New capacitor with 19-times energy density | Hacker News Reader

New capacitor with 19-times energy density

livescience.com

106 pointstromp58 comments

> we were able to achieve an energy density of 191.7 joules per cubic centimeter

Which is still quite a bit lower than

https://en.wikipedia.org/wiki/Lithium-ion_battery

> 900–2,490 J/cm3

And I don't see anything here about self-discharge rates, which are a big limitation for using capacitors for energy storage over any period of time. Does anyone know if the full study has that information?

The main use of capacitors in circuits (if we go by capacity) are bypass capacitors. These are essentially small buffers of electricity that can charge and discharge quickly when a jolt of power is needed downstream.

Without bypass capacitors every sudden draw of power would pull the supply voltage of the whole system down, especially if your power source (battery) can't react that quickly or has limited discharge rates.

Without having looked into the plans here, I'd expect a super-capacitor solution to look like this:

1. You charge your cars of super-capacitors super quickly at special charging spots and drive away

2. The stored charge is then partly used for driving, partly used to charge the cars actual battery and partly used to act as a bypass capacitor to smooth out the load for the battery

3. Eventually the capacitors are empty, but you had a short charging process that filled part of your batteries

I don't really think of bulk capacitance as bypass capacitance, but here the role is similar (on longer time scales).

Cars are unlikely to have huge supercapacitors useful for changing the charge much, too. (And there's not much point in moving power stored in a capacitor to be stored in a battery).

The main use of this is to provide short duration power, as you describe. That is, lithium batteries have great energy density but not as great power density or cycle life as capacitors. So a small amount of bulk capacitance can improve battery performance and reduce stresses on the battery.

This is more useful for short range vehicles like plug-in hybrids, because they have smaller packs. Larger vehicle packs tend to have plenty of power available, as a side effect of being designed to store tons of energy.

Makes me wonder if there might be a viable hybrid with no battery, just a supercapacitor. It wouldn't allow you to reclaim energy off hills or the like, but give the advantage hybrids have in city driving without the downside of that battery pack that will wear out in time.

Some diesel busses have a bank of capacitors that is charged when braking that can then run the onboard electricity for a while.

If an EVs batteries aren't mostly full, the limitation is in the amount of power that can be pumped into them. There isn't really a reason to charge capacitors directly or in any special way.

Even for high discharge I'm skeptical, lithium titanate batteries are probably much better, their durability and extreme charging and discharge rates make them half way to capacitors already.

> the limitation is in the amount of power that can be pumped into them

You could say that. You could also be more pedantic and say that this comes at the cost of some decrease in battery life (and efficiency) due to the resistance resulting in some amount of heat in order to facilitate this high-speed charging process.

But I don’t know if the ROI on improved battery life would be worth the complexity, size, weight, and cost of a two-stage charging process. As battery tech improves, that resistance continues to come down.

this comes at the cost of some decrease in battery life

What does? I said the limitation is usually getting enough electricity to charge them fast. Most of the time EVs are not being charged as fast as they could be.

Bypass capacitors aren't really used to prevent a whole-system voltage sag - they are there to deal with the effects of high-frequency signal switching.

The traces supplying power aren't theoretical, so they have a nonzero impedance. This means that when there is any change in power demand due to transistors switching, the trace will try to resist this change in supply current. An increased power demand leads to a very local voltage drop, a reduced power demand leads to a very local voltage spike.

Bypass capacitors are placed as close as possible to the individual chips to reduce the impedance between the capacitor and the chip. You even intentionally use smaller capacitors because the bigger ones have a higher impedance.

Does it make sense to compare a capacitor to a battery? It is a bit like comparing SRAM to Flash imo.

I think some of the areas of a car to be potentially improved with this kind of advancement are: regen braking efficiency, size and weight, battery health.

Does it make sense to compare them as the lines between them blur? Absolutely! That's the best time to compare.

And the same is true for DRAM and flash. There have been several moments in time when the tables have tilted where battery backed RAM PCI cards have made sense, or where sticking persistent Optane chips in high speed DRAM slots has.

Indeed, DRAM and Flash have experienced a convergence - in server and mainframe land there is storage-class memory that functions as a cache or LUT between DRAM and Flash.

I believe we are a bit farther from a true hybrid capacitor. As for true hybrid memory I believe the likeliest solution for embedded applications will be one of the different analog in-memory compute techniques currently being iterated on.

Yes because ultimately the goal of capacitor research is to make it more similar to a battery in terms of storage.

And battery research goal is to make them behave more capacitor-like, eventually they will cross their paths.

a big limitation for using capacitors for energy storage over any period of time

Who is saying that's the point of this capacitor?

The first sentence in the article is:

> A new material structure could revolutionize energy storage by enabling the capacitors in electric vehicles or devices to store energy for much longer, scientists say.

Yep still much lower than Li-ion batteries but within punching distance for sure. I also don't think this is intended to compete with traditional battery tech as much as making it even better in applications where supercapacitors currently excel at.

