Geology Liveblog: Magmatic Ore Deposits, or Everybody Loves Sulfur

[wolffyluna]

It's time for another geology liveblog! This time we're focusing on magmatic ore deposits, what you find in them, and some notable examples of them.

But first up, we have some terminology.

Terminology: What is an ore, why does sulfur matter, and why geologists want the D (and the R)

So, 'ore' isn't... necessarily a geological term. Something is an ore if mining it and extracting what we want out of it is economic to do. Something can have a lot of something, and still not be an ore. Olivine has an awful lot of iron in it, for example, but you're not going to mine it unless something has gone seriously wrong somewhere. But mining magnetite is a thing you could totally do. A lot of ore minerals are sulfides, because they tend to a) concentrate things to an amount where you could mine them and b) tend to suck up goodies we wouldn't find in other phases.

This tendency for good stuff to be in sulfur is often explained using Goldschmidt's Classification. Goldschmidt came up with four catergories that elements tended to fall in (with the note that a lot of elements fall into more than one bucket.):

• Atmophiles: These are volatile elements that like to be in the atmosphere, eg N, O, H.
• Lithophiles: These are elements that like to be in silicate rocks, eg Mg, Al, K.
• Siderophiles: These elements like being in molten iron, in the core of the Earth, eg Ni, and the Platinum group elements (or the PGEs, because geologists love TLAs (...three letter acronyms.))
• Chalcophiles: These elements love to be with sulfur, eg Cu, Ni and the PGEs.

Ni and PGEs are the elements that are often found in dry magmatic ore deposits, which we'll be focusing on.

Now, to concentrate an element into something that could be called an ore in magmatic deposits (among others) there are two important factors: D and R.

D is the partition coefficient. It's measured by dividing the concentration of something in one immiscible fluid by the concentration in another. A high D (above 1) means that that element prefers the fluid on the top of the divisor, with higher numbers meaning higher preference. For example, the partition coefficient of Pt between sulfide and silica melts is 10,000. Maybe more. Pt's preference for sulfide is so strong it's hard to measure. D explains why some elements concentrate in sulfide minerals. But it doesn't necessarily control how much things concentrate by.

The factor that's important for how concentrated things get is R. R is the ratio between the sulfide and the silicate melt. R is tricky, because you simultaneously want it small (so that the sulfide "sees" a lot of silicate melt and gets very concentrated in chalcophile elements), but you don't want a small amount of sulfide (because one nugget of sulfide is a sad mine.) Often, that means you sulfides that have been in an environment with an absurd amount of silicate melt.

Magmatic ore deposits tend to form from mafic (mantle like) melts that have had a high amount of melting, that have been saturated in sulfide.

They need to be high degree melts because the metals in magmatic ore deposits (nickel and the PGEs) tend to only melt when the rock around them has mostly melted. Nickel tends to only go into the melt when things have melted a lot because it tends to be in olivine, and olivine doesn't melt easily. As for the PGEs-- PGEs don't like doing things. And one of those things they don't like doing is melting. So you need to melt a lot to get any in there.

Sulfide saturation is tricky, because most of these magmas melt because they're rising into a lower pressure area, and sulfur gets more soluble at lower pressures. So you often need to add sulfur into the system somehow.

But once you have got saturated in sulfur, it forms an immiscible fluid with the silica (like oil and water). The sulfur is quite dense, and sinks, picking up chalcophiles along the way. So you often get this big layer of sulfide, with some silicate crystal floating in it, towards the bottom of magma chambers.

Noril'sk: Terrible Place, Sensible Deposit

Norilsk is terrible. It's the most polluted city in the world, and I think at one point it was a gulag. But, as a magmatic PGE deposit, it does at least make sense.

Norilsk formed at the same times as the Siberian Traps. The Siberian traps is a huge plain of basalt, that came up onto the crust through feeder zones like Norilsk.

Norilsk is notable out of those feeder zones because that huge amount of magma came through a layer of evaporite. Evaporite layers have a lot of sulfide. Which meant that as Norilsk formed, the sulfide in the evaporite layers "saw" an awful lot of silicate melt, and grabbed all the chalcophiles out of it. Norilsk is such a rich deposit because of this, that it's often said to contain the entire periodic table (...which is part of why Norilsk the city is so polluted. They just has a lot of stuff they don't know how to deal with.)

