Sciencemadness Discussion Board - Catalytic Converter PGM Recovery

Catalytic Converter PGM Recovery

MrHomeScientist - 21-8-2013 at 05:41

Because the thread that I posted this comment in was locked (rather inexplicably, which has become a disturbing trend lately), I am reposting my information and questions here. I was hoping to get some comments on the method, answers to my bolded questions, and when the experiment gets underway post my results (which people can't do now that the thread is locked, so please do not consider this a double post).

====================================

I might be getting my hands on a catalytic converter, and was going to try my hand at platinum group metal (PGM) recovery. I found a nice "beginner's guide" that I was going to base my procedure on (attached below). What it boils down to is this:

1) Grind catalytic converter material down to a coarse powder.
2) Incinerate material in a furnace to burn off road dirt and organic residues (if it isn't a brand new converter).
3) Treat material with HCl, followed by HNO3 (separately!). This removes any base metals (this step might be unnecessary).
4) Treat material with aqua regia (3:1 HCl:HNO3) to dissolve platinum and palladium. Rhodium remains insoluble. (how would I separate this from undissolved ceramic?)
5) Precipitate platinum from solution as the bright yellow ammonium hexachloroplatinate, using a saturated solution of NH4Cl.
6) Precipitate palladium from solution as the bright red ammonium hexachloropalladate, by adding a saturated solution of KClO3 (my source recommends sodium chlorate, but I have the potassium salt and I can't imagine that would make a difference). Technially, it is the chlorine produced that causes this compound to form.
7) Calcine both the PGM salts in ceramic dishes to reduce them to their metals - driving off ammonium chloride. Ramp up the heat slowly to avoid vaporizing any metals. (although, doesn't this step conflict with (2), above? If we want to heat slowly to avoid vaporizing metals now, wouldn't incinerating the honeycomb earlier be doing some vaporizing?)

Here is another guide that has the same overall process, but differs a bit on terminology. e.g. it calls the brick red precipitate ammonium hexachloropalladate rather than a chlorate salt like the attachment does. Theirs seems to make more sense, so I went with that name above.

Any comments on this method? Suggestions, things to avoid, etc.? I have no intention of making money off of this, I simply want to explore the process and get some palladium and rhodium for my element collection!

Attachment: PGM Recovery from Catalytic Converters.pdf (282kB)
This file has been downloaded 19482 times

blogfast25 - 22-8-2013 at 04:58

Nice finds, MrHS. I think you've got everything you need there for a decent first run. Getting hold of an old catalytic converter may not be so easy though.

BleepingAmazing will probably think all this is not 'alchemical' enough and will want to 'invent' his own process, possibly one that involves getting splattered with steam/H2SO4. Each to his own, I guess!

[Edited on 22-8-2013 by blogfast25]

froot - 22-8-2013 at 05:33

Quote:

4) Treat material with aqua regia (3:1 HCl:HNO3) to dissolve platinum and palladium. Rhodium remains insoluble. (how would I separate this from undissolved ceramic?)

I would try NaOH, molten or hot solution.

Quote:

7) Calcine both the PGM salts in ceramic dishes to reduce them to their metals - driving off ammonium chloride. Ramp up the heat slowly to avoid vaporizing any metals. (although, doesn't this step conflict with (2), above? If we want to heat slowly to avoid vaporizing metals now, wouldn't incinerating the honeycomb earlier be doing some vaporizing?)

I would imagine step 2 involves temps up to 500 degC in air while step 7 involves much higher temps. If you have a flame directed at the metal and the colour of the flame changes, chances are the metal is slowly vaporising via the flame and in this case no flux is used to protect the values. The slower velocity the flame the better. If it's done in an oven, even better.

Remember that these cat converters have been subjected to hot high velocity exhaust gases for hundreds of hours so I wouldn't stress too much about loss by vaporisation.

Edit: found some info here: http://metals.about.com/od/properties/a/Metal-Profile-Rhodiu...

Quote:

The first step in extracting rhodium from the ore, is precipitating precious metals such as gold, silver, palladium and platinum. The remaining ore is treated with sodium bisulphate NaHSO4 and melted, resulting in rhodium (III) sulfate, Rh2(SO4)3.

Rhodium hydroxide is then precipitated out using sodium hydroxide, while hydrochloric acid is added to produce H3RhCl6. This compound is treated with ammonium chloride and sodium nitrite to from a precipitate of rhodium.

