What did I do wrong?

[Bertho]

I started a batch of very old through hole HP PCBs
Picked out the boards with high yield
With hammer and chisel and air hammer removed all components
Cut up the boards to easy handled size
Weighed the batch: 4.23 kg

I did a test batch of 712g of PCBs:
I stripped the solder with hot HCl
I leached using HCl Cu-2 with a tiny amount of peroxide to start it and then an air bubler.
Dissolved with AR, precipitated with SMB
Rinsed,rinsed.., boiled in HCl, rinsed...
Yield: 2.4g = 0.34%

I did the reminding 3518g of PCBs
The only difference was using HCl & Clorox to dissolve the gold.
The yield: 2.0g = 0.06% !!!!

What went wrong the second time?
Several comments first:
I am careful in general and let things settle, regardless, the basic procedure was similar except the second batch probably was a longer time in the copper stripping bath.
I have good stannus testing solution that I always carefully use and it is verified good.
I had a problem with the thermostat failing and some of the boards were boiled in the HCL solder stripping process.
I was fighting with a sludge problem but it got dissolved with the hot HCl & Clorox.

Besides the "what went wrong" question, under what conditions does the stannus test fail to detect dissolved gold?

I might have found the missing gold: I just tested the Cu-2 solution and it at once gives a black percipient with the stannus test.
How can I easiest recover the gold from the HCL/Cu-2 solution. I have about 2 gallon / 6 liter?
What are the false positives using stannus?
Bertho

 

[kernels]

Unless the boards are all identical, my best guess would be that in the original test batch, you had 1 spectacular board mixed in with many average boards, bringing the yield up significantly over the average.

To recover dissolved Gold, your best bet will probably be to cement onto Copper. With the obvious caveat that some more Cu will obviously dissolve into the Cu(II)Cl leach.

 

[nickvc]

Another option if you have some low grade board material to hand add it to the solution, as it strips the copper and leaves the gold any gold in solution will be cemented back out of solution, two birds one stone.

 

[Bertho]

Thanks for your suggestions.
Under what conditions will the stannus test fail to detected dissolved gold?
What can cause a false positive?

 

[butcher]

Stannous chloride SnCl2 is used to reduce gold from solution, just like the other reducing agents we use.

Just like using SMB sodium metabisulfite to reduce gold from solution.

When we use SMB we are changing the gold ions (the salt of gold), back into elemental gold metal, the gold atoms clump up together and gain weight which precipitates as a gold metal powder...

When we use stannous chloride we are changing the gold ions (the salt of gold), back into elemental gold metal, the only difference is that the small clusters of gold metal atoms develop an electric charge, so the gold becomes like a bowl full of magnets continually pushing each other around in solution so that the gold will never settle, the gold is already reduced to a metal...

Just like using SMB when using stannous chloride to reduce the gold ions back into metal , if there is oxidizer in solution (excess nitric acid, excess chlorine...) we will not be able to reduce the gold to metal, so the stannous will not show a violet color even with gold in solution, as the oxidizer either keeps the gold dissolved, prevents the reduction, or lets the gold re-dissolve again as the reducing agent is trying to react to reduce the gold.

Some reducing agents although fairly selective for gold but can also reduce other metal ions to metal, like copper, overusing SMB with a solution loaded with copper ions can reduce copper in solution, copper is one metal that gives color to solutions, as the copper is reduced say from Cupric to cuprous it can give a brown color to the stannous chloride test (we call the false positive).

When will SnCl2 not work, if there is no gold in solution, or if the gold in the solution cannot be reduced (because the reducing agent cannot overcome the oxidizing agent in solution)...
The reduction of other metals like copper can affect the color of the solution.

Stannous chloride is a reducing agent, it can also displace metals from solution, for example, silver making a white milky precipitant of silver chloride.

What makes SnCl2 a great test for gold is that it makes a colorful colloid of the reduced gold metal, stuck forever to be in movement in solution, keeping the gold clusters a certain size, that gives a certain color to the solution as the gold reflects the light.

 

[g_axelsson]

One reason stannous might not detect gold is if there isn't any there. Your missing gold might still be in the solids. If some dissolved while etching the copper it can be redeposited among the circuit board scrap.

If there were some base metal in the sludge with the foils it can cement back the gold out of solution. Especially if you had a problem dissolving the gold in the sludge using HCl + Cl. I would take a small sample of it, add real aqua regia this time and then test the solution with stannous.

Depending on what metals there might be left on the circuit boards, gold can accumulate in one spot.
I had a bunch of ceramic CPU:s and two fiber based in the same AR solution once, the majority of the gold cemented back on the circuit board of the fiber CPU:s. It was just luck that I found it before I threw out the remainder. The ceramic CPU:s were totally stripped but the gold redeposited in a hidden spot. I'm sure there is more hiding inside the fiber board and I have to incinerate it before I can get to it.
http://goldrefiningforum.com/phpBB3/viewtopic.php?f=60&t=17697

Göran

 

[Johnny5]

I started a batch of very old through hole HP PCBs
Picked out the boards with high yield
Weighed the batch: 4.23 kg

I did a test batch of 712g of PCBs:
I stripped the solder with hot HCl
I leached using HCl Cu-2 with a tiny amount of peroxide
Yield: 2.4g = 0.34%

I did the reminding 3518g of PCBs
The only difference was using HCl & Clorox to dissolve the gold.
The yield: 2.0g = 0.06% !!!!

Bertho

.3% is about normal for these boards. I've processed lots of them, and currently have about 100lbs (populated weight).
It sounds to me like you cemented the gold, when you used the AC. We've had people on here try to strip pins using AC, and all they end up with is black pins, and no gold in solution.

 

[butcher]

Many modern circuit boards not only have copper traces on both the top and bottom of the board, but they can also have stacks of copper traces in between layers of the fiberglass, places where acids will have a hard time reaching, good hiding places for the gold to plate out of solution onto the remaining copper hiding from your sight.

 

[Bertho]

Thank you guys for detailed replies.
Things are seldom simple!
In the attached picture there are three samples. I am happy that I bought the "sample plate". It makes it so much easier to compare and test.
The left one is my HCl/Cu2 liquid that was used on the PCBs
The center one is with a drop of stannus chloride added. It is not actually that black
The third one is what the liquid looks like after I added a piece of clean pure copper to cement out the gold that might be hiding there.
What happened?? Why would extra copper turn the liquid solid black?
There is only a very small amount of a black deposit sticking to the copper. I am used to see like a uniform sheet when cementing.
I see no way to test with stannus chloride if there is gold in solution. I plan to let it sit a day and then siphon off all except the bottom 10 mm or so and then dilute it so I can see through the liquid.

 

[butcher]

CuCl2 green + SnCl2 --> CuCl (dark green-brown) + SnCl4
The green cupric chloride is reduced to cuprous chloride while the stannous chloride is oxidized to stannic chloride.

Similar to the false positive from (copper in solution) and overuse of SMB...

CuCl2 + Cu --> 2 CuCl
The cupric chloride (green) attacks metal to form cuprous chloride (brownish)

 

[Bertho]

Thanks Butcher.
So in summary, there is presentably no gold there. Just a false positive.
I will continue hunting to see if I can find the missing gold.
Bertho

 

[butcher]

Any gold in solution will cement out of solution with the copper buss bar, or with reusing the solution on the next batch of copper to dissolve.

If you did have any gold dissolved in solution it would be in the green CuCl2 solution, the darker brown solution of CuCl will not hold gold.

 

[Bertho]

I am waiting to see if it has precipitated. The liquid is too dark to see so I want to make sure it has settled before I siphon off most of it and then dilute it so I can see the precipitant.
Bertho

 

[anachronism]

I may be alone (and blind) out here however I am seeing the two right hand readings as gold. Anyone else?

 

[butcher]

Although the dark reaction can be an indication of gold.
It is also the color of cuprous chloride.


You could try a drop of the solution on (damp) paper with the stannous and dilute it to see if it is violet or brown.
You may try diluting to see if a white powder of copper I chloride powder precipitates when diluted, it may also lighten the color to see if it is actually a brown reaction of the reduced copper in solution or violet color of reduced colloidal gold coloring the dark brown cuprous chloride solution.

You can also try the test with a crystal of ferrous sulfate in your spot plate.

CuCl2(aq) (green) + FeSO4(aq) (greenish-blue) --> FeCl2(aq) (yellowish) + CuSO4 (blue)
Basically no reaction or precipitant occurs in the reaction above, although the color can chanes to more of a yellowish color because of the reduced iron, in solution...
If the solution holds gold a brown ring of reduced gold will develop around the crystal of copperas.

http://jimlaurwilliams.org/projects/seychellePaper/index.html

 

[Bertho]

Great info and amazing link.
Thank you!

 

[anachronism]

Ahhh that's "AP" isn't it Butcher- I see what you mean.

 

[butcher]

I do not see much of a reason to try and test the cupric chloride solution for gold, where reusing the solution on more copper-based scrap will cement gold from solution, or we can cement out the gold on a clean copper buss bar.

Not the best indicator of gold in solution, but in a green solution of cupric chloride, if there are high oxidizers and gold is in the solution, I can normally see a kind of yellow halo-ring at the top of the edge of the solution where the solution touches the glass.

 

[Bertho]

Butcher wrote:
You may try diluting to see if a white powder of copper I chloride powder precipitates when diluted, it may also lighten the color to see if it is actually a brown reaction of the reduced copper in solution or violet color of reduced colloidal gold coloring the dark brown cuprous chloride solution.
-------------
As you expected the cuprous chloride precipitated.
It is like magic to see the ink-black liquid turn transparent light green with a white precipitant.
There was no cemented out gold after several days with a copper bar so I have to hunt for it at other places.
Also no reaction with the stannus chloride test.

I now suspect what happened: I dissolved the gold slivers, flakes, and tiny pieces of gold floating around. It was all mixed with some soldermask that peeled off when the PCBs were boiled in HCl, plus the sludge from the processes.
I am assuming that I used too much Clorox to eventually dissolving the gold and the sludge.
I carefully checked with the stannus chloride that there was no gold in the solution's before adding the solution to be killed and disposed. I did not know about the failure modes of the stannus test that Butcher described. :-(
Afterwards, to kill the copper, I added some steel bars and it has been sitting for a few days. Now if the gold is in there it will be mixed with all the cemented out copper.
Of course, my next question is how to proceed to recover from that mistake.
Bertho

 

[kernels]

You could take a small sample of the Copper that possibly contains Gold and dissolve away the Copper with HNO3, if you see some fine powder left over ,then that could be Gold.

You could also melt it all into a anode for electrolytic Copper cell.

You could even re-dissolve the Cu powder with AP if you wanted.

I would certainly do a small amount first to try and see if there is actually any Gold in there.