Solvent Extraction Preferences

[Rmwatson78]

I have been reading up a lot about solvent extraction of gold bearing hydrochloric acid solutions and I was curious about the preferences from some of the more experienced refiners and chemists. Given the choice, would you prefer to use Butyl Diglyme or MIBK? What in your opinion are the pros and cons for each?

 

[Lou]

TBP better

 

[Gratilla]

I'm sure Lou has (very) good reasons for his choice. TBP (Tri Butyl Phosphate) is a long established solvent, but can be "wide spectrum" under certain conditions ie carrying over silver and some other metals that may be in your "gold bearing hydrochloric acid solutions".

If you're serious about clean separations, I'd recommend you carefully study phase diagrams for your solvent short list and whatever metals you might have in solution in addition to AuCl3. For rich gold solutions consider DEM (DiEthyl Malonate); this solvent is heavier than water and does not suffer phase inversion as the gold builds up in solution, making separation easier.

A key word search of the forum should turn up some interesting information.

 

[Lou]

Not as stable. All of them have draw backs.

 

[Deano]

All solvent extraction chemistries have trade offs.

Probably the safest to use is MIBK in that many people react badly to it and thus stop using it very quickly and so limit their exposure.

TBP is good but the carrier odour puts me off.

If you are willing to carry out more contacts with solution then BDG affects me less than other solvents.

This field has to be rated on factors other than the technical excellence of the solvent, personal reaction is of great importance.

Deano

 

[Lou]

Deano, you can use IsoPar M and have no odor, or simply skip the diluent completely and use it neat like I do. It should not be used in long term installations with nitric acid much above 8 M due to the risk of ill-defined yet explosive "red oil". It is indeed wide spectrum but one can selectively back extract the living hell out of it with very little loss of target metal units. Thankfully, I have a lot of experience with TBP.

I don't know why people find MIBK off putting, aside from the flammability--I love the smell of it and it's viscosity is less than the others, so less issues in disengagement of phases. Hexone is quite useful for refractory metal separations, less so for precious metals. DBC is good.

Diethyl malonate smells like apples (and is a listed DEA precursor to most barbiturates if I remember correctly, so probably not best for people to order here unless ordered in small quantities, like from perfumer's apprentice). I've employed DEM on several occasions. Do I find it much better than dibutyl carbitol--no, not at all, after spending some time with it and the ICP-OES.

All of the organics more or less need oxalic acid and a good reflux, sulfites do not tend to work well.

n-octanol/MIBK mix is probably the best (solubility wise) but does need a diluent due to viscosity and is not oxidatively stable so it's in HCl systems with ORPs less than 650 mV. You can use TBP in aqua regia, butyl diglyme as well (in diluted AR).

Of course, wise people going for recovery use resins for gold extraction and elution. SX systems are very expensive and often have more issues up front.

And Gratilla, if one is worried about phase inversion, then one needs to pony up and buy a spectrophotometer.

 

[Deano]

Many of the users of MIBK as an extractant for AAS work developed what was locally called MIBK headaches.

These were similar to the gelignite headaches obtained from touching your head with gelignite contaminated hands.

It was a relief when the DIBK/Aliquat 336 extractant was developed and rapidly displaced the MIBK.

Most users of DIBK dislike the smell but can work with it.

I have always opted for alternative recovery methods if they are available, I don't think that organic fumes are going to be to your health's advantage.

It is interesting how people react in totally different ways to organic odours.

About half of people cannot smell cyanide and of those who can there is a majority who perceive it as bitter almonds but others smell all sorts of things, I know some who smell it as similar to chocolate.

It is all a very personal thing.

Deano

 

[Geraldo]

Deano, you can use IsoPar M and have no odor, or simply skip the diluent completely and use it neat like I do.

OK, I was willing to go along with a lot of this, but in 20+ years crawling around a lot of filthy plants, I have never, ever seen anyone use TBP "neat." Whether in a mixer-settler, or even a gentle air-stirred column, that is a recipe for mayonnaise.

I would say overall in troubleshooting, that beyond the ubiquitous silicic acid screw-up (a problem in leaching and settling/filtering, should never get to S/X) around 90% of SX problems can be traced to insufficient diluent and phase modifiers. Basically, kerosene and decanol are your friends, and people who don't carefully monitor organic viscosity and disengagement characteristics (ultimately kerosene and phase modifier levels) will have SX failures - phase disengagement, carry-over, compromised extraction efficiency etc.

Or are you talking about a shakeout in a sep funnel? Even then, I use diluent and octanol/decanol but then I am trying to simulate something in a mixer-settler, perhaps some folks don't if they are doing analytical work?

As an aside, a large proportion of the remaining 10% of problems I have seen have been screw-ups in O:A ratio. Again, using TBP "neat" would be problematic there too, unless you didn't care to control the concentration of the stripping solution.

All of the organics more or less need oxalic acid and a good reflux, sulfites do not tend to work well.

What are you using oxalic acid for? Are you stripping gold from TBP with oxalic acid? I have seen that in lab tests, but I am not aware of an operating gold SX plant using oxalic acid (although it may well exist - I have read that rare earth plants in China often use oxalic acid, and they may use it in gold operations too - not that popular in Western countries AFAIK).

Thanks.

Best Regards, Geraldo

PS good (purified) kerosene doesn't smell, and good (purified) decanol only smells if it gets very warm. Some of the products of damaged organics DO smell really bad, but with some care a decent SX plant will barely have any odor other than the stray smells of strong acids and such. I found the gear oil used for some mixers was stinkier than acres of uncovered settlers.

 

[Lou]

I'm talking sep funnel Geraldo not mixer settler batteries, not centrifuges, not pulse plate columns. I'll bet you have a higher Reynolds number in a sep funnel over any settler, so mayo is worse. I lab pilot most things in a 2L FEP sep funnel which is probably about as big as any one doing a few ounces will need.
I don't have issues with emulsions with TBP at bench with what I work with but you're in mining/primary not secondary, right? It's not just viscosity / interfacial surface tension that cause problems--it's also what's in the feed, pH, ionic strength, tramp. I'm not working with 200K liters at 5 ppm influent.

I agree that diluents and phase modifiers can definitely be helpful on process scale but I try to keep them out of my lab if doing sample prep as they can give low results. They're your friends til they're not. Example: say you go to use kerosene/alkyl sulfide for Pd extraction--you'll have big issues unless you pretreat the solvent system. Been there, dealt with that. Too much free sulfides = scum. Why do you think they used to use chloroform instead of aromatics which can have amines (like pyridine) or paraffins (sulfides from the refinery) for a diluent for TBP? In halocarbon you don't have Pd co-extraction and that's why. Most of that is now supplanted with Alamine 336 in newer flow charts. Purified "good" kerosene still isn't good enough, I still have to clean it all up. I like low aromatics and free sulfide below 15-20 ppm.



Where are you Geraldo, Oz? Asia? Maybe it's a Western Hemisphere thing...anyway,

Oxalic acid is used as a reductant out of SX reagents if you're after high purity metal. See Mintek's flow chart for both Harmony and Peñoles. There's a couple iterations of that first PDF out there. I believe it's 1-octanol they're using. Or you could stir with ferrous sulfate, pressure strip etc.
http://www.saimm.co.za/Journal/v104n06p339.pdf (note they're prob stripping with dilute HCl)
http://link.springer.com/article/10.1134/S0036023612110083

I've done it direct from organic, with TBP, hexone (not the CN complex), DBG, DEM, and Et2O. Usually the resultant product is fairly compact gold sand.


The use for oxalic acid in rare earth schemes is for oxalate precipitation. REEs and Th come down. Most of the lanthanides are done with TBP out of nitrate.

Thank you for all of your points--they are well made.

 

[Geraldo]

I'm talking sep funnel Geraldo not mixer settler batteries, not centrifuges, not pulse plate columns. I'll bet you have a higher Reynolds number in a sep funnel over any settler, so mayo is worse. I lab pilot most things in a 2L FEP sep funnel which is probably about as big as any one doing a few ounces will need.
I don't have issues with emulsions with TBP at bench with what I work with but you're in mining/primary not secondary, right? It's not just viscosity / interfacial surface tension that cause problems--it's also what's in the feed, pH, ionic strength, tramp. I'm not working with 200K liters at 5 ppm influent.

I agree that diluents and phase modifiers can definitely be helpful on process scale but I try to keep them out of my lab if doing sample prep as they can give low results. They're your friends til they're not. Example: say you go to use kerosene/alkyl sulfide for Pd extraction--you'll have big issues unless you pretreat the solvent system. Been there, dealt with that. Too much free sulfides = scum. Why do you think they used to use chloroform instead of aromatics which can have amines (like pyridine) or paraffins (sulfides from the refinery) for a diluent for TBP? In halocarbon you don't have Pd co-extraction and that's why. Most of that is now supplanted with Alamine 336 in newer flow charts. Purified "good" kerosene still isn't good enough, I still have to clean it all up. I like low aromatics and free sulfide below 15-20 ppm.

Thanks for your insight Lou!

Yes, I work in mining and primary recovery, although of course many of those operations now look to secondary recovery for $ and to improve their "environment" cred. For example, Teck's Trail smelting complex recycles electronics through the lead smelting circuit (technically through slag fuming).

And for SX the feed is absolutely a big factor in operation, but of course those things are supposed to be controlled or there are unit operations to deal with negative feed issues - like seeding, settling and filtration to remove silicic acid, or carbon treatment of feed solutions to remove stray organics.

The sep funnel thing is interesting, at least for hand shakeouts. If you are young, spry, energetic, you can have pretty vigorous mixing. If you are old and lazy like I am now - well, not so much. The mechanized sep funnel testing at one Cu plant I visited was more like very gentle rocking.

Overall, I haven't had a lot of success replicating plant physical problems using sep funnel tests. I have made mayonnaisse though, using a synthetic leach solution and a mix of Lix 63/D2EHPA/Shellsol, because I forgot the decanol. I have never used TBP neat - all the TBP I have seen seemed gooey-er than many other SX reagents, so I just assumed there would be more trouble. And as I suggest, I have never seen a plant using any of these reagents "neat." That was all I meant - I wasn't trying to be disrespectful. Hope that is clear.

I have NO personal experience with PGMs. My SX experience is limited to Cu, Ni, Co, Ag, Au, In and Ge, plus some pilot plant experience with Zn (although the "pilot" plant was bigger than commercial Ag, Au, In and Ge plants...) and an unpleasant turn with some halide S/X from Zn solution.

Where are you Geraldo, Oz? Asia? Maybe it's a Western Hemisphere thing...anyway,

Oxalic acid is used as a reductant out of SX reagents if you're after high purity metal. See Mintek's flow chart for both Harmony and Peñoles. There's a couple iterations of that first PDF out there. I believe it's 1-octanol they're using. Or you could stir with ferrous sulfate, pressure strip etc.
http://www.saimm.co.za/Journal/v104n06p339.pdf (note they're prob stripping with dilute HCl)
http://link.springer.com/article/10.1134/S0036023612110083

I've done it direct from organic, with TBP, hexone (not the CN complex), DBG, DEM, and Et2O. Usually the resultant product is fairly compact gold sand.


The use for oxalic acid in rare earth schemes is for oxalate precipitation. REEs and Th come down. Most of the lanthanides are done with TBP out of nitrate.

Thank you for all of your points--they are well made.

I am currently in Trail, BC. I have seen operations primarily in Canada, the US and Europe.

Every trip (either as a visitor or a paid consultant) with Penoles has fallen through over the years for various reasons so I have never crawled through their operations. I once was just 4 days from getting a tour of their Naica crystal cave (and had even purchased my own scuba gear) before they rescinded their offer - that saddens me to this day!

I had almost forgotten about reductive stripping. I know it's a thing, just not a popular thing. The Mintek Minataur process you referenced doesn't actually use reductive stripping - they scrub pregnant organic with HCL, strip with DI water, do a fine organic removal on the strip solution (activated carbon, I believe) and then precipitate with oxalic acid, although in other papers I have seen them reference using SO2 since oxalic is so expensive. The closest to reductive stripping I have seen commercially was using sulfurous acid for stripping in a precious metals recovery circuit at a Cu plant.

I know that TBP is very stable, but have you noticed any accelerated organic degradation from reductive stripping with oxalic acid?

Thanks again for your insights and experiences - very interesting!!!

G

 

[Lou]

I never answered your last question but will soon.