NATURE PODCAST

Podcast: Improving battery charging, and harnessing energy from the air

Listen to the latest science news, with Nick Howe and Benjamin Thompson.

This week, machine learning helps batteries charge faster, and using bacterial nanowires to generate electricity from thin air.

In this episode:

00:46 Better battery charging

A machine learning algorithm reveals how to quickly charge batteries without damaging them. Research Article: Attia et al.

07:12 Research Highlights

Deciphering mouse chit-chat, and strengthening soy glue. Research Highlight: The ‘silent’ language of mice is decoded at last; Research Article: Gu et al.

09:21 Harnessing humidity

A new device produces electricity using water in the air. Research Article: Liu et al.

16:30 News Chat

Coronavirus outbreak updates, the global push to conserve biodiversity, and radar reveals secrets in an ancient Egyptian tomb. News: Coronavirus: latest news on spreading infection; News: China takes centre stage in global biodiversity push; News: Is this Nefertiti’s tomb? Radar clues reignite debate over hidden chambers

Never miss an episode: Subscribe to the Nature Podcast on Apple Podcasts, Google Podcasts, Spotify or your favourite podcast app. Head here for the Nature Podcast RSS feed.

Nature Briefing

An essential round-up of science news, opinion and analysis, delivered to your inbox every weekday.

Transcript

This week, machine learning helps batteries charge faster, and using bacterial nanowires to generate electricity from thin air.

Host: Benjamin Thompson

Welcome back to the Nature Podcast. This week is an electrifying episode, as we’re talking about fast battery charging…

Host: Nick Howe

And harvesting energy from the air. I’m Nick Howe.

Host: Benjamin Thompson

And I’m Benjamin Thompson.

[Jingle]

Interviewer: Benjamin Thompson

Now that I’ve got a young family, a few people have asked me, ‘Are you going to buy a car?’ I’m a big fan of public transport, but when we’re in the middle of a long trip, carrying loads of stuff, automobile ownership is sometimes tempting. Should I decide to go down that route, I would most likely look to buy an electric car, but these do seem to have some issues. The phrase ‘range anxiety’ is often used in association with electric vehicles, and this is the worry that you’ll run out of power and end up stranded during a long journey. There are also concerns that electric cars can take too long to charge when you’re out on the road. To charge a battery – put very simply – you need to pump electricity into it. To speed up the charging, you can increase the level of power you’re putting in. But there is a catch – for some sorts of batteries, such as the lithium-ion batteries found in many electric vehicles, charging too fast can result in permanent damage to the battery. So, what is the optimum speed to charge a battery? That’s what Will Chueh from Stanford University in the US is trying to find out.

Interviewee: Will Chueh

So, let me provide the analogy of filling a bucket of water. So, say your objective is to fill the bucket in 10 minutes to 80% full. There are infinitely many ways to do that. For example, you can fill it at a constant rate so that it tops off at 80% in 10 minutes. You can also fill it very slowly initially and fill it very quickly later, and also the opposite, you can fill very quickly initially and fill slower later – all sums up to 10 minutes to 80%. So, the way you deliver the water to the bucket is exactly what we’re trying to optimise for fast charging. How do you deliver 80% worth of electrons into the battery in 10 minutes?

Interviewer: Benjamin Thompson

In this case, rather than looking at an infinite number of methods, Will concentrated on a little over 200 different protocols tested on small cylindrical rechargeable batteries a bit larger than the AA batteries you might use at home. Will and his colleagues took a three-step approach. First, they repeatedly charged and discharged a batch of 48 batteries, each with a randomly assigned charging protocol. Second, they assessed the health of the batteries, but to do this, they needed a trick. The batteries in their study have a lifetime of about 1,000 recharging cycles. To fully assess their health, Will would have had to cycle them 1,000 times which would have taken a while. Instead, he took a bit of a shortcut.

Interviewee: Will Chueh

What we have done here is to use machine learning to basically recognise early signs of battery failure so we don’t have to charge it 1,000 times. We just have to charge it 100 times. So, basically, what we’re doing here is to cut the time by a factor of ten, just by having not to fully cycle the battery to failure, but instead we looked for early signs of failure and we make predictions based on that.

