OriginOil Achieves Live Algae Oil Extraction Process at Bench Scale
Targeted Growth Increases Algae Oil Content by 400%; Partnering with UOP on Algal Biojet

AIST Developing New Lithium-Air Battery; “Lithium Fuel Cell”

Aist liair1
Long-term discharge curve of the newly developed lithium-air cell. Source: AIST. Click to enlarge.

Researchers at Japan’s AIST (National Institute of Advanced Industrial Science and Technology) are developing a lithium-air cell with a new structure (a set of three different electrolytes) to avoid degradation and performance problems of conventional lithium-air cells.

The newly developed lithium-air cell has shown a continuous cathode discharge capacity of 50,000 mAh g-1 (per unit mass of the carbon, catalyst and binder). By comparison, conventional Li-ion batteries offer 120-150 mAh g-1 (active material + conduction assisting carbon + binder), and conventional lithium-air cells offer 700-3,000 mAh g-1. The research was presented earlier this year at an Electrochemical Society of Japan meeting held in Kyoto.

Lithium-air cells, which dispense with the intercalation cathode of lithium-ion batteries and use a catalytic air cathode in combination with an electrolyte and a lithium anode instead, are attractive because of their theoretically very high energy capacity. (Earlier post.) However, one of the serious problems with lithium-air cells reported to date is that a solid reaction product (Li2O or Li2O2), which is not soluble in organic electrolyte, clogs the air electrode (cathode) in the discharge process. If the air electrode is fully clogged, O2 from atmosphere cannot be reduced any more.

The AIST researchers used an organic electrolyte on the anode (metallic lithium) side and an aqueous electrolyte on the cathode (air) side. The two electrolytes are separated by a solid state electrolyte (lithium super-ion conductor glass film, LISICON) so that the two electrolytic solutions do not intermix. Only lithium ions pass through the solid electrolyte, and the battery reactions proceed smoothly.

Aist liair2   Aist liair3
Schematic diagram of a conventional lithium-air (oxygen) battery. Source: AIST. Click to enlarge. Configuration of new AIST lithium-air cell. Source: AIST. Click to enlarge.

AIST confirmed that the discharge reaction product is not a solid substance like lithium oxide (Li2O), but lithium hydroxide (LiOH), which dissolves in the aqueous electrolyte; clogging of the pores does not occur at the carbon cathode. Furthermore, as water and nitrogen do not pass through the solid electrolyte (the partition wall), there are no unwanted reactions with the metallic lithium anode. During charging, corrosion and degradation of the air electrode is prevented by using another cathode electrode exclusively for charging.

Aist liair4
The new cell uses an exclusive cathode for charging. Source: AIST. Click to enlarge.

Metallic lithium is used as the anode, and an organic electrolyte containing lithium salt is used on the anode side. A lithium-ion solid electrolyte is placed in between the two electrolytic solutions as a partition wall to separate the cathode and anode sides. An alkaline water-soluble gel is used as the aqueous electrolyte for the cathode side and the cathode consists of porous carbon and an inexpensive oxide catalyst.

The discharging reactions proceed as follows:

  1. Reaction at the anode: Li→ Li+ + e-
    Lithium ions dissolve into the organic electrolyte as lithium ions (Li+) and the electrons are fed into the conductor wire. The dissolved lithium ions (Li+) pass through the solid electrolyte into the aqueous electrolyte on the cathode side.

  2. Reaction at the cathode: O2 + 2H2O + 4e- → 4OH-
    Electrons are fed from the conductor wire, and oxygen from the air and the reduction reacts on the surface of catalyst in the porous carbon to produce hydroxyl ions (OH-). They meet with lithium ions (Li+) in the aqueous electrolyte and produce water-soluble lithium hydroxide (LiOH).

The charging reactions proceed as follows:

  1. Reaction at the anode: Li+ + e- → Li
    Electrons are fed from the conductor wire, and lithium ions (Li+) in the aqueous electrolyte of the cathode side pass through the solid electrolyte and reach the surface of the anode where metallic lithium precipitates.

  2. Reaction at the cathode: 4OH- → O2 + 2H2O + 4e-
    Oxygen gas is generated. Generated electrons are fed to the conductor wire.

The new lithium-air batteries allow for continuous operation if the aqueous electrolyte on the cathode side is exchanged and metallic lithium is resupplied to the anode, e.g., by means of cassettes. The researchers say that this concept can be taken as a “lithium fuel cell.” By retrieving LiOH from the aqueous electrolyte in the air electrode, metallic lithium can be recovered easily and reused as fuel.

The researchers suggest that the technology holds great potential for automotive applications. At a filling station, the driver of a vehicle thus equipped could exchange the aqueous electrolyte for the air electrode and refill the metallic lithium for the anode in the form of cassettes, and then continue driving without waiting for batteries to be recharged.

AIST says that the new lithium-air battery needs further technical improvement toward practical use. Generally, there are two directions in this new lithium-air battery research, one is for a rechargeable lithium air battery and the other is for a lithium fuel cell.



Great progress, but I'd be keen to try using sodium instead of lithium.

You'd lose a fair bit in energy density, but it would be mega cheap.


If you use sodium, it'll be much harder to have that porous carbon wall that only lets sodium through.
But then, of course, they'll probably change a little more than just lithium into sodium: most scientific challenges are met one day or another, they'll find a way...

Roger Pham

Perhaps NaOH is too caustic, and Sodium metal is too dangerous, to the point of being outright spontaneously combustible, much more dangerous than lithium.

If this Li-air battery will work, it will be a game-changing development that will make BEV's a dominant propulsion means...being both chargeable at home for daily use and quickly refillable for long trips.


This is absolutely shocking. 50,000 mAh/g is 300 times the capacity of the best Li-Ion battery. My car battery has about 50,000 mAh, but it weighs about 10 kg! Why talk about Sodium? How much could a coffee cup sized, 200 mile lithium-air battery, with no cobalt, cost? This will make EESTOR a forgotten footnote.

AIST has a very big reputation, and unlike EESTOR, they publish. They have to because the Japanese government and industry work so close together. They want electric cars and they won't put up with scientific fraud. They also helped develope the HRP-2 robot -http://www.youtube.com/watch?v=1tiOs0vlJig

The chart shows clearly that maximum capacity depends on a low discharge rate. Therefore, at that rate the battery would have to be bigger to provide enough power. In that case, a shoebox sized battery weighing about 20 kg, would give more than 12,000 miles range at 4kWh/mi(Tesla). Charge it or replace the anode and electrolyte once every year. You wouldn't need to do this on long trips unless your driving cross country.

A close look at the AIST web site press release ("developing" link) at the beginning of the article, shows that they have a complete working battery in the lab. The discharge rate of .1 A/g over 20 days is reasonable. At 3 volts that's about .3 Watts/g. In that case if you want 100 horse power you'll need 7,500 watts/.3W/g = 25 kg (15,000 mi range). The press release indicates the scientific work is finished and they only need to configure the physical shape to provide battery or fuel cell models.


Seems like it has to be a typo....like two too many zeros.
If true, fantastic!

Henry Gibson

Neither lithium of sodium are very dangerous compared to gasoline. Sodium, as proposed, might well become the fuel of the future, but recycled lithium is also possible. A combination of sodium and potassium, can be a liquid fuel at ordinary temperatures.

Electric cars with sodium are being sold now with the ZEBRA battery. Even modern lead batteries are suitable for electric cars. Just get someone to build TATA NANO equivalent electric cars. Range extenders, which even so will be seldom used, allow modern lead batteries to be used. My automobile went zero miles yesterday. ..HG..


This seems well published and there should be plenty of interested parties to replicate the results and perhaps improve on the design.

Very promising. The planet looks slightly more interesting.


The AIST press release says the capacity is only for the cathode. So to get the capacity of a real battery, you have to add the anode, electrolytes, and packaging. The anode is pure Li, which probably will supply as much current as the cathode will take. Assuming this we can calculate how much Li is needed to supply 50,000 A-h to a 1 kg cathode. The discharge formula shows one electron is needed for each atom of Li.

So 50,000 A-h x 3600s/h x 6.242x10e18e/A-s = 1.124 x 10E27e. This equals 1.124x10E27 Li atoms. How much mass is this? First get the number of moles = 1.124x10E27 Li x 1 mole/6.023x10E23 Li = 18.7 moles. The atomic weight of Li is 7, so 18.7moles x 7g/mole = 130g. So we would need 130g of Li anode for each 1 kg of cathode for a total of 1.13kg. It seems to me it would need a lot of electrolyte to transmit or dissolve the LiOH. So add 500g of electrolyte, and about 200 grams for packaging and you have a battery that weighs less than 2kg.

So the real capacity may be closer to 50,000 Ah/2kg = 25,000 Ah/kg. At 3 volts we have 3 x 25,000 Ah/kg = 75,000 Wh/kg. Compare this to an off-the-shelf Zebra battery at only 90 Wh/kg. So instead of the 12,000 miles from a 20kg Li-Air battery (my earlier post), we might get 6,000 and need to charge it once every six months.


The U.K. researchers were looking at
results of up to 4,000 mAh g-1,


so I could believe maybe 5,000 mAh g-1, but not 50,000.


Looks like it really is 50,000.
Following the link by clicking on the word "developing" at the begin of the article.


H2 + O2 --> H2O has (without any loss) an energy density of 142 MJ/kg = 39kWh/kg. That is per kg H2.

That any practical system (obviously with some losses, and not only fuel) could do so much better is hardly believable.


At this time, this super unit could be clasified as a potential, ultra compact, slow release, very high energy e-generator or a highly superior recyclable primary battery but not yet a rechargeable secondary battery nor an ESSU.

However, given its inherent extremely high energy production capabilities, it could become a superior option for future BEVs, specially when coupled with a super capacitor (ESStor?)to supply high initial start up and accelleration power and to recouperate more breaking energy.

As a stand alone replaceable unit (primary battery) it could be ideal for portable communication units, e-books, laptops, iphone, music players, etc.

Wonder if this is not what Toyota was talking about some months ago when they mentionned that current lithium batteries would be replaced by greatly superior technology units within a few years.

ESStor may have serious competition but it may retain an interesting niche market for extremely quick charge/discharge ESSUs with almost unlimited duty cycles.

Which technology will be commercialized first? It looks like ESStor units will hit the market place by late 2010. This highly superior recyclable battery may not be on the market place for another 5+ years.

It looks like BEVs may have all the power and range required to kill ICE vehicles by 2015 on shortly thereafter.

Very interesting decade ahead.


The link states:

"capacity of approximately 9000 mAh/g when it is discharged in the air at a discharge rate of 0.1 A/g"

.1 A/g is a very low discharge rate. They are talking about two different implementations, a rechargeable battery and a lithium air "fuel cell", I think some of the projections are getting mixed up.


“Furthermore, by using an alkaline aqueous solution in place of an alkaline water-soluble gel, continuous discharging up to 20 days at the discharge rate of 0.1 A/g in the air has been realized. The discharge capacity of the cell was approximately 50,000 mAh/g (shown in Fig. 3).” The gel produces 9,000 mAh/g, while the water solution (alkaline aqueous) produces 50,000. I guess it’s because it’s easier for the ions to move around in water than in a gel.

.1A/g isn’t so low. This is 100 A/kg or 1000 A/10kg and 2,000 A/20kg. Although my first post I mistakenly said 1000 hp is 7,500 watts, it’s really 75,000 (as in my second post). However nobody needs 100 continuous hp in a car. The average hp needs to be only about 10 hp. In that case my conclusion is the same. So a 20 kg AIST Li-Air battery would give 10 hp at .1A/g for about 6,000 miles. You might need a small EESTOR super-duper capacitor to provide for accelleration, because the required higher current draw would decrease the capacity according to the long-term discharge curve in the press release. The energy of H2O is a good question and I’d like to know how that works.



Even if you assumed that an average 20 hp were required to keep a fair size vehicle going at 100 to 110 Km/h, the combination of super cap + lithium-air with a potential endurance of 3000 hours could be the ideal power source for most BEVs.

Traveling for 3000 hours @ 100 Km/h = 300 000 Km. Not bad at all between cassettes change. The ESStor type ESSU would keep going for many more cassette changes.

Almost too good to be true.
What would happen to our ICE vehicles if that ever comes around?


Even at 20 hp you still have a fantastic range. My calculations may be rather rough, but even if the AIST Li-Air battery is no better than the "normal" Li-Air battery, it's still about 25 times better than the Li-Ion batteries on the market. The Li-Air battery has been the Holy Grail for many years, but researchers haven't been able to make them work without clogging up the cathode. The fact that AIST researchers solved this problem alone, makes it a fantastic breakthrough.

Roger Pham

The following is a more realistic expection of Li-Air battery, based on theoretical calculation of maximum energy available:

"Theoretical and Practical?
The determination of the theoretical maximum capacity of a Lithium-air battery is complex, and there isn’t a flat statement of fact in the Handbook of Batteries , Third Edition as are many more well developed chemistries. To provide the most accurate value for the maximum capacity, BD asked Dr. Arthur Dobley to provide an expert opinion, which we quote as follows:
“Specific capacity:
n For lithium metal alone 13 kWh/kg.
n For the lithium and air, theoretical, 11,100 Wh/kg, not including the weight of oxygen, and 5,200 Wh/kg including the weight of oxygen. This was checked by calculation and agrees with K.M. Abrahams publication ,JECS 1996.
n For the Lithium air cell, practical, 3,700 Wh/kg, not including the weight of oxygen, and 1,700 Wh/kg with the weight of oxygen. These numbers are predictions and are made with the presumption that 33% of the theoretical energy will be obtained. The battery industry typically obtains 25% to 50% of the theoretical energy (Handbook of Batteries). Metal air batteries are higher in the range. Zinc-air is about 44% (Handbook of Batteries, 3rd Ed. pg 1.12 and 1.16 table and fig).

We selected a conservative 33%. You may quote these numbers above and make any comments with them. The theoretical numbers are similar to the numbers in the ECS 2004 abstract. ( The difference is due to mathematical rounding.)

Below are references you can use:
n The patent by K.M. on non-aqueous Li air: K.M. Abraham and Z. Jiang, U.S. Patent 5,510,209 (1996)
n The first paper on non-aqueous Li air: K.M. Abraham and Z. Jiang, J. Electrochemical Soc., 143, 1 (1996)
n The second paper on non-aqueous Li air: K.M. Abraham, Z. Jiang, and B. Carroll, Chem. Mater., 9 1978-1988 (1997)
n Read, J. J. Electrochem. Soc. 2002, 149, A1190-A1195
n Read, J.; Mutolo, K.; Ervin, M.; Behl, W.; Wolfenstein, J.; Drieger, A.;Foster, D. J. Electrochem. Soc. 2003, 150, A1351-A1356.
n Dobley, A.; DiCarlo, J.; Abraham, K.M. “Non-aqueous Lithium-Air Batteries with an Advanced Cathode Structure” 41st Power Sources Conference Proceedings, Philadelphia, PA, 2004, 61.
n Handbook of Batteries, 3rd ed.; Linden, D.; Reddy, T.B., Eds.; McGraw-Hill: New York, 2002.

BD will err on the conservative side and use the 5,200 Wh/kg theoretical value which includes the weight of oxygen and the 1,700 Wh/kg practical value, realizing that production cells may be something less. "


One must remember that due to the low discharge current of Li-Air battery, additional Li-ion battery or Ultra-capacitor will be needed for acceleration and braking energy recuperation.


If in practice, lithum-air (batteries or cells) have a one order of magnitude better energy density than the current Li-on batteries, it could, when coupled with super-caps or EEStor's ESSUs, be the first ideal power unit for extended range BEVs.

Let's hope that both units(+ many other similar technologies) can be mass produced by 2015.


Of course I should have considered the theoretical capacity of Li. That is about 13,000 Wh/kg or 3,863 mAh/g. I think what I did earlier was a stoichiometric analysis (moles, etc.), but physical properties obviously limit the chemical kinetics. I don;t know why you would add the weight of oxygen if it's supplied by the air. So, if you want 50,000 mAh/g from a block of Li, you need to divide it by 3,962 = 13g. So a battery with 1kg of cathode needs 13 kg of anode.

This situation remindes me of the Silicon nanorods that Dr. Qui developed at Stanford. They can soak up 10 times as much Li as carbon anodes. What he failed to mention is that the anode comprizes only about 20% of the weight of the battery. So even if the anode weight was reduced to zero, the battery capacity goes up only about 25%. He told everyone that it could go up "several fold" - to me a deliberate deception.

For the AIST Li-Air battery, you'll need the electrolyte and packaging, say another 6 kg. Very roughly, the total mass is 20kg. So 3V x 50,000 Ah/20kg = 7,500 Wh/kg for a Li-Air battery. This is closer to the capacity that other researchers are claiming, except their anodes are getting clogged up. If you take away two thirds for a "practical" battery, you still have 2,500 Wh/kg. This is still about 15 times the best Li-Ion battery. And it would be about 15 times cheaper per kWh.

Range for a 20 kg battery would be about (2.5kWh/kg) x 20kg x 4mi/kWh = 200 mi or 320 km, enough for five days of driving for the average person. It still gives you 10 average hp and if you need 20 hp, get a 40 kg battery and a 400 mi range.


If the 20 kg battery has 1 kg of cathode, then we can get 20,000 g x .1A/g = 2000 A. And 2,000A x 3V = 6,000 Watts = 8 hp. So a 40-50 kg battery, with ultra capacitor would work.


I should have quit while I was ahead. For 1 kg of cathode in a 20 kg battery, we have 1,000g x .1A/g x 3V = 900 W = 1.2 hp. So if you want 10 hp you need 7,500 W/hp/900W/20kg = 166 kg. That's a lot of weight, but you get 2,000 cheap miles.


It is cathode.
No performance about anode is reported. The anode is more critical for Li-air battery.


One problem with Li-Air is the low current required, about 0.1 A/g of cathode, to maintain high energy capacity. However, it's interesting to look at the EV-1 performance -http://avt.inel.gov/pdf/fsev/eva/genmot.pdf

The EV-1 was a 2 seat, 3000 lb car, with lead acid batteries. It had a 90 mile range at a constant 60 mph! This required an average power of 10 kw = 13 hp. For the normal driving cycle, the range was 78 miles at average power of 4 kw = 5.5 hp. No wonder people liked these cars!

Therefore, a Li-Air battery would need to put out an average of 10 hp for good overall performance. If it has 33% of theoretical efficiency, 1 kg of cathode requires 13 kg of anode (Lithium). Including electrolyte and packaging could add another 6 kg. For 1 kg of cathode in a 20 kg battery, we have 1,000g x .1A/g x 3V = 900 W = 1.2 hp. So if you want 10 hp, you need 7,500 W/hp/900W/20kg = 166 kg. At 2.5 kWh/Kg, that's 415 kwh!!

That's not bad for weight and you get 2,000 cheap miles @ 5 mi/kWh. If the driving cycle needs more frequent acceleration, the range might be reduced to 1,500. This is 1.5 months of driving for the average person. So 50 full charge cycles would last about 6 years. But, the battery would last a lot longer if the charge was maintained.

You wouldn't want to discharge it completely anyway because to get max capacity, you need to charge it slowly. A 2,000 mile battery would take (166,000 g/ 20)*.3 W/g = 2,500 Watts maximum. 400 kwh/2 kw = 200 hours!

The comments to this entry are closed.