Copper Chloride ("AP") cleanup - Experiment writeup

Gold Refining Forum

Help Support Gold Refining Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

haveagojoe

Well-known member
Joined
Aug 1, 2014
Messages
155
As Copper Chloride users will know, one of the drawbacks we encounter when releasing gold foils from escrap is that the solution becomes saturated with copper and eventually the etching process will slow down, and white crystals of Copper (I) Chloride will begin to accumulate in at the bottom of the reaction vessel.

Using an air bubbler to replenish the oxygen level in the solution will help to keep the Copper in solution as Copper (II) Chloride, which is what gives it the characteristic green colour. However after prolonged use the solution will turn very dark as it becomes completely saturated and the white crystals will form anyway.

The common approaches to dealing with saturation are either to decant the solution and allow Copper (I) Chloride crystals to form, after which they can be removed by further decanting or by filtering; or alternatively the solution can be elongated by adding more Hydrochloric acid to dissolve the crystals back into solution. Neither of these approaches are entirely satisfactory, since the former produces solution which is still close to saturation, and the latter produces a larger volume of solution, which in time accumulates as an ever-growing volume which consumes more and more Hydrochloric acid at not-insignificant expense.

A further issue is encountered when dealing with iron-bearing escrap, since when Iron is dissolved into the the solution it forms Ferrous Chloride. While Ferrous Chloride is not a major problem in the first instance since it also dissolves copper, it does so at a slower pace, the solution cannot be maintained by bubbling oxygen through it and the Iron does not drop out as chloride crystals as Copper does. Commonly this is addressed by using a magnet to sort Iron-bearing scrap from non-Iron bearing, and using a different method where Iron is present. However other metals may also be present in the scrap material, such as Nickel, Tin, Cobalt, Lead, etc which cannot be sorted with a magnet. Some - such a myself - opt to skip the extra step of sorting with a magnet, instead simply accepting that the solution will become fouled by this type of scrap. Doing so is uneconomical and results in an accumulation of "spent" acids which require treatment before they can be discarded.

Having accumulated a number of bottles of spent solution, and unwilling to discard them, I was interested to read a suggestion from Shark that metals could be removed from the spent solution using electrolysis:

https://goldrefiningforum.com/threa...p-processing-circuit-boards.34718/post-374650

Intrigued, I did some research and discovered that not only copper but also other contaminant metal salts can be removed from solution by forming an electrolytic cell to plate them out as solid metal, leaving behind relatively clean Hydrochloric acid suitable for reuse. This raised the possibility that the Ferric Chloride issue could be effectively addressed.

The reaction which takes place when current is applied reduces the aqueous metal salts (such as Copper Chloride) to metal on the cathode (negative electrode), and releases Chlorine gas at the anode (positive electrode) by oxidation.

Since Chlorine gas is noxious, I decided to perform my first experiment at a very small scale: 150ml of saturated Copper Chloride solution in a beaker, using two graphite pencil leads as electrodes with a spacing of around 15mm. The voltage required is low, since the reduction potential of copper is just 0.34V+, though for practical purposes in a real-life scenario between 2V and 3V is optimal. I used my bench power supply set at 2.5V, and as the reaction started the current draw was around 100mA.

Even at such a small scale, the production of Chlorine gas was noticable, rising as a continuous stream of fine bubbles from the pencil lead anode and giving off a distinctive pungent smell. However, after just a few minutes Copper metal deposits became visible on the cathode, and after 4 hours the solution was noticably clearer and the pencil lead was encrusted with a thick, uneven layer of copper with a remarkable appearance:

copper on pencil lead.jpg

The deposited copper weighed just over 0.5g.

When I briefly increased the voltage to 5V, the cathode instantly turned black as other contaminant metals such as Iron and Tin were co-deposited along with the copper. They have higher reduction potentials, so at the lower voltage only copper would be deposited until the solution became almost copper-free, after which the other metals will be deposited in ascending order of reduction potential.

Deeming the initial experiment a success, I decided to set up a slightly larger one, using a piece of flattened copper pipe as the cathode and a graphite artist's drawing stick about 8mm in diameter as the anode. I had read that Stainless Steel could be used for the anode but since it may cause Chromium to be introduced into solution I decided against it, since Chromium is toxic.

Again using my bench power supply, I ran 3V through 4 litres of saturated Copper Chloride solution with an electrode spacing of around 20mm, which drew a current of around 1A. I was surprised that very little Chlorine was produced and the deposition of Copper on the cathode was much slower. I put this down to the composition of the graphite stick I used for the anode, since it would contain clay for hardness and strength, which would reduce its current-carrying capability.

After 12 hours the solution appeared slightly clearer and had stratified somewhat, with the clearer part near the surface close to the electrodes while the lower half remained quite dark. The deposits on the cathode were much more uneven, forming "dendrites"- curious-looking growths of metal which stretch out towards the anode. This can be a result of higher current density so I learned that at 3V, 20mm spacing was a litle too close. If I had left the cell to run much longer the dendrites would have made contact with the anode, forming a short-circuit, so I ended the experiment, with partial success.

For my third experiment I decided on a smaller volume of 2 litres of solution, and rather than subjecting my bench power supply to further inhospitable conditions in my semi-outdoor workspace I made a dedicated power supply, using a 2500mA 5V wall adapter and an adjustable voltage step-down buck convertor which I set to 2.5V. Buck convertors are available very cheaply and essentially perform the function of a bench power supply, although they lack current-limiting. Again using the graphite drawing stick as the anode and a fresh piece of flattened copper pipe cathode, but this time at a wider spacing of around 60mm to avoid dendrites from short-circuiting the cell, I measured the current draw at around 150mA. Again the reaction was slow and very little Chlorine was produced, so I felt confident to leave the setup unattended. I checked after an hour to find a modest layer of Copper deposition on the cathode, which this time was more even in composition, possibly due to the lower current density due to the increased electrode spacing. With the slow production of Chlorine gas and reduced risk of dendrite short-circuiting, I am satisfied that this slow-paced reaction will be relatively safe to leave for an extended period to gradually clean the solution of contaminant metals, so I intend to leave it to run for the next few days, checking it regularly, and will add further reports of the results to this thread.

While I am tempted to purchase a larger graphite anode for a future setup to achieve faster deposition of metals, at this stage I am inclined to think that the difficulty in dealing with the increased rate of Chlorine gas production may outweigh the benefits of faster cleanup of the solution. Since the "AP" process itself is slow-paced, I think a slow cleanup of the solution might be an appropriate match in terms of demand for supervision.

Comments and suggestions are welcomed.
 
Last edited:
While I am tempted to purchase a larger graphite anode for a future setup to achieve faster deposition of metals, at this stage I am inclined to think that the difficulty in dealing with the increased rate of Chlorine gas production may outweigh the benefits of faster cleanup of the solution. Since the "AP" process itself is slow-paced, I think a slow cleanup of the solution might be an appropriate match in terms of demand for supervision.
For a fairly inexpensive graphite anode, look for graphite "gouging rods". Metal workers use them to cut steel. You can usually find them where welding supplies are sold. They come in various diameters. They will be coated with a thin layer of copper that provides conductivity in their normal use. You can peel this thin layer of copper off physically starting at one end. Just peel it off, leaving the bare graphite behind. Be careful, as the graphite isn't overly strong.

Dave
 
For a fairly inexpensive graphite anode, look for graphite "gouging rods".
Proper 10mm electrodes are only a few pounds for half a dozen, I'll probably get some in a week or two to try running in parallel. I'm also considering graphite felt which would have a very high surface area. But for now I'm content to take it slow with the one I've got.

I've seen people on Youtube extract carbon rods from zinc-carbon batteries, which I would have tried if I could find one.
 
Last edited:
Yeah, those type of batteries are few and far between these days. Mainly found in the big, boxy 6V batteries. The downside is they've been immersed in the battery's electrolyte, so when you apply heat (voltage) to them, they start oozing electrolyte that's been absorbed in them. Gouging rods are just clean graphite.

Dave
 
For a fairly inexpensive graphite anode, look for graphite "gouging rods". Metal workers use them to cut steel. You can usually find them where welding supplies are sold. They come in various diameters. They will be coated with a thin layer of copper that provides conductivity in their normal use. You can peel this thin layer of copper off physically starting at one end. Just peel it off, leaving the bare graphite behind. Be careful, as the graphite isn't overly strong.