Interviewer: Benjamin Thompson

And this is the final step. They fed these early signs of failure into a model which could then identify the protocols that best balanced fast charging while minimizing damage. The protocols that showed most promise could then be put up for another round of testing with a fresh set of batteries.

Interviewee: Will Chueh

So, what the computer is trying to do here is to identify the patterns in the performance of these protocols and trying to make decisions on what to look at next. So, you can basically imagine our program as a decision-maker and the decision-maker is trying to say, okay, I tested this charging protocol, it sucks so let’s not do it again. Or we do this protocol and it works really well, and we say, okay, let’s do it again or look at its neighbouring protocols. So, essentially, it’s a way to try to hone in on the best charging protocol by learning from the behaviour of all 200 possible combinations that we’re looking at.

Interviewer: Benjamin Thompson

By repeatedly going through this three-step sequence, the team were able to quickly whittle down the options to find the best way of charging these batteries. The method that came out on top rather surprised Will.

Interviewee: Will Chueh

Typically, what is assumed is the best way to charge the battery is you start by putting in the energy really quickly, then you slow it down and you taper it off. This is what it is done in today’s electric vehicles. When we went through our platform here, we actually found out it’s better to actually keep it relatively flat, meaning that it’s better to inject the energy in relatively constantly. So, it was rather unexpected, I think, what we found, but because it was done in a statistically meaningful manner, we’re quite confident that this was a viable outcome.

Interviewer: Benjamin Thompson

In this case, it seems that slow and steady really does win the race, but what does this mean for batteries in general? Can we expect this result to change the way that electric vehicles are charged? Well, perhaps not yet. Will emphasised that this result might only hold true for this one type of battery with its particular form factor and chemistry. There are a wealth of rechargeable battery types in addition to the specific lithium-ion battery he tested, each with their own strengths and weaknesses. He says it’s the method rather than the result which is key here, and hopes his testing approach – the three-step machine learning platform – can be used to develop better batteries in future.

Interviewee: Will Chueh

What we have realised here is not just a better charging protocol, it is rather a very generalisable platform which can be used to optimise many aspects of the battery, one of which we demonstrated, which was picking the best way to charge a battery in ten minutes, but there are really countless other possibilities in which this platform can benefit. For example, what is the best chemistry? What is the best shape, the form factors? There are many choices and decisions to be made for a battery and all of them take time, and what this platform, I think, can help with is to really save time and therefore get to the answer faster.

Interviewer: Benjamin Thompson

That was Will Chueh from Stanford University. You can read his paper over at nature.com.

Host: Nick Howe

Later on, we’ll have further updates on the coronavirus outbreak, and we’ll be finding out how radar is uncovering ancient secrets in Egypt – that’s coming up in the News Chat. Before that, it’s time for the Research Highlights, read this week by Dan Fox.

[Jingle]

Dan Fox

It might not be as tuneful as Walt Disney imagined, but scientists have deciphered the ultrasonic singing mice use to communicate with each other. Mice emit a variety of squeaks to communicate, but researchers have previously struggled to interpret the high-pitched sounds. Now, a team have used machine learning software called DeepSqueak to analyse data from over 100,000 mouse vocalisations. The data were collected in a specialised recording chamber that let the researchers track which mice were squeaking and what all the mice were doing at the time of each squeak. Male mice were found to make distinct sounds depending on whether they were fighting, fleeing or chasing females, and the researchers were able to see how mice reacted to each other’s chirps, changing their behaviour in response. Squeak a look at that paper in full at Nature Neuroscience.

[Jingle]

Dan Fox

Depending on your diet, soy may already have replaced the meat and milk on your table, but soon it could be replacing the glue holding your table together. Common glues used in the assemblies of building materials like plywood or particle board are made from petroleum, but the earliest known wood glues were made of proteins, such as soy flour. If soy adhesives are to make a comeback, they will need to become strong enough to meet building standards. That’s what a team of researchers from Beijing Forestry University have been investigating. They found that combining soy protein with borate salt and a toughening agent in water created a strong and environmentally friendly adhesive. The boron-based glue is three times stronger than plain soy protein and also has anti-microbial and fire retardant properties. Keep your eyes glued to that research at Green Chemistry.

