An old thread from the moderators forum about formate reduction, gold purity and cleaning up gold

[4metals]

Quite some time ago the moderators had a bit of a technical refining chat in the moderator’s forum. It started with a question that Lou asked me about the silver formate process and evolved from there. I thought it would be appropriate to share this thread with the open forum and see if it can generate more discussion. When any responses die down, I will copy this thread to the library. This particular thread ended up discussing methods to clean up “almost fine” gold into .9999 fine gold.



Feel free to comment.

 

[Lou]

Do you have the price per ounce processed for the sodium formate silver method? What are your chemical costs?

I'm considering switching up processes and this is the proposed process methodology for incoming karat:

AFTER weigh assay and melt:

1. Atomize the gold
2. Instead of aqua regia digest, part the silver and copper with nitric acid
3. Residual gold is boiled in nitric until no color takes to the solution
4. Crude ~99%+ gold is cast into anode and run through Wohlwill to 4N+ (probably close on 5N)
5. Silver/copper solution is treated with hot sodium formate and silver filtered off.

I'm debating on using Ti CP 2 as the cathode or just 24 K sheeting for the cell. All the material is already purchased on assay.
I plan on making the electrolyte about 85 g/L [Au] and will change it when deposit quality deteriorates.

Here's where I'm curious:

1. As I understand it, you're getting 4N doing sterling and keeping the gold residue, but will I get 4N silver from a solution that's say 30% Ag 60% Cu bal Fe, Ni? That is, how good is the purity down to low concentration silver solutions (with high base metal contamination)?
2. Do you precipitate the palladium before using sodium formate? At your conditions, any Pd from white gold will precipitate with the silver using sodium formate and DMG-DMSO isn't quantitative on straight nitric acid solutions. We run into the problem of Pd contamination, just like with the Thum cell.


Presently we just bulk digest the powder in aqua regia or chlorate/HCl (best for high silver), chill, dilute, sulfamic to kill NO_X, filter and then add chelants and gas with SO₂. The AgCl (with PbCl₂) still contains traces of gold and I quite hate dealing with it and final accountability on the gold is a bother because we smelt the AgCl to silver and the gold is ending up in the flux (?! I have no idea how).


Thanks for your thoughts all!

 

[Harold_V]

I'm no help, but I'm having a hard time understanding how gold would be absorbed by flux when it's in the presence of silver.

Harold

 

[Lou]

You and me both, but the issue is, I know there's gold in the AgCl even after rinsing very well and it's not reporting in the silver. It only comes to a few grams per thousand ounces karat (if that).

 

[Harold_V]

Are you using soda ash in the flux?

Harold

 

[Lou]

Yes. Soda ash, borax, fine charcoal, magnesia.

 

[Noxx]

2. Do you precipitate the palladium before using sodium formate? At your conditions, any Pd from white gold will precipitate with the silver using sodium formate and DMG-DMSO isn't quantitative on straight nitric acid solutions. We run into the problem of Pd contamination, just like with the Thum cell.

I guess that it's a matter of pH and temp ?

 

[Noxx]

Lou, you remember writing this :

Sodium, potassium, and ammonium formate will all reduce platinum and palladium, with no discrimination. The fact that this does not reduce the metal down is because a.) pH is too low, b.) temperature is too low. I would be certain to keep the temperature down when doing this reaction and maintain careful pH control. Most of my reductions with formate for PGM group metals are done at about 80*C at pH >6.

and HAuCl₄, what we offer here, we offer. I for one am grateful to 4metals for the formate reduction. Beats hydroquinone, and it's cheaper.
Lou

Reducing Silver nitrate to metallic silver

 

[Lou]

I do remember writing it after I tried the reaction on silver nitrate made from cement silver. My question is more on liquors containing a few percent Pd rather than traces, specifically on material coming from a dental lab. Pd loses/gains electrons more easily than platinum and is more prevalent (any Pt that comes in as jewelery alloy is separated).

Dilute solutions of Pd/Pt need high heat to reduce quantitatively. It seems a probable contaminant if the Pd content is high. pH control will be very important.

