will flame out, like shining from shook foil...

The photographs unveil the secret of these "uninteresting"-looking rocks. The colors represent inclusions, mineral samples drawn from far below the planet's crust, a direct record of the earth's interior that can't be obtained from any other source.

Kimberlites are the crystallized remnants of primeval magma, molten rock which pushed up from depths of 200 kilometers and more below the earth's surface, from the changeable and largely mysterious region known as the mantle. Charged with volatile gases, this magma moved under great pressure: Some geologists believe it jetted up at speeds of mach 2, twice the speed of sound.

Sometimes it found a path all the way to the surface and blew skyward in small volcanoes. In other cases, it was trapped in the crust and dissipated its force by squeezing like toothpaste into fingerlike fissures, eventually cooling and hardening. These deposits remain in narrow dikes or fissures laced under the surface, dikes which may end in even narrower pipe-like conduits, called diatremes.

Wherever it came to rest, this roiling proto-rock brought plenty of souvenirs along for the ride, samples from down deep and samples from every crash and bend along the way. When the magma cooled, these fragments were encased, captured forever as a record of the places left behind. As such, they represent a veritable window into the mantle.

The most exotic of kimberlite's inclusions are diamonds.

Diamonds form at depths of over 150 kilometers, down where pressures are high enough to compress carbon into its densest form. Kimberlite magma carried these rare crystals up to the surface, where they were first stumbled on in streambeds.

Then in 1868, diamonds were discovered on farmlands in a remote part of Cape Province, South Africa, far from any river's banks. In the ensuing scramble, these gems were traced to pipes of blue roc —diatremes. The material and the mining town that sprang up were later named after the British Colonial Foreign Secretary of the time, Lord Kimberly.

Diamonds appear very rarely even in kimberlite; when found, only the tiniest fraction are valuable. "A really rich sample," says Duff Gold, Penn State professor of geology, "would contain about 100 carats of diamond per 100 tons of rock." A carat, he adds, weighs a fifth of a gram. Even so, the link between kimberlite and diamond has propelled intense research (not to mention mining) in South Africa since the 1880s. Recent discovery of large kimberlite deposits in Canada has even caused a bit of latter-day diamond fever, with land claims being filed all over extensive areas of Saskatchewan and the North West Territories.

Peter Deines, Penn State professor of geosciences and a member of Mbalu-Keswa's thesis committee, covets diamonds for non-commercial reasons: he studies their isotopic composition to learn about the earth's carbon cycle.

Carbon, Deines explains, has two stable isotopes, carbon-13 and carbon-12. "That is, the element carbon contains two types of carbon atoms, whose atomic mass is 13 and 12, respectively."

In a chemical reaction, these two atoms behave slightly differently. In the larger context of geochemical processes, this means different manifestations of the ubiquitous element have different ratios of carbon-13 to carbon-12. Or, as Deines puts it, "the 13/12 ratio of carbon in the sea is different from that of carbon in the atmosphere and that in plants.

"Thus," Deines continues, "we can learn about the processes by which a material was made based on its carbon isotope record. We can also trace the movement of carbon through the geochemical cycle."

Carbon does not stay put. Like the other elements, it is in constant motion from one big geochemical reservoir to another, cycling slowly from the crust to the mantle and back to the surface.

Kimberlites — and the diamonds they bring up — offer Deines a reflection of the carbon in the mantle — the thick hot-rock substrate below the Earth's crust that takes up 80 percent of the planet's volume, extending 3000 kilometers down, all the way to the dense sphere of the core. By looking closely at diamond isotopes, Deines hopes to sharpen the picture of the insides of the Earth.

Over the course of the last two decades, he relates, the geologist's view of those insides has been considerably changed.

"What we have learned is that our old picture of the mantle was pretty simple. Since the mantle is presumably very hot and homogenizing processes occur in it, we thought that its composition should be the same everywhere."

Now, through isotopic studies and other chemical analyses, "we have found that the mantle is in fact very complicated. It is not homogenous. The range of variations in the 13 to 12 ratio is very large. This was completely unexpected."

Deines offers three possible scenarios. The new data, he says, could reflect changes continually occurring in the mantle itself. It could be read as a remnant of the original formation of the Earth. ("Meteorites and mantle rocks have nearly identical 13/12 ratios.") Or, says Deines, the variations he and others have seen could be an indicator that diamonds are the remnants of carbon from the surface that was returned to the Earth's mantle.

"There's no clear picture yet," he concludes. "But there are some exciting possibilities." If, for example, carbon really is subducted to form diamonds, how do these diamonds sometimes sink to depths of 400-500 kilometers?

Much more commonly than diamonds, kimberlites ferry a host of other rocks and minerals to the surface. These less glamorous inclusions may provide crucial clues to the mantle's composition and its nature.

These foreign fragments, called xenoliths, aren't always tiny. A sample proffered by Deines requires two hands to hoist. It is as dense as a hunk of cannonball and similarly rounded, smoothed like a pebble by the churning of its upward journey.

