Nadja Drabon on a New Lens into the Hadean Eon
Transcript
Oliver Strimpel
This is Geology Bites with Oliver Strimpel. We don't know very much about the first 600 million years of Earth history, which we call the Hadean eon, nor indeed do we know that much about the beginning of the Archaean Eon that followed. But that is beginning to change as researchers deploy the modern tools of geochronology and geochemistry. There are two main questions being investigated. The first concerns the crust. When and by what mechanisms did it become something like the crust we have today, with its thick continents and thin subducting sea floor? The second question, which is closely related to the first, concerns life. When did the Earth first become habitable? In an earlier episode, Peter Cawood talked about the first question, and, in particular, about when modern-style plate tectonics may have started; and in another episode, John Valley discussed the early Earth, suggesting that the Earth might have cooled more rapidly than was previously thought, and thus become habitable very early in its history. Nadja Drabon’s research is also directed to the study of the early Earth. By analysing a newly discovered trove of extremely ancient zircon crystals, she has given us a new lens into the Hadean and early Archaean eons. This has enabled us to witness some of the processes that created and destroyed the Hadean and early Archaean crust. We can then use this knowledge to make inferences about when the Earth first became habitable. Nadja Drabon is Assistant Professor of Earth and Planetary Sciences at Harvard University. Nadja Drabon, welcome to Geology Bites.
Nadja Drabon
Thank you so much for having me.
Oliver Strimpel
Before we talk about the newly discovered zircons, can you remind us as to why zircons are so useful as probes of the early Earth?
Nadja Drabon
Yeah, so zircons are an incredibly powerful tool for four different reasons. So firstly, zircon is really abundant and in felsic crust. Secondly, it's very durable. So, it's very resistant to alteration and weathering. Then thirdly, of course, we can date it, and I believe your listeners have heard plenty about zircon uranium-lead geochronology already. And then lastly, these tiny zircons actually capture a lot of geochemical information, so that includes isotopic records as well as trace and refinements that tell us about crustal conditions experienced during zircon crystallization.
Oliver Strimpel
OK, good, we'll get into that in a moment. But just to clarify, it's the zircon crystals themselves that are indestructible and we're dealing with detrital zircons here, so these are zircons that got eroded out of the igneous rocks in which they originally formed, and then later on got incorporated into sedimentary rocks. In an earlier podcast, Ulf Linnemann talked about using detrital zircons to trace the assembly of Europe in the Paleozoic. Here we're dealing with much older detrital zircons to probe processes that happened in the Hadean and early Archaean. How old are the oldest zircon crystals, and how old are the oldest intact rocks?
Nadja Drabon
This is really a problem we face when we study early Earth history, right? Because, as we go further back in time, we lose the rock record. So, the Archaean Eon represents 1.5 billion years of Earth history - so, a really significant chunk of Earth’s history. But only 5% of rocks are of that age, so the oldest rock is 4.03 billion years old. This rock is from the Acasta Gneiss complex in Canada. This leaves the first over 500 million years of Earth's history without a rock record. We find a few locations around the globe where there’re Hadean zircons. But the oldest zircon that we have found is about 4.4 billion years old.
Oliver Strimpel
OK, so that's not much, but still older than the actual oldest rock itself.
Nadja Drabon
Yeah, that's about 400 million years older.
Oliver Strimpel
So, can you tell us about the ancient zircons that were known before this new source that I just mentioned was discovered?
Nadja Drabon
Yeah, so the most famous location with Hadean zircon is the Jack Hills in Western Australia. So, the first Hadean zircons there were discovered in the ‘80s, and since then, people have dated about 200,000 zircons. So, that's really a huge amount of work. And, of these 200,000 zircons, about 10,000 were Hadean in age. And what is really so great about this location is that it allowed us to first start addressing some of these big questions that we have about the early Earth. So, those questions include: When did the first crust form? The early Earth was initially covered by a magma ocean, so when did we start forming crust? And when did we start forming oceans on this crust? What was the composition of that crust, and does it tell us anything about geodynamics? So, one big question that I'm trying to address in my research is the evolution of tectonic processes. The Earth is the only planet in the solar system that has plate tectonics. We don't really know when plate tectonics started and why it started. And then the last big question, of course, is when did life evolve, and that question has really two components: When did Earth first become habitable, and then when do we actually start finding physical evidence for life? The problem with the Jack Hills is that it's only a single location, so that, of course, it's not very representative of the whole Earth.
Oliver Strimpel
So, tell us about the newly discovered zircons you found in the Barberton Greenstone Belt in South Africa.
Nadja Drabon
We found a new location with Hadean zircon in South Africa in the Barberton Greenstone Belt. So, the Barberton Greenstone belt is between about 3.6 to 3.2 billion years old. And, within this greenstone belt, we have dated about 100 different sandstones throughout the unit, and we found one single sandstone that contained Hadean zircons. And, the sandstone is called the Green Sandstone Bed, because it's very green. So, it contains a lot of chromium-rich minerals, which gives it that green color. And, what is so fascinating about this unit is that the zircons we find in them have ages that do not correspond to any known felsic igneous rock within the Barberton Group. So, it really yields information about a terrain that has since been lost. And this terrain has an age between 4.15 to about 3.31 billion years, so that's 800 million years of history that we can tap by studying the ages and the geochemistry of these zircons. Being able to look at the complementary record to the Jack Hills allows us to start testing some of the hypotheses that were put forward based on the results from the Jack Hills. And, what is nice about the Green Sandstone Bed is that the rock experienced lower temperatures compared to the Jack Hills, so it's somewhat better preserved. Unfortunately, what is not as good as the Jack Hills is that, in the Jack Hills, between 3 and 12% of the zircons are Hadean in age, versus, in the Green Sandstone Bed, it's only half a percent of the zircons.
Oliver Strimpel
OK, so let's talk about the way you analyzed these zircons from South Africa. In the previous episode with Geology Bites, I talked to John Cottle about a technique for rapid and self-consistent dating and measuring of trace-element abundances in crystals such as zircons. Did you use his methods?
Nadja Drabon
We did not use his method. So, John Cottle used laser-ablation split-stream analyses, which allows you to measure the ages and the hafnium isotope composition, perhaps also trace elements of exactly the same amount of material. And, we did not do that in part because we needed much more precise results. As I mentioned, only a half a percent of the zircon are Hadean in age. So, first we analyze as many zircons as we can at a relatively rapid speed, to just be able to find the oldest grains. And then we take the mounts with the zircons on them, and we put them into a SHRIMP.
Oliver Strimpel
Let me interject that this SHRIMP is not a crustacean, but an acronym for an instrument called a Sensitive, High-Resolution Ion Microprobe.
Nadja Drabon
Yeah. The SHRIMP allows much more precise dating. So, we dated the old grains and then we also used SHRIMP to do trace-element analyses. And this is really important to use SHRIMP for that, because some elements that we are studying have incredibly low abundances, but also because some of the elements that are particularly important have interferences. So, you need really high resolution to really analyze the correct peaks.
Oliver Strimpel
So, you performed then several geochemical analyses on the zircons. Can you tell us about those?
Nadja Drabon
I used three different analytical approaches. The first one is: we measured hafnium isotopes to get at crust-mantle differentiation. Then we measured trace and rare earth elements, which tells us about the composition at the magma at zircon crystallization. And then lastly, we used oxygen isotopes, which can tell us about the presence of a hydrosphere in the Hadean.
Oliver Strimpel
Let's go through these one at a time. Let's start with the hafnium analysis. How does that work and what can the results of that analysis tell us?
Nadja Drabon
When we study hafnium isotopes in zircon, what we really focus on is the decay of 176 lutetium to 176 hafnium. So, lutetium and hafnium have very different behavior. Lutetium is relatively compatible, so it's going to preferentially stay in the mantle, versus hafnium. It’s much more incompatible, so it's going to preferentially go into the crust. What we ultimately end up measuring is the 176 hafnium (so, the daughter product of that decay) over the stable 177 hafnium, which is relatively constant through Earth history. So, because of this different behavior, when we look at these different reservoirs, they show different changes in the daughter products, the (176 Hf / 177 Hf) through time. Because the mantle contains more lutetium, the increase of the (176 Hf / 177 Hf) is going to be much more than, for example, continental crust, because continental crust has less lutetium. That means, it produces less 176 Hf and that means that the increase in the ratio of (176 Hf / 177 Hf) is going to be smaller. Zircon is a really fantastic tool to track what the hafnium isotopic composition was of the magma at the time the zircon crystallized, because when zircon crystallizes, it takes up to several weight percent of hafnium, but almost no lutetium. So, it essentially freezes the signal, the isotopic signal, of the hafnium isotopes at the time it crystallizes. So, depending on the ratio, the (176 Hf / 177 Hf), we can now say: was that crust extracted from the mantle like immediately prior to zircon crystallization essentially, or did the zircon actually form by remelting of old crusts? So, as I mentioned again, older continental crust has a different change in the ratio of (176 Hf / 177 Hf). So, it really deviates a lot from the evolution of the mantle, so if we take this old crust and re-melt it, the zircon can again capture the signal, and we can use that to infer the zircon form from this other cross or from more juvenile, so like newly-formed crust, from the mantle.
Oliver Strimpel
Wow, that's really impressive. We can actually use the hafnium isotope ratios in zircons to tell whether Hadean and early Archaean crustal rocks, which themselves have long since disappeared, were made directly by melting of the mantle or by melting of pre-existing older crust. So, can you tell us about the results of your hafnium isotope analysis?
Nadja Drabon
Yeah, so in the Green Sandstone Bed, when we look at the oldest zircons, we can see that there were extracted from the mantle about 4.2 to 4.4 billion years ago. And, for the next 4 to 600 million years, they do not record any evidence of new mantle additions. Instead, it seems like that, that initial crust that was extracted from the mantle, stayed relatively isolated. There was some internal remelting, but there were no new additions of mantle material. And then that signal abruptly changes about 3.8 billion years ago. Anything younger than 3.8, we see a lot of new additions from the mantle, so a lot of juvenile felsic crust production, together with some remelting of older crust. But that isotopic signal of that really old Hadean crust is essentially lost after about 3.8 billion years ago.
Oliver Strimpel
Oh, that's fascinating. So, it points to a change that happened about 3.8 billion years. You're already getting into the early Archaean. Let's talk about the other analyses first, before we circle back and then talk about what you managed to conclude from putting all this together. So, the second analysis that you mentioned earlier was about looking at the trace and rare earth element abundances that you measured in the zircons.
Nadja Drabon
Yeah, that's right. So, trace and rare earth elements have been used for decades at this point, but it's only really been the last 10 years that researchers have recognized that they really provide a lot of information in terms of the setting in which the zircons crystallize. So, for this Craig Grimes analyzed zircons from felsic igneous rocks in different tectono-magmatic settings. So that includes: depleted mantle (so, mid-oceanic ridges); undepleted mantle (so, that's something that would be represented by ocean islands like Hawaii or Iceland); and then lastly, he looked at continental arcs (so, subduction zones, right? That's a very different mechanism. You're pushing crust below a continent in that setting and you have got this type of hydrous-melting signature). So, Craig Grimes ultimately found that there's some very specific trace elements that can really distinguish these different tectono-magmatic processes. And these include uranium, niobium, scandium, cerium, and terbium. So, what do our zircons show, right? That's the big question here. And, for this, long-lived crust that was extracted from the mantle and the Hadean that I mentioned earlier, the question is really: OK, this crust was really long-lived, but how did it form initially? Did it form in a type of arc setting, which would be evidence for plate tectonics? Or did it form in a more mantle-dominated setting? That is something we cannot distinguish by using hafnium isotopes. But now, by using the trace elements, we actually find that these early zircons have a really strong mantle signature and very little evidence for hydrous melting, that we would expect, for example, in a subduction setting.
Oliver Strimpel
Because you'd expect the hydrous in a volcanic arc in a subduction zone, because that would normally be happening under the ocean.
Nadja Drabon
Yeah, because you've got the oceanic crust that is interacting with the ocean and then that crust that is hydrated is being pushed underneath another crust, and that water is being used in the melting process, and it imparts a very characteristic geochemical signature. So, by looking at the Green Sandstone Bed, I mentioned earlier, the hafnium isotopes, that we have evidence that there was initial crust extraction and that that crust was really long-lived. By using the trace elements, we're able to say that that initial crust was extracted from the mantle, and so it doesn't have that strong hydrous-melting signature. This is really remarkable because this oceanic-type crust was stable for several hundreds of millions of years, which is something that is really not typical when you look at the modern environment today. So, when you think, for example, about oceanic crust, oceanic crust typically doesn't get older than 200 million years, versus Hadean crust that formed the source to the Green Sandstone Bed was stable for hundreds of millions of years. So, this is already starting to give us some indication that some of the early Earth geodynamic processes may have been rather different to today.
Oliver Strimpel
So, you also analyze oxygen isotopes. What does that tell us?
Nadja Drabon
Oxygen isotopes and zircon tell us something different. Oxygen isotopes and zircons are really sensitive tracer of fluid solid interactions, so it tells us if the magma from which the zircon crystallized has assimilated a crust that was altered at low temperature, or potentially something that incorporated sediments. So that implies there was some kind of hydrosphere already at the surface of the Earth. It can also tell us if that interaction with the potential hydrosphere happened at high temperature.
Oliver Strimpel
Is it possible to explain in simple terms why interaction with the hydrosphere would affect the oxygen isotope ratios?
Nadja Drabon
So, this is something that is based on an isotope equilibrium process that really depends on the mass of the oxygen and you have different exchange between the material that you're altering and the fluid. And that's going to vary depending on the temperature.
Oliver Strimpel
But this is an entirely inorganic process…
Nadja Drabon
Correct. So, we already know from the Jack Hills that most of the zircons have a mantle-like signature, and a few of them do show evidence for alterations with low temperature fluids, so that there's some evidence for hydrosphere. But here the question, of course, always is: OK, it's only one location. So, was their global ocean, or was it maybe just something more localized? In South Africa, what we find is also a lot of zircons that that elevated oxygen isotope signature, so this is providing us with a better picture of a potentially more widespread hydrosphere already in the Hadean.
Oliver Strimpel
OK, so can you put the results of these three sets of analysis together? The hafnium isotopes, the trace and rare earth elements and the oxygen isotopes, and tell us what you were able to learn about the crust of the early Earth from the South African zircons?
Nadja Drabon
Yeah, sure. So, in South Africa we learned that there's really two phases in the crust evolution. Prior to 3.8 billion years ago, we have cross formation by extraction of material from the mantle, and this material from the mantle was extracted and was kept relatively isolated, so we don't see a lot of juvenile additions. It's relatively stable. But we do find evidence that there was some present water. And then after 3.8, we see this really dramatic change where we now start seeing a lot of new juvenile additions. The crust becomes really unstable. You're still mixing some older material, but we do start seeing this really more hydrous-melting signatures.
Oliver Strimpel
So, is this a period of history that we did not have access to before then, when we were just looking at the Jack Hills?
Nadja Drabon
It is really providing a complementary record. So, tying the heavier isotopes and the trace elements, it’s starting to add a lot of different puzzle pieces to this picture of the Hadean that we are trying to put together. What was the earliest Earth like? Was it very similar to today, or was it different? And there's this really wonderful study by Annie Bauer. And she looked at hafnium isotopes from a number of different locations around the globe, and she found that this evolution in the hafnium isotopes showing really long-lived crusts during the early Earth until about 3.8 to 3.6 billion years ago. And then suddenly we see this destabilization and more juvenile crust additions. But again, hafnium isotopes can't really tell us if that long-lived crust was formed by some kind of plate-tectonic process or something different. So now tying the trace elements, we can say that at least for the Green Sandstone, that we see this transition of something that is more mantle-dominated until about 3.8, and then the signatures start becoming a little more similar.
Oliver Strimpel
Does that suggest that plate tectonics might have started at that time, or are there other explanations for what we've seen in the record?
Nadja Drabon
That's a really good question, because that's something we have to be really careful about. So, starting at 3.8 and the trace elements, we start seeing a hydrous-melting signature. And the thing is that there's different processes that could explain that type of signature. One way is the standard way that we see today, right? Plate tectonics. That's the tectonic model that we are used to, where (we talked about this earlier already) where we are putting hydrated oceanic crust into the mantle. An alternative way how this might have happened during the early Earth, is that when you form crust at the surface and right at the surface it's interacting with the hydrosphere, potentially you're generating subsegments that are being deposited. If that crust is buried, so imagine just a lot of really heavy mantle material coming up to the surface, so the crust thickens, and it starts sinking down. We call that sagduction or partial convective overturn. You might be able to generate a similar signature in the trace elements, and the hafnium isotopes.
