Sciencemadness Discussion Board - sodium fluoride from calcium fluoride and sodium phosphate

sodium fluoride from calcium fluoride and sodium phosphate

tetrahedron - 9-11-2012 at 10:02

here's a breakthrough on my quest for soluble fluorides without HF. the solubility product of fluorapatite¹ (Ca₅(PO₄)₃F) is around 10^-60, while for fluorite² (aka fluorspar, CaF₂) this value is several orders of magnitude larger, around 10^-10. this led me to contemplate the following theoretical reaction:

10 CaF₂ + 6 Na₃PO₄ ---> 2 Ca₅(PO₄)₃F + 18 NaF

to be carried out in boiling aqueous solution, driven by the precipitation of fluorapatite. i found evidence that this approach could work:

Quote:

CaF2 can be converted to fluorapatite in phosphate solutions at various temperatures ranging between 25 and 75°C in the pH range of 6.5 to 8.5³

i think the mechanism involves two steps:

1. 3 CaF₂ + 2 Na₃PO₄ ---> Ca₃(PO₄)₂ + 6 NaF

2. Ca₃(PO₄)₂ + CaF₂ ---> 2 Ca₅(PO₄)₃F

the solubility product of calcium phosphate¹ is around 10^-30, so the first step must be the bottleneck. the second step should proceed almost quantitatively. both an excess of CaF₂ and a fine grain are probably in order, given the risk of the less soluble Ca₃(PO₄)₂ coating and inactivating the particles.

¹ Ionic Product and Solubility Product
² Selected Solubility Products and Formation Constants at 25oC (note: this source gives slightly different K_sp values for the above-mentioned compounds)
³ Transformation of Calcium Fluoride for Caries Prevention

any thoughts, suggestions, and offhanded attempts at debunking my scheme are warmly welcome =D

[Edited on 9-11-2012 by tetrahedron]

blogfast25 - 10-11-2012 at 07:25

For CaF2 this resource:

http://www.ktf-split.hr/periodni/en/abc/kpt.html

… gives a value of 3.45 x 10-11 for Ks, so not hugely different from your value and still much, much higher than values for calcium orthophosphate or fluorapatite.

There’s no real reason to expect that such a displacement reaction from an insoluble ionic compound to an even more insoluble ionic compound can’t proceed (see e.g. the home lab preparation of KOH solutions from K2CO3 and Ca(OH)2, thanks to the isolubility of CaCO3) but expect it to be fairly slow. Also, NaF isn’t highly soluble (about 4 g/100 g water, fairly flat over temperature), so unless you get the concentrations really right you’ll still have to leach the NaF from the solid gunge.

It’s also a recipe for seriously etched glass ware!

And if you really are going to dabble in the chemistry of soluble fluorides, be prepared to gloved and goggled up: they're nasty...

[Edited on 10-11-2012 by blogfast25]

Lambda-Eyde - 10-11-2012 at 07:35

Very nice work, now carry it out and give us experimental results.

Quote: Originally posted by blogfast25

It’s also a recipe for seriously etched glass ware!

I thought this only occured with HF? Or at least to a lesser extent with fluorides only?

Quote: Originally posted by blogfast25

And if you really are going to dabble in the chemistry of soluble fluorides, be prepared to gloved and goggled up: they're nasty...

This right here. Don't underestimate fluorine and its compounds!

Eddygp - 10-11-2012 at 07:49

A good idea would be to melt them to get some interesting reactions, (maybe not fluoroapatite as the phosphate would change) but it isn't as easy to do.

blogfast25 - 10-11-2012 at 07:53

Quote: Originally posted by Lambda-Eyde

Very nice work, now carry it out and give us experimental results.

I thought this only occured with HF? Or at least to a lesser extent with fluorides only?

This right here. Don't underestimate fluorine and its compounds!

There is HF present (in small amounts) in that set up. That’s because the equilibrium F- + H2O < === > HF + OH- that wasn’t mentioned (HF is actually a quite weak Bronsted acid), also rules. Very alkaline conditions would push that back to the left but these conditions weren’t recommended.

Re. fluorine compounds, it depends a lot what compounds we’re talking about : non-stick pans are coated with Teflon remember? But soluble ionic fluoride compounds are very toxic: that too is due to hydrolysis because the resulting HF is quite soluble in cell membranes and once inside human (or animal) tissue it seeks out Ca2+ ions in our electrolyte balance, precipitating them as CaF2. Burns that are painful and hard to heal…

[Edited on 10-11-2012 by blogfast25]

[Edited on 10-11-2012 by blogfast25]

blogfast25 - 10-11-2012 at 08:16

Quote: Originally posted by Eddygp

A good idea would be to melt them to get some interesting reactions, (maybe not fluoroapatite as the phosphate would change) but it isn't as easy to do.

