[Lanthanide]

I've got some need for some sodium nitrite, NaNO2, and the only place I've found to buy it is in the form of 'prague powder', which is commonly used for preserving meats. The stuff I have is 6.3% NaNO2 and the rest is NaCl, although there are also "preservatives" and a pink food colouring, which I'm discounting as being at negligible concentrations.

I really just want to reclaim as much of the NaNO2 as possible, but it doesn't need to be pure. If I could get 50-60g per 1kg of the prague powder I'll be happy. So having say 130g of solid with 50g of that being NaNO2 is fine.

So my method is to dissolve 1kg of powder in 1kg of water while mixing thoroughly, then filter out the water and dump the precipitate, which theoretically should be pure NaCl. Then boil down the water to ~200ml, filter out the water again and dump the precipitate, and then finally evaporate all of the remaining water in an evaporating dish, theoretically leaving 63g of NaNO2 mixed with some amount of NaCl. In reality I'd expect to get much less than that, due to various inaccuracies in the method.

NaNO2 dissolves at 80.8g per 100ml at 20C, while NaCl dissolves at 35.9, and at 100C NaNO2 dissolves at 160g per 100ml while NaCl has only increased to 39.2g/100ml. So for this reason the method is carried out with boiling water, to encourage that all of the NaNO2 is dissolved in preference to the NaCl. As they have a sodium ion in common, getting the NaNO2 to dissolve in preference as much as possible is desirable. As a result of using the boiling water however, throughout the method the water is constantly evaporating, which is seen in my results below.

Anyway, I have performed this 3 times so far, but I can't determine how much NaNO2 is in my final product. The first couple of times I didn't take as many measurements as I should have, but for the last time through I weighed everything at every step. The result I have gotten seems to be impossible, however, so I am at a loss to explain it.

I'm using digital kitchen scales with 1g resolution, I can't gaurantee that they are 100% accurate but they generally seem to be very good - weighing 1000g of water comes pretty much on the 1L line on my measuring jug.

The following measurements are of the chemicals only, the weights of the various containers used has been subtracted already.

Starting water: 1000g
Starting mixture: 1000g
Water after boiled: 986g
Water and mixture initially mixed: 1971 (assume 1g of mixture lost, so 972ml water now).

Decant and filter the water into a separate container.
Remaining precipitate: 754g (there is also an unquantifiable amount of water still in here, as I cannot separate it all out - it's like quicksand).
Solution transferred: 1217g, approximately 900mls.

After boiling the solution down, the weight is now 779g, showing 438ml of water have been boiled off.

Decanting the now boiled solution, leaves a precipitate of 347g behind, again with an unquantifiable amount of water also present.

The decanted solution weighs 411g. After evaporation, the solids weigh 162g, so 249g of water were evaporated off.

162g solids per 249g of water gives a solubility of 650g per litre of water. The solubility of NaCl at 100C is 392g/L, while NaNO2 is 1600g/L. The proportions of NaCl and NaNO2 required to produce this level of solubility is 21.36% NaNO2 to 78.64% NaCl:
21.36% * 1600 = 341.76
78.64% * 392 = 308.2688
= 650.03

From the 162g of solids, 21.36% gives 34.5g of NaNO2, and 78.64% gives 127.5g of NaCl.

However because the solubility of NaCl in water at 100C is 39.2, and I had only 249g of water, the maximum amount of NaCl present, assuming 0% NaNO2, should only be 97.6g, yet I have a whole 30g more of solid to account for? 

I based this method off one a guy posted on another forum, but instead of evaporating at the last step, he chilled the solution down from 100C to 0C, making crystals precipitate out of the solution - going from 160g to 80.8g of solubility for NaNO2, while the NaCl only changed from 39.2 to 35.9g. However I don't believe that the crystallization method would work - the much more ready solubility of NaNO2 makes me believe that unless you have a 100% saturated solution of NaNO2, cooling the solution down would preferentially result in NaCl crystallizing out, so in fact using 200ml of this final solution cooled down to 0C would, in my opinion, not actually crystallize any NaNO2 at all, only NaCl.

Are my conclusions for the original method involving crystallization of the solution correct, or would this method actually work? If it would work, why?

Can anyone point out where I am going wrong in my calculations, or any other suggestions for improvement in my method? How can I work out the concentration of NaNO2 present in my 162g sample? If this whole endeavor even going to be practical at all?

One thought I have had is that I should use 500g of mixture to 1L of water, so that the first decantation/filter step will leave behind only a small amount of undissolved salt, greatly lessening the 'quicksand effect' and allowing much more of the solution to be obtained.

[Train]

What kind of water are you using? 

And, what are you going to do with the sodium nitrite once you have it?

[Lanthanide]

I'm using tapwater, however I live in Christchurch, New Zealand, and we have some of the purest tapwater in the world (no added chlorine - no need). From the city council website:
"Christchurch's water is not considered to be "hard water" and does not require treatment for softening for washing machines and the like.".

I am going to use the sodium nitrite in another chemical reaction.

[billnotgatez]

is there some sort of titration you can do that is specific for one or the other compound

[Lanthanide]

I don't know, is there?

[billnotgatez]

I thought I might find something at this link but seems not.
Maybe the references has something

http://en.wikipedia.org/wiki/Sodium_nitrite