Oliver Strimpel
“Sagduction”. as opposed to subduction, a “sag” So it's just like a sag; a block just falls off.
Nadja Drabon
Yeah, it just falls off. So sagduction was potentially possible in the Archaean, because the heat flux on the mantle was higher, so the crust was generally hotter, and because of that, the crust started behaving in different ways, potentially allowing for that sagging — versus today, the heat flux on the mantle is much lower, which makes the crust much more rigid and does not allow for that sagging behavior. Sagduction has been proposed for ancient rocks, but we don't see it today. What exactly does the signature look like? I hope that one day some modelers will go and model that and can tell us if there's ways to distinguish the two different signatures.
Oliver Strimpel
I said in my introduction that you can use the knowledge you're unearthing about the early crust to make inferences as to when the Earth first became habitable. What does all this have to do with habitability?
Nadja Drabon
I think that there are three important takeaways in regards to habitability. So, the first one relates to the presence of water. We all need water, so the question is when did we start forming a hydrosphere? When did we start forming an ocean? And based on previous results and the Jack Hills and our new results now in South Africa, we can say that at least by about 4.3, 4.2 billion years ago, there was probably a hydrosphere present. The second one is about the creation of topography. So, today the most effective mechanism of creating topography is through plate tectonics and convergent settings. This becomes really important when you think about what the conditions at the surface were. Today, with creating topography, we initiate silicate weathering and that draws CO2 out of the atmosphere. It fixes it into carbonate. That makes the climate much more moderate. So, if we don't have plate tectonics, we have volcanos still producing CO2 again into the atmosphere, making it really hot. But you don't have a good way to draw that CO2 down. So, one question is, OK, when did we start creating all of that topography? And it looks like in terms of mechanisms, it seems like starting at 3.8, there might have been much more efficient mechanisms of creating that topography, so, just really stabilizing the climate. If the climate was really changing and fluctuating up and down, you can still have life, but the conditions are going to be more favorable for more substantial biomass once you start stabilizing the climate. And then, the last evidence is more direct. Beth Ann Bell in the Jack Hills studied the mineral inclusions of zircons and she found at least one zircon that has two graphite inclusions. So, graphite as carbon and carbon could potentially be evidence for life. And she measured the carbon isotope composition. So, carbon has heavier and lighter isotopes, so carbon 13 and carbon 12 and life always prefers the light isotope. If the carbon isotopes are relatively negative, this could potentially be evidence for life, and these graphite inclusions that she measured, she found carbon atom value of minus 24, which is very much consistent with life already being present during the early Earth. And I'm saying consistent, because it's just one inclusion, so we need to be really careful. There's some abiotic mechanism that could produce carbon isotopes like that. But, this is something that I'm hoping to do in future research, is search for more graphite inclusions and measure the carbon isotopes to see if we continuously get these light carbon isotopes that could be reflective of early life. Or, if we find more graphite inclusions, do we see a range in carbon isotopes, which would speak to more abiogenic mechanisms?
Oliver Strimpel
You mentioned how useful it would be to look for other locations. Are you actually hunting for new sources of ancient zircons?
Nadja Drabon
I think there absolutely are more locations with Hadean zircons, and in a lot of my research projects we just analyzed Archaean Sandstones full of zircons. We are always hunting for more locations with Hadean zircons.
Oliver Strimpel
Do you have any promising leads at this point?
Nadja Drabon
We haven't found any new locations, but we did some more analyses on one location in India and one in Wyoming, and here we know that people have found old zircons. So, we went in, and we are just trying to analyze thousands of zircons just to see how old can we get? How old is the oldest zircon going to be? And I'm happy to report we found 14 new Hadean zircons in India and at least one in Wyoming.
Oliver Strimpel
Nadja Drabon, thank you very much.
Nadja Drabon
Well, thank you.
Oliver Strimpel
For more about Geology Bites, as well as pictures and illustrations that support this podcast, go to geologybites.com.