THEY call these the Barren Lands, and it’s easy to see why. From a helicopter it looks as though the glaciers departed yesterday, leaving a wasteland of jumbled boulders and naked granite outcrops. There is no trace of human presence, but water is everywhere – puddles, ponds and lakes cover half the surface.
In the midst of all this emptiness 300 kilometres north-east of Yellowknife – the capital of Canada’s Northwest Territories – lies Lac de Gras, half-covered by ice even in midsummer. On one of its small islands is a gravelled airstrip large enough for a Boeing 737. Nearby sits an open mine pit and a small cluster of blue and white buildings. This is Diavik diamond mine.
A decade ago, after the first diamond strike just a few kilometres from Lac de Gras, this area was ground zero for one of the greatest mineral rushes the world has ever known. Prospectors staked the perimeters of their claims – some 20 million hectares in all – by pitching wooden stakes out of the doors of helicopters as they roared low over the tundra. Already Diavik and its nearby competitor Ekati, the only two mines up and running so far, have produced more wealth than a century of gold mining, the previous mainstay of the north.
Advertisement
Diavik is 60 per cent owned by the UK-based multinational mining giant Rio Tinto and may be the richest diamond lode in the world, yielding up to 5 carats per tonne of ore, more than five times the world average. From the many tonnes of rock that are mined each day come about a hard-hat’s worth of diamonds, about a third of which are gem quality. At this rate, the company plans to extract more than 7 million carats this year, valued at over CAN$500 million (US$408 million). “The economic returns are unparalleled in the mineral industry,” says Scott Cairns, district geologist for the Yellowknife region.
But this desolate area is not only home to hundreds of mine workers and prospectors, it is also visited by people who are not as interested in making money, though they are interested in diamonds. Geologists are here to find out what Canadian diamonds can tell them about the Earth’s history. The burgeoning industry has sparked huge investment in geological research. “When academic interest and commercial requirements coalesce we can help each other out,” says Alan Jones, a geophysicist at the Dublin Institute for Advanced Studies and the Geological Survey of Canada.
To discover a diamond reserve, you need to find the signature of one of the planet’s rarest, most violent volcanic events. Carbon only forms into diamonds at temperatures low by upper mantle standards and high pressure, and these conditions are only found hundreds of kilometres below ground. But few diamonds ever make it to the surface because as they rise, the pressure on them decreases and the crystal begins to rearrange itself into graphite.
To survive as far as the surface, diamonds have to be carried upwards fast on a surge of magma. When gas-rich magma explodes upward at nearly the speed of sound, it rips debris from every stratum it passes through, including diamond-bearing regions. At the surface the gases blast out a narrow cone-shaped hole in which the magma and debris cool and solidify, creating a vaguely carrot-shaped tube of volcanic rock called a kimberlite “pipe” (see Graphic).
The pipes are relatively small – usually only a few hundred metres in diameter – and very few and far between, so finding one is tricky. Prospectors normally look for indicator minerals in river or glacial sediments – garnet, for example, is often associated with diamonds. They then look upstream – or up-ice – for the kimberlite they eroded from. This is how small-time prospector Chuck Fipke found the first Lac de Gras kimberlites in 1991.
Nowadays exploration companies use more sophisticated methods: aeroplanes equipped with magnetometers or gravity meters search for unusual patches of rock. But these aerial surveys are expensive, so companies are looking for ways to focus their search. “The real problem facing the industry is one of area selection,” says William Griffin, a geochemist at Macquarie University in Sydney, Australia, who has worked extensively in the region. “Where should you be looking, and where would you be completely wasting your time?”
Ancient cratons
And that’s where the scientists come in. No one knows exactly how diamonds formed, but we do know that they are found in the oldest parts of continents, known as cratons. These are regions where the rocks have not changed since the Archaean, more than 2.5 billion years ago. The Earth’s hard outer shell, or lithosphere, is unusually thick here, perhaps providing the high-pressure conditions that diamonds need.
Mining companies and the Canadian government are hoping that understanding more about the Archaean might point them to cratons containing economically mineable diamonds. For geologists, of course, the chance to understand more about Earth’s adolescence is interesting in its own right. “Some people say that’s one of the few big questions left in the earth sciences,” says David Snyder, a geophysicist with the Geological Survey of Canada in Ottawa.
Snyder thinks that diamond lodes could be a result of our planet’s adolescent efforts at plate tectonics. Today plate movement is driven by the simmering heat of the upper mantle, but things were different in the Archaean. The Earth was still cooling down after its formation, and the rocks contained high levels of heat-generating radioactive isotopes. Many scientists think the mantle was at a rolling boil rather than gently simmering, and if that was the case, then the way continents formed would have been very different.
When two plates collide now, one usually plunges deep into the mantle. But in the Archaean, the drifting plates would have been warmer, less dense and thus more buoyant says Michael Bostock, a seismologist at the University of British Columbia in Vancouver. As a result, one plate would have tended to slide nearly horizontally under the other. Such “shallow subduction” – still seen today in a few places off the South American coast – could have created the thick continental lithosphere of today’s cratons.
Snyder is trying to confirm this theory by imaging the Earth’s crust. He has an array of small solar-powered seismographs strung across north-west Canada’s Slave Craton, which measure seismic waves coming up from Pacific earthquake hotspots. Wire spikes on his equipment keep peregrine falcons from pecking delicate components and metal boxes prevent caribou from nibbling the wiring. His efforts have paid off. By analysing delays in the arrival times of waves from earthquakes happening at known times, Snyder has found evidence that the lithosphere of Slave Craton has two layers, with a sharp boundary around 120 kilometres down (Lithos, vol 71, p 529). Snyder and Bostock think these layers are the result of two plates stacking during a subduction event some 2.6 billion years ago.