Hmmm... in a melt you're not really taking advantage of the magnitudes of difference in the solubility product, which is what this 'wet' proposal relies upon. In a fusion you'd rely more on Heats of Formation to determine what mix of compounds you end up with after solidification. Different thing altogether.

tetrahedron - 10-11-2012 at 09:05

thanks for the tips guys.

Quote: Originally posted by blogfast25

in a melt you're not really taking advantage of the magnitudes of difference in the solubility product, which is what this 'wet' proposal relies upon.

there's also the kinetics aspect, which might be favored by high temperatures. does anyone know the melting point of anhydrous Na₃PO₄?

Quote: Originally posted by blogfast25

Na₃PO₄ powder is sold as a degreaser. its solutions are basic, so the formation of HF in my procedure should be inhibited. i have a box but i couldn't find it, so i made it myself from phosphoric acid, adding enough soda until fizzing stopped, and then some, and crystallizing it out by cooling. this was quickly washed and used as is without further purification or measurement, and mixed with a 150% molar excess of washed pottery grade CaF₂, filled with enough distilled water to make bumping (pretty bad due to a heavy sludge) manageable, and boiled twice for ~30' with cooling in between. the product was decanted, the supernatant reduced some more and is now out to crystallize. i have more precise figures but before disclosing them i want a yield. a picture of the crop so far (contrast enhanced):

the crystals have the appearance of serrated spearheads and a spicy aftertaste (kidding!)

edit. according to this video on the growth of Na₂CO₃, the needle-like backbone could be due to this compound, while the phosphate and/or fluoride contribute to the sawtooth ('dendritic') effect. here the crystal habit of NaF in the presence of cosolutes is discussed. it seems octahedra or cubes are pretty much the norm.

Quote: Originally posted by blogfast25

NaF isn’t highly soluble (about 4 g/100 g water, fairly flat over temperature)

i shouldn't get my hopes up of finding any cubes during the cooling then. the gunge looks unchanged, but i'll see what i can squeeze out of it.

[Edited on 11-11-2012 by tetrahedron]

blogfast25 - 10-11-2012 at 09:20

Quote: Originally posted by tetrahedron

Na₃PO₄ powder is sold as a degreaser. its solutions are basic, so the formation of HF in my procedure should be inhibited.

Look at it like this: if the scheme works, you end up with a saturated solution of NaF. Trust me, that etches glass albeit slowly. The equilibria are what they are and you have some HF in there.

Just how bad the attack would be depends on temperature and duration, primarily.

One reason why I’m not enormously interested in synthing NaF is that it’s not very useful because it’s not very soluble. You may have to evaporate large volumes of NaF bearing solution to obtain relatively little solid product. Which ideally you then need to convert to NH4F (dry distillation with NH4Cl – see Brauer) to have something useful. Also, NH4HF2 is quite cheap.

[Edited on 10-11-2012 by blogfast25]

Vargouille - 10-11-2012 at 09:50

Since potassium and ammonium fluorides are much more soluble than the sodium version, potassium being the most soluble of the three, perhaps attempting the experiment once over with potassium or ammonium phosphate (either purchased or made analogously to the sodium version with something like KOH or NH3 as a base) will achieve a better yield of a soluble fluoride. This way you can avoid a laborious leeching process. On the other hand, the etching of any glassware used is likely to increase as well, since the increase in fluoride concentration lends to an increase in HF concentration.

blogfast25 - 10-11-2012 at 10:28

Good points, Vargoulle.

12AX7 - 10-11-2012 at 18:18

I expect the reaction would proceed in a melt as well.

Given the low solubilities, an autoclave would be handy here. Given also that glass may be corroded (additional to the weak presence of molecular HF, wouldn't the presence of F- also complex silicon to form hexafluorosilicate ions?), a metal container may be advisable. Is steel corrosion-resistant under the influence of fluoride? How about stainless? Whatever the material, pipes can be purchased, electric heater cord applied and the assembly monitored from a distance, under a well-weighted-down blast shield.

Tim

plante1999 - 10-11-2012 at 18:34

Iron handle extremely well hydrogen fluoride and probably fluoride too. Most of my fluoride/fluorine related experiment are done in iron with satisfying result. I guess stainless steel is good but iron do the job so well and is so cheap that I don't think it worth using stainless steel.

