AgPd contacts refining

[orvi]

As the material I recieved few days ago is very nice and illustrative, I decided to share my refining venture of the lot.

85,5 grams of AgPd contact cuttings, on ferrous carrier (altough looks like copper, it is only copper plated ferrous alloy).
Beads itself are 29-30% Pd by weight. So I expect around 15-18g of Pd in this batch.

I started by dissolving the ferrous carriers in dilute nitric acid (ca 25%) - great caution should be taken during this step as reaction is very exothermic and tend to runaway and foam a lot. When cold, nothing spectacular happen, but after the solution warm itself due to heat of the reaction, it goes exponential - to the boiling point in less than 2 minutes in my case
I was aware of this, so I prepared large 2L beaker for the reaction of only around 20-30g of iron. When reaction ceases, you are left with ferrous colloid goo, which is then diluted with cold distilled water and concentrated sulfuric acid is cautiously added in small portions with stirring until it gets liquified and vast majority of iron hydroxy-nitrates goes into the solution as ferric sulfate.

If you don´t overadd nitric, you are left with just beads of the AgPd alloy, and only trace AgPd goes into the solution - you conveniently scavenge traces of Pd with DMG later, and if you are after tidbits of silver, you can add NaCl solution and let the solution settle for few days and harvest the AgCl. Simply decant the iron solution from the beads and you done majority of the cleaning process from base metals

Be aware of the fact that nitric undergo deep reduction when reacting with iron, so mainly NO is produced as reaction byproduct, and you exchange nitrates to sulfates when you add sulfuric acid - reviving the free nitric, which is then able to react. So overall, very little nitric is consumed, if done cautiously and properly - that way nitric is only used as oxidant, and all iron and other base metals are left in solution as sulfates.
In my case, I just used straight dilute nitric at the start, because ammount of material was small and I did not have time to play around with sulfuric aditions to save around 20 mL of nitric

Then, beads are dissolved in 50% nitric. Unless you do not heat the solution strongly, the reaction is very slow.
After full dissolution, I diluted the brownish red solution (I love that colour ) with distilled water and poured the AgPd solution slowly into vigorously stirred ca 15% HCL. You need to use ammount that conveniently converts AgNO3 to AgCl and also Pd(NO3)2 to H2PdCl4, and you should be left with excess HCL in solution - otherwise bigger ammount of not very soluble PdCl2 form, which stay locked in the AgCl and it is quite a pain to macerate it out of the precipitate.

Then, AgCl was filtered on fritted glass filter, sucked from majority of the reddish brown liquor, placed in the beaker again and covered with ca 5% HCL to start washing the precipitate from locked Pd. I gave this procedure 3 rounds, and in the last wash, I obtained only slightly orange filtrate, and tan coloured AgCl (obviously some Pd remained in it). It need to be said that this procedure is nowhere 100% efficient, but it is quick for obtaining bulk of the Pd in more than 95% purity (which is sufficient to me to sell it).
There are better ways how to resolve these two elements, but I wont discuss them here now.

Tomorrow, I will proceed to formate reduction of the Pd in the solution


Material on the left, AgPd beads with majority of the iron carriers gone on the right.


Dissolution of the beads on the left, and filtration of pay-juice on the right


In the end of the day, nice to see that colour again we will see how moch we could squeeze out of it.

 

[orvi]

So, I proceeded to formate reduction. Due to foaming during the process, and increased volume due to neutralization of excess acid with NaOH, I used 3L beaker as reaction vessel.

To the nice brownish-red juice (lot of free acid from washing the AgCl), I added around 30 ml of formic acid - which was quite an excess, and I will discuss later why maybe a bit dangerous.

Then, pH was adjusted with solid NaOH to the value of 2. Simultaneously, as excess of HCL was neutralized, solution heated up by the reaction heat, and comfortably reached over 70 °C.
When I was close to the ideal pH range, every addition of NaOH created small burst of bubbles. Not sure if the reduction begins and quickly ceased - or it was just NOx gasses coming out of the reaction.
Anyway, when the proper pH is reached, solution start to emmit quite a bit of gas - CO2, and minor evolution of also some NOx gasses (judging by brownish colour). I heated the reaction to nearly boiling point (and simultaneously it was stirred with magnetic stirbar) and hold this temperature until all Pd dropped properly. It took somwhere like 2-3 hours. It gradually changed colour from deep brownish-red to very faint reddish colour and in the end it stayed pale brown - just the leftover bit of ferric ions. Stannous negative (after re-acidifying the drop of juice, of course), DMG negative.
On the edges of beaker, some Fe oxohydroxides "rusted"out, and some little floaty "rusty" layer formed - due to higher pH of the solution, than iron can withstand
In the end, I left the suspension of Pd sponge to cool. And strange observation happened.

When it cooled to like 50 °C, all the Pd suddently during few minutes flocculated and spontaneous bursts of gasses started to be emmited from the suspension. It was going like a lava lamp - heavy Pd precipitate bursted upwards, follow by stream of bubbles... Altough in fume hood, I was able to smell faint ammoniacal smell, not quite ammonia, but more like old dimethylformamide or formamide. And as it gradually proceeded in bigger bursts, I decided to quickly filter the solution, which was apparently reacting on the catalytic surface of palladium.
It is well known that formic acid together with palladium catalyst can reduce variety of organic and inorganic molecules... And there are also nitrates present in the solution, which theoretically could be reduced to ammonia. This combined with formates can somehow catalytically form formamide (maybe ?). Long story short, something was clearly happening... And also knowing the fact that formic acid could be reduced in some conditions to CO, not CO2, I quickly ended up the reduction party

After filtration, I washed the precipitate with water (I would maybe do better washing it with few % HCL) and dried it in the beaker on a hotplate.

Result is 18,67 grams of black Pd powder, which is surprisingly clean (by XRF measurement of the powder). I am bit curious how no silver ended up in my Pd... Maybe our XRF need a little callibration
We will see after melting How much of the weight comprises of salts and other junk, which XRF can´t read.


Onset of the reduction on the left and Pd powder swirling in the beaker on the right


Endpoint of the reaction with rusty goo on the edges and Pd "snowfall" in the beaker

End result and XRF reading of the powder - strangely pure, we will see definitely after melting

Both waste liquids treated with DMG and after stirring for few hours passed through the same filter - practically zero Pd (white DMG without any signs of yellow) was in both first nitric leach of iron carriers and also leftover solution after HCOONa reduction.

 

[orvi]

So in the end I have it all melted down After problems with oxy/propane and subsequently with induction furnance, we have finally end result. 17,50g Pd button, XRF purity 98%, as one incident happened before with it... Which was very frustrating back at time when it happened.

I took small sample to assign melting loss. Around 1,5g of powder, into the smallest quartz dish, and melting it down with oxy/propane just to see the weight at the end. But unfortunately, power of the oxypropane we have is limited by 5L oxygen concentrator, so it took ages to heat the dish to temperature. Finally I was able to melt it for few seconds and then BOOM... Tip of my burner flashbacked and completely exploded brass all over the dish, me and the catchpan. I was not expecting this to happen, but it was clear that it can. Nozzle was quite wide in diameter, because it was drilled, and red-heated brass has no structural strength at all.
So I obtained button heavier than powder I started with, and purity went down from 99,4 to 85 %. So this weren´t very helpful at a time, but told me no silver is present.

 

[FrugalRefiner]

Nice looking button. I hope you're OK.

Dave

 

[Alondro]

Nice detailed explanation of the entire process.

Found something interesting: Hydrothermal Solubility of Palladium In warm solutions with high NaCl concentration at a pH of 6.5 under high pressure, palladium was very soluble. There was also an observation for very low pH (about 1.5) at a low NaCl concentration that Pd also became very soluble. TOO hot, however, and the solubility plunged.

I wonder if the best conditions for separating Ag and Pd in solution might be a combination of high NaCl concentration plus HCl. Though, since this experiment took place at high pressures, this likely alters the chemistry.

At any rate, the study identifies new potential sources for PGEs: modestly warm hydrothermal brine deposits. Super-salty hydrothermal vents with internal temps between 300-700C could carry a huge amount of PGEs, depositing them in minerals as the steam bursts out and all the minerals precipitate.

I'm also curious about using phosphate to separate Ag and Pd. Silver phosphate is virtually insoluble in water (only 0.00065g/L), but very soluble in aqueous ammonia.

I can't find solubility data on palladium phosphate, though. If there's a solubility split there, say, either palladium phosphate is soluble in water or insoluble in aqueous ammonia, then there's a potentially easy and cheap way to separate the metals.

I'd have to look up the method for reducing them back to the metals. There is a biological staining method which uses silver phosphate reduction to silver metal to label certain tissues, so a method does exist already. I just need to find it. Silver phosphate melts without decomposing, so the reduction method has to be a chemical one rather than a thermal one. Or, could one use a flux which grabs the phosphate, maybe with calcium carbonate, to form calcium phosphate? Or iron/iron oxide, to form the very stable low-energy compound iron phosphate?

 

[orvi]

Nice looking button. I hope you're OK.

Dave

Yeah, I was wearing hard cotton long sleeved "jacket", which I use when doing small scale melting in the dish. Much more comfortable than protective "silver" suit, and enough protecting for this scale of the melt.
I was more angry than hurt/injured, as I mess with it for good hour (technical problems with practically every tool/piece of equipment needed), and in the end flying brass kill it all

Thanks.
I know pretty well that melting Pd in quartz dish in air isn´t regarded as good practice, but I do not have the equipment such as vacuum furnance to do it right way. And this "quality" is more than sufficient for selling.

I observed that unless you do not have larger quantity of Pd, it does not erupt adsorbed oxygen often. But what it really hates is changing between reducing and oxidizing enviroment.
You can very easily melt the calcined powder in induction furnance without any spitting at all, but as you touch the surface with graphite rod, it immediately turn itself into firework One touch for about half second, and 0,5g from that 17,5g button goes flying into the dish and surrounding mineral wool. I was prepared that this could potentially happen anytime, so I contained it all nicely.

Same with torch melting, which is in my opinion very tiring and wasteful, as it always spits some material. If you change fuel/oxygen composition during the melting or at the end of the melting (to cover it with reducing atmosphere), also spitting like mad. If you do this transition slowly, it can be tamed tho.

 

[orvi]

Nice detailed explanation of the entire process.

Found something interesting: Hydrothermal Solubility of Palladium In warm solutions with high NaCl concentration at a pH of 6.5 under high pressure, palladium was very soluble. There was also an observation for very low pH (about 1.5) at a low NaCl concentration that Pd also became very soluble. TOO hot, however, and the solubility plunged.

I wonder if the best conditions for separating Ag and Pd in solution might be a combination of high NaCl concentration plus HCl. Though, since this experiment took place at high pressures, this likely alters the chemistry.

At any rate, the study identifies new potential sources for PGEs: modestly warm hydrothermal brine deposits. Super-salty hydrothermal vents with internal temps between 300-700C could carry a huge amount of PGEs, depositing them in minerals as the steam bursts out and all the minerals precipitate.

I'm also curious about using phosphate to separate Ag and Pd. Silver phosphate is virtually insoluble in water (only 0.00065g/L), but very soluble in aqueous ammonia.

I can't find solubility data on palladium phosphate, though. If there's a solubility split there, say, either palladium phosphate is soluble in water or insoluble in aqueous ammonia, then there's a potentially easy and cheap way to separate the metals.

I'd have to look up the method for reducing them back to the metals. There is a biological staining method which uses silver phosphate reduction to silver metal to label certain tissues, so a method does exist already. I just need to find it. Silver phosphate melts without decomposing, so the reduction method has to be a chemical one rather than a thermal one. Or, could one use a flux which grabs the phosphate, maybe with calcium carbonate, to form calcium phosphate? Or iron/iron oxide, to form the very stable low-energy compound iron phosphate?

I like to do it diligently, when I decide to put something together to be publicly seen including mistakes and mishaps, that everyone could take an advantage and learn from them.
There is quite a bit of methods for solubilizing the noble metals in high chloride media, and also in chloride melts. But high pressure is quite limiting to the process scale, as these solutions are extremely corrosive on practically anything.

New possible deposits around salty hydrothermal vents - very nice. I very appreciate insights on geology of new possible deposits.

I am afraid phosphate does not get you very far. Phosphates in general are poorly soluble compounds, if transition metals are involved. Thing is, in ammonia, Pd is also very nicely soluble, forming tetraammino complex.

 

[kurtak]

Then, pH was adjusted with solid NaOH to the value of 2.

Per the bold print - why sodium hydroxide instead of sodium carbonate

I ask because all the instructions I have ever seen/read for reduction of Pd with formic (& therefore used/followed) have called for Na2CO3 for Ph adjustment rather then NaOH

Peoples instructions I have followed - Lou - 4metals - GSP - freechemist - OwlTech --- they all used carbonate rather then hydroxide

I am not an actual "chemist" - so I though the carbonate played an actual roll in the reduction

Apparently not

I guess I am just wondering if there is an advantage/disadvantage to the base used for Ph adjustment ?

Kurt

 

[kurtak]

(I would maybe do better washing it with few % HCL)

Just as a side note (&/or for it is worth) full strength (31%) HCl will dissolve ultra fine Pd & will do so at room temp let alone heated

The reason I know that is when I was processing CATs - some of them (with HIGH Pd) the Pd would instantly go into solution when pouring the HCl into the bucket of combs & before adding any Cl

Not saying its a real problem - just a potential - depending on how fine the particles - HCl concentration &/or temp

Kurt

 

[Yggdrasil]

Per the bold print - why sodium hydroxide instead of sodium carbonate

I ask because all the instructions I have ever seen/read for reduction of Pd with formic (& therefore used/followed) have called for Na2CO3 for Ph adjustment rather then NaOH

Peoples instructions I have followed - Lou - 4metals - GSP - freechemist - OwlTech --- they all used carbonate rather then hydroxide

I am not an actual "chemist" - so I though the carbonate played an actual roll in the reduction

Apparently not

I guess I am just wondering if there is an advantage/disadvantage to the base used for Ph adjustment ?

Kurt

I guess its easier to control with Carbonate, much easier to overshoot with solid NaOH.
At least that is how I see it.

 

[Yggdrasil]

Just as a side note (&/or for it is worth) full strength (31%) HCl will dissolve ultra fine Pd & will do so at room temp let alone heated

The reason I know that is when I was processing CATs - some of them (with HIGH Pd) the Pd would instantly go into solution when pouring the HCl into the bucket of combs & before adding any Cl

Not saying its a real problem - just a potential - depending on how fine the particles - HCl concentration &/or temp

Kurt

Would not that be because PdO is soluble in HCl?

 

[orvi]

Per the bold print - why sodium hydroxide instead of sodium carbonate

I ask because all the instructions I have ever seen/read for reduction of Pd with formic (& therefore used/followed) have called for Na2CO3 for Ph adjustment rather then NaOH

Peoples instructions I have followed - Lou - 4metals - GSP - freechemist - OwlTech --- they all used carbonate rather then hydroxide

I am not an actual "chemist" - so I though the carbonate played an actual roll in the reduction

Apparently not

I guess I am just wondering if there is an advantage/disadvantage to the base used for Ph adjustment ?

Kurt

Solution I was left with was practically pure Pd, with traces of copper and iron. Therefore, as you add NaOH granules directly, no big ammount of intermediary precipitate of BM hydroxides appear. Solution does not foam, you can go much more quicker, and as reaction heats itself from neutralization, re-dissolution of intermediary precipitate of BM hydroxides is much quicker.
However, if I have solution heavily contamined with copper, nickel or worse-iron, I prefer strong NaOH solution from start, then solid bicarbonate to the end. Reason is, with NaOH granules, as they touch copper/nickel/iron solution - immediately hydroxides precipitate on the surface of the granules, blocking further dissolution of NaOH and overall dissolution of NaOH/reaction with excess acid is very slow.

This can be overcome with addition of say 20-30% NaOH solution in thin stream - if you know approximately how much free acid you have, you can pour directly proportional ammount of hydroxide at once. It will heat very considerably (caution is necessary), fastening re-dissolution of intermediary BM hydroxide formation. Then, I continue with solid bicarbonate or carbonate till I reach the pH desired.

All of this is just for sake of saving time in my perspective. Point is just to raise pH to appropriate value. Nuances accompanied with this are explained above. You can go with carbonate from the start and result will be the same at the end

And as it is nicely pointed out, carbonate is milder base than NaOH - so if you overshoot considerably with carbonate, there is less risk of dehydrating formed BM hydroxides to insoluble oxides. Altough very unlikely, beginners could theoretically face this.
Direct use of sodium formate is also small added benefit - you save that little bit of "neutralization hassle", because you already adding the salt without the need to create it "in situ".

 

[orvi]

Would not that be because PdO is soluble in HCl?

From what I know and experienced, PdO is rather resistant to acid attack. Therefore many efforts are made to reduce possible PdO back to Pd either in reducing atmosphere of the furnance or heating to high temp (900-1000°C) and then covering with reducing atmosphere.

Hydrated palladium oxohydroxides from mild "neutralization" of PdCl4 2- complex anions, as are formed during bromate hydrolysis are more easily dissolved back with acid.

 

[orvi]

Finally, I was able to recover the silver and trace Pd from the AgCl which was collected. And it measured 0,1-0,2% Pd in content. I obtained around 40-50g of this contamined Ag, so in it I lost around 50-100mg Pd, which represents less than 0,6% of overall Pd content. Overall, not bad for that easy separation approach, and I am sure that it can be tuned even better