Evan Smith on Diamonds from the Deep Mantle

This is Geology Bytes with Oliver Strumple. Diamond is a high pressure form of carbon. Most diamonds grow at depths of 150 to 200 kilometers before being ejected up to the surface through deep volcanic pipes called kimberlites. A few diamonds originate from much greater depths but they were thought to consist only of small, heavily fractured specimens with many inclusions and so rarely of gem quality. But in 2016, Evan Smith and his co-workers discovered that some of the largest and most highly prized gems were also formed deep in the mantle at depths of 600 to 650 kilometers. Since such diamonds are effectively messengers from the deep mantle, they have become some of the most valuable gemstones from a scientific standpoint as well. What can we learn from them about the deep mantle and the processes that led to their formation at such extreme depths? Evan Smith is a senior research scientist at the Gemological Institute of America in New York. Evan Smith, welcome to Geology Bytes. Thank you. It's a pleasure to be here. I appreciate you having me. We've been mining diamonds for centuries. How come this ultra-deep class of diamonds was only discovered so recently and how did you and your colleagues identify them? So we've been mining diamonds for a long time but we've only developed a geological understanding really since the development of electron microprobe and we could really analyze inclusions in diamonds and develop an understanding of how they form. Now there are some diamonds that we know more about than others and the kind of diamonds that I started studying were ones that typically are not available to researchers. So people have known for a long time that larger and higher quality diamonds that seem to have few inclusions appear to stand out from other kinds of gem quality diamonds. They look different in their physical characteristics but it had been a long time that people had noticed that and no one had been able to take a detailed look at why that might be the case. So what I was able to do with my colleagues was parse out some of those individual samples and take a detailed look at what their inclusions are and we recognize that actually this is another class of diamonds, another variety of diamonds that form an distinct way. So it was actually the inclusions and you had to actually do some microscopic analysis in order to notice that this was a distinct class from the other gem quality diamonds. Yeah and when I just say distinct class we're talking about in general type two diamonds. So diamonds have nitrogen as their most abundant impurity and some diamonds have enough that you can see it with infrared spectroscopy. Some diamonds have too little nitrogen to see with infrared spectroscopy and we call those kind of nitrogen poor or nitrogen free diamonds. Those are type two. Overall those are a very rare category maybe one or two percent of mine diamonds are that nitrogen poor category but when you look at say diamonds over 10 or 20 carats that make up the best and biggest gemstones there's a huge proportion of them that are type two. So it stands out like a sore thumb like why is there that strange relationship with size and quality and color of diamonds. That's the reason why gemologists started to regard them as a maybe distinct and special category and it took geologists a long time to understand why those larger quite often type two diamonds might be different and the difference comes down to studying their inclusions and getting a sense of the environment that those diamonds form in and it turns out that they form in a completely distinct environment from all other kinds of diamonds. So some of the distinctive features of these diamonds were actually used to come up with an acronym which is the way these diamonds are commonly referred to. Could you explain that? Yeah, absolutely. So I kind of loosely refer to this category as type two but if you want to actually have a meaningful conversation about them you have to recognize that there are some type two diamonds that don't fit in that category and there are some diamonds that are clearly in that category that aren't quite type two. So a more meaningful encompassing of their whole physical characteristics set is put into this acronym called Clipper. So the diamonds that I'm referring to are cullinan like that's the C. So cullinan is the largest gem quality diamond ever found. It was 3,106 carats in its rough state and this is sort of a type diamond for this category. So diamonds that are like that diamond cullinan like overall as a group tend to be large, they tend to be inclusion poor, they tend to be very pure so they have very little nitrogen. They are irregular in their shape so they're very seldom a nice crystal shape and they're quite often very heavily resorbed. That's the R. So they're sort of dissolved and etched in their outer surface characteristics. Okay, so clipper C-L-I-P-P-I-R, clipper diamonds. That's right. Kind of alluding to clipper ships but spelled a little bit differently to distinguish them. How do we actually know that they originated 600 to 650 kilometers below the surface which is about three times deeper than the vast majority of diamonds that we find? So the depth of origin for clipper diamonds comes from recognizing the mineral inclusions that they sometimes contain and there are a couple key phases. One is majoritic garnet. So this is a special high pressure form of garnet that contains an excess of silicon. So normally you have three atoms per formula unit of silicon in a garnet structure and majoritic garnet has a little bit more than that. So there's sort of a coupled substitution of silicon, aluminum, sodium, titanium, chromium into the garnet structure that makes it majoritic in its characteristics. And we know this is something that can't form in the lithospheric mantle. It requires