New technology: the Air battery

[FreepZ]

Here's an interesting article in New Scientist:

Air battery to let electric cars outlast gas guzzlers

Basically, it's some new battery technology that, some time down the road, may provide significantly more range to EVs (assuming that the scientists can resolve all for the with this new technology). Note that  Lithium Air batteries are not new, but getting them to work is!

(Thanks to TTXPG Matters for the article link.)

[ttxgpfan]

Thanks for the mention!

But a few hours later I posted about a new Zinc-air chemistry.  Non-toxic, plentiful, won't catch fire when exposed to moister, and twice the density of current batteries at almost an 1/8 the cost.  The start-up company is looking to do grid applications, so they are aiming for a 10,000 cycle life span.  They already have a small pack up to 2700 with no issues.   2700 cycles?  That's 270,000 miles in an Empulse.  I'll take them now, thank you.

[00049 (AKA SopFu)]

But with air batteries you don't need the cycle life that you do with lower energy densities.  For example, if you used the same weight Li-air battery as the Li-Ion battery, you would have a 1,000 mile range out of the empulse.  If you *only get 100 cycles out of that, it's still 100,000 miles before the battery reaches end of useful life (which would just be an 800 mile range instead of 1,000...).  Lots to think about, though.  Charging time won't get shorter with just new chemistry - so to charge from dead a 1,000 mile range empulse would take 100 hours instead of 10.  But at least there will be more options for charging - i.e., if you forget to charge one night you would still be able to get to work in the morning, or options for grid load leveling.

I would be willing to bet that Li-air batteries will be commercially viable within 5 to 10 years.  That means if you wait to get an EV for 2 or 3 years, its value will drop significantly once the Li-air batteries come out.  So buy now!  Er, I mean GIVE ME MY BIKE!

[protomech]

Another thread started on IBM's lithium-air .. maybe we could consolidate all the individual technology threads into one thread per chemistry..
http://brammoforum.com/index.php?topic=1216.0

Researchers in Korea and Italy claim to have found a Li-Air electrolyte that doesn't degrade as quickly with more cycles. This is the same sort of thing that IBM claims to have found, so perhaps we have multiple researchers proceeding along parallel tracks.

The problem has been, as the researchers put it, that lithium-air batteries have an end-point of lithium peroxide (Li2O2), which forms through an intermediate oxygen radical. That radical is very reactive and will generally decompose the electrolyte that shuttles charged ions around between the battery's two electrodes. If it's not possible to avoid the reactive oxygen, the authors reasoned, the best thing to do is to change the electrolyte to something that doesn't react with oxygen.

Some preliminary research in this area had been done, but the initial materials would only conduct charges well at temperatures above 70°C. The authors came up with a mixture of an ethylene glycol derivative (tetra(ethylene) glycol dimethyl ether) and a complex lithium salt, LiCF3SO3. This worked well at room temperature and, perhaps most significantly, the authors found it went through oxygen reactions so quickly that they couldn't detect any reactive oxygen intermediates. "Equally importantly, the peak corresponding to LiCO3+—one of the most likely products of electrolyte decomposition—is not seen," the authors note.

This chemical stability also translated to stable performance, with the behavior on the 20th charge/discharge cycle being difficult to distinguish from the 100th. It also performed well across a variety of charges, from half an amp/gram of electrode material up to 3A/g.

Note that if we get 6-10x improvement in energy/mass and energy/volume, we can probably have packs that give 300-500 mile freeway ranges. 100 cycles without change, say the pack goes to 300 cycles EOL .. that's 90-150k freeway miles.

***

A poster with a supposed PHD thesis on Li-Air drops in:

Sad news for you guys, the 13,500 Wh/kg is bullshit. It uses the mass of the empty electrode, before the battery has been discharged, which is when the whole cell is at it's lightest. The appropriate mass to normalize by is the mass of the filled electrode, which is notoriously hard to measure directly.

However, they can calculate it by assuming that every electron they collected during discharge was converted to Li2O2 (one of the best-case scenarios). Doing so puts the upper limit on energy density down at ~3500 Wh/kg, because that's the best you can do with pure Li2O2 (against Li). This number can be considered the theoretical upper bound on Li-air energy density.


Now for the good news - the cycling stability results, if they hold up, are incredibly good news and I promise you everyone in the field will be looking at them closely. Still, this is a 10+ year away technology, if we manage to make it work at all...

And again:

discussion about 3500 Wh/kg vs lithium 150 Wh/kg

Unfortunately, even this comparison isn't fair. 3,500 Wh/kg is for just the cathode - no anode and no packaging, which your Li-ion number of 150 Wh/kg does include. IF, and it's a big if, we can package Li-air as well as Li-ion, you'd be looking at a number around 1000 Wh/kg for the final product. This number is a lot more flexible given different assumptions, though, so I tend not to use it as much.

Still, a 5-6x "specific energy" (energy/mass) improvement isn't anything to complain about, if we actually pull it off.

discussion about how energy/volume isn't much better than lithium PDF presentation

I'm pretty sure I was at that talk, thanks for the blast from the past

You are absolutely correct that energy density (that is, energy/volume, as opposed to specific energy) is very important, both for consumer electronics and for automotive applications. That slide/presentation, though, is getting a bit ahead of itself.

First off, the way I'm reading it, they're pitching that Li-Air is far worse than Li-ion (66 L for Li-ion vs 166 L for Li-air). There's a couple of reasons for this, but including a Li-conducting glass to protect the Li metal certainly makes the numbers a lot worse, and I'm not sure it's necessary. They don't justify their volume estimate at all, so I wouldn't read too much into it. There's also been quite a few leaps in understanding just how important electrolytes really are since that presentation.

From a theoretical standpoint, the energy/volume of a Li-air battery could reach a few times that of Li-ion (the best in this regard is Zn-air, which is used today in hearing aid batteries - it's just not rechargeable yet). However, estimating a real value of volumetric energy density is quite tricky, because it ultimately depends very strongly on numbers that aren't well bounded, relating to how thick each electrode can be, how well O2 can be transferred into the battery, and how clean the air entering the battery must be (getting rid of water, if we have to do that, is very, very hard). It also depends on the intended application: very slow discharges can probably get better energy/volume than applications requiring high power.

tl;dr: Theoretically, Li-air beats Li-ion on energy/volume, but not as much as energy/mass, but we really have no idea if we'll be able to reach that in a real system.

[ttxgpfan]

That is a very good and interesting point about charge cycles, capacity, and range.  I have noticed discharge rates follow a similar "curve".

I got an interview with EOS energy storage and wrote a post a month or so back.  Their Zink-air battery is completely new, is about 2.6X as dense as li-ion with some weigh penalty, but at most has the potential for 1/3C charge and discharge.  Pair it with capacitors and it would be pretty effective in something like a Tesla, but I don't see it working in elmotos.  Otherwise than that, I see it as the choice over Li-ion.

Quick math says that in order for elmotos to compete with MotoGP bikes they need 58kWh packs in bikes that weigh the same.  More quick math says that is about 8X the density with the same weight.