Nano aluminium offers fuel cells on demand – just add water

From New Scientist


Hydrogen could provide an alternative to battery power

Tomohiro Ohsumi/Getty

By David Hambling

The accidental discovery of a novel aluminium alloy that reacts with water in a highly unusual way may be the first step to reviving the struggling hydrogen economy. It could offer a convenient and portable source of hydrogen for fuel cells and other applications, potentially transforming the energy market and providing an alternative to batteries and liquid fuels.

“The important aspect of the approach is that it lets you make very compact systems,” says Anthony Kucernak, who studies fuel cells at Imperial College London and wasn’t involved with the research. “That would be very useful for systems which need to be very light or operate for long periods on hydrogen, where the use of hydrogen stored in a cylinder is prohibitive.”

The discovery came in January, when researchers at the US Army Research Laboratory at Aberdeen Proving Ground, Maryland, were working on a new, high-strength alloy, says physicist Anit Giri. When they poured water on it during routine testing, it started bubbling as it gave off hydrogen.

That doesn’t normally happen to aluminium. Usually, when exposed to water, it quickly oxidises, forming a protective barrier that puts a stop to any further reaction. But this alloy just kept reacting. The team had stumbled across the solution to a decades-old problem.

Hydrogen has long been touted as a clean, green fuel, but it is difficult to store and move around because of its bulk. “The problem with hydrogen is always transportation and pressurisation,” says Giri.

Slow reaction

If aluminium could be made to effectively react with water, it would mean hydrogen on demand. Unlike hydrogen, aluminium and water are easy to carry – and both are stable. But previous attempts to drive the reaction required high temperatures or catalysts, and were slow: obtaining the hydrogen took hours and was around 50 per cent efficient.

The new alloy, which the team is in the process of patenting, is made of a dense powder of micron-scale grains of aluminum and one or more other metals arranged in a particular nanostructure. Adding water to the mix produces aluminium oxide or hydroxide and hydrogen – lots of it. “Ours does it to nearly 100 per cent efficiency in less than 3 minutes,” says team leader Scott Grendahl. Moreover, the new material offers at least an order of magnitude more energy than lithium batteries of the same weight. And unlike batteries, it can remain stable and ready for use indefinitely.


Full-size, hydrogen-powered tanks might also be an option

U.S. Army photo by David McNally

Read the Full Article Here.



newest oldest most voted
Notify of

Sounds cool. Anymore information available?

Bemused Bill

Sounds cool? Sounds like you are going to have to produce aluminum…a massively intensive coal-fired guzzle of electricity. Sounds like it will be hugely expensive, use much more resources and be much more ‘polluting’ for the same result as ordinary cheap fossil fuels.
It might have its uses in orgs. like the military where money is no object for a performance upgrade? But it will never compete domestically without massive subsidies and will most likely put more shock-horror Co2 into the atmosphere anyway…what am I saying, the watermelons will love it!

Old England

Maybe but with lithium based batteries releasing 150 – 200 kg of CO2 in manufacture per kWh of storage capacity it may be a superior alternative.
A Tesla 100 kWh battery releases 15 – 20 Tonnes of CO2 in manufacture – about the same as 16 years driving a petrol or diesel car!! (IVL study for Swedish Government, published 4 weeks ago).

The aluminum oxide produced is somewhat less expensive to recycle back into Al, but it does take a lot of electricity and the process produces CO2 and while this isn’t really a bad consequence, the greenies won’t understand and will surely object once they figure out the fuel is Al and water and not just water alone.
If you had an infinite supply of real cheap electricity, Al can be a very safe and economical way to store energy, although I prefer releasing its energy by burning it with Ferric Oxide.


And unlike batteries, it can remain stable and ready for use indefinitely.

… and unlike batteries it can not be recharged.
“Old England”:

Maybe but with lithium based batteries releasing 150 – 200 kg of CO2 in manufacture per kWh of storage capacity it may be a superior alternative.

Any comparison to batteries is irrelevant since it is NOT a battery. It is a one time process.
Like any other source of hydrogen you need to ask what the orginal energy source is . This material , even the basic aluminium does not exist as a existing resource. Bauxite is plentiful but needs reducing to make aluminium which you will later oxidise using water.
So the REAL source of energy here is whatever you use to smelt the bauxite, plus whatever other processing is needed to get the nano alloy.
It will obviously be less efficient than electrolysis of water. So is it any safer?
This product reacts with water by not forming an oxide layer. How does it react with air and/or water vapour in air? Fine aluminium powder is one of key ingredients of thermite. I’m not sure driving around with 20kg of thermite as a fuel source would be a good idea.


In welding large iron together like rail road rails. The thermite requires a magnesium strip to be lit with a torch to get it hot enough to cause the chain reaction.

Old England

@ greg – my comparison was of the CO2 emissions – in response to ‘BB’s comment “Sounds like it will be hugely expensive, use much more resources and be much more ‘polluting’ for the same result as ordinary cheap fossil fuels.” – polluting in the sense of CO2 emissions — But it may be Less “polluting” in that sense than electric cars, i.e. releasing Less CO2 during manufacture …. and Despite “polluting” being the antithesis of what CO2 means to life on earth.
I disagree with your comment that “Any comparison to batteries is irrelevant since it is NOT a battery.” – it is highly relevant when discussing ‘green’ ‘CO2-reducing’ energy sources for automotive power and how much “pollution” (plant food) they produce.
And I’m sorry that you object to electric cars being exposed for what they mean for CO2 emissions. Where a “green dream” proposes alternative energy or power sources to “reduce” CO2 emissions it is important to see if that is correct or not – it typically isn’t.

Mark - Helsinki

explosive powder that also passes the brain barrier, yay.

Crispin in Waterloo but really in Ulaanbaatar

Bemused Bill
Yes making aluminum needs a lot of energy, but once refined, it takes far less to recycle it.

Recycling aluminum is less energy intensive, as long as it is mostly metal and a little oxide (mostly only the skin). In the case of the one-time hydrogen cell, all Al is tranformed into oxide and hydroxide, thus back to the full energy intensity…

Crispin in Waterloo but really in Ulaanbaatar

Yeah, well…they don’t have to separate the ore from the junk around it. I think it is a little disingenuous to say it uses the same amount of energy as digging and processing new ore.
I look forward to a system analysis of the energy loop for this technology. If nearly all the energy put into removing the O2 is really all given back in the form of liberated hydrogen (and there are sound reasons why it should be possible) then a storage system would really different in the future.
I know there have been cell phones demonstrated that ran on ethanol. It contained a tiny fuel cell generating electricity and heat. I wonder how much heat is generated in the hydrogen release.
With wind turbines, in particular, generating trash electricity that must be ‘dealt with’ we now need a way to expel O2 from AlO2 on a variable and ad hoc basis. If that can be done, any renewable generator could process ‘powder’ from one form to another nearby and it can be carted away or run through a generator at the same site.
At ten times the energy density, a ‘container’ could store an enormous amount of energy in stable powder form.


Crispin, a few percent less power. Not a big difference.

Clyde Spencer

Crispin in Waterloo but really in whocares,
You said, “Yes making aluminum needs a lot of energy, but once refined, it takes far less to recycle it.” What you seem to have missed is that it turns the aluminum alloy into in a hydroxide, which is pretty close to what the original ore was. That is, it will take a lot more energy than simply melting an aluminum beverage can of the same weight.

The original article continues with, “…it can be made from scrap aluminium (sic), which is relatively cheap.”
So, how much energy is consumed to turn scrap aluminum into this alloy – what is the EROI and ROI?


Think Iceland my friend

I think you might be misunderstanding. And so may I. I don’t hear that this is using aluminum as a fuel, but as the matrix for a fuel cell reaction. One might say a catalyst.
It isn’t clear from the article which is correct.

I think you might be misunderstanding. And so may I. I don’t hear that this is using aluminum as a fuel, but as the matrix for a fuel cell reaction. One might say a catalyst.
It isn’t clear from the article which is correct.
Also, others in this particular thread are talking about a one-time use. Nothing in this article says that. Some people’s reading comprehension is really bad.


Fantastic. Scratch Tesla boot liver. Add water and electrodes. Get car to nearest power point when flat.

Kenneth Mitchell

Free hydrogen sounds really good – until you realize that in order for it to be practical, you’d have to compress it and bottle it. Because the H2 isn’t going to fizz out of the water and into the cryo tank by itself.

Javert Chip

So what happens to the liberated O? Lots of that stuff hanging around isn’t good.

Kenneth Mitchell

The liberated O2 is always useful. You could go ahead and bottle that, too.

Curious George

There is no liberated O. Al forms Al2O3, liberating H.

Crispin in Waterloo but really in Ulaanbaatar

Thanks for that detail Curious.
I didn’t look up what it made and assumed it was simpler.
The liberation of O2 comes from reversing the process and re-using the Al.
As to the liberated H2, the idea is not to bottle and store it but to use it immediately.


Since there is no such thing as a free lunch, the absolute best you can hope for is that the energy to return the aluminum to a pure metalic form will equal the amount of energy released by oxidizing it.
In reality, since nothing is 100% efficient, it will take a lot more energy to recycle the aluminum than you get from this reaction.

Pop Piasa

I would, however, like to try it as a booster for the diesel engine in my truck.

D. J. Hawkins

Unless it’s consumed as generated. Then you might be looking at 30-50 psig which you could easily accommodate in the generating chamber.

Kenneth Mitchell

The 3rd law of Thermodynamics says that it isn’t going to be practical to consume the hydrogen at the time it’s generated. No matter how good the system might be, it’s not a perpetual motion machine. The purpose of generating hydrogen is to store it and use it later on, either in a fuel cell or in some combustion process.

D. J. Hawkins

So, I guess they weren’t really running that toy tank after all???

Kenneth Mitchell

D.J., overthrowing the 3rd law of Thermodynamics would require re-writing most of what we know about physics and chemistry. I’m OK with that, but in this case “Extraordinary claims require extraordinary evidence.” I’m not _really_ from Missouri, but, “Show me!”


Kenneth Mitchell, what is the difference between using hydrogen at the time it’s generated and storing it for use later on. If you can use it later, you can use it now, as long as you can control the reaction so it delivers the hydrogen you need at the time you need it.

D. J. Hawkins

Further, how exactly does the entropy of a well-defined system at absolute zero enter into this discussion??? Nobody is getting something for nothing. Energy goes into making the aluminum alloy, about a quarter is available as the water-alloy reaction goes to completion, the net entropy of the universe increases. Maxwell’s demon doesn’t have to show up for work.


The system doesn’t supply free hydrogen. The aluminum is consumed.
There has been a lot of work on aluminum for energy storage. link

… an electric vehicle with aluminium batteries has the potential for up to eight times the range of a lithium-ion battery with a significantly lower total weight … link

They’ve been working on it for a while and there’s always something that makes it impractical. Based on that, I wouldn’t bet the farm on aluminum for hydrogen generation.

Walter Sobchak

This is really silly.
1. You take Al2O3 and turn it into Aluminum, which is very energy intensive.
2. You turn the Aluminum into a special form of powdered Aluminum using even more energy.
3. You combine the powdered Aluminum with water to produce Hydrogen (an exothermic process I believe), and Al2O3.
3. You either burn the hydrogen in a heat engine or run it through a fuel cell.
That is a four step process to turn electricity into electricity, while wasting about 90% of the input energy. Even environmentalists can’t be that stupid, can they?


You’re putting too much faith in environmentalist being intelligent when their ends justifies the means. They will sell this to the ignorant masses as the cleanest energy ever invented. Ignoring how much energy is used to produce it…that they’ll not tell the masses about.


When you understand that after crushing Bauxite the first step is to desolve the Aluminum Oxide into a chemical solution. That is where the gangue is filtered out. There isn’t any need to go to the smelting of it to aluminum that is the energy intrusive part. When doing nanoparticles alloys.


The stuff you get out of Bauxite is Al2O3. If you know a method for turning aluminium oxide powder into aluminium alloy powder without massive energy input I am very interested.

Crispin in Waterloo but really in Ulaanbaatar

No one is claiming it doesn’t take a lot of energy to create the stored energy.
The fact is it is a very energy dense product. That is what makes it exciting, not come claim for free energy. Energy from coal-fired power plants can operate cars or trucks with this sort of storage. Instead of powering a car for 300 miles with a battery it could go 2500 miles between charges or exchange. That is impressive. It is a serious competitor to ceramic super capacitors. I would buy such a car. I have lots of nuclear power available to me in Waterloo. Wind too, if it is blowing.
It would revolutionise Formula E racing and blow the doors off a gasoline-powered F1 car.


Where did you get the notion that this thing could go 2500 miles between recharge/replacement? Don’t forget to that you need to store enough water to react all that aluminum as well.
Secondly have you thought about the practical problems with this “battery”?
If we operate an unpressurized mode, then the hydrogen has to be produced as needed.
Think of it. Sitting at a stop light, light turns green.
1) Water gets sprayed onto the aluminum. Wait for reaction to get going.
2) Released hydrogen must somehow be piped to the cylinders in the exact amount needed.
That’s enough of a delay to make driving difficult, perhaps even dangerous.
Beyond that, since you are creating hydrogen as needed, the aluminum has to be stored dry. What happens to it if oxygen gets into the canister where the aluminum is being stored?

@Mark: On demand may be a misnomer. I expect that there easily could be a reservoir tank to hold 15-20 seconds worth of Hydrogen to smooth out just such situations. Would that make it no longer an ‘On Demand’ process? I’m not in to hair splitting.


Hydrogen cannot be liquefied by pressurizing. All storage of hydrogen is at very low temperatures requiring special metallurgy since normal steel is brittle as glass at these temperatures.
For vehicles the storage tank must be in thousands of psi to provide sufficient energy


Which is why the article said it is Hydrogen on demand by the 2 components being transported separately and mixed as needed.


Ammonia as a carrier of hydrogen is looking like a viable alternative.

Compared to hydrogen as a fuel, ammonia is much more energy efficient, and it would be a much lower cost to produce, store, and deliver hydrogen as ammonia than as compressed and/or cryogenic hydrogen.[60] The conversion of ammonia to hydrogen via the sodium-amide process,[68] either as a catalyst for combustion or as fuel for a proton exchange membrane fuel cell,[60] is another possibility. Conversion to hydrogen would allow the storage of hydrogen at nearly 18 wt% compared to ~5% for gaseous hydrogen under pressure. link

There has been quite a bit of research on using ammonia plus gasoline or diesel in internal combustion engines. Here’s an example. It reduces the use of diesel by about half.
Fracking has made any transportation fuel other than oil non economic. That includes ammonia. With more expensive petroleum, ammonia, produced from renewable electricity, could become viable relatively quickly.

chris y

Hydrogen loaded carbon nanorods are an excellent carrier of hydrogen, and have additional energy stored in the carbon bonds. Easily handled liquid over a wide range of temperatures, easily stored in low cost tanks.
Octane, for example…

someone want to check my maths?
If we had aluminium metal lying around it’d be great but here’s the thing with chemistry and physics.. the values of energy and the rules of energy conservation are well known and have to be taken into consideration. Now sure if you have a definate need to use hydrogen it’s fine, but if the idea is to save energy we have to look at the overall energy budget – see below (chemistry warning 😉
We start looking at converting the raw ore to aluminium
Reaction 1 : The reduction of bauxite to Aluminium
2. Al2O3 –> 4. Al + 3O2
this reaction is the one that reduced the ore to Aluminium metal and requires breaking the latice energy, demanding an huge input of electricity – currently smelters use about 13 kilowatt hours (46.8 megajoules) of electrical energy to produce one kilogram of aluminium; the worldwide average is closer to 15 kWh/kg (54 MJ/kg). I’ll use the worldide value in calculations..
The reaction these guys chasing to make hydrogen is (reaction2 ):
2. Al + 2. H2O –> 2. AlOH + H2
so starting with 1 kilo of Aluminium that cost us 54Mj or electricity we need to find out how much energy the hydrogen released will give us when burned
(Aluminium’s atomic weight is 27g/mol meaning 1 kilo is about 37 mols of aluminium )
(Hydrogen is atomic weight is 1g so from 1 kg of Aluminium we’d get 37 mols or 37 grams of hydrogen.)
then we want to burn it to use the stuff in (reaction 3)
2H2(g) + O2(g) –> 2H2O(g) + energy – the energy released is 142 Mj/kg
(142MJ/kg is the whiz bang goal of hydrogen as an energy source, for comparison petroleum is 57MJ/kg)
but we’ve not got 1 kg of hydrogen, we’ve 37 grams so dividing 1000 g / 37 grams tells us what fraction of energy we’d get of the 142MJ
we’d get .. 5.26MJ
So for a cost of 54MJ of energy to convert aluminium ore to aluminium, we’d have 5.26MJ of energy released burning the Hydrogen which was released when water reacts with Aluminium
this is why so often in chemical circles hydrogen is jokingly referred to as ‘Hydrogen is the energy source of the future.. and always will be’ – it’s just too damned hard to break the chemical bonds and always incurs a cost greater than the energy it delivers. Petroleum is numero uno because it’s just ‘there’ lying around and basically free. Once you start ‘making energy’ there’s always a cost and it’s usually at a greater expense.
This idea above basically means you’d waste 9 units of energy to produce 1 unit of Hydrogen power – that’s a big cost 🙁 Or putting it another way, you could use 1 kilogram of petrol (energy density 57MJ/kg) to make 37 hrams of hydrogen with 1/10th the energy of the petrol you just burned.


Frankly I’m terrible at mathematics and like how you broke that down.

Phil Rae

I agree with your general premise but don’t agree with the stoichiometry…….what is Al(OH)? Aluminium hydroxide is Al(OH)3 – although even that is just a simplification since there are many bizarre aluminium compounds. Al(OH)3 is just a hydrated form of alumina (Al2O3) which is found commercially as bauxite.
Al2O3 + 3H2O –> 2 Al(OH)3
So, my simpler stoichiometric expression (see reply to Willis below) is the better one to use IMHO
2Al + 3H2O –> Al2O3 + 3H2
or, if you prefer, you can use
2Al + 6H2O –> 2 Al(OH)3 + 3H2
But, any way you slice it or dice it, the correct stochiometric ratio is 2 moles aluminium yields 3 moles hydrogen, which on a mass basis is a ratio of 9:1 i.e. 9kg of aluminium will yield 1kg of hydrogen assuming quantitative conversion. That works out at an output:input ratio of 25-33% depending on whether we use your figure or Willis’ for energy required for aluminium manufacture.


“This idea above basically means you’d waste 9 units of energy to produce 1 unit of Hydrogen power”
1 unit Hydrogen input. Fuel cells systems are about 40% efficient. Not sure where they are today, but consumption was around 13 standard liter per minute per kW.
It might be a whole lot smarter, and cheaper, to bypass the aluminum and charge the relatively large battery that will already be in that FCEV to handle the energy transients that the on-demand system can’t.


(Aluminium’s atomic weight is 27g/mol meaning 1 kilo is about 37 mols of aluminium )
(Hydrogen is atomic weight is 1g so from 1 kg of Aluminium we’d get 37 mols or 37 grams of hydrogen.)

Out by a factor of two, molecular weight of Hydrogen (H2) is 2 so from each kg of Al you’d get 74 gas of H2.

Crispin in Waterloo but really in Ulaanbaatar

I have a minor quibble which is that the energy released you gave is the HHV, which is not useful energy. The LHV of H2 is 120 mJ/kg so this changes your result if you are comparing it with KWH consumed.
The second quibble is bigger: the production cost of aluminum from mining is not the appropriate denominator. The cost of recycling the AL2O3 to AL is, and it is not equal to producing new material. This has to be acknowledged from a logic point of view. You cannot find the correct answer and % returned without considering the appropriate cycle.
The oxidised aluminum is in a highly accessible form. It just needs energy to be added. An innovation may reduce the energy needed to releasing the oxygen and using it for combustion which would reduce the energy input considerably depending on what you oxidised. It would transfer the oxidation to a different, cheap, disposable material like carbon or iron.
If is correct that the amount of energy needed to remove the oxygen is large, and that is not important on its own. If the returned energy is large, it is a ‘so what?’ moment.

Phil Rae

Phil…… not Phil Rae
Please get the stoichiometry right Phil. Your number ISN’T correct! You get 111 grams of hydrogen for each kilogram of aluminium. So the ratio, as I already pointed out twice on this thread, is 9 kg of Al to produce 1 kg of H2. Chemistry works!
Having said that, this idea is still a complete waste of time. Gasoline is cheaper, has more than triple the energy density and is much more useful as a fuel.

thanks Phil, I see the error (H2 not H) my bad., Crispin, you’ve got top raise the temp to over 2000C to extract aluminium that way -costing more energy again which is why electrolysis is used.

Actually this got me thinking, a friend sent me this some days back and I jumped on the cost of this energy – more goes in than we get out but he raised the point that if it was hydrogen *specifically* that was needed it was still viable. Not sure what you’d need hydrogen for, but let’s say you do (zeppelins ?)
When I was a kid we used to amuse ourselves stuffing aluminium scraps into coke bottles, we’d pour in caustic soda solution, screw the cap on and run like hell.. The hydrogen released built up pressure pretty rapidly and there was no need for exotic alloys !
Of course the usual disclaimers : don’t do this at home kids. Still, we also used the same cheap process to fill balloons and it’s probably quite a reasonable way to get hydrogen fast. Now days I’d modify this process by passing the gas through sulfuric acid to neutralize any hydroxides lifted by the steam that’s generated, drying the gas at the same time. Lots of ice water too – the reaction is very exothermic.


No – you just feed the H2 into a fuel cell stack where it generates electricity and pure water. No need to recompress or store .. just run the reaction to generate H2 as you need it.
All FCVs use storage batteries to moderate the production and storage of electricity by the fuel cell. Excess above what is needed immediately to drive the wheels and power lights and electronics on board the vehicle is stored in the battery, which acts to supply large amounts of electricity on demand such as high acceleration

Tom Halla

Interesting. But how much does that puppy cost, and is it actually scalable? And what is that “other metal”?

Rhoda R

And what density of H are we talking about – usable densities or little puffs that have to be compressed.


We’ll have to wait and see but an energy technological breakthrough would be welcome. Then we ban get rid of ethanol.

Curious George

Nothing new. Put a piece of alufoil in water and add some lye. It is scalable. All we need now is aluminum-growing trees, not hydrogen-growing trees.

Gunga din

Aluminum foilage?


Yes, aluminum foliage; that used to be popular on Christmas trees.


A basis for a long-range, quiet, “conventional” submarine? Assuming that the bubbling isn’t too noisy.

Randy in Ridgecrest

Well, so much for the high strength aspect – highly reactive aluminum! So, you end up with a pile of alumina, maybe there is a recycling use for that

Pat Frank

Aluminum is made by electrolyzing bauxite (Al2O3). You only get out the electrons you’ve already put in.
Thermodynamics says that the cycle (bauxite –> Al metal –> bauxite again plus H2) will take net energy. You’d be better off generating the hydrogen directly, and leaving out the aluminum step.
The chemistry and the metallurgy are interesting. The whole story about energy production is a publicity crock.

True. But that is true for all batteries. The point is not that this produces new energy, but rather it is a way to store energy safely and for extended periods. Periods longer than the parts of the engine itself will last.

It takes about 13 kilowatt hours (46.8 megajoules) of electrical energy to produce one kilogram of aluminium in a modern efficient smelter, more for older units. So aluminum has lots of latent energy in it.
And yes, this would work as a one-use battery …
So let’s assume that all of the Al is converted to AlO via the reaction
Al + H2O —> AlO + H2
So it seems that one mole of aluminum will yield two moles of elemental hydrogen. Hydrogen has a molecular weight of 1, and aluminum is 27. So 27 grams of aluminum will yield 2 grams of H2.
Burning hydrogen releases about 140 kilojoules per gram. 27 grams of aluminum will yield about 280 KJ of hydrogen energy. A kilo of aluminum will yield 1000 / 27 * 280e3 = 10.37 megajoules of energy
Now, one kilowatt-hour is 3.6 megajoules of energy. So at the end, we get about 10.37 / 3.6 = 3 kilowatt hours of energy back out of the aluminum … and we put in 13 KWhr to get the aluminum.
So yes, there will be specialty uses for this procedure, but it’s a very poor battery that only gives back about a quarter of the energy you put into it …

Jeff in Calgary

To make it worse, your calculation assumes smelting typical aluminium. I suspect that this fancy Nano Aluminium alloy would take even more energy. And now you are hauling around all this Aluminium plus the non-reactive other metal in the alloy. Not only is it not energy efficient, it may end up being useless when scaled to size.


Willis are you assuming too much about the smelter? Aluminum is produced by converting ore to aluminum oxide, which would be the starting point here.


Mike N,
Wrong and wrong.
An aluminium REFINERY produces alumina (aluminium oxide) from bauxite (ore).
An aluminium SMELTER produces aluminium metal from alumina.

Crispin in Waterloo but really in Ulaanbaatar

Accept the correction and restate your point. You are correct.


Even if it’s a poor battery, could it be useful when paired with intermittent power sources, such as wind or solar? The energy produced from such sources could be used to charge an array of batteries, which could then be used to produce aluminum during times the wind isn’t blowing or the sun isn’t shining. It involves exchanging one type of battery for another, even poorer one. But the aluminum can be stored indefinitely in unlimited quantities, and does not lose it’s charge over time (as long as you can keep it dry.) I guess it all depends on the cost.

Richard Bell

While ‘refreshing’ exhausted aluminum pellets does help with the exploitation of intermittent supply, it is much more effective for the exploitation of intermittent demand.
The following assumes decisions are dictated by price, not politics:
A difficult problem in power systems is scheduling all the generating assets to supply all of the loads (Unit Commitment). Many units can be started and connected to the grid with little delay and only cost money while they are producing power. They are typically driven by gas turbines or diesel engines. Many units can only be started and connected to the grid after an extended start up process and require an extended shutdown process after their scheduled commitment is over. For planning purposes, these units have a starting delay and a minimum scheduled time in the plan. These are the large thermal units, steam from coal, steam from nuclear, and combined cycle plants (natural gas fired gas turbines with a heat recovery steam generator in the exhaust stack). Nuclear powered generators with reactors susceptible Xenon poisoning have an additional minimum period where they cannot be scheduled after completing a scheduled commitment. Hydraulic turbines have a short start up delay and a short shutdown process. Wind and solar can only be scheduled within the limits weather prediction, but errors in prediction will cost the renewable operators, whether they produce too little power or too much.
Now we add the alumina to aluminum converters that accept the exhausted aluminum fuel pellets and refreshes them for another pass through a fuel cell:
The converter plant operators design their converter plants to sacrifice some overall efficiency to maximize their flexibility to run at fractional capacities, as interruptible power contracts have lower electricity costs than regular contracts. The converter plants run at their highest capacities during the periods when, without variable demand of converter plants, the largest, cheapest generating stations would be forced to shut down. As other demands on the grid increase, the converter plant operators are asked to scale back their power demands. So long as the average demand for aluminum pellets is matched by the average production and the stock of aluminum pellets neither completely fills up nor completely empties, all electricity can be produced by the largest and cheapest sources so high cost intermittents like wind and solar are only economical for off-grid applications.


“Willis Eschenbach
August 7, 2017 at 4:38 pm
It takes about 13 kilowatt hours (46.8 megajoules) of electrical energy to produce one kilogram of aluminium in a modern efficient smelter, more for older units. So aluminum has lots of latent energy in it.
And yes, this would work as a one-use battery … ”
But could an easily recycled battery.
And terms of energy, 13 kw hour could be as little 5 cent per Kw hour, but call it 10 cents
1.30 per Kg in terms of energy cost. In terms of consumer cost call $3 per kg.
So 1 kg aL is about size of 1 lb of butter, and waste will bigger than that.
When fill up, you want get the butter, and dump off the waste- or waste more or problem
in terms of mass and volume. And also have the mass of water- later.
“So let’s assume that all of the Al is converted to AlO via the reaction
Al + H2O —> AlO + H2
So it seems that one mole of aluminum will yield two moles of elemental hydrogen. Hydrogen has a molecular weight of 1, and aluminum is 27. So 27 grams of aluminum will yield 2 grams of H2.
Burning hydrogen releases about 140 kilojoules per gram. 27 grams of aluminum will yield about 280 KJ of hydrogen energy. A kilo of aluminum will yield 1000 / 27 * 280e3 = 10.37 megajoules of energy
Now, one kilowatt-hour is 3.6 megajoules of energy. So at the end, we get about 10.37 / 3.6 = 3 kilowatt hours of energy back out of the aluminum … and we put in 13 KWhr to get the aluminum.”
Well need about 60 Kw hour. or 20 1 lb butter volume- fits in breadbox. Weighs: 20 kg..
Cost $60. Waste is more massive and larger volume.
How much water. 9 kg of water has 1 kg of H2 and 8 kg of O2.
Using up 20,000 gram Al gives 1482 grams H2, so need 13,3 kg of water, and add 11.8 of O2
to waste. Or 20 kg of Al becomes 31.8 kg of waste.
The details of waste seems problematic. If difficult the mass and volume involved, may suggest
having more. Say having 5 times as much put in “blackbox” and replace blackbox- rather re-fill it.
Or more like oil change than gasoline fill up. Or not self serve gasoline. But could also get pallet
delivery, if one has the tools. And one could get a month supply and change blackbox every week. Or maintenance every few days [remove waste, mostly] and some of overall once an month. You getting by the ton and getting rid of waste by the ton.
Anyways it is expensive, but it seems more recyclable, but requires less space and less mass-
and if run out of power, it may not require towing it or recharging it

Phil Rae

In general, I agree with your post that this isn’t a particularly big bang for your buck but it’s not actually as bad as you stated.
From a stoichiometric perspective, you should get 3 moles of H2 from from 2 moles of Al rather than your oversimplified 1 for 1 (Aluminium has a valency of +3 normally, oxygen of -2). Note also, that the 27 and 1 mentioned in your post are actually “atomic weights” rather than “molecular weights” – hydrogen has a molecular weight of 2, actually. Anyway, the equation should be:
2Al + 3H2O —> Al2O3 + 3H2
So, 54g of aluminium provides 6g of H2 (assuming 100% conversion) and, based on the energy equivalency you provided above, this means you get ~15.6 megajoules worth of energy (in terms of hydrogen) from the decomposition of 1 kg of aluminium. This equates to 4.3 kWh of output from an initial input of 13kWh to produce the aluminium in the first place so the ratio is about a third back rather than the quarter mentioned in your post.
However, all of this needs to be compared to a (US) gallon of gasoline which provides ~33kWh of energy at very low cost. It’s easy to see why hydrocarbons are such wonderful fuels and why we have used to provide the essential energy we need and should continue to use them for the foreseeable future.

Price of aluminum = $0.86/lb = $1.89/kg
Energy in 1 kg Al = 15.6 MJ
Energy cost of hydrogen from Al = 1.89/15.6 = $0.12/MJ
Price of natural gas = $2.77/MMbtu = $0.0026/MJ
Energy from this new amazing technology is just 46x more expensive than natural gas! Anymore bright ideas from mad scientists?comment image


Much of the “wasted heat” is given off in the exothermic reaction itself:
2Al + 3H2O —> Al2O3 + 3H2 + ΔH
where ΔH is the enthalpy change.
Equation is similar to that of elemental sodium and water:
2Na(s) + 2H2O → 2NaOH(aq) + H2(g) + ΔH
where this ΔH represents an explosion.

Old England

It might seem at first glance that “Free” renewable enrgy such as solar or wind could be used to smelt the aluminium, and thus take energy costs and energy balance out of the equation – that is until you realise that an aluminium smelter needs constant power 24/7, power that is not reliably available from wind or solar.

Leo Smith

worse the conversion of hydrogen to mechanical power is likley to be no better than 40%

Tom in Denver

Yes but how many grams/kilometer will I get, and how much will it cost to refill my tank with nano Al

Wayne Austin

Since when does the US Army get to patent anything?

Javert Chip

Apparently, for quite a while…

Rhoda R

The government has the right to patent anything they develop. Sometimes they keep the patent and sometimes they let their research partner take the patent. It really depends on a number of factors.

Peer reviewed research papers have become such a scandal and the “research” so useless that the new gold standard has become patents.


You mean like the aussie who patented the wheel a few years back just to show how dumb the Patent Office is.


I wonder what the heat of reaction for the oxidation of the Al is. I bet you end up spending a lot of the reaction energy budget heating the reactants. This, of necessity, means your energy storage/recovery efficiency drops.
This system is actually one half of a battery. It generates hydrogen via a redox reaction which (presumably) goes into a fuel cell to generate electricity.
If any significant amount of energy is spent heating the reaction chamber, you will not get as much useful energy out as you spent refining metallic aluminum from alumina. This, as we know, is the Achilles’s Heel of energy storage systems.

D. J. Hawkins

You would lose that bet. The reaction was spontaneous at STP.


That’s the input energy to start the reaction. He was asking about the waste heat energy generated by the Aluminum-water reaction.
I used to put metallic calcium into a test tube with water, to make hydrogen balloons. There was no heating of the water needed beforehand, and the reaction generated waste heat. This is the same principle, and probably about the same waste heat as the reaction in this discussion, along with a comparable amount of energy needed to manufacture the pure metal from ore.


I win! That the reaction goes at STP, says nothing about how exothermic the reaction is. And, yes, the oxidation of Al is hugely exothermic.


TonyL – Like thermite for instance?


Aluminum thermite is actually manufacturing Bauxite. Take Aluminum metal plus iron oxide and get a mixture of aluminum oxied and iron oxide = bauxite

D. J. Hawkins

You said: He was asking about the waste heat energy generated by the Aluminum-water reaction.
He said: I bet you end up spending a lot of the reaction energy budget heating the reactants.
Care to re-evaluate what you said?
You said: I win! That the reaction goes at STP, says nothing about how exothermic the reaction is. And, yes, the oxidation of Al is hugely exothermic.
Your original post was about “heating the reactants”. That’s about the activation energy barrier. How exactly did you win?

Michael S. Kelly

Or you could just use the ferrosilicon/sodium hydroxide/water method that has been used since the early 1900s (by the military, no less), and save yourself a lot of development work.

Peter Morris


D. J. Hawkins

It took far less than an hour for posters here to ignore the link and raise questions answered in the longer article. Right, Forrest, @4.28pm?

D. J. Hawkins

Allow me to quote:

Forrest Gardener August 7, 2017 at 4:28 pm
Interesting. Presumably the Al is consumed in the reaction. So Al + H20 -> H2 + ???
Or is it just that the Al(OH)3 is no longer produced at the surface?
Or something?
Forrest Gardener August 7, 2017 at 4:39 pm
I see that the products of the reaction are H2 and Aluminium Oxide or Hydroxide, so that answers my first question. Maybe the structure of the material prevents the products of the reaction forming a coating. Or maybe the structure repels the products from the as yet unconsumed reactants.

So, did you or did you not jump into your first comment without reading the full article?

Gary Pearse

15kWh/kg to make ingots, after mining, processing and transporting alumina. Then producing nanoparticles, producing other metals in the same form. Whew. Then what do you do with the spent solids. The mix might be non recyclable at reasonable cost. Oh well, maybe they can make $1500 ashtrays for their Navy Jets and save money that way.


First HT, thanks CTM, and I am glad it got everybody thinking. I am just wondering how much Aluminum and water that little tank needs to bring with it to move. I agree that for military purposes the sky is the limit but it would be interesting to see what “Chemistry” can bring to batteries.

Kaiser Derden

pretty sure they can get water locally … so the obvious advantage of not having to schlep the liquid fuel all the way to hell and back … but still I could see some useful military uses (money no object type of stuff) …

Peta from Cumbria - now Newark
Peta from Cumbria - now Newark

… late…..


It’s not really a battery that stores energy in it and is more a fuel cell that holds a catalyst material that reaacts with water to release on demand hydrogen that can burn to power thing’s. Silicon Metal hardens Aluminum as do other light metals. That as namo particles would create a loose bond of the metal elements, that a water molecule of nearly the same size would cause a short across the hydrogen and oxygen on a microscopic scale with the bond breaking as the oxygen is attracted to the Aluminum and releases most of the hydrogen that doesn’t become a hydroxide with the metals…much like an acid solution would dissolving a metal into ionic bonds that releases hydrogen. The exception here is that the oxygen is trapped by the metal creating oxides and hydroxides that are insoluble in just water as a byproduct until it’s depleted all of it’s surface area and reaction ceases. The question is, are the byproducts able to be recycled back to the alloy nanoparticles?


I await some genius to tell us that coal fired power generation is the BEST thing known to humans – by a country mile
another genius could tell us that CO2 is beneficial to our planet and all who live upon it.

Kaiser Derden

well Mr. ALL CAPS genius … CO2 is beneficial to our planet …

Kaiser Derden

and you do realize your comment is way off topic … but to be expected from a ignorant troll I guess …


and Kaiser old mate you failed to realize that I was being sarcastic.


What a blatant SCAM!
I suppose you are meant to dig the Aluminium oxide out at the end, dry it, and send it off to a smelter and post treatment? Gimme a break!

Kaiser Derden

still have to burn the hydrogen right ? to produce power or electricity … could see it being useful in some specialized emergency or military scenarios … but for day to day use … no so much …


“still have to burn the hydrogen right ?”
Or a fuel cell to produce electricity.


Found this from May 16 2007….
New process generates hydrogen from aluminum alloy to run engines, fuel cells
“I was cleaning a crucible containing liquid alloys of gallium and aluminum,” Woodall said. “When I added water to this alloy – talk about a discovery – there was a violent poof. I went to my office and worked out the reaction in a couple of hours to figure out what had happened. When aluminum atoms in the liquid alloy come into contact with water, they react, splitting the water and producing hydrogen and aluminum oxide.
“Gallium is critical because it melts at low temperature and readily dissolves aluminum, and it renders the aluminum in the solid pellets reactive with water. This was a totally surprising discovery, since it is well known that pure solid aluminum does not readily react with water.”
Read more at:

David Chappell

Apart from the Hindenberg factor of lots of H in a confined space, how do you stop the reaction once water has been added to the nano-Al? It seems to me there’s an analagous situation to the problem of solar panels catching fire and the sun continuing to shine.

By turning off the water? That’s how I stopped my carbide lamp in my caving days. As long as the aluminum doesn’t ignite, you’re golden. 🙂


Is that a real question? I mean you know how to stop a diesel engine. You stop the flow of diesel.

Larry D

“Revive … the hydrogen economy” The “hydrogen economy” has never been a thing, just a dream.
This discovery, if it’s scaleable, would require a fill-up with water and the prepared aluminum alloy powder. Think about it. And you have to clean out the aluminum oxide too.

Cartridge swap. Sort of like a laser printer cartridge or acetylene cylinder swap.

No, the Al is not consumed in the process, it’s a conversion. But not to worry – it can easily be converted back using energy from the H that was produced in the process, and then you can do it all over again. An endless supply of H, the only drawback is that you never get to use it for anything else. But best to check first just how much Al the H would be able to covert back. Maybe it’s not quite endless.

Peta from Cumbria, now Newark

It bugged me that Willis’ figures were a bit low….
From very first principles of a Coulomb being x number of electrons, a kilo of Al being y number of atoms and each atom needs 3 electrons, you need to run just under 30,000 amps for an hour to get a kilo of Ally.
Does a cell work at 2 Volts – so that gives 60 kWh per kilo. Says I.
Then I goes check and surprise, I got it right, ish –
lowtechmagazine tells us…

Aluminum (from bauxite): 227-342MJ (63,000 to 95,000 watt-hours)

From here:
Is this between six and nine times the energy density of gasoline?
Promising stuff.
Very energy dense and it burns *very* hot (do they still weld railway tracks with it, certainly launch rockets) so it could make an exceptionally ‘Carnot efficient’ engine.
The exhaust ‘gas’ (carborundum basically) would be rather abrasive tho.
But and and, an aluminium electrolysing cell uses about the same amount of graphite as the aluminium it produces – they need to get the oxygen out the way as fast as fooking possible before it re-reacts with the aluminium they’ve just made.
So how much energy comes from (= is required to make) a kilo of graphite? Gotta add that in.


“burns *very* hot”
Not a useful characteristic. The main problem with all heat engines is that it isn’t practical to run them at optimum efficiency because the materials used won’t take it. Jet fuel for example has a adiabatic combustion temperature of 2,900 K, but no turbine material yet invented can take more than about 2,000 K.

Crispin in Waterloo but really in Ulaanbaatar

If it generates a lot of heat, then that heat could be used to drive a steam turbine generating electricity at an efficiency about of 35-40%. Add that to the 33% calculated above and it is about 70% efficient overall, providing H2 and electricity.


It would take a pretty big tank to make room for a boiler, a steam turbine, a generator, a condenser and a cooling tower.

Charles Nelson

The ‘New Scientist’ could easily be renamed ‘New Witchcraft’….it really has become a fringe publication.

Moderately Cross of East Anglia

Dr Strangelove does a valuable calculation to provide a non-chemist with a useful comparison, but please cut the guys doing the work some slack as at least they are doing their best to provide a possible alternative to the insanity of current battery technology – which is doing the health of the poor kids digging the materials out of the ground no end of harm.
I wonder what the comparison to gas would be if you did the calaculation Dr Strangelove made for the full cost of running an electric car, from material extraction to processing plus producing the “magic” electricity which comes out of the plug into the battery stuffed into the car?
Any attempt to escape from the dead end battery technology and its social consequences seems worth considering.


“Any attempt to escape from the dead end battery technology”
Dead end battery technology? I bet you’ve never owned a cordless drill, or one of those $15 toy helicopters.

Moderately Cross of East Anglia
Well said, there’s more to all this than simply the numbers. And as this was discovered by accident, what’s to say it doesn’t lead to something else, or that something else isn’t discovered by accident.
It does demonstrate, however, that despite the human races conceit, we don’t know nearly as much as we think we do.
The urbanised greens are deliberately blind to the conditions where rare materials are mined, as long as it suits their EV objectives. Nor are these materials called ‘rare’ without reason, they’ll run out far quicker than fossil fuels.
And all this for a trace gas essential to life on earth that was naturally, but accidentally sequestered, quite possibly by catastrophic events.
And at the risk of boring everyone as I have said it before, it seems an extraordinary coincidence that man happened along just when CO2 was at (one of 2(?) periods) it’s lowest atmospheric concentration in the planets history and started liberating CO2 by burning fossil fuels. If I were religions, I might hail it a miracle. But I’m not, so perhaps it might be considered the most fortuitous coincidence the planet has ever witnessed.


