Scientists can now make lithium-ion batteries last a lifetime | Hacker News Reader

Scientists can now make lithium-ion batteries last a lifetime

computerworld.com

163 pointsjanvdberg55 comments

A lot of folks are being dismissive of batteries with a longer life, because devices that use batteries also become obsolete; conveniently, it happens on about the same schedule for some people.

But, that misses the point. A battery that lasts dramatically longer means you can deploy batteries in ways and places that couldn't be deployed before, and you can use them in ways we don't currently think "I know, I'll use a battery!" Power for the grid is one of those areas. We don't store generated power right now, because it is too expensive. Partly it is too expensive because batteries are inefficient and horribly short lived. A battery that lasts 25 years could be matched with solar panels that last 25 years (or more) to provide a very effective replacement for grid power. This is just my own take on what a much longer lived battery might be used for. I'm certain there are tons of implementations that aren't obvious to me.

Even if a battery with a 20 year life cycle never makes it into a phone, it would still be revolutionary.

Can somebody skilled in chemistry confirm the article's claim? Because if it is true, then this is a technical revolution. In most devices I own, the battery is usually the only thing that I need to replace at some point (which is especially unpleasant with Apple devices).

I do research in Li-ion battery anodes. High performance materials with exceptional charge capacity, cycling capability, safety, and cost are frequently discovered, but the materials almost never have all of those properties at once, and that is essential to getting a product to market. Most of the new materials with high charge capacity or good cycling performance are very exotic, and thus it is too difficult or expensive to mass produce them.

In my opinion, hybrid Si/C anodes show the most promise as a material that would realistically be found in Li-ion batteries in the next decade. Nanowires/nanospheres do not seem as promising to me, unless we can find a way to cheaply produce them.

How expensive is too expensive? If you can increase a battery's lifespan by 100x, certainly there would be applications for it even if the production cost was 50x that of current tech. You'd still be doubling your lifetime value, right?

Lifetime value also depends on the rest of the device. Many devices also become obsolete at about the same time their battery reaches the end of its life. If the battery lasts 10 or 100 times as long as the rest of the device, the effort expended to extend its life may have been wasted.

Of course, if device manufacturers could be convinced to build devices around a small range of standard battery sizes, then maybe we could move to a Bring Your Own Battery model of powering our gadgets?

Most consumers probably want a thinner device however...

Agree. My current battery in my Macbook Pro reports 85-88% health. That's a significant loss but my laptop is long overdue for an update (when Apple will finally get of their ass and make a new MBP). Stopping the battery deterioration is neat but if it isn't cheap, it wouldn't be worth it.

I'd rather see extended capacity.

I was thinking, in particular, of grid battery power. Even if the battery is 'dated' it's still useful, and it can just sit and do it's job. As long as it is cost-efficient when it is installed, it makes sense to install it.

The conclusion about Li-ion batteries is not supported by the paper. The paper tests the electrodes in a capacitor configuration, not a Li-ion battery configuration. Li-ion battery electrodes are damaged more by the swelling caused by Li-ion diffusion into and out of the electrode than they are by the electromagnetic forces between electrodes.

It's hard to tell if the article is just confusing capacitors and batteries, or if the researchers consider using the coated nanowires in a capacitor configuration as a proof-of-concept for use in Li-Ion batteries.

From the article:

"All nanowire capacitors can be extended from 2000 to 8000 cycles to more than 100,000 cycles, simply by replacing a liquid electrolyte with a... gel electrolyte," the researchers wrote in their paper.

The result: commercial batteries that could last a lifetime in computers, smartphones, appliances, cars and spacecraft.

My takeaway is that nanowires like this are applicable both in capacitors and in batteries; they've determined a way to make them much more durable in capacitors and if they can duplicate that within batteries then it's going turn out to be an even larger discovery than it already is.

If they were actually working on this only for batteries, then presumably capacitors are simply a test methodology that allows them to test hundreds of thousands of cycles within a reasonable time period.

I'm not a chemist, but I have seen enough articles about revolutionary battery technology that I won't believe it until I see it.

Really. This would also solve one of the biggest problems of battery-based grid scale storage. Li-ion batteries are one of the best solutions to storing renewable energy for use in off-peak periods in almost all metrics except longevity. The cost/benefit for grid-scale storage is over a time period of decades and hundreds of thousands of cycles, usually far longer than li-ion is expected to perform.

energy density is my biggest issue with Li-ion... even if it is probably one of the best...