The precipitate is dissolved in hydrochloric acid and the solution is heated until residual contaminants are burnt off, leaving behind the pure rhodium metal.

Maybe you could apply this to your process.

[Edited on 22-8-2013 by froot]

Fleaker - 22-8-2013 at 06:58

Gross oversimplification and you're liable to get in trouble with catalyst.

Leaching PGMs out of a honeycomb should be done as a hobbyist endeavor--don't think you can get the recovery or the value to make it compare with arc melting the catalyst with a collector metal ( BASF is one company doing this, in Seneca, SC).

) Grind catalytic converter material down to a coarse powder.
Not necessary, as most of the PGM is in the wash coat. Increasing surface area of the material by grinding increases solution holdup. The material should be crushed, ground, and sieved only if it's being prepared for sampling and sending out to the smelter.
2) Incinerate material in a furnace to burn off road dirt and organic residues (if it isn't a brand new converter).
Incineration is always a good step but must be done at the right temperature; otherwise refractory PdO and Rh2O3 form that do not dissolve well in an aqueous leach.
3) Treat material with HCl, followed by HNO3 (separately!). This removes any base metals (this step might be unnecessary).
It is unnecessary. You will not have enough of a reduction of base metal contamination--all this accomplishes is making a base metal solution that may have Pd in it (Pd will dissolve in HCl, if the Pd is fine enough and the HCl is concentrated and air is let in).
4) Treat material with aqua regia (3:1 HCl:HNO3) to dissolve platinum and palladium. Rhodium remains insoluble. (how would I separate this from undissolved ceramic?)
Aqua regia is a bad lixiviant for this material. Better to use 6 M HCl and a chlorine source. Rhodium doesn't remain insoluble. With the right chlorination conditions, over 90%+ of the Rh goes in with the Pd and Pt.
5) Precipitate platinum from solution as the bright yellow ammonium hexachloroplatinate, using a saturated solution of NH4Cl.
Well, given the solubility curve for ammonium hexachloroplatinate, what will probably happen is nothing. Unless you concentrate your solution (think about the mass of catalyst to lixiviant ratio), all you are doing is making an ammonium-containing liquor that's expensive to dispose of via normal waste treatment. Better to use borohydride, or perhaps aluminum powder as the reductant.
6) Precipitate palladium from solution as the bright red ammonium hexachloropalladate, by adding a saturated solution of KClO3 (my source recommends sodium chlorate, but I have the potassium salt and I can't imagine that would make a difference). Technially, it is the chlorine produced that causes this compound to form.
Again, forget about this step from dilute solutions. There's really nothing quantitative about PGM separations via classical hydrometallurgy.
7) Calcine both the PGM salts in ceramic dishes to reduce them to their metals - driving off ammonium chloride. Ramp up the heat slowly to avoid vaporizing any metals. (although, doesn't this step conflict with (2), above? If we want to heat slowly to avoid vaporizing metals now, wouldn't incinerating the honeycomb earlier be doing some vaporizing?)
This is what you do AFTER the material is refined. For the purposes of the amateur enthusiast, this shouldn't be done; you can clean up the Pt/Rhwell enough, and the Pd is easily refined. In grand scale, this is the best way to get PGM sponges.

You don't want to use molten NaOH. Dangerous as all heck, it dissolves the PGMs.

As for the rhodium refining sequence described...it's a lot more involved than that. It's also not practical to use bisulfate unless at the lab scale. The amounts of SO3 produced are very dangerous.

I think your goal should be making a cemented product of Pd/Pt/Rh. Then you can try separating them.

Separation of Pd/Pt/Rh is not easy but there are numerous ways to do it and avoid using ammonium chloride.

MrHomeScientist - 22-8-2013 at 09:34

Awesome information, thanks for the posts. That gives me a lot to think about. Fleaker, with your valuable metals experience I was hoping you'd respond

This is most certainly a hobbiest endeavor, I have no delusions about making my fortune doing this. I simply wanted to have fun with the process and recover some elements for my collection.

