Processing RAM chips

[ahmadbayoumi]

Hell ALL,

Last week I was processing about 675g of IC chips from RAM sticks.

I followed Pat's guide exactly as mentioned in his PDF.

The yield was 0.4g of gold powder and I believe that this is a very low yield.

I'm not sure where is the gold gone?

I used more than 2.5 ml of HNO3 to dissolve all gold wires but only managed to get 0.4g.

I used small increments of HNO3 to be sure that there's no excess nitric.

And after every small addition of HNO3 (0.5+0.5+0.3+0.3+0.3+0.2+0.2+0.2+0.2 ml) I heated the solution for about 30 mins. for each HNO3 addition to speed up the reaction using a glass watch to avoid losing values and then check for remaining gold so that I can see if more HNO3 is needed to avoid excess nitric.

I don't know where's the gold as I was expecting at least 2g however I got only 0.4g

Steps:

1 - Incinerated
2 - Crashed (didn't use ball mill)
3 - Removed metallic parts/pins and removed Si wafers
4 - Washed/Panned
5 - Diluted HNO3 (30%) warm bath
6 - Washed 5 times with distilled water to remove any HNO3 traces
7 - Washed 5 times with tap water to remove any remaining HNO3 traces
8 - Warm HCL bath
9 - AR (Small increments of HNO3 55%) with heat (20-30 mins.) after each addition with total 2.7 ml of HNO3 55% and used a glass watch to avoid losses
10- Diluted with tap water and left the solution to cool
11- Filtered
12- Used SMB to precipitate gold
13- Left the solution over night
14- Got 0.4g of gold powder

Need you help PLS

 

[ahmadbayoumi]

 

[ahmadbayoumi]

 

[FrugalRefiner]

Looking at your first picture, it appears some of the chips are not completely incinerated. This will leave carbon which could be tying up the gold.

Dave

 

[solar_plasma]

Your ICs are not incinerated completely, there is carbon left and even bigger parts. Carbon adsorbs/absorbs (it's hard to me to remember what is what of those) some gold. Bigger parts can still hide gold bondings.

Are you sure to have denoxed properly?

If you had a lot of low yield ICs in this batch, you might have got it all.

 

[dannlee]

Looking at the 'removed silicon wafers' photo shows many of the chip 'cores' made into charcoal but left whole.

You'll also find PMs on the kovar conductors you've separated, every gold bonding wire leaves behind some amount where it was welded.

There is also a small amount to be reclaimed from the silicone dies themselves, again from the bonding wire attach points.

You've lost nothing until you decide to discard parts of the process. The bonding wires get scarred and kinked up and quit behaving like 'gold', they'll attach to glass and plastic to avoid clean rinsing, they'll get tossed out of the condensed material with too energetic swirl of water, and on and on...

I don't water wash chip ashes without at least three catchments, I use falling water (thin stream) and a tall form beaker inside a deep, large clear pyrex dish with that inside another broad pan - and that inside a new plastic cement mixing tray from HD or Lowes. Watch the water you use - often warm water will bubble with trapped air, do trial runs w/o values and make sure the water never turns 'foggy' as each little bubble can lift gold out and away. Don't try to do it all at once, you can make a process out of it and process it in stages, so at each stage you can check the tailings for values and adapt your method. Letting the grey water settle out overnight and gently rewashing it can flag if you've rushed the process too much, the glass sand in a beaker swirled and tapped at a slight angle for a while will collect the heavies where you can see them in a bottom edge...

 

[mls26cwru]

for ram chips, i wouldnt expect a whole lot of gold out of them... I would have to check my numbers, but i had a pretty poor yield from a pound that i processed. I think it was .5g-1.0g per pound. you may have gotten most of it.

 

[ahmadbayoumi]

Thank you all for your replies

FrugalRefiner and solar_plasma,

Yes it seems some chips are not well incinerated. However, I'm not sure what do u mean by tying up the gold ? Is it possible for carbon to absorb or tie visible gold wires ?

And if so, why It required 2.6 ml of HNO3 55% to dissolve these wires ?

Or carbon can absorb dissolved gold (AuCl2) ?

For denoxing, I didn't have to denox as I only used as HNO3 as required in small increments

For the used ICs, you can check the below pics.

dannlee,



Yes, I'm still didn't process Si wafers and metallic parts/pins. But I believe that I should have got more gold from the IC chips alone.

For your washing method, would you PLS explain more your steps and method details using pics if it's possible.

mls26cwru,

I got 0.4g out of 675g (~1.5 lb), so I should have recovered at least 0.75g. Thus, I'm still missing almost double what I have managed to recover so far.

 

[FrugalRefiner]

Thank you all for your replies

FrugalRefiner and solar_plasma,

Yes it seems some chips are not well incinerated. However, I'm not sure what do u mean by tying up the gold ? Is it possible for carbon to absorb or tie visible gold wires ?

And if so, why It required 2.6 ml of HNO3 55% to dissolve these wires ?

Or carbon can absorb dissolved gold (AuCl2) ?

