Watching continents slowly break apart, and turbo charging robotic sniffers
26:04 Share

This podcast is supported by the Icahn School of Medicine at Mount Sinai, the academic arm of the Mount Sinai Health System in New York City, and one of America's leading research medical schools. What are scientists and clinicians working on to improve medical care and health for women? Find out in a special supplement to Science Magazine prepared by the Icahn School of Medicine at Mount Sinai in partnership with Science.

Visit our website at www.science.org and search for Frontiers of Medical Research-Women's Health, the icon school of medicine at Mount Sinai. We find a way. This is the Science Podcast for November 8th, 2024. I'm Sarah Crespi. First this week, staff writer Paul Vussen talks about hunting for signs of the rifting that's in the process of tearing the African continent apart. It'll just take a few million years.

Next on the show, researcher Nick Denler discusses speeding up electronic noses. These fast sniffing devices could one day be mounted on drones to help track down forest fires before they're large enough to spot with a satellite.

This week's staff writer, Paul Vucin, wrote about the coming breakup of the African continent, you know, in a few million years. Hi, Paul. Welcome back to the podcast. Hi, good to be here. Yeah, so we don't know exactly when this is going to happen, but I think just to start with, Pangea might be a good touchstone for people. It's a supercontinent that began breaking up about 200 million years ago.

But the planet is never done shifting. Late tectonics continue. And in this case, we're going to be talking about something called rifting in Africa, which is kind of like breaking up a continent, right? Yeah, that's exactly what it is. It's, you know, how a continent begins to die, how I like to think of it. So why Africa? What do we know about what's happening there that makes us think that it's not as it's not going to stay forever together?

For decades, Africa, in particular, Eastern Africa, has been the model for how continental rifting takes place because this is one of the most active rift zones or the most active rift zone in the world going from Ethiopia and heading downward. You know, you have parts of it that are already well into the process of spreading and splitting apart.

and then kind of changes as you go further. But so this has been an object of study for Western scientists, especially coming in for decades. One of

One thing you mentioned in the story is that there's a river that doesn't lead to the ocean, which is kind of a hint that rifting is happening. There are lots of things that can create depressions in the landscape. But yeah, this is one of those that really starts to kind of cause these rift valleys, famous rift valleys that really where humanity evolved kind of entirely stems with the breakup of Africa. So interesting. Why do you call it dry rifting? So how is this different than wet rifting? I guess the opposite.

What would be kind of the shorthand for magma? So, you know, one of the big questions of how continental rifting works is it was assumed for a long time that you need

some sort of magma to weaken the crust. So melting it from below, right? Yeah. So, you know, when to break apart a continent, that is not an easy thing. The kind of normal plate tectonic stresses have a hard time, you know, the stuff from just that comes from like the other side of the plate sinking into the mantle because you have to break apart both the crust and the upper part of the mantle. It's not just the crust.

So, you know, if you start to kind of seed magma into the crust, maybe you weaken it and that causes breakup. But there are parts of Africa, and this is also seen on kind of older rift zones that have already broken apart, like say the mid-ocean ridge of the Atlantic.

They see no evidence of magma, but rifting is happening. So what's going on there? That's dry rifting. So is that why you're writing the story now? And we've known about the rifting in Africa for a long time. There's this question of how far does the Eastern African rift zone extend? One of the things that come with studying active rift zone is you can study how it starts, how it spreads, kind of what does rifting look like?

At the very start, we call it incipient rifting. And that is going on right now in Botswana. It's one of the places. So that's where I went with scientist Fulorin Kholawal and his collaborators who has been studying this process there. So just the very beginnings of a rift. Wow. It's not something you think you'd get to see because how often does it happen? But I guess, you know, part of all these processes are frozen somewhere along their timeline, somewhere on Earth, right? Yeah.

Well, not frozen, just moving very slowly. Sorry, sorry. Frozen in non-geological time.

Exactly, exactly. So what did you see when you went to the Suwanna? Mostly a lot of sand. You know, the landscape is kind of draped in Kalahari sand. You know, it's not kind of a region of high relief, dramatic mountains like the rest of the rift zone in the north where you have these big valleys. The work they're doing is really kind of you infer from kind of small rises in the slope, then you get a backhoe.

dig a trench, look for kind of shifts in the sediments to indicate there's a fault, like a split in the rock. And if that fault kind of matches up and, and,

is indicative of movement that seems to indicate rifting. Also, they hunt for quarries. Any other way you can kind of get into the rock itself to look for evidence of the extent of the rift and the faulting. Yeah, so what kind of evidence were they able to find or what kind of testing did they do there? So right now, the Eastern African Rift Zone is known to and kind of takes a bend,

to the west when it gets near the Indian Ocean in the south and goes into Botswana to the Okavango Delta, famous Okavango Delta. Researchers I was with have been finding is that

