Copper-based thermal paste

[AmberClad]

Why are copper-based thermal pastes much less commonly used than silver/aluminum-based ones? I'm only seen one type made by GeIL, and even that is hard to actually find in the retail market.

Doesn't copper have superior thermal conductivity? Is it a cost issue? Something about the molecular structure that causes it to fill in the microscopic gap between the proc and the heatsink less effectively?

 

[Steve]

I thought silver was more conductive. It's definitely more marketable, it's got more bling.

 

[AmberClad]

I think pure silver is more conductive, but the "silver" thermal compounds on the market have very little silver content compared to the other materials in them.

 

[PottedMeat]

reactivity maybe? maybe it's hard to find a material to suspend pure copper in without losing the benefits of high thermal conductivity?

 

[CycloWizard]

Originally posted by: PottedMeat
reactivity maybe? maybe it's hard to find a material to suspend pure copper in without losing the benefits of high thermal conductivity?

This is my guess as well. Copper oxides are very good insulators (as I found out when my car battery connections corroded a while back ). Aluminum and especially silver are more resistant to oxidation and have similar or better conductivity.

 

[aka1nas]

Before Arctic Silver was readily available online, they used to have Copper Ice, which was cheaper and for whatever reason, less effective.

 

[Cogman]

Silver is a better thermal conducture the Copper is. It is fairly close but still better. Maybe I have not been looking hard enough, but I have not seen aluminum based compounds (especially ones that are ranked at any level)

A lot of people confuse electrical conductivity with thermal conductivity, They are NOT the same, As I recall, Diamonds have the highest thermal conductivity of any material, but they are electrical insulators.

The wiki has a small list of thermal conductivity of materials. Copper is 401, Silver is 429, Aluminum is 237, and Diamond is 900-2320 W/(m*K)

 

[Veramocor]

Originally posted by: CycloWizard

Originally posted by: PottedMeat
reactivity maybe? maybe it's hard to find a material to suspend pure copper in without losing the benefits of high thermal conductivity?

This is my guess as well. Copper oxides are very good insulators (as I found out when my car battery connections corroded a while back ). Aluminum and especially silver are more resistant to oxidation and have similar or better conductivity.

-5 chemical engineering points CycloWizard.

Aluminum oxidizes far more easy than copper. The kinetics are very quick. The only reason aluminum doesn't fully oxidize is that a barrier of corrosion product prevents any oxygen from diffusing to fresh metal.

If you were to make an aluminum powder I would suspect small particles of aluminum would be almost all aluminum oxide (high surface are to react with oxygen.

Tony

 

[CycloWizard]

Originally posted by: Veramocor
-5 chemical engineering points CycloWizard.

Aluminum oxidizes far more easy than copper. The kinetics are very quick. The only reason aluminum doesn't fully oxidize is that a barrier of corrosion product prevents any oxygen from diffusing to fresh metal.

If you were to make an aluminum powder I would suspect small particles of aluminum would be almost all aluminum oxide (high surface are to react with oxygen.

Tony

P'shaw - that's chemistry! I hate chemistry. No, chemical engineers are not chemists. (Just anticipating the usual flow of questions that I get when I tell people I know just about zero chemistry. )

 

[highwire]

Agreed, oxides of silver are probably more conductive thermally than copper oxides. I know that is the case electrically. And silver is also less likely to react with the carrier or surfaces to produce nuisance products. But, copper is in the running thermally.

Re: Aluminum's secret, its invisible oxide film.

If you were to make an aluminum powder I would suspect small particles of aluminum would be almost all aluminum oxide (high surface are to react with oxygen.

Tony

The rapidly formed oxide formed on aluminum is very thin. Even in the form of common aluminum pigment (microscopic flakes), the oxide is only a small percentage of the metal.
As you may know, aluminum is used in propellants and explosives. In this application, usually, the finer, the better. Only the metallic aluminum is useful in the reaction. So, again, I suspect, not much oxide form even with very small particle sizes.