Dave
Didn’t someone run this experiment last year with a Platinum anode and a carbon cathode ?
 
As Copper Chloride users will know, one of the drawbacks we encounter when releasing gold foils from escrap is that the solution becomes saturated with copper and eventually the etching process will slow down, and white crystals of Copper (I) Chloride will begin to accumulate in at the bottom of the reaction vessel.

Using an air bubbler to replenish the oxygen level in the solution will help to keep the Copper in solution as Copper (II) Chloride, which is what gives it the characteristic green colour. However after prolonged use the solution will turn very dark as it becomes competely saturated and the white crystals will form anyway.

The common approaches to dealing with saturation are either to decant the solution and allow Copper (I) Chloride crystals to form, after which they can be removed by further decanting or by filtering; or alternatively the solution can be elongated by adding more Hydrochloric acid to dissolve the crystals back into solution. Neither of these approaches are entirely satisfactory, since the former produces solution which is still close to saturation, and the latter produces a larger volume of solution, which in time accumulates as an ever-growing volume which consumes more and more Hydrochloric acid at not-insignificant expense.

A further issue is encountered when dealing with iron-bearing escrap, since when Iron is dissolved into the the solution it forms Ferrous Chloride. While Ferrous Chloride is not a major problem in the first instance since it also dissolves copper, it does so at a slower pace, the solution cannot be maintained by bubbling oxygen through it and the Iron does not drop out as chloride crystals as Copper does. Commonly this is addressed by using a magnet to sort Iron-bearing scrap from non-Iron bearing, and using a different method where Iron is present. However other metals may also be present in the scrap material, such as Nickel, Tin, Cobalt, Lead, etc which cannot be sorted with a magnet. Some - such a myself - opt to skip the extra step of sorting with a magnet, instead simply accepting that the solution will become fouled by this type of scrap. Doing so is uneconomical and results in an accumulation of "spent" acids which require treatment before they can be discarded.

Having accumulated a number of bottles of spent solution, and unwilling to discard them, I was interested to read a suggestion from Shark that metals could be removed from the spent solution using electrolysis:

https://goldrefiningforum.com/threa...p-processing-circuit-boards.34718/post-374650

Intrigued, I did some research and discovered that not only copper but also other contaminant metal salts can be removed from solution by forming an electrolytic cell to plate them out as solid metal, leaving behind relatively clean Hydrochloric acid suitable for reuse. This raised the possibility that the Ferric Chloride issue could be effectively addressed.

The reaction which takes place when current is applied reduces the aqueous metal salts (such as Copper Chloride) to metal on the cathode (negative electrode), and releases Chlorine gas at the anode (positive electrode) by oxidation.

Since Chlorine gas is noxious, I decided to perform my first experiment at a very small scale: 150ml of saturated Copper Chloride solution in a beaker, using two graphite pencil leads as electrodes with a spacing of around 15mm. The voltage required is low, since the reduction potential of copper is just 0.34V+, though for practical purposes in a real-life scenario between 2V and 3V is optimal. I used my bench power supply set at 2.5V, and as the reaction started the current draw was around 100mA.

Even at such a small scale, the production of Chlorine gas was noticable, rising as a continuous stream of fine bubbles from the pencil lead anode and giving off a distinctive pungent smell. However, after just a few minutes Copper metal deposits became visible on the cathode, and after 4 hours the solution was noticably clearer and the pencil lead was encrusted with a thick, uneven layer of copper with a remarkable appearance:

View attachment 65966

The deposited copper weighed just over 0.5g.

When I briefly increased the voltage to 5V, the cathode instantly turned black as other contaminant metals such as Iron and Tin were co-deposited along with the copper. They have higher reduction potentials, so at the lower voltage only copper would be deposited until the solution became almost copper-free, after which the other metals will be deposited in ascending order of reduction potential.

Deeming the initial experiment a success, I decided to set up a slightly larger one, using a piece of flattened copper pipe as the cathode and a graphite artist's drawing stick about 8mm in diameter as the anode. I had read that Stainless Steel could be used for the anode but since it may cause Chromium to be introduced into solution I decided against it, since Chromium is toxic.

