Lizard-inspired building design could save lives

Benjamin Thompson

Welcome back to the Nature Podcast, this week, creating a recyclable resin for 3D printing…

Lizzie Gibney

…and how a page from a lizard’s playbook could help mitigate disaster. I’m Lizzie Gibney.

Benjamin Thompson

And I'm Benjamin Thompson.

<music>

Benjamin Thompson

3D Printers come in lots of different types, as do the materials they actually use to print. One group of materials are known as photopolymer resins, which are essentially liquids made of molecules that bind together the polymerise and crosslink when exposed to light, creating a strong, solid material. Resins can be used to accurately print all sorts of shapes and are often used when companies are prototyping products, and some have found uses in dentistry and orthodontic settings. But although versatile, there are concerns about the sustainability of resin-based printing.

This week in Nature, researchers demonstrate a new photopolymer resin that may help overcome this issue, and what’s more, they made it using a chemical from a rather unusual source. I spoke to one of the authors of the new work — Andrew Dove from the University of Birmingham here in the UK, and he laid out some of the issues with current 3D resin technology.

Andrew Dove

So, the resins are often made from petrochemical resources, so they're mined from oil. And also, because what you're doing is you're taking the resin and making a crosslinked polymer, they're very difficult to recycle. And so really, these materials continue to contribute to the plastic-waste challenges and crisis that's going on.

Benjamin Thompson

And so what was the specific problem that you were looking to solve?

Andrew Dove

Fundamentally, it was a challenge that had been my head for a good few years, was how can you make a photocurable resin for 3D printing that then you can do polymerize back to something that you can then directly re-photocure again? So being able to create a closed loop, that was the key challenge we were looking to address. But when we wanted to address that, we wanted to address it in a more sustainable way. So we were thinking around bio-based feedstocks to form the material from the starting point as well.

Benjamin Thompson

And to achieve this then, you turn to a molecule called lipoic acid, which as I've learned is sometimes used as a supplement by bodybuilding enthusiasts. Which on the face of it is perhaps somewhat of an unusual choice. Why have you looked at that in the first instance?

Andrew Dove

I mean it's interesting you say that, we actually buy our lipoic acid from a health supplement company. I tend to buy it myself, so I think they think that I'm some bodybuilder who is getting through a lot of lipoic acid, whereas actually, we're making resins from it. Lipoic acids a really interesting molecule because of how its set up. It has a cyclic disulfide, which you can polymerize to make the backbone of the polymers. And the reason why it's particularly interesting because those disulfide bonds are quite dynamic. So we figured when we understood from the work that had been done previously, that you actually can get that polymerization to go backwards relatively easily. So unlike something like polymerizing an acrylic, where you’re going from a double bond to a single bond, that's very hard to go back up that hill thermodynamically, these closer to equilibrium polymers, are much more easy to access the back reaction from the polymer to the monomer.

Benjamin Thompson

So you can break this bond then and form these kind of straight monomers that you can chain together. And you can also go back the other way as well. But actually, you had to adjust it and your resin is made out of two compounds based originally on lipoic acid.

Andrew Dove

If we just polymerize the lipoic acid, we'd end up with a linear polymer that wouldn't actually be robust enough to display the source of resolution that we get in the 3D printing. So we realised with the acid group that lipoic acid offers, we can do some very simple chemistry to create molecules that have two or three or even more lipoic acid units that can then act as crosslinkers between the chains to make the polymer that results from that much more robust.

Benjamin Thompson

And so you put your lipoic acid-based resin through his paces then using a regular 3D printer. And in one of the tests you printed this little toy tugboat called 3DBenchy and this is a benchmark test model for 3D printers which apparently is quite hard to make.

Andrew Dove

Yeah, that's correct. If you go to a 3D printing conference, you'll probably see everyone showing you their 3D printed Benchy photographs. It's a particularly challenging structure cause of the overhangs and resolution and holes that are within it. I think it’s designed to be aesthetically pleasing, but also technically challenging to print

Benjamin Thompson

And how did your resin perform then printing this little toy tugboat?

