PORTLAND, Ore. -- Carbon is predicted to replace semiconductors, conductors and even insulators in future electronic devices and its already happening one application at a time. The most recent conquest by pure crystalline carbon--graphene--is a new composite material that offers a more efficient and less expensive way of cooling electronic devices than copper alone.
By fabricated heat spreaders from a unique copper-graphene composite and connecting it to microchips using an indium-graphene interface film, heat is dissipated 25 percent faster than with the pure copper heat spreaders used today. The techniques is also cheaper to produce, since it uses less copper that conventional heat spreaders, and will be particularly useful for electronic devices that generate a lot of heat, such as lasers and power semiconductors.
According to North Carolina State University researcher Jag Kasichainula, the thermal conductivity of both his copper-graphene composite, which is attached to the electronic device, and the indium-graphene interface film, are higher than copper, and yet are cheaper due to the exscalating price of copper. The copper-graphene composite can be deposited in films as thin as 200 microns, according to the effective medium approximation (EMA) method of modeling thermal conductivity and determining the interfacial thermal conductance between copper and graphene.
Kasichainula has also created a blueprint for manufacturers wishing to quickly retool for fabrication of the graphene-copper heat spreader using an electrochemical deposition process.
I guess there is a graphite and there is a graphite...the link you quote @Rwatkins contains the following bullet point: Our graphite is GREATER THAN 4 X COPPER AND NATURAL GRAPHITE...so one graphite is 4 times better than another graphite, this is for those who are not yet confused yet ;-)...Kris
Now I am a little baffled. A look at
finds a graphite product claiming 4X improvement over copper (although fine print reading shows that 4X is only in one plane and the perpendicular to that plane is "only" near copper).
For a bit of an education, for those unfamiliar with issues such new technology can take, we should all remember the time it has taken to commercialize FETs from SiC (discussions and work started in the mid 1970's, first commercial devices only a couple of years ago). The challenges are similar, a problematic chemistry that depends heavily on processing means and can very immensely due to what seem like obtuse causes.
It would be much efficient to remove heat using water micro-pipes...that 25% advantage that Colin is quoting in the article is pretty small...generally a new technology requires at least 2x improvements upon the existing one...Kris
Yes, graphene should help with heat, power consumption, cost, speed and weight. There are still many engineering hurtles to clear, and lots of room for innovation, but that's where the fun is-isn't it?
As a one time ASIC designer, I welcome the improved heat dissipation using graphene. I remember having a major battle with heat and power on the devices. The end result was clock gating, parallel paths with slower clocks, in general a pain. If only we could get power into the core and get the heat out using graphene, time will tell..
I eventually envision different formulations of grapheme replacing the metal, semiconductors and insulators in mobile devices, the most amazing benefit of which will be ultra-light weight. Ever notice how heavy your smartphone is compared to a styrofoam block that same size? The weight is mostly from all the copper on circuit boards and for the interconnect inside the chips. With those alone replace with grapheme, you smartphone could become almost as light as that block of styrofoam!
Regarding controlling the band gap, RPI team report tuning the bandgap of graphene anywhere from 0 (a pure metallic conductor) all the way to .2 eV (a semiconductor suitable for infrared detectors) by simply controlling the amount of humidity inside the package:
Rice University reports turning grapheme into an insulator by bonding it to a monolayer of hydrogen:
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.