Again using my bench power supply, I ran 3V through 4 litres of saturated Copper Chloride solution with an electrode spacing of around 20mm, which drew a current of around 1A. I was surprised that very little Chlorine was produced and the deposition of Copper on the cathode was much slower. I put this down to the composition of the graphite stick I used for the anode, since it would contain clay for hardness and strength, which would reduce its current-carrying capability.

After 12 hours the solution appeared slightly clearer and had stratified somewhat, with the clearer part near the surface close to the electrodes while the lower half remained quite dark. The deposits on the cathode were much more uneven, forming "dendrites"- curious-looking growths of metal which stretch out towards the anode. This can be a result of higher current density so I learned that at 3V, 20mm spacing was a litle too close. If I had left the cell to run much longer the dendrites would have made contact with the anode, forming a short-circuit, so I ended the experiment, with partial success.

For my third experiment I decided on a smaller volume of 2 litres of solution, and rather than subjecting my bench power supply to further inhospitable conditions in my semi-outdoor workspace I made a dedicated power supply, using a 2500mA 5V wall adapter and an adjustable voltage step-down buck convertor which I set to 2.5V. Buck convertors are available very cheaply and essentially perform the function of a bench power supply, although they lack current-limiting. Again using the graphite drawing stick as the anode and a fresh piece of flattened copper pipe cathode, but this time at a wider spacing of around 60mm to avoid dendrites from short-circuiting the cell, I measured the current draw at around 150mA. Again the reaction was slow and very little Chlorine was produced, so I felt confident to leave the setup unattended. I checked after an hour to find a modest layer of Copper deposition on the cathode, which this time was more even in composition, possibly due to the lower current density due to the increased electrode spacing. With the slow production of Chlorine gas and reduced risk of dendrite short-circuiting, I am satisfied that this slow-paced reaction will be relatively safe to leave for an extended period to gradually clean the solution of contaminant metals, so I intend to leave it to run for the next few days, checking it regularly, and will add further reports of the results to this thread.

While I am tempted to purchase a larger graphite anode for a future setup to achieve faster deposition of metals, at this stage I am inclined to think that the difficulty in dealing with the increased rate of Chlorine gas production may outweigh the benefits of faster cleanup of the solution. Since the "AP" process itself is slow-paced, I think a slow cleanup of the solution might be an appropriate match in terms of demand for supervision.

Comments and suggestions are welcomed.
The dendrites were Sn maybe?
 
The dendrites were Sn maybe
Just copper I think, they are normally expected when the current density is high, they occurred when I was running it a 3V with the electrodes quite close together. The solution is to separate the electrodes further apart and/or reduce the voltage. Tin can co-deposit with copper at 3V so you might be right that it encouraged the dendrites, but either way just running it slower is the answer.
 
Didn’t someone run this experiment last year with a Platinum anode and a carbon cathode ?
Seems like rather a costly anode... graphite is fine and should last a very long time, and copper is also fine for the cathode as it won't dissolve in the solution when it's under current load. When it's switched off though it should be removed from the solution, and I discovered will then oxidize extremely quickly, turning green-white in air almost immediately. Using old copper pipe is handy as I can just toss it in the metal bin and replace it with a fresh piece rather than cleaning it off. I probably won't even bother hammering it flat next time I change it.
I'd be interested to read the other thread if anyone can link it.
 
Seems like rather a costly anode... graphite is fine and should last a very long time, and copper is also fine for the cathode as it won't dissolve in the solution when it's under current load. When it's switched off though it should be removed from the solution, and I discovered will then oxidize extremely quickly, turning green-white in air almost immediately. Using old copper pipe is handy as I can just toss it in the metal bin and replace it with a fresh piece rather than cleaning it off. I probably won't even bother hammering it flat next time I change it.
I'd be interested to read the other thread if anyone can link it.
I thought the anode was a gram bar and it was non reactive that why it was used .
 
I thought the anode was a gram bar and it was non reactive that why it was used .
Platinum can be used for electrodes but usually only for very high end applications where it's absolutely necessary, it's over $30 a gram and a gram bar is only about 15x10mm and less than 1mm thick; surface area is the main factor for an electrode so graphite rods are most commonly used and widely available. Graphite is also available as solid sheets or felt mats- the felt has the highest surface area but there can be issues with the electrolyte creeping up the mat and eating away at the clamp which holds it.

For anyone wanting to do a cheap mini experiment I would recommend the pencil leads idea, they worked really well. The best way to extract them from pencils is by soaking the pencil in water to make the wood expand and then squashing it gently with pliers to find the join between the two halves. A half-decent electrode could be made by binding a few pencil leads together with copper wire at the top end.
 
The anode bar may have been bigger , I could have sworn it was under a 100$ for that anode ( too rich for my blood ) i started looking for that thread it was a while ago . Still looking , …
 
I have seen the carbon gouging rods used a dozen at a time in a 50 gallon drum. Kept to low power and in a bucket to one side, the copper could be scraped off and collected. I still use gouging rods quite a bit as they are easy to find locally and fairly inexpensive.
 
Update after 24hrs

I had checked the cell earlier in the day and was unimpressed by the rate of copper deposition. On checking the cathode at the 24 hour mark I discovered that there was bare copper showing across about half of it- evidently the solution had been eroding the deposits away faster than they were arriving. I felt this must indicate that the current transfer was not high enough in the solution.

I checked the voltage across the electrodes and got a reading of only 2.25V- I had measured 2.5V at the buck convertor but apparently there is some voltage drop along the cable with the crocodile clips on, which admittedly is unneccessarily long.

I adjusted the buck convertor to give 3V at the crocodile clips.

I also found another graphite art stick which I used to make a second anode to run in parallel. This one was marked "6B", the softest grade, so should be close to pure graphite. Running the pair in the cell side-by-side the difference between them is noticable- the new one produces a fine stream of chlorine gas bubbles, while the old one - presumably a harder grade - produces only a very slight amount of foam around the top. This would appear to support my notion that the clay content was negatively affecting the conductivity. After a few minutes there was a faint whiff of chlorine in the air near the cell, but I judged it not enough to be a major concern in my semi-outdoor workspace.

I will check the cell periodically over the next 6 hours or so. I'm confident that the rate of deposition will be much improved with the new extra cathode and higher voltage.

I will need to improve my arrangement for holding the electrodes in position. Currently they are just hooked over the side of the tupperware box containing the solution and I would be happier if they were held more securely.

The main thing I have learned so far is that the cell needs to be tuned quite carefully- small variations in voltage and electrode position can affect the performance quite a lot. This is easy to do with the bench power supply but a bit more fiddly with the buck convertor since it requires the voltmeter. Also it's clear that pure graphite anode material would be much better than my current makeshift ones.
 
My cathode has developed a fairly advanced case of dendritis this morning :D

dendrites.jpg

I think the dark colour indicates some co-deposition of other metals, I'm still running at 3V so it could be Iron, Tin etc. The solution is still pretty dark. I switched the cathode for a fresh piece of copper.
 
Update at 6 days

Today the solution is quite clear but still has a slight green colour; Chlorine gas production is much less and the cell has ceased to deposit copper. Now the deposits are jet black and powdery, and don't adhere well to the cathode. This means the copper has been almost fully removed and the reaction is now depositing the other contaminant metals such as Iron, Tin, Nickel and Cobalt. I adjusted the voltage down from 3V to 1.8V as there was some bubbling from the cathode indicating hydrogen production. When hydrogen is being produced the acid becomes weaker so it needs to be minimized.
I filtered the solution to remove black sediment, and replaced the cathode for a fresh one. I scraped the copper off the used cathodes and dried it. I was surprised to find that the dry weight of the copper depositions from just 2L of solution was 37.58g, which suggests that the solution held around 20g of copper per litre. However copper chloride can apparently hold up to 750g per litre before reaching saturation in ideal conditions, and in practical terms well over 400g.

copper deposits weighed.jpg

I will leave the cell running at 1.8V to try to remove as much of the other contaminant metals as possible since this is the real objective of the experiment. The solution will need periodical filtering to remove the black sediment in order to avoid redissolution.
 
