Sciencemadness Discussion Board - Extraction of potassium chlorate from old weedkiller

Extraction of potassium chlorate from old weedkiller

Boffis - 12-6-2018 at 14:16

I recently found two old and partly used tubs of sodium chlorate weedkiller of unknown vintage in the remains of a shed at our recently purchased house. Both tubs contain sodium chlorate type weedkiller that according to their labels are 50 and 52% sodium chlorate respectively mixed with fire suppressants. A quick analysis with silver nitrate confirms that the bulk of the additive is a chloride, apparently sodium chloride since it gives no precipitate with sodium perchlorate. One also contains a small amount of a substance that causes the crystals to “ball up” when first added to water, probably an anticaking agent, while the other contains a small amount of insoluble white powder that presumably functions similarly.

A series of experiments was undertaken to try and identify the best route to the recovery of a useful product from these two tubs. The first experiment simply consisted of boil 100ml of deionised water, pre-weighing and then adding the crystals until no more would dissolve. Allowing for a small amount of undissolved material saturation at 90-100°C saturation was achieved at about 63-64g of the crude material. The solution was filtered hot and cooled but no crystals formed even after several hours so the solution was chilled in the fridge overnight. To my surprise still no crystals formed. Mmmm, the recovery of sodium chlorate was going to be tricky!

While the first experiment was cooling I tried boiling a mixture of 63g of the crude weedkiller and 32g of potassium chloride in 100ml of water but it would not dissolve and so more water had to be added (about 35ml) and when filtration was attempted it rapidly crystallised in the Buchner funnel. This required messy rinsing and re-dissolving in a larger volume of water, filtering and then evaporating of the final overly dilute solution. Not a very succesful procedure.

The following day the clear solution from the first experiment was mixed with a hot solution of potassium chloride comprising 32.8g of potassium chloride dissolved in 70ml of water. The solution required filtering as the potassium chloride clearly contains an anti-caking agent too. An immediate white crystalline precipitate formed. The suspension was cooled in the fridge to about 4°C for about 4 hours and then filtered, the cake being sucked as dry as possible and dried in a warm place to give 32.13g of crude potassium chlorate or about 79-80% of theory. Dissolving a little in deionised water and testing with silver nitrate solution gave a strong precipitate but nothing like the density of the original material.

A further experiment was carried out where 70g of crude chlorate weedkiller were dissolved in 135ml of hot water and filtered hot while 32.0g of potassium chloride were dissolved in 70ml of hot water and similarly filtered. The two solutions were mixed hot and allowed to cool slowly. Delicate snowflake like plates soon began to form and then settle to the bottom. The solution was finally chilled for several hours in the fridge to 4°C and the crystals filtered off to give 28.94g of dried crude potassium chlorate or about 71-72% of theory. The crystals were not washed because they are rather soluble in pure water so they will be contaminated with some adhering sodium chloride solution and a further recrystallization will be required.

The filtrate from the first and third experiments were mixed together and evaporated down on a water bath in a shallow bowl until a film of crystals formed on top and then cooled. The crystals were recovered and dried as before to give a further 16.16g of crude potassium chlorate bringing the overall yield up to near theoretical, though the second crop demonstrably contains significantly more chloride.

Since the two-solution approach seemed to be easier to conduct due to the lesser tendency of the solutions to crystallise while being filtered this become the method pursued in the further experiments. In the earlier experiments a 25-35% excess of potassium chloride was used but this appears unnecessary and for the “preparation scale” experiments this was reduced to less than 10%.

The next attempt utilised 180g of crude weedkiller dissolved in 300ml of water and filter hot. Then a separate solution consisting of 66g of potassium chloride dissolved in 150ml of boiling water was prepared and filtered hot. The two solutions were mixed while still fairly hot and no crystals formed immediately. As the solution cooled crystals formed on the surface and then sank, producing a snowflake effect that lasted for several hours. The solution was finally chilled in the fridge to 4°C for several hours to complete crystallization. The filtered and dried crystals weighed 72.95g (about 70.4% of theory).

The filtrate (about 420ml) was evaporated down in a glass pan until a film of crystals had formed on the liquid and it didn’t dissolve when pressed below the surface. The solution and crystals were poured into a large beaker and allowed to cool (now about 380ml). When cool, it was chilled in the fridge for several hours and filtered. The drained and dried crystals weighed 29.61g giving an overall yield of 102.56g or about 99% of theory of crude potassium chlorate.

I am going to repeat the last run at 300g of crude chlorate mixture and take some photographs for the prepublication section. I will also investigate the best method for recrystallisation in the near future and attempt to analyse the product for chloride which should be the "acid test" of purity.

fusso - 12-6-2018 at 14:40

Shouldnt this be in Gen Chem instead of RaAA?

Bert - 12-6-2018 at 17:34

Sodium chlorate is soluble in methanol around 52g per 100g of methanol @ 25C.

