Alloys to store hydrogen?

[Enthalpy]

Hello everybody!

You know hydrogen is thought as a way of storing energy, before transformation in a fuel cell for instance. Storing liquid hydrogen isn't very easy because it needs really cold temperatures, and pressurizing gaseous hydrogen needs heavy tanks with little capacity, so research is being done to adsorb hydrogen in a solid in the hope to store much hydrogen in a small volume under reasonable conditions.

My own two-cents suggestion is to use abnormally light metallic alloys for this purpose. That is, alloys whose volume is bigger than the sum of the volumes of their constituents. My hope is, of course, that more free room is available to hydrogen in the alloy then.

 

Alloys known for their abnormally low density (and modulus) include:
- Bell bronze (20% Sn, 80% Cu)
- Invar (36% Ni, 64% Fe)

 

Also Mn-Cu (vibration damping alloy) has abnormally low modulus, but I haven't found its density.

 

----------

 

Both Ni-Fe and bell bronze alloy elements with slightly different molar volumes. Is this a key to abnormally low alloy density?

 

On the other hand, Ni-Ti has in some metallurgical states a very low Young's modulus but still an abnormally high density, so Young's modulus isn't a reliable indicator.

 

----------

 

Shape memory alloys (like Ni-Ti) change their volume over temperature. Provided they store an interesting amount of hydrogen, the density change could be a means to release the gas at will.

 

----------

 

Well, I guess competitors like zeolite are already better than any metallic alloy, but anyway, that was my message in a bottle.

Marc Schaefer, aka Enthalpy

[Enthalpy]

Magnetostrictive materials change their volume as a consequence of the magnetic field. Only 60ppm in one dimension for cobat, but 2,000ppm for Terfenol-D (which isn't exactly cheap). Provided they store a significant amount of hydrogen (their Young's modulus is aberrant), they might release it at will by the action of a magnetic field.

 

If an electromagnet requires too much power, just pass the material (wire, sheet...) near a permanent magnet.

 

Well, trying isn't difficult, so it can be worth an experiment.

 

Marc Schaefer, aka Enthalpy

Edited May 15, 2014 by Enthalpy

[Enthalpy]

One Mn-Cu alloy is well documented: the M2052, with mass% Mn-22.1Cu-5.24Ni-1.93Fe.

Its density is 7250kg/m³, 5% less than 7628kg/m³ if adding the constituents' volumes. Its Young modulus is also 47GPa instead of 198GPa of Mn and 130GPa for Cu. But it's a vibration damper, as opposed to bell bronze and invar.

Besides providing room in the lattice for hydrogen (just a hypothesis), it has two transitions, one below and one above room temperature, which might help release the hydrogen. Ni and Fe serve mostly to adjust the transition temperatures.

Marc Schaefer, aka Enthalpy

[Moreno]

Aluminum is much more convenient way to store energy than hydrogen.

This is if you're not satisfied with biofuel which is as well conditionally carbon neutral as hydrogen.

So why hydrogen?

Edited November 9, 2018 by Moreno

[John Cuthber]

On 10/5/2013 at 6:55 PM, Enthalpy said:

Alloys known for their abnormally low density (and modulus) include:
- Bell bronze (20% Sn, 80% Cu)
- Invar (36% Ni, 64% Fe)

No, they are not.
Invar is slightly denser than iron.

Bronze is a bit denser than tin, but not as dense as copper- much as you would expect.

You can make an argument for lead having a rather low density. It's about as dense as silver, but the atoms in it are about as heavy as those in gold, so they must be spaced out more widely. If they are spaced out more, they aren't densely packed.



Back at the topic...
If you have some alloy which absorbs hydrogen then that absorption is going to be accompanied by a release of energy.

You have to remove that energy as heat when you "refill the tank".
Even if the absorption energy is only 1% of the combustion energy that's a lot of heat.
Say you want a tank which holds as much energy as the petrol tank in a typical car.
That's about 50 litres.
Petrol stores about 35MJ/L
So the full tank holds 1.8GJ
If you "waste" 1% of that as lost heat then you have 18MJ to remove.
Filling the tank in 5 min gives you 300 seconds over which to lose the energy.
 

That's about 60KW you have to dissipate in order to store the hydrogen when you fill up.

Have fun.
 

[Enthalpy]

Yes, both bell bronze and invar are lighter than the sum of the constituents'  volumes let compute. I did the computation.
Don't mistake bell bronze for other compositions, and remember that nickel is denser than iron.

I don't know if absorbing hydrogen must release energy. A dissolution can release or absorb energy. 1% of the combustion heat of hydrogen would be a lot for a low-energy process, for instance the evaporation of hydrogen at 20K absorbs only 916J/mol while the combustion releases 286kJ/mol.

Anyway, 60kW is as much heat power as a petrol engine releases, and cooling it is long solved. It doesn't need me. And did you check how lukewarm the storing metal gets when absorbing hydrogen? Even if 1mol metals stores 1mol hydrogen, released 1kJ/mol heats only by 40K, so cooling may not be necessary during filling.

[John Cuthber]

2 hours ago, Enthalpy said:

for instance the evaporation of hydrogen at 20K absorbs only 916J/mol while the

Yes. An you can look at that the other way round.

Because the heat needed to vapourise hydrogen is so small, the boiling point of hydrogen is very low.

However we want something where hydrogen sticks to it reasonably well at room temperature.

And that means that the heat of absorption must be in the same ballpark as the thermal energy .

The thermal energy is about RT per degree of freedom

r=3.81 J/K/mol

So 

We know that , at 20C i.e . 293K the thermal energy is about 2.5 KJ/mol per degree of freedom

 

I'm fairly sure that only the translational freedoms count here (the sorbed gas might still vibrate + rotate)

So that's about 7.5 KJ/ mol.
And that's the binding  energy where a molecule of hydrogen has a roughly 50:50 chance of being in the gas  phase.
We want most of it stored in the metal.
So we need a factor of "a few more times" such that the Boltzmann distribution puts almost all the gas into storage.

We are not up to a sorption energy of 15 to 30 KJ/ mole.

That's something like 10% of the combustion energy.

So your engine which- as you say delivers something like 60KW- has a fuel store that needs you to dissipate something like half a megawatt of heat.

(and it has to do it without the "sponge" getting significantly hot- otherwise the gas simply won't stick.)
So rather than the problem of cooling a 60KW engine running at , say, 250C in an air flow of a few tens of metres per second, you have to remove 500 KW of heat from something that you want to keep below something like 50C and  which is stationary.

 

On 11/11/2018 at 10:54 AM, John Cuthber said:

Have fun.

 

[Enthalpy]

I've seen figures like 15kJ/mol for adsorption heats. Measures from serious storage candidates would be better. That's 1/20th the combustion heat.

So if the engine uses the hydrogen in 3h to produce 60kW with 50% efficient cells, and the user tanks in 6min, 3*60kW=180kW must be evacuated.

So what's the trouble? We have hydrogen in the tank, an excellent fluid for that task. Once the heat is at the tank's walls, blowing some air removes it. It's as much heat as from a 240HP thermal engine, except that this heat is already at the atmosphere.

I won't put my time in this. Banal engineering.

[John Cuthber]

An engine runs hot.