Andrew Dove

We get some really beautiful prints out of this resin, the resolution is really, really high. I mean, we didn't just jump straight in and pour a resin into a printer and decide to try and print Benchy. We do a lot of tests before then to optimise the level of lights in particular that we expose for each layer in the print, and a lot of testing to really improve what sort of resolution we can get.

Benjamin Thompson

And how did it stack up to commercially available resins based on petrochemical say?

Andrew Dove

So we're very much at the softer end. But there are commercial resins that we highlight in the paper that are comparable mechanical properties to our materials. But one of the real beauties of this is that we can use a really wide range of different linking groups and alcohol groups to go with that lipoic acid that then infer very different properties. So we can go from things that are really very soft, up to things that are still, on the scale of materials, fairly soft but much stiffer.

Benjamin Thompson

Obviously, you've used this resin to make 3D structures. But the key aim was to make something that can be recycled and reused. What did you see here when you use your lipoic acid-based resin?

Andrew Dove

Yeah, we saw that it worked really quite effectively to be recycled, because we can de-polymerize. So we can undo the polymerization to go back to the starting materials that formed the resin. We can also actually go one stage further back and go back to the chemicals from which you make the resin to fully recycle it properly. It's a really simple process, so we grind the material up and then we quite simply reflux it in a solvent for two hours, and then remove the solvent and you have the resin back it’s about a 91% yield of which the cyclic disulphide recovery is about 96% of that.

Benjamin Thompson

I mean, that's pretty high numbers, but it's not 100%, right? For complete reusability. Do you think it's possible to get there or the laws of chemistry and physics against you at that stage?

Andrew Dove

You're always in equilibrium. So you do have a slight limitation on what you can achieve. That is the major area for improvement for us.

Benjamin Thompson

Of course you wanted to make something that was recyclable. And you did show that it could be used again and again. You printed multiple versions or iterations, I suppose, of the little Benchy tugboat, what properties did they show? Were they comparable? Did they show any degradation or anything like that?

Andrew Dove

The properties of the reprinted polymers are almost identical to the property of the originally printed polymer, it was slightly outstanding in the way but I think it's a demonstration that we've gone really back to the resin. So you know if you're going backwards and forwards of the same chemistry, you should get the same properties out at the end. And we do.

Benjamin Thompson

Obviously, your work is on quite a small scale, right, this is an experimental paper, a proof-of-concept paper. Do you think this work can ultimately be scaled up?

Andrew Dove

Yeah, it's incredibly simple chemistry actually, it's one of the things that really appealed to me when we started working on this was that actually, you can pretty much do this in a bucket — it's very scalable chemistry.

Benjamin Thompson

It's worth noting, though, I suppose that it’s not totally benign at the minute though, some of the solvents and reagents, you know, in your paper aren't necessarily the most kind of friendly to humans, is that something you're investigating or something that can be changed?

Andrew Dove

Absolutely something we actually already spent some time looking at. So we use DMF dimethylformamide for the de-polymerization chemistry. We tried loads of greener solvents, but they're just not as efficient. So we're still looking to optimise that further. But my argument always is, okay, we're using that solvent, what we can recycle that solvent after it's been used. So once it's been removed, we can use it for the next set of de-polymerizations as well.

Benjamin Thompson

So, you've shown evidence then that this system works. What are some of the big questions, maybe big hurdles that need to be overcome do you think sort of moving forward?

Andrew Dove

Cost is always the question. You know, lipoic acid is a relatively inexpensive chemical. But compared to what's currently used, it's relatively expensive. So that's a major cost. I think it's a supply and demand issue. The other thing we're actively looking at is how do we make much stiffer versions of this? The majority of resin sales in this area are on much stiffer materials. And so the challenge for us is how do we achieve that with this type of technology that we can circularise.

Benjamin Thompson

And where do you see this resin being used then if these hurdles can be overcome?

Andrew Dove

We've got a lot of really interesting ideas because of what this unlocks in terms of that resin recyclability. I think anywhere where you have a lot of prototyping going on, if you can get the materials properties, right for that application, I think that would be something that will be very beneficial because you're making prototypes that then can be recycled back into resin and remade into new prototypes. And we're also really interested in looking at these as potential biomaterials as well for medical device-type technologies.

Benjamin Thompson

And are there any other bodybuilding supplements do you think that could ultimately be used as 3D resins? I don’t know.