Ahmad, if I've understood what was explained to me, carbon can both absorb and adsorb gold that is in solution, not the visible metallic gold wires. So as you dissolved your gold, some may have been absorbed into the carbonaceous residue (like water being soaked up by a sponge) and adsorbed (like food crumbs sticking to the outside surface of the sponge). It ties up your gold. Some of your nitric probably went to dissolving metals other than the gold you were targeting.

For denoxing, I didn't have to denox as I only used as HNO3 as required in small increments

You've done well to add the nitric in increments with the goal of using just enough, but in actual practice we almost always add just a bit too much trying to get the last bits into solution.

I got 0.4g out of 675g (~1.5 lb), so I should have recovered at least 0.75g. Thus, I'm still missing almost double what I have managed to recover so far.

It all depends on the ICs. Not all ICs contain gold. Some use aluminum bonding wires instead of gold. You could process pounds of them and get no gold at all.

If you still have the solid material that remained after you dissolved your gold, dry it and incinerate it. Bring it up to a dull red heat. Stir it as you incinerate to expose all the material to the air to ensure complete conversion of the carbon. Then try processing any remaining material one more time. If you've lost any gold, you should find it there.

Best of luck,
Dave

 

[Barren Realms 007]

On the chip's that you did not completly incinerate you could have some of the bonding wires still left in the chip pieces. You should finish the incineration on those as well.

 

[Harold_V]

On the chip's that you did not completly incinerate you could have some of the bonding wires still left in the chip pieces. You should finish the incineration on those as well.

One consideration is that any aluminum that may be present will cement dissolved gold, leaving it behind in the waste.
To ensure that values are not locked up, fully incinerate, to the point where free oxygen applied to the heated mass does not create glowing embers. The mass should be heated to at least 1,200°F. All carbon should be eliminated before dissolving values.

Once incinerated, I strongly recommend a wash with HCL. That would dissolve any aluminum that may be present. Rinse until the rinse water is clear, and allow to settle well before decanting. Filtering this solution would likely be troublesome.

Once well rinsed, I recommend a second incineration. Follow this incineration with a wash in dilute nitric, to remove any transient base metals. Rinse well, testing for the presence of silver.

Once well rinsed, dissolve the values by the solvent of your choosing (I used aqua regia, although I did not process escrap).

I couldn't help but note the general purple color of the material. Sure does make one think there's colloidal gold present.

Harold

 

[Barren Realms 007]

On the chip's that you did not completly incinerate you could have some of the bonding wires still left in the chip pieces. You should finish the incineration on those as well.

One consideration is that any aluminum that may be present will cement dissolved gold, leaving it behind in the waste.
To ensure that values are not locked up, fully incinerate, to the point where free oxygen applied to the heated mass does not create glowing embers. The mass should be heated to at least 1,200°F. All carbon should be eliminated before dissolving values.

Once incinerated, I strongly recommend a wash with HCL. That would dissolve any aluminum that may be present. Rinse until the rinse water is clear, and allow to settle well before decanting. Filtering this solution would likely be troublesome.

Once well rinsed, I recommend a second incineration. Follow this incineration with a wash in dilute nitric, to remove any transient base metals. Rinse well, testing for the presence of silver.

Once well rinsed, dissolve the values by the solvent of your choosing (I used aqua regia, although I did not process escrap).

I couldn't help but note the general purple color of the material. Sure does make one think there's colloidal gold present.

Harold

I had the same feelings.

 

[g_axelsson]

Reading Harold's post I took another look at the pictures and it suddenly hit me.

Your ashes looks too dark to me and the traces of large pieces of carbon among the dies means that you probably had a lot of carbon among the ashes too. When you dissolved the gold some of the gold chloride was absorbed by the remaining carbon. The phenomena is known as preg-robbing ( https://gsa.confex.com/gsa/2004RM/finalprogram/abstract_72717.htm ).
If you still have the material that you treated with AR I would recommend to incinerate it again until it is pale all the way through (and you can add the pieces of carbon from the beaker with the dies). Then treat it with AR again to see if you can leach any more gold out of it.

I don't think there is any colloidal gold in your washed concentrate even though the pictures are a bit to the purple side. Up to that point the only thing that was done to the material was incineration, crushing and washing and that shouldn't create a colloid. I think the color tone in the pictures is from a less than perfect white balance in the camera.

I've more or less restated a lot of what's been told in this thread already. Credit goes to those that already posted, I'm just the messenger. But there is one thing I noticed that no one brought up yet.

Steps:

1 - Incinerated
2 - Crashed (didn't use ball mill)
3 - Removed metallic parts/pins and removed Si wafers
4 - Washed/Panned
5 - Diluted HNO3 (30%) warm bath
6 - Washed 5 times with distilled water to remove any HNO3 traces
7 - Washed 5 times with tap water to remove any remaining HNO3 traces
8 - Warm HCL bath
9 - AR (Small increments of HNO3 55%) with heat (20-30 mins.) after each addition with total 2.7 ml of HNO3 55% and used a glass watch to avoid losses
10- Diluted with tap water and left the solution to cool
11- Filtered
12- Used SMB to precipitate gold
13- Left the solution over night
14- Got 0.4g of gold powder

Step 8, the warm HCL bath? Did you pour off the HCl afterwards or is that the HCl that made up the beginning of AR treatment? If you washed the concentrate in HCl then you probably dissolved quite a bit of gold. You need to incinerate when switching acids or you will dissolve some metal. The porous ashes would be phenomenal to hold onto nitric acid and that would form AR when adding HCl.

I once made an experiment. I dissolved some pins in nitric acid, then washed it once with water. When I added HCl the foils dissolved fully, proving to me that there need to be very little nitric to dissolve a lot of gold. I think that in low concentrations the NOx gases is absorbed by the liquid and forms nitric acid again which dissolves even more gold.

Göran

 

[patnor1011]

What Goran said, also Harold mentioned the same.
Step 8 is problematic. If you treated concentrate with HNO3 no matter how many water washes you do introducing warm HCl will dissolve some of your gold and that will be lost.
I do not use HCl and when I use Nitric wash and wash concentrate with water I never do HCl wash, next step is straight AR leach.
If you feel HCl wash is needed then you must incinerate concentrate before using it. Always incinerate before using HCl and Nitric washes for cleaning concentrate.

If you do HCl wash first, it may introduce some silver chloride to your concentrate. That is why I would go for just Nitric wash first then AR.

Inspect pins you collected with magnet, some of them might have wires attached.
I recently sent PM to another member where I pointed another thing. Proper incineration is a must - IC have to be white and crumbling without much force applied - in just squeezing them in fingers. Time spent incinerating will return in time saved crushing and sieving and repeated incinerating and so on.
Second plus in this is what I observed - if you crush not properly incinerated material your gold wires are incredibly soft and will smear on pins if force is applied say using simple mortar and pestle. I have had my own share of golden coloured pins which result in loss of easy gold. I mean it is easier to get bulk or nearly all of your gold as wires and not to have most of it attached - beaten to pins. You then get most of it at first go and can toss magnetic pins in your stock pot bucket to get any remaining bits of gold from tip of pin where wire was soldered later and without need to spend more acids and time to get unnecessary smeared gold from body of a pin. Hope it is understandable enough if anyone have suggestion how to say all this in proper plain english feel free to correct me, it is early morning here and I did not slept last night.

 

[dannlee]

ahmadbayoumi: You got many different makes and models of RAM, if there was aluminum bonding in a few sticks you still should have more yield from 675g of IC chips.

Just a note... We use the phrase 'Lost values' often though usually it means just not in the main drop:

Cycling acid washes through a 'stock' pot, acting to saturate waste washes with copper does reclaim fractions of PM's (Precious Metals) carried off by overlapping process traces;
Repeating the water separation & settling out of lighter ashed glass portions usually can capture the 5-20% of wild wire escapes from a careful first wash;
Dissolving the base metal legs & heat plates completely is a nuisance but finishes reclaiming a batch nicely.

Once I've gotten the volume of concentrates reduced I've used large chromed ball bearings rolled/spun around in a glass bowl (by hand) to better wash out the dark color of charcoal in the concentrate 'sands', puts the char into solution as a black ink that won't settle out, gentle wash/decant steps then brings out the light grey and white silica sand color - yes it mauls the soft gold badly but since the next step will be AR any gold carried away in the ink and discarded to waste must be extremely small.

Also on that first Aqua Regia step I'll decant and water wash then repeat a weak AR wash two or three times, each time bringing to a simmer heat, until the AR wash remains crystal clear and/or tests no-gold using stannous. For small lots I'm talking covering the concentrates with simmering distilled water, adding 10 or 15cc of HCl and a few drops of Nitric. This seems to work to recover most of gold trapped in the texture of the glass and/or dregs of the incinerated epoxy in the glass pores, and while increasing the bulk of fluid needing to be evaporated/deNoxxed it SURE gives peace of mind on actually discarding the depleted concentrate.

 

[ahmadbayoumi]

Thank you all very much

Very informative replies .. Need to read them all carefully then will give my feedback

Thanks a gain for all your usual support.

 

[mls26cwru]

I just finished processing a pound of ram chips, and the yield was 0.3 grams. However, I did not process the silicon dies or metalic legs yet even though i would not expect much from them.

M

 

[necromancer]

let's not forget the resale value of tested working ram.

20 sticks of 1GB DDR2 ram at $3.00 each would be $60.00. with much less work involved.

to test you only need a motherboard + power supply + the power button & leads from a computer case

 

[jason_recliner]

You also need a drive from which you can boot a utility like Memtest.
Motherboard memory tests are very rudimentary, extending only as far as whether they can "see" the address. Once RAM has been pulled and probably not handled correctly, you can get all sorts of problems, the most common being bleeding between adjacent cells. I don't consider motherboard acceptance as "tested".

 

[marlyn]

There are also some gold traces found on memory boards. Are they worth to be processed (separated from gold fingers and black chips) via AP, for example?

Sorry if this was answered somewhere, already. I did not find it. I'm currently in forum reading / evaluating phase with some first experiments behind. Thanks for understanding.