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

[Tzoax]

I am preparing the next batch of IC chips to test - Flatpacks 4-sided square thick 27x27mm.


I already tested them once (second page on this thread) and i managed to get 0.7g of gold from 576.5g of this kind of IC chips, that is about 1.21g of gold per kilogram.


The differences from testing then and now:
-i collected about 3 times more chips of this kind than from the first time testing, now i have 1635.3g so the results will be more accurate
-i will be using a new way of processing them, like described
-my skills are better than before, so i expect less losses
-i will recover gold from every part of chips - silicon dies, iron and copper wires, heatsinks, powder with gold bonding wires.
-i will picture step by step of what i am doing to make it more understandable.

After burning the IC chips with torch, i baked them on a hotplate until they becomes white and soft.


After several sieving, and separating the wires, heat spreaders, and silicon dies from powder this is the picture of materials that i will process.


I will post the pictures as i progress.

 

[kernels]

Looking forward to seeing your results from this, will be interesting to see if your improved processing method leads to 30% improvement in yield here !

 

[Tzoax]

Looking forward to seeing your results from this, will be interesting to see if your improved processing method leads to 30% improvement in yield here !

Thanks, we will see. Today, i started to make a concentrate out of ash/powder containing gold bonding wires.
So, after first water wash, the volume of powder was about 350ml, i dried the powder, incinerated again, sieved again (separated some more copper/iron wires), and now i am washing again, the volume of powder is now about 50ml. I will continue making the concentrate less volume as i can because the less carbon in concentrate - the more gold will be at the end (because of carbon absorption of gold ions from AR solution).

 

[Tzoax]

This is a picture of final concentrate - 70.8g of concentrate out of 1635.3g of chips.


In a 2l beaker i added 300ml of 16-18% HCl and gentle boiled for about 2 hours.


After decanting HCl i washed the concentrate with a lot of water several times.


I made poor mans AR with a 300ml fresh 16-18% HCl and 2 spoons total of KNO3, adding a few grams at time until all of the gold bonding wires dissolved. Neutralized with urea. The picture after decanting AR solution.



Filtering cold AR solution.


That is what i have done so far... now i am dropping gold with SMB , after washing i will post result.

 

[Tzoax]

The result is 2.4g of gold out of 1635.3g of this kind of chips.


Previous results when processing with diluted nitric acid method:
0.7g of gold out of 576.5g of chips === 1.21g of gold per kg of chips
Now:
2.4g of gold out of 1635.3g of chips === 1.46g of gold per kg of chips

That means 0.25g more of gold per kg, or 20.66% more gold!

This is only from concentrate, i will process the wires, plates and silicon dies for more gold to add to result.

 

[danieldavies]

excellent work. keep it up.

 

[4metals]

So your concentrates ran 3.38% gold and the starting material ran 0.14% gold.

So a ton of this material would yield 45 ounces.

I guess chip picking is a worthwhile pastime.

 

[Smack]

I got 2.6 g per lb. from the last batch of N/S bridge, that's just from the black epoxy tops.

 

[Tzoax]

excellent work. keep it up.

Thank you, i will do my best.

So your concentrates ran 3.38% gold and the starting material ran 0.14% gold.

So a ton of this material would yield 45 ounces.

I guess chip picking is a worthwhile pastime.

In my opinion the best benefit of knowing how much gold each ic chip type have is when choosing to buy any electronic scrap. That way i could know approximately the value of ic chips on the boards, especially when i see telecom or medical boards first time in life, i am forming price/value of the board depending of my tests/knowledge so i choose to buy better stuff no matter what type of e-scrap is, chips are chips, i mean same type of chips are same type of chips. Couple of times i had opportunity to buy a lot of boards, for example couple hundred of poker boards, 30 cents each, but i wasn't sure is it worth, i saw that boards have lot of one type of ic chips but i had no idea how much it is worth. Then i got idea to test each type of ic chip and memorize yields, and next time i would know what and why to buy.

I got 2.6 g per lb. from the last batch of N/S bridge, that's just from the black epoxy tops.

Thanks for sharing!
I guess you are weighting chips along with fiber base, otherwise you should have about double better result.
I take off the black tops-chips, and based on their weight i always have about 11g per kg.

 

[Tzoax]

I am processing silicon dies - 18.2g out of 1635.3g of chips.

Gentle boiled for about 1 and a half hours, with 150g of 16-18% HCl and few grams of KNO3.

After decanting and neutralizing with urea, i added little cold water, AR solution begins to sediment silver chloride.

The look of processed silicon dies.

After totally cooled, i filtered AR solution.

I heated beaker, added 1g of SMB, stirred, now i am waiting gold to settle. There is so little of gold, settling down will take some time...

 

[g_axelsson]

Have you tested the solution? There might be some palladium still in solution from the silver braze.

Göran

 

[Tzoax]

Have you tested the solution? There might be some palladium still in solution from the silver braze.

Göran

Thanks. No, i am adding all remaining solutions after gold drop to a stockpot to accumulate PM's, then i am dropping it with copper powder. Now i am focused on gold, i am not testing silver and palladium content in IC chips, one day maybe i should. For now i am saving all of the residues from IC chips (copper and kovar based internal wires, heat spreaders, and silicon dies for future refining. I am curious how much of gold bonding wires are trapped in these wires, all along with gold from wire joints, so i decided to process all of the gold from this kind of chips i am testing wright now (27x27mm thick flatpacks) and as i am waiting for gold to drop (from silicon dies) i separated the copper (left) and kovar (right) based wires for refining.
In this batch of chips (1635.3g) there is 73.1g of copper based wires and 18.9g of kovar based wires.

 

[Tzoax]

This is a picture of dropped gold from 18.2g of silicon dies, it is obviously very small amount of gold. I also processed the 18.9g of kovar based wires in poorman's AR, it also looks barely visible gold after drop. When i finish processing 73.1g of copper based wires and heat spreaders i will merge all of that gold, and add to obtained 2.4g of gold. For now, it seems that a GREAT majority of gold is in a concentrate. I was hoping that there would be some more gold in this wires because of the joint points between gold bonding wires and kovar/copper wires, it was logic to me that when gold bonding wire brake from base metal wire, there must be some gold residue on it, but for now it looks that it is so small amount of gold comparing with concentrate.

 

[kernels]

I love your thread man, really enjoying reading your updates, doing some great work!

 

[jason_recliner]

Not sure the size of your beaker, I'll take a stab that this looks in the realm of 50-100mg. Not much for the work, perhaps, but definitely interesting.
I would like to say that I really appreciate this thread and your objective posts.

 

[Tzoax]

Thank you for support!
I just finished testing gold content in all residues from ic chips beside concentrate, and my conclusion is that there is minor gold content, in my opinion it is not worth processing (only) for gold, but if processing to recover silver and palladium, it makes no reason not to recover the gold along with it.

 

[archeonist]

This is by far one of the best posts on this forum! Thanks so much for your effort. I am picking chips for a year now and I now have a reference of what I can expect from processing them. It really is fantastic that you shared your method in creating 30% more gold from refining. Once I process my chips I will post some results here.

Thanks again!

 

[Tzoax]

This is by far one of the best posts on this forum! Thanks so much for your effort. I am picking chips for a year now and I now have a reference of what I can expect from processing them. It really is fantastic that you shared your method in creating 30% more gold from refining. Once I process my chips I will post some results here.

Thanks again!

Thank you archeonist! I am glad that my results will be useful to you. Right now i am preparing next material for testing - SOLDER BALLS FROM BGA CHIPS. It is one of the most interesting material for me because i always wanted to know how much gold there is, and there is so little information about it. Thanks to Patnor's document he provided in his thread here http://goldrefiningforum.com/phpBB3/viewtopic.php?f=52&t=11827&p=211524&hilit=bga+solder+balls#p211524 there is about 2.24g of gold for kilogram of this material. So i started today removal of solder balls from BGA chips, i will collect couple of hundred grams and i will test it.

 

[kernels]

Looking forward to see you results from these.

 

[Tzoax]

This is a picture of starting material inside of beaker.



I added 20% HCl and started to simmer boil the solution. This is picture of solution after 2 hours.



This is a bottom view - the solder balls are slowly dissolving.



This is top view - the crust formed.



After 3 hours of total simmer boiling i stopped the process since of lacking my free time, i will continue today. In a meanwhile it looks that most of the tin dissolved. When solution cooled the lead crystals formed like in the picture, probably some of the tin crystals either.



Today i will continue simmer boiling for few more hours and i will filter the solution while still hot to get rid of most of the tin and lead. After that i will boil the material in water and filter again.

My plan is that i repeat the whole process again with fresh HCl to make sure the most of the tin and lead is dissolved and filtered. After that i will incinerate remaining material altogether with filter papers and process it with AR.

If there is anything near 2g of gold per kilogram of solder balls that means it is a very good material for saving, easy to find (graphic cards, motherboards, newer RAM memories etc.) and easy to collect. Also, one of the best benefit in my opinion will be gold recovery from solder balls from flip chips (the ones with silicon die at top that does not have gold bonding wires) - if there is about 2g of gold per 1kg of solder balls - flip chips will be a great and rich source of solder balls. I have a bunch of them, if this test shows good results i will start saving solder balls for future refining.

Alex