Bob Hazen on the Evolution of Minerals
Note, transcripts are not fully edited for grammar or spelling.
Oliver Strimpel
This is Geology Bites with Oliver Strimpel. We're familiar with the idea of new rock types emerging during the history of the Earth. Most obviously, the silica rich felsic rocks such as granite that characterize continental crust accumulated during the course of Earth history. Granite only forms in certain specific tectonic settings, such as above subduction zones, and in mountain belts. This is where partial melting and fractionation of more primitive rocks, richer in iron and magnesium, occurs to generate the granitic rocks, which are rich in minerals such as quartz and feldspar. But what about the minerals themselves? Have they been around since the earth formed, or did they too only appear on the scene later as a result of some geological process? The question of how and when the minerals evolved is a relatively new subject and was and continues to be pioneered by our guest today. Bob Hazen is Senior Staff Scientist at the Earth and Planets Laboratory of the Carnegie Institution for Science and Professor of Earth Sciences at George Mason University. At a Christmas party in 2006, a well-known biophysicist asked him the question, “Were there clay minerals in the Archean?” This triggered a fascination for the subject and resulted in an intense research effort on mineral evolution that has been his focus ever since. Bob Hazen, welcome to Geology Bites.
Bob Hazen
Thanks so much, Oliver. It’s wonderful to be here.
Oliver Strimpel
So did you know the answer to the question?
Bob Hazen
Well, you know, I didn't. And that was a shock. In fact, I'd never heard a question like that before. You know, mineralogists had long thought about minerals absent the context of their geological history. They were physical and chemical objects, and so when Harold Morowitz asked me this question, he was thinking from a very different perspective. He was thinking about the origin of life, and in many models of the origins of life, minerals play a role. Now, if you say clay minerals played a role in origins of life, then clay minerals had better be around early in Earth's history. That's what he was asking. I never heard a question like that. It was completely original and it got me thinking. My gosh, where clays present. And what about all the other 5700 minerals that have been approved and identified and codified? Which of those were present? Which ones have formed more recently? That's how mineral evolution the idea began.
Oliver Strimpel
What exactly do you mean by mineral evolution? Are we talking about an increase in the number of different minerals present?
Bob Hazen
So mineral evolution actually means an increase in the diversity of minerals. But also in the changes in their distribution and other changes that are maybe more subtle. The sizes and shapes of minerals, they're trace elements and isotopes, all different kinds of properties of minerals that through different stages of Earth history have been different from what they are today, and presumably those differences and changes will continue into the future.
Oliver Strimpel
But, with over 5700 minerals, I can well understand why mineralogists would want to impose some fairly simple order and use an easily defined classification scheme.
Bob Hazen
Yes, of course, classification is absolutely essential to every science, and the International Mineralogical Association does a heroic job in separating and distinguishing so that we all can talk about these 5700 different kinds of minerals, and in fact that number grows by 100 every year. So it's a, it's a huge job, so each mineral is defined by two characteristics. It's idealized chemical composition, and it’s idealized crystal structure, physics and chemistry. It's a very efficient system. Each mineral is defined by the minimum amount of information, literally in bits of information that you need to distinguish quartz from diamond, from pyrite, and from any of the other 5700 plus minerals.
Oliver Strimpel
So you're looking at more than just those simple criteria, those structural and chemical criteria. Can you give me an idea of what some of these other characteristics are?
Bob Hazen
So if you think about minerals, each mineral specimen is a time capsule. It's incredibly rich in information. It has minor and trace elements. It has a whole variety of isotopes. It is fluid and solid inclusions. It has a size. It has a shape. It has optical properties. It has magnetic properties. It has all sorts of information-rich attributes, which distinguish each specimen from every other specimen. So in the International Mineralogical Association, they've tried to simplify that so that we have 5700 kinds that we can talk about. But in our system, the evolutionary system of mineralogy, we embrace all of that information-rich complexity, because that's what tells us the story. That tells us the story of each mineral specimen. Where it arose, how it's changed through time, what its neighboring minerals might have been doing.