Once-upon-a-time I had a Government contract to modify several large aircraft. One of the tasks dealt with some 0000 gauge(big) aluminum cables. The oxide is a problem for connections especially when many 100's of amps are involved. I was concerned. But, the answer was at the local industrial electric distributor - a big swagging tool, special terminals and this zinc metal goop to butter the connections with prior to swagging on the terminals. Oh yeah, wire brush to clean, bright metal first. Problem solved in style. All for an oxide film no one can see.

 

[JRich]

Originally posted by: Veramocor

Originally posted by: CycloWizard

Originally posted by: PottedMeat
reactivity maybe? maybe it's hard to find a material to suspend pure copper in without losing the benefits of high thermal conductivity?

This is my guess as well. Copper oxides are very good insulators (as I found out when my car battery connections corroded a while back ). Aluminum and especially silver are more resistant to oxidation and have similar or better conductivity.

-5 chemical engineering points CycloWizard.

Aluminum oxidizes far more easy than copper. The kinetics are very quick. The only reason aluminum doesn't fully oxidize is that a barrier of corrosion product prevents any oxygen from diffusing to fresh metal.

If you were to make an aluminum powder I would suspect small particles of aluminum would be almost all aluminum oxide (high surface are to react with oxygen.

Tony

That's why we use BCl3 when we plasma etch aluminum to scavenge the AlO2 of the Al before we hit it with Cl2

 

[highwire]

That's why we use BCl3 when we plasma etch aluminum to scavenge the AlO2 of the Al before we hit it with Cl2

I give up, what are we making here?

 

[highwire]

Well, Cl2 is going to make the Al disappear, so this must be some kind of chem milling, right?

 

[JRich]

Originally posted by: highwire

That's why we use BCl3 when we plasma etch aluminum to scavenge the AlO2 of the Al before we hit it with Cl2

I give up, what are we making here?

The Al interconnects for chips I work in a fab.

 

[PlasmaBomb]

Cool. Do they not use copper in a lot of the interconnects now though?

 

[mmonnin03]

JRich, you must work for Intel?

I work for Micron and we use Copper interconnects, at least at our Fab in Manassas.

For one thing, copper migrates. Copper Will Not stay where you put it. Put it in some solution for thermal paste and heat it up....it is probably useless at transfering heat.

And besides just because a metal is good at transfering electricity doesnt mean it is the best at transfering heat. Properties somewhat go hand in hand but they are different.

 

[JRich]

I work for Freescale. They use Al for the interconnects with W vias.

 

[silverpig]

Originally posted by: JRich
That's why we use BCl3 when we plasma etch aluminum to scavenge the AlO2 of the Al before we hit it with Cl2

ECR?

We use Cl2 to anisotropically etch silicon, and BCl3 for GaAs I believe.

Or is your process isotropic?

 

[JRich]

It was a RIE/ICP etch. The process is anisotropic.

 

[mmonnin03]

Freescale, a section of motorola...occording to Wiki on Oct 6, '06.

Ahh I wish I knew more about the Etch areas to know about what chemicals they use during.....oh I might get my arse fired if I keep talking. I may not be an expert in Dry/Wet etch but I may know enough to cause some trouble.

 

[JRich]

Freescale is it's own company now. It just recently went from a publicly traded company to a private one.

 

[chusteczka]

Originally posted by: AmberClad
Why are copper-based thermal pastes much less commonly used than silver/aluminum-based ones? I'm only seen one type made by GeIL, and even that is hard to actually find in the retail market. ...

The development of x-ray and photographic films produces a by-product comprising a silver slurry.
See: Silver Recovery

I have always assumed this is where the thermal paste manufacturers obtain their materials from, which is most likely more readily available than a similar substance with copper.

 

[mmonnin03]

Hmmm I dont work in the photolithograhy department but I doubt we would be using something as conductive as silver in our resist films. I cant think anywhere that we use silver in the fab nor have even heard of that metal even mentioned once.

 

[JRich]

There is no silver in our fab. That would be bad.

 

[Necrotech]

Ummmm You could use a suspended synthetic graphite

The average thermal conductivity is 470