Sciencemadness Discussion Board - Electrolysis of alkali salts at STP

Fairly easy to make, all things considered.

It's harder than you think (urea + propylene glycol gives ~9% yield, which must be separated out of a tarry mess.) It might work better on larger scales, though.

elementcollector1 - 4-3-2014 at 19:55

Adding a catalyst of ZnO is supposed to give 98% yield...

blogfast25 - 5-3-2014 at 04:32

Adding a catalyst of ZnO is supposed to give 98% yield...

You have a reference for this synthesis and catalysis?

BromicAcid - 5-3-2014 at 04:35

But, my question remains, Can this be done with sodium? And if not, Why?

As elementcollector1 said, the solubility of the sodium salts in non-aqueous medium is significantly worse than the lithium salts severely limiting the number of non-aqueous options you have to work with. The classic example is the electrolysis in nitrobenzene which is often quoted but the whole procedure is misunderstood. It is not simply an electrolysis in nitrobenzene but you must fuse sodium chloride and anhydrous ammonium chloride to get a complex salt to perform the electrolysis, from there the material deposited at the anode is said to be explosive. Again, simple enough in theory but dastardly difficult in practice.

With regards to the preparation of propylene carbonate from urea and propylene glycol, that is another mess waiting to happen. I have not seen (but would be glad to be shown) anyone get anywhere near the yields claimed in the patent in actual practice. My own experience with producing ethylene carbonate by the urea/diol/ZnO process is that you end up with a mess from that whose separation involves filtration, crystallization, followed by vacuum distillation (BP Propylene Carbonate 242C @ ATM).

Non-aqueous routes are exciting, plus they can be tested quickly and easily with a power supply so long as you have your solvent and your salt. But selecting the proper counter ion and solvent... that's the crux of the issue. I would love to see some of your attempts however. I myself had a great many flings electrolyzing lithium chloride in DMSO before after quite a bit of searching I discovered I was making methyl lithium. And what beautiful methyl lithium it was

Metacelsus - 5-3-2014 at 05:59

Adding a catalyst of ZnO is supposed to give 98% yield...

The 9% yield I got was with the ZnO.

elementcollector1 - 5-3-2014 at 06:13

Unfortunately, the old reference I had died for whatever reason. However, I did find this: http://www.doiserbia.nb.rs/img/doi/1451-9372/2011/1451-93721...
It uses metal carbonates instead of metal oxides.
How did you prepare your ZnO? Was it calcined?

TheChemiKid - 5-3-2014 at 07:10

Has anybody tried just electrolyzing in a hydrocarbon such as pentane, hexane, or heptane?

WGTR - 5-3-2014 at 07:42

I've tried two cations that I can remember off the top of my head; lithium, and calcium.

My experience with lithium was somewhat disappointing. Don't underestimate how dry the acetone (and lithium chloride) must
be. I was never able to get more than just a thin film of lithium at one time, just enough that it would "fizz" when I
dunked the electrode into water. Never enough to actually collect.

I tried electrolyzing calcium chloride in ethanol. That didn't work too well either, but again, I got enough to react with water
when the electrode was submerged in it. One thing I found was that the process did not work too well if the alcohol was
absolutely dry. I had to add a drop or two of water to it to get things going. The process was a clever balance between water
content, current density, and salt concentration.

Zyklon-A - 5-3-2014 at 07:47

TheChemiKid, there are many criteria for it to work as a suitable solvent (e.g. it can't react, must allow salt to ionize ect.)
blogfast25, excellent link, very informative thank you.

I will buy my Propylene carbonate, I've got some $50.00 designated for this project.

Quote:

At a minimum, try and dig up the references of that section of the paper, before trying anything...

I read all the references that he listed pertaining to the subject.

Quote:

My experience with lithium was somewhat disappointing. Don't underestimate how dry the acetone (and lithium chloride) must be.I was never able to get more than just a thin film of lithium at one time, just enough that it would "fizz" when I
dunked the electrode into water. Never enough to actually collect.

Hmmm, well that's about what I would expect. I will keep the container completely sealed off, (to keep it from absorbing water.) How much current did you use? Voltage?
At least it worked, that's incourageing.
Since I already have lithium, I may try to get sodium from NaClO₃, as shown in blogfast25's link. (The reason that I changed the subject of this discussion.) Potassium is another goal....

[Edited on 5-3-2014 by Zyklonb]

Metacelsus - 5-3-2014 at 08:11

How did you prepare your ZnO? Was it calcined?

I prepared it by heating zinc hydroxide (from the reaction of zinc chloride and sodium hydroxide). I did not calcine it.

A full report of my synthesis is in this thread.

Zyklon-A - 5-3-2014 at 16:05

Ok, I'm about to purchase this propylene carbonate. Does it look like it will be a good choice?

Oscilllator - 5-3-2014 at 16:18

Ok, I'm about to purchase this propylene carbonate. Does it look like it will be a good choice?

I dont see why not. Don't be surprised though if the electrolysis does not work at first, since there will no doubt be water and other contaminants that react with the propylene glycol.

Zyklon-A - 5-3-2014 at 18:49

No doubt, I will have to dehydrate my propylene carbonate and whatever salt that will be electrolyzed (probably not the right term.) All experiments will be controlled in a dedicator, and will be handled only occasionally, and briefly.
I couldn't find any information on dehydrating propylene carbonate.

Oscilllator - 6-3-2014 at 00:02

I imagine that if it only contains a small amount of water you could start electrolysing it, and LiOH would form until the propylene glycol was fully dehydrated, then Li woul be produced. Its quite possible that I am wrong, but its worth a shot I think. As for how to dry it, distill over sulfuric acid? H2SO4 at 180 degrees will dehydrate just about anything.

Zyklon-A - 6-3-2014 at 05:37

That's what I would do, but I don't have a distillation setup....
It's quite close to anhydride, so I might be able to use it as is.

blogfast25 - 6-3-2014 at 06:17

Quote: Originally posted by Oscilllator

I imagine that if it only contains a small amount of water you could start electrolysing it, and LiOH would form until the propylene glycol was fully dehydrated, then Li woul be produced.

Actually, if there is water it will be electrolysed first because H3O+ is much, much easier to reduce than Li+.

So start running a low voltage that gives you some current, thus eliminating the water, then increase potential to start reducing Li.

[Edited on 6-3-2014 by blogfast25]

Zyklon-A - 6-3-2014 at 06:32

That's a good idea. Thanks.

blogfast25 - 6-3-2014 at 06:39

That's a good idea. Thanks.

And DO report back: I think we're all dying to know how this pans out! :cool:

It would probably help to dry the solvent overnight over anh. MgSO4 and oven dry the LiCl at 120 C for a couple of hours, just prior to use. Oven dry your cell too...

[Edited on 6-3-2014 by blogfast25]

Zyklon-A - 6-3-2014 at 07:10

Definitely, I'll report my progress here, and in the Successful experiments topic, if it works

. If not, You'll likely find it here.
MgSO₄ will be used to dry the propylene carbonate. I plan on first using NaClO₃ to make sodium, to avoid have to buy another chemical. (I don't have any LiCl.) NaClO₃ may prove to be hard to dry.... A day in the oven should do it.
[EDIT] Damn, I just ruined my 500th post.... International Hazard!

[Edited on 6-3-2014 by Zyklonb]

MrHomeScientist - 6-3-2014 at 08:22

That does sound very exciting! I'm also eager to hear how it goes. I was trying to pursue a route to neodymium metal by electrolysis in an ionic liquid solvent (choling chloride & urea), but that's stalled because I can't get the damn thing to stay liquid. Over time crystals start forming. Propylene carbonate may be an alternative to try, if it works for you.

blogfast25 - 6-3-2014 at 13:34

So the deep eutectic solvent didn't really work out?

Zyklon-A - 14-3-2014 at 12:21

OK, my PC was ordered last week, and should be here soon! I am going to build my power supply today. I haven't seen any references for the optimal voltage and amperes. 10 volts seems to be used by some people, but the amperes shouldn't matter unless the metal product reacts with the solvent and must made faster than it reacts.
How many volts/amps should I use per 100mL PC, with dissolved NaClO₃?
I hope I can try this sometime week or next, as soon as it arrives.

Marvin - 14-3-2014 at 12:48

Your plan is to dissolve an oxidiser noted for explosive tendencies in an organic solvent and pass a current through it in the hope of generating sodium metal. Is table salt really that expensive?

Zyklon-A - 14-3-2014 at 12:56

Perhaps you should read through this topic. Why would I need table salt? Are you suggesting I make a downs cell? I am trying to avoid high temps, here.

blogfast25 - 14-3-2014 at 13:45

How many volts/amps should I use per 100mL PC, with dissolved NaClO₃?

Is NaClO3 really the only available Na salt soluble in propylene carbonate? In the absence of any other anions you have to wonder what will happen at that anode...

Marvin has a point: NaCl appears to be soluble in propylene carbonate and would be the first choice, unless you can come up with some serious reason against it.

[Edited on 14-3-2014 by blogfast25]

Zyklon-A - 14-3-2014 at 16:43

It seems like they would have mentioned it if NaCl worked. That would have been my first try, but I assumed it wouldn't work, or else someone would have used it. I'll give it a try, maybe it will work.

Marvin - 15-3-2014 at 04:09

I may try to get sodium from NaClO₃, as shown in blogfast25's link.

If by reference you mean the 1975 electrochem paper then according to the abstract they used sodium chloride, perchlorate NaClO₄, tetrafluoroborate and hexafluorophosphate.

blogfast25 - 15-3-2014 at 05:29

http://link.springer.com/article/10.1007%2FBF00608791

Abstract

The alkali metals were electrodeposited from various solutions in propylene carbonate (PC). Li, Na, K, Rb and Cs were electrodeposited from solutions of their chlorides and AlCl3 in propylene carbonate (PC) at ambient temperature. In addition, lithium was deposited from LiCl, LiBr, LiClO4, LiPF6, and LiBF4 solutions in PC. Sodium was reduced from NaClO4, NaPF6, and NaBF4 solutions. Potassium was obtained from KPF6 solution.

A new process is proposed for the production and electrorefining of the alkali metals at ambient temperature. Alkali metal amalgam from the commercial mercury-chlorine cell is transferred into an AlCl3-PC electrorefining cell, where the alkali metal is dissolved anodically from the amalgam, and deposited in a pure form at the cathode.

MrHomeScientist - 17-3-2014 at 05:55

Quote: Originally posted by blogfast25

So the deep eutectic solvent didn't really work out?

Sorry blogfast, I just now saw your reply. I've made two batches of the deep eutectic now, the second one after drying my choline chloride in a dessicator for a week. They both will melt into a liquid while being heated, but afterward they slowly start to solidify. My first batch is full of little needles, and the second is mostly solid now. I need to decant the liquid portion and continue experimentation with that, but I just have been unable to do stuff in the lab for the last 2 months. I definitely haven't given up, it's just slow going.

Zyklon-A - 17-3-2014 at 06:45

Wow, thanks for the catch, blogfast25 and Marvin . 'Sodium chlorate' wasn't just a typo, I actually got confused, and was about to make some....

My PSU (that I am making) can put out 6.2 volts and 15.3 amps. I can also bring the voltage up to ~10, and keep the amperage the same. Which should I do?

BromicAcid - 17-3-2014 at 13:17

From the abstract it sounds like you need the aluminum chloride complex the same as when electrolyzing solutions in nitrobenzene. Interesting, wonder if the same potentially explosive complex coats the anode during this as well of if that is an artifact of the nitrobenzene.

blogfast25 - 17-3-2014 at 13:35

Quote: Originally posted by BromicAcid

From the abstract it sounds like you need the aluminum chloride complex the same as when electrolyzing solutions in nitrobenzene.

Are we talking AlCl4- here?

@MrHS:

Keep calm and carry on!

@Zb:

Probably very, very gradually increase voltage and see how amps respond. At first you may still be electrolysing minute amounts of water present in the mix, as it is much easier to reduce than Na+.

[Edited on 17-3-2014 by blogfast25]

Zyklon-A - 31-3-2014 at 09:30

Has anybody ordered from NEOBITS before? If so, how long does it usually take for the shipment to arrive?

[Edited on 31-3-2014 by Zyklonb]

elementcollector1 - 31-3-2014 at 10:12

No idea on the order, but be patient - in the olden days, orders could take months.
Does anyone have any reference to the solubility of other compounds such as rare earth salts in propylene carbonate? If they are, this would be a sweet route to neodymium, cerium and lanthanum, and also possibly samarium.

Brain&Force - 31-3-2014 at 12:09

http://www.sciencedirect.com/science/article/pii/00134686828...

Not a free article, but somebody has tried reducing europium, ytterbium, and samarium in the solvent, but they were reduced to their divalent forms rather than the free metals. Apparently the trifluoromethylsulfonates are soluble in propylene carbonate.

deltaH - 1-4-2014 at 02:03

I have an idea for you to try if you so wish, let's call it pulling a half-Pok

Why not use magnesium as a [sacrificial] anode, an alkali hydroxide as electrolyte (also supplying the metal to be reduced at the cathode) and propylene carbonate as solvent?

The idea is to oxidise magnesium at the anode and form magnesium hydroxide sludge (as it's very insoluble) and reduce whatever metal from the hydroxide you employed at the cathode. This may well give you an ambient electrolytic route to the alkali metals.

I am unsure whether alkali hydroxides would react with propylene carbonate at ambient conditions. However, according to the reference I've attached, it seems common practice to employ sodium hydroxide as a catalyst in carbonate exchange reactions, so for example refluxing ethylene/propylene carbonate with methanol and sodium hydroxide yields ethylene/propylene glycol and dimethyl carbonate. So I think it might be ok provided you don't run your cell too hot.

The nice thing about this is that you shouldn't need to run the cell with too much power (relatively speaking) because you make use of the large reduction potential of magnesium metal. This should also help you to not generate excessive heat when running the cell, though some heating is unavoidable. Excercise caution especially as propylene carbonate is high boiling; a heat sink on your cell may be a good idea if you're planning to put through substantial current!

This way you also don't need to worry about making the anode out of anything special and I assume magnesium is easy enough for the amateur chemist to obtain.

Attachment: Organic carbonates.pdf (379kB)
This file has been downloaded 517 times

[Edited on 1-4-2014 by deltaH]

MrHomeScientist - 1-4-2014 at 05:46

The idea is to oxidise magnesium at the anode and form magnesium hydroxide sludge (as it's very insoluble) and reduce whatever metal from the hydroxide you employed at the cathode.

Insoluble in water perhaps, but do you have any data on its solubility on the solvent in question?

deltaH - 1-4-2014 at 06:30

Quote: Originally posted by MrHomeScientist

The idea is to oxidise magnesium at the anode and form magnesium hydroxide sludge (as it's very insoluble) and reduce whatever metal from the hydroxide you employed at the cathode.

Insoluble in water perhaps, but do you have any data on its solubility on the solvent in question?

I'm afraid not and indeed solubilities is at the core here. I don't expect it to work for lithium though as lithium hydroxide behaves similarly to magnesium hydroxide as solutes go, both of which tend to be rather insoluble.

But large cations are usually more soluble in organic solvents than smaller ones, so potassium hydroxide is probably ones best bet and if it were me, I'd start there and then perhaps try sodium hydroxide if I can get it to work with potassium.

On a similar note, Ba²⁺ and K⁺ have similar ionic radii, so if one could get it to work with potassium hydroxide and if you can get it, one might be inclined to try barium hydroxide.

Sadly, I beleive the goal of lithium or sodium via this route will remain ellusive, but hey, even 'just' obtaining potassium or barium metal would be rewarding, no?

[Edited on 1-4-2014 by deltaH]

Zyklon-A - 1-4-2014 at 06:51

Sounds great! I shall try potassium first.

deltaH - 1-4-2014 at 06:59

Aside from poor solubility, I also think dissolution may be very slow! In fact, if one is going to do this in appreciable amounts, having some prills of hydroxide stirring round in your electrolyte may be a good idea so as to replenish it in solution as it is consumed, just in case the solubility is very low, which is most likely! Hmm... this is starting to sound like a challenging build, GOOD LUCK!

[Edited on 1-4-2014 by deltaH]

Zyklon-A - 1-4-2014 at 07:35

Yeah, well I probably have plenty of time, while I wait for my PC (propylene carbonate), I don't have KOH at the moment, but I can make some perhaps.
Got some NaOH though.

deltaH - 4-4-2014 at 06:46

I found this datasheet (see attached) for DMSO that lists solubilities of a wide range of inorganic salts. I figured, failing good data for propylene carbonate, DMSO is start for a comparison as it is also a highly polar aprotic solvent. But take note, with a lower dipole moment of 4 D, compared to propylene carbonate's 4.9 D!

From this data it appears that my suggestion to start with potassium hydroxide was a poor one as it is less soluble than sodium hydroxide in DMSO, with solubilities of 13mg KOH /100g DMSO and 35mg NaOH / 100g DMSO, respectively, both of which are very low.

But lithium halides have no issue, lithium chloride happily dissolving at 10.2g / 100g DMSO! Calcium chloride, on the otherhand, is listed as insoluble. Magnesium chloride hexahydrate is listed as 1g /100g DMSO, presumably the water present helps somewhat in the case of the hexahydrate salt.

Unfortunately, no data for magnesium hydroxide solubility even in DMSO

So, while experimentation will yield the answers, we have some basis to speculate that lithium as lithium chloride may indeed work with magnesium as an anode, particularly if one has excess lithium chloride so that it can replenish and maintain saturation as it's reduced as well as helping to surpress the solubility of the magnesium chloride forming on the anode.

Also, be warned, if your concentration of electrolyte is very high, do start with a low voltage and work up as I think you may find your cell highly conductive which may result in too much current flowing!

Attachment: Solubilities of substances in DMSO.pdf (1.2MB)
This file has been downloaded 1518 times

[Edited on 4-4-2014 by deltaH]

Zyklon-A - 4-4-2014 at 07:34

Unfortunately, I can't download it at the moment because I'm on a public computer, but I will as soon as I get home.
I think alkali halides will be easiest, mostly because they are easy to get, and cheap.
I've got lots of magnesium so an anode will be easy to make....
Should I use SS as a cathode?

deltaH - 4-4-2014 at 14:26

I've been meaning to ask what you plan to use as the cathode

Yeah SS should be fine, except it's not a great conductor and may get warm if you operate at high current density. Consider using nickel or plated nickel if you have something like that.

By all means, do also experiment with alkaline cells as well (especially if you already have sodium hydroxide), the data I found is for DMSO and while it is suggestive, we still don't know for a fact what the propylene carbonate will do.

When is your propylene carbonate supposed to arrive anyhow?

BromicAcid - 4-4-2014 at 15:37

I have done electrolysis of lithium chloride in DMSO dozens of times. But you do not get lithium metal that way... you only get ... a surprise.

Zyklon-A - 4-4-2014 at 15:54

Great, it looks like potassium perchlorate, nitrate and iodide are quite soluble.

When is your propylene carbonate supposed to arrive anyhow?

I don't know, my dad ordered it on his paypal account, so I can't track it.

BobD1001 - 4-4-2014 at 22:28

Well, I just ordered 1Kg of Propylene carbonate from Alfa Aesar, hopefully Ill receive it shortly, and get some time to experiment with it!

deltaH - 5-4-2014 at 07:52

BromicAcid, I wasn't suggesting one use DMSO, I was using the data as a comparison to try to infer something with regards to solubilities of inorganic salts in propylene carbonate. I would guess that if you made any strongly reducing metal in DMSO, you would quickly reduce it to dimethyl sulfide... a surprise indeed! Let me guess, it stank like &%#@

Zyklonb, I would strongly suggest you abstain from any oxidising anions which would probably react violantly with alkali metals. I'd say nitrates are definately out!

Halides should be fine and hydroxide are a maybe.

Remember the trick here is that you not only want the electrolyte to be soluble, you simultaneosly want the magnesium salt of that anion to be poorly soluble. I'd stick to the simple halides and possibly play with hydroxide just out of curiosity.

What you planning for a power supply?

The thought also occured to me that if you manage to dissolve some hydroxide in propylene carbonate, maybe you may want to try an aluminium anode as well, hypothetically forming sodium aluminate precipitate (if using sodium hydroxide off course). If that works, it would be even cheaper and easier than magnesium. However, I think halide electrolytes are out with aluminium as aluminium halides are probably very soluble in these kinds of solvents.

[Edited on 5-4-2014 by deltaH]

Zyklon-A - 12-4-2014 at 18:25

Quote:

Zyklonb, I would strongly suggest you abstain from any oxidizing anions which would probably react violently with alkali metals. I'd say nitrates are definitely out!

I guess, many of my potassium salts are oxidizing, so I thought that would be cheapest.

Quote:

What you planning for a power supply?

Hopefully I will buy this soon.

Oscilllator - 13-4-2014 at 00:27

I really dont think sodium hydroxide is a good idea, because it ALWAYS has some water in it - a few percent at least. Therefore you will probably have problems forming sodium metal.
@DeltaH - Dont you think using sodium hydroxide and an aluminium anode is a bad idea? Sodium hydroxide reacts spontaneously with aluminium as I'm sure you know. Or does the reaction require water?

deltaH - 13-4-2014 at 01:41

Nice current capability, but I'm concerned about the single out 5V. I think for Mg anodes this might be overkill, in short, a lot of money to spend on a gamble. Why not start with a few D cells and experiment, then later you can source the right kind of power supply for your needs with greater confidence, knowing better how your system behaves?

Oscillator, I thought at first it would, but then wrote out the equation:

Al + NaOH + 3H₂O => NaAl(OH)₄ + 1.5H₂

So yeah, it's water that's actually getting reduced, the sodium hydroxide is just a spectator, but an important one, as without it the aluminium is kinetically unreactive because of it's passivating oxide layer. In fact, maybe a nice way to dry the solvent with electrolyte in it, provided you have some means for the hydrogen to escape. (PS. google ALICE rocket when you have some time for a cool extreme of this

)

As for the water in the hydroxide with magnesium, all that should happen is that the cathode will bubble off hydrogen in the beginning, as blogfast mentioned, until the water is consumed, a bit of a waste of magnesium, but not end of the world.

***

BTW, the E standard calculated from Wiki's table of standard electrode potentials is a mere 20mV for the cell: Mg(s)|Mg(OH)₂(s)||Na⁺(aq)|Na(s), while this will obviously not be the same for an organic system, it is suggestive that these kinds of cells will be current beasts, requiring very little work to make large currents flow. In such a situation, all the secondary resistances become important, e.g. the need for good electrolytic conduction which implies decent solubility and mobility of the ions, also good electrical conductivity on the external circuit and plates, etc. All these can introduce significant losses in this system because the voltage drop due to the REDOX electrochemistry per say won't be very large at all.

[Edited on 13-4-2014 by deltaH]

Zyklon-A - 13-4-2014 at 08:22

Yes, the 5 volts will be overkill, but I want that PSU anyway for other projects - specifically for large(ish) scale chlorate production, I could lower the volts to ~ 3 or so.
I've got plenty of Mg, so no problem there. I also have quite a lot of MMO, so if that's useful for anything....
I found a PDF which says K was isolated from potassium tetrachloroaluminate (KAlCl₄), in propylene carbonate. He used either potassium or aluminum anodes, both of which worked, at least in this particular case.
This Document has lots of information on electro-depositing reactive metals, and is very useful in this discussion.
Here is the PDF:

Attachment: eScholarship UC item 8sz1w229.pdf (1.8MB)
This file has been downloaded 808 times

BTW, it also gives a procedure overview of making Potassium Tetrachloroaluminate, if you're reading this blogfast25 - if you're still interested in making it.

[Edited on 13-4-2014 by Zyklonb]

deltaH - 13-4-2014 at 10:53

Thanks Zyklon for the paper.

Now we know that potassium tetrachloroaluminates works for these kinds of cells and they got decent amounts dissolved into PC (at least 0.5M).

In your context, it could mean that you can enhance the solubility of whatever alkyl halide salt you with to employ in your cell by adding anhydrous AlCl3.

Chlorine producing cells don't seem like a good idea because of their reported degredation of the solvent, especially in the pressence of water, however, using a sacrificial anode probably solves this problem as no chlorine would be evolved and one would also be able to run at a low potential.

I'm still uncertain whether PC will tolerate hydroxides dissolved in them. My gut feeling says that the hydroxide ion, being a strong nucleophile, may attack PC quickly and in the pressence of small amounts of water, forming carbonate salts and ethylene glycol?? But maybe this stops as soon as you consume the water present?

Anyhow, as things stand, alkyl chlorides, possible with added AlCl3 to enhance solubility, would seem the electrolyte of choice using a magnesium anode so as to not produce free chlorine and drop the powder requirements of the cell drastically.

[Edited on 13-4-2014 by deltaH]

Zyklon-A - 13-4-2014 at 11:14

Yeah, So when it is deposited on the electrode, how will it attach to it? Will it stick to it, or crumble off?
Either way, I think NaCl will be easiest, although I might try a K salt first. Hydroxides aren't known to work AFAIK, so I won't try that first.

deltaH - 13-4-2014 at 11:23

Quote:

Hydroxides aren't known to work AFAIK, so I won't try that first.

Agreed! According to the article, the potassium didn't attach on very strongly and did flake off. Who knows what will happen with other metals. But I think it's important that you establish the solubility of your electrolyte. Doesn't help you use NaCl if it won't dissolve! LiCl may be very soluble though if we infer a similar behaviour to the DMSO data I dug up. We also now know that anhydrous AlCl3 can help you to dissolve more electrolyte if need be.

[Edited on 13-4-2014 by deltaH]

blogfast25 - 13-4-2014 at 11:48

Thanks Zyklon for the paper.

Now we know that potassium tetrachloroaluminates works for these kinds of cells and they got decent amounts dissolved into PC (at least 0.5M).

It also means you need to prepare KAlCl4, no sinecure at the hobby level.

Zyklon-A - 13-4-2014 at 11:51

Yes, but it does give a procedure to make it, it's not easy, but possible.

deltaH - 13-4-2014 at 11:57

I was thinking of 'simply' co dissolving anhydrous AlCl₃ and KCl in reasonably dry propylene carbonate and not so much preparing incredibly anhydrous KAlCl₄ by fusing salts under anhydrous conditions as the authors have done.

In this respect, I'm viewing AlCl₃ as a solvant aid when added to the solvent, enhancing the amount of KCl that would otherwise have dissolved. Am I missing something obvious here?

I know hydrolysis can be a problem for such salts, but surely not when the water present is much less than KCl present?

[Edited on 13-4-2014 by deltaH]

blogfast25 - 13-4-2014 at 12:21

Am I missing something obvious here?

I know hydrolysis can be a problem for such salts, but surely not when the water present is much less than KCl present?

[Edited on 13-4-2014 by deltaH]

KAlCl4 is in fact a eutectic mixture:

http://www.crct.polymtl.ca/fact/phase_diagram.php?file=AlCl3...

... with an astonishingly low MP (about 250 C), for a salt. The point at molar fraction AlCl3 = 0.5 is where the composition corresponds to KAlCl4.

I think your proposed procedure wouldn't work but I'm not sure why. For starters, if it can be done that easily, why bother preparing the KAlCl4 separately as they did in that paper?

Water is of course a great enemy of AlCl3, which causes it to hydrolyse. Minute amounts of water can never be entirely avoided, of course. I would certainly go to every bit of reasonable trouble to eliminate water as much as possible.

[Edited on 13-4-2014 by blogfast25]

Zyklon-A - 14-4-2014 at 09:54

Quote:

KAlCl4 is in fact a eutectic mixture:

In that case, it could be made by heating stoichiometric amounts of KCl and AlCl₃ until they melt right? That's how they did it in the PDF. The great thing is, iy only has to be bought once. The cell only consumes KCl, leaving AlCl₃ in solution. So, one can just replenish the used KCl, by measuring the chorine produced and adding KCl stoichiometricly:

Quote:

Originally posted in said PDF
The process used to study the stability of the solvent with respect to chlorine evolution was the electrolysis of KA1Cl4 in propylene carbonate, Chlorine was produced and evolved at the anode.
Potassium was reduced or deposited at the cathode. The overall reaction involved
is:
KAlCl₄ ↔ K⁺+ Cl^- + AlCl₃
The half reactions at the cathode and anode are, respectively:
K⁺ + e^- → K, and Cl^- + Cl^- → Cl₂ + 2e^-.
Assuming a 100% current efficiency, two moles of potassium were deposited
for every mole of chlorine evolved.This proposed chlorine evolution anodic reaction was quite attractive
from the point of view of raw material considerations. Theoretically only
KCl is consumed in the electrolysis, It can be replenished by simply
adding KCl into the system. Hence this process is capable of producing
valuable chemicals, potassium and chlorine, from a relatively cheap
source, KCl. The most important consideration of all, however, is that
the system can be operated at room temperature.

This is exactly what I've been looking for!

[Edited on 14-4-2014 by Zyklonb]

12AX7 - 15-4-2014 at 00:04

Willfully stepping into the middle of a thread without reading;

Can you get propylene carbonate from lithium batteries?

Nevermind that they already have elemental lithium in them to begin with...

LiPo batteries are normally charged to 4.20V absolute maximum, with the claimed reason that, any higher and lithium metal is deposited, which is a shorting hazard, bla bla, fire explosion bla bla. Might it be feasible to, say, remove the lithium anode and separator membrane, and use that as the cathode in a solution with LiCl, and graphite (or something) anode? The added separation between electrodes, of course, helping prevent shorts, and the added volume providing for more lithium content.

In response to the posts I see above; I'm surprised KAlCl4 doesn't make aluminum!

Tim

blogfast25 - 15-4-2014 at 04:35

Quote: Originally posted by 12AX7

Can you get propylene carbonate from lithium batteries?

It would appear so. How many batteries will you have to break into to get a ml of this solvent though? Which then will need purification, of course.

A week ago I extracted some Li from an Energizer AA, it reminded me just how hard and time consuming cracking open these little fortresses actually is.

Zyklon-A - 15-4-2014 at 08:38

Extracting Li from batteries is a pain it the @ss. I assume that getting PC from batteries is even more of a time waist. But you could try. This weekend I plan on making anhydrous aluminum (III) chloride, which will be used to make potassium tetrachloroaluminate. I might have to buy a nafion membrane, which as seen here, is likly going to be rather expensive. Do I really need that membrane? Or will a simple membrane do the job?

blogfast25 - 15-4-2014 at 08:43

Do I really need that membrane? Or will a simple membrane do the job?

Not strictly speaking, no. But it does prevent chlorine moving from the anode towards the cathode.

Zyklon-A - 15-4-2014 at 11:33

Hmm, I wonder what membrane is used in lithium batteries. Perhaps it could be used, I don't have any of the remains of the Li batteries that I've taken apart, but I still have two or three fresh batteries... Seems like big companies can afford to put a good membrane in a battery, considering how expensive they are in relation to the amount of Li in them.

Here's some more information useful to this discussion:
http://epub.uni-regensburg.de/22953/1/ubr11901_ocr.pdf

You have to pay to view this whole paper, but here's some of it:http://pubs.acs.org/doi/abs/10.1021/j100639a008

Here's something else I found:Potassium Carbonate as a Salt for Deep Eutectic Solvents.
It has to do with Room temperature ionic liquids. I haven't read the whole thing yet.

Here's a short paper on topic: Investigation of the Electrochemical Windows of Aprotic Alkali Metal (Li, Na, K) Salt Solutions
Here's a patent that shows a method for preparing high-purity propylene carbonate and for simultaneously making passivated electrodes.
For $278.00, you can buy four liters of PC

[Edited on 15-4-2014 by Zyklonb]

blogfast25 - 15-4-2014 at 12:21

This weekend I plan on making anhydrous aluminum (III) chloride, which will be used to make potassium tetrachloroaluminate.

Don't forget to report that here now. I'm planning Al + 3 HCl for the next, next week end.

Interesting links, BTW...

Zyklon-A - 15-4-2014 at 13:06

OK, I plan on doing 2 Al + 3 Cl₂ (g), rather than HCl (g). Chlorine is slightly cheaper to make than gaseous HCl, at least for me. H₂SO₄ + 2 NaCl → 2 HCl(g) + Na₂SO₄ is more expensive for me, as I only have < 1L 98% sulfuric acid which cost me ~ $20, while I have a gallon of con. HCl(aq) which cost me only ~ $6.
4 HCl(aq) + Ca(ClO)₂ → CaCl₂ + 2 Cl₂ + 2 H₂O. How are you going to make your HCl(g)?

[Edited on 15-4-2014 by Zyklonb]

plante1999 - 15-4-2014 at 13:14

Don't forget to dry the chlorine, sulphuric acid works great for that.

Zyklon-A - 15-4-2014 at 13:18

Yes, I might bubble the chlorine through sulfuric acid, or just use a CaCl₂ drying tube, (which I already have). If sulfuric acid is used, it will only be diluted right? With the possibility of dissolved chlorine, which can be gassed off with heating?

plante1999 - 15-4-2014 at 13:27

Yes, you simply got to boil it down to re-use it.

blogfast25 - 16-4-2014 at 04:45

How are you going to make your HCl(g)?

[Edited on 15-4-2014 by Zyklonb]

NaCl + H2SO4 ('Draino; is good enough)

Zyklon-A - 16-4-2014 at 09:14

Ah, that's what I figured, I will do a small experiment with making AlCl₃ today, maybe a few grams. I also will make a chlorine generator to make a lot of it, but that will be much later.

elementcollector1 - 16-4-2014 at 10:22

So, if the cell only consumes K⁺ and Cl^-, while leaving AlCl₃ alone, what exactly is the point of preparing anhydrous KAlCl₄? Seems like it would amount to the same result in terms of ions in solution. It would certainly be something to try...

Zyklon-A - 16-4-2014 at 10:50

I'm not entirely sure to be honest. But I doubt they would have gone through the trouble to make it if it wasn't needed. Whatever the case, I'm going to make it, as I want to stick to what I know works. I could skip the AlCl₃ and just buy it, but making it is much cheaper.
Buying potassium tetrachloroaluminate is out of the question, I could only find it at SIGMA, and even there, it appears to be discontinued.. Whatever that means.