For example: it would be nice to charge your cellphone battery only once a month even if you need to replace the battery once a year

The issue you have isn't with Li-ion -- it has one of the best volumetric energy densities of all commercially available technologies today. [0] Even experimental technologies don't promise more than 2x.

No, your issue isn't with Li-ion, your issue is with physics. As long as power draw stays within an order of magnitude of current mobile devices (note: backlights and radios put a hard lower-bound on this, and devices are already approaching it asymptotically) your dream is not possible without some different technology altogether.

One of those different technologies might be remote wireless charging.

[0]: https://upload.wikimedia.org/wikipedia/commons/c/c9/Secondar...

Apple won't be using those unless they also have higher energy density than regular cells - they will make their devices thicker over Ive's dead body ;)

Removable batteries have been around for at least 75 years, so I assume that if there were an infinitely-lived battery, it still wouldn't be in an Apple device.

I don't understand the logic in this. How does "we don't want a removable battery" imply "we don't want an infinitely-lived battery"?

Planned obsolescence can be seen as a revenue stream to some.

It's not planned obsolescence. Batteries that require tools to replace are commonplace in phones and tablets, because they allow the device to be made smaller and stronger.

Apple will replace an iPhone battery for $79; Independent repairers charge about half that. You can do it yourself with a screwdriver and a suction cup.

I think the point is that nobody is going to want to use their current iPhone in 7.5 years, let alone 75. That is partly because people like shiny new things, but also because your current iPhone will lose iOS compatibility long before that time.

The iPhone 4s and iPad 2 are still getting updates, nearly five years on. I wouldn't be surprised if some of those devices are still plodding along in 2020, particularly iPads being used in PoS applications.

Imagine how long electric cars would last if the only part you really need to replace is the battery. Not to mention, you have a chance to decouple the parts that don't have or need a technical revolution ( the motors ) from the parts that do ( the battery ).

Fewer moving parts is not "no moving parts". Even non-moving parts move in automobiles - the whole thing moves; add vibration and heat stress, and there will probably never be a zero-repair vehicle (excepting heroic or shiny variants)

One of the biggest problems with electric vehicles right now is resale value. If the battery did not degrade, resale value would be much much higher.

If resale prices were higher, people would be more willing to lease and buy new electric vehicles, and then those vehicles would be available on the used car market, and available for hacker types to do interesting things with them [1]

1. http://cafeelectric.com/stretchla/ ; http://www.autoblog.com/2016/04/19/classic-firebird-restomod...

Not as big a problem with Teslas, apparently. The battery is trivial to replace, and they hold their resale value well enough so far.

Not to mention corrosion. In areas that use heavy amounts of calcium chloride in the winter to melt snow and ice, a brand new vehicle can have its frame and body work rust out and compromised in as little as five years.

In the abstract, one could believe that an electric car would be uncomplicated. But that doesn't seem to be true.

For example a few days ago we discussed the BMW i3. Electric means simple? No. Someone in that discussion posted a link to a 40 min video of the i3 production line. I had no idea of how much effort was involved. https://news.ycombinator.com/item?id=11532982

Yes, electric cars are inherently simpler than internal combustion vehicles. But they're still quite sophisticated, with quite a lot that can break or become obsolete.

There's a whole like more than the battery that needs a "technical revolution". People have been embracing recent developments such as automatic crash braking and Apple CarPlay. It would be difficult to retrofit all the sensors needed for automatic braking and lane change warnings.

So we need modular cars that people can update every few years.

Or crush them all and get new ones I guess...

What he/she probably means is that you don’t need a battery that lasts for decades in a device that would become obsolete in five years. It’s a waste of valuable resources, assuming that this battery would cost more to build.

But in may cases the device becomes "obsolete" because its battery no holds adequate charge and is intentionally not replaceable.

For example, that's certainly the case with Microsoft Surface Pro [1,2,3,etc.] computers as MS made it clear that the design life of the battery was about 3 years (and isn't user replaceable). That's absurd for a device with an initial cost of >$1500.

If the battery lasted 20 times as long, I'm sure many such devices would be kept in service for much longer than 3 years, for many users there's no need to "upgrade" after so short an interval.

>For example, that's certainly the case with Microsoft Surface Pro [1,2,3,etc.] computers as MS made it clear that the design life of the battery was about 3 years (and isn't user replaceable).

That the "design life of the battery is about 3 years" and it isn't "user replaceable" doesn't mean the device become obsolete -- just that you need to go to a shop to have the battery replaced.

Wait, is no one considering that these batteries can be recycled?

Apple or third parties can reuse them a lot of times provided that they share a form factor and there are no significant advances to capacity that would require throwing the old ones away...

With all this talk about plastic bags in the other thread, I'm surprised people think no one wants an infinite battery in a device that will be obsolete in 5 years...

5 years? More like 2.5

Because there's no rationale for excluding either other than to limit the product's lifespan.

Devices 75 years ago weren't designed to be millimeters thick.

If you have a removable battery, you need the walls of the battery to be thicker to prevent puncturing it. You have to have extra thickness for the battery cover and for the wall that goes against the battery. You need a contact interface for the battery instead of a flat flex cable.

There's no way to make an iPhone as thin as it is with a removable battery. You can argue that no one needs a phone that thin, but regardless, I don't think it has anything to do with planned obsolescence.

Limiting the product's lifespan is one rationale, but to say that there are no others is a little unfair. Off the top of my head, these are all within the realm of possibility:

- Easier/cheaper manufacturing process.

- Increased structural stability.

- Make the device smaller/thinner/lighter.

- Apple is concerned that end users will screw something up, so they want their service staff to handle battery replacement.

I'm sure the internal discussions that led non-user-replaceable batteries was more nuanced than "let's milk those suckers for all they're worth".

There will be a public rationalization for why they can't include an infinitely-lived battery, too. They'll probably say "power curve" somewhere.

I can't imagine that an iPhone with a battery door would cost more than pennies more to manufacture. iPhones are fragile, and have always been; if they can charge you $79 for a new battery, they can charge you $79 to fix the battery door. Smaller/lighter/thinner? You're talking a razor-thin space savings that adds up to a major inconvenience for users. Users have been operating battery doors for a very long time without electrocuting themselves or losing fingers. They're about as dangerous as an SD slot.

> I'm sure the internal discussions that led non-user-replaceable batteries was more nuanced than "let's milk those suckers for all they're worth".

I'm sure Snidely Whiplash doesn't work there. That a decision both so profitable and user-hostile was made is neither an incidental part of a general striving for perfection, nor undiscussed.

I worked on the iPhone at Apple and sat in many meetings where the product design and electrical engineering leads would almost come to shouting matches over fractions of millimeters. The engineering was fascinating. Everything inside the phone wants to be away from everything else. For example, the radios have power amplifier chips that are hot spots and need to be kept away from the battery to avoid exceeding its thermal limits. The antennas need to be kept away from noisy chips that would affect their reception. There are external connectors and SIMs invading the case. Once that is all crammed in, they need to drop test the whole thing and hope it doesn't bend or shatter.

It's really challenging to cram that much stuff into a constrained enclosure, and all of the layers add up. As I understand, a removable battery takes up a lot of area, because you need clearance below it and space for the mechanical fasteners. It also breaks up the pristine surfaces that are a big part of Apple's design language. The individual industrial design, product design, electrical, and RF design/antenna engineers responsible for each part aren't thinking about sales in two years. They want to make their piece as good as they can and hit the schedule.

Lifetime cycles are an issue for Li-ion batteries.

For example, lets look at the Tesla Powerwall numbers. The Powerwall warranty covers 85% of capacity with 740 cycles in the first two years -- whichever comes first. Then it covers 66% capacity with 1,087 cycles in three years, and finally, it covers 54% of capacity for 2,368 cycles in five years. http://www.greentechmedia.com/articles/read/is-teslas-powerw...

Running the numbers for the Powerwall, 85%, 66%, and 54% of 7kWh is 5.4, 4.6, and 3.8kWh respectively.

(740cycles x 5.4kWh) + (1087cycles x 4.6kWh) + (2368cycles x 3.8kWh) = 17994.6kWh

Thus the cost per kWh stored is

$3000/17994.6kWh = 16.7 cents/kWh.

Not included is the shipping and installation cost, which could add another $300 - $500, which would run

$3300/17994.6kWh = 18.3 cent/kWh.

From what I've read[1] about lithium-ion batteries, it seems to me that if you don't actually need to utilize the full capacity and don't need to have the full capacity always available, you can get some pretty drastic improvements in battery lifespan by limiting the charge to 70-80% of maximum capacity and trying to avoid discharging below 40-50%.

That doesn't mean you can't go outside those ranges in special cases, just that it's best for the battery to spend as much time as possible within that range. For a Powerwall, I'd take that to mean that charging it to 100% every day with solar then draining it to 10% every night recharging an electric car is going to reduce its lifespan pretty fast, but charging it to 80% daily and using it to reduce your grid draw without fully discharging may let the unit last for years with minimal decline in capacity.

It's like a lot of things - if you're constantly bouncing against hardware limitations, there's a good chance you're putting more stress on the hardware and are going to see shorter lifespans for it. It may not be readily visible, but effectively batteries are a moving part subject to wear so treat them as such and plan for it.

[1] http://batteryuniversity.com/learn/article/how_to_prolong_li...

That's true, but I wouldn't call the effect drastic (as long as you stay in normal voltage ranges).

For example, your link says that 100% depth of discharge lasts 300-500 cycles, 50% lasts 1,200-1,500 cyles, 25% lasts 2,000-2,500 cycles, and 10% lasts 3,750-4,700 cycles.

You might think wow, 3,750-4,700! That's way more cycles than the 300-500 of a 100% discharge cycle! This is true, but usually the important comparison is total kWh stored, not total cycles. Each cycle of a 100% loop is 10x more valuable than a discharge of a 10% loop. So the proper comparison is not 3,750-4,750 vs 300-500, it's 3,750 vs 3,000-5,000. So the shallow discharge cycle comes out on top, but not by a huge margin.

Compounding this is the fact that lithium ion batteries are always losing capacity while being stored at room temperature. So this is another advantage of deeper cycling of batteries. You get more value out before they age.

Anyway, you likely knew this, but I thought I would share in case it was educational for others.

If this can extend the cycles to 100,000 (from the article), this brings it the cost down dramatically. Using a conservative half of 100k:

(50,000 x 5.4kWh) = 270,000kWh

$3300 / 270,000kWh = 1.2¢/kWh, more than a factor of 10!

OS X tells me my battery needs to be replaced on my 15" rMBP, it's apparently at under 80% charge capacity now.

According to "System Information", I've done a mere 359 cycles - and according to their website, "Your battery is designed to retain up to 80% of its original capacity at 1000 complete charge cycles." [0]

Of course their warranty on defective batteries conveniently runs out after just one year. I've been meticulous about only charging it with genuine ($80+) chargers, including at home, at work, one in the car, and one in the carry-on. Someone should tell them about this new tech.

0: http://www.apple.com/batteries/service-and-recycling/

Mine has about 650 cycles on it, and the battery currently lasts about 30 minutes, and dies at around 55% (reported by the OS).

There exists already a lithium battery technology commercially available that claims 20 000 cycle life (that is more than 50 years of daily full discharges). Obviously, there are at least currently some drawbacks with lithium titanate compared to the standard lithium technologies, namely cost and power density.

http://www.ev-power.eu/LTO-Tech/ http://www.ev-power.eu/LTO-technology/

(I have no personal experiences with the technology)

Last night an article from popsci.com was posted that covers the same news

UCI Researchers Accidentally Make Batteries Last 400 Times Longer

https://news.ycombinator.com/item?id=11547756

http://www.popsci.com/researchers-accidentally-make-batterie...

Now if they can only make them work in multiple devices... What good is a "lifetime" battery for a device that's no longer used in 3-5 years.

Not of much use with planned obsolescence still being pretty much the standard.

With device support being 2 years at best, and even FOSS being unwilling to go a decade without a ground up rewrite, i'm not sure this will be much worth for mobile devices...

Bad times ahead for OPEC

According to scientific journalism, cancer and HIV have been cured over 100 times. Scientific journalism is monetization of hype.

No, you are just reading too much into the headline. It doesn't say "company ready to produce batteries that last a lifetime" it says "scientists can make a battery last a lifetime". If you read more into it than that, that's your fault.

I am highly skeptical that you can find any articles from reputable publications claiming cancer or HIV have been cured, let alone 100.

The funnel is: Headline -> Visit -> Ad print -> Ad click

They don't care about the content just the ad revenue. If reporting about dog food brought revenue they would be reporting about dog food.

Reply on news.ycombinator.com