Andrew Dove

I haven't looked through the bodybuilding catalogue, not being in that area, but perhaps it's inspiration for future research.

Benjamin Thompson

That was Andrew Dove, to read his paper, where you can see a picture of what 3DBenchy looks like, head over to the show notes for a link.

Lizzie Gibney

Coming up, how destroying a building helped researchers design safer structures. Right now, though, it’s the Research Highlights, with Nick Petrić Howe.

<music>

Nick Petrić Howe

If you've tried to find a place to live recently, you will know that competition can be fierce. But to acquire decent real estate, some parrots have even resorted to killing. The green-rumped parrotlet may look unassuming, but in a study spanning 27 years, researchers found that in 9% of the nests, adults were attacking baby birds or eggs. Most of the attacks seem to be driven by housing shortages, as childless parrot couples would kill infants to try and evict their parents. Whereas if one of the parents simply died, it was more likely that a parrot would just move in and adopt the offspring. The researchers concluded that both adoption and infanticide seem to convey a benefit to the parents, as future breeding was limited by the available homes. Relocate yourself somewhere peaceful and read that research in full in the Proceedings of the National Academy of Sciences of the United States of America.

<music>

Imagine a cauliflower. Great, now ask yourself, why does that look like that? Well, according to a new study, it's the result of 2,500 years of domestication, and a smattering of genetic changes. Researchers analysed 971 genomes of cauliflowers and related plants like broccoli to understand the curly plants evolutionary history. They found three genes that seem to underlie how it got those tight curls and whirls that make up a cauliflower heads. These changes were encouraged by the humans who grew, what was them broccoli, to give us the cauliflower we know and perhaps love today. The researchers hope that by understanding these genetic changes, it could help us grow better cauliflower in the future. Enjoy that research roasted with a bit of cheese, in Nature Genetics.

<music>

Lizzie Gibney

Next up reporter Dan Fox has a story about how to make buildings more resilient.

Nirvan Makoond

Not a lot of sleep. So we were all quite worried. Of course, we expected a certain result and we were well prepared for it we’d performed a lot of simulations. But there are a lot of uncertainties in simulation which is why we need to test the building. Yeah, so not a lot of sleep and very anxious.

Dan Fox

This is Nirvan Makoond from the ICITECH institute at the Universitat Politècnica de València in Spain, discussing how he was feeling the night before a key experiment. Here's what happens next.

<collapsing sounds>

That was the sound of that €120,000 experiment partially collapsing. Fortunately, that's exactly what Nirvan and his colleagues wanted to happen.

Nirvan Makoond

Very relieved, and the agreement between what we observed during the test and our predictive simulations was surprisingly good. And it surprised even us. That's what I wanted to say, I think it surprised even us. Very relieved and very happy.

Dan Fox

This team of researchers are interested in why buildings collapse, and how those disasters could be mitigated. And this experiment published this week in Nature has just demonstrated the success of their new idea, a controlled collapse that prevents the whole building falling down. Curious about how you stop a building from collapsing once it started, I called Nirvan up and asked him to tell me a bit more about why buildings collapse in the first place.

Nirvan Makoond

Many different kinds of threats can cause a building to collapse. This can be extreme weather events such as floods, landslides, earthquakes, hurricanes, it can be explosions. It can even be ageing, or due to deterioration, construction errors, design errors. But often, a building collapse starts in a particular part of the building and then propagates to other parts of it.

Dan Fox

So what's the scale of this problem?

Nirvan Makoond

I would say very big, because whenever there is a building collapse, actually the whole society around it is affected, lives are lost, the costs are great. An important aspect is that in the current global context, this is getting worse due to climate change. We are seeing more frequent and intense extreme weather events and also there is– there are rising geopolitical tensions.

Dan Fox

So what are the existing solutions to this problem?

Nirvan Makoond

All measures included in codes and the current, let's say best practices in the field of structural engineering, basically all focus on preventing any form of collapse after an initial failure. So this means for example, if a structural component is lost, the current measures focus on ensuring the building is well connected as a whole. So the loads it was supporting can be redistributed to other parts of the structure. So it focuses on preventing the initiation of collapse.

Dan Fox

So what are the limitations of these existing solutions? And how would you do things differently?

