Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Graphene creates electricity when struck by light (extremetech.com)
91 points by kennjason on Oct 7, 2011 | hide | past | favorite | 16 comments


Original MIT description: http://www.mit.edu/newsoffice/2011/graphene-thermoelectric-1...

Current had previously been seen in response to light, but was assumed to be a photovoltaic effect. This new work says it is a "hot-carrier" response.

"The material’s electrons, which carry current, are heated by the light, but the lattice of carbon nuclei that forms graphene’s backbone remains cool. It’s this difference in temperature within the material that produces the flow of electricity."

It is interesting because it is a response to temperature differences and responds to wide frequencies and power ranges of light.

Don't sell your coal mine stock yet, "It is still unclear if it could be used for efficient energy generation. It’s too early to tell."


>Don't sell your coal mine stock yet, "It is still unclear if it could be used for efficient energy generation. It’s too early to tell."

isn't graphene made from coal? maybe this discovery might even further increase coal demand?


Coal is made of carbon molecules, and so is graphene. But then again so is diamond, and graphite (pencil lead). So it's not easy to convert coal->graphene (though it's probably not as hard as graphite->diamond :)

https://secure.wikimedia.org/wikipedia/en/wiki/Allotropes_of...


Coal is largely dirty (impure) carbon. Its easier to pull carbon out of thin air than refine it from coal.


Can anyone with an actual background in thermoelectrical effects tell us why this is not just a consequence of graphene's Seebeck coefficient?

As near as I can tell:

1. They used 850nm to heat one side of the graphene pn junction

2. Photons have an absorption depth, so the side facing the laser heats more than the side away from the laser.

3. Seebeck effect manifests.


They didn't heat one side, they heated all of it. The electrons became mobile and moved from the n-side to the p-side. What's interesting is that the electrons got exited but the nuclei did not. So the nuclei stay in their pattern while the electrons flow.

Edit: all I know is what I read in the press release :) Here's the actual paper http://arxiv.org/abs/1108.3826


> What's interesting is that the electrons got exited but the nuclei did not.

This statement can't be right. If the temperature of the material went up, then the nuclei became exited. And if they didn't, then the result is identical to the photoelectric effect.

Edit: By the way, what you've described corresponds to "freeze-out" in old-fashioned semiconductor terms. As you drop the temperature of a diode, the free electrons and holes fall back to their low energy state and it turns back into a vanilla semiconductor. Heating it up makes them mobile again.

Edit2: > Edit: all I know is what I read in the press release :)

Yeah, it was clear that the PR was not written by someone knowledgeable. I'm sure there are MIT folks on this board so I'm hoping one of them pokes the right colleague so that we get a clearer answer, because it sure sounds like there's something interesting going on there.

Thanks for the arxiv pointer, I'll try to look this weekend.


sp332: If you are still reading, thanks again for the link.

I haven't delved deeply, but I believe the paper is aimed at exploring whether graphene's photocurrent mechanism is in fact thermoelectric in nature. If you look at Eq 1 in the arxiv paper, you will see that they start by calculating the Seebeck coefficient from first principles. (My original comment above turns out to be exactly the issue.)

The reason that this is interesting: The authors are saying that there is an interesting Seebeck effect as a consequence of graphene's material properties, especially its odd density of states.

To the point you raised: The electron population does absorb photons. But rather than transporting across the pn junction due to the increased energy of the absorbed photon, it collides with other electrons and then these electrons form a population with an elevated temperature among themselves. The electrons do not couple with the lattice, so they don't transfer energy to the lattice all that quickly. Hot electrons from the high-temperature ensemble then have a probability of making it across the pn junction.

Like I said above, I haven't fully digested the paper (and I doubt I will, it's not in my critical path), but that was my quick take. Any other readers, please correct me at will.


Am I the only one who wishes we could give these folks a few billion dollars so they could stop issuing press releases and just make something already?


The graphene researchers from yesterday got $2 mil at least! http://news.ycombinator.com/item?id=3080058

But yeah, money going into graphene research will certainly be on the rise over the next few years :)


This effect was seen before, but had been dismissed as a photoelectric effect. http://www.mit.edu/newsoffice/2011/graphene-thermoelectric-1... And here's the actual paper: http://arxiv.org/abs/1108.3826


"Since its discovery in 2004 (by pulling a thin layer off a piece of pencil graphite using sticky tape)"

So can I just do that and get a layer of graphene ? I mean is it really that simple ?


You'll get a bunch of graphite and some microscopic bits of graphene. But yeah, it's the easiest way to get some for experimenting. https://secure.wikimedia.org/wikipedia/en/wiki/Graphene#Draw...


So could a space elevator in theory be self-powering from the cable?

I mean there is going to be 36,000km of it.


Maybe. But they'll have to deal with the fact that a space elevator would effectively "ground" the ionosphere first.


I guess you don't actually need to ground it. The Earth end of it could hover above the ground or it could be attached to some insulator.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: