Wire bonding gold extraction power transistor

[AlanFry]

Hello everyone,

For a couple of months, I spend some time reading articles and passionate subjects on this forum, which I really love — time for me, now to ask for the first time some advice to the community. As a short intro, I'm Alan and I work in the advanced electronics industry in East Europe.

5 years ago, a friend enrolled in the electronic industry and give me a fairly great amount of NPN power transistors (3kg total). This kind of component has been massively used in the 90s in spatial applications in power amplifier architecture but are now obsolete because of the advent of III–V semiconductors like LDMOS, AsGa, or GaN-based PA offering better power and "cost" density.

The transistors I get are made of 2 plated flanges (base and emiter) mounted in a hermetic ceramic package (white ceramic, probably beryllium oxide BeO which looks like a danger to me). An interesting feature is that the datasheet mentioned an internal gold metalization. As the transistor is capable of driving a high amount of current, the internal wiring is excepted to be well established.

For that reason, I choose to open one and as expected, the bonding wires are well present in great quantity. Since I'm not familiar now, with chemical extraction, I choose to collect the gold of bonding wire by hand using an electronic microscope and scalpel with a mask and goggles just in case of hazardous dust. In 30 minutes using 6 units, I collected an amount of gold wire that I can't weigh at the moment as my electronic scale died.. I share a picture of the nude transistor and the wire collected I will appreciate having feedback on the potential of the yield of this. I'm wondering about the remaining quantity of gold present on the plating that of course, I can't extract with a scalpel...

Has anybody already processed and cooked that kind of transistor or have some feedback on the quantity (order of magnitude) of the gold present?

As a newbie, I want to collect the gold wire bonding "by hand" for the moment before starting anything related to the chemical process and try to form a button from that.
I'm wondering if someone can give me some advice on what kind of material/system to choose to melt gold and form buttons. My first guess would be to use an induction heater system, but is anybody has already tested that,
and is that adapted for my needs?




Have a great day

 

[MicheleM]

Welcome Alan! beautiful transistor, because I do not see other metals , for the gold plating i would go directly with HCl (+ very little adding of HNO3) , or if you can't get Nitric acid (HNO3) then i would go with HCl ( +little adding of bleach) , both reaction are dangerous, they produce nitric oxide and Cl2 respectively, so you should use safety precautions (like a fume hood, or stay outside using lids and mask) that are well described in this Forum

 

[GoIdman]

Hi Alan,

Those transistors look very nice and since they are not commercial grade, you should keep some of them intact since the collectible value could well exceed the PM value.
I have worked in electronics for more than 10 years but didnt come across such beauties along the way( mostly commercial electronics)
IMHO the leftover base could be striped in a sulfuric cell to avoid large amount of toxic fumes, after that you could refine the stripped gold dust and leftovers.

Since there is a large amount of copper(or silver as a matter of fact if its military stuff) as the base of the heatsink, the dye is "welded" to it so there should be plenty of contact for recovering almost every gold from the surface.
You could also recover the heatsinks as well.
Search the forum for this process, then check out sreetips videos about the cell.

Be safe

Pete.

 

[Lino1406]

The external yellow looks like beryllium bronze, not gold

 

[Yggdrasil]

The external yellow looks like beryllium bronze, not gold

To me it seems plated, but Beryllium Bronze for electrical connections often are Gold plated.

 

[AlanFry]

Welcome Alan! beautiful transistor, because I do not see other metals , for the gold plating i would go directly with HCl (+ very little adding of HNO3) , or if you can't get Nitric acid (HNO3) then i would go with HCl ( +little adding of bleach) , both reaction are dangerous, they produce nitric oxide and Cl2 respectively, so you should use safety precautions (like a fume hood, or stay outside using lids and mask) that are well described in this Forum

Thanks for your answer Michele, Do

 

[orvi]

Sometimes I do find some of these, and they can be very, very rewarding. Usually, they are RF used, and because of it, white ceramic is very likely to contain BeO. Ones we came across, the "legs", or base metal plates are some Cu-Mo alloys (they bend), or tungsten containing (they brake on bending). CuMo tend to have more values. I did not personally analysed them (base metal composition), so I could be wrong, but this is what I learned/heard from more experienced refiners here. As it is practically certain that the ceramic is BeO containing, I didn´t personally processed them (ones I found on the boards from spectrometers and high grade telecom/RF).