I used to work for a hard drive disc manufacturer, substrate for disc was aluminum. Here’s the issue I see from my experience working there.
We would grind the aluminum disc down to the proper thickness, this left us with a lot of aluminum suspended in a slurry slowly converting to hydrogen and aluminum oxide. First of all this slurry had to be stored in tanks that were vented and monitored for H2 to ensure it did not reach an explosive levels. This is an issue that would probably be fairly easy to deal with in a hydrogen burning car.
Second issue is a bit more problematic, our aluminum slurry could not be allowed to dry out. Converting aluminum into hydrogen and aluminum oxide is an exothermic reaction, let it dry out slowly and that stuff will auto ignite. One of our engineers did an experiment, he put some in a dish with a temperature probe sitting in it. As the slurry dried out it slowly heated up to 130F at which point temperature increased rapidly and the sample started to burn. We disposed of the slurry by loading into a hazmat truck, hazmat driver was not allowed to stop until they reached their destination which was a waste burning facility ~4hrs away. That guy was driving a bomb with the detonator consisting of nothing but time. Before anyone says it’s not that bad, yes it is and we had a couple of very exciting events to prove it.
This can be a real serious issue in a car, can the car be designed to safely handle the “waste” fuel to the point a complete idiot can safely own and operate it? Can the gas stations “safely” handle the waste until it was picked? Are there enough hazmat certified truckers around to safely transport the waste? How far can a trucker go before the load is no longer safe to transport? Figure that last one out and you’ll know the maximum distance allowed between nonexistent reprocessing facilities that will have to be built.
There’s more cost associated with this fuel then just the conversion of Al-H2 and back to Al again. There’s going to be a large cost factor into safely designing a car to use this fuel and the cost of safely handling/transporting the waste for reprocessing.

Clyde Spencer

Darrin and all,
Something that I haven’t seen mentioned so far is that hydrogen is notorious for embrittling metals (especially steel) and is prone to leaking because of its low molecular weight. Thus, even if the economics problems can be addressed, there are still a lot of engineering and materials science issues that need to be solved before the hydrogen technology can be applied commercially.

A detailed analysis of the Al + H2O reaction for hydrogen production was published in 2008, the summary of their conclusions were as follows:
“Aluminum Required: 9 kg Al per kg H2 assuming 100% yield
Gravimetric Hydrogen Capacity: 3.7 wt.% (materials only)
Volumetric Hydrogen Capacity: 36-46 kg H2/L (materials only)
Reaction Kinetics: 2 x 10-4 g H2/sec/g of Al – from published data to date
Cost: $7 per kg H2 (based on the cost of electricity for aluminum production considering only the reduction of alumina to aluminum step)
The current DOE hydrogen storage system capacity targets are a hydrogen gravimetric capacity of 6 wt.% and a hydrogen volumetric capacity of 45 g H2/L (27). It is clear from the analysis presented in this White Paper that no aluminum-water reaction system can meet these targets. Additional negative factors are the high cost of hydrogen from this process, and the amount of aluminum required for large-scale vehicular applications.”
“The November 2007 commodity price for aluminum is $2.36 per kg. At this price, hydrogen from an aluminum-water hydrogen generation approach would cost approximately $21 per kg H2. Even assuming high volume production, the DOE target range for hydrogen cost of $2-3 per kg H2 would not be met.”
Emphasis mine.

Hydrogen is a pain in the butt. Electricity is piped into everyplace, including your home and condo.
Recharging stations for electrics will be nowhere near as plentful as those required for a fleet of gasoline vehicles.

Ian L. McQueen

No one has (yet) mentioned the very high cost of producing those “micron-scale grains of aluminum”. Has anyone here ever purchased aluminum powder? The cost of reducing solid metal to powder is enormous.
That’s a major problem with all schemes proposed by economists, lawyers, sociologists, etc,, that they have no practical experience, the kind that a good engineer should have.

Crispin in Waterloo but really in Ulaanbaatar

It would never be produced as a solid first. When the molten metal is ready, it would be sprayed vertically down into a nitrogen atmosphere cooling chamber producing the particles directly. That is similar to how lead shot is made (they use a screen for that).

Clyde Spencer

Producing aluminum nitride?


You mean “the molten alloy”. Which is hard to make without any of the metals being a solid first.


Don’t tell Musk. He’ll turn it into another massive tax credit scheme.

michael hart

Aluminium has merits, but the solid waste products, aluminium oxide or hydroxide, are messy and awkward to handle. This may seem trivial to some, but dealing with non-fluid compounds is hugely problematic for something that needs to be scaled up to a vast industrial scale. It won’t happen.


Nothing new under sun. Mercury catalysis brings hydrogen production from aluminium and water. Trace of mercury salt induces AL2O3 formation due to air humidity from aluminium surface- you may try it. Hydrogen formation from aluminium was used for balloon filling at start of 20. century. Nanno tricks are useless,

There is a very well known and massively researched principle in metallurgy which is called hydrogen in metals. It doesn’t sound very sexy, but about 28 years ago I had reason and opportunity to read up on it as much as I wanted to.
What is it? Hydrogen in metals is a phenomenon in which hydrogen atoms go into metal crystal lattices and keep on going in, and keep on going in, and keep on going in. The metal sort of absorbs the hydrogen but not chemically. The hydrogen eventually causes problems in the metals and causes metal parts to fail. These failures is the reason it has been studied so extensively, as scientists try to understand it and to learn how to prevent it.
What I am thinking here is that the nano state of the aluminum is possibly (probably?) allowing the water in some way to access the hydrogen that is INSIDE the aluminum already.
If this is the process that is bubbling the hydrogen out, then I can tell you that the wrong people to get involved is the physicists. This would be a metallurgical effect, a crystalline effect. Physicists will not have any experience with this at all.
ALSO: They found this with aluminum, true. It may not be unique to aluminum, though. I’d suggest they nano-treat other metals the same way and see what happens. My bet is copper and silver might work, too. But for vehicles, if this is real, then aluminum would be best, due to weight factors.
This kind of reminds me of the accidental discovery of white-light LEDs. Also the accidental discovery of graphene.

. , Crispin, you’ve got top raise the temp to over 2000C to extract aluminium that way -costing more energy again which is why electrolysis is used.


Try looking up thing’s… Aluminum melts at 1,200 F. High temp aluminum alloys usually under 1,500 F. Aluminum Hydroxide melts at around 540 F. As the vat of aluminum hydroxide melts by the use of a cathode and anode of carbon electric arc the Aluminum Hydroxide acts as a flux floating on top of the molten aluminum, as the molten aluminum pours out of the vat…more aluminum hydroxide is added in a contentious manner. To create an alloy other metals as oxides or hydroxides ot other compounds are added during the process and only slightly higher temperatures are required. This is from a 1933 updated in 1954 edition basic college chemistry book on my desk. What this takes in kwh I have no idea. We don’t know the alloy of the nanoparticles or the melting point to get it done to kwh. And you can melt aluminum and added metals in a microwave furnace in crucibles at lower temperatures than the temperature it takes to melt the alloy after its created. Making nanoparticles alloys is done in a different process from the conventional process.