Sciencemadness Discussion Board


 Pages:  1    3    5

neutrino - 8-12-2005 at 03:26

I didn't know cheap storebought lye and crappy electrodes could give that kind of purity. Nice. :cool:

Does anyone think they can answer my electrode question?

Current density?

Magius - 21-12-2005 at 20:41

At neutrino: As for current density, all the literature and patents I've read on Molten electrolysis points to high current density. For example This site makes a passing reference to high current densities of molten salts. In aqueous solutions, low current densities are desired to prevent errosion and side reactions, but in molten electrolysis...? It appears things are a tad different. I'm not sure, AP chem didn't teach much electrochem.

Now, call me ambitious, but within the end of January, I'll be recieving materials for the castable refractory mix mentioned in the Furnace thread, and aside from building an actual furnace for, building a downs cell is next on my list. The current design is some thing likeThis with the heating element coiled around the container, which is most likely going to be a coffee can. Both the annode and cathode will have the outsides coated with something nonconductive, probably... more castable refractory mix if it adheres to piping. The real problem lies in the production of chlorine. What material stands up to chlorine at 700C?

neutrino - 22-12-2005 at 03:21

Your first link is broken.

I'm still playing with sodium electrolysis. My current setup of two thick copper wire electrodes isn't holding up too well. The copper seems to be in tact, but there is a heavy nonconductive oxide layer on both (?) electrodes. Not sure how it would have gotten on the cathode, especially with the sodium there. The oxide kills my yields; at 20A I should have .3g/min, instead I have a total of 1g for several hour-long runs.

Well anyway, I'm planning on etching the oxide off with some sulfuric acid and plating a more resistant metal onto them. Would Ag or Pt hold up under these conditions? I was thinking about attaching some odd Pt foil scraps I had lying around, but I have no way of keeping them on the electrodes. Any suggestions?

The_Davster - 22-12-2005 at 13:05

Neutrino, despite the powersupply you are using is rated for 20A, it is not delivering that. 20A is the max current it could apply, the actual current is dependant on the voltage applied and the resistance of the sodium hydroxide. You could move your electrodes closer together(not always possible), or put another supply in series, so that instead of 5V you have 10V, that should double the amperage pulled.

neutrino - 22-12-2005 at 14:38

I would think the resistance of a molten salt wouldn't be high enough to cause such a large drop in yeild. I still think that the extreme resistance of the electrodes is the main cause.

In any case, I only have one PSU because these things are a little expensive. I will try to plate the electrodes with Ag and hope that changes things.

A qiuck question for those plating gurus here: do I have to start with a platinum salt to plate the metal, or is it possible to use some other electrolyte?

[Edited on 23-12-2005 by neutrino]

The_Davster - 23-12-2005 at 23:17

Arent your electrodes copper? Resistance of them is pretty much a non-issue here, especially considering the length of your electrodes. Get an ammeter, stick it in series with your cell, and experiment by manually bringing the electrodes closer together and further apart, you will see a difference in the current pulled.

EDIT: Realized you are talking about resistance of the oxide layer, not the actual metal, but the resistance of the salt is still important. Copper is really not ideal electrode material, you get the oxide even under moderate heating of the bare metal. If nickel does not form a sponge, it would be worthwile to try plating on copper.

[Edited on 24-12-2005 by rogue chemist]

len - 22-7-2006 at 20:20

As a kid I was one of 3-4 children at school fascinated by chemistry and doing his own chemistry experiments. We showed off to each other about what we could make - as you can imagine the experiments had to be sensational. But very few of them produced something I can now be proud of. About the most interesting thing I managed to make was sodium.

I placed about 0.5gm of sodium azide (NaN3) in a test tube, tied a party balloon to the top, clamped it and heated with a bunsen burner. After 20 seconds, pop, the balloon filled up (about 1/3 full). And in place of the white powder, a shiny globule, the size of a small pea appeared. The globule gave a good speedboat-on-fire display when dropped into a bucket of water.

As the years passed, the most important thing that helped improve results is not theory, but patience and perseverence. If you are not prepared to invest time (and money) dont even bother, you will get nothing.

Heres my procedure to produce sodium. Similar results have been posted, but this is simple, complete, and with a bit of a difference. The aim is to get, in about half an hour, a shiny specimen of Na sitting in a tube, to show to your friends.

You MUST wear a face mask. You need: a fully charged car battery, a long iron skewer, a jar of NaOH (drain cleaner), a hot plate, a stainless steel mug (you can get them at discount shops for $1), a long glass tube (about 5mm diameter), a suction device on it, some parafin, car starter leads, a clamp with stand. Do not substitute any other power source for the car battery, you need about 100Amps to get the sodium forming quickly so a) it doesnt dissolve/oxidise, b) it forms before the melt solidifies.

It strikes me that many people earlier in the thread were using far too low a current. It takes a lot of Coulombs to deposit even a little bit of substance. Theoretically 100000 are needed to deposit one electron mole, or 23gms of sodium, but that is a minimum. With cell efficiency capped at 50% (and 40% more likely) you will need about 1000C for 0.1gm Na (about the weight of a globule). At 10A that is 1min40s which is far too long conisdering the competing processes. At 100A its 10s, much better!

Place the NaOH in the mug and heat on a hot plate till it melts (wearing the face mask!) Attach the +ve of the 12V battery to the mug with one of the starter leads. The negative must be clipped to the skewer with the other starter lead.

The NaOH will melt in stages, first a pool will appear down one side due to uneven heating, then the pool will extend, until the whole mix has just melted. Some brown-grean discolouration will appear in the NaOH melt, due to attack of the SS and formation of Fe2+ and Fe3+, dont worry about it. If a crust remains on the surface of the melt due to surface cooling, push it into the mealt with an iron rod. Now the melt is at its most viscous and this is the perfect time to produce Na.

Dip the skewer into the middle of the melt, about 2cm, and clamp, at the same time turn off the hot plate. After about 30s globules of shinny Na will start to appear, adhering to the skewer. Move the skewer to coalesce them as much as you can. Do not conduct the experiment in a droughty area. The H2 produced at the cathode must be allowed to bathe the Na as much as possible. I added some Ar gas at times, but that does not seem essential. The melt will soon start solidifying, and before that happens, suck up as many globules as you can into the glass tube - they will look like mercury in a thermometer. You do not want any of the NaOH sucked up, so apply suction until the globule is not quite gone. Start with the smaller globules first. They can be sucked up, and then sucked back into other globules, to make one big globule. Dip the end of the tube into parafin, and you will have Na forever to show off to your friends.

Why turn the hot plate off as soon as the NaOH fuses? Why not collect more Na.? The setup I described is cheap and quick, the price for that is that it does not have the required temperature stability for a continuous process. I have found that most of the Na you get is produced, just after the melting. After that the mixture loses viscosity, and the sodium formed in microdroplets produces a constant whirl cathode-to-anode in the melt. At the anode it is destroyed. I then made a note that tight temp control is required, and am working on this. I hence read in this thread, and in a book bromicacid posted that tight temperature control is indeed required in the reaction - to within 5C! Go 25C over and no Na is produced, as I observed. NONE of the other 10+ sources on the Castner process mention this (and this includes a large chemcial technology encyclopoeadia series devoting several pages to the Castner process).

PS Has anyone read the original acount of Na production by Davy. Its full of interest, passion, and tips on how real discoveries are made. The stuff presented in school chemistry classes, is just a dried up stale version of what chemistry really is about. Most people who teach chem. shouldnt teach. Not only do they bore you to death, you will never succeed with what they teach you.

