Process for ceramic DIP IC's

[snoman701]

I just want to make sure, I've spent a while searching the forum and can't find a process that's agreed upon.

1) Crush to expose bonding wires, silicon chip, and hopefully au braze.
2) Roast lightly to remove traces of oils or other chemicals that could interfere with process
3) Digest tin in heated HCl bath
4) Decant HCl, wash chips, test wash/HCl with stannous. Roast to rid chips of chlorides.
5) Digest copper and other base metals in heated nitric, diluted to 30% or so
6) Decant nitric/cu nitrate, wash chips in DI, test wash/cu nitrate with stannous
7) Add dilute HCl to cover chips, heat
8) Add small amount of nitric acid (1 mL / g expected yield?) to create aqua regia, when all visible braze has been dissolved, decant AR & wash chips with DI.
9) Test AR to confirm Au uptake.
10) Drop with SMB

I'd be appreciative of any feedback on process, before I start.

 

[Geo]

1. Break chips into pieces small enough to ensure the acid can get under the lids (in half).
2. Place chips in container that holds them with at least two thirds of the container free space.
3. Add HCl to almost half the container volume.
4. Add water to bring liquid level to half the volume of the container.
5. Add nitric or nitrate slowly and in increments.
6. Keep adding nitric or nitrate until either all the metal dissolves or the addition of more has no reaction.
7. ensure all the metal has dissolved.
8. Decant the solution into a clean vessel leaving all solids in original vessel.
9. Re-treat the solids with a small amount of AR to ensure all the metal is dissolves. An excess of oxidizer is recommended for this reason. Add this to the original solution.
10. Denoxx and drop the gold with ferrous sulfate preferably.

This should be the fastest way besides cyanide to get the gold out. Don't worry too much about any tin as the excess oxidizer will overcome it. There's no need to pre-treat them even though you can. It just creates more waste in my opinion.

 

[snoman701]

Why drop with ferrous sulfate?

 

[Topher_osAUrus]

Its better for dropping from dirty solutions than smb.
The contaminants that ppt with copperas dont drop with smb.

...man, i should have thought of that in my reply to aga..

 

[Geo]

When dropping gold from a very contaminated solution using SMB, the sulfur dioxide gas produced can precipitate copper sulfate. Ferrous sulfate drops the gold in a much cleaner state but still not very pure. Once it has been precipitated with ferrous sulfate, it can be dissolved with HCl and bleach or AR and dropped with SMB relatively pure.

 

[Topher_osAUrus]

Geo, Do you know what ph range that happens with copper sulfate?
On the chart steve posted it seems copper chloride is in a bit higher of a ph range.


Knowing that answer isn't really that big of a deal I suppose, just more of a curiosity. If there was a particular way one could manipulate a mildly dirty solution to reduce the chance of coprecipitation.

Edit: I read that terribly wrong, i was reading off the ph range for Pd ppt from a chloride solution.

 

[nickvc]

I believe if you read Deanos threads he alludes to a method that will always drop 999 gold if you do so at the right PH :idea:

 

[Geo]

It actually takes a lot of copper in solution and a lot of SO2 created. When both are present at the same time, you will get copper to drop. If the PH is low enough, CuCl stays in solution. I must be a caveman when it comes to refining. I seldom try to finesse and adjust PH unless it is absolutely necessary. I seldom use base material and opt for dilution. I can always evaporate excess water.

 

[g_axelsson]

It actually takes a lot of copper in solution and a lot of SO2 created. When both are present at the same time, you will get copper to drop. If the PH is low enough, CuCl stays in solution. I must be a caveman when it comes to refining. I seldom try to finesse and adjust PH unless it is absolutely necessary. I seldom use base material and opt for dilution. I can always evaporate excess water.

Copper sulfate is water soluble so it will not drop with SO₂ (SMB) It is CuCl that is dropped as a white chloride.
The SO₂ is reducing water soluble CuCl₂ into CuCl and if there is too much water and not enough Cl-ions it will drop as a white precipitate.

CuCl is soluble in HCl or other high chloride solutions, including saturated salt (NaCl) solutions. The pH has nothing to do with it.

Göran

 

[anachronism]

I believe if you read Deanos threads he alludes to a method that will always drop 999 gold if you do so at the right PH :idea:

Yeah we've both alluded to that on a number of occasions Nick but I guess people like dropping their gold dirty.

 

[Geo]

I agree. I apologize if I have been misleading. I am recovering from the flu and am not very focused. I remember discussing this before some years ago. I can't remember how the discussion ended. I was getting more brown precipitate than there was suppose to be gold. It turned out to be copper. For some reason I was thinking copper sulfate but it could have been copper oxide. Which ever it was, using SMB in a copper saturated solution leaves you with very contaminated gold. If you look back at the first gold bar I made and posted, it turned out being tested at 18K and the rest was copper. I boiled the gold powder in water but at the time, I didn't wash using acid so it could have very well been copper oxide.

 

[snoman701]

Ok, still having difficulty knowing when to stop adding nitric.

I have no problem getting gold to go in to solution. Stannous proves that, unfortunately, getting the legs on the IC's to dissolve completely took forever. I heated the solution gently (glass was hot to the touch). What I thought was a lack of nitric acid, must have been passivation, or just a slow reaction.

I waited until there was no more macro bubbling, then added a little more nitric. 1 mL increments.

I added too much nitric.

Instead of fighting it with ferrous sulphate, I denoxxed by dropping in a piece of copper busbar before I went to bed.

My expectation was that when it was denoxxed, it would stop dissolving the busbar. At that point, anything below copper on the reactivity series would be out of solution. I could then decant off a majority of the solution, and then dissolve the gold. Well, I needed to go to bed, and left the copper busbar in there. When I got up this morning there was basically a brown sludge in the flask...the entire busbar dissolved.

The good thing is that the gold dropped, as the liquid that I filtered out of the copper sludge tested negative.

The bad thing is that an unknown amount of gold is mixed with copper powder. Sucks, as I really want to know how much gold i can expect per pound of this specific IC.

So I'm scrapping this batch and starting over. By scrapping, I mean I've got the gold and the copper in my filter papers...I'll incinerate the filters next time I'm melting silver. I can then add the copper and gold mixture to my silver, basically inquarting it. I don't care to waste the nitric right now.

So, starting over with a new batch of IC's. I even broke the legs off these ones, so all that's there to dissolve are the copper traces in the ceramic shell, the bond wires and the braze. How much nitric should I realistically expect to use per pound of chips? How long should it take for an addition of nitric to become exhausted? I was basically waiting for it to go from the big bubbles (macrobubbles) that one sees on fresh nitric addition to the microbubbles that are basically a simmering solution.

Do the acids (just HCl and HNO3) ever react with the ceramic components, or are they quite inert?