Sodium chloride is soluble at around 1.3g per 100g of methanol @ 25C.

Never tried it, if you do, proceed carefully and in small ammounts.

I have certainly heard of using chlorate as a weed killer, but these products were gone from the consumer market a long time ago in USA. They used to sell Sodium chlorate based "solidox" pellets for use as an O2 generator for a single tank torch system here, haven't seen a can of the stuff in decades.

Boffis - 12-6-2018 at 22:39

Chlorate based weedkillers have been withdrawn in Europe years ago but tubs of the stuff still turn up in old garden sheds and farm buildings up and down the country as in this case.

Potassium chlorate is much easier to purify than the sodium salt as it has a very steep solubility/temperature gradient (about 3.5g/100ml at 5 C and 50g/100ml at 100 C) and it is non-hydroscopic.

I haven't tried using organic solvents; again it depends on the solubility gradient rather than absolute solubility and if the gradient is rather flat, as it is with salt and KCl in water, you would end up having to boil of the methanol. That sounds like a receipe for a disaster. Water is always my first port-of-call when looking for a solvent to recrystallise anything and in this case it clearly works well.

woelen - 13-6-2018 at 02:30

I would convert it to KClO3. Just dissolve all of it in as little as possible of water, allow insoluble stuff to settle and then mix the clear solution with a concentrated solution of KCl. If white solid appears, then boil, until everything is dissolved again, and then slowly allow to cool down.

From the remaining liquid you can get more KClO3 by boiling it down and then letting it cool down again.

Herr Haber - 13-6-2018 at 06:11

Ahhh... The good old days of sodium chlorate and sugar

These were the days when you knew what your weed killer would decompose into.
Nowadays it's organophosphorous compounds. Sigh...

Interesting that your NaClO3 contains fire supressants. I never encountered that though someone mentioned V2O5 being used for that purpose.

hissingnoise - 13-6-2018 at 07:54

Quote: Originally posted by Herr Haber

Ahhh... The good old days of sodium chlorate and sugar

Yeah, wallopping chlorate/fuel charged steel pipe ends shut with a lump-hammer with carefree abandon?

How I managed to not blow my face off, at least, will remain an abiding mystery...

I'm just lucky I hadn't access to sulphur, at the time!

XeonTheMGPony - 13-6-2018 at 08:24

Quote: Originally posted by hissingnoise

Quote: Originally posted by Herr Haber

Ahhh... The good old days of sodium chlorate and sugar

I am so lucky to grow up with people that would actually just teach me that stuff with out the bs, abstinence only fails in every way that is possible whether it be knowledge or fun in the sack!

I saved a friend from removing his hand from that knowledge!

Boffis - 13-6-2018 at 10:10

Quote: Originally posted by woelen

I would convert it to KClO3. Just dissolve all of it in as little as possible of water, allow insoluble stuff to settle and then mix the clear solution with a concentrated solution of KCl. If white solid appears, then boil, until everything is dissolved again, and then slowly allow to cool down.

From the remaining liquid you can get more KClO3 by boiling it down and then letting it cool down again.

That's precisely what I have done and described above. Its just that there is very little information on this web site about the actual nitty-gritty details of the process. Exactly how much water for a unit of mixed salt etc.

What you do with it once you have purified it is up to the individual. Most of the chlorate I use is as a chlorinating/oxidizing agent with hydrochloric acid and to make the occasion batch of snappers for Christmas crackers.

unionised - 13-6-2018 at 11:18

Quote: Originally posted by Bert

Sodium chlorate is soluble in methanol around 52g per 100g of methanol @ 25C.

Sodium chloride is soluble at around 1.3g per 100g of methanol @ 25C.

I dare say that's right, but I'd not like to have to check.

Boffis - 14-6-2018 at 13:06

Quote: Originally posted by Bert

Sodium chlorate is soluble in methanol around 52g per 100g of methanol @ 25C.

Sodium chloride is soluble at around 1.3g per 100g of methanol @ 25C.

Never tried it, if you do, proceed carefully and in small ammounts.

OK, I tried it....

I am still here and so is my lab. I heated to boil 50g of crude weedkiller and 100ml of >99.5% methanol. Filtered hot and cool in a shallow open bowl so that the methanol can evaporate (boiling it down sounds like attempted suicide). The cake was tipped out onto a watch glass and dried. It weighed 42.5g. I estimate that about 1g remained adhering to the walls of the flask so 6.5g of dissolved solid in 100ml of methanol. The methanol has cooled and evaporated a little already and crystals are forming. I think your 52g figure is per litre not 100ml.

@fusso; I posted this here because the thread is about a method for aquiring a useful, pure chlorate salts from deliberately adulterated material in old weedkiller. There is probably a good deal of this still kicking around in other garden sheds, garages and utility rooms in rural homes and farms all over Europe since it was still on sale at farm store in the UK as little as 3 years ago.