My results of specific types of IC chips, flatpacks and BGA

[Tzoax]

I prepared concentrate and i rinsed it with water, this is picture from the bottom of beaker.



Then i added silicon dies and fiber "wool like" material (that i previously removed) and i rinsed all of it with water again. It took me many hours to manually separate silicon dies and fiber, next time i will grind everything to ashes because this type of IC chips does not contain internal plates and wires (they have but they are very thin and small) and it just could be milled to powder, rinsed and processed. Now material is ready for AR - and i don't have to worry about tin and lead issues because i removed solder balls manually at start.

 

[cosmetal]

Other reason i don't process BGA solder balls together with BGA chips - when tin from solder dissolves in HCl - form of gold that is released (from it) in solution could easily be lost during washing with water (since gold from solder is not heavy like gold bonding wires). I don't know form of gold that is in solder - it could be fine powder or small flakes - but in any case it is easy to lost during washing - i am not sure if this is true, it just seems logic.

The gold in the solder is alloyed with the tin. Gold is easily dissolved in molten tin, when the solder solidifies gold and other minor metals "precipitates" as tiny particles at the crystal boundaries between the tin crystals. When the tin dissolves in acid the tiny particles of gold is freed from the solder and forms a black sludge, like black gold particles that never wants to settle from a dirty drop

Göran

Goran,

Is this the AuSn 80/20 solder you have been referring to in other posts?

Thanks,
James

 

[g_axelsson]

Other reason i don't process BGA solder balls together with BGA chips - when tin from solder dissolves in HCl - form of gold that is released (from it) in solution could easily be lost during washing with water (since gold from solder is not heavy like gold bonding wires). I don't know form of gold that is in solder - it could be fine powder or small flakes - but in any case it is easy to lost during washing - i am not sure if this is true, it just seems logic.

The gold in the solder is alloyed with the tin. Gold is easily dissolved in molten tin, when the solder solidifies gold and other minor metals "precipitates" as tiny particles at the crystal boundaries between the tin crystals. When the tin dissolves in acid the tiny particles of gold is freed from the solder and forms a black sludge, like black gold particles that never wants to settle from a dirty drop

Göran

Goran,

Is this the AuSn 80/20 solder you have been referring to in other posts?

Thanks,
James

No, this is gold dissolved from ENIG into the solder. Usually < 0.5g/kg or the solder joint becomes brittle. For exact numbers of real world data just check Tzoax's previous posts in this thread.

I might be off on the numbers, I just pulled it from my memory.

Göran

 

[Tzoax]

James, the source of gold in BGA solder balls cames from gold plated pads from PCB. When BGA IC chip is soldered to gold plated pad - solder/gold alloy is made.







I think that 80/20 solder is reffered to braze used to connect silicon die to ceramic chip.

Alex

 

[g_axelsson]

Actually, it is a solder (as has been discussed in another thread http://goldrefiningforum.com/phpBB3/viewtopic.php?f=60&t=26762&p=284049#p284049 ) and was used to seal the lid on ceramic capsules, melting point 280 C. To braze the die to the ceramic body an Au/Si braze was used.

Göran

 

[cosmetal]

James, the source of gold in BGA solder balls cames from gold plated pads from PCB. When BGA IC chip is soldered to gold plated pad - solder/gold alloy is made.

2.png

1.png

3.jpg

I think that 80/20 solder is reffered to braze used to connect silicon die to ceramic chip.

Alex

Alex,

Excellent illustration and pictures. Very much "a picture is worth a thousand words."

James

 

[cosmetal]

Actually, it is a solder (as has been discussed in another thread http://goldrefiningforum.com/phpBB3/viewtopic.php?f=60&t=26762&p=284049#p284049 ) and was used to seal the lid on ceramic capsules, melting point 280 C. To braze the die to the ceramic body an Au/Si braze was used.

Göran

Goran,

Your "thousand words" contain a "million words" worth of expert knowledge. Rest assured that I am not implying that you are "long-winded."

With my feeble mind, sometimes I need to read your posts at least twice before they "sink in."

Never change . . .

James

 

[patnor1011]

Let me rephrase.
I do not really think it should be "processed". I mean there is so little gold it won't pay for the cost associated with trying to recover it. HCl used to dissolve that tin will cost more than any gold recovered from that solder.
I simply put them in AP and let them in till all tin is dissolved. Then I remove IC and continue with what process is decided. As for the little bit of gold left over from dissolving solder, it can sit in AP till there will be enough there to warrant any recovery.

 

[Tzoax]

Actually, it is a solder (as has been discussed in another thread http://goldrefiningforum.com/phpBB3/viewtopic.php?f=60&t=26762&p=284049#p284049 ) and was used to seal the lid on ceramic capsules, melting point 280 C. To braze the die to the ceramic body an Au/Si braze was used.

Göran

Sure it is, thank you for correction.

Alex,

Excellent illustration and pictures. Very much "a picture is worth a thousand words."

James

Thank you James, i am planning to buy USB microscope in near future.

Let me rephrase.
I do not really think it should be "processed". I mean there is so little gold it won't pay for the cost associated with trying to recover it. HCl used to dissolve that tin will cost more than any gold recovered from that solder.
I simply put them in AP and let them in till all tin is dissolved. Then I remove IC and continue with what process is decided. As for the little bit of gold left over from dissolving solder, it can sit in AP till there will be enough there to warrant any recovery.

As i remember you made x-ray test and result was 2,2g of gold / kg of solder balls...
https://goldrefiningforum.com/phpBB3/viewtopic.php?f=33&t=22951&start=150#p270023

Sample I got xrayed was 2,2g/kg but that was just tin balls not a piece of copper traces. All melted together. I still have quite a lot of it. :wink:

My result was 1.42g/kg.

But, after all - does it payoff is determined individually by many factors.

This is a top view of beaker with warm AR solution. You can see that wool like fiber material gets on top surface of solution. After reaction is done i repeated one more time just in case if there is still some of the gold dissolved.



When reaction started i took a snapshot of the bottom of beaker to capture image of gold bonding wires.



After neutralizing solution i cooled it down and filtered it with charmin plug. I think the most of it is silver chloride.



After first filtration there was still very little of silver chloride so i repeated filtration one more time and now i have crystal clear AR gold bearing solution that is ready for gold dropping. Of course i added some water to help SMB drop.

 

[Tzoax]

I finished second test of mixed small BGA IC chips.

Out of 240.5g of IC chips i recovered 1.35g of gold.









Link for details about first test i made with this kind of IC chips:
http://goldrefiningforum.com/phpBB3/viewtopic.php?f=33&t=22951#p241264

Comparing results from two tests:

Test No. 1: 233.7g of IC chips - 1.1g of gold - 4.71g/kg
Test No. 2: 240.5g of IC chips - 1.35g of gold - 5.61g/kg

The difference between this two tests was that in Test 1 i measured starting material with solder balls (that was still soldered to IC's) and in Test 2 i measured weight of starting material without soldered balls (i previously removed them with chisel and i am processing solder balls separately - like separate material).

In short - my results are next - if starting material is weighted and processed with solder balls - 4.71g/kg, and if starting material is weighted and processed without solder balls - 5.61g/kg.

Alex

 

[Tzoax]

Another difference between these two tests i forgot to mention is that in Test 1 i removed silicon dies - i didn't process them, and in test 2 i processed them.

Alex

 

[mls26cwru]

I am just finishing my test up tonight as well on this same type of material... I just have to melt the gold down. I'll add my results and test parameters and a pic or two so we can have another data point!

 

[Tzoax]

I am just finishing my test up tonight as well on this same type of material... I just have to melt the gold down. I'll add my results and test parameters and a pic or two so we can have another data point!

That's great news Mike. In the name of all of us looking forward to see your results - thank you for sharing.

Alex

 

[kernels]

Nice work Alex, those are very high yields for those ICs! I struggle to get them into the 4's consistently. There seems to be a lot of variation in those small BGA packages.

 

[Tzoax]

Nice work Alex, those are very high yields for those ICs! I struggle to get them into the 4's consistently. There seems to be a lot of variation in those small BGA packages.

Thank you Hein. Yes, it seems there are lot of variations of BGA and small BGA IC chips...but there is a confident prediction about their gold content/yields - at least ball park yields - for (big) BGA chips (black tops) in my case there is always about 10+ g of gold/kg. And small BGA's is about 5g/kg (processed with solder mask and without solder balls). I think that if small BGA's are processed without solder mask - only black tops - the results would be pretty same/similar like big BGA IC chips.

Alex

 

[mls26cwru]

Okay, here are the results:

1.6g for the chips in the picture below. I used mostly small BGA from server boards...i pulled out a couple that had metal legs, so they did not get processed. There are almost no RAM chips used in this batch.

Now here is the kicker because I had a brain fart... I forgot to measure the weight before I removed the solder in an HCl bath. After that, I dried them and got a weight WITHOUT solder of 203g.

So final yield on chips alone: 7.9g/Kg or 3.6g/lb. (chip dies and solder was not processed)

If anyone knows an average weight% of solder to chips of this sort, we can make some corrections... otherwise I will get another batch in the works soonish and figure something out.

 

[Tzoax]

Amazing results Mike!

In my last test i also weighted material without solder balls, and i think that most obvious reason for such a difference in yields (5.61g/kg and 7.9g/Kg) is because of ratio of RAM BGA's - in my batch the most of BGA chips was from RAM memories, and in your batch there are almost no BGA RAM chips - comparing size of BGA chips - your batch contains much larger BGA chips than my batch - which lead me to conclusion - RAM BGA chips are lower yield than rest of larger BGA's (from your batch).

More tests like this - better ball park yields prediction. Thank you for sharing.

Alex

 

[Tzoax]

I just bought 24 telecom server boards - 19.4 kg. I paid them 105 euros - 4.34 euros each board - or 5.37/kg. I did paid a lot for them...maybe i will not reclaim the value...but i am intending to process all of the IC's from them and see the gold yields from these telecom boards.

This is list and pictures of boards with weight and quantity.

1. 7 boards - 800g piece
2. 1 board - 400g piece
3. 4 boards - 500g piece
4. 1 board - 900g piece
5. 3 boards - 1100g piece
6. 1 board - 500g piece
7. 4 boards - 1050g piece
8. 3 boards - 950g piece

 

[Tzoax]

 

[cosmetal]

I just bought 24 telecom server boards - 19.4 kg. I paid them 105 euros - 4.34 euros each board - or 5.37/kg. I did paid a lot for them...maybe i will not reclaim the value...but i am intending to process all of the IC's from them and see the gold yields from these telecom boards.

This is list and pictures of boards with weight and quantity.

1. 7 boards - 800g piece
2. 1 board - 400g piece
3. 4 boards - 500g piece
4. 1 board - 900g piece
5. 3 boards - 1100g piece
6. 1 board - 500g piece
7. 4 boards - 1050g piece
8. 3 boards - 950g piece

1.jpg

2.jpg

3.jpg

4.jpg

5.jpg

Alex,

If known, and not too much trouble, can you list the manufacturer's name and the actual, or approximate, manufacture date of the boards?

Thanks!
James