Last edited:
Update at 6 days

Today the solution is quite clear but still has a slight green colour; Chlorine gas production is much less and the cell has ceased to deposit copper. Now the deposits are jet black and powdery, and don't adhere well to the cathode. This means the copper has been almost fully removed and the reaction is now depositing the other contaminant metals such as Iron, Tin, Nickel and Cobalt. I adjusted the voltage down from 3V to 1.8V as there was some bubbling from the cathode indicating hydrogen production. When hydrogen is being produced the acid becomes weaker so it needs to be minimized.
I filtered the solution to remove black sediment, and replaced the cathode for a fresh one. I scraped the copper off the used cathodes and dried it. I was surprised to find that the dry weight of the copper depositions from just 2L of solution was 37.58g, which suggests that the solution held around 20g of copper per litre. However copper chloride can apparently hold up to 750g per litre before reaching saturation in ideal conditions, and in practical terms well over 400g.

View attachment 66008

I will leave the cell running at 1.8V to try to remove as much of the other contaminant metals as possible since this is the real objective of the experiment. The solution will need periodical filtering to remove the black sediment in order to avoid redissolution.
That's some nice copper powder to strain wet with 400 mesh and use to clean the stockpot or if you clean it up with some waste sulfuric to clean other base metals out and use to precipitate gold.
 
That's some nice copper powder to strain wet with 400 mesh and use to clean the stockpot or if you clean it up with some waste sulfuric to clean other base metals out and use to precipitate gold.
Yes I was thinking about using it to precipitate gold, though it would need purifying further, I will see what can be done with sulfuric- I don't produce waste sulfuric at the moment but I have it fresh, presumably it could be quite dilute for this?
I have seen @kurtak 's posts regarding copper powder as a precipitant and the method of stopping short of a complete drop to achieve good purity. I would definitely like to try it.
I think if I used the lower voltage from the start, the copper would already be quite pure as there would be fewer co-deposits. It seems like 1.8-2V is the sweet spot, initially depositing copper slowly on its own, and then when the copper is out, depositing the other metals with minimal hydrogen production.
 
Last edited:
presumably it could be quite dilute for this?
Yes. It will be slower but will remove most of the more reactive metals.
I first strain a bit to fine powder and treat that with sulfuric. Then rinse clean until all salts and acid is out. Store in distilled water. And pippette it out when needed.
I have seen @kurtak 's posts regarding copper powder as a precipitant and the method of stopping short of a complete drop to achieve good purity. I would definitely like to try it.
His posts made me try it and if you take it slow at 50 degrees C and keep a bit of gold in solution, you can get great results. Then I use SMB for the last bit.
And it works good to get free nitric out before precipitating with SMB, when you see brown fumes after adding SMB, switching to copper saves on SMB.

Great thread by the way, I have over 30L of copper chloride to clean up and don't want to treat it as waste but recover the copper and the HCl.
Some safety to keep in mind with the chlorine gas if i do it.
 
Yes. It will be slower but will remove most of the more reactive metals.
I first strain a bit to fine powder and treat that with sulfuric. Then rinse clean until all salts and acid is out. Store in distilled water. And pippette it out when needed.

His posts made me try it and if you take it slow at 50 degrees C and keep a bit of gold in solution, you can get great results. Then I use SMB for the last bit.
And it works good to get free nitric out before precipitating with SMB, when you see brown fumes after adding SMB, switching to copper saves on SMB.

Great thread by the way, I have over 30L of copper chloride to clean up and don't want to treat it as waste but recover the copper and the HCl.
Some safety to keep in mind with the chlorine gas if i do it.
You can also dilute the solution with water and then set up a self-shorting 'battery' setup like I do, with an iron bar and a copper rod connected outside the solution with a copper wire. Copper will cement out on BOTH sides, and often form quite lovely copper 'feathers' and branching clumps of bright crystals, meaning they already have very high purity.

Then you're left with a solution of mostly iron chloride and whatever other more reactive base metals were already in solution. At that point, I usually just neutralize the solution and let it dehydrate down to salts and concentrated metal brine either in my greenhouse with the heater blowing onto it; or over my brick furnace on 'low' heat (just adding enough wood to keep the embers going to warm the iron top plate).

Gotta neutralize before drying, because acidic chlorides will NEVER dry. They're so hygroscopic that in humid weather they actually INCREASE in volume as they pull huge amounts of water from the air.
 

Latest posts

Back
Top