Some further questions on your comments:

1) Makes sense. The values are all on the surface anyway.
2) What temperature would be too high? I was probably going to put the material in a crucuble in my aluminum-melting furnace (currently charcoal powered, but I'm working on a propane version). So it wouldn't be directly exposed to flame. It looks like PdO forms at 350 C though, so this might be excessive heat.
3) Agreed.
4) Why is aqua regia bad in this case? Because it does not dissolve Rh? HCl + Cl₂ should be simple enough to implement - I have a gas washing bottle that I could charge with catalyst, fill with HCl, and bubble chlorine in from the bottom (although I've never been able to use this without causing suckback!).
5) & 6) I had planned on concentrating the solutions, based on my other reading, but had neglected to include that step here.
7) I didn't really understand your points here. Since this was how I was planning on getting my metal powders, how would further heating after this final refining step help?

Cementing out all the PGMs at once would certainly be simpler, but I wanted to separate them as much as possible. Wouldn't separating a mix of metal powders just be the process I outlined above? That's what the honeycomb is coated with in the first place, after all. I can't think of any physical methods of separation, but then again I hadn't researched that avenue. Why should ammonium chloride be avoided?

[Edited on 8-22-2013 by MrHomeScientist]

Fleaker - 22-8-2013 at 18:38

You're thinking right. PdO is what you do not want to form. Now comes the issue of "but how do I de-coke/carburise the catalyst at a measly 350*C?". Good point. Heat it too hot and you have PdO that's not well soluble in acids. Heat it not enough, you have fine carbon black soot that holds value by adsorption. What to do? Bear in mind that most petrol engines are tuned to run rich so that the Pd isn't really oxidized in the converter itself.

Aqua regia is bad because it's bad. The only time aqua regia is indicated for precious metals refining is when you have bulk pieces of platinum, palladium or gold--then you can get a better dissolution time. For fine powders, HCl/Cl2 system is great. Bromine-HBr systems work well too (because you can reflux the bromine and then then sparge with chlorine to rid the system of bromine for re-use).

Also, excess nitric acid means much of your platinum will be dissolved as its nitrosyl complex. This makes cementation exponentially more difficult (can you imagine all the NOx?). It's also expensive, hazardous, and completely unnecessary. Aqua regia does indeed dissolve rhodium, just not very fast. But not much does dissolve rhodium (or iridium) very expediently.

Don't bother bubbling the chlorine unless you really are an ardent admirer of Len1's prepub work--just generate it in situ with hydrogen peroxide. You won't need much and you can conduct this at room temp.

Ammonium chloride should be avoided because of this situation:

1. Waste treatment. Makes ammonia which chelates any other free metals. We've eliminated it as much as possible.
2. Practicality: in working with PGMs, generally, you want the solution as concentrated as possible if using conventional hydrometallurgy. Figure it takes two liters of HCl-Cl2 mix to do a converter. You now have, at best, a 2 g/L Pt solution, 1-1.5 g/L Pd, and perhaps 0.5 g/L Rh. Add ammonium chloride and see what happens. Probably not much of anything! You'd have to concentrate down a lot of solution.
Best to work on the small scale, dissolve all the cemented powders up, and then go about it classically. You do know how to do that, don't you? Brauer is a good start and you can always hit the goldrefiningforum if you need a picture tutorial.

bfesser - 22-8-2013 at 19:22

I just want to say that I'm glad to see discussion on this topic has resumed, and in a reasonable and intelligent manner. I really disliked having to close the other thread.

blogfast25 - 23-8-2013 at 08:29

Fleaker:

You are w/o a shadow of a doubt the resident expert on PGM recovery. Here's a simple question: can amateur recover of PGMs from a single catconverter even be done in a break-even way? Taking into account not labour but all chemical/energy costs? Or should it be seen as a chemical adventure with possibly a very small mixed PGM nugget at the end of the rainbow?

MrHomeScientist - 23-8-2013 at 13:02

After Fleaker's advice on using the HCl/H₂O₂ system, I did some more searching and found this paper:

Attachment: PGM Recovery Rumpold.pdf (960kB)
This file has been downloaded 1020 times

It's a very interesting read. In it, they lay out the equations for PGM dissolution:

Pt(s) + 8H⁺ + 4O^- + 6Cl^- → [PtCl₆]^2- + 4H₂O(l)
Pd(s) + 4H⁺ + 2O^- + 4Cl^- → [PdCl₄]^2- + 2H₂O(l)
2Rh(s) + 12H⁺ + 6O^- + 12Cl^- → 2[RhCl₆]^3- + 6H₂O(l)

