Coollaboratory liquid metal, a new thermal compound

Slaugh · Post by **Slaugh** » Wed Aug 31, 2005 2:54 pm

I found this new thermal compound from Coollaboratory, it's made of liquid metal and it's supposed to beat Arctic Silver 5 by a few degrees with a heat conductance of 82 W/mK. If you understand german, here's the original page. (the previous link is an english translation).

There are more pictures and details about this new kind of thermal compound on the Coollaboratory website. Anybody heard of that company before?

Bluefront · Post by **Bluefront** » Wed Aug 31, 2005 3:09 pm

Looks like a mercury compound to me.....the only metal I'm aware of that's liquid at room temps. Highly poisonous. I doubt it could be imported into the USA.

Why don't they tell the exact composition of this stuff?

cotdt · Post by **cotdt** » Wed Aug 31, 2005 3:32 pm

I'm sure it performs well but it's probably mercury and toxic. Or maybe modified liquid gallium with a lower melting point.

Also it is an electrical conductor I'm pretty sure. Based on specs it can thermally conduct 10X better than thermal compound but still not nearly as well as aluminum or copper.

I'm interested! If only they allow importation into the USA...

cotdt · Post by **cotdt** » Wed Aug 31, 2005 3:54 pm

Those metallic syringes look awesome! It gives me an urge to inject it into the next stranger that walks by...

Slaugh · Post by **Slaugh** » Wed Aug 31, 2005 3:57 pm

They already sell this compound on eBay for EUR 6,90.

I don't know what they use, but it's definitely not mercury:

It exclusively consists of high-temperature conductive metals. "Coollaboratory liquid metal" does not contain any non-metallic additives. For this reason it has the excellent heat conducting capability. It is liquid at room temperature (like mercury), however, is non-poisonous and has a high moistening capability for the most different materials. "Coollaboratory liquid metal" adheres to the cooler and CPU by means of the effective adhesive powers. It does not need any sealing. Of course, also in case of a vertical installation.

But I don't like this part:

"Coollaboratory liquid metal" is optimised in its composition for the application with high-quality copper coolers. Please do not use any aluminium coolers.

cotdt · Post by **cotdt** » Wed Aug 31, 2005 4:09 pm

I wonder if they chip to USA...

Of course those after performance or passive cooling will be using copper or copper-hybrid coolers.

I think the material might be Gallium-Indium-Tin Eutectic.

Slaugh · Post by **Slaugh** » Wed Aug 31, 2005 4:21 pm

Well... Maybe!... I was wondering if this compound had chemical issues with aluminium... But I think it's more an advice than a warning since mostly all good coolers have at least a copper base...

darthan · Post by **darthan** » Wed Aug 31, 2005 4:23 pm

It may not have mercury but that no aluminum warning makes you wonder. When you expose aluminum to air its surface immediately oxidises and forms a hard shell that protects it. If you put mercury on it that strips off the oxidised coating and exposes new aluminum to oxidation which is in turn stripped off. Thus mercury "melts" aluminum. Other metals can probably do this too and so that probably explains why this compound can't be used with aluminum. This is also why mercury can't go on airplanes-they're made of aluminum.

cotdt · Post by **cotdt** » Wed Aug 31, 2005 4:27 pm

darthan wrote:It may not have mercury but that no aluminum warning makes you wonder. When you expose aluminum to air its surface immediately oxidises and forms a hard shell that protects it. If you put mercury on it that strips off the oxidised coating and exposes new aluminum to oxidation which is in turn stripped off. Thus mercury "melts" aluminum. Other metals can probably do this too and so that probably explains why this compound can't be used with aluminum. This is also why mercury can't go on airplanes-they're made of aluminum.

Yeah I think you're right. Well the better coolers have either copper or gold-plated bases so there shouldn't be a problem. When you have a thermal interface this effective, it's probably a good idea to rip off the CPU's heatspreaders and clamp the CPU cooler directly onto it. Someone should try this.

frostedflakes · Post by **frostedflakes** » Wed Aug 31, 2005 5:04 pm

Wow, this has to be one of the most interesting things I've seen in a while.

I may take the plunge and try some...

mathias · Post by **mathias** » Wed Aug 31, 2005 5:33 pm

Slaugh wrote:
"Coollaboratory liquid metal" is optimised in its composition for the application with high-quality copper coolers. Please do not use any aluminium coolers.

Aren't heatspreaders made of aluminum?

