Dating in Space

What is it like being involved in NASA’s space program?

I love it. It’s a privilege and I’m always very excited to tell other people about it. I feel that I have an obligation to share my enthusiasm with the public, as they are the people who are funding the work.

What inspired you to get involved with dating space rocks?

I’ve been involved with studying Mars since graduate school, in particular the Valles Marineris and how this giant feature came into being. I discovered that the rate at which you model extension and applied tectonic forces can really change the ultimate shape of the canyon models you create – and those rates are controlled by the amount of time involved. As a student, I was surprised that planetary science didn’t understand time very well, with geologic maps showing errors of up to 700 million years (or even a billion years in some instances) for estimates of canyon formation. A billion years is a quarter of the age of the solar system...

Hence, from very early on, I thought about how we measure the history of things. Of course, our understanding of the history of Mars ties with our understanding of the chronology of the moon and being able to obtain dates from areas with known numbers of crater counts. In fact, on the moon there is a significant gap in dated samples of known provenance from roughly a billion years ago to around 3.5 billion years ago. Because of this gap, we have trouble constraining the ages of various other planets within the same period.

Also, about a decade ago I was working at NASA’s Jet Propulsion Laboratory (Flintridge, California, USA) where I became very interested in spaceflight hardware and building instruments. Back then, it was widely accepted that performing dating experiments on another planet was impossible and that you had to bring those samples back to the Earth – a very expensive proposition that might only happen once. Therefore, you had to pick the right sample – and characterizing a whole planet from only one sample location would be tricky. Even with the Apollo programs that brought back samples from a number of places (270 kg of rock), we still want to know how old the different surfaces of the moon are. In the ‘Decadal Survey’, NASA’s blueprint for the scientific exploration of different planets, the number one objective for new lunar missions is to obtain more dates to help us understand not only the history of the moon, but also the solar system. During this time, I started thinking about approaches to miniaturizing a range of geochronology instruments.

So the idea of building instruments to obtain these important dates evolved from working at JPL and from thinking about how we can measure in situ instead of the very expensive and difficult sample return method. It’s a political subject; I believe that we still need to do sample return in addition to in-situ dating, but a single sample return is not going to be enough to answer all the questions. With in situ analysis, which for Mars is approximately 10 times cheaper, you could fly missions to many different places for better characterization of a planet.

The final factor that inspired this project is my good relationship with Joe Boyce, an early leader of the Mars program at NASA. We talked a lot over beers and he once said, “Scott, you need to see if you can solve this in-situ dating problem using in-situ chronology.” At the time, it was a crazy idea because no one really thought it was possible, but it inspired me to look into the laser ablation resonance ionization mass spectrometry technique. I actually looked at many different approaches and thought this was probably the best one. So we set out writing proposals and trying to figure out how to make it happen.

How does the technique work?

In all rocks, there are trace abundances of rubidium, including the 87 rubidium (87Rb) isotope that decays into 87 strontium (87Sr) radioactively over time. If you can measure how much 87Rb and 87Sr is in a rock, then you can determine how much time has passed since that rock was formed. However, these two isotopes have essentially the same mass, which makes dating difficult. To measure them with a mass spectrometer, you either need extraordinarily high resolution or you need to engage in some sophisticated tricks to separate the two.

To overcome the problem we use a multistep process. We put the rock in front of our mass spectrometer and our laser beam, and we ablate it to free up neutral atoms, amongst other things, from its surface, including those (Rb and Sr) that we want to measure. We illuminate the explosion plume from the rock’s surface with lasers tuned to ionize the Sr atoms, a process called resonance ionization. At that moment, we use electric field pulses to push just the Sr atoms into the mass spectrometer. We repeat that process a fraction of a second later with lasers tuned for Rb atoms, and push those ions into the mass spectrometer too. Because there’s a time gap between when we inject the Sr and the Rb, we can measure them separately in the mass spectrometer without them overlapping or interfering with each other, which provides an independent abundance measurement. To obtain good statistics, we repeat this one thousand times, while also measuring non-resonant backgrounds. To fully assess the variability in a rock, we ablate 300 different locations on samples of roughly 1–2 cm2.

The only real requirement for sample preparation is a flattish surface or a known surface shape, such as a cylinder, which can be easily created in the lab. Those simple shapes are actually very consistent with existing systems for coring rocks, such as those used by the Mars Science Laboratory (MSL) or Mars 2020.

It takes two to three days to produce one of these measurements, but with the spaceflight designs we’re working on, we think we can speed it up about 250 times, bringing down the time to two hours. The error bars that we get with these dates are typically about 200 million years, which was the case with Martian meteorite Zagami, and a lunar analog, the Duluth Gabbro (1, 2). Since we’re looking at billion-year error bars for Mars and the moon, this 200-million-year bogey is good enough. With other samples we sometimes get this down to the 50 to 89-year range; however, those are for samples such as granites that are not commonly found on Mars, the moon or other planets. We are also always looking for ways to improve the measurement, and believe that the accuracy will continue to improve as we track down subtle instrumental issues.

How does the technique compare with other dating methods?

The pre-eminent method of dating in a lab is to use thermal ionization mass spectrometry (TIMS). This requires someone to take roughly three to 15 (sometimes more) small samples of a rock and grind them up, before processing them chemically to leech out the Rb and the Sr into a liquid state, which is then dried onto a filament. The filament is then placed in the mass spectrometer and heated up until the atoms boil off as ions, which are then measured independently in the mass spectrometer. This TIMS technique is exquisitely sensitive: typically, you see results as good as half a million years or so, compared to our 200 million years. In fact, these measurements are fantastic – and much better than we can do on-board a spacecraft. But...

There are some trade-offs with TIMS. Firstly, there’s a lot of manual labor involved that can’t take place on a spacecraft. Secondly, you’re commonly looking at three to 15 measurements whereas we’re looking at maybe 300 spot measurements. We acquire these 300 measurements in a rectangular pattern, forming an image. From each one of those pixels we get elemental abundance, the Rb and Sr isotopes, and the organics. The image also allows us to map the mineralogy in the rock, allowing us to assess potential biases such as thermally or aqueously altered minerals. For example, sometimes in terrestrial rocks we have observed that the isochron is well behaved for feldspars, but not for amphiboles. Thus, mapping of the minerals in the rock with our instrument allows us to see and understand any “misbehaving” minerals.

In contrast, when you only have the traditional three to 15 measurements, it’s much harder to draw such conclusions. Interpreting the behavior of certain mineralogies in the rock helps us to understand how much heat was applied to it or how much water flowed through it during formation. It’s much easier to do this using our technology than with traditional techniques.

So you’re revealing the story of the rock?

I like to think so. And although the traditional techniques are much more accurate, a 100 to 200-million-year solution is probably good enough. A half-million-year solution in this case may actually be overkill. The real question is: can we answer scientific questions with what we’re proposing? I think the answer is “yes”, which is why this idea is radical and new – and why we’re starting to get more attention!

What other challenges are involved in dating rocks from/in space?

Planets like the moon and Mars are easier to date because they don’t really have plate tectonics or extensive weathering induced by water. Because of this, interpreting the rocks is often simpler. On the flipside, some terrains, particularly on the moon, are very low in Rb and Sr so are very hard to measure. But our recent progress measuring the Duluth Gabbro, a lunar analog (2), indicates that we are capable of addressing many of these terrains.

What were the main challenges developing the technique?

First, we spend a lot of our time focusing on understanding laser ablation, but it turns out there were other parameters we initially considered unimportant, for example, temperature stability in our laboratory. Lots of our lab equipment is sensitive to temperature changes.

Once we sorted out the lab equipment, we focused on laser ablation. Laser ablation is a very complex process, even though we’re just using it to free the neutral atoms. It’s not really the core of what is allowing us to measure the Rb and the Sr; in fact, it’s the second step, resonance ionization (using the second set of lasers), that allows us to separate the Rb and the Sr. Nonetheless, laser ablation is prone to fractionation and can even affect the production of neutral atoms. Scientists often worry about fractionating ions produced from a sample surface, and yet ions are actually relatively rare in this scenario – almost everything that comes off is neutral. While it’s commonly thought that the neutrals are not fractionated, when you’re measuring things at one part per million it turns out that those fractionation processes still matter. Thus, getting the laser ablation process to behave is very important.

Next, we needed to study the resonance ionization process – consisting of three lasers for Rb and another three for Sr – that allows us to separate and measure the abundance of Rb and Sr for each mineral. However, even with resonance ionization, it turns out that contaminants (mass interferences from molecules other than Rb and Sr) can still sneak in. They may be at very low abundance levels; however, even one part per million would still be enough to bias your answer. If there are background contaminants in the measurements, they may be there without any combination of the three lasers used for resonance ionization. To address this we take a background measurement whereby we turn the resonance lasers on and off and subtract this from the other measurements.

Finally, when developing a new technique, you just can’t measure something once – you must do it many times to develop any kind of confidence in a result. Sometimes we get results that don’t make sense, requiring us to go back and relentlessly study the issue until we can work the problem out.

Ultimately, we have worked really hard to change negative perspectives and more people are coming around to the idea that it can be done – maybe not as well as with the TIMS instruments, but well enough for spaceflight applications and some terrestrial applications.

Did the negativity make you more determined to succeed?

Yes and no. It’s always very intimidating when an expert in the field and one of the ‘old boys’ tells you you’re barking up the wrong tree. I would ask them why it couldn’t be done. Then I’d take the opportunity to explain what we’re doing. The story that I started to hear repeatedly was that no one else had thought about doing it in the particular way we were thinking of using. That’s when I started to have hope.

Where to next?

Actually, there are additional steps that are closer to the science. Our real focus now is on measuring as many samples as we can, looking for any problems and showing the capability and limits of the technique. We have already identified a couple of areas where we think we can do better and we’d like to explore those to see whether we can improve the error bars that we are getting today. It seems that every year that goes by we seem to be able to do a little bit better than the year before. When we started, we could barely measure easy terrestrial granites, then after continuous improvements we found we could measure harder meteorites from Mars (like Zagami). After further improvements, we were able to measure difficult lunar analogs like the Duluth Gabbro. We continue to iterate on the ever harder samples, to improve our accuracy, and open more of the solar system to potential analysis using this technique.

What will be the challenges of getting the technology ready for spaceflight?

There are three. The first challenge is the laser systems. There are similar laser systems that exist and have been flown, but our exact laser system has not been built in its entirety. We have NASA funding to build it so we’re thrilled to have the opportunity to do this kind of development. We have now demonstrated two of the lasers in prototype form, and are working on implementing the remaining systems. In the next couple of years, we plan to test these in simulated space flight environments.

The second difficulty is sample handling. I’ve said that sample handling is straightforward and that we believe that existing tools like coring-drill and the rock-abrasion tools will work, but doing it remotely on a different planet is always going to be a challenge. We are working with space flight tool manufacturers to overcome potential difficulties and to show how all the pieces work together. It’s always important to fine tune the sample handling approaches and test them in real-world ways.

The third tricky issue is the politics of writing the proposals and going after this kind of science. Usually it hinges on the question of sample return versus doing measurements in situ. There are many people who feel they are inherent enemies of each other; I think in many ways it’s because these two communities are afraid that one or the other of them won’t be able to do the work over the long term. I think that’s the wrong way to look at it because both of these measurements are very important, as I expressed earlier. Bringing back a sample allows you to do incredible work that you won’t ever be able to do in situ. It takes that kind of insight to really characterize another world, so we need to stop having a debate over which one is going to win and instead talk about how we can use both approaches.