Actually, it seems like it may not stop there and that there are indications from earthquakes and now the discovery of some of these faults that it could take another turn and extend into South Africa, you know, maybe heading towards the ocean that way. There are also suggestions that it could go through Namibia all the way to the Atlantic. And they're also seeing kind of researchers are seeing that just the picture, the future of kind of what this will look like is not this kind of clean split.

you know, of Africa and Cuba could be a series of microplates. Madagascar could split again. Just kind of this big archipelago almost in the future. So,

So that's one thing they're trying to do is map out the future of the rift in this area. And then, and I do want to mention, there is a really nice map that's in the news story. So I know this is audio. You can't see kind of like a map in your mind, perhaps very well. So definitely check that out online. So Paul, are they also looking, you know, from space or from drones to see some of these signs of rifting under the surface? Yeah, Fola, who I was there with, he does

a whole host of techniques. So he's a structural geologist. That means he like studies kind of rocks in situ and layers and their context, but also he pulls in a lot of geophysical tools. So he measures kind of, you know, instruments they run on the ground to see detect changes in, you know, electrical connectivity to see faults there. You can use, even like Google earth, you can kind of see it vegetation or change the slope that reveal that,

It's the first suggestion of a fault that then you need to kind of go into the field and verify. You know, kind of this new generation that is really combining geophysics with more traditional geology to try and push the field forward. So you spot something that indicates maybe the rift is there or a fault is there. What do you do on the ground to say yes or no, this is something that I'm looking for?

They look for things like slicken lines, which are evidence of earthquake slippage that happened in recent past. They look for this kind of shifts in the sediments and then they date it to get when it was exposed to light, which can lead them back to when this last earthquakes, when this kind of change happened. If it's more modern, that suggests it's more active. There's

There's always some debate over when it is truly a rift or not. You also have to show kind of the type of faults. There are different kinds of faults. You know, if it's a fault like the San Andreas, that's not continental rifting. Is a rift a specific depth or, you know, how do you define something as a rift? It's always hard to know kind of precisely. There's still some debate about the Okavango Delta, even though it's widely accepted that it's a rift.

A rift zone that, you know, some saying, hey, maybe it's not, you know, you need to see this kind of thinning of the mantle, you know, that helps suggest it. The movement of the fall to kind of showing that you can see this type of movement, you know, that it makes sense in this broader context. But until you get to a certain point when it can't be denied, there's going to be some debate over it. So why is it important to understand where the rift is going, where it is now, what's happening with it? Beyond the kind of pure theory.

curiosity of the future of Africa and how this all works. Yeah, where the continents are going to be in the future is kind of an interesting question, right? Yeah. And one big question that Fole and his team are working on is when you have this tri-rifting, what else is kind of shaping this? What is guiding this? They know that it can be like

old mountain belts that are now buried underneath everything. This have this like weakness in the crust and perhaps it's their features within those mountain belts that reflect when this hydrothermal blasting of the rocks that these rift defaults can take advantage of this weakness.

But also, these represent hazards. So rift zones are dramatic, beautiful landscapes that also have lots of volcanoes. When they mature, they have lots of earthquakes. So say if this is a true rift zone that is going towards the capital of Botswana, they would want to know that. Absolutely. Where else is rifting happening? I guess I had never even thought about where else there are rifts. There are.

aren't a lot of places right now. On land, I guess. I think there's one in the Czech Republic, if I remember right. I know, kind of strange. I don't know that it's really going much of anywhere. The Gulf of California, I believe, is one. You know, there are famous kind of past rift zones, like Mid-Atlantic Ridge is kind of your best example there. How long does it take for a continent to break up? Do we know?

I mean, we're talking millions and millions of years. I'm sure estimates vary depending on the particular environment it's in. It's not anything we have to worry about in the near term. Yeah. So what made you decide to write about this? Well, I've always been interested in continental rifting. And I met Fola a few years ago, who eats himself from Nigeria, working at Columbia now, brings this kind of great melding of...

different parts of geology and geophysics together. And also he's kind of a new generation of scientists who's trying to make sure, you know, his collaborators in Africa, he's bringing in graduate students to be with him in Botswana or whichever country he's working in. This is something we're also seeing from other countries.

scientists working in the region. He also owes that to kind of some predecessors as well. He was kind of the perfect figure to go there with and get a sense of this. All right, Paul, this has been super fascinating. Thanks for talking with me. Yeah, my pleasure. Paul Voussin is the Earth and Planetary Science Reporter at Science, covering everything from the fringes of the atmosphere to the innermost inner core on Earth and elsewhere in the solar system.