At another site, I also found reactions pertaining to the acid/peroxide system:
H₂O₂ + 2H⁺ + 2Cl^- → Cl₂ + 2H₂O
H₂O₂ + Cl₂ → O₂ + 2H⁺ + 2Cl^-
(phew, that was a lot of HTML tags

)

They also talk about cerium recovery, which is something I hadn't seen before and I'll have to try as well since I don't have that one in the collection either. They add sulfuric acid to their solution of hydrochloric/peroxide "to compensate for the hydrogen consumption." They then state:

Quote:

The experiments were carried out with 150 g of catalyst and 570 ml of synthetic solution. To enhance the leaching conditions, the temperature was raised to 80°C. The addition of the hydrogen peroxide was made very slowly to avoid extreme bubble formation.

They do not define "synthetic solution," but I presume this to be the hydrochloric/sulfuric acid mix, to which they slowly add peroxide. They do not state any amounts or concentrations besides these numbers, so I'm not sure where to go as far as practial application (what % peroxide? how much to add? etc.). I should be able to use the above equations as a starting point, although I've never seen the use of an oxygen ion before (I assume this comes from the peroxide).

This seems very difficult to know how much to use for all the reagents regardless, because the peroxide is decomposed by the chloride ions, road grime side reactions, unknown amount of PGM material, etc. Is there any clear indication, like "solution becomes colorless", "bubbling stops", etc.?

Fleaker, no I have to admit I don't really know the "classical" method. I've never done these sorts of reactions before. I will consult Brauer and the gold refining forum as you suggested; I've ran across the latter a few times in my research already. This is all very exciting! There's some very rich chemistry here (pun intended)!

blogfast25 - 23-8-2013 at 15:33

Mr HS:

The first three equations don't make any sense at all: 4 H⁺ + 2 O^2- for instance really means 2 H₂O! "O^2-" can't exist in aqueous solution anyway: as an "acid" water is far too weak for a second deprotonation step.

More importantly, the equations don't account for the oxidations of the metals AT ALL!

But the file is interesting in many other respects.

Re. Ce, CeO₂, even reagent grade, is not expensive. But turning it into metal is quite another thing. Assuming the lixiviate contains CeCl₃ (Ce IV oxidises Cl^- to Cl₂), precipitation as the double salt with K₂SO₄ should be near quantitative by saturating the solution with that salt, after removal of the PGMs.

Edit:

The first equation, assuming chlorine was the oxidiser in acid (HCl) conditions, would simply be:

Pt (s) + 2 Cl₂ (aq) + 2 HCl (aq) === > H₂PtCl₆ (aq)

... although it would mainly serve stoichiometrical purposes.

[Edited on 24-8-2013 by blogfast25]

MrHomeScientist - 26-9-2013 at 18:59

Update
I have my donated catalytic converter now, and even found someone to slice it open for me!

Looks like it was from a VW, with what I assume is a part number stamped on it. It doesn't look all that dirty inside either, which is nice. Once I get some time (so many things vying for my attention now), I'll proceed with experimenting! I want to try both aqua regia and HCl/Cl₂ on small samples and see how each goes.

Magpie - 27-9-2013 at 10:50

At some point I would like to experiment with catalytic hydrogenations. I know that some of this can be done with Pd/C powders, like hydrogenation of unsaturated fatty acids. Would the catalyst from auto exhaust catalytic converters (CC's) be suitable for this and/or other hydrogenations?

Also I heard from one auto mechanic (possibly a crank) who said that the recycling of used CC's is tightly controlled by the government with heavy fines for diversion out of the recycling system. Is this true?

elementcollector1 - 27-9-2013 at 11:17

Quote: Originally posted by Magpie

Also I heard from one auto mechanic (possibly a crank) who said that the recycling of used CC's is tightly controlled by the government with heavy fines for diversion out of the recycling system. Is this true?

