Refining scrap silver

I had 300g of impure sterling silver. This resulted from a mould leak in the bottom of my furnace, so I re-melted the salvaged metal and poured it as grain into water. Subsequent testing indicated that it was probably too impure to use - not surprising considering that the same furnace is used for casting leaded bronze, and that has also leaked on occasion!

300g of this impure silver was added to 150ml distilled water and 300ml commercial nitric acid in a flask with reflux tube; the bottom of this went below some sodium bicarbonate and diluted ammonia solution. It was initially cooled with a water bath to reduce the output of fumes, and then later heated in the same way (to speed up the reaction). The sodium bicarbonate seems to be irrelevant – and the container of dilute ammonia solution should be placed in a larger container holding more ammonia since the white bursting bubbles actually contain some orange gas.

Subsequently I found that one can better scrub NOx gasses in either 10% sodium hydroxide + 3.5% hydrogen peroxide (Nikon process) or 0.2% hydrogen peroxide and 10% nitric acid (Kanto Denka process). An intriguing remark from www.h2o2.com suggested that peroxide could be added to nitric acid in many metal pickling applications, thereby removing NOx at source and regenerating the acid.

After about 40 minutes when bubbling was very slow, the nitric acid / silver nitrate solution was filtered, the excess silver washed and retained, and the silver then precipitated from the liquid with salt (sodium chloride) solution. After a false start doing filtration, I then continued by decanting the liquid and rinsing the white precipitate with boiling water several times, until there was no blue colouration with ammonia (which would be caused by any copper contamination).

On standing, this white sludge would slowly develop a violet-grey surface due to the action of light. I initially tried reducing the silver chloride to silver with aluminium wire and hydrochloric acid; this caused some continuous darkening as silver was precipitated, but seemed very slow and to require much more wire and time. So the wire was removed, the sludge well washed, then finally had 60g dry sodium hydroxide and 30g dry glucose powder added. The glucose was caramelised (judging by the smell) and the sludge turned more or less unevenly grey with pink-grey parts as the chloride was reduced to silver (and regenerating salt solution I imagine). After well washing, I decided to repeat the process, which got rid of the pink grey parts.
After drying, the final result of my labours was a large number of silver stains on workroom surfaces (and fingers), and about 280g of fine light-grey granules, looking very like cement powder. It took a lot of Internet research to find a suitable way of removing the afore-mentioned stains - in the end I used acidified potassium permanganate followed by acidified sodium bisulphite solutions.

280g of this powder after drying was packed into a ceramic shell crucible with a large lump of coke, and heated in a propane-fuelled furnace for 15 minutes to around 1000C. This resulted in the evolution of large amounts of what was almost certainly chlorine and / or hydrogen chloride, which required ammonia solution in the workroom to help neutralize it. Fortunately I was wearing an adequate mask. Additionally, there was a constant rather large blue and green-blue flame coming from the top of the furnace, something that doesn't normally happen.

The crucible was removed, quenched and broken apart to reveal my two slab castings each about 5mm thick, totalling only 260g – so potentially there may have been up to 20g of chlorine present as chloride or otherwise! A strip of one slab, about 10mm wide, was rolled down to 1.5mm, showing that work hardening was very minimal – even rolling a slab 50mm wide was quite easy in a rolling mill without gears. Two large cavities at the upper ends of the slabs had very intriguing triangular ‘bird foot’ shapes made of minute granules – presumably some kind of crystallization phenomenon. They were so intriguing that I decided to keep them, and added the attached pictures. The one above on the left is a plan view, and the one below on the right is a near-elevation, showing 5 stubs radiating from the centre of a 3-axis structure. The sixth runs into the body of the silver. I guess these would make sense for something belonging to the face-centred cubic system, crystallising from the centre along perpendiculars to each of the six faces of a cube.