With appropriate wash procedures and expert handling it is
entirely possible to get over 4N gold with a
single precipitation.
The finer the atomized product you produce and the more you leach the base metals, the purer it will be. On the flip side, the finer the silver chloride powder produced, the more prone to dissolution, and the more likely it is to hold onto your gold solution thus necessitating more dilute HCl rinse which drags silver into the solution....
The real secret to high purity gold is in fact dilution of solution and correct chelating ligands.
Pretend, for instance that you had had a mixture of oranges, lemons, and limes in a large pool of spectator fruit (ions) and water. The lemons are the gold (III) atoms, the oranges and limes are other metals that would contaminate the gold upon its reduction and subsequent nucleation. If you have more dead space (i.e. water) the likelihood of limes and oranges being around the lemons when the lemons are agglomerating together is decreased. This means the limes and oranges don't get stuck with the lemons as they come together because there are less of them present per unit of volume. It's a statistical mechanics problem. Also, the temperature at which one precipitates will also affect the quality because the free kinetic energy of the solution will determine how fast a precipitate forms--generally speaking, the longer the ripening of a precipitate (this is modeled best numerically via spinodal decomposition processes and a nasty set of PDEs), the more prone a precipitate is to have extremely high quality because defects in the forming crystal lattices are less common. Consequently, this means there are less opportunities for a silver or copper or lead, or whatever else ion to "sneak" into the structure. The density of the precipitate also has a bearing on quality--usually the spongier something is (with higher surface area) the more likely it is to adsorb and absorb undesirables. This temperature effect is especially cogent with certain PGM compounds (most of which are seldom seen in refining, but they have structural morphisms that are temperature dependent).
Copper, silver, and palladium are the chief offenders that need to be removed from gold. Why? Palladium has a similar reduction potential and can be dropped with SO2. Copper is liable to produce its cuprous chloride which needs removed. Silver forms its chloride which can also form a soluble complex if the temperature is high enough and the solution saturated enough with chloride anion (this is counterintuitive to what most people would recall of the common ion effect). That's why after the digestion is done, ice is added. Obviously, the lower the temperature, the lower the activity of the solution, the better the removal of silver and other insolubles. There is, however, a trick to the volume increase--you can't add so much that the amount of chloride anion is minimal because that will encourage more silver to go into solution. It's skewed towards low temperature, high dilution. The colder and to a certain extent, more dilute one gets the aqua regia solution, the better the silver will be removed. Obviously, there's a balancing act on how dilute and how cool one can go (more dilute means less cool, more concentrated, means colder). All of this can be modeled based off of existing know-how with solute-solution equilibria and then backed up empirically.
So, how then to get the purest gold in the fastest turn around?
This is the
modus operandi that I consider the best (for karat):
- 1.) Incoming material is melted inductively, a sample pulled and run via fire assay (with nondestructive XRF before hand)
2.) This material is heated and poured into a preheated tundish (which must be heated in its own right)
3.) It is struck by high pressure water hitting the stream of metal at a certain point in its free fall
4.) The material is effectively atomised into a fine powder (think of a beach sand with a slight distribution in size) but NO finer, as it makes settling over long and requires flocculating agents
5.) Fluid decanted or pumped off via diaphragm pump and passed through a cellulosic filter to recover a small quantity of highly fine materials
6.) Granulated product is digested in aqua regia, or HCl-Cl2 mix with careful additions made to prevent a froth over due to high surface area
7.) After digestion is complete (evidenced by Au(III) ion concentration plateau in spectrophotometer, reactor is emptied, shot of sulfuric acid added to remove incidental lead from solder
8.) Ice is added and allowed to mostly melt, this dilutes the solution and drops the temperature, decreasing Ag+ content in solution
9.) Filter pressed with ice cold 0.05 M HCl rinse (look for filtrate to run clear and neg. with stannous)
10.) Solution is treated with various proprietary chemicals
11.) Reduce with whatever reducing agent is desired, preferably while solution is cold (for LSO2 at least).
12.) Filter press the gold, rinse with HCl and ammonium hydroxide.
13.) Heat lamp til dry
14.) Pack in crucible (residual ammonium chloride from rinse is beneficial in removing any last base metals; a tad bit of high purity nitre and borax can also be added) and melt
15.) Weigh and assay
There should be no evaporation to remove nitric. In fact, a degree of free nitric acid can be perfectly tolerated, but really, it's foolish to use more nitric acid than is required, or any more chlorine or chlorate (neither of which are detrimental, but more chlorate means more chloride which means more dilution to rid the silver). Wasting reagents is wasting money. This is why less filtrations are better, less manipulation is better, less of anything that costs time and effort and doesn't have a justified return is far better.
If one was under a less tight schedule but wanted quicker accountability on the silver and better over all purity, one can leach the powder with hot 50:50 nitric acid to remove the silver and the base metals
and then cement the silver and run an ICP to check for gold (tech grade nitric acid sadly has some Cl- in there which will make a very weak aqua regia but will still dissolve very small portion of gold).
Lou
