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

[theitalianhenchman]

I'm fairly sure he did do BGA chips from ram early on, on page 1.

The circuit boards are generally flash Gold plated before the BGA chips are reflowed on, the very thin Gold plating dissolves into the solder during the reflow process, that is what he was trying to recover by dissolving away the Tin.

That bga where from different sources, not only ram, as he writed some comments after, I'm curious because some people says that from ram bga it's up to 5grams per kg, while other say it is barely 1g/kg. Now I understand that it varies from manufacter, age, and so on, but a so big difference seems too much.

Thanks for the explanation; I didn't know that the board were gold flashed and I thought that the only gold from the solder comes from the bonding wire beeing soldered to the board.

 

[kernels]

I'm fairly sure he did do BGA chips from ram early on, on page 1.

The circuit boards are generally flash Gold plated before the BGA chips are reflowed on, the very thin Gold plating dissolves into the solder during the reflow process, that is what he was trying to recover by dissolving away the Tin.

That bga where from different sources, not only ram, as he writed some comments after, I'm curious because some people says that from ram bga it's up to 5grams per kg, while other say it is barely 1g/kg. Now I understand that it varies from manufacter, age, and so on, but a so big difference seems too much.

Thanks for the explanation; I didn't know that the board were gold flashed and I thought that the only gold from the solder comes from the bonding wire beeing soldered to the board.

No worries, bonding wires are never soldered to boards as far as I know, I believe they use a ultrasonic welding process to attach bonding wires to the die and then to the pads / legs inside an IC.

I have processed a lot of BGA from RAM over the last year or so, it has dramatic variation, normally I expect over 2g/kg, best I've got is over 6g/kg. The biggest variation that I get with any e-waste.

 

[Tzoax]

I had a very little free time last couple of weeks...but i am back now, and currently i am separating gold bonding wires from the concentrate - the remaining concentrate is incenerated/grinded/washed again until all of the (the most of) carbon is gone. So in next week first yields results will be there. On this pictures you are seeing 4.5g of gold bonding wires collected (so far) out of concentrate from 536.6g of BGA chips.

 

[Tzoax]

First test - Big BGA IC chips - from graphic cards and motherboards (south/north bridges).
All of the fiber (green) bases was previously removed with heat gun.



I just prepared materials for testing on the way described in the next picture. The reason i separated everything is that i will process iron first, then (when dissolved) i will pour aqua regia (from iron beaker) to another beaker with silicon dies, when done - from that beaker to another with concentrate, and finally i will dissolve gold bonding wires in the last beaker. Maybe all of this sounds crazy but...i wonder will this affect on yields... since (doing this way) i think i reduced chance for something go wrong to minimum. Maybe those small things i changed will affect on yields, maybe not, but i think it is worth to try.

 

[cosmetal]

Tzoax,

I am still learning. So, please be patient with my questions regarding this portion of your excellent thread.

How did you remove the carbon from the concentrates? You said it was “incenerated/grinded/washed again “. Was all of the carbon removed as you were separating the gold bond wires with your gold pan?

When you wash the cons (concentrates), are you using a dilute HCL wash, DH20, or plain H20 from the tap?

Regarding your process, If I understand correctly, you will:

Within the first beaker, you dissolve your iron in AR. Once done, you will transfer this solution to your second beaker which is holding the silicon dies. Are you planning on filtering this solution before transfer to the second beaker?

The second beaker holds your silicon dies. If necessary, will you add more AR or do you plan on trying to use only the original AR+Iron solution to dissolve the silicon dies? Once this second iron+dies solution is created, you will transfer this to a third beaker. Are you planning on filtering this solution before transfer to the third beaker?

The third beaker is holding your cons (concentrates). Again, if needed, more AR is added to create the third solution. Are you planning on filtering this solution before transfer to the fourth, and final, beaker?

The fourth beaker holds your gold bond wires. If necessary, with the addition of more AR, a fourth solution is made. This solution is filtered and the gold is dropped.

I am very interested in your results.

By accident, I created a situation where I had to battle through a batch of incinerated IC powder that contained a mixture of carbon powder, iron, copper, silicon dies, cons and bonding wires. All created due to my impatience. I started with 600 gms of RAM ICs (DIP and BGA type) and, so far, have only been able to retrieve 0.50 gm of gold powder.

James

 

[Tzoax]

James,

How did you remove the carbon from the concentrates? You said it was “incenerated/grinded/washed again “. Was all of the carbon removed as you were separating the gold bond wires with your gold pan?

I repeated inceneration/grinding/washing 5 or 6 times...just to make sure there is no more carbon there - i know because the last time i repeated process the washing water shows not any color - it have to be black/gray color of water if carbon is present. Another check was with microscope - there is no carbon ash visible - only this material on the picture.

When you wash the cons (concentrates), are you using a dilute HCL wash, DH20, or plain H20 from the tap?

For washing/making/reducing concentrate am using warm tap water, spoon, spray bottle (for water) and detergent.
I add one spoon of concentrate (at a time), fill the beaker with water mix it with spoon, add little detergent, mix again, wash spoon (possible gold bonding wires sticked) with spray bottle, and i spray a little more water on top layer of liquid from beaker to "drop down" possible gold bonding wires that might floating there. I repeat this 5-6 times until water is clear, then i remove this concentrate to another beaker, add another spoon and so on.

Regarding your process, If I understand correctly, you will:

Within the first beaker, you dissolve your iron in AR. Once done, you will transfer this solution to your second beaker which is holding the silicon dies. Are you planning on filtering this solution before transfer to the second beaker?
The second beaker holds your silicon dies. If necessary, will you add more AR or do you plan on trying to use only the original AR+Iron solution to dissolve the silicon dies? Once this second iron+dies solution is created, you will transfer this to a third beaker. Are you planning on filtering this solution before transfer to the third beaker?

The third beaker is holding your cons (concentrates). Again, if needed, more AR is added to create the third solution. Are you planning on filtering this solution before transfer to the fourth, and final, beaker?

The fourth beaker holds your gold bond wires. If necessary, with the addition of more AR, a fourth solution is made. This solution is filtered and the gold is dropped.

Yes, you understood it correctly.
First I will decant AR from one beaker to another, then just in case i will make just a little more of AR just to make sure all the gold is dissolved. After that i will wash beaker with HCl and then water and again add all of that to the next beaker. This is very little of material, so there will be very little of chemicals. There is no need to filter it at all stages, i will filter final AR from the last beaker.

By accident, I created a situation where I had to battle through a batch of incinerated IC powder that contained a mixture of carbon powder, iron, copper, silicon dies, cons and bonding wires. All created due to my impatience. I started with 600 gms of RAM ICs (DIP and BGA type) and, so far, have only been able to retrieve 0.50 gm of gold powder.

I would incinerate the all of it again, sieve it, remove iron legs with magnet, wash, and then repeat everything so there in minimum carbon left, then AR.

Thanks for posting, if you have troubles with it you can post pictures here.
I hope i explained this well, if you have any other question please ask.

Alex

 

[Tzoax]

While i was incinerating silicon dies i noticed that braze from most of silicon dies changed color to brownish-yellow:



I have seen this many times while incinerating, but i thought that this color will remain....but as soon i remove silicon dies from hotplate it changes color back again instantly - to gray.



Does anyone know what type of braze behaves like that and why?

 

[Tzoax]

I processed everything like i described.
At the second stage (silicon dies) - the brazes peels off (in form of foils) like in the picture. And it took a while - it seems like reaction is not progressing.



At the end - i neutralized the AR solution with urea, waited to cool and filtered it. But some of the silver chloride went through filter paper and now i am waiting that silver chloride settles down in the form of white sediment so i could decant AR solution and continue.
I am sure that source of this silver is from the braze of silicon dies.



So today i will have results.

 

[Tzoax]

Here is the result of gold yields from processing 536.6g of Big BGA IC Chips.

Dropping gold with SMB...



Interesting thing happened (to me - for the first time) - while i was washing the gold powder (3xWater/3xHCl/3xWater) i noticed that as soon i added HCl to the gold powder - the solution becomes pale purple colored. After several washings the solution became colorless. But....i can not explain why it happened this time. It looked very much like colloidal gold - when tin reducing gold to purple color - but there was no tin at all. I think it is related with silver...



Washed and dried gold powder.




And finally result:



So from 536.6g of BGA IC's i recovered 5.68g of gold.
It is about 10.58g of gold per 1kg of this kind of IC chips.
I expected a little bit more...but i think that 10+g of gold per kg of material is a good result either.

 

[mls26cwru]

That is very close to my averages for the chip tops as well. You're going to have a nice little button there!

 

[cosmetal]

Tzoax,

Very well executed plan. Thank you for your detailed explanations and multiple pictures.

You wrote:

"i wonder will this affect on yields... since (doing this way) i think i reduced chance for something go wrong to minimum. Maybe those small things i changed will affect on yields, maybe not, but i think it is worth to try."

It appears that it didn't increase your yields as we both had hoped. I wanted to see if the silicon dies would up the gold. I wonder if this method (plan) had an effect on increasing other PM content?

James

 

[Tzoax]

That is very close to my averages for the chip tops as well. You're going to have a nice little button there!

Thanks, when i finish all of the chips i prepared for testing i hope a nice little gold bar will see the light of the day.. :G :mrgreen:

It appears that it didn't increase your yields as we both had hoped. I wanted to see if the silicon dies would up the gold. I wonder if this method (plan) had an effect on increasing other PM content?

I did made test of just silicon dies, it is on this same thread, this is a link - http://goldrefiningforum.com/phpBB3/viewtopic.php?f=33&t=22951&start=90#p261522

I processed only 18.2g od silicon dies at that time, and there was gold, but there was so small amount that my old 0.1 scales could not detect the weight. This time (on this last test with BGA chips) i was thinking maybe silicon dies with braze was made with a more gold than braze from other types of IC chips...but it seems they are the same.

So, from now on - when i test the IC chips i will avoid processing them with silicon dies, it will not affect (considerably) to yield results, it only gives a trouble because of the silver - first problem is a time to (partially) dissolve braze from silicon dies - it is very slow, then - not all silver will dissolve because AR could dissolve only "some" of the silver, then the color of dissolved silver in AR is also yellow - like gold chloride - so many times i thought it was gold dissolved.....but at the end - all of that silver chloride is the biggest "problem" - in other words - it saves a LOT of time when processing IC chips without silicon dies. Eventually - when i collect some more considerable amount of silicon dies i will test them again.

Alex

 

[mls26cwru]

I have over 2 lbs of silicon chip dies that I hope to process sometime soon. I will post the numbers if I ever make it to that point... I was playing around with the idea of smelting them, but not sure how well it will work since the brazing nubs are so small. I am assuming I could crush them and add borax, litharge, and some flourospar and it should be like a straight forward fire assay fusion. I'm still reading up on it though, so anyone more experienced with smelting, any advice would be appreciated.

 

[cosmetal]

Alex,

I must admit that this last week is the first time I had available to really "dig" into your thread in detail.

Again, thank you so much for your methodical and detailed work and explanations!

Please keep up the great work which is so valuable to us newbies :!: :!:

James

 

[rickzeien]

Tzoax,

Very well executed plan. Thank you for your detailed explanations and multiple pictures.

You wrote:

"i wonder will this affect on yields... since (doing this way) i think i reduced chance for something go wrong to minimum. Maybe those small things i changed will affect on yields, maybe not, but i think it is worth to try."

It appears that it didn't increase your yields as we both had hoped. I wanted to see if the silicon dies would up the gold. I wonder if this method (plan) had an effect on increasing other PM content?

James

Is it possible that even though the yield did not go up the purity did?

Sent from my LG-H872 using Tapatalk

 

[modtheworld44]

Tzoax

Just out of curiosity,what do you do with all the extra ash waste when you remove it from your gold concentrates?Thanks in advance.



modtheworld44

 

[Tzoax]

Alex,
I must admit that this last week is the first time I had available to really "dig" into your thread in detail.
Again, thank you so much for your methodical and detailed work and explanations!
Please keep up the great work which is so valuable to us newbies :!: :!:

James

Thank you James, it is my pleasure to share results and experience about processing IC chips. It is a large field, with many possible ways of processing it, there are many types of IC chips, and exchanging results and experiences about processing it surely helps a lot to anyone interested in this subject.

Is it possible that even though the yield did not go up the purity did?

The only difference that i made this time (processing IC chips this way) was that i decanted AR from concentrate and silicon dies before dissolving gold bonding wires (i was hoping it will affect the yield but it didn't). The purity is always on a high level at the end (after washing), so i doubt that it makes difference. Last time i checked the purity was 997.

Tzoax
Just out of curiosity,what do you do with all the extra ash waste when you remove it from your gold concentrates?Thanks in advance.
modtheworld44

Thank you for question - it is very interesting subject. Now - i save everything (just in case). I always had feeling that when making concentrate (washing ashes with water) - when decanting water (with ashes) some on the gold bonding wires could be lost.
At first i didn't pay attention to it - i had no experience - a "feeling" how and when it should be decanted. So i start to collect all of the decanted ashes into a big container, and when ashes settled - i inspect the ashes with microscope (after concentrating heavier particles at the bottom by panning) and at first there was some of the gold bonding wires.
But after some time (when i start using detergent and water spray, also improving other details) there was no more gold bonding wires "escaping" from mixture while decanting.
Finally - at this last time - i tested the speed of pure gold bonding wires (from the picture above - gold bonding wires in the spoon ) falling through the water - just to observe how fast they will sink and will some of them stay floating on the surface... also i observed how they reacts on stirring water with spoon....
And they are falling pretty fast...couple of seconds maximum.

So - here is my little "tricks" how to avoid a loss of gold bonding wires while decanting:

-i am using hot water when mixing. Maybe it have no sense but i have feeling that ashes mixes much better and faster in hot than in cold water.

-i am always mixing one (maximum one and half) soup spoon of ashes at a time. This way i can much better "control" concentrate.

-i am using 350ml beaker

-after stirring ashes with spoon (pretty well) i add a little of detergent and stir again. When i remove the spoon i spray it with water (using spray bottle), and i also spray a surface of water.

-i tap a beaker with fingers couple times (on a side) to drop down some of the possible gold bonding wires sticked on the walls of the beaker

-i give it about 10 seconds, then i slowly decant water with ashes to a container - no more than 1/3 of volume (for the first time)

-then i refill the beaker with water, repeat everything from above, and this time i decant about half of the volume. Each time i decant more and more of water

-after 5-6 times of decanting - water becomes clear so i remove the remaining concentrate to another beaker and repeat everything

 

[Tzoax]

I found that there is very little data about SMD transistors here on the forum. For instance - about SOT-23 package. They are everywhere on all types of PCB's -especially on laptop's motherboards and graphic cards. I recently start to collect them and i am planning to test them.



I found documents about metal content inside of them - and according to this - one transistor SOT-23 package have 0.00008275488g of gold. Weight of one transistor is 0.00872g. That means that 1kg of this type of transistors will have 9.49g of gold.

You can do math yourself. Remember to totalize gold from gold bonding wires and gold from dice.



Then, there is SOT-323 package - 9.47g of gold per kilogram.



SOT-89 package - 3.68g of gold per kilogram.



And this is type of transistors that contains no gold.




I think SOT-23 package transistors is a pretty interesting material to test considering they are often used on PCBs.

 

[kernels]

I would love to see the result of these Tzoax, unlikely to bother collecting them, but that seems like a surprisingly high yield.

 

[FrugalRefiner]

That's an interesting find!

But if I did the math right, it'll take over 100,000 to make a kilo.

Still, it's one more source of gold.

Dave