Smelting gold and iron

[SapunovDmitry]

Guys, i need some help once again.
I have recently smelted some .925 gold fillings with a small amount of iron fillings (apx. 25% of iron in mixture).
The flux was the following( per 100 gr. of fillings): 35 gr. of borax, 5 gr. ammonia chloride, 2 gr. of silver).
I ended up with the bar with certain magnetic properties and white top and yellowish bottom. Since i cover my crystalizers with antifire blanket the cooling rate of the metal is low and bar had time to settle and stabilize.
Is that correct that iron and gold are separated in this bar?
The question arises since i see white top and yellow bottom in the bar.
How do i remove iron from the bar through smelting ( i don't want to use aqua regia and other types of wet chemistry).
I can use sodium nitrate but i have graphite crucible and i don't want carbon to be oxidized in the process.
Also i could use ammonia chloride again but am afraid of significant gold losses since hydrochloric fumes could be oxidized into chlorine +water and it could affect the bar. Besides ammonia acts at 340 C and at that temperature gold is not liquid and only surface iron could be removed. I think it is not a problem if most of iron is deposited on the top of the bar (white stuff) but if i am wrong gold losses could be high.

Has anyone any experience with similar cases?
Advices are strongly welcome.

 

[Harold_V]

It is assumed that because silver won't alloy readily with iron, you should be able to melt the material with more silver. As silver has a greater affinity for gold than iron does, the gold should report in the silver, with a layer of iron above. This is one of the rare instances where you can separate metals by melting.

I have not tried this exercise except with a sulfide compound. Iron completely reports in the sulfide layer. When you pour, you get three distinct layers. The bottom one is the values, the mid section is the sulfide, and the top layer is the flux. I recommend you use a cone mold.

Harold