This is some pretty rewarding material to mess/ruin for learning how to extract gold I think. I just say that ammount of gold in bonding wires just partly add to the overall gold content of the transistor. So you are recovering nice and practically pure gold, but not nearly the ammount what is there.
I would advise to ask for toll/fee refinig them here on the forum. There are many very experienced, honest and fair refiners that will do the job for you.
And of course, try to find their collector value. This isn´t everyday stuff like Pentium CPU´s, it could have high premiums, if you have them in great shape and unused.

 

[AlanFry]

Thanks, to everyone for the feedback. I share a picture of the gold wires and an area I scratched to understand the plating material and thickness. From that maybe some of yours will be able to give me some more precision. The plating material is really soft and sticks very easily to the bonding wires.

In synthesis, after having removed manually the bonding wire, the remaining materials are essentially Cu/W composed of the flange, gold from the plating, Silicium or whatever from the transistors units, and ceramic.

As well mentioned by @goldman as all the supposedly gold-plated areas are connected on the "grounded flange" so, in my intuition, the sulfuric stripping seems to be well adapted for de-plating the gold. Is there any advantage to using this method again the HCL and bleach method (pros and cons?)

When using HCL, I suppose that the concentration needed is around 90%. Here, I can have access to sulfuric acid around 31-38%, and for the HCL, I didn't manage to find higher than 23%... Is that concentration of H2SO4 enough for the stripping cell reaction ?

Thanks in advance for your help

 

[eaglekeeper]

I've done a lot of these or something similar and as ORVI said they are decent.... if you have a quantity of them. I use an exacto-blade to scrap the bonding wires into a small container. I run the bodies in a striping cell, then add the bonding wires to the cell slimes to refine. And yes there is a potential that BeO will be present.

Just a hand full I haven't finished preparing yet. There's a few where the gold is only under the caps and there is gold that has alloyed with the solder, but not much though. I don't know if it's cost effective to try to recover that gold from the solder..... I have several Lbs of bodies I've already striped.... and I'm still debating on whether to recover the gold or just take them to the scrapyard as dirty brass/#2 copper.

 

[FrugalRefiner]

When using HCL, I suppose that the concentration needed is around 90%. Here, I can have access to sulfuric acid around 31-38%, and for the HCL, I didn't manage to find higher than 23%... Is that concentration of H2SO4 enough for the stripping cell reaction

I just want to clarify things. Are you saying you have HCl that is 90%? I've never seen that. Typically, "concentrated" HCl is somewhere around 32%, give or take a couple of percent. 90% HCl would fume like crazy!

The sulfuric for a stripping cell needs to be around 90% or higher. Lower concentration sulfuric will attack base metals. like copper, while concentrated sulfuric barely touches it. Sulfuric can be concentrated by heating it, but be very careful! Hot, concentrated sulfuric can ruin your life.

Dave

 

[GoIdman]

Thanks, to everyone for the feedback. I share a picture of the gold wires and an area I scratched to understand the plating material and thickness. From that maybe some of yours will be able to give me some more precision. The plating material is really soft and sticks very easily to the bonding wires.

In synthesis, after having removed manually the bonding wire, the remaining materials are essentially Cu/W composed of the flange, gold from the plating, Silicium or whatever from the transistors units, and ceramic.

As well mentioned by @goldman as all the supposedly gold-plated areas are connected on the "grounded flange" so, in my intuition, the sulfuric stripping seems to be well adapted for de-plating the gold. Is there any advantage to using this method again the HCL and bleach method (pros and cons?)

When using HCL, I suppose that the concentration needed is around 90%. Here, I can have access to sulfuric acid around 31-38%, and for the HCL, I didn't manage to find higher than 23%... Is that concentration of H2SO4 enough for the stripping cell reaction ?

Thanks in advance for your help

Using the sulfuric cell you dont have to digest the base metals thus you will not create chemical waste.
For refining your gold you can use far less chemicals.
The sulfuric acid can be reused after filtering.

Be safe,

Pete.

 

[AlanFry]

@orvi
In my case de pins are "soft" and they are not breaking when they are bend so it's more likely Cu/Mo alloys like you mentionned.
For the flange part, I would go more for copper tungsten Cu-W here because in the 90s Cu-Mo was less common, Idk if there is a method to test it.

For the Ceramic related hazard, I'm lucky because the transistor use ceramic and hermetic package with a metalic cover on the top, so by properly placing a plier on the top, I can crush the cover and open it without breaking the cermaic structure.

I've done a lot of these or something similar and as ORVI said they are decent.... if you have a quantity of them. I use an exacto-blade to scrap the bonding wires into a small container. I run the bodies in a striping cell, then add the bonding wires to the cell slimes to refine. And yes there is a potential that BeO will be present.

Just a hand full I haven't finished preparing yet. There's a few where the gold is only under the caps and there is gold that has alloyed with the solder, but not much though. I don't know if it's cost effective to try to recover that gold from the solder..... I have several Lbs of bodies I've already striped.... and I'm still debating on whether to recover the gold or just take them to the scrapyard as dirty brass/#2 copper.View attachment 51490

@eaglekeeper Nice RF transistors, I see you have also, at the top, (the 4 items) some RF power resistors (thick film resistors), these are used in attenuator, RF combiner or balanced architectures in telecom applications they 'll probably have less gold than your other devices.

Exacto blade is a great alternative, In my case I'm using a watch oiler tool (Watch & Clock Oiler Swiss Grobet Design Oil Tools Number 29 Old Stock NOS for sale online | eBay) with different shape available, really precise and also fragile..

If I understand correctly, this means that in your case, the stripping cell cannot disolve the tin's solder layer or in another words, mask the platted area ?

 

[AlanFry]

I just want to clarify things. Are you saying you have HCl that is 90%? I've never seen that. Typically, "concentrated" HCl is somewhere around 32%, give or take a couple of percent. 90% HCl would fume like crazy!

The sulfuric for a stripping cell needs to be around 90% or higher. Lower concentration sulfuric will attack base metals. like copper, while concentrated sulfuric barely touches it. Sulfuric can be concentrated by heating it, but be very careful! Hot, concentrated sulfuric can ruin your life.

Dave

Thanks for your precisions Dave and Pete, I made a mistake, I was confused between HCL and H2SO4 concentration. Can someone explain to me the scientific reason, that higher concentrated H2SO4 will not touch the copper body in comparaison to lower one ?

 

[Shark]

Thanks for your precisions Dave and Pete, I made a mistake, I was confused between HCL and H2SO4 concentration. Can someone explain to me the scientific reason, that higher concentrated H2SO4 will not touch the copper body in comparaison to lower one ?

Passivization.

 

[eaglekeeper]

If I understand correctly, this means that in your case, the stripping cell cannot disolve the tin's solder layer or in another words, mask the platted area ?

If the solder layer is thin enough I guess it will strip. But....just about all of mine has large puddles of solder and I would have to hit each one with a heat gun (again) and wiper those puddles off. I guess I should have thought about that while it was still hot when I removed component.

@eaglekeeper Nice RF transistors, I see you have also, at the top, (the 4 items) some RF power resistors (thick film resistors), these are used in attenuator, RF combiner or balanced architectures in telecom applications they 'll probably have less gold than your other devices.

Thanks for the info... nice to have members here that can help identify components, because I can only identify the basics stuff. So those at the top are resistors.....🤔. Then you're probably correct about not containing gold.

 

[dpgold]

Hello everyone,

For a couple of months, I spend some time reading articles and passionate subjects on this forum, which I really love — time for me, now to ask for the first time some advice to the community. As a short intro, I'm Alan and I work in the advanced electronics industry in East Europe.

5 years ago, a friend enrolled in the electronic industry and give me a fairly great amount of NPN power transistors (3kg total). This kind of component has been massively used in the 90s in spatial applications in power amplifier architecture but are now obsolete because of the advent of III–V semiconductors like LDMOS, AsGa, or GaN-based PA offering better power and "cost" density.

The transistors I get are made of 2 plated flanges (base and emiter) mounted in a hermetic ceramic package (white ceramic, probably beryllium oxide BeO which looks like a danger to me). An interesting feature is that the datasheet mentioned an internal gold metalization. As the transistor is capable of driving a high amount of current, the internal wiring is excepted to be well established.

For that reason, I choose to open one and as expected, the bonding wires are well present in great quantity. Since I'm not familiar now, with chemical extraction, I choose to collect the gold of bonding wire by hand using an electronic microscope and scalpel with a mask and goggles just in case of hazardous dust. In 30 minutes using 6 units, I collected an amount of gold wire that I can't weigh at the moment as my electronic scale died.. I share a picture of the nude transistor and the wire collected I will appreciate having feedback on the potential of the yield of this. I'm wondering about the remaining quantity of gold present on the plating that of course, I can't extract with a scalpel...

Has anybody already processed and cooked that kind of transistor or have some feedback on the quantity (order of magnitude) of the gold present?

As a newbie, I want to collect the gold wire bonding "by hand" for the moment before starting anything related to the chemical process and try to form a button from that.
I'm wondering if someone can give me some advice on what kind of material/system to choose to melt gold and form buttons. My first guess would be to use an induction heater system, but is anybody has already tested that,
and is that adapted for my needs?

View attachment 51344


Have a great day

Hello everyone,

For a couple of months, I spend some time reading articles and passionate subjects on this forum, which I really love — time for me, now to ask for the first time some advice to the community. As a short intro, I'm Alan and I work in the advanced electronics industry in East Europe.

5 years ago, a friend enrolled in the electronic industry and give me a fairly great amount of NPN power transistors (3kg total). This kind of component has been massively used in the 90s in spatial applications in power amplifier architecture but are now obsolete because of the advent of III–V semiconductors like LDMOS, AsGa, or GaN-based PA offering better power and "cost" density.

The transistors I get are made of 2 plated flanges (base and emiter) mounted in a hermetic ceramic package (white ceramic, probably beryllium oxide BeO which looks like a danger to me). An interesting feature is that the datasheet mentioned an internal gold metalization. As the transistor is capable of driving a high amount of current, the internal wiring is excepted to be well established.

For that reason, I choose to open one and as expected, the bonding wires are well present in great quantity. Since I'm not familiar now, with chemical extraction, I choose to collect the gold of bonding wire by hand using an electronic microscope and scalpel with a mask and goggles just in case of hazardous dust. In 30 minutes using 6 units, I collected an amount of gold wire that I can't weigh at the moment as my electronic scale died.. I share a picture of the nude transistor and the wire collected I will appreciate having feedback on the potential of the yield of this. I'm wondering about the remaining quantity of gold present on the plating that of course, I can't extract with a scalpel...

Has anybody already processed and cooked that kind of transistor or have some feedback on the quantity (order of magnitude) of the gold present?

As a newbie, I want to collect the gold wire bonding "by hand" for the moment before starting anything related to the chemical process and try to form a button from that.
I'm wondering if someone can give me some advice on what kind of material/system to choose to melt gold and form buttons. My first guess would be to use an induction heater system, but is anybody has already tested that,
and is that adapted for my needs?

View attachment 51344


Have a great day

How do you plan to deal with Be? Incineration in a fume hood would change BeO ?

 

[Lino1406]

Passivization.

To my best knowledge, copper will be attacked by concentrated sulfuric acid

 

[Yggdrasil]

To my best knowledge, copper will be attacked by concentrated sulfuric acid

90+% as used in the Sulfuric stripping Cell, will not, dilute hot Sulfuric will attack copper if I’m not mistaken.

 

[Shark]

I think AlanFry was asking why dilute sulfuric was more aggressive on copper than concentrated sulfuric.

 

[Yggdrasil]

I think AlanFry was asking why dilute sulfuric was more aggressive on copper than concentrated sulfuric.

Aah, rookie mistake from me.
Commenting on the latest post.

Alan I’m not entirely sure, but I think it is a combination on passification and solubility. Because if it get hot enough it starts dissolving anyway.
So it is important to keep the Sulfuric “dry”/ (it soaks the water straight from the air) and cold.

Someone will set me straight if I’m mistaken I hope😀