Probably the same people who regulate melting the zinc out of a penny.

violet sin - 27-9-2013 at 13:25

with regards to the recycling of cats thing. (purely speculation) but the insurance company may want a FULL return on a wrecked car. meaning diversion for other uses amounts to theft in their eyes. but I do know for certain ( mechanics I have asked ) that some people try to buy a used one to put BACK on their car at a lower cost. guy told me I can't buy one from some one and have it put on my car, it has to be new. I just wanted to know about getting some material for this process and he looked at me like I was a few min away from doing some crystal meth and coming back with a freshly hacksawed CAT for him weld to my car( which didn't need a new one). so I did a little reading and it seems mainly environmental protection laws about certified used and aftermarket replacements.

http://davisconverters.com/statelaws/California-State-Laws.s...
"Beginning July 10, 2008, no used converter can be legally advertised for sale, sold as replacement, or installed in California after this date."

http://www.dec.ny.gov/chemical/87411.html
"A used catalytic converter is not a new aftermarket catalytic converter, or a replacement part as defined below. This prohibition does not apply to the sale of used catalytic converters for recycling purposes."

the last reference makes it pretty clear(in new york), for cash = no prob, for USE = illegal.
-Violet Sin-

Magpie - 27-9-2013 at 18:49

I have not done any research on this myself but what you are saying Violet Sin makes sense, ie, that there are laws against reuse of a used CC on an automobile. Another mechanic was showing me a barrel of used auto parts he was going to sell as scrap metal. Sitting in this barrel (outside the shop) was a used CC.

Traveller - 27-9-2013 at 20:31

It was a more lucrative business when Rhodium was up at $10,000 per Troy ounce. The crash in 2008 knocked the price way down and it has never recovered. It got back up to $3000/oz. in 2010 for a while but for a long time now it has been on a steady decline from there and is now close to $1000/oz. Rumour has it that new technology in CC's has cut down drastically the requirement for Rhodium, while requiring more Palladium.

Update 2

MrHomeScientist - 14-10-2013 at 07:36

To do some initial testing, I broke off a few small pieces of the converter material, totaling 1.9g, and placed these in a test tube. I added 20mL of 9.5M hydrochloric acid, and it actually started reacting with the honeycomb material! It formed lots of bubbles and turned the solution almost completely black. From smell it seemed like the gas was chlorine, but I could have mixed that up with HCl vapors and I didn't try any definitive testing to determine its identity. After leaving it to react for two days, I added a roughly equal amount of water to dilute it and discovered the solution was actually a deep green.

Surprising, since I thought the honeycomb would be unreactive ceramic. Could the color come from nickel, perhaps?

I suppose the next step would be to filter off the solids and treat these with nitric acid to see if there is any reaction there. Then, aqua regia to get the PGMs in solution. I know aqua regia was frowned upon earlier in the thread, but I wanted to do some small scale testing to see if I could recover anything at all.

elementcollector1 - 14-10-2013 at 08:13

Quote: Originally posted by MrHomeScientist

To do some initial testing, I broke off a few small pieces of the converter material, totaling 1.9g, and placed these in a test tube. I added 20mL of 9.5M hydrochloric acid, and it actually started reacting with the honeycomb material! It formed lots of bubbles and turned the solution almost completely black. From smell it seemed like the gas was chlorine, but I could have mixed that up with HCl vapors and I didn't try any definitive testing to determine its identity. After leaving it to react for two days, I added a roughly equal amount of water to dilute it and discovered the solution was actually a deep green.

Surprising, since I thought the honeycomb would be unreactive ceramic. Could the color come from nickel, perhaps?

I suppose the next step would be to filter off the solids and treat these with nitric acid to see if there is any reaction there. Then, aqua regia to get the PGMs in solution. I know aqua regia was frowned upon earlier in the thread, but I wanted to do some small scale testing to see if I could recover anything at all.

That green actually reminds me of Cr(III) - see if you can add some NaOH and NaOCl and get a yellow solution.

plante1999 - 14-10-2013 at 09:29

I don't find it really surprizing as your honeycombs doesn't look like the ceramic one. They look more like the metal varient, which I forgot the composition.

blogfast25 - 14-10-2013 at 12:29

Quote: Originally posted by elementcollector1

That green actually reminds me of Cr(III) - see if you can add some NaOH and NaOCl and get a yellow solution.

It could also be plain Fe²⁺. With NaOH and NaOCl that would give red/brownish Fe(OH)3 precipitation.

If no precipitate forms and you get a yellow solution, add some lead nitrate (or acetate), to form definitive PbCrO4, a yellow precipitate that's hard to mistake for anything else (in this context).

Quote: Originally posted by plante1999

I don't find it really surprizing as your honeycombs doesn't look like the ceramic one. They look more like the metal varient, which I forgot the composition.

Revisiting that photo, that cross section looks very metallic, indeed. If so, you'll need a ton of acid to dissolve that away!