Nirvan Makoond

It has been shown to be effective for when an initial failure is very small, so very localised. However, in most cases of disastrous building collapses, we see that actually, this initial failure can be quite large. And it's something we cannot design against to completely prevent collapse. So our approach, we try to arrest a collapse once it has already initiated. So we look at the problem of limiting the extent of a collapse, rather than completely preventing its initiation. So in practice, it works by controlling the sequence of components that fail during the collapse process. What this means is, for instance, ensuring that connections or beams fail before columns fail, which are basically the main load bearing elements of the system. A great metaphor, I think, for– for our approach is the way that lizards shed their tails to escape predators. For certain normal functions, a lizard’s tail is perfectly attached to its body. However, when a predator has grabbed the tail of the lizard, the lizard can activate a particular movement, which basically causes the tail to break off. And similarly, when the building is operating normally, we ensure full connectivity and allow loads to be distributed so that the building can work as it is intended to. However, when a collapse has already started, what we ensure is that the failure is controlled, and that the most important parts of the building further away from the collapse remain intact.

Dan Fox

So you started by testing this idea in computer models, what did those teach you?

Nirvan Makoond

The computational models really taught us that connectivity is actually causing more collapse propagation. So basically, it helped us really look at how the loads are transferred during the collapse, and how different building design choices cause more or less propagation. So this, this idea that a building being well connected together can also cause a greater extent of propagation. We have thought about the others had thought about it, demolition experts hadn't seen it. But really, it hadn't been studied, so the computational simulations allowed us to understand this phenomena better, but also to prove that hypothesis.

Dan Fox

And then you built a physical model. But unlike most research, you build a full-size two-story building. So why was it important to have a test at that enormous scale?

Nirvan Makoond

First of all, because the nature of that phenomenon itself, to understand it, scaling it down introduces a lot of uncertainties in the analysis. And for example, we cannot scale the effect of gravity, but scaling down a structure really influences the accelerations and therefore the outcomes of that test are not reliable for that highly dynamic and complex phenomenon. Also, the test also serves as a form of proof of concept. So the construction sector is very conservative, and is very risk averse for good reason no, because of the consequences where the building collapse. However, this also means that introducing changes in the way things are done really requires convincing a lot of people and basically scaled down tests are not convincing enough.

Dan Fox

Could you talk me through what happened on the day of the full-scale experiment when you collapsed your model?

Nirvan Makoond

Our test had two phases. The first one was to ensure that despite our design modifications, we were still able to prevent a collapse from initiating after a small initial failure. And so the in the first phase, we actually removed two columns, and no collapse opens, really ensuring that our building complies with requirements covered in codes. And then in the second phase, there was a sudden removal of the third column that triggered the collapse. And once this collapse started in the initial phases, as the collapsing parts were falling down and the building was still very well connected, everything was being pulled towards where the initial failure was starting. And then basically, our hierarchy or sequence of failures that was planned intended in the design activated. And this cause basically just part of the structure to collapse down while the other was able to remain upright.

Dan Fox

Did you learn anything from the from the scale model that you hadn't predicted?

Nirvan Makoond

Yes. The real test definitely allowed us to assess reliably the level of damage in the impact part of the structure. And that is quite important for how you could use that remaining part or if it is still safe to perform evacuation and rescue operations there. So the real test also allowed us to evaluate that with more certainty.

Dan Fox

Given the success of the test, how easy would it be for engineers to put this type of design into practice in buildings in the real world?

Nirvan Makoond

So, there are two aspects of this. There’s the design aspect and there’s the constructability aspect. Actually, to implement this in a real construction project. It's all low-tech solutions, all construction details that are currently used, there's no fancy devices or anything special you need, which means it has a high potential for impact, it can be implemented really easily at relatively low cost. However, the design aspect, in its current stage of development, it requires a lot of simulations, and quite high-fidelity simulations, which practitioners are not accustomed to. So there's quite a bit of work still to be done on developing simplified methods to be able to implement this in practice, from a design perspective. But implementing it in an actual construction project, once the design has been made, is actually very straightforward.

Dan Fox

So what's next for this research? Can this be applied to different types of buildings where you're going to take it from here?