[Edited on 24-7-2006 by len]

excook - 25-7-2006 at 02:05

This will be my 1st and only post (most likely) but I decided that to save you gentlemen alot of wasted effort I'd give you my method of making sodium metal. The equipment needed: battery charger; large one with wheels ($100.00), stainless saucepan (1 quart or less in size), several 1/2" 6inch (or longer) carriage bolts ( stainless is best but galvenized works just as well), 1/2 nut, 2 feet of 1/2" steel conduit, 4 or 5 feet of copper wire (12 gauge or heavier), hot plate (temp. adjustible), several stainless steel teaspoons; wood to cover the ends of the handles and a means of attaching it to the spoons (screws, wire, etc.) a stainless bowl to fill with mineral spirits to drop the metal in, and ideally a box with a light a glass front, a place to insert your arms, with a exhaust hose that will take the fumes out of the room you set up in. and some more stuff that i'll remember as I go along.
First thing put your pan on the hot plate, weld (or some-how attach) the nut to the end of the conduit. Screw the bolt to this giving your self at least 2 inches of adjustment. To turn the bolt you'll need pliers; don't rely on your fingers to do it as it will be hard to turn when it's hot. Bend the conduit so the rounded head of the bolt will be suspended so as to just touch the surface of the melted sodium hydroxide (red devil lye and make sure it is dry; not a solid mass in the container)and attach it to your bench so it does not touch the pan. The copper wire is used to connect the battery charger leads to the pan and the conduit. You can use the chargers clamps, but after a couple of times they will be ruined. The negative lead goes to the bolt; the positive to the pan (you can reverse them when the metal production slows to keep produsing). prepare the pan (even if its new) by first washing it, then take muratic acid; wet a paper towel with it, wipe the spoons, bolt and the inside of the pan with it. Then soak some more paper towels in toulene and wipe everything with it. Everything must be clean. When dry turn on hot plate on high, put pan on it and pour 1/2 can of lye in it to melt; when melted pour the rest of the can in to melt. Put on a longsleeve cotton shirt put on some latex gloves and put on some cotton gloves over them; tape your sleeve ends so no flesh is exposed, the gloves won't fully protect your hands, but welding gloves (and the like), are almost to clumsy to wear.If not using a fume box, protect your self from flying lye and the fumes which will leave a coating of lye on everthing including your hair.
When 1st can of lye has melted put a spoon in it (to raise its temp.) put the bolt so its head is just touching the surface; a little off-center if needed to manuver the spoon around just don't let it contact the pan. Set the battery charger at the middle value (Around 30 amps) and turn it on. If everything was clean and the lye was dry, and attemps at hurrying the melting with a propane did not occur(it puts to much moisture in the lye with the burning propane) The sodium metal will form along the bottum of the bolt head, grow in size and then will breakoff and will go to the out side of the pan. They will keep combining and will grow to a size that will almost fill the spoon you will use to get it out of the pan and into the other pan with the solvant.
The lye is also forming hydrogen gas; the bubbles caused by excess electricity from the charger, These bubbles then shortout; you'll see a a electrical spark which ignits the hydrogen gas, causing flying molton lye. When the charger is on; it will produce enough heat to keep the lye melted,but, be prepaired for the Charger to shut down to cool-off. When this occurs turn the hotplate up to keep the lye melted. As the lye starts to produce less, add a couple of tablespoons of lye to keep it going. To produce one big piece from all of the little pellots that you will have collected; get a 1 cup stainless measuring cup, turn the hot plate on high, remove the pellots from solvant and place on paper towel and qiuckly absorbed most of the solvant. put the pellots in the measuring cup and with a pair of pliers; start swirlng the pellots; never letting the pellots to stay in one spot to long so they don't ignite. when the metal in molten and in one piece; quickly dump it into a tall glass filled with solvant; it needs to be tall enough that the metal cools and hardens before it hits the bottom. Cut away the green lye that will stick to the sodium. this method will work well if:
everything is dry and clean Use the metal carefully and legally

Organikum - 25-7-2006 at 03:01

Cyrus wrote:
If the sodium hydroxide gets too hot, it will just melt more NaOH

He got the idea. Thats exactly why the original "Castner Tiegel" is such a genius invention. The solid NaOH at the walls and at the bottom of the reaction chamber provides protection against corrosion, insulation AND temperature control.


len - 25-7-2006 at 04:27


He got the idea. Thats exactly why the original "Castner Tiegel" is such a genius invention. The solid NaOH at the walls and at the bottom of the reaction chamber provides protection against corrosion, insulation AND temperature control.

What it does is add some heat capacity to achieve better temperature control. Have you tried it out in practice, on a home scale rather than industrial Castner cell scale? Try to melt the NaOH just in the centre so it doesnt spread out to the walls, and then keep it just like that with current regulation. You'll be constantly adjusting current and turning your hot plate on and off. The heat conductivity of the NaOH is just too great. Either youll overshoot and melt the whole thing, or it will all solidify on you together with your sodium.

Whats important here is the ratio of heat capacity to heat conductivity (or time constant of the system). The heat capacity goes with the volume, conductivity with the area, so the problem does'nt scale.

We have electronics now, thats what its for. We can just sit back and watch the Na form :)

I dont see the point in the elaborate HCL/toluene cleanup suggested two posts above. Your NaOH will likely contain far more impurities, and even that will be insignificant compared to the impurities formed when the SS starts dissolving.

[Edited on 25-7-2006 by len]

Organikum - 25-7-2006 at 05:21

Using the pot as electrode is flawed. You obviously didn´t get what I wanted to say. Have a look at the original Castner Tiegel and you might understand.

len - 25-7-2006 at 05:37

Originally posted by Organikum
Using the pot as electrode is flawed. You obviously didn´t get what I wanted to say. Have a look at the original Castner Tiegel and you might understand.

Ive seen it many a time, and I know what you are saying. But Ill repeat what I said because you did'nt understand that. You cant achieve it manually on a small scale (i.e. in a SS mug). Ive tried it, you've no control. The molten region either quickly spreads out to the walls, or it all solidifies on you. In fact a central anode is a red herring here. One could still have all the benefits of the partially thawed NaOH conducting the electrolysis at one spot on the wall. It will still be equally difficult to achieve.

Anyway all this theorising is pointless cause you obviously have'nt done it. If you want, go do it, and post pictures of the melt and the pile of sodium you have made. I meanwhile will concentrate on something more achievable.

[Edited on 25-7-2006 by len]

len - 29-7-2006 at 04:33

Right well, Im developing ideas for a steady-state electrolysis to produce sodium in the Castner process.

The first thing to decide is whether to adopt a topside or bottomside cathode. The advantage of topside is simplicity, disadvantage is feasibility. First consideration is Organicum's idea of a bell over the cathode to collect Na. The idea of having it NaOH coated on the inside to prevent shorting by the sodium seemed feasible at first.

Alas I think it has a fatal flow. Density of Na is 0.95'ish, NaOH melt about 2.0. So about 50% of the Na is under the liquid. As the Na collects this will necessitate further lowering the bell into solution. Its tip will get hotter and eventually a short will form. Any ideas appreciated?

not_important - 29-7-2006 at 08:01

That's another problem with small setups, removing the sodium if you are making more than a small bit. When the apparatus is large enough it's easier to get access to the molten sodium for removal.

When I was much younger, the Ls and I tried this route. We used a couple of ceramic cups that we'd drilled holes in the bottoms of for bells. The cathode bell got a second, smaller hole that received a length of small steel tubing, bent into a squared U. The other end ran through the cork of a bottle containing mineral oil and having a second tube to pull a vacuum on. Oh, and the steel tube ended high in the bottle, not touching the oil. Ever so often apply a bit of vac to the out tube and pull a dollop of liquid sodium over into the oil. The steel tube was hot enough from conducted and intercepted radiated heat that the the sodium stayed liquid in it.

We had it working to some extend, running into the problems with eregulating the temperature of the NaOH but able to make 100 to 200 grams before thing went foobar. Then we ran into a source of sodium ingots, and drpped the electrolysis.

I remember that adding other salts to the NaOH would help, but it's so long ago I'm not sure what they were. Sodium carbonate comes to mind, at about 5% total, but that could be wrong and I'm sure there was a second salt added. They both lowered the melting point slightly, and reduced the tendency for the sodium to disperse in the hydroxide. We got it out of some turn-of-the-century inorganic chemistry book that had an odd lab technique with an industrial bent.

len - 1-8-2006 at 03:16

Does anyone know thermal conductivity of molten/solid NaOH?

not_important - 1-8-2006 at 03:48

This does have some information

My book on moltensalts in chemistry, which might have such information, seems not to want to be found.

The_Davster - 23-8-2006 at 13:57

I melted 50g or so of NaCl and electrolysed it today. When the salt is molten it is a nice glowing orange. Used nickel electrodes. Really annoying, the melt solidifies around where the electrodes are placed instantly. After fiddling with it a bit I actually was able to get a small lump of Na, which on contact with the air caught fire. I dont think my temp control was good at all(electric furnace on full) as after a while the melt just gave off white smoke when it was electrolysed. I am pretty sure that this is because the sodium was actually boiling off and reacting with the air. Nickel electrodes were corroded badly. Melt still stuck in crucible, so I am not sure of the damage to it. Got lots of the grey crap like in NaOH electrolysis.

Molten NaOH is much much easier to manage.....

[Edited on 23-8-2006 by rogue chemist]

len - 23-8-2006 at 15:47

rogue chemist, from what I read pure NaCl is unsuitable for electrolysis because at its melting point 801C Na actually forms an emulsion in it and wont separate into globules. CaCl2, something like 10% I believe, is added to lower the mp to somewhere around 650C. I believe it will be worthwhile trying that, keeping the current above 50A, and using thermally insulated Ni electrodes,counteracting the increased resistance by upping the voltage a bit. Also at this temperature its probably worthwhile keeping the Na under a bell through which you poke your electrode. The solidifying effect at the walls of the bell will isolate it from the solution.

The_Davster - 23-8-2006 at 16:04

Yes, there was definate emulsion forming, I will have to give the eutectic with CaCl2 a quick try. Seems like a very small ammount of Na excaped the emulsion in this attempt however.

Fleaker - 23-8-2006 at 17:17

Originally posted by len Alas I think it has a fatal flow. Density of Na is 0.95'ish, NaOH melt about 2.0. So about 50% of the Na is under the liquid. As the Na collects this will necessitate further lowering the bell into solution. Its tip will get hotter and eventually a short will form. Any ideas appreciated?

You can coat the collection/receiving pipe with a refractory (alumina-based) clay. At that temperature, it's electrical conduction is minimal but IIRC the sodium attacked the clay. Alternatively, one could position the electrode such that any sodium formed will flow into a bell with a narrow neck so that the majority of the sodium is stuck in the narrow tube.

Magpie - 23-8-2006 at 19:32

This evening was my first attempt at making sodium. I used a 5cm diameter food can (smaller than a soup can) and the PSU (12VDC) salvaged from an old computer. Cathode was a 1/16" iron wire (coat hanger) with a 1 cm loop in the end. See photo attached.

I poured about 3cm of granulated NaOH into the can and melted it with a bunsen burner. I felt I must measure the cell reistance before I dare apply electrical power so did so with an ohmmeter. I checked it several times and read 30-35 ohms each time. This seemed high but I really had no idea what it should be. This would be safe for my 12VDC (4a max) PSU so I applied electrical power. After a few minutes I checked the electrode loop for sodium but nada. Then I noticed that the PSU cooling fan was off so assumed I'd blown the PSU. I terminated the experiment at this point. I checked the PSU a few minutes later and it appeared to be fine.

So, I'm not sure what happened but wonder if a circuit breaker internal to the PSU shut it down. I really need to get an ammeter so I can more closely tell what's going on. It would also be nice to have a continuous temperature reading of the melt.

Any comments and/or suggestions are welcome.

[Edited on 24-8-2006 by Magpie]

[Edited on 24-8-2006 by Magpie]

[Edited on 30-1-2007 by chemoleo]

sodium.jpg - 63kB

12AX7 - 23-8-2006 at 21:49

If it's an ATX power supply, you need to add a jumper so it thinks the motherboard says "ok".

I've overcurrent-ed a switching supply or two before, I think. I'm not sure how they shut down, if it's a latch or what, and what resets it.

Next time, try the 5V output instead, or use a 1 ohm resistor in series.


chromium - 23-8-2006 at 22:35

Most computer PSU-s (no matter ATX or older) switch outputs immediately off when overcurrent occurs. If switched off mechanically or disconnected from mains for less than minute they are OK again. Maybe your electrodes touched each other for moment or maybe resistance of melt at higher voltages is much lower. Resistance of electrolytes is not the same at all voltages and your ohmmeter might be using very low voltage.

[Edited on 24-8-2006 by chromium]

The_Davster - 23-8-2006 at 22:48

I have pulled 18A out of a PSU line rated for 9A, they lie about max rated amperages. Ran it at this current for 3 days, it still works.

Stick a 10 W 10 ohm resistor on the 5V line, and be sure you connect the green and black wires. I think one of these is what Tim mentioned.

12AX7 - 23-8-2006 at 23:47

Eh, I once tried running a chlorate cell from an AT supply.

First try: it seemed to work. After some time, I discovered the 5V rectifier had failed shorted (probably after overheating). Replaced with beefier part.
Second try: choke on 5V output burnt to black. (30A rating with mere 16AWG wire?) Power supply still operational, but 5V went open circuit. Modded filtering scheme.
Third try: finally, the power transistors exploded...

My best guess is the cell wanted about 50A, which should be within the VA capacity of the supply...but go figure...

Also keep in mind the V/I characteristic of an electrolytic isn't ohmic, hence my suggestion of a series resistor.


len - 24-8-2006 at 05:04

Double triode (forget your part number) I agree. From my experience using an AT supply for Na electrolysis is like using nickel cadmiums to jump start you car. Not enough voltage and not enough current. I have ever only ever had success with a well charged car battery.

You need to aim at about 1 globule/10 sec to compete with air-oxidation diffusion (1 globule/10sec = 0.1gms/10 sec ~ 0.0005 mole/sec ~ 50 Coulombs/sec === 100 Amps @50% cell efficiency, but this is a theoretical maximum 150Amps more likely).

So you need 100A. The voltage drop in leads etc generally means you need more than 5V to push this current through in home setups.

If you really are desparate about using an AT supply you can parallel them (some of them will cut out in this mode, depends on the type). You can also series connect. You can even up the voltage by varying the sampling resistor used as feedback from the 5V supply rail to the PWM chip (its generally easy to find).

Magpie - 24-8-2006 at 11:58

I am fortunate to have a spool of resistance wire (1 ohm/31") so did some testing of my computer PSU with various resistance coils.

Ignoring the label limit for the amps for the 12v and 5v outputs I tested the 12v at 12a (P=144w) and the 5v at 25a (P=125w). I assumed that since the PSU is rated for 145w these power levels should be OK. They were, i.e., there was no shutdown of the PSU.

The attached photos show the two test coils under power.

[Edited on 30-1-2007 by chemoleo]

5v25aPSUtest.JPG - 51kB

Magpie - 24-8-2006 at 12:00

And here's the photo for the 12v test:

[Edited on 30-1-2007 by chemoleo]

12v12aPSUtest.JPG - 49kB

Elawr - 24-8-2006 at 12:12

Has anyone tried an arc welder? Many units for stick welding supply a few volts at hundreds of amperes. Most will supply direct current. I have used arc welders on occasion, but never for electrolysis. Perhaps sodium could be made. Another robust source of high amperage DC are those heavy-duty devices designed to fast-charge car batteries as well as jump-start car engines.

len - 24-8-2006 at 16:22

An arc welder is suitable current-wise, but its voltage is too high, typically 30-40V open circuit, plus you need the wire-fed DC variety. It can of course be rewound. A heavy duty jump starter would be just right.

Fleaker, thanks. I think that an alumina coating is more difficult than castners solution, which is having an underside electrode, plus it would be attacked as you say. I have though of the sodium outflow tube you mention, trouble is NA would have to flow down so the recepticle must be located below the surface of the NaOH melt!

[Edited on 25-8-2006 by len]

Magpie - 24-8-2006 at 20:05

Success this evening! :D with both 5v and 12v, but 12v was better. I also closed the cathode loop to 1/2cm ID. I don't think sodium at 308C has enough surface tension to stretch across the 1cm ID loop I tried previously.

There was the popping and throwing of hot NaOH previously mentioned by others. My hood glass served me well here. I used the series current limiting resistance coil of 0.2 ohm for the 5v and 1 ohm for the 12v. This was nice as I never had to worry about shorting out the electrodes.

Here's a photo of the few BB's I made with 12v. As you can see these are of varying purity:

Na BB's.jpg - 122kB

Elawr - 25-8-2006 at 04:52

Bravo Magpie!! I am seriously impressed with your sodium apparatus! From the picture it appears your yeild is substantial to say the least. Have you weighed it? I want to know how long it took to make your sodium. You were running about 12 amperes at 12V DC through your melt, right?

Great job! Thought about scaling the thing up?

Magpie - 25-8-2006 at 09:04

Thanks Elawr but I really just used the ideas of those ahead of me on this forum. As far as production rate with 12v it seemed like I was getting a BB about every minute. I haven't weighed them but will try to melt them together & purify later today. And yes, you are correct: the shorted electrode current would be 12a when using 12v. Of course this drops due to cell resistance and electrolysis when making sodium.

I tried to make sure I could always see some solid NaOH in the melt. As stated by others it is likely critical to keep the molten salt temperature at or near the melting point to avoid solublizing the Na in the NaOH melt.

The device is crude and there are several problems that you have to live with:

1. temperature control - trying to keep it even and near the melting point of NaOH (308C)
2. popping of H2 with attendant shooting bits of NaOH
3. low production rate for those interested in serious quantities of Na

On the plus side it is a great introduction to sodium making at a very low cost. The way we go though computers everyone likely has at least one old PSU laying around.

