Organic Transistors....

F

Flux

Regular
what if manufacturers could build a CPU with organic melanin transistors. Doped Melanin is better than inorganic because while at certain voltages it is a poor conductor but at higher voltages it switches(and gives off light when it does) to its "on" state and becomes almost as conductive as silver. The only problem is it is unstable in the air(thus needs to be sealed/coated). With the source of melanin being cheap as dust (cow eye,most mammalian hair,bananas,algae,etc) if one can refine the process of obtaining high purity melanin and dope it with impurities like boron and bromine. The cost of the transistor would drop by 100 fold.

Being an organic compound we can probably biuld a transistor with melanin and other compounds.we could probably engineer microbes to biuld transistors or IC from the nanoscale up.
 
So um... real infections instead of digital ones? :) But surely, there are issues with "growing" hundreds of millions of them as well as switching speed?
 
3dilettante said:
Superconductors aren't all that useful as transistors. Having no resistance, they don't turn off.
Click to expand...

You can have other ways to stop/regulate/reroute voltage other than resistance.

Image this. When you have the switch "on" there is no thermal limit. Having no resistance you can clock it thousands of times higher and still have no heat.


Now making it organic means the transistors can be grown.

interesting stuff.
 
Flux said:
You can have other ways to stop/regulate/reroute voltage other than resistance.
Click to expand...

Can you name one?

Image this. When you have the switch "on" there is no thermal limit. Having no resistance you can clock it thousands of times higher and still have no heat.
Click to expand...
The reason why semiconductors are useful is that they can do both "on" and "off".

Superconductors don't do "off" because superconduction means resistance is 0 while "off" means resistance>>0.

If a transistor channel that is supposed to be "off" is cooled to the point of being superconductive, it can't be "off" because current is flowing through it.
 
3dilettante said:
Can you name one?

I am not the engineer. If a R&D/engineering team was payed to find a way to control voltage by a means other than resistance I am sure they could. I am not qualified to answer that question.

The reason why semiconductors are useful is that they can do both "on" and "off".

Semiconductors are useful because they can under certian condition become more conductive(voltage=Si,GeAs light=Se,Te Magnetic=CdS) yet can control voltage by the resistance of its off state.

A transistor's on state is when it has voltage going through it and on to the function(switch on).If the voltage is rerouted(via a magnetic field,resonant inductance,quantum tunneling etc) and does not go to its function it is now 0(switch off). Same result as the resistance based transistors.

But this is moot because melanin is not a superconductor. But in its on state it becomes almost as conductive as silver.:oops:

Superconductors don't do "off" because superconduction means resistance is 0 while "off" means resistance>>0.

There are ways to control voltage other than resistance. Like inductance,quantum tunneling, spin(using an magnetic field).

If a transistor channel that is supposed to be "off" is cooled to the point of being superconductive, it can't be "off" because current is flowing through it.
Click to expand...


I am sure if microchip manufacturer hired a team of skilled R&D solid state engineers,time and a apropriate budget they would figure it out.

Melanin is not a superconductor, but its better than silicon in almost every way.

Obviously if a material has no resistance you can't use a resistance based transistor design for that material.:???:
 
A transistor's on state is when it has voltage going through it and on to the function(switch on).If the voltage is rerouted(via a magnetic field,resonant inductance,quantum tunneling etc) and does not go to its function it is now 0(switch off). Same result as the resistance based transistors.
Click to expand...
Can you give a link to any practical implementation of these other methods? I'm not sure any of them are restorative, and some of them are highly unpredictable.

What is the mechanism of what you are touting? How exactly does the completely uncontrollable tendency of a particle to pass through a barrier at random make a useful transistor?

But this is moot because melanin is not a superconductor. But in its on state it becomes almost as conductive as silver.
Click to expand...

What does it take exactly? How much voltage, how quickly, and at what cost in power?

I am sure if microchip manufacturer hired a team of skilled R&D solid state engineers,time and a apropriate budget they would figure it out.
Click to expand...
Hard to say that without some real comparisons. Apparently one is easy to mass-produce, and can be made to switch very quickly.

The plastic and organic transistors I've seen thus far do not do that well when it comes to speed.

Melanin is not a superconductor, but its better than silicon in almost every way.
Click to expand...
I'd like for you to give me some real analysis on this. I can say spaghetti is superior to silicon in almost every way if I don't need to back it up with a thorough comparison.
 
Flux said:
You can have other ways to stop/regulate/reroute voltage other than resistance.

Image this. When you have the switch "on" there is no thermal limit. Having no resistance you can clock it thousands of times higher and still have no heat.
Click to expand...

The voltage is not "routed" anywhere, as _xxx_ pointed. And the "off" state of a transistor works based on voltage barriers, not resistance. The resistance of a material gives you an idea of how long it takes for the e- to "crash" against a nucleus, loosing is energy into heat, which is (being over simplistic) why you can express power disipation as I*R^2.

In a transistor, the e- can't pass to the other doped material because they can't reach the higher level of energy. When you polarize it in DIRECT-INV, you lower the barrier of the first junction to a level where e- can pass by, and reduce the other barrier to a very low state, allowing the e- to "flow" through the material:

_____/------\
...................\
,,,,,,,,,,,,,,,,,,,,\___________

This is more or less what you would see on a transistor, where axis X represent in distance, and Y axis represent energy. The heat that is generated is due to e- that get caught by the h+ left on the p-dopped material. That is why the way to improve a transistor heat condition is to make the p region as thin as possible.

This is what I've known and been teached, but some or all of it might be wrong. Yet, I don't get a clear picture of what you're proposing could be done with organic transistors.

edit: a better picture of what I was talking about: Polarized n-p-n transistor
 
MatiasZ said:
The voltage is not "routed" anywhere, as _xxx_ pointed. And the "off" state of a transistor works based on voltage barriers, not resistance. The resistance of a material gives you an idea of how long it takes for the e- to "crash" against a nucleus, loosing is energy into heat, which is (being over simplistic) why you can express power disipation as I*R^2.

In a transistor, the e- can't pass to the other doped material because they can't reach the higher level of energy. When you polarize it in DIRECT-INV, you lower the barrier of the first junction to a level where e- can pass by, and reduce the other barrier to a very low state, allowing the e- to "flow" through the material:

_____/------\
...................\
,,,,,,,,,,,,,,,,,,,,\___________

This is more or less what you would see on a transistor, where axis X represent in distance, and Y axis represent energy. The heat that is generated is due to e- that get caught by the h+ left on the p-dopped material. That is why the way to improve a transistor heat condition is to make the p region as thin as possible.

This is what I've known and been teached, but some or all of it might be wrong. Yet, I don't get a clear picture of what you're proposing could be done with organic transistors.

edit: a better picture of what I was talking about: Polarized n-p-n transistor
Click to expand...

http://www.physorg.com/news7488.html

Oh look someone has created a spin based transistor using carbon nanotubes. Not so impossible isn't it. Again it can be done and has been done. But again no one wants to risk a untried technology yet.

There are several companies developing plastic chips. Its not exactly fringe.

http://www.sciencenews.org/articles/20030830/fob5.asp

There are N-type organic transistors and P-type

"His team's new class of molecules assembles into semiconductors of both p- and n-type. A rod-shaped organic molecule made of six thiophene units forms the basis for each type of material. Each thiophene, in turn, is a ring of five carbons and one sulfur. When the researchers replaced the rod's two end thiophenes with a perfluoroarene group (a ring of six carbons decorated with fluorines), the organic molecule behaved like an n-type semiconductor. When the researchers instead replaced the next two thiophenes from the ends, the molecule behaved as a p-type semiconductor."

http://www.techweb.com/wire/story/TWB20031215S0016

Organic transistors can be grown in fibers. Now could this be cheaper than the alternative inorganic transistors' fabrication?...

"we have demonstrated flexible organic TFTs formed on fibers...This technique is compatible with textile-manufacturing; this work therefore represents an important step towards the realization of a viable e-textile technology."

spinotronics is already being prepared for the Ipod.

http://www.textually.org/textually/archives/2006/05/012508.htm


Hey there is even a quantum tunneling transistor that can switch at 1THz with extremely low power.

"The very fast device may run at a trillion operations a second, as have other, more primitive tunneling devices. This is roughly ten times the speed of the fastest transistor circuits currently in use. Actual speed has not yet been measured, says Simmons, because it is "not easy to measure such high speeds, which are near the limits of what can be measured with conventional equipment." The extremely fast device also runs at extremely low power -- tens of millivolts and microamps -- as compared with the few volts and milliamps needed by transistors currently in use."

http://www.sandia.gov/media/quantran.htm

Here is another diagram of a tunneling transistor.

http://physicsnow.org/png/html/tunnel.htm

single electron transistors are organic carbon nanotube. Inorganic semiconductors are showing their age.

http://pubs.acs.org/cen/topstory/7928/7928notw3.html

"The demonstration "is a very significant [advance] in developing the technology for assembling carbon nanotube-based devices," said Deepak Srivastava, a senior scientist and technical lead in computational nanotechnology at the NASA Ames Research Center. "People have always talked about using wet chemistry for assembling molecular electronic components into precise locations," he said. "This is a first proof of the principal."

This can be the first step towards organism/machine hybrids(aka cyborgs)
 
The advantage silicon has over the examples of plastic and organic transistors is that it's many times faster, and currently much more densely packed.

For the tunneling and spin transistors, there's the example where there is one transistor at supercold temperatures, with a lab set up to handhold just one transistor, which leads to a single transistor switching 100 times faster (maybe not, there are fast semiconductor ones in similar setups that do switch at high speeds) with no requirement of reliability or use in a functioning device.

The silicon transistor and billions of fellow transistors in a standard mass-produced chip will just have to get by without the lab and the gallons of coolant.

Transistor switching speed is unlikely to be the primary limiting factor in future chips. There are silicon and other semiconductor transistors that switch in the hundreds of GHz range. Signal propogation, something limited by the the speed of light, is more likely to matter.

I haven't seen any conclusive proof any of these alternatives has the same mix of reliability, speed, and mass manufacturability.

Perhaps in twenty years they might, after silicon scaling stops.
 
You must log in or register to reply here.
Back
Top