 

[Noxx]

Is it possible to drop those PGM first then ?

You can always go the S/X route...

 

[4metals]

I would bet the finely divided gold in the silver chloride is reacting with the chlorine when it cracks off the silver in the melt. How high in concentration does the gold get in the slag? With atomized gold I've never seen the gold concentration in a sugar reduced silver bar over 1%.


The palladium issue in the silver nitrate is alleviated with a dilute DMG solution injected on the feed side of a pump. The DMG grabs the Pd and the filter traps it. (Works well in a silver cell too)

I have found the formate chemistry to be cheap but the complications due to relatively high concentrations of sodium formate in the waste stream are costly to remove. There is a peroxide oxidation step required to successfully treat the waste that is finicky and takes an inordinate amount of intelligent supervision. Factoring in the fact that the intelligent supervision is you Lou I have no doubt you can do it. The guy I set up to do the formate process had his silver to gold ratios change drastically so he went over to silver cells but he does clean up the last silver when a cell solution is changed out with formate and the silver comes out over 999 even with high copper.

I've been experimenting with dropping gold from HCl, chlorate solutions with SO₂ and stopping the SO₂ before all of the gold is dropped. The balance of the gold in solution is gassed after the gold dropped first has been filtered out. The first gold can be 9999 and the balance lower, but never under. 9995. It seems that when the gold reaction goes to completion, some copper comes down at the very end with the gold. Very little but enough to drop below 9999. I'm thinking very cold acid (to get the soluble silver chloride filtered out) followed by gassing a very cold acid to almost completion, will get 9999 in 1 drop with SO₂. High Pd may necessitate a DMG filtration before gassing.

I was never a big fan of electrolytic gold refining, too much inventory held too long.

 

[Lou]

We're well below 0.1% Au content in the flux. I'll probably ball mill it all with the ashes and dust at the end of the year and collect the values in silver with fluorospar to thin the whole mess out. Your explanation is much like mine, I figured that the NaCl and nascent Cl2 produced during the smelting process is scrubbing the gold to the floating flux layer.

If I were going to clean up the sodium formate-laden raffinate, I think most of it could be oxidized to CO₂ electrolytically and finished with peroxide.


I've long noted the same phenomena you have, that is, the gold that comes out first is always the highest quality (it clumps up fastest, filters and rinses best). This makes sense from a physical perspective, because statistically speaking, when there is a high [Au (III)] there are more gold ions per unit volume so when they agglomerate there is less chance of incorporating in baser metals.
One thing I've been experimenting with is a modification of Chris's EDTA-based chelant/precipitant. Chris sent me a rough write up on how he used it to produce 4N gold from karat scrap. I sent off to Brenntag for a sample of the product to evaluate. He used a Dow Chemical product called Versenol 120 which is a strong chelant for valence (II) metals like Pd, Ni and Cu. It has a hydroxyethyl moiety that is selectively attacked by Au (III) while the EDTA structure is a classic chelant for iron (Versenol is chiefly used in de-scaling operations). I find that adding 30-60 mL chelant/L of 250 g*L-1 [HAuCl₄] after the bulk of the gold has been precipitated maintains the high quality of the gold through the end of the reduction as well as speeding it up. The temperature should not exceed 120 F. The pH upon addition is usually 0.5-1 or so and never rises above 4. Chris originally used it by itself as a precipitant for gold, but it's $13/gallon in bulk and it's only 40% w/v solution in water. This makes for very expensive gold precipitation but it does indeed produce 4N gold. The con to it, as it is oxidized, CH2O is produced as the gold attacks the hydroxyethyl part of the molecule and this formaldehyde makes for the potential of generating carcinogenic chloromethyl ethers.

I always chill and dilute the solution to remove AgCl. I sometimes put a 0.5 cm layer of cellite in the filter to remove the silver quantitatively because the AgCl produced when the solution cools and dilutes is too fine and passes through most practical filter paper (1-3 cfm PP is what I use). The cellite filter aid is easily removed when you smelt the silver (the borax makes it into a glass). It's important that you remember that AgCl is more soluble in hyper chloridic conditions (contrary to common ion effect) or when potassium cation is around (ever try KClO₃ and HCl for high silver alloy?). I find that ice does the job well, it dilutes the solution and cools it down. It's more to filter, but more AgCl comes out because the AgCl₂^- complex is broken upon dilution.

