Graphene has >10x better conductivity than copper

Tzupy · Post by **Tzupy** » Thu Feb 28, 2008 8:29 am

Link to article at PhysicsWorld: http://physicsworld.com/cws/article/new ... 8ED5798677
If graphene 'flakes' could be mixed with aluminium, we would get aluminium-graphene fins that are light as Al but better conducting than Cu.

AZBrandon · Post by **AZBrandon** » Thu Feb 28, 2008 9:22 am

Based on the article it sounds like it's both got very exciting prospects as well as a long development time before we'll see it in use anywhere.

jaganath · Post by **jaganath** » Thu Feb 28, 2008 10:33 am

is the conductivity of the heatsink fins really the biggest bottleneck in heatsink design? i would have thought it was the transfer of heat to air (still very inefficient).

jhhoffma · Post by **jhhoffma** » Thu Feb 28, 2008 10:55 am

jaganath wrote:is the conductivity of the heatsink fins really the biggest bottleneck in heatsink design? i would have thought it was the transfer of heat to air (still very inefficient).

Especially in passive or low-aiflow situations.

But whatever gets the OC-freaks another 10MHz FSB bump...

pipperoni · Post by **pipperoni** » Thu Feb 28, 2008 2:16 pm

In SPCR terms, I think the benefits would be more subtle and come in the form of thinner fins (thus presenting less air resistance) and less need for heatpipes to distribute heat evenly across the cooling surface area.

It would be neat to have a thermal super conductor, then you could do some silly things like make the side wall of your case your heatsink, you could even attach the CPU die directly do it (just don't punch your case too hard!).

I think graphene will have the same challenges as using plain old graphite as a heatsink, the thermal conductivity is good in two of the three axes and the material will be difficult to machine or otherwise manipulate to really be production worthy.

derekva · Post by **derekva** » Thu Feb 28, 2008 2:53 pm

How about magnesium? Works really well in old Porsche and VW motors...of course it can catch fire if it gets too hot.

"Hey, I just hit 5GHz air-cooled on my Opteron!"
<WHOOOSH!>
"Crud"

-D

jaganath · Post by **jaganath** » Thu Feb 28, 2008 3:50 pm

"Hey, I just hit 5GHz air-cooled on my Opteron!"
<WHOOOSH!>
"Crud"

Felger Carbon · Post by **Felger Carbon** » Thu Feb 28, 2008 5:13 pm

pipperoni wrote:In SPCR terms, I think the benefits would be more subtle and come in the form of thinner fins (thus presenting less air resistance) and less need for heatpipes to distribute heat evenly across the cooling surface area.

The Cooljag Falcon 92-cu already has .02" thick fins. Which bend really easily. Thinner, and you get lotsa cuts and lotsa lawsuits. Thin does not always win! (signed FC, poet laurentide)

jhhoffma · Post by **jhhoffma** » Thu Feb 28, 2008 5:42 pm

derekva wrote:How about magnesium? Works really well in old Porsche and VW motors...of course it can catch fire if it gets too hot.

But it looks so pretty from far away.

I grew up near Dow Chemical's main manufacturing plant and they have a huge climate-controlled dome for storing magnesium ingots. I remember one of my organic chemistry instructors (at a local university) tell me that if that building caught fire, you'd see it from Detroit, about 130 miles away!!

BillTodd · Post by **BillTodd** » Fri Feb 29, 2008 5:31 am

Thermal conductivity of magnesium is 150 W/(m.K)
Ally ~200
Copper 385
Diamond 900

The measurements of Graphene are of molecular thickness membranes (they would make interesting, and lethally sharp, heat-sink fins

)

I wonder though, if these membranes were layered whether they would retain their huge conductivity or lose it completely and become like graphite, a poor heat conductor (just 5 W/(m.K)).

albatros_la · Post by **albatros_la** » Fri Feb 29, 2008 8:12 am

There are carbon fibers which can have a thermal conductivity up to 1100-1200W/m/K if properly treated (just search for K1100 fibers). Usually they're used to manufacture high conductivity CFRP panels, so with an epoxy resin matrix (chronically low conductive). Nowadays there are some chemical industries able to carbonize full CFRP panels and turn them all to graphitic state in a manner which gives them really high thermal performance (the processes are usually called "chemical vapour infiltration and carbonization" and give to the material anisotropic high performances [let's say up to 400W/m/K in plane and about a tenth of the same value out of the plane]). By the way the production processes are really expensive and the material costs really a way more than for the simple and cheap aluminium and copper. This type of new frontline materials are currently under development and their domain of application is the aerospace industry, where there are a lot of concerns about lightweight structures able to withstand high heat loads. Personally, I prefer the aluminium: it's reliable, cheap, and it weights enough (so it's also a good sound dampener and a not-too-bad thermal capacitor). I would like to think that the main effort must be focussed to low power/high efficiency electronic devices which require littler sinks and could operate finely also with the today's sinks.

pipperoni · Post by **pipperoni** » Fri Feb 29, 2008 9:03 am

Felger Carbon wrote:The Cooljag Falcon 92-cu already has .02" thick fins. Which bend really easily. Thinner, and you get lotsa cuts and lotsa lawsuits. Thin does not always win! (signed FC, poet laurentide)

I'm secretly waiting for a heatsink that can double as a french fry maker.

FartingBob · Post by **FartingBob** » Tue Mar 04, 2008 2:32 pm

How long before we have to start wearing thick gloves when handling our heatsinks?