As far as will I scale up, well, I really don't need that much sodium so this will supply my needs for now. Building a really nice production cell that would solve the above problems would entail quite a bit more time & money, IMO.

whoa there

Magpie - 25-8-2006 at 14:25

I'm going to have to revise my earlier claims about producing sodium - at least of any purity. I placed my BB's in near boiling mineral oil (>250C) and they did not melt. When a BB is placed in water it vigorously emits H2 gas, but does not float and catch on fire like it should. They are grey and I assume are a mix of sodium and NaOH. I do remember seeing some very small droplets of silvery metal, however, but no significant quantity of them.

So, I'm beginning to see the light on what len has been saying about the requirement for high current. Tell me again how this NaOH forms by oxygen diffusion. Is this via the molten NaOH?

The_Davster - 25-8-2006 at 14:34

I bet there is just a coating of NaOH covering the sodium and keeping the sodium from melting into a single blob. A NaOH coating could also be increasing the overall density of the BB above that of water, causing it to sink. I've tossed lumps of sodium the size of several peas into water, they have never caught fire for me, perhaps some yellow sparks, but no fire.

Crush them up a bit, then melt.

len - 25-8-2006 at 17:52

Magpie, well done on making Na, but you will certainly be able to improve on that. The stuff I made is shinny like mercury (I sucked it up into a galss tube, and its still like that in my garage) and when I chucked some globules which remained in the melt after it solidified into water, the NaOH covering quickly dissolved off and a molten globule started skidding on water shooting of sparks from the burning H2, like Na should

The trick as you say is high current. A tiny drop of Na has a large surface area to volume ratio and its at that boundary where the destructive processes (O2 oxidation, emulsification in NaOH) occur. I believe that you want to form Na fast so the globules are small for a very short time, and you win the race against these processes. Slow rate of formation will lead to impure Na, and if too slow, to no Na at all.

[Edited on 26-8-2006 by len]

Magpie - 25-8-2006 at 18:39

len if you have read through this thread you know that there is a multiplicity of experiences and understandings of how electrochemical sodium is best produced. So forgive me while I dig up some background, but I'm trying to understand the root problems. My goal is to find the cheapest, easiest way for the home chemist to produce a few grams of high quality sodium.

Organikum enlightens us with drawings and operating conditions for the Castner Tiegel cell. I find it especially attractive that the cell voltage is 4v, as this would be perfect for the computer PSU 5v capability. Now to the amperage requirement. I understand what you are saying that time is not your friend and you must severely limit (<10 sec) the residence time of any formed sodium in the molten NaOH. Castner Tiegel does not do this. It uses a tight mesh screen to prevent gas bubble movement from one electrode to another. Is this just to prevent H2 movement or does it work for the O2 being produced at the anode also? Is the anode the source of the O2 that is fouling up sodium purity? Would the screen prevent this? I understand that the H2 produces a nice cover gas for the sodium to protect it from atmospheric oxygen and water. Castner Tiegel also provides for this. So if what I have stated is true it seems like we should be trying to optimize a Castner Tiegel cell on a small scale. In this way we would not have to depend on high currents and quick removal of sodium. I'm not implying that this would necessarily be optimum for every home chemist. It depends on whether you like to spend your resources on crafting a small scale Castner Tiegel cell, or buying a larger power supply, such as the battery charger on wheels recommended by excook. Or, as you have stated, a husky car battery that could always be brought to full charge when needed.

len - 25-8-2006 at 19:19

Magpie, most of what I read said the screen in the Castner is to prevent Na globules from penetrating to the anode, and keep it in the upside-down bell. I did not use it in my set up since I had a large anode-cathode separation.

I dont believe I state anything at variance with Castner. Castner cells use a high current of the order of thousands of amps, they can not be easily scaled down. I dont mean the Na must stay in contact with NaOH for < 10s. I meant that small globules should become big ones on that time scale. After that the Na floats in the NaOH, but it has a much smaller surface area exposed to it. The O2 just comes from the oxygen in the air, to which the Na exposes itslef once it floats. I believe the biggest source of impurity in your Na is 1) the Na had partially dissolved in the surrounding NaOH 2) it has oxidised. To cure the problems I suggest up the current and keep closer control on temperature. Between +5C to +15C over melting point of NaOH (I think its 318C when pure)

I have read most of the thread, and it is still my opinion that most people have been using too little current. I back that up by saying, and I dont want to cause offence, most results in the thread have been disappointing. Nobody has demonstrated any substantial sodium production (by this I mean more than a few random globules) or a useful apparatus. If you can give me an appartus to produce relatively pure Na with small currents, thats great, but I dont believe it can be done without going to fancy cells and inert gases.

I find the real worth of postings on this forum over say professional chemical literature, is that they lie here less, people dont mind saying when something has not worked out, or descibing things in real terms. It would be true to say my profession, physics, is dying a slow death because of all the lies (if you dont publish faster than the next guy - you perish, and if theres nothing worthwhile to publish - you make mountains out of molehills)

[Edited on 26-8-2006 by len]

Magpie - 25-8-2006 at 21:05

len thanks for clarifying the purpose of the screen in the Castner cell (preventing liquid sodium migration). It's been a while since I read those old posts closely.

Assuming the atmosphere is the main route for oxygen contamination of the molten sodium some kind of cover over the sodium surface would help - such as inert gas or at least a bell or tube. Inert gas is getting fancy but a cheap, cleverly designed bell or tube should be achieveable. Incidentally, I measured the boiling point of my mineral oil at 330C. Pure NaOH melts at 318.4C. I wonder if paraffin or even petrolatum would work as a cover.

Temperature control is another key issue as you have noted. One could rig up a heating jacket with a nickel sheathed thermocouple and a PID controller. Again I hope we can find something simpler and cheaper.

Caution: Don't make the mistake I did and place your glass thermometer in the molten NaOH - even briefly. My thermometer's bulb now has a rough surface and I'm afraid to use it for critical applications. It's been retired. :(

[Edited on 28-8-2006 by Magpie]

len - 25-8-2006 at 21:37

Magpie, your idea of a liquid sheath as opposed to a gaseous one is interesting. I know paraffin wont work, I recently tried it as an oil bath medium for my nitrobenzene. It boils too low - starts fuming heavily at 225C and boling at 250C. Some mineral oils will boil much higher, but I believe boiling oil with lye is a way to make soap. Maybe some oils dont react that way, dont know, maybe someone does? My handbook also lists them all as highly carcinogenic, if so their fumes at 300C cant be good. I am also currently working on the same problem as you.

not_important - 25-8-2006 at 22:38

Mineral oils are out and out hydrocarbons, they won't saponify as vegetable oils will. There may be some slow reaction with NaOH, I suspect it will be more likely if oxygen is freely available.

This could be a way to help control the melt temperature. If the container was made deeper, vapours from the boiling mineral oil could be cooled with a condenser coil. Not only would oxygen be excluded, but by tweaking the oil you might be able to use its boiling to limit the rise in temperature.

Even with good reflux of oil vapours, you would not want to use flame as the heat source, unless you did extra work to keep flammable vapours away from the flame.

Note that mineral oils are a mix of compounds, and thus have a boiling range rather than a boiling point. There are different boiling ranges available, you might have to adjust that of whatever oil you can find by distilling off the lower boiling components.

12AX7 - 26-8-2006 at 11:03

Are silicones inert around Na, NaOH?


Fleaker - 26-8-2006 at 21:30

Under what conditions Tim? The temperature will dictate much of the reactivity. At 110 celcius, sodium won't touch PTFE, double that and it's serious degradation. I think silicone would handle it well, but sodium has amazing heat dissipation qualities: it easily warms up and it easily cools down.

I'm still not entirely sure why so many people are trying the NaOH method? Probably because it is cheaper and more forgiving I suppose (i.e. lower temperature). Still, NaCl/CaCl2 does work and it will work well--just expect some problems early on in the process. As many of you know, for manufacturing sodium to be worth your time, you must put a lot of amperage into the system, otherwise it is simply too slow.

Oh @ len, what I was referring to was what worked in the coffee-can-Castner. I'm sure I've posted about that before in either this thread or some other. The electrode still needs to be covered with an insulating mix, and it's hell to get it into the side of the cell without having the salt leak which is rather fluid and pretty volatile. It's long since destroyed/lost/thrown away, but if you'd like, I can describe the details/dimensions as best as possible?

Magpie - 26-8-2006 at 22:46

Duratherm S, a silcone oil, looks like it might have possibilities. See Duratherm S

[Edited on 27-8-2006 by Magpie]

len - 29-8-2006 at 04:53

Fleaker, I serached this forum for you and Castner (or sodium) and couldnt find anything. You may be thinking of another forum?

Fleaker - 30-8-2006 at 16:42

I could have sworn I posted something about it. Sorry for wasting your time then! I'm now set on sodium (through Ordenblitz I found a reasonable supplier and got 3kg of it) so have no need for a Castner, however, I can still draw out the design which worked, albeit not as efficiently as I would have liked. Just need to get hold of a scanner :\

Also, does anyone know of a cheap supplier for CsCl, like say 1000g for 150USD or so. I've been wanting to make cesium for quite a while now.

The_Davster - 30-8-2006 at 17:25

Cesium salts, cheap:o. Let me know if you find any, I have never seen any for less than a dollar or two per gram. Companies in china may have some for cheap, assuming you can find a company that carries it.

I wish you luck in isolating metallic cesium, the stuff instantly catches fire in air. Firsthand experiance there unfortunatly.

not_important - 30-8-2006 at 23:41

Originally posted by Fleaker

Also, does anyone know of a cheap supplier for CsCl, like say 1000g for 150USD or so. I've been wanting to make cesium for quite a while now.

Cheapest I've see in several years, but I haven't been looking very hard; about twice what you want to pay.

neutrino - 31-8-2006 at 07:13

The order page is on ordinary http, not https. In other words, the web site sends your credit card information over an <i>unencrypted</i> connection when you order.

I wouldn’t order online from them.

Fleaker - 31-8-2006 at 12:45

They seem to be the cheapest place I can find too (exluding 25kg sacks from China) I plan on doing the reduction of CsCl with calcium metal under high vacuum. It obviously cannot be performed with sodium metal due to the Na distilling over with the Cs.

@ Rogue Chemist: Don't worry, I'm quite familiar with its spontaneous combustion in air. I will do it in vacuo then fill with argon as I fill ampoules with it, just a scaled up procedure of that in Brauer's. Anyone who would like to see the video of Cs burning automatically in air can see it here:

Hopefully the link works.

12AX7 - 31-8-2006 at 20:49

Ooh, toasty. And that's like what, two hundred bucks of metal, up in smoke all its own!!


tumadre - 13-9-2006 at 22:35

we almost need a separate thred on NAOH vs. NACL

I am continuing work on my hot NACl project

my current plan is to make a four cell system, the steel cathode from the bottem through the ceramic,
riser pipe above it, and carbon rods(12 per cell) surrounding the riser pipe.

I plan on 80 amps @36 volts if the salt can take the current / salt resistance, if not, then just short out the fourth cell.
yes thats 12 moles per hour, ~260 grams per hour.

limitation: the ceramic material itself, and sealing the ceramic/steel at the base of the cell.
whether I use three or four cells is determined by the strength and practicallity of the cell structure, if one centimeter is thick enough then use four cells,
if 2 cm is needed then three would be better, the space I have avalible inside my oven is 9 inches wide, 12 deep by 9 high

another difficulty is if the calcium plugs up the tube off the riser pipe

the Cl2 is 144 liters/ hour, absorbed by slacked lime (CaO) to become bleaching powder (CaOCL).
the slacked lime is in a rotating 8 inch PVC pipe 6 feet long, I would be pulling the CL2 through the Cao, from the oven, and a completely sealed setup at that

the oven is fire brick, the kind used 70 years ago in the local creosote plant long gone ( BTW Not insulating brick)

idea: put the heating element in the liquid salt.

i plan on running this for 40 hours straight, to get about 25 pounds of sodium, ~120 pounds of bleaching power

next question: how much is 25 pounds of sodium worth?

i hope to get the ceramic fired by the 20th , and the cell running by the 22th of september
the power source is 30.5 volt transformers rectified, and a combo of inductors or caps to get the most power.

12AX7 - 14-9-2006 at 09:51

A couple hundred dollars. I heard of 6lbs logs going for about $200 on eBay, a good deal at that.


Zinc - 4-10-2006 at 05:57

A little OT but does K dissolve in KOH if it is heated to a certain temperature as Na dissolves in NaOH?

Improvised production of Na metal

Chainik - 7-12-2006 at 22:24


The best and safest way to produce sodium metal is to make what I call a SodaBomb. It looks like a pipe bomb but is used to produce Na instead.

A steel pipe that is threaded on both ends, ~2in. dia x ~12in. long with matching end caps is needed. It is essential to remove whatever coating there is (most likely zinc) so that only the iron contacts the molten Na.

The top cap needs to have steel or other suitable tubing welded to a hole in the cap, it needs to be ~12in. long so the heat wont affect the vacuum tubing. Next a hole is drilled for the graphite anode which needs to be long enough to reach well into the molten NaCl sol. Then attach a wire to the steel tube as the Steel tubing itself will be the cathode.

Now the Bottom cap is attached and the apparatus is filled to within 1" from the top with NaCl and a suitable heat source applied. When the NaCl is thouroghly liquid carefully attach the top cap and apply vacuum to the reaction vessel (an aspirator is best). then apply the correct DC current (you need to work this out for yourself).

When the reaction is done let the vessel cool down with vacuum stil applied and when cool submerge the entire vessel in a nonreactive fluid, (i.e. mineral oil, kerosene etc.) Then you can remove the end caps and cut the Na out or if possible heat the oil sol. slightly to melt out the Na metal.

Sound fairly plausable to me, though i am not an expert. I can see some logistical difficulties in making a tight, yet electrically insulated seal between the steel cap and the graphite rod, which could sustain 800C temperature for a number of hours. Some kind of fire proof cement would probably work. Heating coul be supplied by a propane furnace with a few burners to distribute heat around the lenght of the pipe, which would be secured in place vertically. A recirculating water aspirator station should provide the sucktion nesessary to remove the chlorine gas. Water in the station should be periodically drained and replenished, as i would imagine it would absorb the chlorine. Current could be provided by a 12V car battery connected to a charger perhaps?

I would love to try this, but have to build a furnace first...


12AX7 - 8-12-2006 at 07:36

How does the chlorine bubble out without re-burning with the sodium? Sodium does float.

The combination of molten salt and vacuum seems rather silly to me. Salt has a rather high vapor pressure.


BromicAcid - 8-12-2006 at 18:12

Sounds like a pain in the butt. And drawing off the chlorine under vacuum using an aspirator? That would put a slight vapor pressure of water in the reaction vessel which would be quickly reacted away drawing in more water vapor (albeit tiny amounts). All of this sounds similar to what I tried with potassium hydroxide/magnesium reduction, if scaled up, the whole apparatus could have been put into mineral oil and product scraped out.

Chainik - 8-12-2006 at 20:38

I suppose the apparatus could be designed with a circular (pipe) shield surrounding the graphite electrode, and the chlorine outlet positioned in the area between this electrode and the shield. But then isn't the whole thing is really starting to become a downs cell with suction chlorine removal?

The only advantage of doing it this way, IMO, is the vessel design (metal pipe and screw caps - readilly available) and the product removal (cooling the pipe down and opening it under mineral oil...

tumadre - 21-1-2007 at 15:12

The easiest way is the way I tried to finish before moving on with my life.

A steel/copper funnel with pipe drawn off the top suspended over the steel electrode (-) inserted through the bottom of a ceramic vessel. and graphite electrodes (+) suspended in the liquid salt by the copper screw connection to the top of the rod.

To start the system from dry salt, expenditary rods are inserted so that they touch the steel rod, and using a 4 or more car batteries to provide the initial current to heat and melt the salt.
BTW this will only work if the ceramic vessel is at least a few liters in volume as you will crack the ceramic.

any clay/ceramic can be used if only one cell is used and if the ceramic is not exposed to any electrical potential diference.

yes you can position an iron pipe over the graphite anode and suck off the CL2, use a copper rod and ceramic perforated disks to hold the graphite and copper rod centered in the pipe so it don't touch the steel pipe, and send the electricity down the copper rod to the graphite, extend the pipe ~30 cm and afix the copper rod to a silicone seal through the steel pipe

more later-

indigofuzzy - 14-2-2007 at 04:56

Out of curiosity, would glass stand up to having molten salt and/or molten sodium metal in it? If so, than with some time and glass blowing skills a cell could be made with the electrodes passing through the glass, which would keep the electrically isolated.

BromicAcid - 14-2-2007 at 04:58

Of course it wouldn't stand up to sodium hydroxide on a chemical level for very long but glass does stand up to sodium. The problem though being the heat required to keep everything liquid with a different electrolyte and the glass becoming soft under that heat. At least that's the main problem I see with the idea.

ziqquratu - 28-2-2007 at 17:07

Ahoy sodium enthusiasts!

Just stumbled across this article, which I find DAMN interesting. The is (as the title suggests) designed for lab scale work, and gives nearly 100% cathodic efficiency (ie. almost 100% yield of sodium based on current used). An added feature which I find nice is that the anodic efficiency is much lower, 20-30%, meaning less chlorine is produced than would be expected.

The feedstock is a 30:70 mixture of sodium chloride and aluminum chloride. The temperature is around 250 *C, and the currents they use are anything between 1 and 10 amps. So, using 1A, approximately 1mol of sodium can be produced per 24h!

The reaction works because it uses beta alumina as an ion-selective separator. And this, as far as I can see, is the main sticking point - does anyone know anything about how one might go about making or (easily) obtaining a cylinder of the stuff? The other issue is that they start off with some sodium in order to prepare the alumina surface, although there seem to be enough claims in this thread that SOME sodium can be produced easily enough, so this may not be a real problem.

Anyway, here's the article:

Design, construction and operation of a laboratory scale electrolytic cell for sodium production using a β-alumina based low-temperature process

Journal of applied electrochemistry 2002 vol. 32, no12, pp. 1383-1390

Attachment: Electro_Na_Alumina.pdf (338kB)
This file has been downloaded 1313 times

not_important - 1-3-2007 at 13:43

This place makes an "inside-out" version of that, for injecting sodium into molten metal, using NaOH as the feed material.

They sell preforms of beta''-alumina with several dopings, Na and K included.

tupence_hapeny - 26-3-2007 at 21:08

This is the 'new' method of operating a modified 'castner' type process. It uses methanesulfonyl chloride as the electrolyte, and operates above the melting point of elemental sodium (~100C) and according to the patent operates between 100-200C, to produce elemental sodium and chlorine from a divided cell (using glass matting or a glass frit to divide a H-Cell). As 100-200C is rather attainable (as is methanesulfonyl chloride see the patent: Methane(g) (LPG) + Chlorine(g) + Sulphur Dioxide(g) is allowed to react in a glass tube @~15-100C under UV irradiation [sunlight/low pressure mercury vapour lamp/high pressure mercury vapour lamp) this process would appear to be the most viable for the production of alkali metals at home. Anyway, for the references, see this page (my post is at the bottom):

Can someone please advise me if the required gaseous mixture could be generated by mixing sodium hypochlorite and sodium metabisulphite (in the requisite ratios) and adding Hydrochloric acid to generate both the Sulphur dioxide(g) and Chlorine(g)? IF SO, may I suggest passing LPG (Methane = ~95-97%) through the reaction vessel to collect the gasses thus generated and passing the desired ratios directly into the reaction tube? Alternatively, add HCl to the bottom of the reaction tube (or attached reactor) and then add the required volume of LPG? With a pressure equalised addition funnel, it should be possible to add additional reagent precursors as required.

NB if the condensate from the reaction can be collected in an inner or outer container, in such a manner as it does not come in contact with the spent precursor/HCl mixture, it should be possible to remove it cleanly without contamination.

not_important - 26-3-2007 at 22:19

It's not easy getting very strong solutions of sodium hypochlorite, and there seems to be a fair amount of chlorine lost to solution; I assume that the same problem would exist with SO2. The HCl + MnO2 route seems to be better for making chlorine, perhaps a mix of MnO2 and metabisulfite would be better.

I'd keep the gas generation entirely separate from the actual reaction area, plumb the gases over into a reaction vessel. I know you want to use the HCl produced, a closed loop system with active pumping may be needed to do that, but I think a properly design might work OK.

I'm not sure you could get good enough control over the gas ratios that way, but it's simple enoug it might be work a try. Note that sunlight will work in place of a lamp, you could run the reaction in one or more condensers (in series). You'll need to vent the gases to the external world to prevent pressure buildup from non-reactive gases.

12AX7 - 27-3-2007 at 11:55

Originally posted by tupence_hapeny
Methane(g) (LPG)

LPG is propane. Methane is natural gas, rarely stored in cylinders (methane is supercritical at room temperature so is a refrigerated liquid) and widely available through domestic gas plumbing.


not_important - 27-3-2007 at 14:20

Ah, you are correct 12AX7, I missed that - read it as LNG which is not sold in stores this side of the asteroid belt.

tupence_hapeny - 27-3-2007 at 21:37

I missed it too, 12AX7... Perhaps I got too excited looking at the low temp sodium:( Anyway, provided you are on a gas grid, LNG is actually more available (in most areas) than LPG, simply take it from the gas line. I welcome criticism of the idea, such criticism may help design a workable mechanism by which to make low temp (fairly low tech - face it, the gasses can be piped in through stainless steel lines and fittings into a SS collar (and probably top and base) at the ends of a glass tube. Yeah, sunlight is supposed to work.

It also appears that some of the co-electrolyte can be reused after replenishing the solution with NaCl. I noticed that my links are NOW down, so here they are again:

Methanesulfonyl Chloride

Low temperature Alkali Metal Electrolysis using Methanesulfonyl chloride:

I personally find the fact that the links are down somewhat surprising and somewhat unusual, however, I have noted that their (Free Patents Online) website gives the notice that it is closed for some reason

[Edited on 28-3-2007 by tupence_hapeny]

Edited again, I think it may be easier to produce a known volume of gas (SO2 form Metabisulfite and Cl2 from salt) through electrolysis, at least then you could cut down on the cost of the entire process (extra reagents) and establish some control over the rates of generation. Face it, if you can't generate Cl2(g) from salt solution, you probably shouldn't try to make Na from the same:P

[Edited on 28-3-2007 by tupence_hapeny]

JohnWW - 31-3-2007 at 16:40

Is there a link anywhere to PDF versions of those two patents, which would have the diagrams and illustrations, which are omitted from the HTML versions on those two sites?

not_important - 1-4-2007 at 03:24

Originally posted by JohnWW
Is there a link anywhere to PDF versions of those two patents, which would have the diagrams and illustrations, which are omitted from the HTML versions on those two sites?


Do a search, prefix the patent number with the country code. You get a list of matches, clicking on one results in the entry for that patent being shown. Then select the 'Original Document' tab, and after that the 'save full document' link above the patent display.

indigofuzzy - 2-4-2007 at 01:06

Ok, something inside me says this sounds *way* too easy.

Hypothetically, one just needs natural gas (for the methane), a reaction that makes chlorine, sulfur burned in oxygen, and to mix the gases in proper amounts in a glass tube, and place this in sunlight (or under a blacklight tube), yielding the necessary ionic fluid? Then dissolve the required salts and electrolyze to your heart's content?

Is it really that simple, or did I miss the big huge GOTCHA somewhere?

not_important - 2-4-2007 at 02:52

Burning sulfur to get SO2 also gives some SO3, plus there's all the other stuff in air, some purification will be needed. You need to have some control over the proportions of the gases in the mixture, or do some design work to come up with a plant that can tolerate a lot of variation. You'll need to distill the product, and it is an acid chloride.

A blacklight tube isn't very intense, and if you read the patent it's more in the blue range that you want anyway. High intensity mercury or metal halide lamp is more like it.

You need to keep the electrolysis rig rather dry.

More than likely it can't compete on an economic basis with the established processes. And, yes, there may be problems with one or both parts of the full process, many patents include a lot of handwaving and don't-look-behind-the-curtains; they may not be practical when filed and are either done in the hopes that they can be improved or to CYA in case someone else manages to come up with an improvement.

len1 - 10-5-2007 at 00:24

I carried out some prelimanry experiments on NaOH and NaCl/CaCl2 electrolysis. Turns out ther latter is a much better bet , unexpectedly, despite the 280C higher temperature. There reasons:

1) Experiment has shown that molten Na is not substantially more reactive at 580C than at 330C. The presence of Cl does not substantially change this.

2) Forming Na in molten NaCl is not an issue, it forms easily, it doesnt dissolve. The only problem is collecting it.

3) Molten NaCl does not attack glass/porcelain, making collection much easier

4) Molten NaCl does not spit, there are no nauseous hydroxide fumes. If you get some on you its only good for you.

The picture shows Na globules being collected in an inverted tube under an inert atmosphere. You can not do this with NaOH.

The only difficulties are:

1) Some bubbles of Na (about 0.5cm sphere, 0.1g each) miss the test tube and break like a shinning sun on the surface (picture). Placing the -ve electrode inside the tube, and or using gauze does not right this for obvious reasons.

2) The Na collects in the form of a multitude of such globules in the tube. They are covered by a stubborn coat of melt/carbon particles which are very difficult to dissoldge in getting them to coalesce under molten paraffin.

[Edited on 10-5-2007 by len1]

NaCl-Na.bmp - 461kB

Magpie - 10-5-2007 at 13:44

Very nice len1. Would you mind providing more details of your cell or are you waiting until you get a production model developed?

I would like to know cell design, materials of construction, method of heating, keeping the inert atmosphere, removal of chlorine, etc.

You say:

Placing the -ve electrode inside the tube, and or using gauze does not right this for obvious reasons.

I must be missing the obvious as I don't understand this statement.

len1 - 10-5-2007 at 21:09

Originally posted by Magpie
Very nice len1. Would you mind providing more details of your cell or are you waiting until you get a production model developed?

I would like to know cell design, materials of construction, method of heating, keeping the inert atmosphere, removal of chlorine, etc.

You say:

Placing the -ve electrode inside the tube, and or using gauze does not right this for obvious reasons.

I must be missing the obvious as I don't understand this statement.

Thanks Magpie for the nice words. Your guess was spot on, my real aim is to see if I can make a small jig for making sodium continuously. Im not really satisfied with these results so per se, except as experimentation to help fix the final design. My aim in posting this is to see if anyone can contribute useful info as per how to coalesce dirty Na globules, and interaction of molten Na with glass.

I used an ordinary high-temp furnace, inverted for crucible use. The front shield was a ceremic tile. The crucible itself is porcelain. The -ve electrode was a stainless steel wire inserted through a 6mm OD glass tube. The reason for the latter, is to avoid the Na forming/floating to the surface following the wire - this would otherwise happen. The chlorine removal mechanism was my lungs! Seriously its not a huge problem for a small set up. If you make 2gms of Na, you release 0.05 moles Cl2 gas, or 1.5l worth. For comparison, 1 liter of liquid chlorine, 5% NaOCl, used to wash your floor releases up to 13gms of chlorine! The tube had Ar blown into it periodcally, being cold and heavier than air, it displaced the O2 in the tube.

The liquid Na wets almost anything including glass, which it attacks also, hence the black coating you can see. The coating - whose composition I dont know, does not seem to conduct electricity, and once formed protects the glass, which was just superficially corroded. I would be interested to know if you or anyone else has any references to this effect of liquid Na on glass. Using a metal tube to collect the Na (which you could also do in NaOH) would overcome this, but it cant be used around the negative electrode.

Placing the -ve elctrode in the center of the tube has the Na wetting the tube, crawling around it, and being electrically connected to the -ve, starts formation of globules outside the tube. Len

[Edited on 11-5-2007 by len1]

Magpie - 11-5-2007 at 10:45

How much of the -ve electrode is actually exposed to the molten NaCl? That is, not encased in glass. I presume that is the only part of the electrode that is generating Na.

Could this active part of the -ve electrode be the terminus of a J-shape. Then let the Na crawl up the wire and float to the larger glass tube for capture and removal.

Then possibly you could make the capture tube of metal as it would not be electrically connected to the -ve electrode via Na bridging.

But these are geometry issues about which it is very hard to give any meaningful help and about which you are surely in the best position to solve.

209 - 17-5-2007 at 09:54

I am intrested in making a little sodium metal myself. Being that I don't have a power sourse of 5V @ 50 amps (maybe the arc welder?) I dont really have a power sourse that is good enough to create the juice required. I do have a 12 volt transformer out of the wall that may work. Through personal experience I have found that the totse forums really arn't that accurate:D:mad:. So I don't know if the 12 volt transformer will work. Will it?

b_d_Dom - 17-5-2007 at 13:03

Originally posted by 209
I am intrested in making a little sodium metal myself. Being that I don't have a power sourse of 5V @ 50 amps (maybe the arc welder?) I dont really have a power sourse that is good enough to create the juice required. I do have a 12 volt transformer out of the wall that may work. Through personal experience I have found that the totse forums really arn't that accurate:D:mad:. So I don't know if the 12 volt transformer will work. Will it?

If it is just a basic transformer then no. Unless it has a built in rectifier it is outputing AC which is no good for electrolysis.

You can build a rectifier for it if you have some diodes lying around, or just find yourself a DC power source.

If you have any questions on this subject just send me a U2U and I should be able to help.

Jdurg - 14-6-2007 at 05:46

It seems a bit odd that the Na is "attacking" the glass because there are many chemical suppliers who sell sodium metal in sealed glass ampoules. This metal is obviously liquid when it is poured in there and the glass is perfectly fine. The only alkali metal I know of that attacks glass when molten is lithium. Sodium, Potassium, Rubidium and Cesium all seem to be fine in the presence of glass. I wonder if there are impurities in the glass, however, that might be reacting with sodium as there are more varities of glass in this world than there are people.

tumadre - 14-6-2007 at 09:24

AFIK The issue is the sodium dissolves in the glass at that temperature.

anyone know if it it reacts at all?

I don't believe there is a reaction with the soda-lime glass, but whatever the case that is far too messy.

try a ceramic tube say 15 cm long bonded to a glass tube, the glass tube makes a 90 degree bend and goes into a negative pressure inert gas atmosphere, and 'suck' the sodium out of the cell.

Put a copper cooling tube down the glass tube, so it cools the sodium to about 300-200C, so it don't melt the glass.
The 3-5 mm internal diameter ceramic tube will drop the temp from 550-600C to the 300/200C.
Regulate the vacuum to whatever is needed to keep the salt/sodium level to about 5 cm up the ceramic tube.

The sodium is the negative electrode

[Edited on 14-6-2007 by tumadre]

len1 - 18-6-2007 at 21:04

Originally posted by Jdurg
It seems a bit odd that the Na is "attacking" the glass because there are many chemical suppliers who sell sodium metal in sealed glass ampoules. This metal is obviously liquid when it is poured in there and the glass is perfectly fine. The only alkali metal I know of that attacks glass when molten is lithium. Sodium, Potassium, Rubidium and Cesium all seem to be fine in the presence of glass. I wonder if there are impurities in the glass, however, that might be reacting with sodium as there are more varities of glass in this world than there are people.

I also have some shinning Na in a clear glass ampule. I sucked it up from some of the globules formed when NaOH
was electrolised, i.e. at 300 degrees.

The sodium we have here is not at 300 degrees though, its at 600. There is clearly a black coating, you can see it in the
photo, on the pyrex tube. The coating only exists at those points were the Na touched it, not at others.

not_important - 18-6-2007 at 21:24

When the glass is hot enough some sodium can diffuse into it. Once there it can exist as a metal, or reduce some of the ions in the glass. Depending on what type of glass is being used I ould not be surprised to see a little reduction of Si, to the element or Si(II), or boron; iron and manganese in ordinary soft glass would also undergo reduction down the the metals.

Note that glass starts becoming conductive when hot enough, although not really conductive. This has been used to prepare small amounts of very pure sodium for absorption/emission studies, a electric light bulb was used as the electrode, the sodium formed on the inner surface and fresh sodium ions diffused in from the moltan batch it was immersed in. The glass gets quite dark, and the inner surface becomes coated with a sodium mirror. The rate of production is too slow for making useful amounts of sodium, but the related method using beta alumina mentioned earlier is fast enough.

len1 - 24-6-2007 at 16:29

I can also add that the molten sodium in the NaCl/CaCl2 mix has very high attraction towards any glass or ceramic surface. It wets it as a thin layer, and then reacts. This is evidenced by a dark black patch at the bottom of the ceramic vessel in the picture at the point where the -ve electrode touched it. Even though quite a bit of density differential exits between the melt and the sodium (2:1) forcing the Na to float, it was energetically more favourable for it to wet the bottom completely, and react with it, as it formed, then float to the top. It clearly was also able to crawl along the already reacted patch.

garage chemist - 2-8-2007 at 12:43

There is a documentation on how to build a DIY castner cell on versuchschemie:,9576,0,-Na+aus+NaOH+elekt...
A fine stainless steel wire mesh is used as a diaphragm with good success. This really seems to be necessary, as industry uses it as well.
The cell is heated electrically- the wire wound on the glass fiber cloth is the heating which is wired in series to the cell after the NaOH has been molten. It stays at the operating temperature when 7A flow through cell and heating. Every hour the sodium is scooped off with the special spoon. The large white blob in the second to last picture is the sodium.

[Edited on 2-8-2007 by garage chemist]

len1 - 9-9-2007 at 16:16

The previous post describing a home-made Castner cell on a German forum was the perfect precursor to what I wanted to achieve, which is a good quality permanent apparatus to produce Na on-tap. I believe I now have this.

Although the pictures in the previous post were an inspiration, the text wasnt, Ive never learnt any german, and after reading the Babelfish translation I was so confused I simply tried reading the original text using the similarity of some German words to the english to make sense.

