Reducing Silver nitrate to metallic silver

Gold Refining Forum

Help Support Gold Refining Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I got the method 2 years ago from a sterling refiner processing 30,000 ounce batches, we scaled down to 1000 ounces a day because we melt 350 ounces of karat gold scrap in with the 1000 ounces of silver (essentially inquarting) and digest the alloy in 50% nitric. The gold doesn't react and is filtered out and cleaned up in another nitric dissolve and melted with a borax, soda ash, manganese dioxide flux which scavenges any remaining silver and produces a 999.5 fine bar consistantly. The silver is processes as described and comes out 9999+ every time. The gold assay comes from a JM assay and the silver I assay with a Volhard titration in house because we sell the silver ourselves. We take care not to include any bars of gold with PGM's in the process because it will effect the gold purity, as it too will not dissolve completely. There is some on line literature about the method but it is sketchy and I go by the procedure I've outlined above. As with any process, familiarity will allow you to know when the drop is done just like you can look at a gold precip and know by looking that you've gassed enough. (of course I always have the guys check with stannous just in case) Recovery from the acid is complete and accountability is excellent. Just watch the pH, if the solution is stable at 1.5 when you start adding the reducing solution, and the reducing solution has been reduced to 4, you're OK. We make 55 gallon batches of reducing agent so it's on tap when we need it, it has never gone bad. One more thing, the digestion gets hot, which I'm sure you know, so heavy walled 316 stainless tanks hold up the best. Polypro is fine for the mixing and storing the sodium formate solution. Once you see the sponge this drops, you'll be hooked.
 
4metals said:
The silver is processes as described and comes out 9999+ every time.

the silver I assay with a Volhard titration in house

I have several questions, but this is the one that really caught my eye.

How on earth are you able to get enough accuracy from a Volhard titration to say that the silver comes out to be 9999+ Fine? I ask this with some experience, since I have run many 1000s of Volhard titrations in my life.
 
Um, I can buy 99,9% at the top end. Just because you used 20.21 mL of titrant, does not 4 significant figures make!!

It's possible I suppose, but I would not feel comfortable claiming this. This 10X increase in purity also brings in such factors as:

  • inconsistency of the glass buret's diameter, and therefore volume along a distance

    volume and therefore concentration differences based on temperature fluctuations

    error on the chemist's part in making the titrant, delivering the titrant, sample preparation, etc.

    the lower bound of where the indicating reagent can be seen (I posted on this already).

Can you give me details of your method, I'd like to dissect it.

Thanks!!
 
Here's the method and a cut and paste from the spreadsheet I set up to calculate the working factor for the thiocyanate as calibrated with proof silver. I've calculated the titrants strength and the silver proof weights so a class 'A' pipette is used to deliver the first 75ml followed by up to 10 ml from the capacity of the burette. Now I know that there's an inherent error in pipettes and burettes, and variations of standard temperature and pressure cause volume fluctuations, but I've set up a refinery and lab for a guy who isn't about to spring for an ICP or an AA, and considering the spitting and corrections needed when using a classic fire assay on fine silver the titration produces the most consistent result, and it is reproducable. I am however open to all suggestions.

Volumetric Silver Titration
“Volhard’s method”
The analysis of silver bullion by titration is a very accurate for high silver bullion and fine silver. Samples which are soluble in nitric acid are perfect for this method. The converted silver chlorides from the aqua regia refining process also work well as the gold can be calculated gravimetrically after the parting process and the silver can be calculated via titration. This method eliminates cupellation and proofs required to correct for furnace losses. It also is preferred for fine silver assays providing the titrant is calibrated against fine silver when it is made up.
The basis of the reaction is as follows:
AgNO3 + KSCN = AgSCN + KNO3 (1)
3KSCN + Fe(NO3)3 = Fe(SCN)3 + 3KNO3 (2)
Reaction (2) does not occur until (1) is completed, therefore the addition of a small amount of a soluble ferric salt serves as an indicator by the formation of red ferric thiocyanate in the presence of an excess of the thiocyanate. When the titrant enters the solution it will produce a red streak which will re-dissolve on mixing until all of the silver has been converted and the solution turns red signifying the end point. Chlorides interfere by reaction with silver so all reagents, glassware, and water must be chloride free.
Potassium thiocyanate Prepare and standardize a solution. (M=N) 5 g KCNS Diluted to 1 L. Standardize against 0.5g fine silver.

