New Geochemistry Technique Yields Clues about Earth's Earliest Days

Half a century ago, in lab nicknamed the "Lunatic Asylum" in the Charles Arms Laboratory of the Geological Sciences, the late Gerald Wasserburg constructed the first-ever digital mass spectrometer. That device, dubbed the Lunatic I, revolutionized the field of geochemistry by increasing by an order of magnitude the precision with which isotope ratios could be measured; isotopes are the "flavors" of elements and vary based on the number of neutrons they have in their atomic nuclei. The Lunatic I ultimately was used to make high-precision measurements of the first lunar samples obtained by the Apollo missions.

Now, in the neighboring Seeley W. Mudd Laboratory of the Geological Sciences, a building crowded with graduate students and bulky mass spectrometers, Assistant Professor of Geochemistry François Tissot carries forward the Institute's tradition of pioneering geochemistry.

Tissot, who started as a faculty member at Caltech in 2018, set out to accomplish a goal thought to be impossible: probing the isotopic ratio of the element uranium inside a single crystal of zircon. Variations in this isotope ratio can change ever so slightly the calculated age of a sample, making this analysis a condition sine qua non for accurate dating. But single zircons crystals had long been assumed to be too small to be subjected to uranium isotope analysis. In work published in the Journal of Analytical Atomic Spectrometry in August, Tissot described his analysis of 31 individual crystals. The work confirmed some hypotheses—and ruled out others—regarding the conditions on Earth immediately after the planet formed.

Recently, Tissot answered a few questions about his research and newly developed methods:

What kind of questions are you hoping to answer with your research?

I'm an isotope cosmochemist by training. Among other things, we want to know where the material that makes up the solar system came from. What type of stars gave birth to the building blocks of our galactic neighborhood? The raw material that makes up our world is the result of billions of years of what we call "galactic chemical evolution," a term that describes the change in composition of our galaxy as nuclear reactions in the interior of stars produce heavier elements from lighter ones, like hydrogen and helium. When massive stars die, they shed material and sometimes even explode, blasting these newly produced atoms out into the galaxy. With high-precision analysis of the isotopic compositions of these new elements, we can learn more about the star or stars that birthed the material that eventually created our home.

How can you study something that occurred more than 4.5 billion years ago?

Good question. It's not like we have a piece of the solar system when it formed, right? Well, actually we do, in the form of meteorites, which are rock fragments left over from the birth of the solar system that eventually crossed Earth's orbit and fell on our planet. Not just any meteorite is suited for study, though. Just as rocks on Earth have undergone countless transformative processes over the past 4.5 billion years that have erased most of the information about the solar system's birth, meteorites from Mars or other planetary bodies are poor recorders of the earliest times of our solar system. But a subset of meteorites, called chondrites, have been subjected to far fewer changes. They have never been molten and essentially look like an aggregate of loosely related constituents, potentially formed at different times and places in the solar system. Analyzing these constituents provides us with critical clues about when, and under what conditions, they were formed. For example, was a sample formed somewhere that was hot and poor in oxygen? Most likely, it formed close to the sun. Does a different sample contain a lot of water? Then it probably formed in the outer solar system, where there are cooler temperatures that are necessary for water to exist.

How do you analyze a meteorite?

Our work on isotopes is slightly different than what I just described and more aggressive. We take a piece of meteorite, inspect it, clean it, digest it in acid, and break all the chemical bonds between its constituents until it is completely dissolved. Depending on the question we are trying to answer, we then isolate one or more elements and determine their isotope compositions. Some elements are great chronometers, others tell us about oxygen levels, and still others tell us about a meteorite's DNA, so to speak; if two meteorites have the same composition, they formed from the same parent material.

For example, let's say we look at a chondrite. We see a lot of small round spherules, called chondrules, as well as some bright white-to-pink inclusions that look nothing like the chondrules. These are calcium- and aluminum-rich inclusions, or CAIs for short. Why is that interesting? Calcium and aluminum are known for being very refractory; that is, if you heat up the major elements that form a rock, these will be the last ones to evaporate or melt. Conversely, they're also the first ones to condense when going from hot to colder conditions. As the solar system formed, it cooled, and calcium and aluminum were the first elements to transform from a gas to a solid. These CAIs are the oldest solids that we have ever found. Dating CAIs is how cosmochemists were able to determine the age of the solar system in the first place.

How do you figure out how old these minerals are?

Through isotope ratio measurements. It's much like carbon dating, only using heavier elements with much longer half-lives, like uranium. When uranium decays, it turns into lead, so we can study the amount of lead and uranium in a sample and get an idea of how old the sample is. We also study the ratio of different isotopes of lead found in the sample. Elements exist in multiple flavors, or isotopes, based on how many neutrons they have. Measuring the ratio of one flavor to another in a sample gives us information about how the sample formed, letting us piece together information about the sample's age as well as what components were injected into our neighborhood to generate the cloud that would eventually condense into our solar system. So, we can track the type of stars that were around the solar system and exploded and polluted what would become the solar system.

What are the challenges you face in this research?

For our recent study, we moved away from meteorites and worked on zirconium silicate, or zircon, a mineral known for its resistance to being changed chemically. Some zircons formed on Earth more than 4 billion years ago. They are the oldest samples of the primordial Earth that we have, and studying them has changed—and continues to change—our understanding of the early Earth.

We wanted to measure uranium isotopes trapped in zircon. The challenge, however, was sample size. We had individual crystals of zircon that don't yield enough material to study. As I mentioned, to process a sample, we digest it in an acid bath until all of its chemical bonds are broken and then strain it through a resin that separates out specific elements. What we're left with is a single element in a solution. And that's where the difficulty lies. There's a fundamental limit on the volume of solution we can analyze with current machines. It is around a quarter of a milliliter. You might think: Why not simply dilute the sample and make a larger volume? The catch is that we need the concentration of the element we want to study to be high enough to generate a signal big enough to measure with a mass spectrometer. That concentration is about 5 to 10 nanograms per milliliter. Anything less, and you won't be able to get usable data.

Individual grains of zirconium silicate often contain no more than 1 or 2 nanograms of uranium. To reach a concentration level of 10 nanograms per milliliter, one has to work with a volume of solution close to the lower limit of what the instrument can handle. Until our work, most people in the community thought this measurement was plain impossible.

How did you solve this problem?

We cut the volume of our sample down to a fifth of a milliliter and pushed the instrument to its limits by using a different "spike"—a spike is an artificial isotope that allows us to monitor the quality of our measurement. It took a great deal of testing and development, but with these careful adjustments, we were able to measure uranium in a single zircon crystal. Contrary to what others have claimed, it turns out that this is not an impossible task, just one that requires a lot of meticulousness and patience.

What has this allowed you to discover?

As we announced earlier this year, we tested uranium from 31 zircon samples that date back to the Hadean, Earth's first geologic eon, which ended about 4 billion years ago. The eon is named after Hades, the Greek god of the underworld, for the hellish conditions that are thought to have existed at the time, with oceans of lava, a total absence of water, and so much radioactive material that nuclear fission—as performed by humans in nuclear plants—was occurring naturally at Earth's surface. Quite the picture! If nuclear fission were really occurring so widely, however, we should have been able to find evidence of it easily in these samples. The burning up of 235U drastically changes the ratio of the two long-lived isotopes of uranium (238U and 235U). Instead, we found that zircons from the Hadean have exactly the same isotopic composition as meteorites and other Earth rocks that have not experienced any nuclear fission reactions. We found no evidence of natural nuclear reaction at all. This indicates that conditions back then had more in common with the current Earth than previously believed.

What's next?

Now that we know how to measure zircons, the challenge will be to scale up our data production in order to study specific periods of Earth's history in detail. The method we developed shows that we can get more reliable age information and maybe tease apart geologic events that are, so far, impossible to resolve from one another. We also found that uranium isotopes might be a tracer of the amount of oxygen in magmas. More work is needed to turn this into a full-blown tracer that could allow us to understand the how magma generation and the oxygen level of the earth's crust changed through time.

_{Image: Cathodoluminescence image of a Hadean zircon crystal, similar to the ones studied in this work. This particular grain is 4.06± 0.1 billion years old. Credit: Elizabeth Bell}