Hacker News new | past | comments | ask | show | jobs | submit login
A long-lasting, stable solid-state lithium battery (seas.harvard.edu)
157 points by macinjosh 27 days ago | hide | past | favorite | 59 comments

Every time there is some news on some research breakthrough in a technological field, lots of people comment how it is unrealistic, not yet a working product...

Yes, it's research if they had a product they would have founded a company and made lots of money already. I suspect these comments are due to the fact that many here are in software, where an idea can essentially instantly transition from research to production. In hardware that can take decades (and there might be lots of deal breakers discovered along the way).

Now if the complaint is about the press releases reporting on those breakthroughs. That is a problem with the pressure on scientists to work on relevant problems (often strongly called for by the same people who then complain it's not a product yet), and demonstrate impact.

Indeed, not a working product. Working lab-only 'hardware'. The "methods section" in the Nature paper is fascinating:

"The anode-solid electrolyte-cathode for half-battery or anode-solid electrolyte-anode for symmetric battery were pressed together in a homemade pressurized cell at 467 MPa and kept at 50–250 MPa during testing." "All batteries were assembled in an argon atmosphere glovebox and the galvanostatic battery cycling test was performed on an ArbinBT2000 workstation."

A lot of things which look good in the lab don't make it to production because issues not highlighted in the original research though.

I've seen this in software as well, trying to implement novel algorithms from papers.

For example, one raytracing denoising algorithm I tried worked wonders, except it required the entire rendering to be serialized (raytracing is "embarrassingly parallel"), alternatively use a per-pixel instance of the data structure, leading to many GB of extra memory required for larger images. Results looked great in the paper, not so much in the real world...

All the standard comments about battery breakthrough articles apply. Who here is tracking which emerging battery technologies will hit scale manufacturing first and how much of a difference will they make for cars, stationary storage and personal electronics from a cost, weight and volume perspective?

All the standard comments about battery breakthrough articles apply.

Yes. 1) It's in Nature, which used to have a good reputation but has published flaky battery articles, 2) the abstract doesn't indicate that they built a usable battery, only that they designed one, and 3) the phrase "up to" is used on performance numbers.

Take it for what it is: scientists chipping away at different problems related to storing energy in solid state batteries. There is a lot of that happening. It adds up to companies finding their way to the market eventually. Battery longevity and charging times are two of the problems addressed by this particular solution.

From the lab to mass production in a giga factory while being cost competitive with other solutions is not something that is going to happen overnight, regardless how good the technology is. It requires lots of investment, planning, prototyping, etc.

However, there are several companies gearing up to start battery production around the 2025-2030 time frame that are targeting mass production that have had lots of investment and are at this point beyond the prototype stage. This includes companies backed by car vendors including Toyota, VW, GM, etc. to the extend of many billions. That suggests that if their collective efforts pay off, we could see solid state batteries creeping into the market over the next eight years. I don't expect it to displace current non solid state battery technology until some time next decade at the earliest.

Cars may not be the first logical application. Planes are a much more lucrative target where the price of the batteries won't be as much of an issue and the increased safety and energy density are highly desirable properties. And the market is much smaller, making it possible to target it with a bit more modest infrastructure. There are several electrical planes in or near production right now where e.g. 2x capacity increase would be very meaningful in terms of range. E.g. Bye Air just announced a 500 mile, eight seat twin prop a few days ago. It's not going to be cheap. It would make perfect sense to use relatively expensive batteries in such a plane.

Stationary storage is not an obvious initial market for solid state batteries either until they become dirt cheap and mass produced. Energy density is just much less of a concern for those and instead cost per watt is going to be the key driver there. For commercial vehicles and (taxis, trucks, vans, heavy equipment), charging speed and cost are the main drivers.

Faster charging speeds means capacity becomes less important. If you can top up a 500 mile battery in 10-20 minutes, maybe a 200 mile battery is good enough for most. You can get that capacity with most EVs today and charging times are getting there as well. I know I don't last that long without a bath room break at least. Solid state batteries will be nice to have in this market but not essential.

IMHO solid state high performance batteries will find their way into EVs eventually probably starting with high end, overpriced vehicles.

Thanks for sharing your perspective!

...for mice!

I am also interested in seeing a tabular summary, but it seems to be quite a big task to collect all or most candidates. The published paper alone lists quite a few references, each one possibly offering an alternative electrode composition.

Greatest battery of all - Mitochondria!

While it's always good news that there's research going on in this field, the litmus test is still: do they have a large scale production process developed yet?

Currently the real revolution is the increasing density and thus cost-effectiveness of LiFePO4 batteries, which recently reached $100/kWh in certain applications and made the DoE revise its target from $80/kWh to $60/kWh.

High energy density and quick charging are nice-to-haves and would enable electric airliners, but what should be the focus for now is cost-effective grid-level energy storage - LiFePO4 are inching closer to that with every passing year.

I like it, but I'm also afraid of it.

Lithium metal reacts explosively to water, and it's impossible to avoid a vehicle battery cracking open in a bad enough accident. Physics is a stern task master.

Existing lithium ion batteries are pretty dangerous, as we've seen from accidents involving Teslas and the like: but so are gasoline autos.

It would be hard to persuade me that these batteries aren't substantially more dangerous still.

That all said, rapid recharging could be a useful feature for stationary storage batteries, and the higher energy density is a nice bonus, although space isn't really the limiting factor for storing renewable energy. There's a place for this technology, somewhere.

I'm always happy to read about new battery chemistry, but always come back to the slogan: Goodenough's batteries are good enough.

It’s not the lithium which is dangerous with li-ion batteries though. It’s the flammable liquid electrolyte. That’s what gives you explosions and fire. I think the lithium does make it harder to put out after it has caught fire though, but it’s not generally the reason why they catch fire.

With next generation batteries, some have been shown to be completely fire proof, as they don’t all use the same flammable electrolyte as Li-ion. With some, you can cut them in two without anything happening.

I’m not at all worried about the lithium in lithium metal batteries. The key is reaction rate. A pure chunk of lithium metal can be dangerous because all the metal is available for reacting to happen nearly instantly. A microsopically thin slice sandwich between non-reactive materials is not a big concern. With catastrophic damage you may have thermal run-away. But fire departments in countries with high share of EVs, like Norway, have already shown they can handle it well, and they consider EVs far safer than gasoline cars.

Sure, a lot depends on the exact formulation of the batteries.

It still makes me nervous. Most of what you said was about lithium ion batteries, which is irrelevant here. I gather that you know that, and I do as well, but I thought it was worth spelling out for the general reader.

It's not irrelevant at all, you mentioned them yourself, and he spent more time discussing lithium metal. What's irrelevant is your nervousness.

Doesn't LiFePO4 have substantially the same electrolyte yet is self-extinguishing?

It's worth noting that the solid electrolyte in these batteries is based on phosphorus pentasulfide -- not particularly flammable, but prone to releasing clouds of extremely toxic hydrogen sulfide on contact with water. Exit one problem, enter another.

Hydrogen sulfide is dangerous in a contained area, the smoke from a smallish fire producing it would be something to be aware of, not a catastrophe.

One big whiff of concentrated hydrogen sulfide will kill you very dead very quickly. People tend to think they would smell it first, but if it's concentrated you'll be dead before the smell even registers.

Extremely unlikely.

Not at all.

This is a known hazard when working on waste-water networks. There are a few deaths every decade, despite nearly everyone having gas detectors (and being told to use them before going into a manhole).

At high enough concentrations, H2S just shuts down your smell receptors.

People who die in such situations don’t not smell the hydrogen sulfide. They just either have no way out quickly or they keep going in spite of the danger. I’ve been in such situations.

a contained area, like ... the cabin of a car?

The giant fire would be the much bigger problem if you are still in the car.

(The fire is smallish in a 'amount of material' sense and huge in a 'you don't want to be anything resembling close to it' sense)

Always good to be cautious when there is that much energy just "sitting there."

It is a clever hack to add a layer of electrolyte that inhibits dendrite growth. Interesting question how thin they can make that layer before it no longer stops dendrites.

For those wondering, a "dendrite" is solid metal (in this case lithium) that precipitates out of the electrolyte during discharge, favoring the path of current through the electrolyte. If the dendrite reaches the cathode it creates a short circuit inside the battery. This problem can result in fires and other issues. When the battery is charged, the electrolyte chemically reabsorbs the metal and "eats" away the dendrite. Leaving behind a "hole" where it had been in the electrolyte. Since the whole doesn't conduct current, it reduces the current delivery capacity of the battery. The electrolyte they have chosen in the paper can actually "fill in" the holes to some degree (they call it "self healing") which minimizes battery "memory" about previous cycles and extends the number of times you can recharge a battery.

