The assay / refining lab..

[4metals]

GSP

I was just the opposite of your experience, of the thousands of assays I've performed 90 % were jewelery related (cupellations and fusions) and the balance e-scrap. Your tips and suggestions, all excellent, are the things I have personally experienced as well but are the things learned by actually doing assays. Like feathers and the way a bead looks if it has Platinum in it.

For this reason I feel strongly that any serious assayer should have a bookshelf with books on classical fire assay at his or her disposal. I've been fortunate to have on my shelves the following books, all well worn, and all providing a well needed direction from time to time.

Fire Assaying by Shepard & Dietrich
Analysis of Noble Metals by Beamish & Van Loon
The Metallurgy of Gold by Sir T.K. Rose
A Textbook of Fire Assaying by Bugbee
The Sampling & Assay of the Precious Metals by Smith

and finally Practical Assaying by Mitchell I have a 5th edition printed in 1881, it has a terrific section on blowpipe assay techniques.

 

[goldsilverpro]

I have the first 4 books, hardbound. I prefer the Shepherd and Dietrich and the Bugbee. I think both have been reprinted and are available at a fair price.

The last one is available here in pdf: HERE
I also have about 20 other old assay books, in pdf, that I downloaded from Google books.

 

[4metals]

Ok now it's time to do some refining in our little lab.

Lets start with karat gold jewelry the no left over method.

The no grief method for refining jewelry scrap.

This method is for refining karat jewelry scrap by inquartation. The silver is recycled and used over and over again to inquart more karat. The plus of this method is you will never end up with un-dissolved pieces of gold, and if you are a jeweler you can semi-refine and reuse the gold without the aqua regia step.

1. Weigh your scrap and either melt it and assay to obtain an exact assay or estimate the fine gold content.
2. Add fine silver equal to 3 times the fine gold content and melt together, do not add flux, and pour the stirred alloy into cold water.
3. The resultant metal is called popcorn shot or cornflake shot mostly because it has a large surface area which helps it dissolve quickly in the next step.
4. Digest the scrap from step 3 in a mixture of 1 part distilled water to 1 part nitric acid.
5. When the reaction is complete filter the solution and rinse in distilled water. The residue on the bottom will be mostly gold but it looks like coffee grounds.
6. Take the coffee grounds, rinse them into a beaker with distilled water, decant the excess water and heat the beaker with the gold in a solution of 1 part distilled water to 2 parts nitric acid.
7. Filter the residue and collect on a filter paper. The second nitric dissolve will have dissolved the majority of the remaining silver and rinsing well here will directly affect the purity of your product. For that reason I recommend vacuum filtration. It will assure the water containing dissolved metals is separated from the gold and can be rinsed clean.
8. The mud remaining in your filter is very high grade gold usually 99+% with the primary contaminant being silver. Depending on what you are doing with your gold will determine what you do next. If you are making alloy for jewelry you can melt the gold with a flux containing Manganese Dioxide which will scavenge the remaining silver and can produce gold approaching 999. This can be alloyed with silver and copper to make nice new alloy, having only semi refined the material.
9. If your goal is 9995 gold, you will now need to refine this high grade gold. If you are refining it further, skip the melting step 8 because the granular gold which looked like coffee grounds will dissolve very fast if you don’t melt it.
10. Before we proceed with the final refining process, let’s attend to the silver from the inquarting process.
11. Mix all of the acid from the first and second nitric treatment, and add sodium chloride (salt) to drop the silver as a chloride. It will come down like heavy white snow. After it settles add a pinch more salt until you are sure no more silver is in solution.
12. Filter the liquid to collect the silver chloride and rinse it well. Here again a good vacuum filtration will benefit the end product.
13. The acid can be treated as it contains no significant PM’s.(waste treatment)
14. The silver chlorides can be reduced to metal by either the frying pan method or the sugar method, Both of which have been discussed on the forum.
15. Now we can concentrate on the final refining of the gold. Place the gold in a mixture of 4 parts hydrochloric acid and 1 part nitric acid. Add the acid slowly and continue adding until all of the metal is digested. This will go quickly because of the small particle size of the coffee ground appearing gold you started with.
16. Add ice to the acid before filtering, this will lower the solubility of the traces of remaining silver and allow them to be filtered out easily.
17. The acid from this step will be a bright red color as all of the base metals were leached out in the nitric digestions. Carefully in a container which is only half filled so the reaction has room to rise, add urea prills. Be careful here, until you have a feel for how reactive urea can be, add slowly and stir gently. When the prills you add begin to float at the surface of the acid and not dissolve, you have added enough.
18. If you didn’t add enough urea to kill off the nitric, it will burn off when you begin to precipitate the gold so it’s wise not to add too much.
19. Now begin to add your precipitant. Sodium metabisulfite will drop cleaner gold than ferrous sulfate. (mostly due to rinsing) Add slowly and stir gently and test with stannous chloride until the precipitation is complete.
20. I have never had to drop out the lead and tin from the acid but some members may prefer to do it and I would appreciate some input from those who have done this more than I have. I take precautions to keep the stannous solution away from the aqua regia, some guys actually dribble it into the acid and at the same time are adding tin. I place a drop of the acid to test on a paper towel and add the stannous to the towel. A drum of burnable paper waste stored in the refinery is burned and processed as sweeps when it accumulates.
21. When the gold is dropped, it is again rinsed with hot distilled water, and then cool distilled water. (Some use ammonium hydroxide here, I usually don’t) Vacuum filtration here helps the purity of the product.
22. This gold can be melted, this time without flux, only a small pinch of borax to assure all of the gold pours out of the crucible will be needed.

As an exercise in learning the art of refining, it is a good practice to follow this method before you start with shortcuts. This method also assures the silver in the karat you are refining is recovered as well, something direct aqua regia does not assure. The chlorides coming off direct aqua regia refining contain gold which has to be recovered as well. I would also melt the gold product of the semi refining first step (at least once) just so you can see just how pure it is at that stage of the process.

 

[goldsilverpro]

4metals,

Everyone does things differently. I know that everything in your procedure will work. However, in several areas, I prefer to do things in another way.

10-14 - Although I have produced and converted silver chloride many 100s of times, I have learned to not produce it unless it is absolutely unavoidable to do so - e.g., the AgCl that is produced when directly dissolving karat gold in AR. Unless the AgCl is processed immediately, and not aged, it is often difficult to get a 100% conversion. If you have several 100 oz of aged AgCl, it can also filter and rinse very slowly. It can become very pasty and it tends to pack tightly in a vacuum filter.

I much prefer to use copper (buss bar is best) to cement out the silver. It only takes 2 steps (cementation, filtering), whereas AgCl takes 4 (precipitation, filtering, conversion, filtration). Copper cementation removes 100% of the silver. You knows it's finished when a drop of HCl or salt water produces no white cloud. Even though cementation only produces 99% silver, at best, that is fine if you're using the silver for inquartation. With aged AgCl, I doubt if the purity is any better, considering the rinsing problems.

BTW, what is the frying pan method?

15-18 - I'm a huge believer in using the HCl and HNO3 separately, rather than premixing them. Generally, the gold powder is first covered with an excess of HCl. The HNO3 is then added in small increments. When an increment stops fizzing, more is added. When a small addition produces no reaction, the gold is dissolved - this assumes that an excess of HCl was added to start with. After doing this a few times, knowing when you hit the endpoint becomes second nature. If, when learning, you overshoot the nitric, just boil it down as normal.

This method eliminates the need to boil the solution down or to add urea. I don't like urea at all for this application. If you add too much, it will form a precipitate. I also felt that I got purer gold when I didn't use urea.

A very slight excess of nitric acid will cause no problems when you use SMB to drop the gold. The SMB will first react with the excess nitric and then it will drop the gold with no problems. If there is much more than a slight excess of nitric, this process won't work well, since it will take a ton of SMB to kill the nitric. I actually prefer using sodium sulfite. It seems to stink less than when using SMB.