Don't go anywhere. Next up, we talk high-speed robotic sniffers with researcher Nick Dedler. Japan's Noster specializes in postbiotic gut microbiota metabolite-based pharmaceuticals research to treat metabolic and immune-related diseases. Noster's products include biosynthesized GMP bacterial preparations and QMEC, the world's first HYA-50 metabolite postbiotics healthcare supplement.

Their analytical services include liquid chromatography, mass spectroscopy, metabolome analysis, and state-of-the-art genetic sequencing for gut microbiota analysis. Visit www.noster.inc to discover how Noster can help you. This podcast is supported by the Icahn School of Medicine at Mount Sinai, one of America's leading research medical schools. Icahn Mount Sinai is the academic arm of the 8-hospital Mount Sinai Health System in New York City.

It's consistently among the top recipients of NIH funding. Researchers at ICON Mount Sinai have made breakthrough discoveries in many fields vital to advancing the health of patients, including cancer, COVID and long COVID, cardiology, neuroscience, and artificial intelligence. The ICON School of Medicine at Mount Sinai. We find a way.

We've talked a lot about odors on this podcast. I'm basically fascinated by the sense of smell. It's this amazing scientific puzzle. How does this enormous collection of odor molecules, we call it the smell space, how do all these different chemicals interact with receptors in our cells and then somehow get translated into a sensation of smelling a cookie or a burning tire?

Another question is, are humans really as bad at smelling? You know, do we give ourselves enough credit for how good we are at smelling? And then here's a new one. How fast do we or other animals detect odors? And can we replicate that speed in a robotic nose or in a robot? This week in Science Advances, Nick Denler and colleagues wrote about creating a fast electronic nose. Hi, Nick. Welcome to the Science Podcast.

Hello, so happy to be here. Yeah, I'm so excited to talk about the sense of smell in a new way. This is not something I really considered before, faster than the speed of smell. So how fast do animals smell? How much do we know about that? Obviously, that's very different for different animals and different species, but generally it's actually really fast. So for example, a

Several insects, such as grasshoppers and fruit flies, they have transduction times of a few milliseconds and less. And that makes them really fast at reacting and navigating through complex environments. We figured this out in insects. Do we know anything about the mammal side? Yes, it's interesting because it has long believed that the sense of smell in mammals is really slow, but it has been disproven recently.

So this major study that came out a couple of years ago has shown that mice, they can discriminate between pairs of smell also down to the timescale of milliseconds. And actually also humans are really fast, not as fast as mice, but they're really fast as well. Why might it be useful for robots or electronic noses to be fast?

faster to do this like high-speed smelling that it seems that animals can do? It really depends on what you want to use such a machine for. Traditionally, gas sensors have been used for many things that did not require the high speed. For monitoring of air quality in office building, it's sufficient to be, you know, in the time scale of seconds to minutes.

or if you want to build a smart fridge, you also don't need to be that fast. A smart, wait, a smart fridge? Yeah. So if you want to recognize if a certain food is about to spoil, you might be able to smell that, right? So there's actually companies producing those smart fridges that have the sensors inside. They will tell you exactly when your product is going off. Okay. But yeah, time is not that important. It's not, you know, minutes is fine.

No, it's true. However, if you want to place such an electronic nose on a mobile platform, like a robot or a drone, and you want to use it for navigation purposes, then it suddenly becomes really, really critical to be fast. I think we should take a second here to talk about what an electronic nose is. When we talk about electronic nose, it's a kind of a very special use case. And it just, yeah, why don't you just tell us what exactly that means in this scenario? Yeah, so an electronic nose is,

basically an electronic device that can perceive smells, right? So often it has different sensors on them and you have electrical periphery to read them out and to control the sensors. And you can design this for different use cases and depending on what you want to do exactly, the design might look different. What's the range of orders that an electronic nose could smell? Is it

Five? Is it a thousand? Again, it really depends on how you design the device. So in our case, we only had eight sensors on our electronic nose. So the space will map maybe 10 or 20 different odors that you can distinguish at once. However, there are different designs that might not be particularly fast, but they can map 300 or more odors at once. And what's the technology there where the chemical meets the sensor? Like what's happening in that

So that will be a little bit more technical, but the sensors that we are using, they're based on the metal oxide technology. And so this sensing principle is based on a little oxide layer that changes its conductivity depending on how different gases in the air react with it. And so that means that it's not a lock and key type mechanism like a receptor. Like you can just have signals that you're looking for that are, you can set it up. This is the signal I want and then just send it to the sensor.

train your kind of the downstream of the detection to signal, oh, this smell has hit the sensor. And it's a little bit more agnostic to the molecules. Absolutely. Yeah, that's correct. So each of those sensors will give you

basically a one-dimensional sensor response. And if you have multiple of those sensors that are all kind of broadly receptive to different odors, you can get this fingerprint that might be characteristic for an odor and then learn that downstream. This is something that also comes up a lot with doing odor research is producing odors, controlling where they go, timing them. All that stuff is like, it's a whole technology setup that also can be really intense or intricate.

Yes, that's correct. It's actually extremely difficult to deploy odorants in a precise way. And fortunately, we had a great collaboration with the Francis Crick Institute and

And they have been performing those experiments with the mice. And we managed to use their setup. And what it consists of is basically a system of different, really fast odor valves. And with that, they can shatter those odorants very precisely. You can then design your stimulus as you want it to be. Yes, so you can say pulses or alternating or...

Or just, yeah, how fast using this setup of these tiny precise pulses of odor and your little electronic nose, how fast were you able to detect smells with this setup? So one of the key results was that we showed that you can classify short odor pulses down to 10 millisecond duration if you have, let's say, five different odorants.

And the second key result is that we can decode correlations between two odor pulse strains up to frequency of around 40 Hertz. And this is the result that's been shown in mice. So we match this performance in mice on a very similar experiment. How does that translate into, you know, a mobile robot? It's a miniature sensor. It's on board. What can you do with that? There is a

Quite a few applications that I described in the paper, but one of my favorite ones is not mentioned so far, and it's more in ecology. You can use those fast gas sensors, for example, for the early detection of wildfires. So if you, a bit of context, wildfires, they act globally as one of the largest CO2 emitters.

And conventionally, they are detected via satellites or small monitoring airplanes. However, for them to be seen from the air, they need to already have a certain size. And if the fire is below the tree canopy, it's really hard to detect them from above.

And there, fast gas sensors, that could be a great solution for this. And in fact, there is some startups already in Germany that are doing this. They're deploying those relatively fast sensors in the whole forest and then detect the fires when it comes up.

But actually, since they're stationary, you need a really big infrastructure if you want to cover a large area. And I believe that with the sensors and the electronic gnosis that we developed, you could place those on a drone and you could patrol the forest from the air for the smell of the fire.

And if that reduced the infrastructure while increasing the area. Could you use this ability to detect changes between the odors to kind of do locations or track down a smell?

I would say so. I think you can use this for any application with a similar mechanism. So literally anything where you perform an active search based on the smell. So odor navigation or olfactory navigation. Why is it important to animals to be able to detect odors quickly and discriminate, you know, different odors in a really speedy way?

If we look into the field of fluid dynamics and if we study how these odors are dispersed in the air, we notice that the processes are much more complex than we naively would expect. For example, it might sound quite intuitive that the odor intensity is following a gradient between the source and the receiver, but that's actually not true in most cases. So odors, they come in intermittent packages, which are produced by

micro and macro scale dynamic processes in the air. And those packages themselves, they can be extremely short and there might be lots of time passing between two of them. This makes it really important for animals to sense these cues rapidly and process them as efficiently as possible. Because it's going to be danger, it's going to be food, it's going to be weather, all that different kind of stuff. Yeah, it could be anything from looking for food or looking for a mate or sensing, yeah, smoke.

So what made you decide that studying odors or electronic noses was for you? Why do you work in this field? Yeah, so I'm a physicist by training and I got my undergrad in Switzerland. Before and during my master's, I've done a few internships where I've experimented a lot with novel sensor designs. So new types of cameras or also gas sensors. And I always found it really fascinating that the sense of smell is so underexplored.

especially considering that chemosensation is likely the oldest and most fundamental sense that has evolved in animals. For me, I was super excited to start this PhD in that field that attempts to combine insights from physics and biology, but also electrical engineering and computer science. This has been a wonderful conversation. Thanks so much for talking with me. Thanks.

Thanks for having me. Nick Dendler just this past week successfully defended his PhD. His work was conducted while in the biocomputation group at the University of Hertfordshire and the International Center for Neuromorphic Systems at Western Sydney University in Australia. You can find a link to the science advances paper we discussed at science.org slash podcast.

And that concludes this edition of the Science Podcast. If you have any comments or suggestions, write to us at [email protected]. To find us on podcasting apps, search for Science Magazine, or you can listen on our website, science.org/podcast. Do also answer our survey. We have that running right now, and we definitely would love to hear from you in that form as well.

This show was edited by me, Sarah Crespi, and Kevin McLean. We had production help from Megan Tuck at Podigy. Our music is by Jeffrey Cook and Wenkui Wen. On behalf of Science and its publisher, AAAS, thanks for joining us.