Sciencemadness Discussion Board

Separation of NaClO3 and NaClO4 following the thermal decomp method...

Varmint - 12-5-2014 at 05:24

In the us government document AD003174 "Thermal Decomposition of Alkali Metal Chorates" the following statement is made at least two times:

a. Sodium perchlorate and sodium chlorate cannot
be separated from one another by any known simple method.

I have a hard time understanding this with the wide difference in solubility. Why wouldn't a fractional crystalization process work?

At any given temperature/volume, water will hold less NaClO3, so you dissolve at elevated temp, ride the temp curve down until you approach the point where the NaCLO4 is ready to start dropping, and you get virtually 100% NaCLO3.

Yes, it's a pain to ride this "curve" especially if you don't know the relative starting concentrations, but close observation and determination ought to afford reasonably pure compounds.

What an I missing? some Chlorate/perchlorate common ion effect? It doesn't make sense.

DAS

[Edited on 12-5-2014 by Varmint]

Bert - 12-5-2014 at 07:11

With enough care, the two chemicals can be separated.

Here's a patent on a combined, continuous thermal decomposition and batch differential crystalization separation process.

http://www.google.com/patents/US2733982

It is possible the degree of separation was not considered acceptable for mil-spec.

jock88 - 12-5-2014 at 07:52


Two beauties at the end of this thread to read.
http://www.sciencemadness.org/talk/viewthread.php?tid=29164#...
Then study the mutual solubility graph here:
http://www.oocities.org/capecanaveral/campus/5361/chlorate/n...

You are not really seperating Na chlorate from Na Perchlorate in your application but rather Sodium Perchlorate from Sodium Chloride. All Chlorate should have been destroyed from the heating.

Sodium Chlorate Production

Zyklon-A - 12-5-2014 at 08:42

I don't intend on derailing this topic, but I don't know of a better place to ask this question.
How does one make sodium chlorate from electrolysis of sodium chloride?
With KClO3 production, it's very easy, as the chlorate is nearly insoluble. But in the case of sodium chlorate it is much more soluble.
One idea I had was to put 290 grams of NaCl (5 mols) in 317 mL of water. Only 120 grams will dissolve at ~80°C. when electrolysis is complete this will result in 530 grams of NaClO3 (5 mols). 530 grams of NaClO3 will dissolve perfectly in 317 mL of water. So the simple indicator will be, that as soon as the NaCl at the bottom completely goes into solution you will have only 120 grams left and the rest is chlorate. There are several problems though: Namely, the water isn't just a solvent, it enters the reaction, thus it's used up. The empirical reaction is close to as follows: NaCl + 3 H2O → NaClO3 + 3H2. So I could just measure the hydrogen given off to calculate how far the reaction has gone right? Actually No. This reaction is not nearly 100% theoretical, it's more like ~30-45%. So much more hydrogen will be given off (per amount of chlorate produced) then the equation suggests. Since I have now way of knowing when all the chloride is converted to chlorate, and since I have no way of knowing how much water is present it seems like it will be nearly impossible to know when the reaction has gone to completion.
Sure, I could just add KCl to the solution once most of the NaCl has gone into solution, and precipitate KClO3:
NaClO3(aq) + KCl(aq) → KClO3↓ (s) + NaCl (aq). However I don't want KClO3, I want NaClO3. I know NaClO3 is more dangerous and should not be used in pyrotechnics. I already have lots of KClO3 (and KClO4) which I use for pyro. I want NaClO3, not for anything specific, as of yet, but more for the challenge of making it.
Does anyone know how this could be done?
I believe Mailinmypocket works at a sodium chlorate plant, but industrially they have methods to calculate the efficiency of their plant based on their PSU and trial and error I guess.
My PSU can give 5 volts and 20 amps, and I know how to figure out based on several laws (of which I've forgotten the names) how long it will take, but since I don't know the conversion efficiency, I can't really expect it to be very accurate.
Any idea's?

[Edited on 12-5-2014 by Zyklonb]

hissingnoise - 12-5-2014 at 09:30

Quote:
I know NaClO3 is more dangerous and should not be used in pyrotechnics. I already have lots of KClO3 (and KClO4) which I use for pyro. I want NaClO3, not for anything specific, as of yet, but more for the challenge of making it.
Does anyone know how this could be done?

Simply run the reaction until saturation is assured and cool the soln. to precipitate NaClO3?

[edit] This will, of course, require regular addition of fresh salt to the cell . . .



[Edited on 12-5-2014 by hissingnoise]

Varmint - 12-5-2014 at 10:01

Jock88: Absolutely not, the referenced document states plainly that there is always residual chlorate using the thermal decomp method, and it varies based on temp/time, but never gets any better than about 7% left unconverted.

That seems odd to, and the natural question is, if you were able to sequester the 2 (end up with pure NaClO3) will you again end up with 7% remaining from that decomp run, eventually (asmptotically) approaching zero chlorate residual?


macckone - 12-5-2014 at 10:05

http://www.oocities.org/capecanaveral/campus/5361/chlorate/d...

Usually people either produce chlorate by electrolysis or
they make perchlorate and destroy the chlorate by reduction.

Depending on what you need your chlorate for, the perchlorate
contamination from crystallization may not matter.

The reverse is generally not true. Ie. chlorate contamination in
perchlorate is usually a problem (ie. in pyrotechnics, oxygen candles, etc).
Hence the link on destroying chlorates in perchlorates.

Zyklon-A - 12-5-2014 at 10:44

Is it really safe to heat sodium chlorate? I wouldn't do this. Potassium chlorate is much more stable, and less likely to explode, and therefore safer to try. I still would't do that because perchorate is cheaper than chlorate.

Thanks hissingnoise, I'm not sure exactly how to ensure saturation, but I guess only add NaCl already dissolved in water. But this would also lower the concentration of chlorate, so I could do some tests, with a small quantity of the solution and see how much chloride it will dissolve. Then based on that factor, and an equivalent quantity to ensure saturation of chloride, without diluting the concentration of chlorate in the solution. and continue until a decent amount of chlorate has precipitated. Then add a very small amount of water, to make sure that no chloride precipitates upon cooling (NaCl solubility curve is extremely small with large changes of temps.)
Then allow to cool, and collect the chlorate - dissolve as much more chloride as possible and repeat until the amount of desired chlorate is reached.

jock88 - 12-5-2014 at 15:09



How does one make sodium chlorate from electrolysis of sodium chloride?

This link will explain alot
http://www.oocities.org/capecanaveral/Campus/5361/chlorate/c...

You can make Na perchlorate simply be adding saturated Sodium chlorate solution to a Sodium perchlorate cell and obtaining Solid sodium perchlorate from the bottom of the cell.
I believe commercial plants add acidulated brine (salt solution with a certain amount of HCl acid added) to the solution being worked on. This keeps pH correct and adds salt at the same time to keep concentration of Chloride/Chlorate at a point where it some chlorate will fall out of solution when the solution goes into the crystallizer.

Varmint - 13-5-2014 at 03:03

My preferred path is KCl --> KClO3, but the ease of conversion to K from Na via "double-decomp" makes the use of Na still worthy of consideration.

The core issue is anode life. While I'd love to use the platinum anode for electrochemical conversion of ClO3 --> ClO4, the prospect of killing my expensive anode with low ClO3 concentration (missing a liquor topoff or whatever) just makes that path unattractive. So I recalled having a copy of AD003174 on hand, and began studying in earnest.

The apparent difficulty of separation of the NaClO4/NaClO4 caught me off guard. Even Bert's tacit agreement that it should not be impossible (or perhaps even difficult) was encouraging, based on solubility "difficulty" wouldn't seem to be an operative word. So, I'm left to conclude there is something unusual about the NaClO3/NaClO4 structures that causes them to crystalize out more in tandem than their solubilities would indicate.

Given that, I thought it would be educational to try and run the process, even considering starting with pure ClO3/ClO4 and combining them in order to separated them and record it carefully. But like everything else, this could ignore what might be important issues like chloride or hypochlorite residuals, or other liquor components. Let's face it, the US government chemists didn't say it was impractically difficult for no good reason.

So, I'm at something of a crossroads, I'd sorely love to learn why it is difficult, but the paractical side of me says use the evidence provided by others and choose a more practical path.

Thank you all for your feedback.

DAS

Zyklon-A - 13-5-2014 at 05:51

Quote: Originally posted by Varmint  
My preferred path is KCl --> KClO3...

The core issue is anode life. While I'd love to use the platinum anode for electrochemical conversion of ClO3 --> ClO4, the prospect of killing my expensive anode with low ClO3 concentration (missing a liquor topoff or whatever) just makes that path unattractive.

I think it's a good idea, to use an MMO anode to make chlorate. Then use a Pt anode to oxidize it to perchlorate. It will be rather expensive, and thats an investment you should decide if you want to take. I do have a lot if MMO if you want to buy some, but I don't have any Pt.

Varmint - 13-5-2014 at 08:39

Zyklonb:

I might invest in MMO at some point, but right now I'm happy turning carbon rods into dust that's easily filtered. It pleases me to know my "only" impurity is essentially inert, and easily removable.

The issue with Pt is the accelerated wear when ClO3 gets low (same case during ClO3 production when Cl gets low), so the core point is I'm trying to avoid investing in Pt, ever.

Seems like the only "magic" anode is PbO2, and being a bit of a stubborn jackass, I might just set my sights on developing a robust lead based anode, and if I happen to come up with something workable, then everyone wins.

In fact, I rise to the challenge, I'm going to put some effort into PbO2 anodes, the promise of going from chloride to perchlorate is just too tempting not to give it a go.

DAS

[Edited on 13-5-2014 by Varmint]

Zyklon-A - 13-5-2014 at 08:47

OK, FYI my MMO isn't expensive. I sell it for $3.00 per sq inch. It's mostly the Pt that kills.

Zyklon-A - 13-5-2014 at 09:09

So I'm running a sodium chlorate cell now. 780 mL of saturated NaCl solution is being electrolyzed with MMO electrodes (both anode and cathode) at 5 volt and 20 amps. This will yield 462 grams of chlorate if isolation was possible without adding more NaCl. Of course for reasons discussed above, it isn't so I won't really be able to see what the yields are. I intend to make this like a continuous process, and just add the chloride as needed. Anyway, I'll post more once something starts to happen.

jock88 - 14-5-2014 at 11:18

Since Sodium Perchlorate has a number of hydrates this will need to be noted if yields are being weighed and calculated. Which hydrate forms depends on the solution concentration and temperature as your product crystallizes. The usually hydrate that forms (between zero and 50°C) is the mono-hydrate. This mono-hydrate must be heated to 60°C to obtain NaCl04:0.8H2O. It must then be heated above 150°C to obtain anhydrous Sodium Perchlorate. Melting takes place at 472°C and decomposition starts at 490°C and ends at 520°C. (Reactivity of Solids 3 (1987) 75-84)

I think the way to go for Pt anodes is to simply clamp some Pamp Suise one gram bars between two pieces of Titanium (this actually works as the current 'jumps' the gap ok) and going from Chlorate to Perchlorate. The anode should last for years.


Useful page back up .
Account of making Perchlorate with lead dioxide
http://oxidizing.typhoonguitars.com/chlorate/leaddiox/mmold....

[Edited on 14-5-2014 by jock88]

[Edited on 14-5-2014 by jock88]

hissingnoise - 15-5-2014 at 02:57

Quote:
I think the way to go for Pt anodes is to simply clamp some Pamp Suise one gram bars between two pieces of Titanium (this actually works as the current 'jumps' the gap ok) and going from Chlorate to Perchlorate. The anode should last for years.

Titanium, itself, is not inert and can't be used as anode in chlorate cells!

Varmint - 15-5-2014 at 03:39

hissingnoise:

Correct, Ti is a "valve" metal and will passivate when used as an anode.

This is why it is an ideal substrate for MMO, PbO2, and other anode candidate materials.

The effect is really interesting, as the MMO (for example) decays with use and flakes off, the TI passivates, and if in constant-current operation, all that current is carried by the remaining uncompromised MMO. In essence, the titanium portion of the electrode "doesn't exist" so far as current flow is concerned.

It becomes something of a runaway reaction, the higher current density causes more MMO to fail, and as more fails, more current is "directed" to that which still survives, increasing the failure rate.

Once the last bit of MMO has been erradicated by this ever increasing current density, the current flow stops somewhat abruptly as the titanium is fully passivated, and if I recall correctly, the standoff voltage of the passivation is on the order of 35 or 40 volts.

DAS

[Edited on 15-5-2014 by Varmint]

jock88 - 15-5-2014 at 06:22

Quote:
Quote: Originally posted by hissingnoise  
Quote:
I think the way to go for Pt anodes is to simply clamp some Pamp Suise one gram bars between two pieces of Titanium (this actually works as the current 'jumps' the gap ok) and going from Chlorate to Perchlorate. The anode should last for years.

Titanium, itself, is not inert and can't be used as anode in chlorate cells!



I omitted the word PLATINUM from my post (silly me).
Obtain a piece of platinum bullion. A one gram piece in the (bad from ebay) picture has a surface area of about 1.5 x 0.9 x 2 (sides) = 2.7 cm squared. This will give a roughly a 'one amp' anode.
It should last for a very long time imo.
The junction between the Pt and Ti (Pt clamped to Ti) will conduct current OK when the whole thing is kept under the solution surface as shown by pdfbq over on amateurpyro.org

$T2eC16VHJHEE9ny2rTbkBRDc5mt5Ug~~60_3.JPG - 38kB

Zyklon-A - 15-5-2014 at 09:53

Quote: Originally posted by Zyklonb  
So I'm running a sodium chlorate cell now. 780 mL of saturated NaCl solution is being electrolyzed with MMO electrodes (both anode and cathode) at 5 volt and 20 amps. This will yield 462 grams of chlorate if isolation was possible without adding more NaCl. Of course for reasons discussed above, it isn't so I won't really be able to see what the yields are. I intend to make this like a continuous process, and just add the chloride as needed. Anyway, I'll post more once something starts to happen.

Ok, well, it turned out that 20 amps isn't enough to keep 780 mL of the solution at a good enough temperature (65-80°C). It hovered at ~50-60°C, so then I put the cell in a water bath and kept it on "melt". Well after leaving the cell for about three hours on "melt", I found that it melted my crappy plastic cell. So now I have it in a smaller glass container, which holds about 500 mL of solution. It operates at closer to 80°C which is good.
Quote:

Titanium, itself, is not inert and can't be used as anode in chlorate cells!

Like Varmint and jock88 said, Ti passivates as an anode. Which is good, otherwise, if you got even one chip in the inert layer of MMO, Pb (IV) oxide, Pt or whatever you're using, it would deteriorate from the inside out.

jock88 - 15-5-2014 at 15:34


You could just rap the larger container is some glass wool (house insulation).
The increase from (approx) 50 to 80 degrees in not going to make that big of a difference in current efficiency.

hissingnoise - 16-5-2014 at 02:57

Quote:
Once the last bit of MMO has been erradicated by this ever increasing current density, the current flow stops somewhat abruptly as the titanium is fully passivated, and if I recall correctly, the standoff voltage of the passivation is on the order of 35 or 40 volts.

Thanks Varmint, the passivation layer is more robust than I'd assumed . . .
Quote:
I think the way to go for Pt anodes is to simply clamp some Pamp Suise one gram bars between two pieces of Titanium (this actually works as the current 'jumps' the gap ok) and going from Chlorate to Perchlorate. The anode should last for years.

Contact area would be a problem here as the electrolyte would incline to seep between surfaces (despite clamping) making Ti/Pt contact very weak, and greatly reducing the reaction-rate . . . ?

[edit] Incidentally, Pt wire, 500mm x 0.5mm weighs around 2.2 gm!


[Edited on 16-5-2014 by hissingnoise]