I wonder what the comparison is in J/kg? I could see lithium & capacitor hybrid systems being viable in EVs. The cybertruck's battery housing is full of plenty of deadspace, and if these caps have 1/10th the energy density you could squeeze in an extra 10+kwh of storage. Caps typically have a much longer lifespan, so I could see them acting as a great buffer for the rapid charge/discharge of regen braking.

I recall following a project about 15 years ago where someone converted a car to electric using a lithium battery for range and a bank of ultracapacitors to increase peak power and handle the energy from regen. Here's the page on the project: https://www.metricmind.com/ac_honda/main2.htm

Great link! It looks like the caps he used were ~ 3 J/cm³, so 1/65th the density of the ones in the article.

The source article is Enhanced Energy Storage [...] Ferroelectric Ceramics via Domain Engineering https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10381779/

It seems like they didn't really look at leakage current, but did apply a 10 Hz sine wave to it for about a day (~1 million charge/discharge cycles) and it didn't degrade notably.

I don't think the intent is to use the capacitor as a substitute for battery energy storage, but as a complement. I don't think these would be a good fit for EVs, but for supercharger stations maybe? Charge/discharge efficiency and cycle lifetime will probably be higher than a battery, and in stationary applications, cost matters more than weight or volume. Of course cost depends on production volume, which isn't something you can measure in the lab, so who knows.

It might also pair well with battery chemistries that are good for energy density but poor on power density, like lithium-sulfur or LFP.

For an EV, they could be useful in stopping/starting. Rather than having regen go to the battery you could first send it to a capacitor which then trickle charges the battery (reducing wear). And rather than pulling directly from the battery, you could pull from a capacitor first to get to cruising speed. That would allow the battery to slowly discharge into the capacitor rather than going directly into the motor at full force.

You'd play to capacitors high cycle life and fast charging/discharging capabilities rather than their density.

Maybe it could be used in a similar fashion for the hybrids out there.

That's also much less than 19x what an EDLC can store. Picking a standard Maxwell part[1] I calculated 36 J/cm3, so it's more like 5x; still impressive, but comparing a lab result to a product isn't fair; there are EDLCs with much higher energy densities in the lab.

https://maxwell.com/wp-content/uploads/2022/07/3003348.2_3V-...

I can think of many uses for a capacitor with close-to-battery energy density. The cap properties (endurance & high charge / discharge rates) are pretty useful in some applications.

I think this would complement batteries.

One big use would be ev road racing.

Cars on a track have high acceleration, followed by extreme braking.

If this energy could be buffered by a large capacitor, the main battery could just be slowly discharged to a capacitor to provide energy for the race. The capacitor would provide and absorb fluctuating energy to speed up and slow down.

I also imagine complex ev battery architecture could be simplified if the energy in and out doesn't have to be radically managed.

I see it somewhat like running an internal combustion engine at one efficient RPM instead of having to run at all RPMs and use a gearbox.

that last bit i find interesting. You not just get to run it at efficient rpm it can also be dramatically smaller and have better acceleration. It could even continue to run after reaching the destination (since you have an engine it might as wel do work for you)

I was using that as an analogy, but there's no reason you couldn't have a weird hybrid car with a generator, a capacitor and electric motors. No batteries.

No reason in principle, but for practical reason that probably will never become common. Seems more likely we'll get "hybrid EVs" with batteries and capacitors working together, capacitors acting as buffers for batteries. Until the technologies converge into a single thing that does both, but I'm guessing that'll take decades rather than years. (Then the only question that remains will be, will they be called batacitors or capatteries? :-)

This is a ferroelectric capacitor.[1] Those have been around for a while, but in small sizes for memory cells, not big ones for bulk energy storage. Actually, it's just a new material for one, not a whole unit. Here's the actual paper. [2]

The paper makes says it might be useful for pulsed energy systems - radars and such, perhaps. It's a ceramic capacitor, so potentially it can have low equivalent series resistance and you can get the energy out fast. The paper doesn't even mention longer-term energy storage. The picture of a car charging port is deceptive.

[1] https://en.wikipedia.org/wiki/Ferroelectric_capacitor

[2] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10381779/

Regenerative braking is always a big deal in these discussions, put power control circuitry also likes a beefy capacitor. If you look at a Macbook power supply, the size of the caps seems to be dictating the thickness of the brick (the transformer is the same thickness but those can be rearranged)

More energy density in capacitors is a big thing, as capacitors are easily the biggest parts you will have to use on your PCB (if we exclude non-negotiable parts like connectors, buttons, displays, etc.)

But I'd like to add that it might take a long while for the results of this work getting into market. For circuit designers it is crucial to know under which conditions and after which time capacitors fail (and in which way they fail). This is all knowledge that has to be generated before such a thing goes to market. And usually you don't wanna be the early adopter on things like these

is it a big deal ? Any components need to pass some conformity tests, and testing capacitors is an established process.

There are bike with supercapacitors out there, so it's definitly not that far from the market.

> More energy density in capacitors is a big thing, as capacitors are easily the biggest parts you will have to use on your PCB

I don't think it's that simple, energy density and power density seem to be roughly inversely related across technologies. i.e the more energy you can pack per unit mass the slower it is to get it in and out. A substantially lower power density could actually make the equivalent component larger or heavier for many purposes.

Super capacitors already exist that bridge the gap between electrolytic capacitors and chemical batteries. If you compare the specific energy with the specific power of each the ratio drastically changes from left to right in this table with old electrolytics beating everything when it comes to power:

https://en.wikipedia.org/wiki/Supercapacitor#Comparison_with...

On the other hand this comparison is per unit mass, and these technologies differ vastly in volume density, so a comparison per unit volume would reduce the differences in power density. I'm pretty sure the order for power density will be the same though.

> energy density and power density seem to be roughly inversely related across technologies. i.e the more energy you can pack per unit mass the slower it is to get it in and out.

This sounds really elegant and simple to the point where people might take it as an axiomatic fact but the very existence of both nuclear fission and fusion stops it from being a general rule. They’re both discharging power from individual atoms and it would actually be great if we had a way to slow down the rate of discharge.

It also has plenty of plain old basic chemistry examples showing it false - lithium ion vs lead acid, gasoline vs fuel oil, etc.

Yes, it's only a high level observation, but one that seems to hold true for electrical storage so far, i.e where there is a closed system with an electrical input and output.

For fusion there are also various direct energy conversion methods for electrical output, but not the reverse, i.e i don't think there is method for nucleosynthesis directly from charged particles.

In a more general sense, sure, power vs energy relation doesn't hold, but when limiting the system to the requirements to be a self contained electrical storage component, this relationship seems to emerge.

> More energy density in capacitors is a big thing, as capacitors are easily the biggest parts you will have to use on your PCB

Not really. Most modern electronics don't need significant capacitance in F. Your selection of capacitors will be dominated by a very large number of physically small, low-value bypass capacitors, one each placed as close as possible to every single power pin. The main thing you care about is their inductance - which is why you want them small and close to the pin. I'm already using the smallest parts the factory lets me without charging a premium fee, so nothing to gain here.

Beyond that there might be a handful of bulk capacitors distributed among the power rails, but those are rare and already rice-sized. Making those smaller wouldn't hurt, but it's not the first property you'd be selecting it on, and the potential gains are pretty much wasted on anything which isn't an iPhone.

There are still some rather bulky capacitors in things like chargers and power supplies, but those aren't exactly space-limited.

Is "if reproduced at scale" the material science equivalent of "in mice" in biology/medicine? I don't mean this in a cynical "this will never work" way, I'm just wondering about the culture and "path to innovation/mass market" of this field.

Advertisement density is pretty maxed out on that site

𝄞 Used, abused, without clues 𝄞

𝄞 I refused to blow a fuse 𝄞

𝄞 They even had it on the news 𝄞

𝄞 Don't believe the hype 𝄞

The word "battery" originally referred to a bank of capacitors. One day, capacitors will reclaim the word.

The Rho Agenda sci-fi books rely on capacitors with absurdly high energy densities to power the tech. Is there any thing in the physics that would prevent a capacitor from having a higher energy density than a battery?

I remember some attempts to achieve that, do not have source anymore. The whole idea was to take two electrodes and grow thick forest of graphene tubes on them massively increasing surface of electrodes. But to get even close to batteries, electrolyte was necessary, however pouring electrolyte onto electrodes modified in such way caused graphene tubes to break off and create shorts. So ultimately went nowhere. Maybe if it would be possible to grow graphene in the electrolyte directly.

I would be very, very leery of storing large amounts of energy in supercapacitors. A li-ion battery pack can go into thermal runaway, but a supercapacitor inherently dumps it's energy if penetrated. Electricity moves at a respectable fraction of lightspeed, the energy dump will be *very* fast.

In fact the books use that rapid discharge property to power weapons.

> a tiny gap in the core increases the relaxation time — a term used to describe the period over which the capacitor loses charge

> materials are layered like pasta sheets in a lasagna

Isn't this similar to how current lithium cells are made? An anode and cathode with a gap layer sandwiched between them (just on a different scale)?

Could be useful for Mach effect thrusters. More transient mass to push against.

Another Better Battery Bulletin. Meh (An exCeptional capacitor circular?)

Not exactly related to the article, but reading it triggered a sudden realization: vampires are just big capacitors. If you put a stake through their heart, you connect all the plates (possibly many layers of them) and they are vaporized into a cloud of dust.

Well, dust and steam, which could be where the whole "turning into mist" comes from.

I got nuthin' for blood and bats.

The "magic smoke" is what happens when your PC gains a scintilla of self-awareness and realizes it entered your home without an invitation.

There's a scifi movie. 5 million years to earth. They do something like that.

It's a good movie.

The idea that vampires die when placing a stake through their heart is so funny to me, because most things die once you've put a stake through their heart...

I think the idea is that vampires don't die unless you put a stake through their heart. Most things have many other ways of killing them.

I thought sunlight and water also kills them.

I think the original idea was vampires don't stay dead unless they're staked through the heart.

But most things die if you put the stake through their brain. Vampires only die if it goes through the heart. I'm thinking of one of Saberhagen's vampire stories where the vampire's car (or perhaps he was a passenger??) has a bomb that goes off, as it goes boom the vampire is worried about whether any of the dash is made of wood.

Reply on news.ycombinator.com