higher pressures. So when you have majoritic garnet in a diamond it indicates that it comes from greater depths. So we have some examples of majoritic garnet in clipper diamonds. We also have examples of calcium silicate minerals, namely brayite. And we think that this is something that is retrogression or inversion product from the higher pressure phase called calcium silicate perovskite or Dave Maui as its mineral name. So again this is a mineral that was trapped in diamond that we think could only have been trapped at very great depths. And I in the 2016 paper wrote a looser bracket somewhere between 360 and 750 kilometers deep in the earth, but I think putting it somewhere within the mantle transition zone or top of the lower mantle is probably about right in terms of the processes and the environment where these diamonds might grow. We also have some information from geobarometry that indicates that they come from higher pressures, but the biggest constraint for the depth for clipper diamonds really comes down to just the observation of these high pressure minerals that could only have formed at great depths. Do you think it's possible that some diamonds might have been able to form at even greater depths? It's just that they just don't make it to the surface or we haven't found them yet? There are some people who believe that there are diamonds formed at the lower mantle core boundary and I wouldn't be so quick to rule that out, but I think at present we don't have any really firm evidence that shows diamonds must come from that depth. And based on our understanding of how these diamonds are forming, by and large they involve subduction and ingredients derived from subducted slabs and the processes going into them seem to occur where things are happening to subducting slabs where they're losing volatiles and we think this is something that makes a lot of sense as a subducting slab reaches the mantle transition zone and lower mantle and it's losing volatiles and diamond formation is part of that process. So you mentioned some telltale minerals such as the majoritic garnet that tells us they actually originated from those extreme depths, but how can they survive if the pressures and temperatures appear on the surface? The diamond itself is a very robust crystal structure. It has an incredible strength and almost chemical inertness. So it can survive across a wide range of geological conditions after it's formed. We know that based on the morphology of clipper diamonds, a lot of them appear to have been broken in the mantle and they've certainly been very strongly resorbed. So a lot of the diamond might actually be missing. It's been dissolved or oxidized away, but the great thing about clipper diamonds is that because their average size is so large, it appears as though we still have a lot of diamond to deal with even after we've destroyed a lot of it. So we end up with diamonds making their way all the way to surface by a couple transport mechanisms. There's some initial stage of upwelling or vertical ascent for the diamond, but we know the final stages of diamond ascent involve kimberlites or related mantle derived magmas that sweep diamonds from a depth of say 200 kilometers all the way up to earth's surface. That journey is not necessarily friendly to the diamonds. It can etch or dissolve or resorb away some of the diamond. But when we see them, we know that they've retained at least some of their diamond material because we have enough to cut and polish. And you mentioned that even though their inclusion poor, which is the IP and clipper, they still have some inclusions. So how do those inclusions survive under the enormous changes of pressure? You think they just kind of explode them below the diamond to pieces? Yeah. So the diamond doesn't change very much in its volume as you change the pressure. It's not very compressible, but the minerals trapped inside it, relatively speaking, are much more compressible and they want to expand as you bring that diamond up to surface. So they exert a tremendous amount of pressure on the surrounding diamond and the diamond can crack locally around an inclusion. It can deform locally around an inclusion in response to that tremendous pressure. But to some degree, that tremendous pressure can actually be preserved in some cases. And we do see inclusion sometimes with one, two, three giga pascals of pressure. Even I think the highest is even over 10 giga pascals of pressure actually preserved at surface conditions inside an inclusion in diamond. So some of those tremendous pressures are another strong piece of evidence that requires the diamonds in those cases to have a sub lithospheric or super deep origin. So these diamonds actually come from depths that correspond to the lower mantle transition zone where we see a change in seismic wave velocity, which is thought to be caused by changes in the mineral phases that are present down there. Up till now, our only source of knowledge about this region has come from seismic data and from lab experiments that try to mimic the temperatures and pressures of the mantle, such as those described by David Colstead in his podcast episode. What new information have we learned about this zone from the clipper diamonds? So first, just to be perfectly clear, we do have other kinds of sub lithospheric diamonds. So we've known since the late 1980s that some diamonds appear to come from the sub lithospheric mantle. And there's a body of literature already in existence before 2016 describing diamonds from Western Africa. And the big one is the Chiuina region in Brazil that contain a large proportion actually of diamonds that look like they come from the asthenosphere or transition zone or uppermost lower mantle. And clipper diamonds add another layer to that story because they tell us that there's actually another variety of diamonds, another suite within that super deep package. And what we're learning from all of those diamonds is the sense of what happens to subducted slabs. Sub lithospheric diamonds seem to have some kind of a genetic connection with subducted materials. They appear to form from volatiles released from subducting slabs. So they could be reducing carbon, they could be reducing water, they could be releasing boron, in the case of type 2B diamonds, which are often blue. So that's sort of the general picture we see from sub lithospheric diamonds. And specifically for clipper diamonds, we're seeing another angle of this subduction story because clipper diamonds, even though they usually don't contain inclusions, when they do contain inclusions, about 70% of the time those inclusions are this unusual metallic phase that we think was actually trapped as a molten metal. It's a mixture of iron, nickel, sulfur, and carbon. And it's likely that that actually served as the diamond growth medium and has now been trapped as little blebs inside some of these clipper diamonds. And that metallic melt is something totally bizarre that we really didn't expect to find. Something we learned more recently in 2021 is that that metallic melt has a direct connection to serpentinization in the sea floor. So it has a heavy iron isotopic signature that we can't explain in any other way than to say that that iron-rich precursor was a product of serpentinization. So ocean water interacting with the oceanic lithosphere before it's subducted down to the mantle transition zone or the top of the lower mantle. So the diamonds, the clipper diamonds are telling us that there's this unusual iron-rich metallic melt produced during serpentinization or that evolves during subduction. And that's an additional part of the subduction story that you can't really get easily from seismic observations or even from experiments to say that this is a phase that evolves. So diamonds give you another window into deep subduction processes to tell you what's happening to the slabs down there. And a big question on everyone's mind is how much material is actually recycled and carried from Earth's surface down deep in the mantle. The diamonds seem to be a product of that process, the conveyor belt that's bringing things down into the mantle. And diamonds and their inclusions are capturing the release of these materials. And they're capturing the sort of hidden secondary arc, if you will, as the slab enters the lower mantle. There's sort of another wave of geological activity that is unseen to us. And it's that story that's captured really well by sublothospheric diamonds. That's fascinating. Is it possible to explain how we can connect that metallic iron nickel material to serpentinization? Yeah, sort of a convoluted pathway. And to be fair, we don't understand it fully. But we know serpentinization loosely looks like this. You've got water circulating through the crust and mantle portion of the oceanic lithosphere. It reacts with olivine and it makes as products. It makes serpentine, it makes brucite, it makes magnetite. And you've got a fluid that's evolving during that process. And because you're oxidizing iron from iron two plus to iron three plus to make magnetite, you're ending up with kind of a reducing fluid in some cases that has hydrogen in it. And that reducing fluid can actually start to precipitate iron and nickel alloy phases. And we know that both the magnetite, that's an iron-rich phase, and the iron-nickel alloy, which is another iron-rich phase, both those iron-rich phases produced during serpentinization are isotopically heavy. We think that one or both of those iron-rich phases are contributing directly to the evolution of this metallic melt that we now see trapped in clipper diamonds. And we don't, again, fully understand how that story of serpentinization in the sea floor ends up making this diamond-forming liquid. There's some kind of evolution process that happens during subduction. You've got a journey of a couple hundred kilometers and a few million years in between that time of serpentinization and diamond growth. And right now that's what we're trying to figure out is how this metallic phase evolves and how widespread it might actually be and what it might actually mean beyond just diamond formation. So the isotopic signature is also other evidence that this material came from the sea floor. But it's interesting that you said that it was actually an isotopically heavier material because I thought that some of the materials that accumulate on the sea floor because they're generated either by weathering processes or by biogenic processes tend to produce isotopically lighter signatures. Yeah, that's exactly correct. So during the deposition of carbon in the sea floor, we expect that when we see lighter signatures, we attribute that to biogenic processes. So we think that that carbon has come about because of the activity of living things. And we do think that some of that carbon, that isotopically light carbon, may actually be subducted and go towards making diamonds with isotopically light signatures, light in terms of their carbon signature. And is that what we actually measure in the diamonds? Yeah, yeah, we can measure the carbon isotopic signature of diamonds fairly easily. And that's one of the nice pieces of information that seems to connect diamonds with subduction. So we see a story of subducting slabs contributing to diamond formation. And that story is echoed by the mineral inclusions, by the carbon isotopes of the diamonds themselves, as well as by things like the iron isotope chemistry and the fluid chemistry in general that we see. Another observation is the boron isotopes in type 2b diamonds. So these are blue diamonds like the Hope diamond. That boron isotopic signature is a little bit heavy. It's heavier than what you would expect to see for simply mantle derived boron. It