Make your own liquid metal compound, galinstan

[mackerel]

Doing my usual youtube scan this morning, the first part of this video may be of interest here. In short, take some gallium, add indium, and it turns liquid. This combo is well known, but this is the first time I seen it done in practice in this way. The addition of a small amount of tin can reduce melting point further.

I've done delids and used a regular 1g tube of Conductonaut which has done about 5 applications so far with an unknown amount remaining. I've limited use to under IHS only but if it were cheap enough, you could use it anywhere (where there isn't aluminium anyway).

Looking first on ebay I found some indium quickly, but no gallium. amazon.co.uk had a few vendors. It will be an area for future research but based on https://en.wikipedia.org/wiki/Galinstan you'd be looking at roughly 3 parts gallium to one part indium, with a smaller than indium quantity of tin. Roughly speaking, I think I could make around 30g of galinstan for a comparable cost to buying a 5g tube of Conductonaut.

There will be some risks and unknowns of course, maybe some ratios are better or worse in this use case. Purity of the samples used might have unknown effects, and that's not even thinking about the potential longevity of it. Given my low usage rate it is just easier for me to buy more Conductonaut if I need it, but this could be a future video for a tech tuber...

 

[Alaric]

I love this place. That's the kind of forward thinking that opens avenues of action, instead of just theory.

 

[Blaylock]

...potential longevity... I'm no chemistry major but Gallium and Indium are both base elements. Aside from them separating the only thing I could see effecting them long term would be a chemical reaction such as tarnishing which should not effect its thermal conductivity. Add in the fact that the two elements are neighbors in the same group (Boron) would suggest that they would play well together for a very long time.

 

[Alaric]

I'm no chemistry major

Me either, but your off the cuff analysis of this situation beats the snot out of my pathetic (lack of) knowledge on the subject. Nice! All I remember from high school chemistry was my teacher (Mrs. Dow) being mildly amused at my mixing pepsin, nitric acid, and some now forgotten ingredients in test tubes, and dropping them from the second floor window to the sidewalk. Also her surprise at how close I got to my goal....Thank God I grew up then and not now. The FBI would have hauled me off. LOL

 

[mackerel]

See my earlier statement more as a generic disclaimer. I just don't know what the risks may be. Even if we look at commercial offerings like Conductonaut or the other one who's name escapes me right now, there is often talk about the longer term stability of such, regardless of the mechanism.

 

[Johan45]

Cool Labs is the other one Mack. Just a quick search here seems Gallium/Indium is readily available https://www.amazon.com/Gallium-99-99-Pure-Grams-RotoMetals/dp/B01M6AYJNV https://www.amazon.com/s/ref=nb_sb_...efix=indium,industrial,173&crid=3W3S9XQ62VSTP

 

[trents]

But the companies who market these liquid metal TIMs probably buy the ingredients in large enough quantities to get significant discounts. It's hard for me to imagine much savings by doing it yourself. Especially when you are only using it occasionally as opposed to a mass production environment. But if you want to try it just to satisfy your own curiosity then that's another story.

 

[Alaric]

I'm a believer in "Because you can".

 

[mackerel]

But the companies who market these liquid metal TIMs probably buy the ingredients in large enough quantities to get significant discounts. It's hard for me to imagine much savings by doing it yourself. Especially when you are only using it occasionally as opposed to a mass production environment. But if you want to try it just to satisfy your own curiosity then that's another story.

Your pricing may be different, but as stated in the 1st post, I estimate I could easily make 25g+ of LM for less than the price of buying a 5g tube of Conductonaut. Of course, this is pointless if you don't need 25g+ of the stuff. You might fairly say you get other stuff than the compound, but syringes, alcohol wipes and cotton buds cost practically nothing. Pricing just doesn't scale well to smaller quantities and it is quite frustrating when they don't often do bigger sizes. MX-4 is an exception to that, and I went for a 20g tube which should last a while, unlike the smaller tubes.

 

[Alaric]

MX-4 is an exception to that, and I went for a 20g tube which should last a while, unlike the smaller tubes.

Made that same purchase for the same reason.

 

[trents]

Made that same purchase for the same reason.

Me as well. They used to sell MX-2 in 30 gm tubes and I would buy them.

 

[Alaric]

If I ever snag a decent mobo for benching it, I have a 980 BE that I'd like try delidding, and some budget liquid metal TIM would be just the ticket.

 

[Blaylock]

Are you guys talking about this tube? (banana for scale)

 

[mackerel]

On original topic, I've gone and ordered some. Don't mind testing this myself. Have ordered 10g of Indium, and 25g of "liquid metal bullion" which is almost certainly gallium based on some research, but there is a bit of uncertainty there. I managed to find that on ebay, rather than the more expensive but named gallium on amazon. Still looking for a small quantity of tin but not sure this is essential.

 

[Alaric]

Are you guys talking about this tube? (banana for scale)
View attachment 197404

Nah, this one:


Bearing in mind that's a size 10 1/2 American made Red Wing boot. If it were a Nike it would be a size 13. LOL

Still looking for a small quantity of tin but not sure this is essential.

Is it possible the tin is some sort of bonding or mixing agent for the other ingredients? I'm now looking at you Blaylock, you chem maestro. LMAO!

 

[mackerel]

Reference the galinstan page earlier, a small quantity of tin is supposed to further lower the melting point than gallium+indium together. The Conductonaut web page does state it contains gallium, indium, tin, has a temperature range starting 10C, and it also gives a density. I think that might be enough clues to deduce a rough ratio of ingredients, which I'll attempt when I have more time. I've ordered some tin now. Actually, far more than I needed, it is quite cheaper than the other two. My limiting factor will be the gallium, as I have the other two in excess. I'll think more about a test plan later, I could afford to make smaller batches of different ratios for test. A temperature controlled environment to check freezing point may be more difficult but not impossible e.g. I could place a sample in a container in a water bath, add ice to slowly cool the bath and observe the freezing point.

Anyone know on what CPU generation did Intel stop using solder and switch to TIM? I need some test subjects...

 

[Alaric]

Haswell and Sandy Bridge.

edit:The above series were the last of the soldered ones.

 

[Johan45]

If I ever snag a decent mobo for benching it, I have a 980 BE that I'd like try delidding, and some budget liquid metal TIM would be just the ticket.

You do realize these are soldered on right?

 

[Alaric]

You do realize these are soldered on right?

Do now-

 

[The Wicker Man]

https://www.reddit.com/r/watercooling/comments/6uf2gc