Ferric alum indicator Dissolve 100g of crystals in 1 liter of distilled water, Add 2M Nitric Acid to clear the solution.
----METHOD----

Dry a small tray containing reagent Potassium Thiocyanate in the drying oven overnight. In the morning place in the dessicating chamber to cool.
Prepare a solution for titrating by dissolving 5 grams of dried Potassium Thiocyanate in 1 liter of distilled water.
Prepare a sample of proof silver (99.99 fine) by digesting approx. 0.5 grams via the procedure below and titrate it to standardize your solution. The validity of this method is you are assuming the silver used to standardize the titrant is fine (proof) silver.
Enter the volume of solution titrated into the spreadsheet, it will calculate the strength of the solution.
Weigh 0.5 gram of unknown sample on the analytical balance and digest in 15 ml Nitric acid. Heat acid to expel the nitric acid. If analysis is for fine silver bullion the 0.5 gram sample should be sufficient.
Dilute the parting acid and rinse waters to 50 to 75 ml and cool the sample in a tray of cold running water in the sink. Warm samples produce errors on the high side.
Add 5 ml of Ferric alum indicator.
Titrate with Potassium thiocyanate to a red endpoint which doesn’t fade on mixing. The trick to this is the speed of the mixing bar. Too fast and the swirling white Silver Thiocyanate masks the endpoint. Slow the mixer down when the endpoint is close so the Silver Thiocyanate swirls lower in the flask and more of the solution, which is where the endpoint reaction takes place, is visible.
Enter the volume of thiocyanate titrated in the spreadsheet and it will calculate the silver purity.
As with any assay duplicates of both standardization and unknowns will increase accuracy.

Standardizing Solution
proof # 1 <><><><> <><><><> proof # 2
0.4243 grams proof Ag 0.4365
76.43 milliliters KSCN actual 78.71
0.005551 grams Ag/ml KSCN 0.005546
0.005549 WORKING FACTOR

unknown #1 <><><><> <><><><> unknown #2
0.4391 g sample digested 0.4265
79.13 ml KSCN titrated 76.86
0.439059096 grams silver in sample 0.426463821
0.999907 assay 0.999915

Average assay 1 & 2 0.999911009
difference between cups 4.16271E-06
 
4metals,

You are obviously a very knowledgeable person and we are lucky to have your participation on this form. I certainly agree that the Volhard method is the best choice for your client but I feel, along with Lou, that the claim of using it to produce silver assays of 4 significant figures is quite extravagant. As with Lou, I would think that 3 significant figures would be considered a worthwhile achievement. Your method is almost identical to the one I used. You have taken almost every precaution I can think of to get the best results.

The number of significant figures in the answer can be no higher than the lowest number of significant figures in any part of your procedure. I learned that in the first week of College Physics.

(1) From the sample and proof weights you provided, it seems that your analytical balance only reads to +/- .0001 grams = 3 significant figures for the .5 gram weighings. Also, there are 2 weighings involved, the proof and the sample.

(2) The tolerance of a 100 ml Class A buret is +/- 0.1 ml. A reading of 90.00 ml would actually be somewhere between 89.9 ml and 90.1 ml. Even if you used the entire 100 ml of titrant, you would be stretching it to get 3 significant figures. As I see it, this is the source of the greatest error.

(3) I assume you achieved a single drop endpoint. I don't know how small the drops are, but I doubt that they are smaller than .02 ml = 3 significant figures.

(4) What is the source of the proof silver? If it is silver that you produced, that would cause a problem. It might be best to buy one of those special high dollar 999.99 Fine Maple Leaf coins.

I'm sure there are others that I haven't thought of.

Also, all these errors are cumulative.

You may be actually getting 999.9 silver from your process. However, if you have based that entirely on a Volhard titration, I think you are fooling yourself.

Chris
 
When I set up the procedure to refine the silver the first production batch was sent out to Ledoux for an ICP. The actual analysis was for impurities and they were subtracted from the total silver so it was an assay by difference. So the original assay and hence the claim was better than 4 nines. When the boss saw the tab for the analysis he said theres got to be a better way. Volhard was the best for much less expense and an inexpensive check for production lots. We use Heraeus silver 9999+ to standardize and run 2 titrations to standardize and for each sample we also run two titrations. The buret is a microburet with 0.01 ml subdivisions and 0.01 plus or minus tolerance. The weakest link is the 75 ml pipette with a tolerance of 0.08 ml. As with any fine metal assay the higher the purity the more difficult the assay.
An old timers indication of purity for fine gold is the cleavage in the bar surface as it cools, exposing the crystal structure. Not quantitative but a good indication, do any of you guys know a trick to infer purity on fine silver?
 
What Chris said ^.


Anyway, 3N is believable from a volumetric titration.

Gravimetric titration is actually far superior and with the right balance and right setup, you can almost take this down to the equilibrium concentration of silver from its thiocyanate salt (which has a low Ksp). You seem smart enough to figure that one out.


Lou
 
With the added info you gave, the potential error is less but, you're right, the pipet is the bad actor. Even though your first batch was proven to be 999.9, it doesn't mean the rest are. If you must produce 999.9, I would get another assay method, in order to back it up. IMHO, the Volhard method just doesn't cut it.

Have you looked into the Guy-Lussac method? It's a pain to set up and a pain to use, but it's considered more accurate than the Volhard method. It was the standard method for silver bullion used by the U.S. Mint. Some use a much simpler combo G-L/V method. They start with Guy-Lussac and finish with Volhard.

There's been a lot of discussion about the "pipe" and the crystal on pure gold and silver on this forum. From my experience, the crystal on silver will appear at least at 999.5 and maybe as low as 999.0.

BTW, the formate refining process is very interesting. I've read about it for years but have never used it. I have found the standard silver cell to be simple and, if it's done right, can produce a consistent 4N product.

How are you melting the silver? Induction? Gas? Was the original sample sent to Ledoux taken before or after melting?

Welcome to the forum!
 
The silver is melted by induction and poured through an induction heated shot box because the jewelery industry, where we do most of our business, uses silver in shot form. We have been getting requests for 100 ounce bars but considering how pure silver spits on cooling we would have to set up an argon blanket to cover the melt and the mold, now we're looking for equipment to set it up. The silver sent out to assay was melted and shotted. All of the feedstock for the refining process is sterling and we assay it with the Volhard titration as well. I prefer the sodium formate reduction of silver to a silver cell because we inquart karat into the starting alloy and produce a high purity gold residue as a side benefit. The only cost is the nitric to digest the base metal in the karat. I have sent out samples of the silver both when we were processing only sterling and again when we inquarted the karat, both assayed 4 nines. We found a lab to assay by difference using an AA and it was cheaper so the boss dealt with it, providing it doesn't become a habit. I've never personally used the Guy Lussac method, it seems more complicated than I'm willing to introduce into our lab, but I will read more about it.
 
I'll have to give this formate reduction a crack sometime. Sodium formate is dirt cheap to make from formic acid and sodium hydroxide. Before I switched over to using hydrazine sulfate and now lately hydrogen gas, I used to use sodium or ammonium formate for reducing Pt, Pd, and the like. It worked well and quantitatively, but I was never happy with the size of the precipitate. Always too fine, even after a prolonged boil.

What's the silver precipitate like?

I'm about to buy some hydroquinone from Shor to use on silver, merely because a little goes a long way and I'm curious if their silver precipitant crystals really are hydroquinone.
 
I just thought I would bring this up.

Hydroquinone is used quite extensivly in anal bleaching.

Lou if you need some and havent purchased yet shoot me a PM ill send ya some.
 
Hahahaha,

What am I supposed to make of that comment? Guess I had it coming after not calling on those hoods. I've got the message, I'll give you a buzz and we'll work it out.

Anal bleaching hmm? :lol: :lol:
 
DNIndustry said:
I just thought I would bring this up.

Hydroquinone is used quite extensivly in anal bleaching.

...


Well, I certainly learned a new term today :shock:

I thought it was a typo, but a little google-fu taught me otherwise... (note to self: if it sounds really weird, think before you google and never start with a picture search!!)


Back on track:
Wasn't sodium formate mentioned somewhere as a nice reductant for silver?
 
Another way you can do it that has the advantage of being simple is dissolve the silver nitrate in distilled water, add just a few drops of nitric acid, and then stir with a sheet of copper. The silver will drop out of solution as you stir. It's a little tedious, but it works.
 
DNIndustry said:
I just thought I would bring this up.

Hydroquinone is used quite extensivly in anal bleaching.

Lou if you need some and havent purchased yet shoot me a PM ill send ya some.

Which end you put the ORP electrode?
 

Latest posts

Back
Top