Hacker News new | past | comments | ask | show | jobs | submit login
New Lithium Battery Design Eliminates Costly Cobalt and Nickel (machinedesign.com)
223 points by doener on Sept 14, 2019 | hide | past | favorite | 76 comments



> Iron fluorides have more than double lithium’s capacity of traditional cobalt- or nickel-based cathodes. In addition, iron is 1/300th the cost of cobalt and 1/150th the cost of nickel.

I'm curious how much the cost of cobalt/nickel is relative to the whole battery's cost.

The lowered geopolitical risks and general availability could be significant regardless.


Some information about battery parts cost: https://about.bnef.com/blog/behind-scenes-take-lithium-ion-b...


Lithium itself is ~79%, Both Nickel and Cobalt are less than 4/5% each.


This comment is meaningless without clarifying

- which chemistry is being discussed

- what these percentages represent: weight, volume, or cost

There is no lithium-ion battery (NMC/NCA) where Li corresponds to 79% of any of those.


What do you mean here? Lithium is a tiny portion of a li-ion cell, about 1g in an 3600mAh 18650 cell. There is significantly more of Nickel, Manganese, and Cobalt in an NMC cell for example.


Perhaps he means 79% of the cost, rather than 79% by mass or volume.


He'd still be very wrong. Cathode materials are less than 25% of the typical cell price.


There is an already cobalt free lithium battery chemistry there LiFePO4 — been in use for more than a decade, and is a preferred one by commercial and heavy duty EVs because of superior battery life at the cost of capacity.

The new chemistry mentioned has no commercial use cases unless it beats LiFePO4 in its niche or surpasses mainstream chemistries, which is a big stretch.

BYD intentionally went full in onto LiFePO4 because they very well foreseen cobalt shortages.


I think the article says iron fluoride potentially has double the capacity as conventional cathodes but in it's raw form isn't stable enough. I think it's a big deal because 'even' a 50% increase in energy density is a huge deal. And doubling means the absolute end of internal combustion powered cars.

Who knows. I remember articles about LiFePO4 back in the late 1990's. Took them a long time to solve the problem of high cathode resistance and put them into production. Seems like there is a huge space to explore here and little to judge what will work.


LiFePO₄ is also useful for replacing alkaline batteries in small devices, because the cell voltage (3.2V) is approximately 1.5V x 2.


Add to that, in real world conditions very few battery packs can reach 160-170Wh/kg (and hold it after years of use,) while 140Wh/kg LFP been a reality for the last 5 years.

They can be more fully charged and discharged, while other lithium cells have to overprovision massively to extend the battery life. This alone makes it so that energy density advantage of NCA cells is thrown out of the window if you count the effective battery pack capacity with BMS sparing cell life.

And in comparison to NCA cells, LFPs are much, much cheaper, exactly because of the aforementioned cobalt and nickel use. Phosphates are, after all, spent by tonnes as a fertiliser.

Add to that that have a very, very rare trait among chemical batteries — they are stable at high temperatures.

This allows for making them work in very hot climates without liquid cooling which would've added a lot to cost and weight.

It also allows them to take more regen current, and therefore they don't suffer from as big range reduction from "regen choke" in hot climates. From the above, we also get their higher charge current tolerance, meaning that they can be charged faster.

Another very useful trait is their resistance to swell, offgassing and high dimensional stability. This allows for bigger, higher volume fill ratio cells, reducing the battery pack size, and thus its housing mass. Near all "brick" cells are LFP for this exact reason.

They age more evenly, unlike other chemistries, and that too contributes to ease of making bigger cells, and lessen the need for battery balancing, reducing charging times.

EV makers are stuck with small 18650s because it is very hard to increase cell sizes without not worrying about swelling. And the lower volume to area ratio of 18650s also lead to noticeable mass disadvantage.


> Add to that, in real world conditions very few battery packs can reach 160-170Wh/kg (and hold it after years of use,)

Here's a tesla battery pack that comes in at 212 Wh/kg if the specifications are accurate: https://www.evwest.com/catalog/product_info.php?cPath=4&prod...

> And in comparison to NCA cells, LFPs are much, much cheaper, exactly because of the aforementioned cobalt and nickel use. Phosphates are, after all, spent by tonnes as a fertiliser.

Is there a good source for LFP that actually is cheaper than plain lithium ion? I've been looking at doing an electric conversion, and it seems like they tend to run about $400-500 per kwh, whereas lithium ion tends to be more like $300-400. I can believe that LFP is cheaper to make, but if that were so, it seems like it ought to be reflected in retail prices unless vendors are selling them at a huge markup. (I've only looked into this recently, so I don't know if battery prices have spiked because of tariffs on China or something like that or this is just the normal retail price.)


> Here's a tesla battery pack that comes in at 212 Wh/kg if the specifications are accurate:

The question is how much of that is overprovisioned for extra battery life

LFP cost more retail, mostly thanks to fewer makers selling on open market, but lower wholesale price on LFP is something quite certain.


I expect that's probably the true capacity; i.e. you wouldn't want to use more than 80% of that or so if you want the battery to last a long time.


> it is very hard to increase cell sizes without not worrying about swelling

thank you. i've always wondered why they aren't just the size of like a soda can, or a thermos.


Although 26650 (26mm diameter x 650mm length) Li-Ion are reasonably common.


Almost all EVs have much better energy density than that. The Model 3 has about 246 wh/kg. Most of the new EVs on the market right now are probably ballpark above 200 wh/kg and pretty much all of them offer warranties of 8 years for 70% of their capacity. That would be about 170 wh/kg after 8 years.

400 Wh/kg batteries are starting to reach the market pretty soon (interesting for aviation).


Remember, the battery pack energy density vs cell energy density is a very different thing. Model S whole pack's energy density is modest 125wh/kg, way, way less than that of individual cells


Just curious, who else besides Tesla is using 18650s? Or even NCA?


Zotye, Trumpchi, few FAW OEM vehicles, mostly for the ease of air cooling


Interesting. I'm not sure I trust NCA enough to rely on air-cooling, but it's hard to deny they're very attractive on the price/performance side.


I meant them using 18650 size for ease of cooling. What I saw were them having battery packs with zigzagging air channels in between few rows of 18650s, nothing about the chemistry they use


Other advantage is LiFePO4 is a lot safer than cobalt based batteries.

My company abandoned cobalt lithium ion batteries in our products because installers will mount our product in utility closets and then people will do things like store cardboard boxes in there.


cardboard boxes. in closets?!?!


They mean that the space should be set aside for the equipment, not used for the equipment and also storage.


In commercial you can sort of trust they won't do that. Or it's on them if they do. But with residential properties yeah no. You go look and there is a closet with a water heater and they've got boxes and cleaning supplies stored in there.


Given the opening reference to safety, I'd assumed the problem work cardboard boxes was fire risk.


Sure. You dedicate the space to the equipment for safety reasons.


What does the voltage depend on? It's interesting and odd to me that it'd be an inherent property of the chemistry.


It is an inherent property of the chemistry. It's a property of the electrode materials and how they release/accept electrons during oxidation/reduction (with voltage being a measure of the energy transferred per electron). Keyword is "Standard electrode potential".


I'll read up on it, thanks. I'd expect that it was a function of the electrode area or another physical property, so this interests me, thanks again.


Off topic, but how did you write a subscript character? The hn format page doesn't allude to any markdown/html/rtf support.


By using the Unicode "subscript four" character (https://www.compart.com/en/unicode/U+2084)

₀₁₂₃₄₅₆₇₈₉


Yeah, I just googled "subscript four" and copied from the first result. There's probably a correct way to do it.


The method I've come up with is to install an IME (fcitx) that has a addon that can search the Unicode database for a character. So, I press a key combo, type "subscript four", and press enter.



Totally different type of cells. LCO and NMC batteries which this would theoretically replace are energy cells with a specific energy of around 200Wh/kg. These are used when total energy matters, but you don't need high charge/discharge rates, such as with a laptop or cellphone.

LFP like you are talking about are considered power cells (along with Manganese Spinel). The specific energy is about half of an energy cell at around 90-120 Wh/kg. These are that cells you use when you want the stored energy quickly, such as with an electric car or power tools.


> cells with a specific energy of around 200Wh/kg.

The talk is about specific energy of the whole pack, not cells alone. After you take into consideration higher wiring, casing, and cooling weight, the mass advantage of NCA cells will not look so apparent.


There's also traditional lithium polymer batteries which tend to use aluminum and copper as their anode/cathode. Of course they aren't king for density or safety but they certainly don't use cobalt. https://www.genstattu.com/blog/what-is-lipo-battery-pack-con...


Nobody wants batteries of lower capacity though. This LiFeF (or whatever it's called) chemistry promises higher capacity.


It depends. Grid level storage is mostly about cost per kwh. not kilos per kwh. Energy density matters for small devices and transportation but these are not the only applications.


It is impressive that the article actually addressed the issue of the cathode expansion, as that was my first worry about using iron, but not one that most people would be aware of. I'd be interested to hear about the cycle life of these, the 300 cycles mentioned in the article doesn't exactly inspire confidence when something like NMC will reach 1000 cycles.


I wonder how these breakthroughs (if they eventually turn out successful and make the previous process uneconomical) affect the “gigafactories” under construction worldwide? Is a change like this going to render a factory useless if it was built for the “previous” process? Or is this something that can be retrofitted quite easily into an existing manufacturing process such as in the Tesla/Panasonic Gigafactory?


See the second to last paragraph of the article.

Dr. Goodenough’s glass based battery is incompatible with the gigafactory, however this would be compatible-ish with existing manufacturing techniques


dang, please change the link to this: https://www.greencarcongress.com/2019/09/201090911-gatech.ht...

Much more information and a link to the actual paper: https://www.nature.com/articles/s41563-019-0472-7


Is it possible to recycle Lithium batteries to recover the metals in them? That should be better for the environment and the metal source does not diminish.


Yes. This is done routinely for laptop and phone batteries.

I googled some numbers for you: https://recyclinginternational.com/batteries/the-unknown-suc...

Apparently close to 100K tonne of lithium batteries were recycled last year which amounts to about half of the volume of batteries that should have reached their end of life around that time.

Not very surprising as the metal is valuable and the recycling process is fairly straightforward. Apparently, recycled lithium has some nice properties that results in slightly better batteries even.

I'd expect that with car batteries, recycling will be very big business. Dumping hundreds of kilos of valuable materials in a land fill is not going to be a thing. They recycle pretty much all of the steel in cars as well.


> Apparently, recycled lithium has some nice properties that results in slightly better batteries even.

What are those?


Without JS, this page is blank. With JS, it is a clusterfuck. Text is below.

New Lithium Battery Design Eliminates Costly Cobalt and Nickel

A new cathode and electrolyte are the key to doing away with these increasingly scarce metals in lithium batteries.

Stephen Mraz 1 | Sep 12, 2019

Lithium-ion batteries have gained in popularity over the last decade based on their higher power and small size. But their popularity is straining the world’s supply of cobalt and nickel, two metals used in lithium batteries. As a result, prices for those metals have skyrocketed.

To develop alternative designs for lithium-based batteries with less reliance on those scarce metals, researchers at the Georgia Institute of Technology have been looking into new cathode and electrolytes to replace the expensive metals and traditional liquid electrolyte with lower cost transition metal fluorides and a solid polymer electrolyte.

“Electrodes made from transition metal fluorides have long shown stability problems and rapid failure, leading to significant skepticism about their ability to be used in next generation batteries,” says Gleb Yushin, a professor in Georgia Tech’s School of Materials Science and Engineering. “But we’ve shown that when used with a solid polymer electrolyte, metal fluorides show remarkable stability, even at higher temperatures, which could lead to safer, lighter, and less-expensive lithium-ion batteries.”

In typical lithium-ion batteries, energy is released during the transfer of lithium ions between an anode and a cathode, with the cathode typically made of lithium and transition metals such as cobalt, nickel, and manganese. Ions flow between the electrodes through a liquid electrolyte.

The Georgia Tech researchers fabricated a new type of cathode from iron fluoride and a solid polymer electrolyte nanocomposite. Iron fluorides have more than double lithium’s capacity of traditional cobalt- or nickel-based cathodes. In addition, iron is 1/300th the cost of cobalt and 1/150th the cost of nickel.

To produce such a cathode, the researchers inserted a solid polymer electrolyte into a prefabricated iron fluoride electrode. They then hot pressed the entire structure to increase density and eliminate voids.

The polymer-based electrolyte is flexible so it can accommodate the swelling of the iron fluoride while cycling and forms a stable and flexible interphase with iron fluoride. Traditionally, the swelling and side reactions have been key problems when using iron fluoride in batteries.

“Cathodes made from iron fluoride have enormous potential because of their high capacity, low material costs, and the broad availability of iron,” Yushin says. “But changes in volume during cycling, as well as parasitic side reactions with liquid electrolytes and other degradation issues, have limited their use previously. Using a solid electrolyte with elastic properties solves many of these problems.”

The researchers then tested several variations of the new solid-state batteries to analyze their performance over more than 300 charging and discharging cycles at a temperature of 122⁰F. They found that the new batteries outperformed previous designs using metal fluoride even when these were kept cool at room temperatures.

The researchers determined that the key to the better battery performance was the solid polymer electrolyte. In previous attempts using metal fluorides, it was believed metallic ions moved to the cathode’s surface to eventually dissolve into the liquid electrolyte, causing a capacity loss, particularly at elevated temperatures. In addition, metal fluorides catalyzed the rapid decomposition of liquid electrolytes when cells operated above 100⁰F. However, at the connection between the solid electrolyte and the cathode, the solid electrolyte remains stable, preventing the electrolyte from dissolving.

“The polymer electrolyte we used was common, but many other solid electrolytes and other battery or electrode architectures, such as core-shell particle morphologies, should be able to dramatically mitigate or even fully prevent parasitic side reactions and attain stable performance characteristics,” says Kostiantyn Turcheniuk, a research scientist in Yushin’s lab.

In the future, the researchers aim to develop new and improved solid electrolytes to enable fast charging and to combine solid and liquid electrolytes in new designs fully compatible with conventional cell manufacturing used in large battery factories.


Firefox' reader mode deals well with the page by the way.


> Without JS, this page is blank.

Works fine in Dillo, which has no JS capability. Doesn't work in FF with NoScript.


So this is pretty cool, as a GT alum. Caveats are:

- The article contrasts the price of iron and cobalt. It’s true, but fluoride is more expensive than iron. Still cheaper than cobalt, though!

- This is a very new chemistry, which means a lot of scale-up work will have to be done before it becomes commercial. Other technologies have stalled at this phase, like nanostructured Si anodes and S cathodes. In particular, polymer electrolytes can be easy to manufacture or not.

- Li-ion batteries typically perform better at higher temperatures. The tests in this article were performed at 50 C, which may not translate to practical performance in an environment which is often below 0 C. Also, 300 isn’t really that many cycles.


When are we getting lithium graphite batteries which are supposed to be not dense and a lot safer than LiPo? I mean, to the point you can buy one for toys, drones, cell, etc


Maybe one day they will eliminate the lithium needed as well


Should I sell my Cobalt futures then? Or is it one of those "battery-breakthroughs" that we are not going to hear again ever?


I’ll believe it when it ships in my phone.

Otherwise the usual response to battery tech news applies: k keep me posted.


You are right, but the pace at which the battery news is coming these days (frequency), has certainly increased many fold.

Can't wait to have 1Kwh under $50 if this "double the capacity" ever actually comes true though!


Many Chinese suppliers are already operating at or below the $100/kWh mark.


It depends on the use case of course. Low cost with no other constraints is good for utilities, medium cost with smaller size and weight for vehicles, high cost/high capacity/small-light for portable devices.

Since the rest of electric cars is essentially designed, I'm looking forward to someone who really cracks the battery problem. It's a real game changer for a large part of everyday life. No wonder the car companies are thinking twice about building new ICE models, the world could change in a minute.

Question: Is there a graphable Moore's Law at work for batteries?


So you don't want to know about _anything_ until it's available inside phones that you can buy today? I personally like to know what may be on the horizon. It's interesting to read about it, and remember it one the day it arrives (or something else does)!


Not the OP, but... Strictly in phones I can buy today? Not really. Plausibly in phones (or something in a normal household) in the next 3-6 months, sure.

I'm just tired of hearing about battery advances in labs that never make it out to me. Just let me know when I should be prepared to change my purchasing decisions.


> Plausibly in phones (or something in a normal household) in the next 3-6 months, sure.

I don't know, I remember reading 25 years ago about this awesome technology called OLED that instead of blocking light it actually lights up and you can even bend the display!


It would be different if we didn't get the exact same story every week, seemingly.


Web developpers had a similar habit. It eventually got a name: "the javascript fatigue".


More than likely you’ll see it in electric vehicles and utility energy storage first.


Any market usage would be nice. Unfortunately what typically happens is there's an article about a huge breakthrough and then we never hear about it again.


and yet battery capacity keeps on increasing...

"We've made a breakthrough in battery tech" is a news story. "We've increased the battery life in our phone by 5 hours" is a news story. "We've increased our profit margins by 5% by eliminating nickel from the battery in this phone" is not a news story that gets reported, but that doesn't mean the advances in battery tech aren't being used.


The stories that get reported are almost never in the product that gets shipped.

It's like the endless stories about new types of transistor. Computers are faster, but our transistors are not made of diamond or nanotubes or DNA.


> and yet battery capacity keeps on increasing...

my phone was lasting about one-day-per-charge 10 years ago... it's about the same today


The processor is atleast 100x more powerful. The display is 100x better. Speakers, cameras, sensors all have advanced. And yet your battery still lasts on one-charge-per-day. That must be surprising.


but it is not 100x as useful...


Might compare capacity?


I would rather have the same phone that I had 10 years ago if it meant that it would last 7 days with today's batteries


I have a feature phone, ironically lasts about 7 days. Recently purchased.


Have you tried turning off your screen, data connection, wifi, gps, bluetooth and NFC? 7 days of battery life is easy...


I haven't figured out how to use my phone with the screen off... what's the trick?




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

Search: