Laptop touch pad Boards and Keyboard traces

Gold Refining Forum

Help Support Gold Refining Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I had never even considered dioxin as a by-product when I incinerate IC:s but of course you're right. So I went on a google spree and dug into some scientific papers on that topic. After a quite shallow reading it seems like there are a few things to consider.

I don't think incineration of IC:s produces so much dioxin. One major part of dioxin is chloride and a quick scan of the net seems to show that the epoxy used in IC:s is low or even free of chloride.
Generally the trend is towards halogen free plastics in the industry just because it can easily form toxic gases in case of a fire.

Dioxin is created between 200 and 800 C but peak production seem to be around 450 C. It is created when partially burned compounds condense on soot with a lot of CO2 and CO present. So sooty flames are bad, keep a lean flame with excess oxygen.

Dioxin is destroyed by burning it for a long enough time at a high enough temperature, 20-40 seconds at 600-800 C seems to be adequate.

Pyrolysing the IC would boil off the volatile part of the plastic and the fumes could be burned in an after burner with long enough retention time. Then the carbon that is left could easily be slowly incinerated in an oven at 700 C with some air circulation. Just about visually glowing red. If you go too high you will sinter the ash (silicon dioxide) into a hard ceramic and you make it harder to crush the chips.

Here are some of the texts I read.

https://www.mdpi.com/2073-4360/15/19/3964https://www.diva-portal.org/smash/get/diva2:142298/FULLTEXT01.pdfhttps://www.jstage.jst.go.jp/article/isijinternational1989/43/3/43_3_461/_pdf
I haven't run as large volumes as Kurt has, only 10-15 kilo per year and with batches around half a kilo. I first pyrolyse the chips and then I incinerate them in a temperature controlled oven. 600 C for five minutes or more depending on batch size, crush, sieve. Anything caught in the fine sieve goes back into the incineration.
To crush the incinerated chips I just put them in a wide necked glass bottle and shake them for 30 seconds (with the cap on). The sieve I uses are fine enough to catch most metal pins, even the fine pitch ones.
Each incinerate - crush - sieve cycle removes about 30% of the remaining plastic and the ash is fine enough to pan it right away (or rather washing with two beakers).
I have also experimented with ultrasonic to crush any small ash particles clinging to the bond wires and had good results. Some day I'll make a proper write-up of my process.

Göran
 

Attachments

  • 20230815_171833 Pyrolyser.jpg
    20230815_171833 Pyrolyser.jpg
    7.8 MB
  • 20231003_180711 After pyrolyser.jpg
    20231003_180711 After pyrolyser.jpg
    7.5 MB
  • 20231018_132143 Before ultrasonic.jpg
    20231018_132143 Before ultrasonic.jpg
    1.6 MB
  • 20231019_230354 After ultrasonic.jpg
    20231019_230354 After ultrasonic.jpg
    1.6 MB
  • 20231023_134248 Guld.jpg
    20231023_134248 Guld.jpg
    1.9 MB
I had never even considered dioxin as a by-product when I incinerate IC:s
Hey there Goran long time no see - so it is REALLY good to see you posting again here on the forum

The dioxins (along with the other toxic gasses/fumes) from the open burning of E-waste is a major concern to the WHO - EPA etc. ect. --- the dioxins at (or near) the top of the list

The vast majority of the plastic in E-waste is the epoxy - whether it be the epoxy in the fiberglass board it self or the IC chips

Besides the dioxins & furans & hydrocarbons there is a whole host of toxic gasses/fumes produced when burning plastic (including epoxy)

Which is the reason I post that it needs to be done properly

Burning E-waste NEEDS to be considered hazardous/toxic just as the wet chemistry part of refining is hazardous/toxic

They both put your health at risk when not done right (but you know that) :D ;);)

Others don't which why we post

Example; - vinegar/salt method is safe --- WRONG

Kurt
 
Thank you all for the info. So to recap (please correct if I'm wrong), I can pyrolyse the ICs in a furnace with a temp of 600C so as to not melt Al (Only Tin/Lead will be melted if I'm correct). An air blower is best in that case to get enough oxygen in the system to burn the gases. A closed pressure vessel with a pipe heading back into the main part of the fire intake is best since it holds the gases longer and sends them to the hottest part of the flame.

Some of your setups show an added chimney to your furnaces, that to keep the gases in contact with heat longer?
 
Thank you all for the info. So to recap (please correct if I'm wrong), I can pyrolyse the ICs in a furnace with a temp of 600C so as to not melt Al (Only Tin/Lead will be melted if I'm correct). An air blower is best in that case to get enough oxygen in the system to burn the gases. A closed pressure vessel with a pipe heading back into the main part of the fire intake is best since it holds the gases longer and sends them to the hottest part of the flame.

Some of your setups show an added chimney to your furnaces, that to keep the gases in contact with heat longer?
The important thing is to not introduce air/Oxygen into the pyrolizing chamber, but burn the gases after with plenty of oxygen after in the fire itself.
And when it is properly Pyrolised you can incinerate it with plenty of oxygen in open fire if needed.
 
Thank you all for the info. So to recap (please correct if I'm wrong), I can pyrolyse the ICs in a furnace with a temp of 600C so as to not melt Al (Only Tin/Lead will be melted if I'm correct). An air blower is best in that case to get enough oxygen in the system to burn the gases. A closed pressure vessel with a pipe heading back into the main part of the fire intake is best since it holds the gases longer and sends them to the hottest part of the flame.

Some of your setups show an added chimney to your furnaces, that to keep the gases in contact with heat longer?
You can do some general research on pyrolizing on the net.

It was devised as a technique to pull volatile gases and liquids out of organic matter for fuel.
Like this wood chip cars in and around WW2, and as a source for Syngas during the WW2.
In this application they introduced super heated steam into the reactor to get better yields.
For our purpose on the other hand, we just want to get rid of and decompose the volatile compounds,
so there is no need for this kind of measures, simple will work fine.
Just make sure the off-gases are properly burned in the hottest part of the fire.
Pick off the Aluminum parts and there is no need to restrict the heat.
The best is an open vessel with no pressure since it will pose no danger of exploding.
 
You can do some general research on pyrolizing on the net.

It was devised as a technique to pull volatile gases and liquids out of organic matter for fuel.
Like this wood chip cars in and around WW2, and as a source for Syngas during the WW2.
In this application they introduced super heated steam into the reactor to get better yields.
For our purpose on the other hand, we just want to get rid of and decompose the volatile compounds,
so there is no need for this kind of measures, simple will work fine.
Just make sure the off-gases are properly burned in the hottest part of the fire.
Pick off the Aluminum parts and there is no need to restrict the heat.
The best is an open vessel with no pressure since it will pose no danger of exploding.
How do I make sure the gases are in contact with the heat for enough time if it's an open vessel? 20secs seems a long time for a simple open crucible. Is it enough to use the chimney pipe on the furnace to get the gasses burnt?

I'm a little confused though, if I need oxygen in the flames to help burn off the gases and oils produced, won't the oxygen get into the open container and affect the pyrolysis process?

Would something like this be enough?

Pyro1.jpg
Maybe use this as a base plate?

1704846380917.png
 
Last edited:
OK since we are on the subject of Pyrolizing ICs, do I have to handle any of these differently? I'm going to start another round of sorting and would need to know what can go with what and what should be handled separately so I can sort once and not 10 times ;) PXL_20240110_010148645.jpg PXL_20240110_003631669.jpg
 
I won't be able to give you good info on how you want to sort them, but can provide more generic details about the items in this picture. FYI, a bit of "the blind leading the blind" here. Main thing is that it's a lot of time and effort to research each individual chip, and even within a same model, changes are made throughout the years. Gold bond wires are gold, but sometimes they use aluminum for bond wires, and I'm not sure what properties might lead to guess that they'd be aluminum vs gold. If you assume all bond wires are gold, you'll be disappointed, but saving time by making assumptions may also lead to throwing away gold.
1. gold corner BGA. The silicon die is mounted to the fiberglass substrate. Bond wires, usually gold, connect the traces on the fiberglass substrate to the top of the silicon die. Then it's encapsulated in black plastic. So, most of the good stuff is in the plastic. I havent tested, but would guess that the gold on the fiberglass substrate is very thin. But, if you separate them, sometimes little nubs of the bondwires are left attached to it.
2. flip chip BGA. The silicon die is mounted upside-down to the fiberglass substrate. This negates the need to have bond wires to connect the traces to the die.
3-4. These are BGAs of some sort. I'm not certain if they can be known to be flip chip or not. #4 specifically is common on hard drive boards, so might be someone knows more about these specific chips.
5, 7. Typical plastic ic chips. Die is usually mounted to a metal frame, pins surrounding but not touching, bond wires connect the pins to the die, and the whole thing is encapsulated in plastic.
6, 8, 11. Also typical plastic IC chips, but also very commonly this specific model and manufacturer. 6 is a type of storage chip for storing firmware of a device. 8 is a clock chip common on motherboards, usually paired with a resonator to provide a stable common clock signal for many components on the motherboard. 11 is a common PC hardware management interface IC, dunno why but many manufacturers use this brand for that function on their motherboards.
9. this package is used for ICs which need up to 8 connections, and sometimes as a power mosfet (a transistor that can handle high power loads, like 16). If it's an IC, it is usually the same as typical plastic IC chips. If it's a mosfet, usually 4 legs of one side makes one connection, 3 legs of the other side are connected with either a relatively thick copper strip, or by a row of parallel bond wires, and the last leg is usually a single bond wire.
10. yep, just a bunch of transformers in a convenient package, for isolation of things like phone lines or ethernet lines.
12. If no polarity markings, these are often the same as the multi-layer ceramic capacitors (MLCCs) shown in your 17, with legs attached and dipped in epoxy. It seems these are more common in older boards, and older boards are more likely to have MLCCs made with precious metals rather than nickel or other common metals. Sometimes you may find ones that are more orange, with polarity markings, and a not as symmetric-looking. Those would likely be tantalum capacitors.
13-14 are types of capacitors. I think both are wet and are considered aluminum capacitors, but I'm not 100% certain about #14 because marketing materials often claim these are "solid state capacitors".
17. In your pic, they are multi-layer ceramic capacitors. Tan, square prism shaped. Old ones used layers of ceramic and precious metals to create a high surface area capacitor. Newer ones use nickel instead of precious metals. People say you can use a magnet to distinguish between the two somehow. Surface mount inductors also come in this size, and are obviously more magnetic than their capacitor look-alikes. Not all of the little "crumbs" this small are MLCCs and inductors, though.
15. Inductor
16. mosfet - rarely have gold bond wires. Unlike 9, I don't think I've ever seen one of these use parallel bond wires for current carrying. At most, it might have 1 bond wire for the signal pin, and even then it may just be an aluminum bond wire.
 
I won't be able to give you good info on how you want to sort them, but can provide more generic details about the items in this picture. FYI, a bit of "the blind leading the blind" here. Main thing is that it's a lot of time and effort to research each individual chip, and even within a same model, changes are made throughout the years. Gold bond wires are gold, but sometimes they use aluminum for bond wires, and I'm not sure what properties might lead to guess that they'd be aluminum vs gold. If you assume all bond wires are gold, you'll be disappointed, but saving time by making assumptions may also lead to throwing away gold.
1. gold corner BGA. The silicon die is mounted to the fiberglass substrate. Bond wires, usually gold, connect the traces on the fiberglass substrate to the top of the silicon die. Then it's encapsulated in black plastic. So, most of the good stuff is in the plastic. I havent tested, but would guess that the gold on the fiberglass substrate is very thin. But, if you separate them, sometimes little nubs of the bondwires are left attached to it.
2. flip chip BGA. The silicon die is mounted upside-down to the fiberglass substrate. This negates the need to have bond wires to connect the traces to the die.
3-4. These are BGAs of some sort. I'm not certain if they can be known to be flip chip or not. #4 specifically is common on hard drive boards, so might be someone knows more about these specific chips.
5, 7. Typical plastic ic chips. Die is usually mounted to a metal frame, pins surrounding but not touching, bond wires connect the pins to the die, and the whole thing is encapsulated in plastic.
6, 8, 11. Also typical plastic IC chips, but also very commonly this specific model and manufacturer. 6 is a type of storage chip for storing firmware of a device. 8 is a clock chip common on motherboards, usually paired with a resonator to provide a stable common clock signal for many components on the motherboard. 11 is a common PC hardware management interface IC, dunno why but many manufacturers use this brand for that function on their motherboards.
9. this package is used for ICs which need up to 8 connections, and sometimes as a power mosfet (a transistor that can handle high power loads, like 16). If it's an IC, it is usually the same as typical plastic IC chips. If it's a mosfet, usually 4 legs of one side makes one connection, 3 legs of the other side are connected with either a relatively thick copper strip, or by a row of parallel bond wires, and the last leg is usually a single bond wire.
10. yep, just a bunch of transformers in a convenient package, for isolation of things like phone lines or ethernet lines.
12. If no polarity markings, these are often the same as the multi-layer ceramic capacitors (MLCCs) shown in your 17, with legs attached and dipped in epoxy. It seems these are more common in older boards, and older boards are more likely to have MLCCs made with precious metals rather than nickel or other common metals. Sometimes you may find ones that are more orange, with polarity markings, and a not as symmetric-looking. Those would likely be tantalum capacitors.
13-14 are types of capacitors. I think both are wet and are considered aluminum capacitors, but I'm not 100% certain about #14 because marketing materials often claim these are "solid state capacitors".
17. In your pic, they are multi-layer ceramic capacitors. Tan, square prism shaped. Old ones used layers of ceramic and precious metals to create a high surface area capacitor. Newer ones use nickel instead of precious metals. People say you can use a magnet to distinguish between the two somehow. Surface mount inductors also come in this size, and are obviously more magnetic than their capacitor look-alikes. Not all of the little "crumbs" this small are MLCCs and inductors, though.
15. Inductor
16. mosfet - rarely have gold bond wires. Unlike 9, I don't think I've ever seen one of these use parallel bond wires for current carrying. At most, it might have 1 bond wire for the signal pin, and even then it may just be an aluminum bond wire.
Thank you so much, that does help a lot. My follow up question would be, can they all be processed together for Silver? From what I've seen and read, that's what you get with these?
 
Last edited:
I'm not sure about silver content. Several of these types of things may or may not have silver, and it's just a design choice by the engineers who designed it, to meet some technical requirement. If you have a blended mix of all this stuff, I would not bet against there being any silver in there. But I also think it might be worth cherry-picking other components if silver is your goal.

Silver is a good conductor of electricity and heat, and often is used to plate the contact surfaces of switches. (ever take apart a keyboard? It's just a big thing of lots of switches ;) ) Can also find it on components made of hard-to-connect materials, like quartz crystal or compressed blend of polymer and metal oxide powders. Manufacturer uses silver to connect the legs to the actual device, so that the end user can just connect to the legs instead of figuring out how to connect a wire to a piece of glass or powder. You can find quartz resonators/crystals in just about all digital electronics.

I'm not recommending this sorting, because this is based on my goals and planned processes. But these are my categories:

ICs without legs - plastic tops from #1, plus #3, #4, and bga RAM chips
ICs with legs - 5-9, 11.
ICs low grade - anything else plastic with pins and gold bonding wire. voltage regs like sot223 packaged lm1117, 3-leg plastic transistors, misc. sot23 ICs. NOT mosfets. I don't go out of my way to get all of them, but if they are easy enough to pop off, I'll throw 'em in a "low grade ICs" bucket. Just I don't want to mix weird plastics like in LEDs and things with the other ICs which all seem to use the same kind of plastic.
Capacitors - MLCCs - 12, 17. Starting to look like dealing with these is above my comfort level, but I held on to these, especially when found on older equipment pre ~y2k
capacitors - tantalum - not pictured, but I do pull tantalum.
gold plated metals - I divide into groups, fully plated (no gaps) and partial plated. I hear some say to further divide into magnetic and non-magnetic pins.


Lastly, I make no claims that any of this is worth the materials and effort to process, either in small or industrial scale. I'm a hobbyist. Building up stock for recovery and refining takes a ton of time, which is nice, because there's tons to learn first, not the least of which is waste treatment. There are many ways to skin a cat, but here we must remember that we gotta do it cleanly during, and clean up afterwards, which may affect what processes you decide to use.
 
That last paragraph really applies to what I'm trying to do, I'd rather go slow and produce the least amount of waste. I'm also doing this as a hobby since I work in IT, had a lot of scrap laying around and it seems like the perfect second part to my other hobby which is lapidary work with stones and mostly Opal. I want to use the recovered metals to produce findings that I can use for mounting my stones in. I just like to learn as much as I can about my hobbies, just like everything else i like to do ;)

I'm going to start with the obvious parts, gold fingers first, then gold plated pins, and move on from there. I'm just trying to do the best job possible with the sorting while I'm there.

I do have to add that I always try and make my hobbies pay for themselves, so I try and do things correctly so the end products are worth something and I can keep the hobby going.
 

Latest posts

Back
Top