[Edited on 11-11-2012 by plante1999]

tetrahedron - 11-11-2012 at 04:10

upon standing outside ovenight a crop of small (0.1 - 1 mm) crystals formed on the walls of the beaker. under the microscope these show an irregular shape, almost amorphous:

Quote: Originally posted by blogfast25

Look at it like this: if the scheme works, you end up with a saturated solution of NaF. Trust me, that etches glass albeit slowly. The equilibria are what they are and you have some HF in there.

unfortunately i didn't notice any etching, so i have no proof of a saturated NaF solution.

Quote: Originally posted by Vargouille

Since potassium and ammonium fluorides are much more soluble than the sodium version, potassium being the most soluble of the three, perhaps attempting the experiment once over with potassium or ammonium phosphate (either purchased or made analogously to the sodium version with something like KOH or NH3 as a base) will achieve a better yield of a soluble fluoride. This way you can avoid a laborious leeching process. On the other hand, the etching of any glassware used is likely to increase as well, since the increase in fluoride concentration lends to an increase in HF concentration.

sodium fluoride is classified as toxic by ingestion but only 'irritating' to skin. handling it in the absence of acids is not especially dangerous. ammonium fluoride is a different beast, it can sublime, etc. let alone the bifluoride. the potassium analog would be a practical alternative, but potassium salts are more expensive.

Quote: Originally posted by 12AX7

Given also that glass may be corroded (additional to the weak presence of molecular HF, wouldn't the presence of F- also complex silicon to form hexafluorosilicate ions?), a metal container may be advisable.

i guess the affinity of fluoride for phosphate must be higher than for silicate.

[Edited on 11-11-2012 by tetrahedron]

Vargouille - 11-11-2012 at 04:48

They are not so much more expensive that using potassium rather than sodium is inadvisable. At 100C, NaF has a solubility in water of 5.05 g/ 100 mL, and 4.04 g/ 100 mL at 20C. KF, on the other hand, has a solubility of about 350 g/ 100 mL at only 18C. KOH is more expensive than NaOH, but only by a few dollars. K2CO3 and KHCO3 are quite a bit more expensive than their sodium counterparts, (and, interestingly, they are also more expensive than KOH, at least on the places I saw them sold, so KOH is probably a better idea) but the increased yield of fluoride would still justify it.

blogfast25 - 11-11-2012 at 07:25

Quote: Originally posted by tetrahedron

unfortunately i didn't notice any etching, so i have no proof of a saturated NaF solution.

Filter off the supernatant solution till perfectly clear. Test for fluoride ions with strong CaCl2 (or equivalent soluble Ca salt) solution. CaF2 when I last saw it forming, forms a kind of slightly gelatinus precipitate. At low concetration of fluoride it may take some time to 'develop'.

How long did you stew your brew for? With such low solubility of one of the reagents its concentration is very low and thus reaction rates can't be very high.

[Edited on 11-11-2012 by blogfast25]

[Edited on 11-11-2012 by blogfast25]

tetrahedron - 12-11-2012 at 17:19

Quote: Originally posted by blogfast25

i gave up on this synth, it's been a disappointment. i might try the potassium analog instead.

Quote: Originally posted by blogfast25

How long did you stew your brew for? With such low solubility of one of the reagents its concentration is very low and thus reaction rates can't be very high.

twice for about 30 minutes. actually the equilibrium constants say nothing about the speed of the reaction. what matters is how fast each calcium compound goes into solution/precipitates. moreover, in the absence of a global equilibrium constant the 'common ion effect' is only approximately right. what's certain is that due to the high phosphate (and possibly fluoride) concentration very little calcium will be present in solution at any time.

Quote: Originally posted by Vargouille

They are not so much more expensive that using potassium rather than sodium is inadvisable. At 100C, NaF has a solubility in water of 5.05 g/ 100 mL, and 4.04 g/ 100 mL at 20C. KF, on the other hand, has a solubility of about 350 g/ 100 mL at only 18C. KOH is more expensive than NaOH, but only by a few dollars. K2CO3 and KHCO3 are quite a bit more expensive than their sodium counterparts, (and, interestingly, they are also more expensive than KOH, at least on the places I saw them sold, so KOH is probably a better idea) but the increased yield of fluoride would still justify it.

my Na₂CO₃ is like $3/kg, beat that with any potassium salt ;p for a small-scale proof-of-concept prep KCO₃ would be preferable though.

[Edited on 13-11-2012 by tetrahedron]