Slaugh wrote:Well... Maybe!... I was wondering if this compound had chemical issues with aluminium... But I think it's more an advice than a warning since mostly all good coolers have at least a copper base...

But why would they want to unneccessarily hurt their sales?

frostedflakes · Post by **frostedflakes** » Wed Aug 31, 2005 5:42 pm

Athlon64 heatspreaders are copper. What makes them appear to be aluminum is a thin nickel coating. This is on their to prevent air from oxidizing the copper, which would hurt performance.

I can't speak for P4, but I'd assume their heatspreaders use the same construction.

EDIT: Went ahead and ordered a tube. I'll be sure to post results when I get it.

sockeye · Post by **sockeye** » Wed Aug 31, 2005 7:01 pm

Weird...I knew this stuff looked familar. Just a few days ago, I was looking at this page, describing gallium and a gallium alloy that you can buy and science experiments you can do with them. The page contains this picture:

Which is the exact same picture found on the coollaboratory page. I think that pretty much tells us what they are selling (probably also where they buy it, and where they got their picture!)

Also included on the gallium page above is a picture of what the stuff does to aluminum:

John

StarfishChris · Post by **StarfishChris** » Thu Sep 01, 2005 3:14 am

If nothing else this stuff looks easy to spread on (= fewer problems) and should keep for a long time. Fortunately most good heatsinks (XP-120 etc) have a copper base, and if you're using a cheap aluminium-based heatsink you probably wouldn't care about getting other paste.

Bluefront · Post by **Bluefront** » Thu Sep 01, 2005 3:27 am

The heat-spreader on every P4 I've seen is aluminum. I have lapped quiet a few.....no doubt about it. I've seen little evidence of any protective coating.

Maybe one of you brave guys can drip some of this stuff on your P4-3.2 Northwood.....report on what happens.

mathias · Post by **mathias** » Thu Sep 01, 2005 6:12 am

StarfishChris wrote:Fortunately most good heatsinks (XP-120 etc) have a copper base

That's nickel plated. Besides aluminum and copper, there's still the question of how it reacts with nickel, silver and gold.

StarfishChris wrote:and if you're using a cheap aluminium-based heatsink you probably wouldn't care about getting other paste.

I wouldn't consider zalman flowers cheap. Also at least zalman and thermaltake heatpipe sandwich coolers appear to have aluminum bases, with the exception of the schooner.

mathias · Post by **mathias** » Thu Sep 01, 2005 6:16 am

Bluefront wrote:The heat-spreader on every P4 I've seen is aluminum. I have lapped quiet a few.....no doubt about it. I've seen little evidence of any protective coating.

At least this stuff might be useful for removing those heatspreaders.

StarfishChris · Post by **StarfishChris** » Thu Sep 01, 2005 7:05 am

FYI: The Allround-PC test used a P4 630, which means that either
- Newer P4s have a different (non-aluminium) heatspreader, or
- The heatspreader is the same across all P4s

Can someone confirm one of the above?

mathias wrote:
StarfishChris wrote:Fortunately most good heatsinks (XP-120 etc) have a copper base
That's nickel plated. Besides aluminum and copper, there's still the question of how it reacts with nickel, silver and gold.

So is the A64 heatspreader (and possibly the P4 one?) so it's arrived 3 years too late if it reacts with nickel.

Poor babelfish translation of some of the compound description on Allround-PC:

Material
Alloy from gallium and Indium as well as small quantities of precious metals

Application/costs
a syringe with approx. 1,5g of the alloy is f?ca. 6,90. erh?lich and should depending upon CCU f?ca. 10. 20 applications are sufficient.

Characteristics
electrically leading!! may come only with copper as well as the CCU into connection!!

mathias wrote:
StarfishChris wrote:and if you're using a cheap aluminium-based heatsink you probably wouldn't care about getting other paste.
I wouldn't consider zalman flowers cheap. Also at least zalman and thermaltake heatpipe sandwich coolers appear to have aluminum bases, with the exception of the schooner.

True. But many heatsinks come with copper bases, including those by Scythe, Arctic Cooling and Thermaltake, and Zalman heatsinks come in copper-only versions. I can't comment on GPUs.

frostedflakes · Post by **frostedflakes** » Thu Sep 01, 2005 8:34 am

On the eBay auction, they have a picture of this stuff being applied to an Athlon64 IHS, so I think it's pretty safe to assume that this will not react with nickel. At least not in the short term. Whether any long term oxidation will occur, I don't know.

Shadowknight · Post by **Shadowknight** » Thu Sep 01, 2005 8:40 am

...

Not touching this stuff with a 10-meter cattleprod. BURNING HOLES THROUGH ALUMINUM?!

BrianE · Post by **BrianE** » Thu Sep 01, 2005 10:17 pm

Shadowknight wrote:...

Not touching this stuff with a 10-meter cattleprod. BURNING HOLES THROUGH ALUMINUM?!

Not burning, it just keeps breaking up the aluminum oxide coating on top.... stuff like mercury (I guess gallium does it too) reacts weirdly to aluminum. Even though we're used to thinking aluminum is a stable, non-reactive metal, it's actually incredibly reactive, which is why untreated bare aluminum always looks whitish-silver rather than chrome-like.

According to the link, that pic is of a sheet of aluminum foil, even though at first glance it looks like a thick plate of it.

The DIY mirror making thing sounds neat.

Slaugh · Post by **Slaugh** » Thu Sep 15, 2005 7:41 pm

Coollaboratory has a new website now and it's multilingual. Here's what they're saying about the aluminium in the online manual:

Do not use cheap aluminum coolers. Their quality is not very high, and aluminum is not resistant against our Liquid Metal. Aluminum could form an alloy with "Coollaboratory Liquid Pro" which would react with the surrounding air's humidity. A black stain forms after a while, showing insulating qualities. ' Therefore: No contact of "Coollaboratory Liquid Pro" with aluminum. NO ALUMINUM COOLERS!

And here's another thing:

Like all metals, our Liquid Metal may act as an electric conductor. This forms no problem if applied according to our instructions, as your cooler is made of conducting metal as well. Make sure "Coollaboratory Liquid Pro" does not touch any electric components. This could lead to shortcuts when turning the computer's power on.

frostedflakes · Post by **frostedflakes** » Thu Sep 15, 2005 8:10 pm

My liquid metal TIM actually arrived about a week ago. The directions that came with it were in German, and I didn't realize there were English instructions online, so I just winged it.

So far temps are no better than with AS5, but as I mentioned, I didn't follow their installation instructions, so this may be why. Next time I take my computer apart I'll try to reapply the TIM and see if I can get better temps.

cotdt · Post by **cotdt** » Fri Sep 16, 2005 12:28 am

frostedflakes wrote:My liquid metal TIM actually arrived about a week ago. The directions that came with it were in German, and I didn't realize there were English instructions online, so I just winged it.

So far temps are no better than with AS5, but as I mentioned, I didn't follow their installation instructions, so this may be why. Next time I take my computer apart I'll try to reapply the TIM and see if I can get better temps.

if you're brave enough, what about testing with the heatspreaders removed?

frostedflakes · Post by **frostedflakes** » Fri Sep 16, 2005 9:03 pm

Temp results are with the IHS removed. I'm using a mobile CPU, so it didn't come with an IHS.

Slaugh · Post by **Slaugh** » Mon Nov 14, 2005 11:56 pm

I found this news on Matbe today. A user put some Coollaboratory liquid metal on a CNPS7000B-Cu LED, and the aluminium support of the heatsink is now in very bad shape!

Here are some pictures of the mess:

(Click on the pictures for a closer view)

dorion · Post by **dorion** » Tue Nov 15, 2005 5:03 pm

I need to get me some of this stuff, great way to ruin someones chrome finish, aluminum and chrome are pretty in the same situation with these things. Also seems like a easy way of making a hole in your case for that new fan.

Badger · Post by **Badger** » Tue Nov 15, 2005 5:21 pm

frostedflakes - update?

frostedflakes · Post by **frostedflakes** » Tue Nov 15, 2005 6:11 pm

Nope, not really. Put the stuff on and didn't get any better temps than with cured AS5. Haven't messed with it since. If anybody wants some I still have a bunch left, if somebody wanted to cover shipping consider it yours.

jojo4u · Post by **jojo4u** » Sat Jan 07, 2006 2:42 pm

German review against 2 other high-end compounds (AS5 and MX1). Test system Silverstone NT-02 on Venice 3000+.
The Liquid Pro compound leads by 1°C, not very much.
Applying is easy - dismounting is another story: The HS and the CPU get glued together, you need quite force to seperate it. Also needed: metal polish (e.g. http://www.nevrdull.com/) to get rid of the stuff.

http://www.dirkvader.de/frame.php?site= ... index.html

Coollaboratory liquid metal, a new thermal compound

Coollaboratory liquid metal, a new thermal compound

Gallium or a gallium alloy.