In the old days, Deines says, when kimberlite was mined for diamonds, miners would spread it out on the ground and let it weather. Soon enough, the soft kimberlite would crumble and "rot," leaving the harder xenoliths for harvest. Unfortunately, this susceptibility to weathering makes kimberlites a problem both to find and to work with. Arnold Doden, a Penn State graduate student in geosciences, is mapping kimberlite intrusions scattered across the range-lands of Montana. It is frustrating work.

"These deposits are very small," Doden says, "maybe a few meters wide. And they are generally so altered at the surface it's hard to tell what they are. Once exposed, they deteriorate pretty rapidly."

That's why the samples being studied by fellow graduate-student Mbalu-Keswa are so important. Unearthed from the Tanoma coal mine in Indiana county, where they were a bit-wrecking nuisance, they were later handed over to Mbalu-Keswa's adviser Duff Gold by mining-company geologist Joseph Tedeski. They were well-protected, and so they are unusually well-preserved.

Gold, a South African, has a built-in interest in kimberlites. He figured the same would be true about countrywoman Mbalu-Keswa, who had come to the United States in 1988 to attend Smith College. He was right. "When he told me about these samples I was very interested," Mbalu-Keswa says.

When she began to slice these kimberlites into thin sections and look at them closely under a microscope, her interest grew.

"One day," she remembers, "I called Dr. Gold over to look. I said, 'Duff, what's this?'"

Not a trace of diamond. But the tiny sections Mbalu-Keswa was painstakingly searching displayed some inclusions that were very different from anything she or Gold had seen before. Inside megacrysts — embedded crystals of pyroxene and garnet — were tiny pockets, the largest a millimeter across, that contained several different minerals each. "Most of these have more than five. I counted twelve in one of them."

The standard inclusion, she explains, is a single mineral. An observer sees a simple mass, stuck like an island in the grainy host crystal. Around it, there is usually a thin fuzzy ring.

"It's pretty easy to figure out what happened," Mbalu-Keswa says. The surrounding crystal grew around a blob of the molten rock from which it precipitated. The ring is left over from the reaction that let these two minerals reach a state of chemical equilibrium: as Mbalu-Keswa puts it, it is the result of two disparate materials struggling "to be happy together."

With Mbalu-Keswa's inclusions, by contrast, "It's a much more complex situation." Not just one mineral embedded in another, but a miscellaneous group — an inclusion like a rock in itself. Furthermore, as she points out, there's no reaction ring surrounding all this mayhem. What circumstance could cause such unusual conglomeration?

It will be a difficult problem to unravel. Only one other researcher has reported multiple inclusions, a Canadian working on kimberlites from Kentucky. And adding to the complexity of the picture is the difficulty of figuring out which inclusions and alterations are "original," dating from the magma stage, and which are "secondary" additions, the results of subsequent magmatic events.

Deines ran into this problem 30 years ago, as a graduate student working with kimberlite from nearby the Tanoma mine, at Dixonville, PA. "That's why I got interested in diamonds," he says, with a wry smile. But, he quickly adds, "Chuma is looking at much fresher sections. Freshness makes all the difference. And we can do things we couldn't do before."

For Mbalu-Keswa, the first step was to identify the elements present in the inclusions using an electron probe microanalyzer, a machine that bounces electrons off the section surface, providing a precise breakdown of chemical constituents. She found mostly phlogopite, a type of mica, along with lesser quantities of the minerals ilmenite, magnetite, apatite, perovskite, and olivine.

The more difficult matter facing Mbalu-Keswa now is to somehow deconstruct these samples—hypothetically, that is—in order to arrive at the composition of the original kimberlite magma. "This is the real importance of what she's got," Gold stresses. "We believe we have some of the original kimberlite magma trapped in these crystals."

"It's like having a pot of soup," Mbalu-Keswa says, "and trying to figure out from the finished product what were the original ingredients. There is tomato present, but was it the fresh vegetable when it was put into the pot, or was it paste?"

Petrologists increasingly rely on computer programs to help sort possible combinations, working out a laborious process of elimination. "There are ways to figure out what can and cannot co-exist," Mbalu-Keswa says. "You eliminate until you find the combination that makes the most sense. Really, though," she adds, "it's an educated guess. No one knows for certain."

Even if she can unravel the story these strange inclusions have to tell, Mbalu-Keswa acknowledges that it will be hard to extrapolate from her small samples to any meaningful generalizations about Earth's mantle at large. "It would be nice if we could find some more samples," she suggests. "Some from Russia, maybe, or Southeast Asia. That would give us a better range."

Still, she concludes, a clear understanding of the Tanoma kimberlites might offer fresh insights into what goes on below the Earth's hard crust.

Chuma Mbalu-Keswa received an M.S. in geosciences in August 1994, and is currently working on a Ph.D. Her adviser, David P. (Duff) Gold, Ph.D., is professor of geology, 339 Deike Building, University Park, PA 16802. Peter S. Deines, Ph.D., is professor of geochemistry, 207 Deike. Arnold Doden is a Ph.D. student in geosciences.

Last Updated December 01, 2004