I was thinking of 'simply' co dissolving anhydrous AlCl₃ and KCl in reasonably dry propylene carbonate and not so much preparing incredibly anhydrous KAlCl₄ by fusing salts under anhydrous conditions as the authors have done.

I have no idea if this will work, although blogfast25 seems to think it won't, so I'll trust him until further experiments explain the "ifs" and "whys".

Quote: Originally posted by blogfast25

I think your proposed procedure wouldn't work but I'm not sure why. For starters, if it can be done that easily, why bother preparing the KAlCl4 separately as they did in that paper?

Water is of course a great enemy of AlCl3, which causes it to hydrolyse. Minute amounts of water can never be entirely avoided, of course. I would certainly go to every bit of reasonable trouble to eliminate water as much as possible.

[EDIT] It seems to me that fusing the salts wouldn't be that hard. Probably $15 worth of equipment and an inert/anhydrous atmosphere. It will require a couple days of planning, a hot torch, and several hours of free-time. Nothing I and many others haven't done, and don't do time and time again.

[Edited on 16-4-2014 by Zyklonb]

deltaH - 16-4-2014 at 11:43

I look forward to reading about your efforts in preparing KAlCl₄.

In regards to making anhydrous AlCl₃, I recall my supervisor once upon a time lamenting having to sublime it for purification. Wiki quotes a sublimation temperature of a mere 180C, so... perhaps you may want to look into methods that prepare it chlorothermally (if they even exist/are feasible) killing two birds with one flaming ball of phosphorus?

...and no... I don't mean igniting ferric chloride and aluminium powder... that would probably be rather suicidal!

[Edited on 16-4-2014 by deltaH]

Zyklon-A - 16-4-2014 at 11:52

What is chlorothermally?

deltaH - 16-4-2014 at 11:57

lol, your link was funny

, what I meant is a thermite-like reaction where you exchange chlorine instead of oxygen, the idea would be to pick candidates such that the reaction is not too exothermic as a gaseous product will be produced (and so could be potentially explosive if very exothermic), namely, sublimed aluminium chloride.

For example igniting zinc chloride and aluminium? Shooting totally from the hip, but you get the idea... (the latter being probably no good anyhow)

Or maybe lead (II) chloride and aluminium... but then who has lead (II) chloride just lying around and probably too exothermic, though does have the nicety that the chloride salt is anhydrous to start with and lead has a very low melting point and high boiling point.

Out of interest, using the NIST Webbook, I determined the standard enthalpy of vaporisation for aluminium chloride to be 121 kJ/mol, so as you can see, that 'sinks' some energy.

Anyhow, if there's no precedence for chlorothermal reactions like these, then it's probably too dangerous or unlikely to succeed to be worth it... was just a side thought.

[Edited on 16-4-2014 by deltaH]

blogfast25 - 16-4-2014 at 13:20

delta H:

Reductions of chlorides instead of oxides are fairly rare for several reasons:

1. the resulting chloride usually has much lower heat of formation than the corresponding oxide, making these reactions less thermodynamically favourable. AlCl3 is a case in point because it's only a semi-ionic compound, with much lower enthalpy of formation than alumina. For chlorides, Al is therefore a fairly poor reducing agent.

2. the resulting chloride is usually much more volatile than the corresponding oxide. This creates problems of containment.

One important chloride reduction does spring to mind: the Kroll process for Ti and Zr. MCl4 + 2 Mg === > M + 2 MgCl2. It is not an 'easy' process.

The reductions of FeCl3 and ZnCl2 with Al I predict (off the top of my head, without calculation) not to be thermodynamically favourable (Delta G > 0).

But because of its volatility, removing AlCl3 from the reagent mix may make these reactions practically possible nonetheless. If the target chloride is itself somewhat volatile this may impose limitations on temperature.

All in all direct synthesis of AlCl3 seems the 'easiest' route by far.

Where oxidic reductions aren't an option for whatever reason, fluoridic ones can be contemplated.

[Edited on 16-4-2014 by blogfast25]

Zyklon-A - 16-4-2014 at 13:29

Ah, yes that makes sense. I will distill it anyway, because the reaction of chlorine and aluminum is very exothermic, and will easily vaporize the aluminum chloride.
BTW, is it possible to do a chloral-thermal reaction with something like sodium chloride and aluminum oxide? Of course it won't really be like a thermite, because it will be highly Endothermic , but if heated very hot, will the low boiling point of aluminum chloride drive the equilibrium to the right?:
Al₂O₃ + 6 NaCl <--> 2 AlCl₃ + 3Na₂O.
I didn't have much luck with regards to making AlCl₃ today, but I will try again. I didn't really have a good apparatus to control the reaction, but I just got a nice SS pipe that should work great!
I will report my progress when it's done.
[EDIT] I didn't see blogfast25's post until now, question partially answered.

[Edited on 16-4-2014 by Zyklonb]

blogfast25 - 16-4-2014 at 13:33