Sciencemadness Discussion Board

Recovery and isolation of Tungsten from Tungsten Carbide?

Ghrrum - 22-4-2024 at 05:37

So I like to think I am just this side of clever, however I'm well aware that most of that is actually hubris, so I figure I should ask folks that are smarter than I to check my thought process here.

First, the references:
https://patents.google.com/patent/US4533527A/en
https://patents.google.com/patent/US4784687A/en

My aim is to make something that is mostly made of tungsten and heavier than lead of equivalent volume, be it cube, sphere, etc. This is just to make something that I think will be nifty and challenging, no larger reason past that.

I'll be working from scrap tungsten carbide tooling crushed 1cm-ish granular or as close to powder using a hydraulic press.

Since I'm dealing with some degree of uncertainty IRT the grade of WC used in the tooling I have to make some assumptions on the purity, I'll assume 100g of WC at 50/50 for initial tests.

I'm looking for input before I start.


j_sum1 - 22-4-2024 at 16:33

Not sure what you mean by 50/50 here. Presumably mole ration of WC.

Without reading your patents, I suspect this is all a bit ambitious. WC is very inert. If you found something to dissolve it then you will end up with a tungsten solution and the challenge of reducing it back to metal. The bigger challenge then will be processing it into the shape you want. I don't immediately see a route that avoids ultra high temperatures.

Ghrrum - 23-4-2024 at 04:29

Abstract from the first article:
"Tungsten can be extracted into solution from tungsten carbide powder residues with aqueous hydrogen peroxide to a greater extent by employing in solution a catalytic quantity of hydrochloric acid, such as a mole ratio to the tungsten preferably in the range of at least 0.1:1 moles of HCl per mole tungsten, and often up to about 2.5 moles of HCl per mole of tungsten at 55° to 85° C. The selection of the mole ratio in practice usually takes into account the grade of tungsten carbide starting material. It is very preferable to use at least 7 moles hydrogen peroxide per mole of tungsten and an extraction temperature of 65° to 75° C."

The end result is Tungsten in solution, which I assume (I get fuzzy around this point) should look like tungstic acid
H2WO4
Liberal application of heat should net WO3 and H2O at the other end.

After that it's adding carbon and more heat to get straight tungsten
2 WO3 + 3 C → 2 W + 3 CO2

Ref: https://en.wikipedia.org/wiki/Tungsten_trioxide
Quote: Originally posted by j_sum1  
Not sure what you mean by 50/50 here. Presumably mole ration of WC.

Without reading your patents, I suspect this is all a bit ambitious. WC is very inert. If you found something to dissolve it then you will end up with a tungsten solution and the challenge of reducing it back to metal. The bigger challenge then will be processing it into the shape you want. I don't immediately see a route that avoids ultra high temperatures.


You are correct that I was referring to the ratio of tungsten to carbon in the scrap. Shape is not critical for my desires, I'd be happy with powder.
I'm comfortable working with high temperatures, 1300-1600C is about my limit, but that is about what most of the pyrolytic process I've been looking at hit.

bnull - 23-4-2024 at 05:59

Let's say you have managed to convert all tungsten carbide to tungstic acid and then to metallic tungsten. What next? The melting point of tungsten is uncomfortably high (~3400 °C). Unless you have equipment for sintering, you'll end up with tungsten powder or, in the best case scenario, blobs of tungsten.

Correct me if I'm mistaken. I assume that the matrix of your scraps is cobalt. That's quite interesting in itself (even if cobalt salts are carcinogenic or very close to it). A problem here would be the formation of cobalt tungstate, which is, I believe, insoluble. With an excess of cobalt--which is the matrix--there will be no free tungstic acid.

But let's say that you first treated the scraps with acid to dissolve all the matrix, and now the carbide is in powder or chunks at the bottom of the vessel. The reaction of powdered carbide with hydrogen peroxide is quite fast, with chunks not so. You have two options here: either you recover the metal by reduction with carbon at high temperatures (>1000 °C) or by electrodeposition. Either way, it's hard to shape the metal produced into the forms you desire.

There are at least two books in our Library that deal extensively with tungsten: Vallance and Eldridge's A Text-Book of Inorganic Chemistry Volume VII Part III: Chromium and its Congeners and Mellor's A Comprehensive Treatise on Inorganic and Theoretical Chemistry: volume 11. The Handbook of Preparative Inorganic Chemistry by Georg Brauer (one of my favorites, halfway to the end of the page) has two processes in pages 1417-1418, of which the second one is more appropriate in your case: reduction with metallic zinc. If you're careful enough, it's quite straightforward.

Again, the only ways to shape the metal into the forms you want require very high (>3400 °C) temperatures or sintering equipment. Or you can embed tungsten in another metal and shape the composite.

As for the "nifty and challenging" part, why not crystals of tungstates? A big, well-formed crystal can be quite a challenge. It may take months and good temperature control (and usually a high degree of purity) to grow one as big as your thumb, which by then you can remove it, clean it as much as possible, and preserve it in a block of epoxy.

By the way, welcome to the forum.

Ghrrum - 23-4-2024 at 07:46

Quote: Originally posted by bnull  
Let's say you have managed to convert all tungsten carbide to tungstic acid and then to metallic tungsten. What next?


Doing the thing is the whole end point for me, if I ever ended up with enough of it I might fill a hollow steel ball with it just to play with and have something comically heavy (king architectural sells hollow steel spheres pretty cheap if you ever need one for silly ideas, I have many silly ideas).
Quote: Originally posted by bnull  

Correct me if I'm mistaken. I assume that the matrix of your scraps is cobalt. That's quite interesting in itself (even if cobalt salts are carcinogenic or very close to it). A problem here would be the formation of cobalt tungstate, which is, I believe, insoluble. With an excess of cobalt--which is the matrix--there will be no free tungstic acid.

But let's say that you first treated the scraps with acid to dissolve all the matrix, and now the carbide is in powder or chunks at the bottom of the vessel. The reaction of powdered carbide with hydrogen peroxide is quite fast, with chunks not so. You have two options here: either you recover the metal by reduction with carbon at high temperatures (>1000 °C) or by electrodeposition. Either way, it's hard to shape the metal produced into the forms you desire.


The presence of Cobalt isn't something that was on my radar, so I'll have to rethink my approach. You have, however, reminded me of electrowinning and that does appear feasible to a degree. (ref: https://www.911metallurgist.com/electrowinning-tungsten/ )
Since I'm recycling the stuff the process changes a bit. Cobalt is slightly more electronegative than tungsten and it does slowly dissolve in HCl to form CoCl2.
My thought process is now to still crush the stuff, and try electrowinning it in a dilute HCl to see what pulls out first and check it with a magnet to see if the result is magnetic (cobalt) or not (carbon or tungsten).
Quote: Originally posted by bnull  

There are at least two books in our Library that deal extensively with tungsten: Vallance and Eldridge's A Text-Book of Inorganic Chemistry Volume VII Part III: Chromium and its Congeners and Mellor's A Comprehensive Treatise on Inorganic and Theoretical Chemistry: volume 11. The Handbook of Preparative Inorganic Chemistry by Georg Brauer (one of my favorites, halfway to the end of the page) has two processes in pages 1417-1418, of which the second one is more appropriate in your case: reduction with metallic zinc. If you're careful enough, it's quite straightforward.


You are my hero for the citations. I never realized the amount of literature available on here.

Thanks for such a warm and well thought out welcome.

Rainwater - 23-4-2024 at 12:33

Recovery and isolation is the easy part. I never got good results reducing it back to the metal, used a welder with carbon rods, and got W metal beads contaminated with carbon. Trying with hydrogen gas unexpectedly relocated equipment all over my lab/yard.
Most assayers who deal with W will take Na2WO4·2H2O.
The real profit is the bits and shafts the carbide pieces where attached to. I save up my bits until i have 5-6 lbs to process at once.

Tungsten alloys are mixed with sodium hydroxide and an oxidizer like KNO3 or CO2 free O2, then heated to above 340c. The evolution of NOx will serve as a reaction indicator. I have played around with forcing the brown gas to go slow, or fast. And found that fast outgassing seams to dissolve more alloy per mol of KNO3 used. And a 3 times molar excess will dissolve most of the W.

This forms Sodium tungstate. Which is dissolved in water, filtered dried and stored.
If you want the metal, you can skip the drying step and titrate with HCl. At netural ph, filter and keep the solids, then expose to CO, or H2 gas at 850c. Once you figure out how to melt it into one of those cool element cubes, let me know. 3400c and controlled atmosphere is hard to do.

unionised - 23-4-2024 at 14:30

It's not often I find myself thinking "the easy way to do this would be electron beam melting".

bnull - 23-4-2024 at 14:54

Quote: Originally posted by Ghrrum  
The presence of Cobalt isn't something that was on my radar, so I'll have to rethink my approach. You have, however, reminded me of electrowinning and that does appear feasible to a degree. (ref: https://www.911metallurgist.com/electrowinning-tungsten/ )
Since I'm recycling the stuff the process changes a bit. Cobalt is slightly more electronegative than tungsten and it does slowly dissolve in HCl to form CoCl2.
My thought process is now to still crush the stuff, and try electrowinning it in a dilute HCl to see what pulls out first and check it with a magnet to see if the result is magnetic (cobalt) or not (carbon or tungsten).

No need to rethink it at all. Just process the material in two steps: (1) a bath of unmixed hydrochloric acid or sulfuric acid, which will dissolve cobalt and leave WC unattacked at the bottom of the flask, which you can then filter and wash with water; (2) a bath of hydrochloric acid with peroxide to dissolve the carbide as suggested by the patents. If peroxide doesn't perform well, maybe a solution of persulfate in sulfuric acid could be used.
Quote: Originally posted by Rainwater  
Trying with hydrogen gas unexpectedly relocated equipment all over my lab/yard.
:P

Ghrrum - 25-4-2024 at 09:12

Quote: Originally posted by bnull  

1. stick it in acid
2. wash it off
3. stick it in acid WITH hair bleach



Paraphrasing, but essentially that, yeah?

bnull - 25-4-2024 at 10:30

In short, yes.