Hacker News new | past | comments | ask | show | jobs | submit login
Cambridge spin-out starts producing graphene at commercial scale (cam.ac.uk)
118 points by airstrike 36 days ago | hide | past | web | favorite | 42 comments

Can someone ELI5 why graphene, defined as "a sheet of carbon just one atomic layer thick", can be sold as a "chip"? Of what use would be a one-atom-thick "chip"?

Also, "graphene-based chips could deliver speeds more than ten times faster than silicon chips" -- wut? The basic reason (I thought) silicon could be used to make integrated circuits is that it is mostly NON-conductive, but can be made conductive in part, so as to form e.g. capacitances, insulated conductors, diodes, etc. But in my ignorance I thought the one thing that graphene has to be, is conductive, completely conductive in all directions. So how can you make the various parts of a circuit in a material that is one big short circuit?

I think they means the graphine as the conducting wires. Someone probably told them wires/bands 1-atom thick connecting parts = smaller chips = faster chips. Id also be interested in what happens in terms of heat. Such small structures conducting electricity Would graphine parts on a chip get hot enough to combust?

Graphene is a better heat conductor as well, so heat ought not to be as big a concern.

Thank you, but the text of that article is, um, science-challenged, for example,

> At room temperature, graphene is also capable of conducting electricity 250 times better than silicon, a rate faster than any other known substance.

Conducting "better", by which I think they are referring to resistance or conductance, has no relation to "speed" of current flow. So that sentence is noise. And I don't think it's possible for current to move 250 times faster even in a superconductor. Current in silicon is already a good fraction of the speed of light.[0]

Their description of a graphene transistor is not much better,

> They then applied a magnetic field to the graphene ribbon, which made them realize they could control the resistance of the flowing current through the ribbon.

This confusing sentence seems to be trying to describe a magnetically-controlled switch, with an applied field throttling the current flow in a graphene conductor. Hall effect?

OK, now you made me try to find something better. The Wikipedia article on Possible Uses of Graphene[1] does talk about transistors[2] but primarily as FETs[3] and says nothing about magnetic control. A magnetic field does produce a Hall effect[4] but this is not the same as a transistor. As for switching speed, [2] says "In 2013 researchers created transistors printed on flexible plastic that operate at 25 gigahertz" which is fast but not faster than silicon.

0. https://en.wikipedia.org/wiki/Speed_of_electricity

1. https://en.wikipedia.org/wiki/Potential_applications_of_grap...

2. https://en.wikipedia.org/wiki/Potential_applications_of_grap...

3. https://en.wikipedia.org/wiki/Field-effect_transistor

4. https://en.wikipedia.org/wiki/Potential_applications_of_grap...

It's obviously a popular text, but the second paragraph includes a link to the actual paper; you could have saved yourself some searching by simply clicking that:


As for the "rate faster" sentence, you are misinterpreting it. It's referring to the amount of charge transferred per time unit, not speed of propagation. See


Reduce R and I goes up.

"Speed" in integrated circuits is mostly determined by RC delay - the need to charge the gate capacitance of the destination transistor, through the resistance of the interconnnect (metal, usually aluminium) and the gate on-resistance of the driving transistor. So lower interconnect resistance would make a real speed difference.

Does the nature article that futurism piece was talking about make it clearer?


Less resistance less heat not compact transistors

Is their graphene going to be a real graphene? If I read Lowe[0] correctly, the primary problem in this space is that you can't really buy actual graphene from any of the vendors.


[0] - http://blogs.sciencemag.org/pipeline/archives/2018/10/11/gra...

Has there been any research on the health risks? I'm sure graphene dust isn't great to be breathing in and if we're thinking of using it to improve a good amount of products and materials I feel this should be studied before that cat is out of the back. I feel like this will be asbestos all over again except much, much worse.

The saying is letting the cat out of the bag.

And yet lots of people use "cat out of the back". Was suggested in the eggcorn database forum in 2010 - http://eggcorns.lascribe.net/forum/viewtopic.php?id=4367

bag + fat fingers + swipe/autocorrect = back

I am quite wary of this kind of thing, but fine graphine dust should be very similar to fine graphite dust of which plenty of experience already exists. Graphite has the same structure as graphine, only less orderly.

Nope, that research olnly happens after novel cancer deaths and their resulting lawsuits. That said, like other substances, it is pretty safe once laminated, as you would likely need to do based on the fragility.

So they're producing the basic materials but have concluded there is no long term business related to supplying graphene wafers. That means there are other companies already in that race or soon to come out with capital. Exciting times indeed.

> graphene could increase (sensor) sensitivity by a factor of more than 30

A *factor of 30 would be unreal. Anything more than 2 in the real world would be amazing.

They mean 30 times as sensitive ie. sensing 1/30th as much stuff. That's what "a factor of n" means btw. Maybe you had in mind "a power of".

If previous sensors required 300 photons and new ones require 10, that would be an absolutely incredible achievement. (Scale that to whatever current state of the art is for camera sensors, of course)

That's what I think of when I hear "30x improvement", at least.

Yeh that's the plain meaning. They will be referring to a particular sensor type. Amazement is no cause to say the claim is "unreal"

Sensing 30 times as much stuff is not the same as sensing 1/30 as much stuff.

The article says they worked around the copper impurities, but I wonder if they overcame the silicon impurity issues that cause low performance and "disappointing results" in most applications.[1] Summary[2][3]

[1] https://www.nature.com/articles/s41467-018-07396-3

[2] https://phys.org/news/2018-11-full-potential-supermaterial-g...

[3] https://www.createdigital.org.au/graphene-performance-reveal...

As a material scientist, I'm reminded of a joke from college which has changed only a little: What's the only thing graphene (used to be carbon nanotubes) can't do? A: Leave the lab

Without this basic availability, no companies that would build on top of it could even start.

Hopefully this news triggers a wave of new investment built on hype with a few companies that actually succeed, enabling another platform for even more new innovations.

Has the band gap issue been worked out?

I just read through their website, they claim it has, using the heading 'configurable':

``` The capability to control graphene layer properties defines its usefulness for the desired application. Whether this be physical (mechanical strength, porosity, thickness of stacked layers), electrical (conductivity, carrier density, band gap availability) or otherwise (thermal conductivity, transparency, flexibility). ```

As a former Slashdotter, I've seen this joke on every single article about "nanotubes".

As a former nanotube research grad student, I have to wholeheartedly agreee though.

I have yet to read a graphene comments thread here or even on Reddit without someone mentioning it.

I find the little differences between U.S- and U.K.- English interesting.

In U.S. English, "spin-out" usually refers to an automobile spinning while out of control, for example when it hits a patch of ice and gets uneven traction.

What’s the cost vs chip grade silicone waffers? Even if they can, they would then need commercial scale process to print the electronics, tip of iceberg right there

No "buy now" button... ie. still vapourware

What? Why would a company selling industrial raw materials have a "buy now" button on their website?

Why? Do you also have the need to buy silicon online?

Well, if he's from a research lab that uses graphene I imagine that option would be useful. As of right now, there are plenty of places to buy semiconductor wafers of which silicon is the most basic and cheapest (and even some forms of graphene like aqueous).

Agreed, this is the kind of thing you work out a contract for.

Right. They are not producing graphene sheets in commercial volumes. They made some 20cm samples, big enough that, if produced commercially, they might be useful.

Yet, that web site is not eager to sell those. It may be a design issue, but the only place one may find anything (no guarantees) is on the contact us link.

If they have samples, I'd expect them to want to distribute those (hell, every time I see a graphene related company I look on how I could buy a sample), but I'm not even sure about what I'll get after contacting them. Is the contact there for investors only? For hiring?

If they want people to start using their graphene for something, they may need to add a "get a sample" button somewhere. They may not be vaporware (there are enough labs producing some kind of graphene out there, there are even sites that are more clear on what they produce and yet aren't clear enough on how to get some), but there isn't a line of people wanting to jump over any obstacle to get it either.

For what would you yourself buy a graphene sample? How much are you willing to pay for it?

I have this project where I will soon need a very thin organic electricity conductor. I will need a few mm^2 of it, and it doesn't have to be graphene, so I am not willing to pay much, and I am probably out of the target market of any such startup. Yet, I'm always tempted to search for a price.

Glencore claims to be a commercial scale vendor but it doesn't accept PayPal.... ie. still vapourware

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