If geeks love it, we’re on it

Researchers demo compelling successor to silicon

Researchers demo compelling successor to silicon

Carbon atoms linked into a sheet of graphene

A group of researchers from a trio of Japanese energy and technology institutes have demonstrated a ground-breaking process (PDF) which permits for graphene growth on a silicon carbide substrate.

“Graphene is a promising contender to succeed the throne of silicon in electronics,” the researchers said. “To this goal, large-scale epitaxial growth of graphene on substrates should be developed. Among various methods along this line, epitaxial growth of graphene on SiC substrates by thermal decomposition of surface layers has proved itself quite satisfactory both in quality and in process reliability.”

Epitaxial growth refers to the process of epitaxy, where a film of crystal of the same structure and orientation as a seed crystal is grown on a base. In this instance, the film of graphene is being grown from a base of silicon carbide through high heat to make a transistor, the basic building block of all electronics.

Graphene, meanwhile, is a hexagonal array of carbon just one atom thick. Graphene has been lauded for having the highest electron mobility of any substance known to man. Electron mobility describes the relative ease at which electrons are compelled to form a current in the presence of an electric field like one might find in a PC component. Graphene also boasts an exceptionally low resistivity at room temperature and has been praised for its possibilities in the burgeoning field of spintronics.

Experiments with the substance, such as those at IBM, have created 26GHz transistors that exhibit FET-like behavior. Field effect transmission has been used in silicon electronics for decades to control the flow of electrons to represent binary ones and zeroes through subtle variations in an electric charge; this process is known as switching.

Though silicon transistors are destined to be as small as 11nm if Intel gets its way, future shrinks below 10nm will leave silicon unable to contain the electrons necessary for transistor switching. It is hoped that smaller and more robust materials like graphene will be able to stave off electron tunneling in the preservation of Moore’s Law.

While other experiments in materials like indium phosphide have yielded 1THz transistors, no single technology has made gains like graphene. Invented in 2007, graphene transistors have already eclipsed the performance that silicon transistors took 40 years to achieve.

In spite of the material’s merits, however, the researchers note that there are obstacles ahead, saying, “The challenges for this graphene-on-SiC technology, however, are the abdication of the well-established Si technologies and the high production cost of the SiC bulk crystals.”

With a price of $1000 to create a slice of graphene that measures no longer than the width of human hair, perhaps “challenge” is too light a word. Even so, researchers continue to develop next-gen materials like graphene to ready the industry for the fast-approaching day when silicon is atomically incapable of being advanced.

Comments

  1. Tim
    Tim By the time die shrinks get down to 11 nm, if graphene or some other material is not yet economically viable, we may have to finally accept the fact that multicore cpus and their speeds have gotten as good as they are going to get.

    With the speeds and power efficiencies we already have at 45 nm and 32 nm, no one should be complaining by the time we get down to 11 nm.
  2. Zuntar
    Zuntar Dr. Thrax, can you post a linky to your source please!!

    I work for Cree, Inc. I kinda know a little (OK, a fair amount) about SiC substrates and epitaxy.
  3. Myrmidon
    Myrmidon I just applied to Cree, Inc. Surely you can put in a good word for me?
  4. Thrax
    Thrax The source has been added to the lead.
  5. Zuntar
    Zuntar
    Myrmidon wrote:
    I just applied to Cree, Inc. Surely you can put in a good word for me?

    YGPM;)

    @thrax Thank you sir!

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!