Over the past few decades, computers have radically changed the way we work, play, communicate and live. At the base of this revolution are semiconductors: materials that can act as either conductors or insulators depending on the energy applied to them. Silicon is the world’s predominant semiconductor material, but science marches on and researchers are excitedly searching for another, superior semiconductor to control the next generation of technology. A recent development by Professor Edward Conrad at the Georgia Institute of Technology could place graphene in that exalted position.
Graphene has ten times the conductivity of silicon, and is already used in some electrical contacts. However, in ordinary graphene samples this conductivity cannot be ‘turned off’, so to speak. To be a capable semiconductor, a material must possess a “band gap”: a significant difference in energy needed to excite a valence band electron into the conduction band. Generally graphene has no band gap: it is much too easy for a valence band electron to enter the conduction band. Some researchers have created graphene in microscopic ribbon shapes with band gaps of 100 meV; however this is not significant enough to use in circuitry and the ribbons are too thin to produce at scale anyway.
Prof. Conrad and his colleagues took a different approach, utilizing “epitaxial growth”. They heated silicon carbide (SiC) to temperatures of 1360°C. Some of the carbide decomposes into graphene, in layers bonded to the carbide substrate. Conrad’s team investigated an electrically inert “buffer” layer, and observed band gaps of over 0.5 eV.
There has been previous research into this technique of growing graphene, but with several key differences. Most teams focused on electronically active layers of graphene, rather than the buffer layer on which Prof. Conrad observed the high band gap. Professor Alessandra Lanzara in the University of California at Berkeley attempted a similar study in 2006, but ended up with a much lower band gap. “It turns out that crystalline order is extremely important to get this band gap, and they didn’t have that” says Conrad. Tight control of the growth temperature was necessary: at 1360°C there was a band gap of 0.5 eV, but at 1380°C there was no band gap at all.
Prof. Conrad’s colleagues at Georgia Tech are already attempting to construct graphene transistors. Of course, it may be while until a graphene-based computer arrives on your desk. Further experiments must be conducted to fully understand the role of temperature on graphene growth. The cost of producing these graphene sheets also must be considered. Conrad is optimistic though, saying that “The first [silicon] transistors they sold were $1,500. The point is, you get the device first, and you worry about the cost later.”
Leave a Reply