Graphene: Fast, Strong, Cheap, and Impossible to Use

http://www.newyorker.com/magazine/2014/12/22/material-question

by ezhil • 12 years ago

218 88 12 years ago

I have no doubt that graphene will change the world, but as the article mentions, it may take a long time to get there.

The thing that bothered me the most was the mention of the patent race. Everything was "patent patent patent" - which means instead of a discovery in lab A spurring a totally new way of thinking in lab B, it restricts lab B from ever thinking that way.

Without the race to patent every single minuscule discovery about graphene and its properties, advancing early discoveries could move at an exponentially faster pace. Don't grab the ball and run with it - grab it and keep it to yourself.

I know all the arguments for patents, but I can't shake the feeling that they do nothing but stifle innovation and discovery by locking up research just when it gets interesting.

(When the measure of success is how many patents are filed, not how many discoveries are made... or how you just beat the Chinese to filing a patent, I'm just a tad less hopeful.)

crusso• 12 years ago

which means instead of a discovery in lab A spurring a totally new way of thinking in lab B

Without the patent system's protections, lab A would be unlikely to divulge its discoveries to lab B.

jacquesm• 12 years ago

That's fine, they're going to have to market their stuff some day.

DannyBee• 12 years ago

So the grants they depend on require divulgence to get the grant. It's not like they are funding themselves ...

swamp40• 12 years ago

If you think provisional patents are stifling innovation, you should see what happens when you tell these researchers that they no longer have the possibility of getting incredibly filthy rich someday from their discoveries - because all patents have been eliminated and a new era of "share and share alike for the common good" is upon us.

warfangle• 12 years ago

Do the researchers themselves even own the patents -- or will the company/university they work for own the patents?

You should see what happens when they realize they aren't ever going to become filthy rich someday from their discoveries... just like everybody else.

baddox• 12 years ago

I imagine lawyers being more upset about that than researchers.

hyperbovine• 12 years ago

Nobody I know in research is in it for the money. There are much, much more direct ways for technically gifted people to get rich--look up "quantitative finance".

locopati• 12 years ago

Hell, yeah! Good thing Jonas Salk buttoned-up his polio discoveries so he could make mad cash.

Oh, wait, he didn't.

"There is no patent. Could you patent the sun?" - Salk

gwern• 12 years ago

You mean, 'the possibility of getting incredibly filthy rich, and the likelihood of making someone else with a more basic patent incredibly rich'? A patent thicket is the paradigmatic tragedy of the anti-commons...

Osmium2• 12 years ago

you should see what happens when you tell these researchers that they no longer have the possibility of getting incredibly filthy rich someday from their discoveries

My personal experience with this is that patents are only used as a kind of status symbol, in that they help researchers acquire new grants/funding ("look, it's marketable!") and make themselves look good to their institution. I'm yet to meet a researcher who's actually made any money off one, and most are by no means adequately compensated for their work.

That all said, I have met former researchers who've made successful start-ups from their academic work, and made money that way, and I'm sure patents were an essential part of that, but by that point they've left academia anyway. For them, I never had the impression that the prospect of making money was the motivating factor for getting into research. It was all just serendipity. Though I'm sure there are exceptions.

rodgerd• 12 years ago

Yes, I've noticed no-one works on compilers unless they can use IP protection to get filthy rich. Or web servers. Or operating systems.

placebo• 12 years ago

This assumes most of innovation is driven by hope of getting incredibly filthy rich, an assumption which I can provide plenty of counter-examples for. Eliminating all patents is of course counter productive as their advantage is obvious, but the current patent situation is far from being optimal (understatement of the day) so I'm certain a whole range of improvements are possible which will encourage innovation and do not necessitate switching over to communism...

mc808• 12 years ago

a new era of "share and share alike for the common good" is upon us.

Bringing discoveries and inventions into the public domain for the good of society is the primary reason for having a patent system. To the extent that it inhibits this outcome, the system is failing.

noonespecial• 12 years ago

Because it worked so well for Tesla. (Nikola that is)

GFK_of_xmaspast• 12 years ago

The Bayh-Dole act wasn't some singularity of research innovation.

higherpurpose• 12 years ago

So much for standing on the shoulders of giants for progress. It's more like standing on your own shoulders.

hiou• 12 years ago

The problem with patents is not whether or not they should exist, but the massive friction involved in licensing them to others. If the patent system had more requirements or controls about requiring a fair system of licensing and ownership division it could serve its original purpose much better in my opinion.

bdamm• 12 years ago

Perhaps there is room for a specialized marketplace?

JackC• 12 years ago

which means instead of a discovery in lab A spurring a totally new way of thinking in lab B, it restricts lab B from ever thinking that way.

This is not necessarily true.

Patents, unlike copyrights, are additive. If you write a book based on the Harry Potter characters, JK Rowling owns the copyright to your derivative work. But if Lab A patents a way of making graphene, Lab B is free to research and patent non-obvious refinements.

So you tend to see a tree of patents by different people:

1. Awesome new invention Foo! (Lab A) 1.A. Small but commercially important refinement of Foo. (Lab B) 1.A.i. More performant version of small refinement of Foo. (Lab A) 1.A.ii. Different way of increasing performance on small refinement of Foo. (Lab C)

... where you definitely need to license Patent 1 from Lab A if you want to sell your own Foo, but if you want to sell a Foo with Lab C's performance improvements you need to license Patent 1.A and 1.A.ii as well.

This is why there is a "race to patent every single miniscule discovery" in the first place -- Lab A getting Patent 1 doesn't remove the incentive to discover refinements. It is not obvious that the search for refinements would happen faster without this incentive.

new299• 12 years ago

In order to develop on Patent 1, Labs B and C should officially license Lab As patent.

As I understand it there's no "research exception" for patents, and they don't just apply to "selling". I'm basing my understanding on this article:

http://www.nature.com/nbt/journal/v25/n11/full/nbt1107-1225....

It's also true that researchers rarely take any notice of patents, and the issue just gets ignored.

AnthonyMouse6• 12 years ago

Patents, unlike copyrights, are additive. If you write a book based on the Harry Potter characters, JK Rowling owns the copyright to your derivative work.

No, they're both the same. One of the exclusive rights under the Copyright Act is the right to make derivative works, so you can't make a book based on Harry Potter characters without JK Rowling's permission. If you do anyway, JK Rowling can sue to stop you but she can't start selling your version herself and taking all the money without your permission. Nobody can sell the derivative work without permission from both authors, which is as it is with patents.

This is why there is a "race to patent every single miniscule discovery" in the first place -- Lab A getting Patent 1 doesn't remove the incentive to discover refinements.

Of course it does. It requires the discoverer of the refinement to obtain the permission of Lab A before they can make any money, and the refined version competes with Lab A's version. So Lab A is better off to exclude the refined version from the market unless they can claim nearly the entire margin from it as they can with the original version, and the patent allows them to do that. Which leaves little or no incentive for anyone other than Lab A to make refinements.

The same thing happens with copyright. Disney makes movies based on ancient fairy tales because they're in the public domain. ABC's Once Upon A Time uses characters from Disney's Frozen because they're the same company. Universal Studios pretty much never makes movies with Disney characters, even though they theoretically could with Disney's permission, because Universal would be doing all the new work even though Disney could claim as much of the revenue as they like.

It is not obvious that the search for refinements would happen faster without this incentive.

That's the fundamental trade off with patents. You grant a monopoly to one entity on the theory that the incentive of the monopoly will get that entity to invest in the research. The benefit of that incentive then has to be weighed against the cost of excluding all other potential innovators from the market for the term of the patent.

This is why patents work best in markets where innovation is slow. If patentable innovations happen only infrequently and/or there are only a small number of major players then the cost of the monopoly is smaller because fewer incremental innovations are being excluded by the patent.

This is one of the big reasons that software patents are so abominable -- the more innovation that would be happening without patents, the higher the cost of issuing patents grows.

saidajigumi• 12 years ago

This is not necessarily true.

In biotech/pharma research, it is absolutely true. If a competitor's patent lands first on targeted genes/proteins/etc. it can shut down an entire lab's line of work permanently. The staff involved usually get reshuffled to other lines of interest, and any work they'd done is simply written off, never to see the light of day.

WalterBright• 12 years ago

The US airframe industry was so hamstrung by patents that innovation and progress shifted to Europe. It took the government to force the industry to accept a patent pool and move forward. http://en.wikipedia.org/wiki/Wright_brothers_patent_war#The_...

throwawaykf055• 12 years ago

This is actually a myth that was recently busted. See this paper, which presents compelling empirical evidence of the aircraft industry from that time to show that there was no such hold-up:

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2355673

Unfortunately, this is one of those myths, like the one of Watt's patents holding up development of steam, that gets repeated so often that people accept it without question.

_greim_2• 12 years ago

it restricts lab B from ever thinking that way

Is it not possible to look at existing patents and build on them, filing new patents which build on the existing ones? What would really restrict lab B from ever thinking that way would be if lab A never filed a patent, and just kept the discovery locked up as a trade secret.

nkurz4• 12 years ago

In principle this would be possible. In practice, the standard advice is that inventors should avoid looking at competitors' patents to reduce the danger of "treble damages" due to "willful infringement". This fear is likely overblown, but I've certainly never heard of anyone excited to hear that their competitor has published a patent because they hope that the new information revealed in the patent will help them to innovate better.

I'm sure it's industry specific, but I've don't think I've ever even heard someone claim that they learned something from a patent that they could not have legally learned in some other way. Given that this spread of knowledge is part of the ideal that is presented for the existence of patents, it would be nice to hear from anyone who works in area where patents are felt to actually help in the spread of knowledge.

I'm willing to believe that promising a monopoly can help to incentivize research and innovation, but I'm doubtful they positively affect the rate at which knowledge is shared.

rayiner2• 12 years ago

advancing early discoveries could move at an exponentially faster pace.

Faster pace towards what though? Someone swooping in, doing the last 1% of the work, and making all the profit. E.g. it's sad how much more money Apple has made on Siri than SRI did. Patents do a pretty awful job of making sure researchers and inventors get compensated. But they're one of the few things to tip the balance in favor of researchers and inventors.

nkurz4• 12 years ago

But they're one of the few things to tip the balance in favor of researchers and inventors.

My guess coming into this thread was that the business people and lawyers would be more pro-patent than the researchers and inventors. Of course one can be both, and I don't know the backgrounds of many participants, but it's interesting that in this thread both you and JackC have strong backgrounds in law. Have you found many researchers or inventors without comparable legal expertise who perceive the balance to be tipped in their favor? And moreover, that think this advantage is maintained rather than threatened by patents?

WildUtah3• 12 years ago

There are a mountain of patents to block innovation in speech recognition. Dragon and many others are ready to knock down any kind of real innovative small company that has success.[0] SRI never had a chance in that market to make it alone and that was understood from the beginning.

Only Apple's and Google's cash and legal muscle makes it possible for new speech recognition tech to be commercialized in the face of patents blocking innovation. Anyone smaller and less intimidating can just be crushed by incumbents.

"researchers and inventors"

That's disingenuous. It's the researchers and inventors that want the patent system out of software, information, and communication.

[0] http://www.nytimes.com/2012/10/08/technology/patent-wars-amo...

swish41• 12 years ago

Great article. IMO, CVD is the most promising scalable, manufacturing method, but the following advances have to happen:

- Able to grow Graphene of arbitrary thickness on a wide variety of surfaces - Achieving CVD at low temperature - Control of grain size, ripples, doping level, and number of layers - Coming up with a cost-effective way to handle to the Transfer step, or eliminating it entirely

As the current market for graphene applications is driven by the production of this material, there is a clear hierarchy in how soon the applications will reach the user or consumer. Those that use the lowest- grade, cheapest and most available material will be the first to appear, probably in a few years, and those which require the highest, electronic- quality grades or biocompatibility may well take decades to develop.Also, because developments in the last few years were extremely rapid,graphene’s prospects continue to improve.

bsder• 12 years ago

I'm surprised that they seem to be concentrating so little on CVD.

Yes, CVD is expensive, but it's not rocket science anymore. We can do all manner of things to pizza-sized wafers of silicon in CVD. Shining lasers just above the surface to heat the gas but not the substrate is one of the easiest ones.

It feels like the labs are all chasing patents rather than actually doing research.

matthewmcg• 12 years ago

Perhaps I am misunderstanding the terminology, but isn't graphene always the same thickness, i.e. one atomic layer?

morgante• 12 years ago

Once again fooled by a headline which is equally applicable to the realtime graphing framework...

Xcelerate3• 12 years ago

I believe you're thinking of "graphite". Although they should make a lightweight version of their framework called "graphene" haha.

Xylakant• 12 years ago

There's a project called "grapehene" which is a nicer ui for a graphite backend. https://github.com/jondot/graphene

deweller• 12 years ago

Ha. So true.

Xcelerate3• 12 years ago

I have a question for someone on here. When I began research in grad school in 2012, graphene had already been discovered eight years previously. I understand that before this happened, a "2D" atomic layer of a material had never been isolated.

What I want to know is: why was this considered impossible before then? I do molecular dynamics simulations (and am in fact performing simulations on graphene currently), and an individual layer of graphene easily stays intact in both a vacuum and other environments. The capability to simulate one layer of graphene was also available way before 2004.

So what was it that made single-layer isolation so astonishing at the time? It seems like it would have been a perfectly reasonable assumption to me (but like I said, I suppose I'm biased by the benefit of hindsight).

jxm262• 12 years ago

I'm not really qualified to answer this, but I recently came across a similar question that might shed some light (old post though).

https://www.physicsforums.com/threads/why-is-graphene-only-a...

tfgg• 12 years ago

Peierls and Landau predicted that 2D crystals can't be stable: http://physics.stackexchange.com/questions/110079/why-a-mono...

Either they made assumptions that don't hold, or graphene actually isn't "stable" as such but is stable enough in reality. I don't know much of the history of quantum chemistry simulations of graphene, but I'm fairly sure they existed beforehand and the experiments were a nice confirmation. Quite often it's possible to simulate things which are thermodynamically impractical, with things like an incomplete thermodynamic sampling in MD, or small periodic cell sizes artificially stabilising systems.

Osmium2• 12 years ago

I do molecular dynamics simulations (and am in fact performing simulations on graphene currently), and an individual layer of graphene easily stays intact in both a vacuum and other environments.

You can make a lot of 'stable' things in molecular dynamics simulations that don't exist in reality ;) You could construct any number of 2D materials that exist quite happily because they're close to some local minima or other, or because of some quirk of the potential forcefields you use. So I don't think "I can make it exist in a computer" is necessarily a good reason for expecting it to exist in reality, even if in the case of graphene it's relatively straight forward. Though useful, classical molecular dynamics is also quite dumb: it's really good for simulating known materials, but not so much for discovering new ones.

As for "2D" materials generally...well, we can make monolayers of a lot of things, sandwiched between other materials. And we can make other things that act like they're 2D even if they're not. There's nothing spectacular about it per se. The difficulty, as with all these materials, is that predicting that something has useful properties and actually being able to manufacture it at scale and relatively defect free is huge. You're fighting entropy, basically. So I think the surprise was not that graphene could exist in principle, but that it could be made reliably.

Xcelerate3• 12 years ago

You could construct any number of 2D materials that exist quite happily because they're close to some local minima or other, or because of some quirk of the potential forcefields you use ... though useful, classical molecular dynamics is also quite dumb

I believe you're thinking of the state of simulation in the 1980s. There's been quite a bit of progress since then ;)

For my work, I'm using a new reactive forcefield that's based off of DFT calculations and has a very close energetic match with experiment. Another forcefield I have used is optimized for structure and almost exactly reproduced neutron spectra for a variety of hierarchical materials, including long-range order (and that was from model systems alone).

Also, QM-MM methods have recently become more popular with additional computational power, so the parts that have significant non-classical effects can be computed correctly without wasting computational power on the parts that aren't going to make a difference.

As with any tool, molecular dynamics works very well if you know the theory behind it and know what you can and can't use it for. Structural carbon simulations is one of the areas where a lot of good forcefields exist (AIREBO for instance).

tim333• 12 years ago

Andre Geim talks about it 21 mins in to his Nobel Lecture -

http://www.nobelprize.org/mediaplayer/index.php?id=1418

enknamel• 12 years ago

I like the sentiment of "Friday Experiments". That's exactly in line with how we have hackathons and hackdays. They have also led to very interesting innovations.

ash• 12 years ago

The original Graphene paper:

"The reported graphene films were made by mechanical exfoliation (repeated peeling) of small mesas of highly-oriented pyrolytic graphite as described in the supporting online material. This approach was found to be highly reliable and allowed us to prepare FLG films up to 10 μm in size. Thicker films (d ≥ 3nm) were up to a hundred microns across and visible by the naked eye."

Electric Field Effect in Atomically Thin Carbon Films (Science, 2004)

http://arxiv.org/pdf/cond-mat/0410550.pdf

serve_yay• 12 years ago

Shoutout to the journal that rejected the research team's findings, once for being impossible, then once for being mundane.

scott_s3• 12 years ago

Forgive me for derailing a technical thread, but I want to comment on the author's description of the building I work in:

A vast arc of glass with an upswept front awning, it is a kind of monument to the difficulty of predicting the future. Saarinen imagined that transformative ideas would emerge from groups of scientists working in meeting areas, where recliners and coffee tables still sit beside soaring windows. Instead, the scientists spend much of the day hunched over computer screens in their offices: small, windowless dens, which seem to have been created as an afterthought.

I don't think he gives it enough credit. Saarinen actually had an egalitarian notion: no one gets a window office, because the shared hallway is one giant window. You can see what I mean in this shot: http://www.research.ibm.com/people/a/alpert/watson_night.jpg

The building is long, not high. The two main hallways that go the length of the building (the inner and outer parts of a semicircle; aerial shot here: http://upload.wikimedia.org/wikipedia/commons/thumb/d/de/IBM...) are floor-to-ceiling window. The spokes that actually have offices are short hallways between these two long, main hallways.

I like this building. I get to see the outside world as I walk around the building - not in glimpses, but in full. During the day, there is a lot of natural light coming from the large hallway windows. More than other office buildings I've been in, I have a constant sense of where I am, physically, with respect to the outside world. I like the notion that everyone - even the executives - get the same kind of office. (Some are a little larger, but not more than twice the size of a normal office.) And unlike cube-farms, we have private offices with doors we can close.

mey• 12 years ago

Do you have any additional notes or articles on this building? Reminds me of what Joel had been focusing on http://www.inc.com/magazine/20080601/how-hard-could-it-be-ad...

scott_s3• 12 years ago

I don't think it's nearly that ideal. I just think it's better than the "some offices get windows, some get none, and all hallways get none" alternative I've seen elsewhere in office parks.

jessriedel• 12 years ago

Got to disagree with you on this one Scott :) It's a simple matter of light-hours. Folks spend a lot more time in their office than walking between them. The light streaming in through those big beautiful windows is wasted on a mostly empty hallway!

scott_s3• 12 years ago

I'm comparing it to Hawthorne, where most people had interior offices and the building had completely interior hallways. I was one of those who had a completely interior office, and it sucked.

If you give some offices windows, then you must take those window spaces away from other public and private spaces.

tempestn2• 12 years ago

I enjoyed this little jab: "It was twice rejected by Nature, where one reader stated that isolating a stable, two-dimensional material was “impossible,” and another said that it was not “a sufficient scientific advance.”"

It's completely impossible, but if you did manage to do it, meh.

pja• 12 years ago

That's peer review in a nutshell for you :)

tempestn2• 12 years ago

The title doesn't seem remotely accurate given the content of the article. Graphene is already being used in several applications, with many thousands more being researched, a number of which were highlighted in the article. As it also discusses, it takes quite a long time for any new material to really become fully utilized; that certainly doesn't equate to "Impossible to Use".

trhway• 12 years ago

Tells a lot about current human civilization:

"The technology for the Fukushima-reactor cleanup stalled when scientists in Japan couldn’t get the powder to work, and the postdoc who developed the method was unable to get a visa to go assist them. "

edwinespinosa09• 12 years ago

What I took most from this is that the continual battle of whether graphene will indeed be important as a semiconductor rages on. I remember back in the day when I first heard about it I wanted to hop in and invest like Texas instrument with silicon. I guess at this point it's a guessing game.

mrfusion• 12 years ago

Are there any other materials in the past that have taken this long to get from the lab to mass production?

quote2• 12 years ago

Have you read the article?

"Aluminum, discovered in minute quantities in a lab in the eighteen-twenties, was hailed as a wonder substance, with qualities never before seen in a metal: it was lightweight, shiny, resistant to rust, and highly conductive. It could be derived from clay (at first, it was called “silver from clay”), and the idea that a valuable substance was produced from a common one lent it a quality of alchemy. In the eighteen-fifties, a French chemist devised a method for making a few grams at a time, and aluminum was quickly adopted for use in expensive jewelry. Three decades later, a new process, using electricity, allowed industrial production, and the price plummeted."

Graphene: Fast, Strong, Cheap, and Impossible to Use

Related Stories