Attack on glass during molten salt electrolysis? / Chemistry / Forums

Forums

4hv.org :: Forums :: Chemistry

« Previous topic | Next topic »

Attack on glass during molten salt electrolysis?

Move Thread

LAN_403

Chris

Tue Feb 06 2007, 06:20AM

Registered Member #8 Joined: Thu Feb 02 2006, 04:34AM
Location: Harlowton, MT, United States
Posts: 214

Tonight I made my first attempt to electrolize a eutetic mixture of lithium and potassium chloride to produce lithium metal, such as used in commercial lithium metal production. It didn't work (it seems like nothing I ever make does

). If I get this to work, I will make another thread about it, talking more about the design and the metal I got. For now I would just like to see if anyone can give me any insight into what happened to the glass and screen barrier I had between the anode and cathode in this setup.

The setup is described in the images below. It resembles almost exactly my larger downs cell for sodium or a large industrial cell, except that this is only a small open setup and is not continuously fed or drained. This salt mixture melts at 500Â°C or less (eutetic mix is only 405 or something), and the melt went very smoothly, with no cracking to the glass tube divider. This temperature is far below even the annealing point of the glass, which is just cheap soda-lime glass from an old fluorescent tube. After the crucible was filled with molten salt, I turned on the argon, and slowly turned up the electrolysis current. Stuff started to happen around 3V, but much more rapidly at 5V which is where I carried out most of the run. No current measurement was taken, but I will do that next time if I can manage to scrounge up a shunt and meter. Chlorine gas was produced at the anode, which is simply a graphite arc cutting electrode. Shiny blobs which appeared to be lithium metal did appear to be finally accumulating later in the run, but then the surface of the bath smoothed out again and it was gone. It is difficult to tell if this was actually metal at all, or just shiny bubbles of salt.

In any case, upon removing the upper part of the aparattus and cleaning it off after the run, the screen (which was made of fine stainless steel mesh, and is typically made of iron gauze in industrial cells) was completely dissolved away below the level of the salt bath. I don't know if this is just unavoidable in such a small cell because it's too close to the anode or what. If this is the case, I would be very concerned about the screen in my big downs cell too, though I should be able to keep the anode above the screen in it if I wish. If all else fails, I guess I could just plate it with gold or platinum. I guess I don't fully even understand the function of the screen in these cells, but I thought it was supposed to repel the ions in the bath to keep them separated.

More worrying is what happened to the glass tube after the electrolysis. The glass turned frosty, and a piece was missing. It turned very brittle and soft and just crumbles away in your hand, almost as if it is partially dissolved. I thought glass was impervious to both hot chlorine and molten chlorides of all sorts. It seems unlikely that is this any kind of thermal damage, since the temperatures involved should have been below even the strain point of the glass, and well below the annealing and softening points. Does anyone know why the glass was so attacked? If it is so easily attacked, I would worry about the glaze on my big cell; if it fails, the alumina based ceramic beneath will be quickly eaten, resulting in possibly catastrophic failure and explosion. What on earth do they use for isolation in the industrial cells if glass cannot hold up? Does the soda or lime in the glass make it basic enough to be attacked by chlorine or something?

Bored Chemist

Tue Feb 06 2007, 12:54PM

Registered Member #193 Joined: Fri Feb 17 2006, 07:04AM
Location: sheffield
Posts: 1022

I think this might be due to the fact that hot glass conducts electricity quite well. In particular IIRC it acts as an ionic conductor (ie ions, rather than free elctrons, travel through it). With the setup you have there is a significant potential across the glass. This may be high enough to drive Li ions from the melt into the glass. I don't know enough about glass chemistry to be certain, but I suspect that glass with added Li would have a lower MP than the original glass.

I believe that one of the few uses for which asbestos is still permited is the manufacture of porous membranes for NaCl electrolysis, but I think that's aqueous rather than molten.

Chris

Tue Feb 06 2007, 03:10PM

Registered Member #8 Joined: Thu Feb 02 2006, 04:34AM
Location: Harlowton, MT, United States
Posts: 214

If hot glass is such a good conductor of ions (remember this is even below the strain point) and cold glass isn't, where does this transition start? Will it conduct electric current like a regular conductor as well or only act as a membrane? If glass is such a good conductor at high temperature then how is it possible that molten salt cells can exist as they obviously need electrical insulation at some point? Ceramic is easily attacked by chlorine and its ions. It seems there is no material in the world which would work, certainly not one that was known in Humphry Davy's day. Will ceramic glazed with silica based glass (the divider material for my large downs cell) hold up in this position during molten NaCl/CaCl2 electrolysis at 600C?

In typical industrial cells, this divider is made of steel (not sure if it's even coated with anything!) and it doesn't seem to cause any problems. The insulation is around where the cathode enters the cell, but it still has potential across somewhere.

Steve Conner

Tue Feb 06 2007, 03:50PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

The original Downs cell is US Pat. # 1,501,756. If you look that up on a patent search engine, you can see exactly how it was done. I think the separator is made of steel or iron. This patent was also quite interesting:

While searching, I saw another patent for an electrolytic refining cell that used glass as a membrane to let sodium ions through. Glass conducts electric current only in the form of ions, but they can be the ions that the glass itself is made of. If you heat a glass rod to red heat with a torch, with an alligator clip lead on either end, you can pass enough 120V current through it to light a bulb.

Hazmatt_(The Underdog)

Wed Feb 07 2007, 02:38AM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

If you're going to use a glass in your apparatus you shouldn't use soda glass. Even if you heated it all evenly, allowing a breeze to pass by will thermal-shock the glass causing it to crack later.

I have thermal-shock problems even with boro glass, but that's because I don't have an oven yet.

If you're looking for something more rugged you're going to need either a heavy wall boro or quartz. Both can be had from glass suppliers like glass craft and Wale apparatus.

Chris

Wed Feb 07 2007, 03:45AM

Registered Member #8 Joined: Thu Feb 02 2006, 04:34AM
Location: Harlowton, MT, United States
Posts: 214

Thanks guys. I think I know now where I can place plated barriers in my big downs cell to eliminate these problems while maintaining insulation between the cathode and screen. The alumina based ceramic underneath the glaze should insulate fine then and be protected from attack. I'm not sure the composition of the glaze that was going to be used on the ceramic parts of the cell, but I know it is rather high temperature, rated to like 1800F (meaning it is probably a composition with a strain point or maybe softening point higher than that, probably very high silica and resiliant). Pure silica glaze would be ideal. I guess I can use a steel tube with the screen brazed on or something for the divider in these little batch cells.

Bored Chemist

Wed Feb 07 2007, 09:25AM

Registered Member #193 Joined: Fri Feb 17 2006, 07:04AM
Location: sheffield
Posts: 1022

Q.
If hot glass is such a good conductor of ions (remember this is even below the strain point) and cold glass isn't, where does this transition start? Will it conduct electric current like a regular conductor as well or only act as a membrane?
A.
At room temperature for some glasses in thin layers like those used in pH probes glass will act as an ionic conductor.
Q.
If glass is such a good conductor at high temperature then how is it possible that molten salt cells can exist as they obviously need electrical insulation at some point?
A.
They don't try to use hot glass as an insulator.

Ceramic is easily attacked by chlorine and its ions. It seems there is no material in the world which would work, certainly not one that was known in Humphry Davy's day.

Please provide evidence that, for example, alumina is atacked by chlorine or chloride. Davy named aluminium so he must have been familliar with alumina.
Q.
Will ceramic glazed with silica based glass (the divider material for my large downs cell) hold up in this position during molten NaCl/CaCl2 electrolysis at 600C?
A.
God knows but I sugest you try it on a small scale first.

In typical industrial cells, this divider is made of steel (not sure if it's even coated with anything!) and it doesn't seem to cause any problems. The insulation is around where the cathode enters the cell, but it still has potential across somewhere.
I'd advise that you look at the patent cited above. (BTW, thanks for that Steve )

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob