Sciencemadness Discussion Board

Tripotassium phosphate

MolecularWorld - 9-12-2015 at 12:32

I couldn't find any threads discussing the preparation of tripotassium phosphate.

A recent thread has shown the usefulness of tripotassium phosphate in salting out ethanol and ammonia. Tripotassium phosphate does not appear to be available as any OTC products, or even from mainstream websites such as eBay and Amazon (at reasonable quantity and cost).

Tripotassium phosphate can be produced from the reaction of phosphoric acid with potassium hydroxide. While I can find phosphoric acid readily available as multiple OTC products, I can't find alkaline potassium compounds OTC (at reasonable concentration and cost). Potassium chloride is widely available OTC, and quite inexpensive (less than $1/lb, if purchased as fertilizer or water softener salt). The question is: how can tripotassium phosphate be made from potassium chloride, without using alkaline potassium compounds?

It is known that phosphoric acid can be heated with the potassium salt of a more volatile acid, to give a phosphate of potassium and the volatile acid. Of the following reactions:

H3PO4 + KCl → KH2PO4 + HCl
H3PO4 + 2 KCl → K2HPO4 + 2 HCl
H3PO4 + 3 KCl → K3PO4 + 3 HCl

...I suspect only the first will go to completion. Equimolar amounts of phosphoric acid and potassium chloride, heated to drive off HCl, should leave monopotassium phosphate. This could be reacted with potassium hydroxide to produce tripotassium phosphate, but again, the idea is to avoid alkaline potassium compounds.

Calcium hydroxide can be found OTC as "slaked lime" or "pickling lime". If monopotassium phosphate is dissolved in water and stirred with a mole equivalent of calcium hydroxide, "tricalcium phosphate" (or something similar) should precipitate, leaving tripotassium phosphate in solution.

3 KH2PO4 + 3 Ca(OH)2 → K3PO4 + Ca3(PO4)2 + 6 H2O

Will this work? Is there a better way to make tripotassium phosphate from potassium chloride?

karlos³ - 9-12-2015 at 13:10

Neutralise H3PO4 with K2CO3, it doesn´t get more simpler than this. Both educts are OTC. I don´t know about your country, but in most both educts are readily available and cheap.

MolecularWorld - 9-12-2015 at 13:24

Where, exactly, are you finding potassium carbonate OTC? The only place I've found potassium carbonate OTC is pottery glaze suppliers, the nearest one of which is over 100 miles away. I've found potassium hydroxide OTC only as a 14% solution, at a price that works out to about $25/lb of potassium hydroxide.

As noted in the opening post, I was already aware of the 'react an alkaline potassium salt with phosphoric acid' route; for me, this is more expensive and no more convenient than the potassium chloride route I outlined above (if it works). And I didn't ask for the simplest method:

Quote: Originally posted by MolecularWorld  
The question is: how can tripotassium phosphate be made from potassium chloride, without using alkaline potassium compounds?
...

Is there a better way to make tripotassium phosphate from potassium chloride?


[Edited on 9-12-2015 by MolecularWorld]

aga - 9-12-2015 at 14:34

I just followed deltaH's recipe and Slowly added (with stirring) 3 mol equiv of KOH to H3PO4

First time i added a load of water, and that just took longer to boil down.

Add some water if it looks too syrupy when you start adding the KOH.

Might not be the Best chemical synthesis explanation ever, however the product worked very well for the application.

Edit:

Drying it is a bitch - highly hygroscopic.

I use a stainless steel dog bowl, dump the K3PO4 into it and stick it in an oven for 1 hour at 150 C to get the dry product.

Needs scraping out and grinding to get a powder.

For approx > 50g quantities you'd then have to stick the powder back in the oven as the time taken to grind it would allow it to suck up atmospheric moisture.

[Edited on 9-12-2015 by aga]

BromicAcid - 9-12-2015 at 16:34

I bought a 5 lb unit of potassium hydroxide from a Michael's arts and crafts store for soap making going on 15 years ago. I just checked their website however and I don't see it. You may be able to find it locally OTC for that purpose but apparently not from Michael's.

The pain in the butt part is that tripotassium phosphate has a damn high melting point. So your mixture is going to rock up on you when you get some heat in it if you start to form your target compound. It would be worth trying to setup for distillation, charge your stoichiometric quantities of phosphoric acid and potassium chloride then add in some water. Equip the setup with a addition funnel and start stripping off water. Continue stripping off water/acid until the pH of the distillate is ~7. Use some AgNO<sub>3</sub> solution to check your product solution for chloride/completion.

This could work, even though the acidity of that last proton is crappy is it ionizeable and you may be able to drive it.

[Edited on 12/10/2015 by BromicAcid]

MolecularWorld - 9-12-2015 at 17:07

Quote: Originally posted by BromicAcid  
This could work, even though the acidity of that last proton is crappy is it ionizeable and you may be able to drive it.

Are you saying this reaction could actually happen?

H3PO4 + 3 KCl → K3PO4 + 3 HCl

This would certainly be more convenient, not requiring calcium hydroxide, though I thought only monopotassium phosphate could be formed in acidic or neutral conditions.

Also: It appears soluble phosphate will also give a precipitate with silver nitrate.

[Edited on 10-12-2015 by MolecularWorld]

AvBaeyer - 9-12-2015 at 18:40

I purchased tripotassium phosphate from Elemental. The price is ok for small quantities. As for KOH, it's widely available on the net - Duda Diesel for one. Potassium carbonate is widely available from soap making suppliers.

The reaction of H3PO4 with KCl will be of limited use. The product will be a mixture of mainly mono- and dihydrogen phosphates. You won't get tripotassium phosphate.

AvB

MolecularWorld - 9-12-2015 at 18:51

There's a few listings on eBay too. I didn't say tripotassium phosphate is completely unavailable, just that it's not available OTC or online as cheaply as I could potentially make it. I'm also aware potassium hydroxide and carbonate are available online, but again, not as OTC or as cheap as potassium chloride.

I looking to make about a pound, for which Elemental wants $32.49 plus shipping. At that price, it's worth it to me to try an alternate synthesis from KCl.

I noted in my opening post that the reaction of H3PO4 with KCl would most likely give only monopotassium phosphate (although it appears BromicAcid may disagree), which is why I included the second step of converting the monopotassium phosphate to tripotassium phosphate with a suspension of calcium hydroxide. This second reaction is the one I'm less confident will work. Anybody?

AvBaeyer - 9-12-2015 at 18:57

Sorry I missed your statement in your opening- apologies. As for the use of Ca(OH)2, I think that you need to take a look a the pKa's of the various phosphate species and the Ksp's of the calcium phosphate species. You may end up with a bunch of partially soluble species that reach a buffer pH. Phosphates do not often behave as expected.

AvB

BromicAcid - 9-12-2015 at 18:58

Wow, my brain just glossed over the whole insolubility of silver phosphate thing, my mind was simply on "check for chloride using silver nitrate".

Regarding the feasibility of the reaction, one of those protons will go with the chloride to make HCl. This is analogous to the known method to make hydrogen iodide, in that method you use phosphoric acid instead of sulfuric acid because the sulfuric acid is a strong enough oxidant to react with the formed hydrogen iodide to give sulfur dioxide and iodine. Although Wolen posted sometime back that he didn't have a good results making HI but he did make HBr easy enough by this method. Anyway, it will work to make HCl but how many of those protons will surrender under even driving conditions (beyond the first) I would not guess with any great certainty.

deltaH - 9-12-2015 at 23:43

Let me first say I think your idea is inspired, MolecularWorld! I particularly like the idea to use lime to generate the basic K3PO4 salt... I think this could be made to work, but I doubt you'd make Ca3PO4, this doesn't normally form from aqueous solutions, instead, I think you will form CaHPO4. That just means you need to use a little more lime.

Just be cognizant that lime is a stubborn reagent. I would suggest boiling the KH2PO4 solution with the lime powder (in 20% excess) for some time. Normally when the lime reacts nicely, you see some change in the physical consistency, for example, you might form copious and voluminous white material with the consistency of a thick cream, if you see this develop during boiling or before, then it's a good sign. Boiling might also help to consolidate the precipitate into particles that are easier to filter, because a cream is a pain.

Finally, even if your final pregnant liquor (salt solution) is contaminated with a little chloride and calcium, so what... it should still work well for salting out. The big danger though is that your form K2HPO4 or mixtures of it with K3PO4 and it would be hard to know that this occurred.

[Edited on 10-12-2015 by deltaH]

woelen - 9-12-2015 at 23:55

You can also use a hybrid approach for making K3PO4. First use KCl and H3PO4 to get KH2PO4+K2HPO4, probably a mix of both and once you have that, complete the process with K2CO3. Then of course you still need the (in your location more expensive) alkaline potassium compounds, but only 1/3 to 1/2 of the amount needed in a full synthesis based on them.

AJKOER - 10-12-2015 at 10:19

An idea is to use a variation of Scheele's process which normal makes as one of its products, NaOH, but by substituting KCl for NaCl, one may be able to produce KOH to react with H3PO4.

For more background on Scheele's process, here is an extract from Atomistry (link: http://lead.atomistry.com/basic_lead_chlorides.html ), to quote:

"PbCl2.3PbO is also supposed to be produced in Scheele's process for manufacturing caustic soda by the action of litharge on sodium chloride solution, thus:

2NaCl + 4PbO + H2O ⇔ 2NaOH + PbCl2.3PbO,

but according to Berl and Austerweil the compound PbCl2.4PbO.2H2O is produced when a normal solution of sodium chloride is employed, thus:

2NaCl + 5PbO + H2O ⇔ 2NaOH + PbCl2.4PbO.

The hydroxychloride formed in Scheele's process turns yellow when heated till it is anhydrous. In 1787 Turner took out a patent for preparing caustic soda by Scheele's reaction, and the ignited basic chloride was employed as a pigment under the name of Turner's yellow, or patent yellow. "

Read also Atomistry comments on Lead monoxide at http://lead.atomistry.com/lead_monoxide.html .

The suggested variation would be to substitute KCl for NaCl for the possible creation of KOH, which is then reacted with H3PO4.
------------------------------------------

There may also be a path using Silver in place of Lead. See http://www.sciencemadness.org/talk/viewthread.php?tid=62026 . The equation would be:

Ag2O + H2O + 2 KCl --) 2 KOH + 2 AgCl (s)

Note, the formed AgCl can be reduced by boiling with the appropriate sugar solution and thus effectively recycling the Silver.
-------------------------------------------

[Edit] A wholly original and speculative path maybe to simply boil a not too concentrated solution of KCl with a mixture of Cu and freshly prepared Cu(OH)2. The creation of CuCl (or a basic chloride) would suggest KOH formation.
-------------------------------------------

However, before doing any of the above, I once purchased TSP (trisodium phosphate) from a hardware store. I would try using it in place of the preferred potassium salt to see if it is worth the effort.

[Edited on 10-12-2015 by AJKOER]

[Edited on 10-12-2015 by AJKOER]

MolecularWorld - 10-12-2015 at 11:15

Quote: Originally posted by woelen  
You can also use a hybrid approach for making K3PO4. First use KCl and H3PO4 to get KH2PO4+K2HPO4, probably a mix of both and once you have that, complete the process with K2CO3. Then of course you still need the (in your location more expensive) alkaline potassium compounds, but only 1/3 to 1/2 of the amount needed in a full synthesis based on them.

Good to know this would work:
Quote: Originally posted by MolecularWorld  
Equimolar amounts of phosphoric acid and potassium chloride, heated to drive off HCl, should leave monopotassium phosphate. This could be reacted with potassium hydroxide to produce tripotassium phosphate, but again, the idea is to avoid alkaline potassium compounds.

I'll probably do that if the calcium hydroxide route fails.

- - - - - - - - - - - - - - - - - - - - -
Quote: Originally posted by AJKOER  
However, before doing any of the above, I once purchased TSP (trisodium phosphate) from a hardware store. I would try using it in place of the preferred potassium salt to see if it is worth the effort.