[Edited on 14-10-2013 by blogfast25]

bismuthate - 14-10-2013 at 12:50

that looks to me like a concentrated CuCl solution , but HCl will only attack copper in the presence of an oxidiser so that should be impossible. Add a solution of baking soda and you should be able to identify the percipitate by color.

MrHomeScientist - 14-10-2013 at 17:48

Wow lots of ideas! And all of them doable with my current stock of chemicals. I'll try and do some testing at some point this week.

Oscilllator - 14-10-2013 at 22:59

I seriously doubt that the green colour is nickel. I have previously tried to dissolve nickel in concentrated hydrochloric acid, and the reaction proceeds at an extremely slow rate unless an oxidiser (I used H2O2) is present. Also, the deep green colour of the solution would mean the nickel would have to be fairly concentrated, and since you only put 1.9g in this test tube, this seems highly unlikely.

froot - 14-10-2013 at 23:56

Did you check the material's magnetic properties?

MrHomeScientist - 16-10-2013 at 18:35

I did some testing with the green solution.

I took 1mL and placed this in a test tube, then diluted with 1mL of water. In another test tube I prepared a solution of several mL of bleach and about 1g NaOH (I didn't make precise measurements).

Combining the two yielded no change for the first few mL (as the solution was neutralized), but then sudden precipitation of a brown goo that I've seen many times before - iron hydroxide. This settled after several minutes, leaving a light brown solution.

I also tried out adding hydrogen peroxide (3%) to another milliliter of converter solution to try that route to convert it to brown iron(III). There was lots of bubbling, but no change in color beyond lightening from the dilution. That tube is on the right in the picture of the settled precipitate.

Also, the raw converter material is indeed magnetic, as you'd expect from the results pointing to iron above.

Finally, I tried adding a solution of a tiny amount of lead nitrate to a sample of the brown solution from picture 3 above. No change at first, but after about 5 minutes I did see a small amount of a precipitate.

So this seems to confirm that my metallic-looking converter substrate is iron with at least a chromium impurity.

blogfast25 - 17-10-2013 at 04:44

Quote: Originally posted by MrHomeScientist

So this seems to confirm that my metallic-looking converter substrate is iron with at least a chromium impurity.

If there really is chromium in there it's probably deliberate, as it increases corrosion resistance of steel a lot: see stainless steel. Good work.

Personally, I would repeat the test for lead chromate by carefully filtering the supernatant solution of that tube on the right. Check its colour. Then add quite a bit of lead nitrate solution. Right now your result isn't 100 % conclusive. Fe(OH)3 has a tendency to peptise.

[Edited on 17-10-2013 by blogfast25]

Gometal - 4-12-2013 at 20:06

The catalytic converter you are trying to refine is called a stainless. The material in side of the converter is called a stainless wrap. Not very common converter or material used in converters. The best way to extract the material is to pick a part the stainless and seperate the powder from the stainless. We yielded 24.6 grams of palladium at a 98% purity on a converter called the Torpedo with a street value of $180.00 give or take depending on who you sell to. There are many converters that yield 100% profits margins the hard part is finding them. We used the same process Fleaker has described previously. You will lose on overall yield but Make up the difference in a higher % of spot market price.

Magpie - 30-8-2016 at 18:43

I bought a used propane catalytic heater at a garage sale a few years ago with the idea that I could recover the PGM from it. A picture of the catalyst is shown below. It is ~0.56 ft2 x 1/2" thick.

Does anybody have any idea what PGM it contains and what amount in grams it is likely to contain?

Jimmymajesty - 2-6-2019 at 13:20

I have <0.3wt% H2PdCl6 in AlCl3/Hcl pH~0.1 solution. I wanted to recover the Pd from a catalytic converter.
This can not be filtered as it is a crystalline mass at room temp, when hot the colloid SiO2 particles clog the filter in no time or it may be too thick even then I am not sure.
I tried to cement out the Pd with Al rods by heating it up till all the AlCl3 went into solution => a black colloidal Pd was obtained that does not settle, goes trough filter paper as well.
Redissolved the solution with Cl2 gas then tried to cement out with copper, I got the same colloid.
Now... what the hell? Is this supposed to be this difficult? This PGM recovery career of mine together with this solution will end up in the sink at this rate pretty soon.

BTW Magpie have you got the answer to you question?