Pd Pt Rh mix industrial cat refining tips ?

[orvi]

Hi
for a longer time, we have a quite complex problem sitting in our refining shop. Solution containing approx.:
34% Pd
9% Pt
2% Rh
rest is base metals mainly Fe, Co, Ni, some trace Pb, no Ag nor Au.
Metals are in form of chlorides in used up AR soln. Some nitric is certainly present.
We aim to recover enriched rhodium fraction, more than 50% by mass would be fantastic, as we work with quite a lot of this material. Also Pt and Pd fractions need not to be super pure, also mixed fraction PtPd would be fine.

I read whole bunch of papers, topics trough forum... And came with some DIY plan how to deal with the issue:

1. drop Pd with DMG - enriching soln with Rh, at pH ca 3 there shouldnt be much Ni coprecipitation, but if some, that is not big issue
2. drop impure Pt with NH4Cl along with some part of Rh as contaminant
3. rest of noble metals cement on something (Zn, Fe ?)
4. calcining impure ammonium hexachloroplatinate and melting remaining metal sponge in NaHSO4 - or leaching Rh with H2SO4 - which approach would be better ? or is there better approach than these ?
5. cemented noble metals leaching similar way as above.

I have no big experience in this field, but have great practical chemistry experience overall. Maybe im missing some point or its wrong in some way...
My question is: will it work ?
Better question: what´s your suggestion ? how could this be done better/more efficently ? is it even possible ?

Thank you in advance for your opinion

 

[nickvc]

Welcome to the forum.
I will say from the off I’m no expert at PGM refining as I have always avoided it when possible.
The one thing I do recall is that PGMs need concentrated solutions to be precipitated with any degree of success.
If you would be happy with cemented PGMs why not use copper buzz bars and agitation to cement the values from the solution which would give you a much richer concentrate to work with if your aim is to separate the metals or you could simply melt assay and sell.
Trying to work with wet chemistry on cats is a losing option in my opinion, we had a member who spent hundreds of thousands of dollars trying it and found it paid better to de can crush assay and sell as the residues left after filtering contained more values than the refiners charges, he was a very large cat buyer so I assume he got really good terms but it never hurts to ask.

 

[JAREL]

If you have that in solution I believe it best to decompose any nitric acid in solution, add water of the liquid is very acid concentrated, then precipitate with aluminum or zinc putting a little at a time until solution is clear and you get no color from stannous chloride test. That precipitate can be then treated to separate the precious metals.

 

[nickvc]

If you have that in solution I believe it best to decompose any nitric acid in solution, add water of the liquid is very acid concentrated, then precipitate with aluminum or zinc putting a little at a time until solution is clear and you get no color from stannous chloride test. That precipitate can be then treated to separate the precious metals.

Problem here is that all the metals will cement out not just the values so back to square one :shock:

 

[orvi]

Thanks for invitation and suggestions also im sorry for my english, im not native speaker

I already have it in the HCl solution (ca 19% - by titration). Leaching was inefective, but prolonged heating get things dissolved (AR). XRF confirmed. already have it like this, so i must figure something out at least i must try hard before giving up
Cementing everything would be the next option, i wondered precipitating PGM black on zinc dust (excess), than stirring it til all base metals and excessive zinc dissolves in remaining acid (no metals insoluble in HCl are present except PGMs). Should be left with relatively nice enriched PGM mud (in theory).
Also i read in this forum that handling lots of DMG precipitate isnt very "friendly" approach to eliminate Pd.
So from the Pd/Pt/Rh drop on zinc i could proceed with hot nitric leaching (should be selective for Pd and remaining base m. junk and relatively fast due to surface area) to eliminate Pd.
The Pt/Rh residue heating in H2SO4 will be the final step.

 

[orvi]

update:
i treated a small sample of soln. (100 ml - roughly 1g Pt/3g pd + 0,1-0,2g Rh) with NaOH to make roughly pH 2-3 for proper DMG drop of Pd. DMG was used just for removing Pd from the sample, in bulk run ill use nitric leaching for PGM mud which i drop with Zn.
After removing Pd, i treated the soln with powdered zinc to precipitate the Pt/Rh remains. After successful precipitation (XRF) of all valuables, i added some more HCl to eat up the excess zinc, filter the mud and dry it.

To this point, everything worked just fine I than placed the Pt/Rh/trace Pd powder into 40-50 ml concentrated H2SO4 and cooked it with stirring at 200°C for 1h. But after dilution and filtration, filtrate doesnt have any Rh. No trace.
I wonder if I should cook it for example for few days (?), or the reaction would finish after couple of hours...
Hard to say, bit confused.
Any thoughts what ive done wrong ?

 

[orvi]

update:

after numerous failed attempts, here is another one
i decided that i "train" my approaches on pure, base metals free batch of Pt/Pd/Rh solution. so i managed to get solution with only PGMs dissolved in ratio roughly Pd50 Pt47 Rh3%.
this was treated with powdered zinc until the solution was free of values. PGM black was filtered and washed well from chloride residues.
to this black ppt, 53% nitric acid was added very slowly in small increments. heating and magnetic stirrer were on and mixture was held at rolling boil for 2 days, with regular aditions of 53% nitric.
after this there comes the dissapointment... in the cake of undissolved metals there was still shitton of Pd (roughly 60-70% dissolved), but strangely enough about 30-40% Pt dissolved also. but afterall, Rh still managed to stay in ppt.
discouraged, i washed the remains of the cake and baked it in conc. sulfuric acid at 200° for 2 hours. after that, the content of Rh in the cake dropped from roughly 3,5% to 1%. also about 1/3 of Pd was gone to the soln. Pt seems to be inert to hot sulfuric acid, which is great news.

i still wonder what is the reason that im unable to dissolve that finely divided Pd in nitric acid. i used to refine Pd contacts, which were solid pieces of Pd and dissolved in nitric with no issue. perhaps as the metals are precipitated chaotically and quickly on the surface of zinc, they block access of nitric straightly to Pd, forming aggregates (???). i dont know what exactly is going on.

thing that comes to my mind was to alloy the PGMs with Al, and than eat Al in HCl, preparing ultrafine PGM dust. this should eliminate the aglomeration issue (if this is the issue).
or "inquart" PGMs with silver, than proceed straightly to the nitric leach, leaving majority of Pt and Rh in powdered form.

anyone experienced similar behavior ?

 

[nickvc]

As I have stated I’m no expert but I wouldn’t go with silver as that will allow all the Pd and some of the Pt and Rh to be dissolved as well, perhaps an expert could confirm that.
My suggestion would be to inquart with copper and then dissolve in nitric leaving a copper palladium nitrate solution to be treated and the Pt and Rh as a mixed powder again ready for further treatment.
The option I would follow would be to melt with copper and assay and sell the bar but that’s a personal choice.

 

[orvi]

good point. copper seems to be legit choice.
i slowly start to disbelieve all statements in textbooks about Pt and Rh being so inert mainly Pt.
i firstly blamed old XRF to show some bullsh... but another new (and certified one) clearly shown, that in hot sulfuric acid (200 °C, 24h leach) Pt is also dissolved (quite nicely).

proceeded with clean PtPdRh chloride solution, i dropped everything on powdered zinc, filtered the PGM black, and cooked it in nitric for 1-2 days. about 60% Pd and 30-40 % Pt gone to the solution. no Rh could be detected in liquid.
the remaining powder (containing ca 3,5% Rh) was cooked in H2SO4 (200 °C) roughly 24 hours. Rh content in the solid residue dropped to less than 0,4 %. that was the good news. Bad news was that sulfuric leaching solution also contained shitton of Pt... :shock:
Discouraged, i diluted the H2SO4 leaching solution with water to ca 25% concentration and dropped PGMs with zinc.
The solid (after complete zinc dissolution in excess acid) contained 10 % Rh, ca 30% Pt and rest Pd.
Well done

yes. youre right. if it was my decission, catalyst will go straightly to the crucible, melted with Pb plus right flux at minimum 1400°C, residue cupelled and sold

it started to be too interesting ill keep you informed if something worth sharing comes out from my experiments with it.

 

[Geo]

To your solution, add a concentrated solution of ammonium chloride. Amounts and volumes are relative. Add the concentrated solution to the pregnant solution in amounts that are relative to the starting amount of pregnant solution. The ammonium chloride does two important things. First it drops Pt as a yellow salt. This is what you are looking for. Add until no more yellow salt is produced. Second, the ammonium binds with the palladium chloride. This allows the Pd salt to be converted to a solid by adding chlorine gas. The Pd salt will precipitate as a red to dull-red salt. None of this effects the rhodium to any great degree, which remains in the solution. I recover the Rh and can clean it somewhat but have never tried to refine Rh. Be aware that finely divided Rh powder will dissolve in concentrated HCl without an oxidizer. Palladium is difficult to cement from acidic solutions. It tends to dissolve in dilute HCl. To successfully cement all Pd from an acidic solution, the PH must be adjusted up. From my own observations, the less reactive a metal is, the harder it is to drop as an oxide. You can dilute gold solution until the solution has no color but still test positive for gold. All of the less reactive metals exhibit the same behavior. In an electrowinning cell, the solution entering the cell has very little dissolved metal initially.