Bushveld: The Massive Weirdo

The Bushveld is a massive layered mafic intrusion in South Africa. There are some layers in this mafic intrusion, like the Merensky Reef and UG-2, that contain most of the world's accessible platinum.

The thing is, the Bushveld shouldn't have that much platinum. There is just too much platinum. The sulfides had to get that platinum from somewhere, and while the Bushveld is big, it's not big enough.

Layered mafic intrusions are intrusions that cool slowly, so that different minerals crystallise out at different times, and form layers. When they're sulfide saturated, you get sulfide layers. Sulfide is really dense, so it sinks to bottom, and you find layers with big silicate crystals cemented by sulfides. You can get different kinds of layering, but the Bushveld has a slightly odd one. The Bushveld has a predictable sequence of layers, that follows the patterns of most other layered mafic intrusions-- that every so often gets stopped part way through, then starts again. The current theory for why that happened is that there was repeated intrusions over time. This can potentially explain some of that extra platinum in it, because it means that those sulfides were seeing a lot of silica, and thus a reasonable amount of platinum.

But there's another potential explanation. All the platinum rich layers are in chromite layers. The thing about chromite formation is that it reduces the magma around it-- in the redox sense meaning it sort of ...spits out electrons into the surrounding magma. Sort of? Lots of things affect the solubility of platinum. One of those things is how oxidised the environment is. If platinum is in a reduced melt, it's way less soluble. There have been some experiments showing but if you crystallize chromite, sometimes get little nuggets of platinum forming around it. Which could potentially explain why there are so much platinum around those chromite layers. As the chromite formed, it sucked the platinum out of solution.

Sudbury: A Mess from a Meteor

I hate Sudbury. Sudbury doesn't make any sense.

Sunbury formed from the impact of a meteorite. This meteorite melted a lot of stuff. That at least make sense.

What doesn't make sense is that it's one of the largest nickel deposits in the world. Because nickel comes from mafic intrusions. Sudbury is incredibly continental, aka felsic.

I have no idea where the nickel in Sudbury came from. It haunts me.

There are a couple of potential explanations. One is that the melting from the the impact let a lot of sulfide see a lot of continental crust and it was enough that there was a non-depressing amount of nickel. Which is possible, but a little weird. The other explanation is that the meteorite that hit Sudbury had a lot of nickel in it, and the melted sulfides just picked that up. Iron meteorites often do have a lot of nickel in them, and the one Sudbury was probably big.

I don't know which explanation I like the least. I don't like either of them. But Sudbury is one of the world's largest nickel deposits, and I guess I have to learn to love this weird mess.

Comments

[yvannairie]

When you say PGEs don't like doing anything, does this also mean they're chemically stable?

[wolffyluna]

Part of it is that they are very chemically stable, but part of it is-- they don't really like being anywhere other than the core, and maybe in sulfides. They're not super happy being in the mantle, but they like being in mantle melts less, which means getting them into crust is challenging.

(The other problem with PGE deposits is that PGEs are in such low concentrations in most places, they have to get concentrated by absurd amounts to get a deposit.)

[yvannairie]

Oooh okay so it's more just. Heavy and viscous and doesn't like moving around. About the only thing I know about platinum is that it's insanely rare.

[wolffyluna]

Platinum is pretty hecking rare. The reason that there's any that's not stuck in the core is because the Earth got hit by meteorites with platinum in them after the core formed.

It's not so much heavy and viscous (though it is both of those things) so much as not really chemically "fitting" into things that aren't molten iron or sulfides. Which means in silicate phases it kind of sits there, not doing anything, because everything kinda sucks at fitting it in. (It's also not super keen on chemical bonding. Platinum group elements are nicknamed the 'noble metals' and while they are not as bad as the noble gases when it comes to bonding, they're not spectacular at it either.)

[brin_bellway]

I'm glad to see that geology liveblogs are back. :)

(Were you dictating this, by any chance? Towards the end there are a lot of things that look like automated-transcription errors, to the point that it gets a bit difficult to read.)

[wolffyluna]

I was dictating the last half, because I thought it would be more "efficient"... and thought I'd caught all the errors. Woops, sorry about that. The errors should largely be fixed now.