16 - Instead of using ice, I prefer to add 3 volumes of water to the AR. This precipitates out the greater bulk of AgCl that was dissolved in the strong AR. It actually might be best to dilute AND use ice. I also prefer dropping the gold from a more dilute solution.

20 - To be safe rather than sorry, I always added a little sulfuric to the aqua regia before filtering. It won't hurt anything.

 

[4metals]

GSP

Thanks for your reply, the copper cementation of the silver is another entirely different approach which brings us to the same end. Excellent.

The frying pan method is when you use dilute sulfuric on silver chloride in a cast iron frying pan. I guess in this age of stainless steel and teflon frying pans I should refer to it as the cast iron frying pan method. I mentioned it only because it can be a benefit to the refiner processing smaller lots.

Concerning urea, I generally do not like to use it, the main reason is that in larger scale waste treatment the urea breaks down at elevated pH to form ammonia which redissolves the copper out of the hydroxides causing discharge limit issues. If the small scale refiners on this forum use the copper cementation followed by the iron displacement the urea will never see high pH so it's not an issue. Another reason for caution with urea is how it can over react and boil over in a second if you're not careful.

The sulfuric addition is a good precaution to eliminate both tin and lead and should be included in the procedure.

 

[goldsilverpro]

4metals,

Have you ever tried using HCl and HNO3 separately instead of pre-mixing the AR? It works great. No urea needed. I haven't had to boil down or use urea in almost 30 years, except the few times I got distracted and added too much nitric. When first trying it, it helps to estimate approx. how much you'll need by first calculating the amount of gold you expect. After you've done it a few times, that's not necessary.

 

[4metals]

GSP

Actually I always add them separately, I add the required HCl based on the rough estimate of 7.5 t.o. of total metal is digested in 1 liter of aqua regia, then I add 25% of the nitric and usually feed the balance from a drip feed. I usually don't have time to sit there and add as required so I never have done it that way. Currently we gas with SO2 so we just burn off the excess nitric with SO2 and don't use urea.

Just out of curiosity, have you ever added Nitric using an ORP controlled reaction. What I'm asking is the digestion controllable by monitoring? Then it is automate able. That could save on nitric, SO2 not to mention the NOx. Just curious.

 

[mike.fortin]

4metals. what is drip tubbing made out of? dont nitric hurt it? thanks. mike.

 

[4metals]

Drip tubing, I use a teflon tubing with a teflon stopcock from a burette, but most plastic tubing will work you just have to change it when it gets brittle. Nitric ages the plastic and makes it stiff so when you notice that the tube isn't as flexible as it used to be, time to change it.

 

[4metals]

Our discussion of bullion cupellation methods set off some interesting discussions, now it's time to discuss crucible assays because we will come across sweeps in our processing of both jewelry and electronics. Hopefully more interesting discussion will result.

Equipment needed for assaying sweeps type materials.

In addition to all of the supplies listed for cupellation you will need;

30 gram crucibles
scorifiers
tongs to pour crucible
scorifier tongs
granulated lead
multi depression conical molds
bone ash (for cleaning up spills in the furnace)
anvil
hammer

The trick with assaying powdered material is to select a flux which can reduce the metals in the sample so they can be collected by the lead. In the jewelery business that requires a "standard flux" a "green rouge flux" and the strong reducing flux for refining residues.

The formulations for these fluxes follow below. The details for mixing the sample and flux are also listed below. When you finish the fusion, you will pour the contents of the crucible into a conical mold to produce a cone shaped lead chunk. If all went especially well it will have an apple green clear slag on top when it cools. The only way you know if a fusion went well, meaning it collected all of the metal in the sample, is if all of the flux and fused sample pours out of the crucible without leaving any small glowing balls of lead clinging to the side or unreacted sweep sticking to the crucible. If the crucible doesn't empty cleanly try the more agressive flux usually needed for green rouge (green because of the chromium) and repeat the process. Keep notes because once you determine which flux works for a customers material it makes it easier to look it up than to figure it out again.