suggests very strongly that the boron we see in those diamonds must have come from Earth's surface and was deposited into the oceanic lithosphere by serpentinization or alteration. One of the reasons that clipper diamonds were not identified as belonging to a different class is that we actually find them in the same locations as all the other diamonds. Is that just a bias because that's where we're looking for diamonds or are these in fact thought to be the only places at which clipper diamonds make it to the surface? That's a really good question. So the fact that we recover clipper diamonds now at places where we mine lithospheric diamonds is maybe a byproduct of the fact that when we're looking for new diamond deposits, we're essentially looking for lithospheric diamonds. They seem to make up the bulk of mined gem diamonds and the tools that we have for finding new deposits like indicator minerals, people looking for minerals that are mantle derived that are accompanying diamonds but are more abundant than the diamonds themselves. So those indicator minerals are exclusively lithospheric. And when we're looking for a deposit and trying to decide whether or not it's economic to mine, we take a bulk sample of that, can relate to see how many diamonds are in it. Usually we're looking exclusively again at the lithospheric diamonds because those are a more abundant population. When we do find clipper diamonds in a deposit, they tend to be much more infrequent and it's really difficult to assess how many are there because they're present in such low abundances. But that leaves open this interesting possibility that maybe there are some kimberlites that appear to be barren but actually contain small amounts of clipper diamonds. And we just don't have a way of detecting the presence of diamond in those deposits. And even if you dug up a big bulk sample, you might not find that many diamonds at all. It's not until you mine the entire deposit that you realize, oh actually we've got some of these larger, higher quality clipper diamonds that do make this deposit economic. But it would be a very risky venture to try that. So nobody does. Do we know anything about how old these clipper diamonds are and how long they took to reach the surface? The age of clipper diamonds is still very much an outstanding question. The fact that we find some clipper diamonds at older deposits like the Premier Kimberlite tells us that there must be some clipper diamonds that are at least 1.15 billion years old. So whatever it is, the process that makes clipper diamonds has been operating for over a billion years. We also find clipper diamonds at younger deposits like at Let's Sing in Lesotho. This is a 90 million year old deposit, but we don't know the age of those diamonds. We just know that they're not any younger than 90 million years. We can take some clues to the ages by looking at other categories of sublithospheric diamonds such as those from Jouina that are not necessarily great gemstones, but also are part of this super deep diamond forming story. So we have some ages of those diamonds. A recent study by Suzette Timmerman found that there are sublithospheric diamonds that appear to have formed a long time ago in the Precambrian, maybe 600 million years, something like that, and those diamonds sat around in the mantle before eventually being brought to surface. So we know that at least in some cases there is a period of mantle residence for sublithospheric diamonds. They're sitting somewhere in the mantle for a few hundred million years before they're brought to surface. And in the case of that study, what was really nice is that they found diamonds of similar ages that were actually on Brazil in the South American continent and some in the African continent and those are obviously divided by the Atlantic Ocean. So they think that those packages of diamonds were actually adhered to the base of the continental lithosphere and that's where they've been residing. They've sort of been sitting at the bottom, stuck, plastered onto the bottom of the continent four millions of years. And as those continents have drifted apart, the diamonds have gone with them because they're stuck to the bottom like chewing gum on your shoe. And you know, it's not until a kimberlite comes along and samples some of that mantle package and brings the diamonds to surface that you know that they're present at all. And actually, you know, you need this interesting turn of events, coincidences that line up to actually bring the diamonds to surface. That's fascinating. So it's really a two-step process, possibly, that they form at these great depths down in the mantle transition zone region. And then some mechanism, maybe it's just convection, brings them up to the base of the lithosphere where they get stuck and then wait for whatever process is associated with kimberlites. I don't know if it's considered to be mantle plumes or what, but until it just happens to be above those. And then so it really conjures up a vision that maybe the entire lithosphere is peppered with these bits of valuable chewing gum, if you like. Yeah, it's possible. I should make a comment on the ascent mechanism for sublithospheric diamonds. So right now, our best idea of the ascent mechanism might involve just the upwelling of subducted materials. So the subducted slab as a whole is denser than the surrounding mantle. That's why it's sinking. But within that slab, there are perhaps more depleted, more magnesium rich regions, less iron, so they're less dense. And some of these packages of rock might actually be less dense than the surrounding mantle and form buoyant diapears. They might not be all that big, but they just have to be big enough to capture packages of diamond bearing rock and bring them up to surface and plaster them onto the bottom of continents. So we think that it makes the most sense if