In any case the key problems with the original experiment (which is perfectly fine for a pilot-run) that I identified were:

1) fairly flimsy nature of apparatus - the Na collector just sits in the melt - if you heat it too high, it will move.

2) The container walls are used as an anode, which is not how Castner operates, leading to wall erosion, as well as making it impossible to use nickel in the anode, by far the most sturdy material when it comes to dissolution under potential in molten NaOH

3) The bottom of the vessel was ramed clay, exposed directly to NaOH at near melting temperature - although the original experiment tried to maintain a temperature differential so the bottom NaOH is solid, in practice its hard to do, and in any case at elevated temperatures the clay will rapidly dissolve (my experinments show molten NaOH attacks clay at the order of 1mm per 10 mins). The dissolution of the clay leads to substantial contamination of the bath, shortens its life, as well as leading to perforation at the bottom with time.

4) The low amperage of the cell resulting from (3) means you have to wait all day to get even a few grams of Na.

The several areas of the original which were used unchanged was

1) The cathode was made of wound Cu tube of short cross section rather than solid. That is a great idea comapred to a solid Cu cathode- meaning much reduced heat loss through Cu notorious heat conductivity, a much more unifrom temperature of the bath, which is essential for any reasonable yield in this case.

2) The Na collector was a slotted pipe into which a thin wire gauze (sold in Woolworths for toasters) was inserted, and held up extremely well.

3) The cell was heated from the outside with insulation provided by fibreglass.

The present construction (of which more will follow) has the following features:

1) A cylinder anode of 1mm Ni sheet of 6cm diameter by 5 cm long attached by adjustable SS bolts to a top circular plate separated from the collector and casing by ceramic insulators scavanged from a radiation heater

2) A NiCr heating element scavanged from the same heater wound around fibreglass matting and providing 300W of power. The fibregalss was held in a wound state around the outside of the cell ny two stainless steel hose clips. Each of the hoseclips also acted as a terminal for the heating element. This provides a sturdy constrcution.

3) An outer radiative-heat container of SS (an indian cofee jar from woolworths) of 15cm diameter and 15 cm long encased the cell with wound NiCr.

4) Holes drilled in the radiative container with terminal posts attached to bolts drilled in the hose clamps allowed the passage of heating current at 240V. Note its independent of electrolysis current, and is only needed at startup. A hole also passed a thermocouple probe held by a screw clamp to the cell body. It was found that 365C was the ideal temperature for bath melting - while 320C was ideal for electrolysis.

5) The cell measured 8cm diameter x 15.5cm and held 1 kg of caustic, which although full when cold, filled it half-way when molten - this reduced heat variation in the bath due to collector protrusion.

6) A lided SS pot 25cm diameter 22cm long filled with glass wool formed the outer insulation. The pot was bolted to a long stand admitting a receptacle which was passed cathode.

7) The cathode was 5cm long 15mm diameter wound copper tubing. Fixed at the bottom (cool end) to a threaded nut and sealed with exhaust cement.

8) Temperatures measured were 100C for bottom of SS pot, 66C for top, 44C for lower end of cathode stem, 33C for cathode, 150C for anode terminals, 270C for top of Na collector, 320C for bath inside collector.

9) The Na was collected by a perforated ladel. This let the bath pass thru as stated - has liquid Na got a high surface tension! A lot was wasted still clinging to the ladel - and you had to shake it really hard to get even half of it to drop into the parafin at 140C

10) The bath current was 47 - 50A. Experiment time was about an hour to produce the sodium shown.

[Edited on 10-9-2007 by len1]

Na1.JPG - 75kB

len1 - 9-9-2007 at 16:18

Here are some more pics. I havent worked out how to post more than 1 at a time

[Edited on 10-9-2007 by len1]

Na2.JPG - 79kB

len1 - 9-9-2007 at 16:20

The power supply

Na3.JPG - 73kB

len1 - 9-9-2007 at 16:21

The sodium

Na5.JPG - 67kB

len1 - 10-9-2007 at 19:24

The ladel

Na6.jpg - 59kB

len1 - 10-9-2007 at 19:25

The Cathode

Na7.JPG - 62kB

12AX7 - 10-9-2007 at 19:46

Very nice! :thumbsup:

len1 - 11-9-2007 at 17:45

Thanks Tim

Following are more pictures of the Na cell internals. This is
the annular Ni anode attached to the Na collector

Na8.JPG - 21kB

len1 - 11-9-2007 at 17:47

This shows the heating coils wound around the outside of the cell, they are needed only to initially melt the NaOH. The 50A current @4.6V is then just perfect for maintaining cell temperature

[Edited on 12-9-2007 by len1]

Na9.JPG - 47kB

12AX7 - 11-9-2007 at 17:48


len1 - 11-9-2007 at 17:49

This shows the radiation shield in place, with protrusions for the thermocouple and the heating coil binding posts.

Na10.JPG - 43kB

len1 - 11-9-2007 at 17:54

No, Ni anode is right. The cathode is of copper. Thats the standard arrangement for a Castner. The anode is were the greatest potential for corrosion exists and so needs to be made of a metal least sensitive to corrosion in NaOH.

Unless you are thinking of a voltaic cell where the carthode and anode are the other way around, but in an electrolysis cell I believe thre labeling is right. The terminal to which the +ve voltage is applied is the anode

Magpie - 11-9-2007 at 18:48

Very nice len. Your apparatus looks well engineered providing ruggedness and reliability in a compact assembly. Yet it probably still minimizes cost and maximizes use of OTC parts. You did do some welding though it looks like?

I'm glad you posted some more pictures. I was having a heck of a time trying to figure out how your cell was assembled! Please post some more.

Where does the "toaster gauze" fit in? We haven't seen that yet have we? BTW what is "toaster gauze?"

Did you assemble the power supply yourself? If so, please provide some details.

(BTW, your reference to Woolworth's is interesting. When I was young we had "five and dime" Woolworth stores. But they no longer exist in the US. I see that Woolworth stores still survive in other parts of the world, however. And in Australia and New Zealand the stores are unrelated to the orginal US stores by that name. )

[Edited on by Magpie]

len1 - 11-9-2007 at 19:21

Originally posted by Magpie
Very nice len. Your apparatus looks well engineered providing ruggedness and reliability in a compact assembly. Yet it probably still minimizes cost and maximizes use of OTC parts. You did do some welding though it looks like?

I'm glad you posted some more pictures. I was having a heck of a time trying to figure out how your cell was assembled! Please post some more.

Where does the "toaster gauze" fit in? We haven't seen that yet have we? BTW what is "toaster gauze?"

Did you assemble the power supply yourself? If so, please provide some details.

(BTW, your reference to Woolworth's is interesting. When I was young we had "five and dime" Woolworth stores. But they no longer exist in the US. I see that Woolworth stores still survive in other parts of the world, however. And in Australia and New Zealand the stores are unrelated to the orginal US stores by that name. )

[Edited on by Magpie]

Hi Magpie, nice to hear from you again. Thanks for the words of encouragement. I had a project for a DIY Castner cell in mind for a long time (I dont think anyone has published a design before) but I couldnt get to doing the pilot experiments to see what worked. There were so many options. Then I saw the link garage chemist posted and thought this is the pilot run I was looking for. I still had to solve a few problems (such as how to isolate the colelctor from the anode) but the pilot showed this approach works on a semi-scale in principle.

Im very interested to see others repeat this project and introduce their own alterations. After all we have 12 pages in this thread so lets have a great outcome.

I have lots more pics but wasnt sure if positing so many would be what people on this forum wanted. Im willing to post more + answer questions, if people want. I dont know how to post several pics in one post.

I did do some welding, but you get change from $100 for welding machines were I live. The bits that had to be welded were the cathode stem to cell base, cell base to cell walls, and cell top to Na collector. The later can easily be avoided though. You could avoid the former by using a cast iron pot rather than a pipe and using a threaded base stem and nut to attach to the bottom of the later. I just couldnt find the right sized pot.

Toaster gauze is stainless steel fine mesh sold here to go in some sort of toaster I believe. To be honest I didnt really care much for the toaster so dont know how it goes in there. You can also get SS gauze from camping stores here.

I didnt know 'woolies' as we call them here is from the US. Yes its alive and well in Australia/NZ, competing with Kmart (your Walmart I believe)


DerAlte - 11-9-2007 at 19:21

Superb work, Len 1. A thread stopper!

Der Alte.

12AX7 - 11-9-2007 at 19:45

Huh, then I'm confused about where the sodium comes from.

So the nickel anode bubbles off -- you said Castner, so this is NaOH? -- O2 and H2O, while the cathode is placed in the center of the shield and periodically harvested?


 Pages:  1    3    5