Author Options:

A new (and potentially real) way to produce hydrogen from water Answered

A report in Science News today describes the results of an extremely detailed molecular simulation (full non-relativistic quantum mechanics) which shows how small clusters of (less than two dozen!) aluminum atoms can catalyze the dissociation of water into hydrogen with very high efficiency. 

The paper has apparently been accepted by Physical Review Letters, but there's nothing up on arXiv :-(

Before the alternative-energy fringe starts jumping up and down, the big roadblock to this mechanism is how to fabricate and distribute aluminum atomic clusters (called "superatoms" in the report; sigh...).  Making "nanoscale" (another annoying word...) clusters of atoms usually involves high temperatures and vacuum systems.   Together, those will consume substantially more energy than you recover from making and burning the hydrogen.

However, if that roadblock can be overcome, then this does become a viable technology for fuel cells:  fill a tank with water, pour in a pouch (or bucket) of aluminum clusters (magic pixie dust :-), and hydrogen starts bubbling out.  It will be interesting to see how this shakes out

Oh, and the simulation makes cool pictures, too!


This is thermodynamically uphill (286 KJmol-1). Where is the energy input?


I don't know.  From the description in Science News, the Al cluster ("superatom") is an electron donor.  It strips protons from water molecules, leaving OH- behind.  The protons plus donated electrons recombined to form H2.  What I don't quite get (not being a chemist myself), is why you don't end up with a rather nasty basic solution, plus highly ionized Al clusters, in the end.

If it's a catalyst it should just give you hydrogen & oxygen. But you'd need to be "shovelling" energy into it...
Unless you end up with hydrated alumina-sludge?
Some details missing in that article.


"Some details missing in that article."  Thank you, Captain Obvious :-)  It was a "news report," not the actual paper.  As I said in my posting, I could not find a preprint of the actual research.  PRL requires payment to view articles (obviously, as it's a publication). 

You're at least as good a Searcher as I am; if you can find a readable copy of the paper, I would really appreciate it!  I have some of the same questions you do.

I had a bit of a look, but I had a busy day. I'll have another go tomorrow.



8 years ago

I think on this one, at best, you'd have to make the aluminum superatoms in situ. Thermite side reactions and the abundance of atmospheric oxygen would limit both the prevalence of this technology and it's application. I'll ignore the amount of energy that would be required to make the superatoms in situ at the moment and just say that the logistical problems this technology has inherant to it are troubling.

If ya want to see what the best catalysts for making hydrogen are, you need look no further then current industrial chemical techniques.

Interestingly, most techniques for making hydrogen involve hazardous chemicals and expensive metals, but the largest and best catalysts are often ignored.
The best catalyst for making hydrogen in the industrial setting is carbon, specifically, activated carbon (which is made by taking regular charcoal and removing the impurities in it).

Once you have an activated charcoal filter, you only need heat, pressure, and steam, and since pressure and heat usually go hand in hand with steam it's not too hard to get ahold of.

Could you provide some links to peer-reviewed articles on how aluminum clusters ("superatoms") are formed?  I wasn't able to find any, and I'm quit interested.

The trouble with the carbon catalyst is just the heat and pressure issue.  The goal with a small fuel cell is that it be (a) portable, and (b) reasonably safe for use in a car.  I'm not sure those constraints can be met, but I would be very interested in being proven wrong.

Thank you for the conference link!  That seems to me a different use of the term "superatom" than in the news article I posted.  Your paper sounds like it's describing something more like a "quantum dot" -- a controlled area of dopant in a semiconductor, which behaves like a large atom with well-defined orbitals.  The things in this top are more like really small atomic clusters (with tens of atoms instead of hundreds), which are free-standing.

That Wiki article is outstanding; thank you!  It not only describes what was used in the simulation I posted, it even talks about how their created in practice.  The fact that they usually have a halogen ion attached to make them chemically active is quite interesting, and probably has bearing on Lemonie's cogent comment about energetics.

Similar in that both use aluminium, but this version seems not to consume the aluminium in a chemical reaction - the reaction you link to needs to be "recharged", because the "liquid metal" is eventually saturated with aluminium oxide.

Some day I hope my question about catalysts is answered...

Is there one for everything? 

Your question is ill-formed, and therefore unanswerable as stated :-)

What do you mean by "everything"?  Do you mean any chemical reaction?  Do you mean some reaction to get to any final state product?

And do you mean are the catalysts involved already known?  Do they already exist?  Or could a catalyst be prepared/designed where one is currently unknown?

Catalysts, in general, either accelerate, increase the efficiency, or overcome an energy barrier of some chemical reaction.  The inputs and outputs remain the same, and the catalyst itself ends up as an additional reaction product (i.e., the net amount of catalyst is unchanged before vs. after).

Many reactions proceed quite happily on their own, without any need for a catalyst.  So, in that sense, not "everything" has a catalyst.

In biology, almost every reaction involving proteins both has and requires catalysts (i.e., enzymes).  The reason is that you are dealing with massively complicated molecules, such that the rate for a given complex reaction happening "automatically" is almost nil.  The enzymes in this case facilitate reactions by exposing the necessary active site in a folded 3D molecule, or bringing the reactants into proximity by preferentially binding to both, or being a good donor or acceptor of electrons (which supports moving ligands around).

I think I liked that one better when they replied with a blank stare and a sigh...

I love the analogies for enzyme functions, generally they're susceptible to innuendos, however they always get clumsy once you use the real terms. Even more proof my answer's ill formed... 

I'm going to suppose the answer to the general question is no, imaginably there are reactions that can't be made simpler or made to happen at lower energies... 

.  Employing the proper crystals (a full moon helps), one can get the superatoms of Aluminum to vibrate and release prodigious amounts of HHO with no external power source. <snicker>

I'm going to guess that this will first see action as a vehicle fuel - a network of water-stations supplying distilled water would be much easier to set up than a network of cryotanks.

As for the supernanoatomicdoobries, would striking an arc between aluminium electrodes produce particles on the required scale as the metal evaporates from the electrodes?

(Not sure how you would collect them, though.  Electroctatics?)

Good question, Kiteman.  I don't know, as I'm not a "soft condensed matter" physicist.  With carbon, doing that gets you all the cool buckyballs and smaller clusters (down to C20 or so); it seems likely that you could do the same with aluminum, but I'm just guessing.

When I looked on arXiv for "aluminum superatom" papers, I ran across several talking about aluminum and gold clusters, but those all ran in the hundreds of atoms range.