Sciencemadness Discussion Board - Magnesium from aqueous electrolysis

Magnesium from aqueous electrolysis

Levi - 23-4-2007 at 08:47

I've got an idea to produce magnesium in small quantities (and likely low effeciency) by the electrolysis of MgCl2 in solution with a buffer of some sort to keep the pH up high enough and prevent my Mg from entering back into solution faster than I can remove it. My idea is to use a sodium acetate buffer and keep the solution as dilute as possible. Ideally, the electrolysis would then form magnesium metal, sodium chloride and dilute acetic acid.

My calculations show that it takes 3.325 volts to decompose MgCl2 but I'm not very confident with it, can anyone verify that this is correct or see an immediate fatal flaw in my procedure? I realize that the magnesium obtained will not be structurally sound as the voltage will be high enough to decompose the water as well; I expect to get a mush-like slop at the cathode. But since most chemical applications are likely to call for powdered magnesium this may actually be advantageous since it eliminates the need for ball milling.

[Edited on 23-4-2007 by Levi]

12AX7 - 23-4-2007 at 08:50

So how do you prevent the "mush" from reacting with water?

3.whatever volts is too much in water at any pH.

Levi - 23-4-2007 at 08:52

Quote:

Originally posted by 12AX7
So how do you prevent the "mush" from reacting with water?

3.whatever volts is too much in water at any pH.

Magnesium will not displace hydrogen from cold water so my idea was to keep the solution as cold as possible.

Explain what you mean by 3 volts being "too much."

UnintentionalChaos - 23-4-2007 at 10:43

Have you ever cleaned a small piece of magnesium and placed it in water? It bubbles (slowly) regardless of the temperature. It is still an alkaline earth after all. What you are doing by electrolysis is adding extremely tiny particles of a reactive metal with none of the normal protective oxide layer into an aqueous environment. You won't produce anything but Mg(OH)2 unless you cut out the water. Hence,

3.whatever volts is too much in water at any pH.

Note the "In water" in what 12AX7 said.

Levi - 23-4-2007 at 10:57

I just dropped a clean (no oxide) firestarter block in ~50°F water and it did not react at any appreciable rate (very tiny bubbles formed slowly). As I stated, I do not expect this to be an efficient process, just that I will be able to take Mg out faster than it is going back in. Granted the small particle size will work against me but I can also get the water temp a lot colder than 50°. I am preparing a solution of MgCl2 presently and will be testing the theory without a buffer at first because I haven't made any sodium acetate yet. I will post results (and pictures) when I have them.

UnintentionalChaos - 23-4-2007 at 11:09

A firestarter block has only a few cm^2 of exposed surface area. The spongy magnesium you are likely to make will have an enormous surface area (most likely several orders of magnitude greater than the block) and will immediately react with the water giving you Mg(OH)2 with (maybe) trace amounts of metallic Mg. I've electrolyzed a Mg salt solution before and you get a nice heap of white hydroxide crumbs and not a trace of metal.

Nicodem - 23-4-2007 at 12:15

Quote:

Originally posted by Levi
My calculations show that it takes 3.325 volts to decompose MgCl2 but I'm not very confident with it, can anyone verify that this is correct or see an immediate fatal flaw in my procedure?

It takes less than 3.325 V to decompose H2O, so there is the flaw in your procedure. It is exactly like when you electrolyze CuSO4 - you get Cu and O2 and barely any H2. That’s because CuSO4 decomposes at a lower potential than H2O. Therefore practically no Mg would form by the electrolysis of MgCl2(aq). You need a solvent that will not decompose electrochemically at the potential required to decompose MgCl2 while also not reacting with the formed Mg.

Zinc - 23-4-2007 at 12:31

I have heared that lithium can be made by electrolisis of a solution of lithium chloride in amyl alchohol or even acetone but in a lower yield. Perhaps Mg can aso be made that way. I know not really aqueous...

Levi - 23-4-2007 at 12:33

I really don't see why decomposing the solvent is a problem for anything except efficiency. I recently "plated" (the result was mush) some copper from a copper hydroxide/carbonate solution and the water was bubbling away without having any other effect than making the copper plate poorly. Anyhow, I just did the experiment with the magnesium and will post the first photo of a 3 photo set. My camera died after I uploaded one

I'll post the others when it's back online.

<a href="http://i14.tinypic.com/33u79ea.jpg">Image 1</a>

Mg(OH)2 precipitated at the anode and whatever substance was plating at the cathode did not precipitate but stuck to the graphite and is visible in the second photo.

The solution was slightly basic with some ammonia impurity and possibly some MgSO4 and trace amounts of sodium chloride. The original solution was made by precipitating Mg(OH)2 from a solution of epsom salt and ammonia. The hydroxide was filtered and washed once but still smelled slightly of ammonia. To the hydroxide was added water and HCl until the solution was clear and Na2CO3 evolved no gas. The electrolysis was done with a 9v battery and a graphite cathode.

[Edited on 23-4-2007 by Levi]

Levi - 23-4-2007 at 13:24

<a href="http://i16.tinypic.com/4e02wl3.jpg">Image 2</a>
<a href="http://i11.tinypic.com/47twjly.jpg">Image 3</a>

Kind of blurry but the gray/white solid is visible on the graphite and a small pile nearby on the paper.

[Edited on 23-4-2007 by Levi]

The_Davster - 23-4-2007 at 14:05

Underground iron structures can be protected from corrosion by cathodic protection, wheras otherwise they would rust from reaction with water and air.
I cannot see why one could not reverse the process with enough power (volts and amps) and force some magnesium out of solution, at terrible efficiency of course.

When doing some stuff with tin/antimony alloys, I found some deviations from what should be expected to react and pass into solution. Whether this is a physical effect from exposure of various metal to the electrolyte or a chemical effect from so much power, I never could figure out.

Other electrolysis experiments, for example, dissolving copper in a nitrate solution. Copper anode, graphite cathode, based on redox potentials copper should just dissolve making Cu2+, and water decompose into hydrogen and OH-. At low voltage, low power, this did happen. However when the amps were cranked to around 10 by use of large surface area and small electrode spacing, and the powersupply was turned to 12 V, strange reactions occured. Oxygen was being given off one electrode, hydrogen the other, despite the reduction potentials for water to oxygen and Cu to Cu2+ being almost a volt different. Copper still was dissolving. Nitrate was even reduced to ammonia after a while.

. None of this was predicted by my redox tables based soley on reduction potentials.

I think the lesson here, is that unless you keep the power low, and the cell at those 'standard conditions', an electrolysis cell is more complicated than looking at a redox table.

So crank the volts and amps up, and hope for unusual results!

garage chemist - 23-4-2007 at 14:57

Magnesium is absolutely impossible to prepare in aqueous solution, fact.
H3O+ is reduced far easier than magnesium ions, thats the reason why it wont work. At high pH, magnesium hydroxide precipitates, removing the Mg2+ ions from solution.

Only at cathodes with very high hydrogen overvoltage this may work. Mercury is the only cathode that could possibly do this.

texaspete - 29-4-2007 at 12:47

I was going to try to make some Mg from MgCl2 in an aqueous with NH3 to keep to pH up, but thought it was a bad idea since Cl2 gas will be produced at the anode. From what I have heard, Cl2 and NH3 don't mix.

Still, as Davster said, overvoltage will produce results against what the redox tables say.

Maya - 2-5-2007 at 09:35

I've plated Sodium out of a NaOH solution before, of course I was using a Hg cathode

dedalus - 2-5-2007 at 10:22

Quote:

Originally posted by garage chemist
Magnesium is absolutely impossible to prepare in aqueous solution, fact.
H3O+ is reduced far easier than magnesium ions, thats the reason why it wont work. At high pH, magnesium hydroxide precipitates, removing the Mg2+ ions from solution.

Only at cathodes with very high hydrogen overvoltage this may work. Mercury is the only cathode that could possibly do this.

Yes, what you said.

The basest metal that will electrodeposit from an aqueous electrolyte is manganese.

indigofuzzy - 3-5-2007 at 11:52

Just a thought, wouldn't keeping the pH up (and therefore alkaline) make Mg(OH)2 even more likely to precipitate?

For the record, I've tried to get magnesium metal out of MgSO4 solution. I've even pumped a whopping 12 volts through it. As expected, Magnesium hydroxide was produced.