Oliver Strimpel
OK. Before we talk about some of these additional rich sources of infomation let's return to the basic statistics for a second, which is the number of different classically defined, structurally defined minerals. As you mentioned, this number has grown over time and continues to grow today. So can we divide mineral evolution into various distinct stages over the history of the Earth?
Bob Hazen
So dividing mineral evolution into stages is exactly how we started. We said what are the different phases? What are the different processes? And this really boils down to physics, chemistry, and ultimately, this is a big surprise, biology because the different stages of mineral evolution represent times when new processes came into play. We have three big eras of mineral evolution. The first is the period of when Earth was first forming and there were meteorites of all different kinds that now fall to Earth that represent that earliest period before 4 1/2 billion years ago, about 400 or so different minerals that we can recognize from those objects. And then there was a long period, maybe 2 billion years, from 4 1/2 billion to 2 1/2 billion years, where most of the minerals form through just purely physical and chemical processes. The various kinds of temperature and pressure regimes, compositions of fluids changing. And so we had granitic crust forming. We had volcanoes. We had various kinds of ocean-deposited minerals, all things that didn't require life. Minerals that might occur on other non-living worlds. But then on Earth, this amazing thing, the origin of life occurred and, more importantly, the origin of microbes that learned how to produce oxygen to get energy through this redox process, and thousands of new minerals arose simply because biology changed the environment of our planet.
Oliver Strimpel
And so did the majority of the minerals that we have around us today actually coevolve with biology?
Bob Hazen
You know, Oliver, we've actually done statistical studies on this recently. What we find is that slightly more than half of all the mineral species, those 5700 or so can form through biological processes. But only about 1/3, In fact, 34% of the minerals occur exclusively as the result of biological processes, either directly or in many cases indirectly producing oxygen, which then causes an oxidation of surface minerals that wouldn't have occurred.
Oliver Strimpel
Otherwise, I can imagine perhaps at the simplest level that we're producing various carbonates as a result of, say, bone-forming processes in microorganisms and you know, hence all the limestone deposits and the cliffs of Dover and so on. But you're talking about thousands of minerals here. So can you give me a flavor of the sort of environments that biology is producing to generate these minerals that we don't get in the inorganic world?
Bob Hazen
Indeed, so most of the biologically mediated minerals are the consequence of atmospheric oxygen. You know what rust looks like? Well, you can rust. In air quotes, the minerals of uranium and molybdenum and copper and cobalt and nickel and and many, many other elements. Lots of those minerals occur at or near Earth surface, where there's an oxidizing environment and so a whole host. There's like 300 uranium minerals that have formed through oxidative weathering. Now there are also quite a number of fascinating minerals that occur in other biological ways. There are minerals that form because of interaction of guano and in urine with the environment. There's a new mineral that was discovered called Tin Nun Q light. It's a carbon-bearing mineral that forms when falcons poop onto burning coal mines. Believe it or not! So it bakes the falcon poop, and that creates new sets of crystalline forms. And there are lots of minerals, of course, that are being mediated by human activities. Mining activities, mine waste dumps and drainage and smelters and just the weathering of archaeological artifacts create new chemical forms that had never before occurred on Earth’s surface. So these are all biologically mediated minerals.
Oliver Strimpel
I want to just jump way back in time for a moment and ask you what were the very first minerals, in fact the minerals that were able to form even before, say, the solar nebula got itself together.
Bob Hazen
Oh, Oliver, I love thinking about this problem and it turns out it had largely been answered by our astrophysicist friends. Because they study tiny, tiny grains of minerals. They're referred to as Stardust, literally because these are grains and minerals that formed long before our solar system. In the atmospheres of very old stars. In some cases those stars were just very energetically shedding their atmospheres. AGB stars, they're called, and they produce things like diamond, of all things, they produced silicon carbide. They produced oxides of silicon and aluminum and magnesium and calcium. Some of those stars exploded. Supernovas produce mineral grains, and those mineral grains come to us as tiny particles in meteorites. You can dissolve them out, and they have huge isotopic anomalies in silicon and oxygen and nitrogen and carbon, and so you can identify them quite easily even though they're very small, and realize that we're looking at particles that in some cases are billions of years older than planet Earth. So that's the beginning. Those are the err minerals. About 20 of them, we think.