One note of caution, the slag on the cone molds often cools and shatters without warning, It can take out an unprotected eye. I always cover the molds with a board in the lab just for that purpose.

If the material is refining residues, which have been burned, crushed and sifted, the last flux listed is required. This method takes some practice because the metals you are collecting in the lead only amount to a few milligrams. So usually it is only done for gold and silver and assays for Pt and Pd are usually sent out to a lab with instrumentation.



Fluxing Formulations
All portions for 30 gram crucible fusions



Standard flux mixture

50 pounds Litharge (1 full can)
11.662 pounds Soda Ash (5925 g)
5.0 pounds Borax (10 waters) 2270 g
1.666 pounds Silica (756.3 g)
1.666 pounds Flour (756.3 g)

Test for lead reduction every time you make a new batch. Fusion should produce a 35-40 gram button of lead. To increase the weight add ¼ teaspoon of flour to reduce an additional 4 to 5 grams of lead. The total amount of extra flour added to get a 35 – 40 gram button, plus the ¼ teaspoon called for in the typical flux mixture equals the standard addition for this batch of flux.

Green rouge (chromium) flux mix

50 pounds Litharge (1 full can)
4.662 pounds Soda Ash (2116.5 g)
5.488 pounds Borax (10 waters) 2942 g
1.925 pounds Flour (873.9 g)
1.645 pounds Carbon (746.8 g)
1.372 pounds Silica (622.9 g)

Typical non-green rouge sweeps

4 tablespoons standard flux
1 teaspoon Litharge
¼ teaspoon Flour (or standard addition for batch)
Stir to mix and add 3 gram sample
Add Silver 3 grams (no Silver for dore)
Mix completely
Sprinkle 1 tablespoon Litharge on top
Sprinkle 1 tablespoon Borax on top
Fuse at 1800 degrees Fahrenheit for 45 minutes
Remove from furnace and swirl to mix
Continue fusion for another 30 to 45 minutes
Pour into cone molds

Green rougs sweeps, high chromium

Same as above but substitute Green rouge flux mix for the standard flux

Reducing flux for difficult fusions and refining residues

3 tablespoons Green rouge flux mixture
1 teaspoon Soda Ash
1 teaspoon Carbon powder
Stir to mix and add 3 gram sample
Add Silver, 3 grams, (no silver for dore)
Mix completely
Sprinkle 1 tablespoon Litharge on top
Sprinkle 1 tablespoon of Borax on top
Fuse at 1800 degrees Fahrenheit for 45 minutes then add:
Add 1 teaspoon of a mixture of 1 part Carbon and 2 parts Soda Ash
Add 1 teaspoon Carbon
Add 1 tablespoon Borax
Swirl to mix and fuse for another 45 minutes
Pour into cone molds


Notice all the instructions end by saying pour into cone molds. That is because after the cone mold the lead is cleaned of any slag, hammered into a cube shape, and cupelled. From this point on you follow the directions of a cupellation assay. The hammering into a cube shape helps remove any adherant slag so only lead carrying the PM's goes on the preheated cupels.

Of course you should perform all assays in duplicate to confirm accuracy.

These fluxes and methods are more tuned towards jewelry sweeps but don't differ greatly from sweeps generated from other processes. The big difference is the fluxing formulations. This is where the guys with expertise in e-scrap should jump in with fluxing formulations for e-scrap. The way I assayed e-scrap was as copper based bullion from a melt at a refiner specializing in that. Those assays involve a scorification in granulated lead before a cupellation so no fluxing was ever needed in my experience.

 

[goldnugget77]

Hi Folks and 4metals
Problems with the calculations for silver and gold

2. Weigh a sample of approx 0.5 grams of karat scrap and place in foil boat
3. Weigh out 1.5 grams of fine silver (to inquart the gold) and place in the foil boat
5. Prepare 2 samples for a gold assay in the same manner, if you need a silver assay prepare a foil boat
the same as above but do not add silver.

Now I have
3 imaginary samples
Sample 1: 0.5 grams of karat scrap + 1.5 grams of fine silver = Total weight 2.0 grams

Sample 2: 0.5 grams of karat scrap + 1.5 grams of fine silver = Total weight 2.0 grams

Sample 3: 0.5 grams of karat scrap = Total weight 0.5 grams

16. It is now time to part the beads in the acid hood. The only beads that are parted are the samples you
added silver to.

The beads that you didn't add silver to are usually gold in color, they are used to calculate the silver
content of the sample by subtracting the average weight of gold from the 2 samples you are assaying
for gold. The calculations will be at the end of the parting section.

Now well imagine that what remains is the following
Sample 1: Total weight 0.3grams

Sample 2: Total weight 0.2.75 grams

Sample 3: Total weight 0.5 grams


This looks wrong
For example
For sample 1
30 divided by 2
I am getting 15
That looks wrong


How would I calculate sample 2
When the amount is 2 and 3 quarters

31. To get a silver number weigh the small bead you did not part. Divide the weight of the bead by the starting
weight and multiply by 100 to get % gold and silver in the sample. Now subtract the average gold percentage
from step 30 from the gold and silver percentage to get the silver percentage. Reported as silver by difference.

This is something I am not clear about

If you need it I can post an Excel spreadsheet for the calculations.

This would be a very big help
Thanks

 

[goldnugget77]

Is the purity of sample 1
600% pure

and sample 2
550% pure

 

[4metals]

The problem you are having is with decimal places. However the reason for running 2 cups for the same sample (duplicates) is to be sure the results are close. Your theoretical example has a spread too large to have faith in either the sample or the assayers technique.

I am out of town and my mini Dell does not have Excel and this site doesn't accept the file-type my spreadsheet produces. I'll post an assay calculation with silver correction when I get home late Friday.

 

[4metals]

OK nugget here's the spreadsheet in Excel 2003

A few things, first the sheet is protected so you can only input data into the yellow colored cells AND you can change the correction factor. I put a correction factor of .999 in the sheet but a good assayer runs a correction with every furnace load. That being said, you can run a proof and put your own correction factor in. If you need to know how to run proofs let me know and I'll post the technique.

The sheet is also locked so you can't accidentally erase or delete the formulas but it is not password protected so you can unlock it and edit it if you like.

 

[goldnugget77]

Hi 4metals
Thanks very much for that program
I tried it out according to the samples I prepared before and I couldn't do it
Here is how it looks

image posted is no longer available but I saw them and responded below. (4metals)

 

[4metals]

Decimal point , decimal points, decimal points. When you put 2 decimal points in one number it is no longer a mathematical expression. All starting weights should have decimals too unless they are exactly a whole number.

 

[goldnugget77]

Hi 4metals
Thanks for taking the time to help

Here I tried one more time
Using an imaginary batch
Did I enter all the yellow fields in the right way
Also is the result of the assay
14.86K for gold
and silver is
40.5%
is this correct
Thanks again

 

[4metals]

Gold sample weigh is the weight before you add silver. In your original question it was 0.5, which would make it 60% before correction. Plus the dore will always lose weight unless it is all precious metals with no base metals when you started.
Using the numbers you started with the answer is 14.86 PERCENT not Karat all answers in %

 

[4metals]

The silver dore weight is .4 there is no silver added to this cupel because you are looking to find the silver content by difference. So you subtract the weight of the gold determined in the 2 cups parted to get the weight of the silver because the .4 is made up of the gold plus the silver in the original sample, and divide that by the start weight to get the silver percentage.

 

[PreciousMexpert]

Hi 4metals
The starting sample weight is 0.5
shouldnt the amount
for
coronet wt Dore
be also
0.5