Nirvan Makoond

The philosophy itself is applicable in principle to all types of building. Honestly how the implementation is completely different. So we are going to implement and even test its effectiveness in other building types. That's really the next line of what we are already working on. Also, as I previously mentioned, there's a lot of work that needs to be done to transform this fundamental research into practical solutions, no developing simplified methods, convincing the board which we hope the test will do. This is more or less our future vision for this work. And we really hope to, to really transform this into practical solutions to really have an impact on society, improving the resilience of our buildings.

<collapsing sounds>

Lizzie Gibney

That was Nirvan Makoond from the ICITECH institute at the Universitat Politècnica de València. If you want to see that partial collapse in action, we also have a video on the Nature YouTube channel which we will link to in the show notes.

Benjamin Thompson

Finally on the show, it’s time for the Briefing Chat, where we discuss a couple of articles from the Nature Briefing. Lizzie, what have you been reading this week?

Lizzie Gibney

So this is about an algorithm that scientists have been using to discover asteroids. So they found using this algorithm twenty seven and a half thousand asteroids and that is as many as were discovered by all the telescopes put together last year. So that's a hell of a lot of asteroids.

Benjamin

Well presumably then researchers had no idea that these existed.

Lizzie Gibney

That's right. So these are completely newly discovered asteroids. So this is a story in The New York Times, and the research came out of the Asteroid Institute — very aptly named. Now how people usually look for asteroids is by trying to spot them in the same picture taken over the same night, you look at the same spot of the sky, and you see these little tiny blips that move when the backdrop of stars stay the same. And that's how you know that you've seen an asteroid. But of course, that means having lots of pictures, this can be a more laborious process. What they did here was essentially harness the fact that asteroids photobomb other shots of the sky already. So they looked in over 400,000 images that already existed in the archives of NOIRlab — a research laboratory in the States. So they found 1.7 billion little dots that came up in just one image. And they use this algorithm to project possible orbits that would connect those dots. And so through this kind of quite heavy computation, they were able to figure out that those dots were actually asteroids that just cropped up in these other images. And they could connect them together to figure out their orbits as well. So that's how they were able to find so many in just existing data that didn't need anything that was new.

Benjamin Thompson

So rather than sort of struggling to get time on a telescope to take more pictures, then you can actually go back at this and say, this catalogue that exists and figure out what these might actually be.

Lizzie Gibney

Exactly. And so it takes quite a lot of computational power. So this was in collaboration with Google Cloud. And it was eight and a half million equivalent of CPU hours. So that's about five weeks, even distributed across lots of different computers. So that's, you know, that's quite hefty amount of computing time. But the benefits, as you say are you don't need to go out there and get new time on a telescope. And it begs the question of what else might already be out there in all of these images that we've got, they've been taken over decades, that could be new discoveries?

Benjamin Thompson

And I'm sorry to bring it down this avenue, but if researchers didn't know that these were even here, like do they pose a threat to us?

Lizzie Gibney

So they found around 150 that are categorised as near-Earth asteroids. So they're orbits come near Earth's orbit, none of them seem to be a threat that we are aware of at the moment. So I think we can spin it actually as a positive in that, we're able to find these new asteroids that some of them come near Earth using this technique. And actually, using this technique on other efforts to find near-Earth asteroids, is probably going to be quite beneficial in the future, because this actually was trying to find asteroids that are further out in the main asteroid belt and further out in the Solar System. And but it found these serendipitously so if it actually goes out and tries to find near-Earth asteroids, we could really boost the number that we find. So there's a telescope in Chile being built by scientists in the States called the Vera C. Rubin Observatory. And one of their mandates that's come from Congress is that they need to find 90% of near-Earth asteroids that are bigger than a certain size, I think about 150 metres across. They didn't think they were gonna be able to hit that before, because they had to take two images, to track each asteroid. So they’d potentially see an asteroid and then they need to take another one to see the movement. They no longer need to do that. So this halves the amount of time it takes, this doubles the amount of space that they can look at. So the scientists, they are now quite confident that actually they are hopefully going to get near that 90% target by using this this algorithm.

Benjamin Thompson

That's kind of neat. I mean, do you think this is a technique that could be applied to other things floating around in space?