that ascent mechanism is tied to the process of diamond formation and also involves subducted ingredients, because that's your best chance to have kind of diamond formation and then the appearance of buoyant materials that can bring diamonds up and then kind of park them at the base of the lithosphere. As I mentioned in the introduction, these diamonds are extremely valuable. And I'd be surprised if your research budget enables you to outbid celebrities and royalty. So how do you get access to them and are you just confined to non-destructive methods such as spectroscopy? So when we know for sure they're part of the Clipper category, we're talking about diamonds that are quite large, a few carats or more, you know, upwards of 10, 20, 100 carats. And sometimes these diamonds are completely flawless and decolor. So they're the best of the best gem diamonds. And absolutely they command the highest prices and you can't buy that on a research budget. You can't borrow it. What sort of price are we talking about in today's market? Many of the diamonds that I've studied are easily in the millions of dollars range. That brings me to my position at the Gemological Institute of America. This is the right place to be if you need to look at rare and expensive diamonds. This is a place where they come through with relative frequency. So the infrastructure of laboratories grading diamonds every year is seeing millions of diamonds come through this grading infrastructure. And this is the way I was able to build up observations. So we were screening all of these diamonds trying to find the ones that are nice gem quality type 2a diamonds, but that also have inclusions in them. So that's very few diamonds. It's something like one in 10,000 of all the diamonds that come through for grading might be of interest to look at for a study like this. So I was able to borrow these diamonds for a period of hours or maybe a day or two, depending on the individual diamond. But these are polished faceted diamonds coming through the lab for grading. And I have the ability to observe them non-destructively. So this means looking at them under the microscope, taking lots of pictures. And the biggest tool I was relying on was ramen spectroscopy. So this is an instrument where you put the diamond under a microscope essentially and line up the inclusion under the crosshairs and then you can shine a laser at it and collect the light that bounces off of the inclusion and get an idea of what mineral it is or in some cases what volatiles like methane or hydrogen are actually trapped with that inclusion system as well. Additionally, there were a few samples that I had that I was able to do more invasive studies on and actually polish them open and expose some of the inclusions to do things like electron micro probe work where you need an exposed inclusion. So those samples were off cuts. So things that had been trimmed off of larger diamonds during the manufacturing process. These might be trimmed off because they're too heavily fractured or because they have big black inclusions in them. And the manufacturer is looking to exclude that "ugly material" from the end product, the polished diamond. And that off cut is something that's extremely valuable for scientific study. Interesting side note on the cullinan diamond. So I mentioned this was a 3,106 carat rough diamond. Its largest faceted piece was just over 500 carats and that means that they cut away a lot of material. If you look at original pictures of the cullinan diamond you can actually see a couple of these black specks in them. So even the type example did have inclusions and the notes from the people actually cleaving this diamond, their very first cleave through the cullinan diamond was to intersect what they called a big black spot, an inclusion inside the diamond. So even the very early people trying to polish diamonds are avoiding these inclusions. That's part of the reason why they seem to have so few inclusions is because people are actively excluding them from the polished diamonds and when we see these fantastic pieces come up for auction, they're completely inclusion free. Well, that's great. So the diamond trading companies like De Beers are actually quite collaborative with you in the sense that the bits that are the greatest scientific interest actually are blemishes from that point of view. Yeah, that's right. In some cases, if they're actually trimming a diamond down using a laser, so they're laser cutting off some of these more included regions of diamond, yeah, they're open to the idea that there might be something valuable scientifically that comes from that piece of diamond. What are you working on at the moment? Right now, I think the most interesting thing I've got is still the metallic inclusions in clipper diamonds. We have an idea of its composition. We know this very exciting tidbit of information that they're actually connected to serpentinization in the sea floor, but like I said, we don't understand the full story of how this metallic liquid evolves, how it behaves, what its life cycle is. What happens after diamond formation? Does this metallic liquid just sink down into the core or is it reacted away? And what does it mean for the whole recycling process and the mantle geodynamics of carbon beyond just diamond formation? I think this is something that we didn't expect, and that's always the thing that you should pay most attention to. You've got to pay attention to those things that don't make sense. So right now, the metallic inclusions just don't make a lot of sense. Evan Smith, thank you very much. Thank you. To see pictures and illustrations that support this podcast, go to geologybites.com, where you'll also find transcripts and a subject matter index of all the episodes. There you can also give me feedback, which I welcome, as well as sign up to get my emails about new episodes.