As supporting evidence, they point out that a type of rock called eclogite can be found in diamond-bearing kimberlites. It forms from oceanic basalt subjected to very high pressures. These findings suggest that carbon-rich sediments from the seabed, taken down by the subducting plate, may have provided the raw material for diamond formation.
If Snyder is right, then finding two-layered cratons with old subduction zones would be key. “We would say that’s where you want to look for kimberlites,” Snyder says. He also suggests how mineable kimberlites might be found (see “From India with love”).
However, there are other explanations for the layering Snyder has seen. Griffin, for example, thinks the lower layer of Slave Craton formed from a plume of magma that erupted deep within the mantle, bringing diamonds up along with it. “We think it just came up and plastered itself onto the lithosphere that was there previously,” says Griffin. This would make it less likely that old subduction zones that did not experience such a plume would contain diamonds. He notes that many of the diamonds at Lac de Gras have minerals within them that could only have come from much deeper within the mantle. No one knows yet which of these theories, if either, is true. “Dave and I have agreed to disagree,” says Griffin.
Follow the graphite
Another possible problem with Snyder’s theory is that some cratons, such as South Africa’s Kaapvaal Craton, look like a single, very thick layer. But that can be explained, says Richard Carlson, a geochemist at the Carnegie Institution in Washington DC, who has worked extensively in the Kaapvaal. He thinks both the Kaapvaal and the Slave began as two-layered structures. The Slave remained frozen at this stage, while in the Kaapvaal – and perhaps other cratons as well – the two tiers gradually integrated, forming the seamless lithosphere we see today. “Whatever happened there, it managed to heal itself better than the Slave Craton did,” Carlson says. But his theory is far from complete: why the different outcomes? “I have no idea,” Carlson says.
Jones has another suggestion for prospectors. He thinks diamonds can be found by looking for graphite at shallower depths. In collaboration with Herman Grütter, a geoscientist with Mineral Services Canada, and others, Jones images the lithosphere by measuring electric currents induced within the Earth by the same solar magnetic activity that causes the northern lights. The team found a zone of unusually high electrical conductance about 80 to 120 kilometres directly below Lac de Gras. They believe this is graphite – a good conductor. And the temperature and pressure at this depth are right for graphite, too. Just below, where the pressure is higher, is the diamond zone. “If that rock type had been deeper down, all that carbon would have been diamond,” says Grütter.
There are other ideas out there too. Thomas Stachel, a geologist at the University of Alberta, has discovered that diamond-bearing Archaean rocks show evidence of exposure to molten fluid from deeper down, and Stachel thinks this may have provided the carbon. The question is what this fluid was made of. It could have been methane-rich material from the deep mantle that oxidised when it hit the crust, releasing elemental carbon. On the other hand, it could have been carbonate-rich fluid that underwent reduction – the opposite chemical process to oxidation – again producing elemental carbon. Stachel’s team is measuring the oxidation state of the rocks enclosing the diamonds to find out which process happened.
The answer could be of immense practical significance, say Stachel. If diamonds formed from a seepage of carbonate-rich fluid, then prospectors should look for rocks that have been exposed to just the right amount of such fluid – too little and there would not be enough carbon to make ample diamonds, too much and the oxidation would destroy the diamonds again. On the other hand, if diamonds formed from methane-rich fluid, then the more exposure to that fluid, the better.
“There is a revolution in the way people think about diamond exploration”
While geologists work out which theory is right, mining companies are already acting on their ideas. For example, exploration companies are venturing beyond the traditional cratonic regions in search of “pericratonic” areas: regions of Archaean crust that may be hidden amidst newer bits of the continent. On the strength of possible Archaean remnants, mining companies have staked claims to every bit of the Melville peninsula in Nunavut (see Map). “All the big boys are there,” says Grütter. Meanwhile, the world’s largest diamond-mining company, De Beers, is working with Jones to apply the same techniques to the Kaapvaal Craton. “I genuinely think there is a revolution in the way people think about diamond exploration,” Grütter says. “And you can see it in the way they are spending their money.”
From India with love
Kimberlite pipes of the same age are often found in rows, according to David Snyder of the Geological Survey of Canada. He thinks cracks in the Earth’s crust allowed kimberlite plumes to surface, and this would explain their linear arrangement. Not all kimberlites are economically mineable, but those of the same age often have similar properties. So Snyder recommends that when mining companies find mineable pipes, they should look for a pattern of other deposits.FIG-mg24814801.jpg
The cracks are probably caused by tectonic plate collisions, he says: models show that a plate being pushed forward as new crust is created in the ocean will develop cracks parallel to the direction of movement, while a plate being sucked down into a subduction zone will tear perpendicular to the direction of pull.
Some of the 48-million-year-old kimberlites at Lac de Gras form neat north-east to south-west lines. Their orientation does not match that of any other geological features, but it does correspond to within a degree or two to the direction in which the North American plate is now moving. And 48 million years ago is exactly when India smashed into Asia, raising the Himalayas and sending a ripple of stresses around the world. Perhaps, thinks Snyder, Canada has that collision to thank for its diamonds.