Lizzie Gibney

Yeah, I don't see why not. I mean, the very fact that this was found in archived images, and that it's essentially just apply an algorithm on top of those, that suggests that there might be a lot more that we can find, just by going through all of this old data. So I think there's a lot of potential there.

Benjamin Thompson

Well, it's super interesting that, you know, there's a lot we don't know about what's in our near astronomical neighbourhood. But speaking of unknowns, let's shift to my story today. But let's look inwards somewhat. So we know that space is fantastically complicated. But a lot of people would say that the human brain is perhaps the most complicated thing in the Universe. And there's a paper that came out in Science and a news article about it in Nature, that describes a way to map a very, very small part of the human brain in astonishing detail. And it's revealed patterns of communication between neurons and all sorts of other things as well.

Lizzie Gibney

And is there a kind of image of this?

Benjamin Thompson

Oh, there are multiple images of this. And this kind of system, which we'll talk about, is available for researchers to look at online. So if you head over to the show notes, you will find a link to the story and where that can be done. And what's happened here is that the researchers took a brain fragment from a 45-year-old woman who had undergone surgery to treat her epilepsy. And this came from the cortex, okay, so part of the brain involved in you know, problem solving and processing sensory signals that sort of thing, right. And it's a cubic millimetre of brain. And within it, I mean, some of these numbers are staggering, within this cubic millimetre, there's about 57,000 cells of different types, 150 million synapses— so these are, you know, connections between neurons— and hundreds of millimetres of superfine blood vessels. And all of this kind of put together is apparently 1.4 petabytes of data.

Lizzie Gibney

And so this image that I've just brought it up now, it is absolutely stunning. So we've got all these colours, which I assume are false colours. It's got these wavy grasslike tadpole patterns, I mean, that that's an incredible shot.

Benjamin Thompson

Oh, I mean, it's absolutely fantastic. But you made the point there, that's quite a small amount, one cubic millimetre is about a millionth of the human brain. And you're right false colours as well. So what's happened here is researchers, they took this tiny fragment of brain, they preserved it and stained it with heavy metals, so you can see the cells easier. And then they cut it into about 5,000 slices, right. Each about 34 nanometres thick and then they imaged these with an electron microscope. And then, as is often the way, they're turned to an AI, right, to try and stitch these together, right. I will say it's not perfect, potentially, because this stitching might not have gone, you know, 100% according to plan. They've manually checked a proportion of it, but they're hoping that other researchers can come in and correct any errors that may have been introduced while the map was being made.

Lizzie Gibney

And so once they've got this image, they've got this map that they've stitched together, what have they actually seen in it?

Benjamin Thompson

I mean, they found some things, some of which have never been seen before. Okay, Lizzie. So, some of the things are unconventional neurons that are described in the story, and they make up to 50 connections with each other, right. And so this is a far and away more than potentially the couple you'd normally find, as the researchers in the article are quoted as saying. But they also found some other stuff as well, pairs of neurons that are almost perfect mirror images of each other, cells that wrap around themselves to form knots. And the team plan to produce maps of other parts of the brain from other volunteers as well. But I think we're a little bit away away from putting it all together. One of the researchers behind the project is quoted as saying they reckon a map of the entire brain is unlikely in the next few decades. I mean, my goodness, the stuff they found already, I mean, what more is to be discovered I guess?

Lizzie Gibney

Well, we've got to figure out what it all does.

Benjamin Thompson

Well, I think that's a reasonable thing to say, because of course, if you want to know how it works, you need to know what it's made of and then you can kind of move down the chain, I suppose. And figuring how the cortex works in particular could, you know, in the long-term help in the treatment of you know, psychiatric and neurodegenerative diseases, potentially. But say, it's the first step in the right direction, but it's all available online for people to have a look at.

Lizzie Gibney

Amazing, I might make that my screensaver for a little while. Well, thank you Ben. And listeners for more on those stories and for where you can sign up to the Nature Briefing to get more like them, check out the show notes for some links.

Benjamin Thompson

And that’s all for this week, as always keep in touch with us on X, we’re @NaturePodcast, or send an email to podcast@nature.com. I’m Benjamin Thompson.

Lizzie Gibney

I'm Lizzie Gibney. Thanks for listening.