Sign in with
Sign up | Sign in

Graphene Enhances Efficiency of Heat Spreaders by 25 Percent

By - Source: NC State

A scientist discovered a way to cool semiconductors much more efficiently than a traditional copper-based heat-spreader can.

According to Jag Kasichainula, an associate professor of materials science and engineering at NC State, graphene could be key to improve simple cooling methods.

The scientists developed a heat-spreader that consists of a copper-graphene composite that was attached to an "electronic device" with an indium-graphene interface film. Tests showed that the increased conductivity by using graphene can dissipate heat about 25 percent faster than pure copper. Kasichainula also noted that graphene is less expensive than pure copper, which could reduce the cost of heat spreaders in the future.

In a paper published in the April 2012 issue of Metallurgical and Materials Transactions B claims the researcher describes test results in which a 200 micron thin film of copper-graphene deliver a thermal conductivity of 460 watts per meter per kelvin (W/mK), up from just 380 W/mK of electrolytic copper at a temperature of 27 degrees Celsius. The conductivity of copper-graphene increases with lower temperatures and decreases with substantially higher temperatures. Kasichainula said that the conductivity was 510 W/mK at -23 degrees Celsius and 440 W/mK at 77 degrees Celsius.

There are 51 Comments. B
Top Comments
  • 32
    Bloob , April 13, 2012 8:09 PM
    Somehow it feels like technology will jump 10 years ahead when we finally master graphene production. It's almost as good as duct tape.
  • 27
    neon neophyte , April 13, 2012 8:10 PM
    Carbon; is there anything it can't do?
  • 20
    neon neophyte , April 13, 2012 8:27 PM
    Quote:
    "The conductivity of copper-graphene increases with lower temperatures and decreases with substantially higher temperatures. Kasichainula said that the conductivity was 510 W/mK at -23 degrees Celsius and 440 W/mK at 77 degrees Celsius."

    So the law of diminishing returns strikes again. What's the conductivity at 100C? Since some hardware gets above 77C...If its worse than copper this should be relegated to lower power devices. Since you know carbon isnt known for its heat conductivity.


    Diamond is one of the better known heat conductors actually. Diamonds being made of carbon.
Other Comments
  • 32
    Bloob , April 13, 2012 8:09 PM
    Somehow it feels like technology will jump 10 years ahead when we finally master graphene production. It's almost as good as duct tape.
  • 27
    neon neophyte , April 13, 2012 8:10 PM
    Carbon; is there anything it can't do?
  • 10
    brickman , April 13, 2012 8:19 PM
    A Graphene CPU cooler in my Corsair Graphite 600t. A perfect match :D 
  • 12
    Bloob , April 13, 2012 8:19 PM
    memadmaxSo, what does this mean boys n girls?You can get an extra 100mhz overclock on ur bulldozer!=D

    It means we might see passively cooled 78xx-series... ( or more likely 88xx or 98xx )
  • 8
    shin0bi272 , April 13, 2012 8:21 PM
    "The conductivity of copper-graphene increases with lower temperatures and decreases with substantially higher temperatures. Kasichainula said that the conductivity was 510 W/mK at -23 degrees Celsius and 440 W/mK at 77 degrees Celsius."

    So the law of diminishing returns strikes again. What's the conductivity at 100C? Since some hardware gets above 77C...If its worse than copper this should be relegated to lower power devices. Since you know carbon isnt known for its heat conductivity.
  • 4
    drwho1 , April 13, 2012 8:25 PM
    While reading this, all I could think off was... Why don't just try to make the whole thing off this material? Would it work without the copper? If possible, how lower temperatures can/could get?
  • 5
    nikorr , April 13, 2012 8:27 PM
    Cool stuff!
  • 20
    neon neophyte , April 13, 2012 8:27 PM
    Quote:
    "The conductivity of copper-graphene increases with lower temperatures and decreases with substantially higher temperatures. Kasichainula said that the conductivity was 510 W/mK at -23 degrees Celsius and 440 W/mK at 77 degrees Celsius."

    So the law of diminishing returns strikes again. What's the conductivity at 100C? Since some hardware gets above 77C...If its worse than copper this should be relegated to lower power devices. Since you know carbon isnt known for its heat conductivity.


    Diamond is one of the better known heat conductors actually. Diamonds being made of carbon.
  • 5
    suddenstop , April 13, 2012 8:33 PM
    shin0bi272"The conductivity of copper-graphene increases with lower temperatures and decreases with substantially higher temperatures. Kasichainula said that the conductivity was 510 W/mK at -23 degrees Celsius and 440 W/mK at 77 degrees Celsius."So the law of diminishing returns strikes again. What's the conductivity at 100C? Since some hardware gets above 77C...If its worse than copper this should be relegated to lower power devices. Since you know carbon isnt known for its heat conductivity.


    Well maybe they wouldn't get that hot anymore. Place a fan on it and as temp comes down the sink pulls more heat. The cooler it gets the cooler it gets. Also, hot areas of the sink would have slightly less thermal conductivity than cooler places - better distributing the heat load across the entire sink. My cpu at 4G stays under 60c and my gpu hangs out at 80c. In both cases this technology would be a massive improvement.
  • 10
    computernerdforlife , April 13, 2012 8:37 PM
    lostmyclan: Your English is broken beyond repair. Even "cooper" can't help you. Perhaps you can "why u shut up".
  • 6
    JamesSneed , April 13, 2012 8:43 PM
    Quote:
    Carbon; is there anything it can't do?


    Nice one.
  • 3
    trumpeter1994 , April 13, 2012 8:47 PM
    Sounds like this could be revolutionary for heatsink design
  • -7
    shin0bi272 , April 13, 2012 9:00 PM
    suddenstopWell maybe they wouldn't get that hot anymore. Place a fan on it and as temp comes down the sink pulls more heat. The cooler it gets the cooler it gets. Also, hot areas of the sink would have slightly less thermal conductivity than cooler places - better distributing the heat load across the entire sink. My cpu at 4G stays under 60c and my gpu hangs out at 80c. In both cases this technology would be a massive improvement.

    True but if youre willing to risk your hardware on it first be my guest. And note your comment about 80C gpu... at 77C its barely any better than plain copper. So if you extrapolate that to temps above 77C it will be worse than copper.

    And Neon: When did they start making graphene out of diamonds? Oh yeah they dont. they use 1 atom thick sheets of carbon which makes it closer to the graphite in your pencil than the diamond on your finger and a poor conductor of heat.
  • 1
    CaedenV , April 13, 2012 9:26 PM
    This is great news! Even with recent die shrinks servers are still having heat issues, this could really help with the 1U markets.
    Also the GPU market could benefit for this as well, perhaps paving the way for high end single slot cards again? Or perhaps better passive heat sinks?
    You still need good air exchange to pump the heat out of your case, but anything that can better transfer the temperature from the CPU to the air is something worth moving towards... especially if it is cheaper than what we are already using.
  • 13
    Kamab , April 13, 2012 9:28 PM
    shin0bi272True but if youre willing to risk your hardware on it first be my guest. And note your comment about 80C gpu... at 77C its barely any better than plain copper. So if you extrapolate that to temps above 77C it will be worse than copper. And Neon: When did they start making graphene out of diamonds? Oh yeah they dont. they use 1 atom thick sheets of carbon which makes it closer to the graphite in your pencil than the diamond on your finger and a poor conductor of heat.


    Graphene is extremely thermally conductive. Carbon has many allotropes, best not to generalize about all of them at once.
  • 7
    830hobbes , April 13, 2012 9:47 PM
    shin0bi272True but if youre willing to risk your hardware on it first be my guest. And note your comment about 80C gpu... at 77C its barely any better than plain copper. So if you extrapolate that to temps above 77C it will be worse than copper. And Neon: When did they start making graphene out of diamonds? Oh yeah they dont. they use 1 atom thick sheets of carbon which makes it closer to the graphite in your pencil than the diamond on your finger and a poor conductor of heat.

    Neon was responding to your statement that "carbon isnt known for its heat conductivity." Clearly, he showed that it is. One of the important determining factors of thermal conductivity is the strength of the bonds between atoms (similar to hook's constant in a spring). C-C bonds are strong bonds. The reason graphite doesn't conduct heat well is that there is very low bond strength BETWEEN layers. In fact, this is a commonly used principle for creating thermal insulators: make lots of layers.
  • -4
    Zingam_Duo , April 13, 2012 10:28 PM
    graphene this graphene that graphene my ass... when are we going to see anything that actually benefits us but articles how great graphene is?
  • 0
    freggo , April 13, 2012 10:57 PM
    computernerdforlifelostmyclan: Your English is broken beyond repair. Even "cooper" can't help you. Perhaps you can "why u shut up".


    Kindly remember this is not a US site. THG is owned by a French company. so let's not criticize non-native English speakers for their use of the -american-English language (how good is YOUR French or German?) but let's judge by the content of a post as it relates to the topic!

    Smile and the world smiles with you :-)

  • -2
    __-_-_-__ , April 13, 2012 10:59 PM
    drwho1While reading this, all I could think off was... Why don't just try to make the whole thing off this material? Would it work without the copper? If possible, how lower temperatures can/could get?

    yeah it would. but it would cost even more then your entire house.
Display more comments