Recycling Silver

[greybeard]

I would think that I could use more current with the CN- complex ion because I would not be generating H2S.

Actually, you will need considerably more voltage to achieve the same current, as the cyanide complex is only weakly ionized. This is in fact why commercial electroplaters normally use a cyanide bath: the silver ion density is much lower than it would be with, say, nitrate, and the resulting deposit is denser and smoother. The silver I recovered by the zinc cementation route was electrolyzed from the nitrate (to get rid of the zinc which had snuck in as oxide) and yielded a dense mass of dendritic crystals; a cyanide bath would presumably have allowed me to go directly to massive silver.

As far as recovering silver from old paper goes, it would seem that just using a two- or three-bath sequence of rapid fixer would be the easiest. You aren't concerned with archival stability (although the fixed-out paper is prized for certain alternative processes) so any fixer would work (straight hypo if speed is not an issue).

If the silver loading is only two grams per liter (as argued above) and the fixer capacity is 30 8x10s per liter, a 100-sheet box will yield considerably more in silver than the cost of the fixer. A two-bath setup will raise the silver loading of the first bath to at least two or three times the "archivally recommended" level, making it even more economic. If you are really provident, just use your "exhausted" print fixer to salvage a few sheets of paper per session, and write off the small amount dragged out in the paper fibers.

 

[kq6up]

Would not the KCN leach out the AgX very fast and thoroughly with just a 1% solution? This might be a moot point as there was no KCN in the lab.

Thanks,
Chris Maness

 

[wogster]

I would think that I could use more current with the CN- complex ion because I would not be generating H2S.

Actually, you will need considerably more voltage to achieve the same current, as the cyanide complex is only weakly ionized. This is in fact why commercial electroplaters normally use a cyanide bath: the silver ion density is much lower than it would be with, say, nitrate, and the resulting deposit is denser and smoother. The silver I recovered by the zinc cementation route was electrolyzed from the nitrate (to get rid of the zinc which had snuck in as oxide) and yielded a dense mass of dendritic crystals; a cyanide bath would presumably have allowed me to go directly to massive silver.

As far as recovering silver from old paper goes, it would seem that just using a two- or three-bath sequence of rapid fixer would be the easiest. You aren't concerned with archival stability (although the fixed-out paper is prized for certain alternative processes) so any fixer would work (straight hypo if speed is not an issue).

If the silver loading is only two grams per liter (as argued above) and the fixer capacity is 30 8x10s per liter, a 100-sheet box will yield considerably more in silver than the cost of the fixer. A two-bath setup will raise the silver loading of the first bath to at least two or three times the "archivally recommended" level, making it even more economic. If you are really provident, just use your "exhausted" print fixer to salvage a few sheets of paper per session, and write off the small amount dragged out in the paper fibers.

If your fixing to remove all silver then you don't need to be archival, but you may wish to do it in near total darkness, to prevent any kind of physical development resulting in fixer not being able to extract all of the silver. Don't forget if you then wash the paper the first few minutes of wash will contain some silver as well..

 

[greybeard]

Would not the KCN leach out the AgX very fast and thoroughly with just a 1% solution?

It would, although potassium cyanide is not the friendliest stuff to have around, and it you are dead-set on being thoroughly responsible, dealing with the waste is something of an issue.

Silver cyanide per se is insoluble, so you need a 100% excess of cyanide to form the soluble complex. If my numbers are correct (and it's been a long day, so don't count on it) a gallon of your 1% KCN would carry almost exactly a Troy ounce of silver, and it would take about 400 sheets of 8x10 paper to saturate it. Of course, that much paper will drag out a considerable amount of solution, so you will need to wash it (preferably in a cascade to minimize the amount of contaminated water) and then recover the silver from all of the resulting solutions.

Personally, I'd rather have a pile of paper saturated with sodium thiosulfate than the same pile containing a measurable amount of cyanide...

If you decide to try this route, it would be best to work out the decontamination procedure before dissolving any cyanide salts (Oxidation to cyanate seems to be the preferred option, but a sensitive test for completeness it going to be in order, too.)

 

[kq6up]

Thanks greybeard. I will definitely take that into consideration. At this point it was more of an academic question because I was unable to find any cyanide on the shelves in our chem department. I also do not want to waste any reagent grade stuff since the low plating grade stuff would work just fine. My next door neighbor growing up was a technician at a plating shop. I used to get all the crazy stuff I wanted when I was a kid. It is amazing I did not off myself in the process, but that with ham radio peaked my interest in science and I wound up getting a BS in Physics. Pure curiosity is a good thing.

Thanks,
Chris Maness

 

[alexhill]

To get rid of the KCN isn't very difficult (but I would rather be using sodium thiosulfate for removing silver from prints). (there was a url link here which no longer exists) has a pdf in it which describes using bleach to break it down and liberate the silver. Just be sure to wear nitrate gloves when using KCN, and don't drink it

This has been a really helpful and interesting thread, thanks guys!

 

[kq6up]

H202 kills it too. As a mater of fact my first experiment was bleaching NH4S203 with H2O2 and it knocked the Silver out of solution. However, it was in the Ag2S form.

Regards,
Chris Maness

 

[greybeard]

If, as I hope, you saved the sulfide, you should put it in a crucible with an excess of KNO3 and fuse it; if what I have read is correct, you should wind up with a lump of silver. My inclination would be to add a bit of borax, too, since it will flux most stray oxides that might be present and help the globules to coalesce. A propane torch should be an adequate heat source, as would acetylene-air, but if you use oxyacetylene it may be hard to avoid damaging the crucible.

I have several grams of product from zinc cementation, and I plan to try the nitrate fusion with it. Unlike most of the non-electrolytic silver recovery processes, zinc cementation is well documented. At high pH, there is better than 95% recovery, and I treat twice with excess zinc, so there should be negligible silver in the waste. The downside is that the zinc is almost certainly less pure than silver obtained through electrolysis, but the nitrate/borax fusion should take care of most of the impurities that I would care about.

 

[alexhill]

There was a question about the strength of hydrogen peroxide, as it (?) degrades over time and can be difficult to judge how much is needed to kill the KCN, where as bleach doesn't have that problem. I haven't had to break down any kcn waste yet (still less than 1 gallon) but the hydrogen peroxide is very exothermic, correct?

 

[kq6up]

The H2O2 with NH4S2O3 was VERY exothermic for OTC strength H2O2. It was just the plain ol' 3%. It took a while to run away thermally, but it easily reached 110 DegF from a 75 deg solution. I would imagine some 30% food grade would boil over.

If greybeard's crucible recipe works, that would probably be the best way to go. Just fry it with high current and dump out the Ag as Ag2S. I couple of amps would get it done in minutes, and just fire the Ag2S in a crucible coffee filter and all. If you don't want to bother with the electricity the H2O2 does the same thing, but the Ag2S is much finer, and it takes a whole week to precipitate out of solution.

Time to get experimenting.

Chris

 

[greybeard]

There is a potential problem (pun intended) with the electrolytic process: you may have an appreciable amount of silver left in solution as the sulfate, which my Chemical Rubber Handbook, 44th edition, shows as being soluble to the extent of 5.7 g/l in cold water and 14.1 g/l at boiling. Both the kinetics and thermodynamics matter, so a simple experiment to test for residual silver might be the easiest way to resolve this question,

It turns out that sodium sulfide is used in patina formulations for metal sculpture, and is (relatively) cheap. Its solutions are corrosive, so it has to ship as a hazardous material, but it may be available locally in larger communities. It seems that about all you would have to do is make sure the solution is comfortably alkaline (to prevent hydrogen sulfide formation) and then add an excess of sodium sulfide.

Silver sulfide is about as insoluble as they come so it should be easily possible to drive the reaction very close to completion. Any excess sulfide will simply oxidize to sulfate in the nitrate fusion step, so there is no need for precise stoichiometry. I don't know what the precipitate would be like, but it if won't filter readily, let it settle for a few days and pour off the supernatant. Silver sulfide is pretty dense, so unless it forms a gel, it should drop out pretty readily.

 

[kq6up]

Would not the presence of the Thiosulfate radical effect the insolubility of of Ag2S. Ag2S does dissolve redily in aqua-regia, but instantly converted to insoluble AgCl. So to paraphrase you, just dump in excess Na2S and you would precipitate out all of the Silver with no electrolysis. That might be the best way, and just fire the filtrate in a crucible with KNO3 and borax to force the silver to bead instead of flow. Does this sound correct?

Chris Maness

 

[greybeard]

Thiosulfate may complex Ag2S as it does AgCl, but I suspect that it doesn't (I'm a physicist, not a chemist.....). If it does, then the Na2S method won't work all that well. In most qualitative analysis schemes, sulfide ion is used to drop out the heavy metals, more or less independent of other stuff present, so it seems like it would be worth a try. If you have ammonium polysulfide in your lab, it would take about ten seconds and a few milliliters of used fixer to find out.

I had a bright idea once: use zinc to extract the silver, dissolve out any excess zinc with sodium hydroxide, and dissolve the residue silver metal and/or oxide) in nitric acid to make silver nitrate. It didn't work---presumably because a lot of the silver was present as sulfide. Just extracting the silver/zinc with nitric acid and electrolyzing did work, as I mentioned earlier. I still have the presumed sulfide, so I plan to try the nitrate fusion on it when I get my hands on a graphite crucible.

As far as the sulfide theory goes, your summary is right on. I think that if I had a filter paper in there, I would ash it in air before adding the nitrate, just so that rapid combustion (nitrate + cellulose -> poof!) wouldn't loft stuff out of the crucible. There will be some zinc present, but if the borax is in excess the resulting glass should be easily water soluble.

 

[kq6up]

I have sodium sulfide in the lab. I will give it a go on Monday. The sulfide is pretty insoluble. It is supposed to dissolve in HNO3, but it did not for me. Maybe it needed to be heated or a stronger concentration. I have a ceramic crucible at work. I am just going to fire it with a propane torch.

Regards,
Chris Maness

 

[greybeard]

Rats. It seems that thiosulfate does indeed complex silver sulfide (along with gold, etc.).

I found a Kodak assay procedure for used fixer that is based on titrating with thioacetamide (motion.kodak.com/.../US_plugins_acrobat_en_motion_support_processing_h243_1209.pdf) but ferricyanide is also needed, presumably to oxidize either the silver or the thiosulfate in order to make the sulfide reaction quantitative.

Hydrogen peroxide from Wal-Mart is probably cheaper than potassium ferricyanide from just about anywhere, so the sodium sulfide route may not be all that attractive after all. If the precipitate from peroxide addition is hard to filter, it may be possible to "seed" the mixture with silver sulfide and then add the peroxide very slowly (dropwise, over hours or days) to permit nucleate crystal growth rather than gel formation. Or just aging it for a while may give a similar result.

 

[kq6up]

I think the electrolysis method was better even if I just dump high current through. The precipitate settled out within a matter of minutes. I would imagine the S(-2) ion is produced by the Anode. I wonder if a salt bridge might prevent the S(-2) from attacking the metallic Silver.

I think the reaction for the H202 is:

H2O2 + 2[Ag(S2O3)2](-3)] = Ag2S + 3SO4(-2)

The H2O2 killing the hypo and dropping Ag2O is what I assumed would happen, and it did kill the hypo. I tested for Ag2O by adding excess NH4OH to redissolve the Ag2O. It did not do anything, so I was unsure of what I had until I did the electrolysis, and determined I had made Ag2S.

Regards,
Chris Maness

 

[willrea]

The sun is finally out so maybe I can start evaporating my spent chemicals.

 

[Kirk Keyes]

To get rid of the KCN isn't very difficult (but I would rather be using sodium thiosulfate for removing silver from prints). (there was a url link here which no longer exists) has a pdf in it which describes using bleach to break it down and liberate the silver. Just be sure to wear nitrate gloves when using KCN, and don't drink it

While this procedure will work, keep in mind the the gas that's liberated by adding bleach (hypochlorite) to a cyanide solution is cyanogen chloride which is a VERY toxic gas. It is a chemical warefare gas known as "CK". Cyanogen chloride is known to penetrate some gas masks. The effect is quite similar to breathing hydrogen cyanide.

This proceedure, if done, must be done with adequete ventilation.

A bit more info is here:
http://www.cdc.gov/NIOSH/ershdb/EmergencyResponseCard_29750039.html