IBM opens up graphene bandgap, edges closer to commercialization
By Donald Melanson 
posted January 28th 2010 1:54PM
posted January 28th 2010 1:54PM
Graphene transistors have long been touted as the next big thing to deliver a true leap in electronics of all sorts, but there's been a few considerable limitations holding them back from fully replacing silicon. IBM now says it's managed to overcome one of the biggest hurdles, however, and has announced that it's been able to open a "bandgap" for graphene field-effect transistors (or FETs). As EETimes reports, that's important because while graphene does have a higher carrier mobility than silicon, it doesn't have a natural bandgap, which has so far kept the on-off ratio of graphene transistors far lower than their silicon counterparts. Of course, IBM insists that its still only just scratched the surface, and says that it's already hard at work on opening up an even wider bandgap, achieving even higher electric fields, further improving the on-off current ratios of graphene FETs.
@Bill Gates
Looks like some failed HTML injection there Mr. Gates.
@Delta
Oh weird, there was a hanging div tag in your comment there a moment ago.
thank God engadget is back to normal news that doesn't have "Apple" in it.
@sumx4182 Do you think the iPad will adopt this technology?
@mhope only if it helps with the menstrual cramping and bloating...
Wooooooooooosh
The sound of this article going over my head.
@Ted2
Whew, I thought I might be the only one.
@NaJaKwa
It says "IBM perfected a cool new toy that threatens to usurp the "round hole/square peg" game [of technology]". Just in a different (really long) language...
@Ted2
Atoms in a crystal share energy levels, which are then referred to as "bands". There's a conduction band (when electrons have enough energy to leave their atoms and go wondering around, conducting electricity) and a valence band (where electrons don't have enough energy to conduct, and are stuck to their parent atoms). The gap between these two bands is called the "bandgap".
Metals have overlapping bands (negative bandgap), so nearly all of their electrons can move around and conduct electricity, making them good conductors. Insulators have large bandgaps, making it difficult for an electron to jump to the valence band conduct electricity. Semiconductors have small bandgaps, meaning that if we give the electrons enough energy, they'll be able to conduct through the semiconductor.
Bandgaps are very important in all aspects of electronics. By engineering junctions (the place where two materials meet has a warped bandgap depending on the materials), we can get some interesting electrical properties. For example, we can make current flow in only one direction and make a diode.
Since graphene had no bandgap (the conduction and valence bands were the same), it would conduct like a metal. That isn't useful, since we have no control over how much it conducts. By altering the material to separate the bands, graphene may have a future as a semiconductor.
@KarlW
Nice explanation. Q: if graphene isn't a natural semiconductor and will require constant alteration (extra cost?), does the "enhanced carrier mobility" really offer a justifiable alternative to silicon?
@KarlW Thanks Karl!
@KarlW
Excellent summary. I feel able to actually ingest the info in this article now, thanks to you!
@Bandigolo
The process they go through to make electronics grade silicon is just insane, so complexity probably isn't much of an issue. Silicon is just really useful because it's the second most abundant element in the Earth's crust (after Oxygen), and because its natural oxide SiO2 (glass) is a great insulator and at the nano level actually lets materials diffuse through it (through glass!). It's a really handy material.
Also:
"Insulators have large bandgaps, making it difficult for an electron to jump to the valence band conduct electricity." should read:
"Insulators have large bandgaps, making it difficult for an electron to jump from the valence band to the conduction band and conduct electricity."
I'm glad the Ipad is done with so we can get our old Engadget back.
IBM stuff? Lazerbeam fusion articles? Yes please!
I can't see graphine going commercial in the next 10 years without some major cost reductions. Silicon is cheap and abundant and we haven't reached the end of what can be done with it nor do I think we will in the next few years. Silicon is and will stay the dominant commercial semiconductor.
@The Shen
You sound much smarter than I, so I will agree with about silicon. I thought graphene's first application would be to replace indium in LCDs. I thought indium supplies are supposed to run out soon.
@The Shen Cost reductions? They haven't even made a prototype yet as it's still in the research stage, how can you even know or complain about what the price would be!?
@Randomness I think he is implying high costs due to not enough future demand and hence no mass production benefits... as with all new technologies. (except apple, always expensive =P). Or maybe the costs of making this artificial bandgap..
My electronics professor told me that it would take another maybe 20-30 years before hitting the silicon ceiling of performance.
I think a discussion in this topic is beyond most readers in Engadget. You would actually need to have an EE degree to actually understand the actually semiconductor physic behind all this (getting my EE degree this May). It is nice that the industry is looking at different materials other than silicon.
@nuclearecko808 I do have an EE degree and work in microprocessor design. This is really cool research-y stuff, but building SRAMs, much less microprocessors, is still a ways off. I think they're referring to what you know as the p-n junctions (pay attention in Solid State Device Physics!). It's a problem for computers when the 0's are 0's and the 1's aren't 1's! :-)
This is HUGE. Silicon days are counted!
Have you heard about the advancements in graphene growth via CVD? Cheap graphene is coming :)
it doesnt matter, there is no way to make graphene at a high enough level for commercial utilization. how to grow graphene is still much under research at major universities and labs. i agree @The Shen that we're a long way from actually using graphene based FETs
This one time, at bandgap...
@Zanzibar BB McFate
Nice...
I heard a rumor that the iPad will support graphene in 2014. Engadget, can you confirm or deny? Maybe this rumor warrants an entire new article.