[Jingle]

Host: Benjamin Thompson

Next up on this week’s podcast, if I was worrying about running out of charge in my electric car earlier, Nick’s been finding out about a technology that may mean I always have charge, using a device that draws energy from the air. Take it away, Nick.

Interviewer: Nick Howe

The air is full of energy. That’s what lightning is – the release of some this energy as charged particles of water bounce off each other. So, there’s energy to be had, but how can we get at it? After all, lightning is not the easiest thing in the world to bottle. The thing is, we don’t even need storms. The charged particles of water in the air themselves can be tapped into and used to generate a current. Several researchers have taken advantage of this, using specially designed super-thin sheets of microscopic nanowires, pulling electricity from seemingly thin air. But so far, these technologies have only demonstrated brief bursts of energy and often have to be near a lot of water. This week in Nature though, a team from the University of Massachusetts, Amherst have shown a similar technology but without these problems. Their solution? Instead of designing and making nanowires out of things like silicon, they took them from a bacterium called Geobacter sulfurreducens. This species naturally grows protein nanowires, as filaments or ‘pili’ on the outside of the cell.

Interviewee: Jun Yao

These type of bacteria, they use these to serve as a nanoscale cable to conduct electrons.

Interviewer: Nick Howe

This is Jun Yao, one of the authors of the new paper in Nature. He’s actually never worked on bacterial nanowires before. Instead, he’s an expert on inorganic nanowires. But when he found out about these structures produced by Geobacter, he got excited at the possibilities. Initially, Jun just wanted to see if these bacterial wires could function similarly to inorganic nanowires.

Interviewee: Jun Yao

So, I simply want to replicate a similar concept by making the simplest sensor, which usually is a humidity or moisture sensor.

Interviewer: Nick Howe

To make his moisture sensor, Jun removed the nanowires from bacteria using a blender and purified them. He then produced a super-thin sheet of bundles of randomly arranged protein nanowires that was less than a tenth of the thickness of a human hair. He could then run a charge through it and see if it could act as a moisture sensor.

Interviewee: Jun Yao

Then accidentally, the student who was working on the sensor device saw that even without supplying the device with an external power, he could still get the electrical current to signal on the device.

Interviewer: Nick Howe

Initially perplexed, Jun set to work replicating and re-replicating these results. Satisfied it wasn’t a blip, he then started to work out where this electricity was coming from. It had to be coming from something in the air, but what? Jun’s search eventually led him to the most likely culprit – water. In fact, the level of charge generated correlated almost exactly with the level of humidity in the room.

Interviewee: Jun Yao

I was excited because at least I knew that this was real.

Interviewer: Nick Howe

But how much electricity was this tiny device actually harvesting?

Interviewee: Jun Yao

I think we can produce several microwatts of power across 1 square centimetre size, and that’s actually a small amount of power, which means it can essentially only power a tiny LED light.

Interviewer: Nick Howe

Now, that may not sound like much, but Jun thinks that’s only a test case for what this technology can really do. Humidity can pass through the layers of nanowires, so there’s a possibility that you can stack a lot of them together and increase the power output.

Interviewee: Jun Yao

We can estimate that if we find a clever engineering strategy to stack them in a 3D configuration, we can actually get the power density larger than 1 kilowatt per cubic metre.

Interviewer: Nick Howe

One kilowatt is about enough energy to power ten laptops, so it’s a little bit more substantial than powering a small LED. Of course, the amount of power that can be harnessed is dependent on the amount of humidity present, but Jun thinks even in desert environments, this device would work and still produce a small amount of power. Julea Butt is a scientist who studies electron transfer in proteins, who wasn’t associated with this latest work. She was intrigued by the new research.

Interviewee: Julea Butt

I think using humidity as a source of energy and electricity is a very interesting idea, and this demonstrates one way that that might be possible. So, they’re clearly, on a small scale, displaying the kind of properties that we’re used to seeing on a larger scale in the approaches that we’ve got at the moment to power things, so that all looks very promising.

Interviewer: Nick Howe

Julea did have concerns, though, about the scalability. If you have to collect the nanowires from the bacteria, will that not be a bottleneck? Well, Derek Lovley, the resident microbiologist on the paper, may have developed a solution for that.

Interviewee: Derek Lovley

So, what we’ve done recently is we’ve taken the gene out of Geobacter and developed a strain of E. coli, which is a very common microbe used for mass production of commodities that’s able to also produce these wires, so we can make them in large quantities. So, I really feel that we can easily make kilogram quantities of the wires with this new strain that we have.

Interviewer: Nick Howe

There are other concerns too, such as how well these proteins will endure in the environment. But Jun thinks they’re actually pretty persistent.

Interviewee: Jun Yao

In fact, we made a device and we put it in the ambient environment for ten months and then we go back to test again and produce the same electricity. It’s bio-derived, it’s green, and yet it seems to be very stable.

Interviewer: Nick Howe

Jun is excited for the future of this research, as he believes this is a sustainable source of power. At the moment though, he’s trying to get to the bottom of exactly how this system works, but in the small scale, he thinks the bacterially grown nanowires could help power things like your phone or your smartwatch. He’s got bigger ambitions, though.

Interviewee: Jun Yao

This thing can scale, why not that you simply can paint, let’s say you paint your entire wall in your home or anywhere, so that you can collect much more power to power lighting or even possibly other electronic devices in your home, and that’s what I see.

Interviewer: Nick Howe

That was Jun-Yao. His colleague Derek Lovley also featured, and you can read their paper over nature.com. You also heard from Julea Butt from the University of East Anglia here in the UK.

Interviewer: Benjamin Thompson

Finally on the show, it is, of course, time for the News Chat, and I’m joined in the studio by Ehsan Masood, Nature’s Africa and Middle East Bureau Chief. Ehsan, thank you so much for joining me.

Interviewee: Ehsan Masood

It’s a pleasure. Thanks very much for having me.

Interviewer: Benjamin Thompson

So, a few stories to cover today then, and I think we should start with the ongoing coronavirus COVID-19 outbreak that is centred, of course, in China. Listeners, as always, this is an ongoing and developing story, so for the latest updates, I recommend you go over to nature.com/news. Ehsan, what’s been happening since the show went live last week?

Interviewee: Ehsan Masood

Some of the things have changed is that we’re moving into more therapy territory now, and we’ve discovered that there are more than 80 clinical trials that are now being tested in China for various kinds of treatments, and that’s interesting and promising. Of course, these things are not going to emerge straight away, and these won’t necessarily be the answer to slowing down the rate of infection, but it’s definitely a promising thing.

Interviewer: Benjamin Thompson

And the WHO are getting involved in these trials.

Interviewee: Ehsan Masood

That’s right, yes. We’ve heard that Soumya Swaminathan, the chief scientist at the World Health Organisation, says that their teams have been taking stock of the many trials and their idea is to draw plans for some sort of clinical trials protocol.

Interviewer: Benjamin Thompson

And what will this protocol do then?

Interviewee: Ehsan Masood

So, the idea is to have some kind of standardisation. I think that’s what the World Health Organisation is most interested in. So, for example, are the trials designed with some sorts of common parameters in mind, things like control groups, randomisation, measurement of clinical outcomes. These are the kinds of things they’re looking at.

Interviewer: Benjamin Thompson

What sort of things then are being tested at the moment?

Interviewee: Ehsan Masood

Well, it’s a little bit of a shopping list, to be honest. So, there’s a couple of HIV drugs that are used to blog enzymes that viruses need to replicate. Then China has launched a couple of trials to test chloroquine, which is quite a well-known anti-malarial drug. There are some stem-cell trials, and then there’s also some trials of traditional Chinese medical therapies, and the WHO’s chief scientist has made it quite clear that the organisation’s involvement, particularly in this aspect, is so that herbal remedies can be evaluated with the same level of vigour that’s expected of pharmaceutical testing.

Interviewer: Benjamin Thompson

Well, it could be a while then, I guess, before these trials come to fruition, but finding treatment seems to be a super important thing because case numbers have been going up and in some cases, have increased quite dramatically since last week.

Interviewee: Ehsan Masood

They did increase dramatically in the Hubei Province, which is, in a sense, the epicentre of the outbreak because the authorities there changed the way in which they’re making their calculations. Right now, we know that we’re up to about almost 1,900 deaths and about 73,000 cases of infection.

Interviewer: Benjamin Thompson

And how were case numbers being calculated then in Hubei which changed things?

Interviewee: Ehsan Masood

So, up to now, diagnosis was happening on the basis of lab results. What’s now happened is that physicians are being able to diagnose cases on the basis of chest images, and that’s certainly bumped up the numbers in that one instance which happened on 14 February.

Interviewer: Benjamin Thompson

And how has that helped physicians in the area?

Interviewee: Ehsan Masood

Well, it’s definitely in response to pleas from clinicians who say they have been overwhelmed by patients with respiratory diseases and they don’t really have time to wait for the lab results, so that’s really been helpful for them in that sense. We also spoke to Michael Mina, who’s an infectious disease epidemiologist and immunologist based at the Harvard School of Public Health in Boston, and he told our reporter that triaging based on symptomatic evaluation and a physical exam is basically the bedrock of hospital-based and clinical triage.

Interviewer: Benjamin Thompson

Well, you talked about some case numbers there, Ehsan, but it seems there is one case in particular in a different part of the world that hasn’t been reported before.

Interviewee: Ehsan Masood

That’s right, yes, the African continent got its first case in Egypt. Up until now, the WHO had a massive push to provide diagnostic testing across 14 different African countries where there were real concerns, particularly around returning Chinese workers who work on various projects in Africa going back and they might be a source of infection. That kind of expected increase hasn’t happened but yes, we’ve got the first case in Egypt.

Interviewer: Benjamin Thompson

So, the virus continues to spread then, and it’s affecting a lot of people. Nature has actually been running a survey of researchers affected by the outbreak and, listeners, head over to nature.com/news to read their stories, but it’s not just people. Events are being affected a well and Ehsan, that’s part of our next story.

Interviewee: Ehsan Masood

It is, yes. There was meant to be an international conference on biodiversity. It was going to happen in Kunming in China, and it had to be relocated to Rome because of coronavirus.

Interviewer: Benjamin Thompson

Yeah, and this is due to happen next week then, and what is this meeting all about?

Interviewee: Ehsan Masood

This is a meeting that’s supposed to agree or begin the discussion to agree on a whole new set of target indicators for biodiversity laws.

Interviewer: Benjamin Thompson

It’s something we’ve covered on the show before. Some of the levels of these losses are staggering.

Interviewee: Ehsan Masood

They are, I mean the headline figure that’s provided by the United Nations Biodiversity Scientific Advisory Body is up to 1 million species are threatened because of human activities.

Interviewer: Benjamin Thompson

What’s due to be decided at this meeting to try and mitigate losses in future?

Interviewee: Ehsan Masood

There has been a ten-year plan for reducing the rate of biodiversity loss. It was agreed in 2010 and the deadline to begin to halt or at least reduce the rate of loss was meant to be the end of this year, the end of 2020, and we’re not going to meet hardly any of the target indicators and this is the problem. And of course, this is a very important year, both for climate and for biodiversity, which is why international organisations and government representatives and scientists are going to start to meet and to figure out well, what’s the next set of targets that we need to start agreeing on and, quite frankly, targets that are actually achievable.

Interviewer: Benjamin Thompson

Well, this meeting, as you say, Ehsan, was due to be held in China, and a lot of eyes were on the country and their response to biodiversity. Why was this, specifically?

Interviewee: Ehsan Masood

For a couple of reasons. So, China is, as so many countries are, a highly ecologically diverse environment, but at the same time, of course, China is rapidly industrialising in its own country and it’s also helping a lot of other countries industrialise too. And we know that a lot of the zones where industrialisation is happening, both in China and outside of China, are areas which are quite ecologically fragile and so the Chinese leadership is extremely conscious of this because its own scientists are telling it that this is happening, and it’s trying to find a way around this. How do you industrialise without completely obliterating your ecologically fragile areas?

Interviewer: Benjamin Thompson

If a new set of targets are to be decided then at this meeting, if they’re being wildly missed in the last ten years, what’s the feeling from people who are there about these being stuck to this time around?

Interviewee: Ehsan Masood

First of all, there’s been no mandatory reporting of progress, so every country that’s a member state of the UN has to issue what’s called a national biodiversity action plan. It’s a sort of statement of intent and principals – here’s what I’m going to do – but it’s not a reporting tool, and so one thing that’s really important going forwards is not only a state of ambition, but actually every year or every few years, what exactly is progress looking like against that ambition, and that’s going to be a really important aspect of these discussions.

Interviewer: Benjamin Thompson

Well, we’ll have to keep an eye out then, Ehsan, and see what is decided at that meeting. But for the meantime, let’s move on to our third story, and let’s head over to Egypt and a story about Tutankhamun’s burial chamber.

Interviewee: Ehsan Masood

It is, yes, let’s head to the Valley of the Kings in Luxor. This is definitely one of my favourite stories of the week – the possible discovery of some hidden rooms around Tutankhamun’s burial chamber. And to be honest, we don’t really know who might be buried in those rooms, what might be in those rooms, but there’s pretty much a lot of speculation.

Interviewer: Benjamin Thompson

Well, let’s speculate away then. What sort of thing are people thinking might be hidden in these rooms?

Interviewee: Ehsan Masood

Well, the big one, if it can be eventually confirmed, is that this could be the lost tomb of Nefertiti, the mysterious queen whose remains have never been found. It’s the holy grail, and is this new find potentially pointing us to the tomb of Nefertiti? That’s the big question.

Interviewer: Benjamin Thompson

Well, what do researchers have to do to answer that question then, and how are they looking in this space in the first instance?

Interviewee: Ehsan Masood

Because Tutankhamun’s burial chamber is quite sealed – there’s only one entrance to go in to and these hidden rooms that have been discovered are leading off of his burial chamber – that kind of rules out using more invasive methods, and so research teams are using more non-invasive methods, and they’re using, particularly, radar.

Interviewer: Benjamin Thompson

Right, and how does one go about finding what may be a room to actually seeing what’s inside it?

Interviewee: Ehsan Masood

That’s really difficult. So, the first thing that has to happen is someone has to replicate this find. So, at the moment it’s only one team. So, of the teams that have gone before, there’s a reasonable amount of consensus that there are some hidden rooms and there’s a degree of consensus as to where those rooms might be. This team has found something really quite large and new and so before anything else happens, someone has to go in there and replicate that. It could be the team themselves repeat their experiment and do it again. It could be that someone else comes in and does it.

Interviewer: Benjamin Thompson

Well, I guess there must be some excitement in the field about this potential find. Is everyone on board with what this may be?

Interviewee: Ehsan Masood

There is excitement, but also not everyone is on board. This is, in part, a story of two quite strong-willed individuals who have different ways of looking at the world and of looking at the world of Egyptology. So, you’ve got one group which is the group that’s behind this find and they put a lot of store in radar and in non-invasive methods and you’ve got another group of Egyptologists who think that so far, whatever we’ve found of value in Egyptology has been found using more traditional methods, and so at some point, someone will have to go in and dig.

Interviewer: Benjamin Thompson

Well, finally, Ehsan, what happens next then? What’s the sort of timeline for this?

Interviewee: Ehsan Masood

There isn’t really a timeline as such, but one thing that really does need to happen is because there’s quite a lot of difference of opinion, not only about what’s in there but even basic things like what methods do you use to make your discoveries and inventions, there almost needs to be some kind of peace conference or a meeting of the research community, just to sit down and just make some basic agreements on how they’re going to move forward. That’s the first thing because it’s quite a field where there’s some strong opinions and differences of views.

Interviewer: Benjamin Thompson

Well, listeners, to read more about these stories, head over to nature.com/news, and all that remains is for me to say, Ehsan, thank you so much for joining me.

Interviewee: Ehsan Masood

It’s a pleasure.

Host: Nick Howe

That’s it for this week. Don’t forget, you can check out our Twitter – we’re @NaturePodcast – or if you’re not a twitterer, then you can send us an email. We’re podcast@nature.com. I’m Nick Howe.

Host: Benjamin Thompson

And I’m Benjamin Thompson. Thanks for listening.