I've yet to evaluate Chris's claim of Pd rdxn by heating the solution after the gold has been filtered out.

 

[Noxx]

Lou,

It's funny that you mention the Versenol since I was reading about it yesterday. I think that it is also available in powder form, which might cost less than the pre-mixed sln. I found that Trisodium N-(hydroxyethyl)-ethylenediaminetriacetate is also manifactured by BASF under the brand name Trilon D. Maybe it costs less ?

 

[goldsilverpro]

I've long noted the same phenomena you have, that is, the gold that comes out first is always the highest quality (it clumps up fastest, filters and rinses best). This makes sense from a physical perspective, because statistically speaking, when there is a high [Au (III)] there are more gold ions per unit volume so when they agglomerate there is less chance of incorporating in baser metals.

I've always noticed the same thing. That last bit of gold is most usually darker and finer. Optically, though, the darkness could partially be related to the particle size.

4metals idea of only ppting the bulk of the gold is a good one - simple and I like it. The rest could then be dropped and simply re-worked. I would think the hard part, though, would be knowing when to stop, especially with SO₂. That could surely be worked out, though.

As Lou said, I used the V-120, alone, to ppt the Au. The V-120 is quite alkaline and the drop was a pH thing (I used 0-6 pHydrion paper). At about a pH of 3.5-4.0, all the gold is out. As above, the final Au is darker but, in this case, I think it's all due to the particle size. To do this, it takes about 1.5 gallons of V-120/gal of AR, based on full strength AR. Excess nitric doesn't interfere - Actually, it seems to make it work better. For reworking gold powder, with no evaporation, that's about 20 oz/gal of V-120 - about $.65/oz. For karat gold, of course, it would be a lot more expensive. Since I could smell a little formaldehyde, I always thought that's what dropped the gold, but I think Lou has other ideas about this. If anyone tries this, be aware that a lot of foam is created when adding the V-120. You can add the first 1/4 with no problems. After that, lots of foam. I put about 1 gallon of AR in a 5 gallon bucket and put that in about a 10 gal plastic tray. I added the V-120 in increments and hand stirred the solution often, but always waited until the fizzing died down on each addition to prevent foamovers. The gold always was at least 999.9 Fine. We had to have that fineness because we were making PGC plating salts from it. Before my discovery, they were paying an arm and a leg for 999.9 bullion, when they could find it. The East Coast office had a bank of Wohlwill cells, but that wasn't cheap either. I would guess that we did at least 5000 oz/month, for at least 5 years, using this method, mainly from electronic scrap. I also used it when I had my first refinery.

When I first experimented with this, I tried lots of the Dow Versene (solids) and Versenol (water solutions of the Versenes) products. I ended up using V-120 because it was a liquid to start with, and I got the best results with it. At that time, Hampshire Chemical made a similar, but cheaper, product called Hampol 120, but it seemed to be weaker. A problem with it, nowadays, is waste disposal. The EPA hates chelated heavy metals because it is so damned hard to break the chelate bonds - chelate comes from the Greek word "chelos", which mean claw. There are ways to break them, though.

Lou's idea is to do this differently. As I understand it, he's just adding enough chelate to tie up the base metals and then using other means to drop the gold. For this, about any EDTA or HEDTA product would work, I would think. I did think of doing this a long time ago, but never got around to it. I like it. Sure cheaper than what I was doing.

 

[4metals]

Been busy working on this high purity single pass refining process and have had a few eureka moments. First off chelation works in holding up the base metals once the dosage is worked out for the base metal content. I atomized some 1000 ounce lots to 100 mesh size digested and precipitated with chelant to Dow's specifications and the fine gold was indeed fine 4 nines worth fine. Then came the waste treatment of the chelated waste. World class pain. Long oxidation process required to break the chelation and not something I would look to institute in a commercial quantity of gold production.