 

[Geo]

To your solution, add a concentrated solution of ammonium chloride. Amounts and volumes are relative. Add the concentrated solution to the pregnant solution in amounts that are relative to the starting amount of pregnant solution. The ammonium chloride does two important things. First it drops Pt as a yellow salt. This is what you are looking for. Add until no more yellow salt is produced. Second, the ammonium binds with the palladium chloride. This allows the Pd salt to be converted to a solid by adding chlorine gas. The Pd salt will precipitate as a red to dull-red salt. None of this effects the rhodium to any great degree, which remains in the solution. I recover the Rh and can clean it somewhat but have never tried to refine Rh. Be aware that finely divided Rh powder will dissolve in concentrated HCl without an oxidizer. Palladium is difficult to cement from acidic solutions. It tends to dissolve in dilute HCl. To successfully cement all Pd from an acidic solution, the PH must be adjusted up. From my own observations, the less reactive a metal is, the harder it is to drop as an oxide. You can dilute gold solution until the solution has no color but still test positive for gold. All of the less reactive metals exhibit the same behavior. In an electrowinning cell, the solution entering the cell has very little dissolved metal initially.

Also, I use tri-chloro pool chlorine powder to generate the chlorine gas. Make sure the solution is acidic and add the pool chlorinator in really small doses while stirring. A very small amount creates a lot of chlorine gas. This is a recovery process and not refining. The resulting Pd salt is not pure.

 

[orvi]

To your solution, add a concentrated solution of ammonium chloride. Amounts and volumes are relative. Add the concentrated solution to the pregnant solution in amounts that are relative to the starting amount of pregnant solution. The ammonium chloride does two important things. First it drops Pt as a yellow salt. This is what you are looking for. Add until no more yellow salt is produced. Second, the ammonium binds with the palladium chloride. This allows the Pd salt to be converted to a solid by adding chlorine gas. The Pd salt will precipitate as a red to dull-red salt. None of this effects the rhodium to any great degree, which remains in the solution. I recover the Rh and can clean it somewhat but have never tried to refine Rh. Be aware that finely divided Rh powder will dissolve in concentrated HCl without an oxidizer. Palladium is difficult to cement from acidic solutions. It tends to dissolve in dilute HCl. To successfully cement all Pd from an acidic solution, the PH must be adjusted up. From my own observations, the less reactive a metal is, the harder it is to drop as an oxide. You can dilute gold solution until the solution has no color but still test positive for gold. All of the less reactive metals exhibit the same behavior. In an electrowinning cell, the solution entering the cell has very little dissolved metal initially.

You are right. I have tried this standard procedure early on. Main problem with this stubborn feedstock is very low Rh content. It is very difficult for me to "enrich" Rh content for example to 20% (which could we sell) without loosing like 50% in the process. I am certain that it could be done properly, but the NH4Cl step and also DMG or chlorine Pd precipitation drag also a good portion of Rh salts to the precipitate. I studied many papers on the Pt drop subject, also read whole bunch of information here on the forum, mainly from the experts in the field, like Lou or 4metals. Along the other refining stuff that i need to do, i have little time to move with this material forward.

Currently, small testing is undergoing - which raised from observation, that Pt is relatively soluble in nitric acid, when finely divided.
Cementation of another PtPdRh solution on zinc formed nice, easy settling precipitate of PGMs. Unfortunately, prolonged boiling in nitric acid shown that reaction very quickly reach equilibrium point, when no more metals are dissolved. Platinum and palladium were dissolved, Rh staying in the precipitate. But it was slow and results were nowhere "usable", as the reaction didn´t reach full dissolution of PtPd anyway.
Equipped with proper gear, we prepared two buttons of PtRh material "inquarted" with copper and aluminium, respectively. PGM content 20%. Just to test this out. No base metals, just PtRh alloy and copper or aluminium.

Theory was to dissolve them in nitric or NaOH, leaving very active particles of PGMs, which could then be leached with nitric to get rid of the majority of Pt. Just enough to pass that 20% mark on Rh, as we could recover sellable Rh. Dissolving the copper "inquarted" button in nitric acid unfortunately produced so fine nanoparticles of PtRh, that they refuse to settle even after a week. Flocculants arent of any help (at least what i have in hand). So i am aware of putting the second Al "inquarted" button to NaOH solution, as it will likely produce the same result.
Now figuring out what to do with this route, if to completely abandon it, or try one more time.
We concluded with any route employed, the main problem will be separation of Pt and Rh from this mix.

I was never "***ked up" so badly by chemistry in my life (doing high end chemistry for 8 years now) after these experiences, i start to think that it will be unaviodable to conduct hydrolysis on this stuff in some step. Another thing, which i never done before. Also, i dont know if it is even possible to conduct the hydrolysis on such impure mix of PGMs (i can get rid of base metals easily).

Maybe one more question:
It is possible in some way to drop Pt and Pd simultaneously with saturated NH4Cl solution and chlorine gas ? This could possibly eliminate two steps of precipitation/filtration, maybe lowering the loss of Rh locked up in the precipitate to somewhat lower percentage. In my position, i have no big benefit/profit from refining Pt and Pd to the pure form (eg 99,9%). My objective is only to produce some reasonably enriched Rh fraction, where i could get paid for Rh. I also want to avoid the manipulation of large quantities of PGM salts to minimum due to hazards, so maybe 1-5% of extra profit on fully refined stuff don´t outweigh the potential risks.


You mentioned, that Rh is soluble in concentrated HCl even without oxidant. This is very interesting observation, thank you for pointing it out

Thank you for your reply.
orvi

 

[Geo]

It can be done in one step but then you would have to reduce and refine the Pd and Pt separately. After the addition of NH4Cl, it is simple to filter the salt out and then add the chlorine gas for the Pd. In my case, I add pool chlorinator powder to generate the chlorine gas. The palladium ammonium chloride is stable in solution until chlorine gas is introduced and then converts to red (NH4)2PdCl6. The process I do is quick and dirty. It is not perfect but it effectively separates the metals. There will be cross contamination of elements in each phase. This is recovered by cementation and collected until enough to process is gathered.

 

[orvi]

It can be done in one step but then you would have to reduce and refine the Pd and Pt separately. After the addition of NH4Cl, it is simple to filter the salt out and then add the chlorine gas for the Pd. In my case, I add pool chlorinator powder to generate the chlorine gas. The palladium ammonium chloride is stable in solution until chlorine gas is introduced and then converts to red (NH4)2PdCl6. The process I do is quick and dirty. It is not perfect but it effectively separates the metals. There will be cross contamination of elements in each phase. This is recovered by cementation and collected until enough to process is gathered.

Thank you for the valuable information. Do you measure some quantity of NH4Cl based on expected ammount of PGMs or just dump in good excess of it (sat. solution) ? I am asking because too much NH4Cl can also cause the drop of other PGMs and junk with hexachloroplatinate precipitate (at least what i read about it and observed in my first trials).

If i dont bother with PtPd mixture to say first (i will end up mixing the two anyway in the final melt to sell them), is there any culprit to process platinate and palladate salts (precipitated both together) to PtPd metal alloy ? I do not intend to refine them, as this will cause a lot of hazardous work for very little profit. If i have them separated in close to pure form (eg 95-98%) by means of two stage precipitation as you do, other element (traces of Pt in Pd fraction) would be considered as impurity at selling, thus lowering the value of whole lot.

What is the composition of pool chlorinator you mention ? I think you mean calcium hypochlorite. Here, it is obtainable, and I think about using it instead of setting up chlorine generator from TCCA/HCl - calcium in chloride solution shouldn´t be a problem (?). One option was to use chlorate, but i don´t quite like the possibility of ClO2 popping and splashing PGMs to me and whole fume hood.

Yeah, i look for something like this rough and straightforward. I would be pleased to have enough time to play with this in more scientific way (that´s what i am used to do in my "real" work). Time for "educated giving-up" on this was while ago. Now i am all into this (unintetionally) and i must somehow scavenge the values from it. Luckily, time is not the big issue, I am not in hurry.

I will post here if any usable result (positive or negative) come out of this situation.
Appreciate your responses.
orvi

 

[Geo]

Trichloro-s-triazinetrione in powder form.
Sodium hypochlorite in powder form will work as well but commercial bleach will dilute the solution.

 

[Geo]

Ammonium chloride in a saturated solution is used to drop both Pt and Pd with addition of chlorine gas. The solution is also used as a rinse to wash the recovered salt. If you use water, it will dissolve some of the salt. Since an excess of NH4Cl is needed, the saturated solution can absorb very little extra salt. It is just natural to use the same solution to rinse the solution from the precipitated salt.

 

[orvi]

So, after a long time I have something new with this refining/learning venture. And maybe worth sharing with community, as these findings could be usable in some way kind of. This time, training on model situation/hypothesis.

As I explained above, we hypothesize about "inquarting" PGMs with copper or aluminium, and then etch base metals out and leach the remaining PGM precipitate with nitric acid in order to dissolve Pd and Pt, hopefully leaving Rh as insoluble mud.

So, two alloys were prepared (aiming on 20% PGM content - roughly accomplished), but this time just PtRh, without Pd (no pure Pd in hand unfortunately). As the platinum will be more tough to dissolve, it should be good quantification of result anyway. Dont mind other elements measured, our XRF starting to be confused a little

First, copper alloy was etched with nitric acid (65%). From the first few ml of acid, it was quite obvious what is gonna happen copper dissolved without issues and also without reqirement for heating - to the black "solution". Precipitate thus formed was so fine, that it refused to settle (whole week settling time). Some bigger particles make it to the bottom, but no progress was seen afterwards. Two different flocculants were tried with no success, and similar result was obtained also with centrifuge. Centrifuge tho produced a small ammount of relatively usable "solid" residue. I was affraid of something like this happening, but it is what it is.

Supernatant (after centrifuge) was measured by XRF to reveal that significant portion of Pt was in solution. But it is important to mention, that solution was full of nanoparticles of the PtRh, so the result is only informative. Also, Os and Au certainly aren´t present, our XRF gun need some understanding for fiction results

This approach could be valuable for quick dissolution of these metals (mainly Pt), maybe in AR (as Lou suggested in one thread), but this precipitate would be very painful to deal with. No filtration recommended, as it would likely fail completely It is what it is, lesson learned.

Next, Al+PGM alloy was dissolved in 20% NaOH solution. This time i was pleased by the results from the beginning.

Reaction was very robust, but not going out of control so quickly, as plain Al would behave. Cooling with water bath and cautious addition of more NaOH is required. Precipitate of PGMs was this time very compact, yet fluffy. Heavy, readily settled to the bottom of the beaker. Precisely what I was looking for the next experiments. But, reality shows if it serves good for the nitric leaching step Hopefully, grain size and crystal structure will help us.

 

[arthur kierski]

Oi
por um longo tempo, temos um problema bastante complexo sentado em nossa oficina de refino. Solução contendo aprox .:
34% Pd
9% Pt
2% Rh
o resto são metais básicos, principalmente Fe, Co, Ni, alguns traços de Pb, nenhum Ag ou Au.
Os metais estão na forma de cloretos na solução AR usada. Certamente algum nítrico está presente.
Nosso objetivo é recuperar a fração enriquecida de ródio, mais de 50% em massa seria fantástico, pois trabalhamos com bastante deste material. Além disso, as frações de Pt e Pd não precisam ser super puras, e a fração mista de PtPd seria adequada.

Eu li um monte de jornais, tópicos de fórum ... E veio com um plano DIY de como lidar com o problema:

1. solte Pd com DMG - solução de enriquecimento com Rh, em pH ca 3 não deve haver muita coprecipitação de Ni, mas se houver, isso não é um grande problema
2. solte Pt impuro com NH4Cl junto com alguma parte de Rh como contaminante
3. resto de metais nobres, cimento em algo (Zn, Fe?)
4. calcinação de hexacloroplatinato de amônio impuro e fusão da esponja de metal remanescente em NaHSO4 - ou lixiviação de Rh com H2SO4 - qual abordagem seria melhor? ou há abordagem melhor do que essas?
5. metais nobres cimentados com lixiviação de maneira semelhante à anterior.

Não tenho grande experiência neste campo, mas tenho grande experiência prática em química em geral. Talvez eu esteja perdendo algum ponto ou esteja errado de alguma forma ...
Minha pergunta é: vai funcionar?
Melhor pergunta: qual a sua sugestão? como isso poderia ser feito melhor / com mais eficiência? é mesmo possível?

Agradeço antecipadamente pela sua opinião

 

[Yggdrasil]

?

 

[arthur kierski]

i would do the folowing---1-cement with copper--the powder from cementatio,dissolve with nitric,making a solution,with pd---the powder left use hcl+peroxide,the liquid will contain pt and the powder left is rh--now ,treat the solutions with the reagents like dmg for pd and the platinum solution with nh4cl--the rh ,purifies your way