Oliver Strimpel
That very much reminds me of the recent podcasts by actually your colleague at Carnegie, Rick Carlson, who talked about pre-solar grains and the isotopic anomalies in those.
Bob Hazen
Imagine looking at a tiny mineral grain much too small to see. Even with the most powerful optical microscope. And yet you're learning about the evolution of stars 6 or 7 billion years ago.
Oliver Strimpel
I want to move on to the next phase now and the minerals that require geological processes, if you like, that take place on the surface of the earth. How, for example, do you actually make a mineral like quartz or a mineral like feldspar?
Bob Hazen
This is called the paragenesis of minerals, and we've just completed a massive study. It was allowed really because of the COVID lockdown. Months and months and months I isolated myself, and studied all of those 5741 mineral species and looked at every different way that they might have formed. Turns out we came up with a list of 5757 varieties—just like Heinz, I guess, 57 varieties—of mineral-forming things like forming from a hot magma, cooling and having the crystals form, or aqueous alteration or heating and squeezing in a process called metamorphism. In some cases, just being formed by evaporation like salt crystals that form when you evaporate salt water. Things like lightning strikes, meteor strikes, impact-formed minerals. Well, it's all over the map. Most of the minerals form in only one or two ways. But it turns out there are some minerals like pyrite, iron sulfide, that form in more than 20 different environments, some of them biological, some a biological, some at high temperature, some at low temperature, some with water involved, and some completely dry. Just amazing. All the different ways you can form some of these minerals.
Oliver Strimpel
OK, let's talk about the various nontraditional information-rich characteristics that you're using in your new mineral organization. Presumably, the minerals formation environment such as those you just mentioned with respect to pyrite is one of them.
Bob Hazen
Oh, Oliver, absolutely. Let's talk about diamond. Everybody loves diamond, pure carbon, and it's in the idealized carbon structure of diamond, according to the International Mineralogical Association. But when we look at diamond, we look at all the differences amongst all these different kinds of diamonds. Diamonds formed in the atmospheres of stars as those atmospheres cooled. They were carbon rich and diamonds precipitated at very, very low pressure but high temperature. So that was billions of years ago and then hundreds of millions of years ago, diamonds were forming deep in Earth's mantle at high temperature and high pressure as carbon-rich fluids were squeezed and heated and the diamonds formed. And some of those diamonds come to the surface and in very rapid eruptions. Kimberlite eruptions, for example. But then still other diamonds form through impact processes. If an asteroid smashes into Earth, one of the things you get is tiny little impact diamonds. Now impact diamonds are quite different from mantle diamonds, which are quite different from stellar diamonds. They have different trace and minor elements. They have different defects. They have different morphologies. All of that information leads us to believe that there are at least half a dozen different kinds of diamond, all conforming to the IMA definition of diamond, but each with distinctive properties because of all that information that each sample gives to us.
Oliver Strimpel
I'm wondering about the clustering and clumping idea. If you like, I can see that with diamond you've got three very distinct origin stories. The stellar origin, the mantle origin, and the meteorite- impact origin. But coming back to your previous example of pyrites, where maybe you're talking about continuoussly varying hydration environments or temperature environments, how do you decide where to place a mineral cluster in that continuum?
Bob Hazen
That's a really important question for us. So what we've done in the case of pyrite. We had assembled a database of more than 10,000 analyses of pyrite and their minor elements. So pyrite is iron sulfide. But iron sulfide almost always incorporates other elements. Cobalt and nickel. You might have some gold in there. You might have silver. You might have zinc. So you basically analyze all these different pyrite samples which look more or less like each other. So you wouldn't, just looking at the hand specimen, you wouldn't be able to distinguish them. But the trace elements tell a story, and when we've done cluster analysis on all of these different attributes, a dozen or more chemical characteristics of each pyrite specimen, we find that they cluster. Certain types of pyrites are characteristic of a kind of volcanic environment, and some other pyrites might be characteristic of a hydrothermal environment or a shallow low temperature environment where biology plays a role. It's those trace and minor elements. In the case of pyrite, that gives us a lot of information. But you need to do the cluster analysis and I want to make a point here. You don't do a cluster analysis on just two elements, so let's look at iron versus cobalt. You take many elements simultaneously. This is multidimensional analysis, and this is why I say we're entering the era of mineral Informatics, which instead of using the old traditional XY plots, we're analyzing our systems, typically in 10 or 15 or 20 different variables simultaneously. And you start seeing patterns emerge, patterns that the human brain could certainly not recognize by looking at a table of 10,000 analyses of 20 different elements. But mathematically you can do that analysis, and out pops these amazing results that there are different what we call natural kinds of mineral. They're all one IMA species. It's all pyrite, but they're different natural kinds.
Oliver Strimpel
That's really fascinating. You know, in a sense, defining a new mineral species. I suppose a little bit like the Victorian naturalist did. They went out in the field, and I guess at some point they had to decide well, is this bird feather configuration sufficiently different from this other bird feather? I don't know Darwin's finches, say, on the Galapagos, and he put them into different species because they were sufficiently different. Is this a similar kind of phenomenological characteristic, or is there something more analytical that comes out of this?
Bob Hazen
Oliver, it's exactly like that. In fact, I feel such an affinity sometimes to these naturalists. In the 19th century mineralogist James Dwight Dana and Charles Darwin. You know, in 1857, that's two years before he published On the Origin of Species, he coined a phrase of two words that are now used all the time in classification. He wrote to a colleague a botanist, he said, “I'm just struggling over this problem of lumping and splitting.” And what he was saying was that if you're talking about the origin of species, you better have a clear idea of what a species is. And do you lump lots of similar things into one larger category or do you split lots of things that have slight differences into separate categories? And the Galapagos finches is one example, and there are many examples in botany where you have to decide if something's a species or just a variety. And of course, there's a certain arbitrariness to that. That's true In mineralogy, too, we have to make decision lines. Are all diamonds the same? Or where do you draw the line? If you have a million diamond specimens, you could say, well, there's a million different kinds here, or you can say it's just one or what we've come up with is. Maybe there's about 6 that are clearly distinct because they arise through different processes. And Darwin was thinking the same way, and everyone who tries to classify things has to grapple with that. And it's a very human endeavor. One hopes that you're doing something that's reproducible and independently verifiable and rigorous and scientific.
Oliver Strimpel
I can imagine this is going to be hotly debated. So as we discussed, there's 5700 odd recognised mineral species today. Once you've completed your clustering analysis and using all the additional mineral characteristics beyond the structural and chemical ones that are traditional, can you guess how many mineral species we'll wind up with?
Bob Hazen
Well, I want to make a distinction. We will still have 5700 mineral species. Species is very formally defined and approved by the IMA, and I absolutely rely on them. I have to be able to use diamond. I have to be able to use quartz or pyrite. Those are the species. But we suggest from our first analysis that right now we know of 10,556 natural kinds. So it's roughly double the number of species, because those are based on the different ways that minerals form. So on average those 5700 minerals form in two different ways, and those two different ways create two different sets of attributes, characteristics. And those characteristics therefore can be used to define, going all the way back to Aristotle, this idea of natural kinds, a true division in nature. Well, of course that's a little bit problematic, isn't it? What's true is everything just a human construct, or are there real divisions? And I would argue, gosh, look at the periodic table, carbon is an element that has six protons. That's different from nitrogen. It's different from oxygen. It's different from every other chemical element. I think those are true divisions. It's a little less obvious when you get to minerals, because there's such continuous gradations of composition and trace orf minor elements and size and shape and all the other attributes. So one has to have maybe in some cases slightly more arbitrary divisions. But I do strongly feel that there are natural kinds in mineralogy, and that our job is to try to find those.
Oliver Strimpel
I'm struck by your statistics that on average the minerals are formed in two different ways, and I'm wondering if we look in this multi-dimensional phase space that you've constructed with things like the hydration environment and inclusions and optical properties and so on. Is there one axis along which we see the most differentiation, if you like, or most common distinction between the two modes?
Bob Hazen
That is a wonderful question. And you know, Oliver, I haven't ever really put it in that way. You've just come up with a really brilliant new way to articulate this. But if I were to say the axis would be related to temperature, pressure, composition space. If you can imagine an axis where you had temperature on one axis, pressure on a second axis. Pressure being the depth within the Earth. Temperature, being something that does get hotter deep in the earth, but it can also have volcanoes at the surface and then composition, which is a multidimensional variable in itself. And if you think of there being different sorts of volumes or regions within temperature, pressure, composition, space, when you find two different ways that mineral forms, it's typically in two different parts of that, what we call TPX volume. Pyrite forming at very low temperature in water versus pyrite forming in incredibly high temperature in a volcanic environment. So they just lie in two different places in temperature, pressure, composition, space. So you have two different kinds. I think that's the simplest answer. It's not the universal answer, but it is an answer.
Oliver Strimpel
How do you handle solid solution series like all olivine?
Bob Hazen
Well, this is really important. The solid solution series. So in the terms of IMA, International Mineralogical Association, you have two different kinds of end member common olivines. One is the magnesium end member called forsterite. One is the iron end member called fayalite. Well, that's all very well and good, two species. But I can show you a rock that has grains of olivine that are slightly compositionally zoned. That is, the ratio of iron to magnesium shifts gradually as you go from the center of the crystal to the outside. And so the center of the crystal can be forsterite and the outside of the crystal is fayalite, right? And this gets really extreme in some complex minerals like tourmaline, beautiful semi-precious gemstone where you can have a single crystal with lots of different colors in it that literally has seven different species of tourmaline in a single crystal. But to us, that's one natural kind. It's one crystal that formed in one environment over some period of time. The composition of the fluids shifted, and so you have different local concentrations of iron and magnesium or aluminum or some other element, but nevertheless it's one crystal. It's one kind. So this is where we do splitting or lumping according to what nature is telling us, whereas IMA quite legitimately has to divide things according to their end member compositions, because otherwise it would be chaos.
Oliver Strimpel
Coming back for a moment to the coevolution of minerals and life. Are there some minerals that can only be made with the help of biology and you named an exotic pair just earlier. However, are there ones that might be observable on exoplanets as a sign of exobiology?
Bob Hazen
So we've been thinking very hard. Is there a mineralogical biosignature? That is some kind of distribution or identities of minerals that would tell you, ah, life has been here or life is here. And I don't think we're at that point yet. There's no smoking gun. There's certainly no single mineral species that if you found it, you'd say, oh, it's unambiguous, we found life. I think the signs are going to be more subtle. I think in some cases it may be not the mineral species, but the morphology. The extreme case of that would be like a fossilized tooth, or an ammonite, or a trilobite, where the minerals are preserving the actual shape and size when something was clearly alive. But they're more subtle examples where you can have, for example, minerals that biology produces. But biology produces nanoscale grains, whereas under the normal circumstances of physics and chemistry you produce much larger crystals. There are lots of examples of that, and so the morphology could be the bio signature.
Oliver Strimpel
On your web page, you have a graph that shows the growth of the number of minerals over history. And I noted that it flattens off as we approach the present time. But aren't humans producing many new materials now, as well as new isotopes that have long half-lives that will be permanently or semi permanently in the record? And do such man-made minerals and isotopes count, if you like, in your classification?
Bob Hazen
So in our original description of mineral evolution, we divided Earth's history into 10 stages, and it's certainly true that in some stages, for example 7, which is the great oxidation event, you see a huge pulse of new minerals. And then other stages, for example stages 8 or 9, you don't see the same kind of change because there's not a lot of new temperature, pressure, composition environments produced during those stages. Stage 10 is the terrestrial biosphere. We do see a jump because of a whole variety of minerals. But not the same magnitude as the great oxidation event. However, are we now in stage 11, which would be the Anthropocene, the period in which human activities change the surface of the planet in significant ways. And one of those ways is producing so many new kinds of solid compounds, mineral-like compounds that will be preserved for millions, maybe hundreds of millions of years in a layer in a strata which will mark humankind. Now, if that's true, some people argue, well, since this is a very well defined geological strata, it's a marker layer. We should define this as a new geological time period, and mineralogical evidence has been used a lot to debate this point back and forth. We argue that there are hundreds of minerals already approved by the IMA as occurring naturally that are now being produced in larger quantities through human activities, especially near mining operations and smelting operations. There are literally 10s or hundreds of thousands of new solid compounds that will be preserved in waste dumps in cities that are buried over geologic time. Huge quantities of things like just bricks and cinder blocks and asphalt, but also different kinds of alloys. Stainless steels and aluminum alloys and so forth. Things that just wouldn't occur naturally, but are now part of the natural environment because of human activities.
Oliver Strimpel
I'm wondering if your classification system might provide a somewhat more objective way of defining whether we are in a new geological period, and how you'd identify the Anthropocene and distinguish it, say, from the Holocene.
Bob Hazen
What an interesting question. So, for example, the mineral hematite, Fe2O3, has been around Earth since the very beginning. There are a number of different natural kinds. It has trace and minor elements as size and shape, but now we're creating huge quantities of steel structures that rust and that rust is hematite, Fe2O3. And I was suspected if you did a cluster analysis on hematite from human artifacts, it would be compositionally rather different from the natural varieties that occur in many different environments. It would have different trace of minor elements. It would have different isotopes. It was processed and mined in different ways.
Oliver Strimpel
You might be responsible for determining where the Anthropocene Golden Spike goes.
Bob Hazen
The Golden Spike there are, of course, people who say it should be isotopic, perhaps tied to nuclear testing, which is a very dramatic spike. It could be tied to the burning of coal and other fossil fuels, which changes the carbon isotope signature. It could be just things like moving dirt around and tunneling and roads and bridges. Agriculture could become related to the golden spike. So there's lots of ideas out there, and I have to say it also has political overtones, as some people want to emphasize the fact that humans are changing the environment, while other people want to minimize that concept.
Oliver Strimpel
What are you working on at the moment?
Bob Hazen
The evolution of mineral evolution has been one of increasing quantitative rigor. We started with a nice story. In 2008 we published a paper, Ten Stages of Mineral Evolution. It's a narrative of planetary history that's very easy to swallow and digest. A lot of people like it. It's been used in museum exhibits and teaching and so forth. Second stage was to build the database. It took us about 10 years to build the mineral evolution database, which had locality, mineral species, and age information. Just three parameters, and we built it with 200,000 different mineral age locality data. But we now are realizing we need to get into the era of mineral informatics. That's all these different attributes and minerals. And so we're building huge databases, we're applying them. And we're preparing what's called the evolutionary system of mineralogy. This is a multi-part system. The first six parts are published or impressed. There'll be another eight or ten parts. A system of mineralogy has a very specific meaning. It means that you take every known kind of mineral and you place it into a conceptual framework. The current conceptual framework has been around for about 200 years. That's the IMA plan of organizing minerals according to their structure in chemistry, and that's not going away. That's really fundamentally important. But the evolutionary system says let's complement that by placing every mineral in its formational historic context, and these are the natural kinds I've been talking about. Well, it's a huge job to go from a qualitative 10-stage evolutionary history of the planet to 5741 stories. Some of which have multiple storylines, and you have to basically define every one of them and see cross relations. And so we're using big informatics approaches. We're using network graphing. You know how you can follow Facebook friends and people who buy certain products and so forth and graph them as a network of interactions? Well, you can do the same thing with minerals. Minerals have friends, minerals associate with some of their friends, and they never associate with other of their friends, and they form at high temperature. They form at low temperature, which are different parts of network graphs, and it goes on and on and on, and it's so exciting, and I could talk about this for hours.
Oliver Strimpel
We'll have to wrap it up for now, but we'll definitely want to check back in with you when you get to your final volume and get your overview perspective of the entire system. Bob Hazen, thank you very much.
Bob Hazen
Oliver, thank you so much. This has been a real pleasure.
Oliver Strimpel
For more about Geology Bites, as well as pictures and illustrations that support this podcast, you can go to geologybytes.com.