Then I had a quantity of fine gold which I had from stopping the process before completion to remove the first dropped .9999 metal and this was the second half of the drop. There was about 2000 fine ounces there and my client was being polite but he needed to do something with that gold. I wasn't looking at it as he was but it was his money tied up in the pile not mine!

So I took a few sponge samples for assay and tried something for S&G's mostly because he had a piece of equipment I could use to do this effectively. I added the gold sponge to the tumbler, added some 25% sulfuric acid, added a counter rotating propeller (this one was titanium) and set them a spinning. After about 15 minutes of extremely effective mixing i shut it down, filtered off the acid and rinsed with distilled water, twice. The sponge was re sampled for assay and it was run on an ICP by difference.

Before the sulfuric pulping process the assay was .99962 average, and after .999993 average!

I did the same thing 2 days later with another 1500 ounce collection I accrued and same thing, in excess of 5 nines. I had used this in the past to clean up gold sponge and we called it pulping. Back then there was no instrumental assaying going on so it was never quantified.

This is cheap, quick, and no waste treatment issues. You should give this a go Lou, I'm sold. I've run it on the gold straight from the atomizing and complete SO₂ gassing process and again .99999 gold. No chelant's cheap, and
simple. Ain't chemistry fun!!!

 

[Noxx]

Nice

 

[lazersteve]

It always seems the simplest solutions are the best. Too bad you had to have all that other expensive equipment assay to confirm it.

Have you tried the 25% sulfuric on lower purity gold sponge to see how well it brings the purity of say 3N gold up?

Steve

 

[4metals]

I have not used it on .999 gold but it should work. There is a great deal of effect from how compacted the sponge is so keeping it moist so it will remain less densely packed and more likely to allow the sulfuric to do its thing.

In 2 weeks I'll be at a refiner who produces around 4 to 500 ounces of fine from stone removal. No cold acid and SMB dropped, it runs around .997. I was tempted to try dropping the digested karat with Versenol 120 to get .9999 but this may do it instead. I have to come up with a cheap tumbler mixer in a size for smaller lots but if it works I will report back.

 

[Lou]

I have not used it on .999 gold but it should work. There is a great deal of effect from how compacted the sponge is so keeping it moist so it will remain less densely packed and more likely to allow the sulfuric to do its thing.

In 2 weeks I'll be at a refiner who produces around 4 to 500 ounces of fine from stone removal. No cold acid and SMB dropped, it runs around .997. I was tempted to try dropping the digested karat with Versenol 120 to get .9999 but this may do it instead. I have to come up with a cheap tumbler mixer in a size for smaller lots but if it works I will report back.

You can bet I'll give this pulping a try. Can you describe more about how the tumbling/pulping process works and the reactor? Titanium is soluble in sulfuric, but I have a small neoprene-lined water tight ball mill with some alumina media. The gold sponge will have to be very well, spongey to mix right. That heavier finer Au at the end of the SO₂ drop probably won't clean up as well because it has a much lower specific surface area and is by its very nature, more likely to be contaminated by statistics alone. I'll give this a shot with Pt, but already output purity on Pt is 4N+

I'm rather amused. I digest gold in HCl and ammonia but never tried sulfuric. I use all polypropylene felt for filtration, so it won't be a tough transition. I'm convinced it is the agitation process that gives such good base metal removal. What temperature parameters?

The Versenol waste treatment is a bear, but you can anodically-crack it, but why do that when sulfuric works so well?

 

[4metals]

The reactor for the dissolving was an oblique cylindrical tumbler similar to this but a smooth cylinder rather than an octagon (but I doubt that will matter)
View attachment 10538

The tumbler ran at about 50 - 60 rpm. Inserted into the front of the tumbler was a high speed propeller rotating to thrust downward. The blade we used was titanium which was coated with what they told me was teflon. The blade was situated just above the solids and the acid was deep enough to prevent splashing. The acid is 25% reagent sulfuric, the balance (75%) is hot water.

The rotation of the barrel combined with the agitation provided with the down-thrusting propeller seems to break up any clumps and provide good contact with the gold. (pulp) 15 minutes seemed to do it for me.