Graphene Won't Replace Silicon in CPUs

CommanderFrank

Cat Can't Scratch It
Joined
May 9, 2000
Messages
75,399
According to a report by Bit-tech, IBM has reversed its initial assessment of the uses for graphene. Initially, IBM touted graphene as a replacement for silicon in computer manufacturing.


Graphene, in its current research state, may not replace silicon as the main material for building CPUs, but continued research should eventually lead it to a strong foothold in the future of nanoelectronics.
 
That's to bad. It had so much potential, but from a processing/manufacturability point of view, they probably thought it wasn't feasible.
 
i saw this coming two years ago - the new bubble will be topological insulators.

they are superior for a number of reasons, and actually have a band gap. Most of the known materials have small gaps (few tenths of an eV - for reference I think Silicon is 1.1 eV) but the search is on for others that are truly insulating.

TI > graphene because a) could in principle have a reasonable on /off ratio b) the high mobility surface states are protected, even in the presence of weak disorder c) novel quantum phenomenon are expected at the interface between TI and other materials - with the potential that they can be exploited.
 
I think they are trying to state that a FET (ie what was shown to got 100 ghz) isn't going to replace your normal PNP or NPN transistor.

This has always been known.

The fact is that Graphene doesn't form a band gap (what you need for pnp type transistors) unless you can use defects, or are able to create ribbons at most ~10 nm.

The problem is that you have to basically control where the defects are on an atomic scale, or if you create a ribbon you need to be able to select how the ribbon terminates. (Zig-Zag vs arm-chair type bonds.)

Unfortunately we are unable to perform such techniques. And of course if we ever are able to directly manipulate single atoms on a large scale (not just using STM) we really don't need graphene...

(Full disclosure, I am a Ph.D. grad student currently writing my Thesis on graphene... So I sort of know graphene to say the least.)
 
There's research going on ATM as well as a number of published articles on creating a bandgap with graphene. The statement that graphene won't replace silicon for transistors is therefore inaccurate. The Slashdot discussion on this earlier today had a lot of good links.

Sure, it won't be easy, but I'm quite sure we'll be seeing graphene being used for transistors :)
 
There are other applications for graphene that could arise sooner. Touchscreens, displays in general, ...
 
Or you can just use materials whose surface states are robust in the presence of defects, so long as they don't break time reversal symmetry....

Really suggest you consider hopping on the gravy train... TIs will be bigger than graphene.

Disclose: I am also a phd student working in condensed matter... nanomagnetism and now thin film growth of bismuth selenide.
 
Really suggest you consider hopping on the gravy train... TIs will be bigger than graphene.

Disclose: I am also a phd student working in condensed matter... nanomagnetism and now thin film growth of bismuth selenide.

Oh we don't need money, yet... But we already have some experiments going on with TIs, just isn't my research. While I can't say graphene is easy, anyone can make a device from exfoliated samples... now good samples, that is different. (Though I must say we use epitaxial graphene from SiC... one of the harder methods to control the growth though it gives some really good samples.) TIs from what I have seen are basically worse then getting a good GaAs sample. IE you need a good sample, then a good isolation, but then you need to access that isolated state... And for my research (Low Temp, High B-Field STM/STS) that would mean ~1 month wasted per sample. And of course seeing as I am done in ~3 months I really don't need to start with anything new...
 
i saw this coming two years ago - the new bubble will be topological insulators.

they are superior for a number of reasons, and actually have a band gap. Most of the known materials have small gaps (few tenths of an eV - for reference I think Silicon is 1.1 eV) but the search is on for others that are truly insulating.

TI > graphene because a) could in principle have a reasonable on /off ratio b) the high mobility surface states are protected, even in the presence of weak disorder c) novel quantum phenomenon are expected at the interface between TI and other materials - with the potential that they can be exploited.

Ummm... WHAT?! Seriously, I just spent 30 minutes looking up over half of what you referred to. After looking into it though, I am now very excited by the the term "topological". :p
 
I think they are trying to state that a FET (ie what was shown to got 100 ghz) isn't going to replace your normal PNP or NPN transistor.

This has always been known.

The fact is that Graphene doesn't form a band gap (what you need for pnp type transistors) unless you can use defects, or are able to create ribbons at most ~10 nm.

The problem is that you have to basically control where the defects are on an atomic scale, or if you create a ribbon you need to be able to select how the ribbon terminates. (Zig-Zag vs arm-chair type bonds.)

Unfortunately we are unable to perform such techniques. And of course if we ever are able to directly manipulate single atoms on a large scale (not just using STM) we really don't need graphene...

(Full disclosure, I am a Ph.D. grad student currently writing my Thesis on graphene... So I sort of know graphene to say the least.)

Does graphene even have defects? I've had a class on semiconductor devices and I'm starting a quantum materials class this semester, so I'll be the first to say I know nothing about practical application. That said, it seems like the reason graphene and nanotubes are useful (tensile strength, etc.), is the lack of any defects in the structure.
 
Work has been proceeding in order to bring perfection to the crudely conceived idea of a insulator that would not only supply inverse reactive current for use in unilateral phase detractors, but would also be capable of automatically synchronizing cardinal grammeters. Such an element is Graphene

The original insulator had a base-layer of prefabulated amulite, surmounted by a malleable logarithmic casing in such a way that the two spurving electrons were in a direct line with the pentametric proton. ... The main neutron was of the normal lotus-o-delta type placed in panendermic semi-boloid field in the stator, every seventh conductor being connected by a nonreversible trem'e pipe to the differential girdlegravity on the 'up' end of the grammeters.

Forty-one manestically spaced grouting phases were arranged to feed into the quantum slipstream a mixture of high S-value phenylhydrobenzamine and 5% reminative tetryliodohexamine. Both of these liquids have specific pericosities given by P = 2.5C.n^6-7 where n is the diathetical evolute of retrograde temperature phase disposition and C is Cholmondeley's annular grillage coefficient. Initially, n was measured with the aid of a metapolar refractive pilfrometer ... but up to the present date nothing has been found to equal the transcendental hopper dadoscope. ... Undoubtedly, Graphene has now reached a very high level of technical development. It has been successfully used for operating nofer trunnions. In addition, whenever a barescent skor motion is required, Graphene may be employed in conjunction with a drawn reciprocating quantom planar arm to reduce sinusoidal depleneration.

DISCLOSURE: I was a student at one time, and I have also seen all of the Star Trek movies.
 
Does graphene even have defects? I've had a class on semiconductor devices and I'm starting a quantum materials class this semester, so I'll be the first to say I know nothing about practical application. That said, it seems like the reason graphene and nanotubes are useful (tensile strength, etc.), is the lack of any defects in the structure.

It can - any amount of contaminant on the surface is a defect. If you mean defect as in missing carbon atom, that i'm not sure about.

If you place the sheet on a substrate, if there are any contaminants they will deform the sheet and this causes scattering. Although there is possibility for this to be exploited as well, in certain circumstances.
 
First thanks for the congrats... still have to finish writing it though. As for a job I have a written offer but need to graduate first...

As for defects in graphene. Oh yeah there are defects. Just type the words graphene defects into google and you will the idea.

Some occur intrinsically in the growth process. IE grain boundary between rotated planes, "puckers", vacancies. But you have to remember that often instead of say a single atom missing on the normal honey comb structure, everything will rearrange itself so that you get say 2 pentagons and 2 heptagons. (stone wales defects)

Honestly when I say defect I usually mean anything besides the ideal graphene. So that also includes any deposited atoms. (There are plenty of papers out there on what those can do.) In addition your dialectic will change the way graphene behaves (again treated as a defect). And heck the gate ends up acting as a giant defect.

In fact one of the biggest problems they had with exfoliated graphene were contaminates from Si02 and residue. Hence device people have moved to creating suspended graphene, which is no easy feat in the slightest.

So basically what I am saying is low mobility graphene usually means it has a lot of defects from the creation process. And if you look at most of the papers out there with Low temp mobility measurements only in the thousands, that usually means that if you looked at their sample under Scanning Tunneling Spectroscopy it would look like crud.

To see some images of graphene defects just type in google images: Scanning Tunneling Microscopy graphene defect
 
Graphene is a fabrication NIGHTMARE. The idea of making an atom-thick structure with CONTROLLED and known properties several million times over PER CHIP is mind boggling. Now, it's pretty mind boggling that chips exist at all so it was worth the effort to research. There all ready are other things coming out, like graphene as a top contact that is very transparent for displays instead of ITO or F-SnO2 etc.

For conventional processors, we'll go 3-D with multi-core first. That is, two (or more) layers of transistors. As for "conventional" materials, GaAs is always an option because of the far higher mobility (less heat from fast switching = more GHz).
 
Sure, it won't be easy, but I'm quite sure we'll be seeing graphene being used for transistors :)
The ability to manufacture on worthwhile scale is what holds back many impractical lab projects. IBM of course is one of the worst offenders when it comes to breathlessly announcing breakthroughs that never show up in commercial scale products.
 
Back
Top