Already done. Store-bought TSP appears to be effective at separating ethanol from a 40% solution, despite a ~20% "sodium sesquicarbonate" impurity. What still needs to be done is to compare the effectiveness of trisodium phosphate to tripotassium phosphate. For that, I need tripotassium phosphate, and purified trisodium phosphate. Purifying the TSP should be as simple as dissolving it in warm water and slowly adding phosphoric acid until effervescence ceases, then recrystallizing (its temperature/solubility curve is perfect for recrystallization).

[Edited on 11-12-2015 by MolecularWorld]

MolecularWorld - 10-12-2015 at 21:41

Best laid plans...
Quote: Originally posted by MolecularWorld  
Purifying the TSP should be as simple as dissolving it in warm water and slowly adding phosphoric acid until effervescence ceases, then recrystallizing (its temperature/solubility curve is perfect for recrystallization).

An attempt to convert the sodium sesquicarbonate in the store-bought TSP into additional trisodium phosphate required more phosphoric acid than expected (as judged by adding phosphoric acid until effervescence ceases). I first thought that the TSP had been cut with more sesquicarbonate than the 20-25% the MSDS claims, but I've since noticed the following claim on the Wikipedia article for trisodium phosphate:
Quote:
Carbonate can only produce disodium phosphate:
Na<sub>2</sub>CO<sub>3</sub> + H<sub>3</sub>PO<sub>4</sub> → Na<sub>2</sub>HPO<sub>4</sub> + CO<sub>2</sub> + H<sub>2</sub>O

This claim seems suspect, and is unreferenced, but it would explain my observation. It would also mean tripotassium phosphate can't be made from potassium carbonate.

[Edited on 11-12-2015 by MolecularWorld]

woelen - 11-12-2015 at 01:01

Wikipedia is right, you cannot make Na3PO4 from Na2CO3 and H3PO4. The reason of this is that H3PO4 is a relatively weak acid and solutions of Na3PO4 are quite strongly alkaline.

If Na3PO4 is added to water, then the following happens:

Na3PO4 --> 3Na(+) + PO4(3-), followed by
PO4(3-) + H2O <--> HPO4(2-) + OH(-)

Such solutions then react with CO2:

CO2 + OH(-) --> HCO3(-), followed by
HCO3(-) + OH(-) <--> CO3(2-) + H2O

You need quite some excess acid to have all carbon dioxide expelled from the liquid.
You can make Na3PO4 again by adding NaOH, until the solution has a pH of appr. 12. You solution contains some PO4(3-), mainly HPO4(2-) and some H2PO4(-), together with OH(-). On evaporation of the water, however, the phosphate will crystallize as PO4(3-). The evaporation MUST be done without air contact, otherwise the solution will absorb CO2 from the air and you end up with a mix of Na2HPO4 and Na2CO3. That is exactly what commercial "Na3PO4" is, it is alkaline phosphate solution allowed to evaporate in contact with air, so that it absorbs some CO2 and hence contains Na2CO3. For most applications such carbonated "Na3PO4" is good enough, it still has 3 Na(3+) ions for each phosphate ion and is quite strongly alkaline.

I forgot to mention in my previous post for making K3PO4 that the end product will contain some carbonate as well, but for most applications this is no issue.

[Edited on 11-12-15 by woelen]

deltaH - 11-12-2015 at 12:17

There's a workaround:

(1) H3PO4(aq) + Ca(OH)2(s) => CaHPO4(s) + 2H2O(l)
(2) 3M2CO3(aq) + 2CaHPO4(s) + Ca(OH)2 =>(boiling)=> 3CaCO3(s) + 2M3PO4(aq) + 2H2O(l), M = Na, K

[Edited on 11-12-2015 by deltaH]

chloric1 - 19-6-2024 at 23:25

Quote: Originally posted by deltaH  
There's a workaround:

(1) H3PO4(aq) + Ca(OH)2(s) => CaHPO4(s) + 2H2O(l)
(2) 3M2CO3(aq) + 2CaHPO4(s) + Ca(OH)2 =>(boiling)=> 3CaCO3(s) + 2M3PO4(aq) + 2H2O(l), M = Na, K

[Edited on 11-12-2015 by deltaH]


Are you completely sure? No need to bring calcium ions to the party! I’m personally fine with a concentrated aqueous solution as I will use it for brazing and soldering. So I will just neutralize phosphoris acid with potassium hydroxide