What that means is that in the lab there are many different ways to do this, this is yet another. So far, none of them have reached the point where you could manufacture batteries that were cost competitive with the current designs.

My prediction has been that flow batteries (batteries that can replace their electrolyte with "charged" electrolyte for "instant" recharging without needing high currents at the charger, are more likely to become practical before all metal Lithium or Lithium-Silicon batteries become practical, but it is super awesome that so many smart people are trying to solve this problem.

Actually, e.g. sodium could be great for energy storage as a metal as well. There are long expired Lockheed patents about how to turn sodium with water directly into electricity. Other processes, that already work on industrial scale could recycle the product of that reaction sustainably back into sodium metal. The energy density of sodium is a bit less than e.g. ammonia, methane or liquid hydrogen but it is quite easy to produce without any fossil fuel consumption, it is easy to handle it at room temperature and normal pressure. The sodium hydroxide solution is not too bad if e.g. spilled on some land or in the ocean - it will neutralize quite quickly and on land even add needed minerals. Yes, it burns with water but otherwise isn't very flammable and you don't get vapour problems as with gasoline (e.g. when you fill a large tank).

Sodium also melts just below the boiling point of water so it even could be feasible to transport using a pipeline (it would just need some insulation) - e.g. you could use it instead of water steam to transfer heat as well. But this is secondary.

If you are interested to read more about how these facts could be used more globally, read more here: https://www.orgpad.com/s/energiewende

Sodium-sulfur batteries are available commercially. They have good energy density but they have to be heated to 350C which makes them impractical for vehicles. They're commonly used for grid storage.


If you mix sodium and potassium, the eutectic melts at -12.6 C.

Lithium-ammonia eutectic ("metallic ammonia") freezes at 90K, the boiling point of liquid oxygen.

Vehicle batteries are heavily compartmentalized. They use many small cells, much like your laptop battery likely used several smaller cells. Cells in a car battery are not much larger. And there are walls between groups of cells, both for structural and safety reasons. Cells are often even charged individually, with parameters tailored to a specific cell and not its neighboring cells.

Because of this, if a battery is badly deformed, chances are that most cells stay intact, and only relatively few cells expose the electrolyte and the lithium.

Also, I heard most fire departments know better than to put out a car fire with water, even with gasoline cars. Car fire extinguishers normally use powder or pressurized CO₂.

>and it's impossible to avoid a vehicle battery cracking open in a bad enough accident

If you watch old car racing competitions, you will see that they were extremely dangerous because of the gas deposit burning at the minimum crash.

How many of those you see today? Close to zero. They let the deposit deform under an accident and it really works.

Electric cars are much safer for humans than ICE in accidents because batteries could displace under people instead of hitting people like engines do.

Electric cars are going to make cars significantly safer by avoiding the accident in the first place.

I prefer having a car that helps me not having an accident than worrying too much about what would happen in a bad enough accident.

> Electric cars are much safer for humans than ICE in accidents because batteries could displace under people instead of hitting people like engines do.

Don't engines do a similar job of going underneath the chassis in high impact collions?

I don't think EV are safer, because there aren't as many data points just yet - but seatbelts definitely help.

Energy is energy. If there's a way for damage to cause the battery to discharge "all at once" there will be an explosion/fire just as bad as the equivalent gasoline fire.

> Electric cars are going to make cars significantly safer by avoiding the accident in the first place.

I see no reason why the fuel source has any bearing on accident avoidance.

EV's are theoretically better at accident avoidance because of their distribution of mass compared with ICE.

e.g. Lower/distributed center of gravity makes cars faster to stop, and easier to turn.

This is one of many advantages. Some others are related to: fewer components i.e. fewer failures, instant torque i.e. better handling, etc.

That said, given that an equivalent accident occurred to an ICE car and to an EV from the side[0], "energy is energy" and likely will result in similar injury/fire.

[0] the front/back of EVs have better crumple zones than ICE.

Lithium iron phosphate batteries are extreme safe though?

Relatively speaking, absolutely.

Both those and other formulations (mostly lithium cobalt oxide) have ionic lithium, hence, lithium ion.

This uses metallic lithium, which again, reacts explosively in the presence of water. The danger of 'standard' lithium ion chemistries is, as a sibling comment mentions, that the electrolyte and anode are flammable. LiFePO is much more stable and unlikely to burn, relative to lithium cobalt oxide: but neither of them will straight-up explode if a naive fire department sprays water on them to try and put a fire out.

I don’t think that’s the huge risk you think it is. The cells are encapsulated and the lithium metal (in very thin sheets between non-flammable solid state electrolyte) is not exposed. Naive fire departments spraying water on an oil gasoline fire would cause much worse damage.

> Goodenough's batteries are good enough

Not really for cars. For cars they still need to become cheaper, safer, lighter, more sustainable, more recyclable, better energy density for more range, better power density for recuperation and faster charging, and more cycles for greater life span.

This aluminium-ion battery looks interesting:


The power density and duty cycles are there. They say the energy density is there once they increase the cell voltage.

Aluminum ion batteries are not actually any better than lithium ion. They’re worse.

Aluminum air batteries are also worse than lithium air batteries.

We have existing lithium ion cell chemistries that allow 350-500 mile electric car range and battery lifetimes well into 500,000 miles, possibly millions of miles.

> Aluminum ion batteries are not actually any better than lithium ion.

This one is. Cheaper, safer, better power density, long cycle life. It has a wide temperature tolerance so may not need much temperature management when in an EV battery pack:


Their batteries are at 1.7 volts currently. They say they can achieve the same voltages as current lithium-ion batteries to be a direct replacement for them.

> We have existing lithium ion cell chemistries that allow 350-500 mile electric car range

Few EVs achieve that. Why? Because the batteries cost too much and weigh too much leading to a more expensive vehicle. Not everyone is going to buy a Lucid Air:


We need cheaper batteries, like this aluminium-ion battery.

I think you'll find that the energy density and energy co-location of a gas tank to be far greater than a battery.

Far more negative outcomes are possible with highly flammable fuel only separated from the outside world by a thin metal membrane.

But fuel requires a spark or heat to ignite , a lithium battery pack can explode or catch fire after being punctured and the internals exposed to air.

Lots of expensive cars have loads of magnesium in their gearboxes. Nobody thinks about that when driving with low gearbox oil levels.

Sounds very promising! It's super annoying how they keep duplicating paragraphs of text tho. I don't like reading the exact same text twice and not realizing it until halfway through.

Looks interesting. It has a second electrolyte that is 'is less stable with lithium', sounds like it you might not want to crush it as might happen in an auto accident.

In a NOVA episode on super batteries, I saw a demo of a lithium ion battery that was safe enough that you could puncture or snip of pieces of the battery packet while it is in use.

Lithium Ceramic Batteries are fairly forgiving, it looks like: https://youtu.be/kJXRyWQgOY4

Can anybody access the paper? I'm curious if they mention anything about energy density or specific energy.

From the abstract:

> Our design also enables a specific power of 110.6 kilowatts per kilogram and specific energy up to 631.1 watt hours per kilogram at the micrometre-sized cathode material level.

Didn't anyone notice the aptronym? The researcher is named Li.


feels a bit overdone, given the freaking press release admits as much. We don't need meme copypasta here.

> “This proof-of-concept design shows that lithium-metal solid-state batteries could be competitive with commercial lithium-ion batteries,” said Li. “And the flexibility and versatility of our multilayer design makes it potentially compatible with mass production procedures in the battery industry.

> and is just around the corner from taking over the world

From the article they don't appear to be making such a claim:

> Scaling it up to the commercial battery wont’ be easy and there are still some practical challenges, but we believe they will be overcome

Do you have anything concrete to say about this technology?

> [ ] it is incapable of delivering current at sufficient levels.

is this a problem? I mean li-ion cells used by tesla only have around 4 volt and ~3000mA. I would consider size per cell/thermal per cell (as you stated warm/cold environments and charging) more important to make a good distinction.

Watching latest Sandy Munro's videos it seems big issue with solid state batteries at the moment is indeed operating temperature. Current one's deployed are operating at 80C which is OK for constantly utilised vehicles, but not personal cars.

In this case, which cross(es) are the most likely candidates for being crossed ?

I choose to interpret the post as a sort of subversion of the usual way this meme is used: none of the boxes are checked, so in this case the battery technology has actually passed the rather rigorous battery-technology-skeptic's filter. The post is rather ambiguous, though.

At least check a few of the boxes. It will lead the discussion properly.

FWIW, this is a meme reference. See: https://craphound.com/spamsolutions.txt.

I have no idea what makes you think they weren't aware this was copied.

Can’t wait till 2050 then!

This story has now been posted on HN for three days in a row. Somebody is pumping it.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact