An organic transistor that mimics a brain synapse

By | January 29, 2010, 12:10am PST

For the first time, nanotechnology researchers in France have developed a hybrid nano-particle-organic transistor that can mimic the main functionalities of a synapse.

Credit: Mil-Tech.com

The NOMFET (Nanoparticle Organic Memory Field-Effect Transistor), as it’s known, is an organic device made of a molecule called pentacene (an organic semiconductor) and gold nano-particles. It exhibits the main behavior of a biological spiking synapse and can lead to a new generation of neuro-inspired computers, capable of responding in a manner similar to the nervous system.

As Mil-Tech reports, Dominique Vuillaume, a research director at CNRS (the French National Science Agency) involved with the project said; “Basically, we have demonstrated that electric charges flowing through a mixture of an organic semiconductor and metallic nano-particles can behave the same way as neurotransmitters through a synaptic connection in the brain.”

To grasp how a NOMFET works requires a quick review of how neuron networks operate. In the nervous system, a synapse is the junction between two neurons, enabling the transmission of electric messages from one neuron to another and the adaptation of the message as a function of the nature of the incoming signal (plasticity). For instance, if the synapse receives very closely packed pulses of incoming signals, it will transmit a more intense action potential. Conversely, if the pulses are spaced farther apart, the action potential will be weaker. It is this plasticity that the researchers have succeeding in mimicking with the NOMFET. A transistor can be used as a simple switch — it can then transmit, or not, a signal — or instead offer numerous functionalities (amplification, modulation, encoding, etc.).

Science Daily explains that the innovation of the NOMFET resides in the original combination of an organic transistor and gold nanoparticles. “These encapsulated nanoparticles, fixed in the channel of the transistor and coated with pentacene, have a memory effect that allows them to mimic the way a synapse works during the transmission of action potentials between two neurons.”

“The output of the NOMFET is thus able to reproduce the deceasing or amplifying behavior typical of a synapse depending on the frequency of spikes,”  Vuillaume recently told Physics World.

The transistor’s performance is comparable to the seven CMOS transistors (at least) that have been needed until now to build an electronic synapse and mimic this plasticity. The devices produced have been optimized to nanometric sizes in order to be able to integrate them on a large scale.

Neuro-inspired computers produced using this technology eclipse silicon computers to perform more complex functions comparable to those of the human brain, such as visual recognition.

The study is published in the 22 January 2010 issue of the journal Advanced Functional Materials, and can be accessed on Scribd.

Kick off your day with ZDNet's daily e-mail newsletter. It's the freshest tech news and opinion, served hot. Get it.

Topics

Christopher Jablonski is a freelance technology writer.

Disclosure

Chris Jablonski

Christopher Jablonski has no business relationships, affiliations, investments, or other actual/potential conflicts of interest relating to the content posted so far on this blog.

Biography

Chris Jablonski

Christopher Jablonski is a freelance technology writer. Previously, he held research analyst positions in the IT industry and was the manager of marketing editorial at CBS Interactive. He's been contributing to ZDNet since 2003.

Christopher received a bachelor's degree in business administration from the University of Illinois at Urbana/Champaign. With over 12 years in IT, he's an expert on transformational technologies, particularly those influential in B2B.

Related Discussions on TechRepublic

Did you know you can take part in these discussions with your ZDNet membership?
24
Comments

Join the conversation!

Just In

RE: An organic transistor that mimics a brain synapse
Steampower 22nd Aug
Really, xanax , levitra Brazil and levitra be bright headed for complete you lots of consummate a brief count back just as Viagra as well Vardenafil.
0 Votes
+ -
The Devil's work!
kd5auq Updated - 29th Jan 2010
The world won't last another 6,000 years!
sad

Don't get your neurons in a wad!
It's great research, just doing a poor imitations of a Nancy Palin reaction.
wink
0 Votes
+ -
Not totally convinced.
CobraA1 29th Jan 2010
Replicating a neuron is one thing - replicating a brain and intelligence are quite another. We don't understand the brain well enough to really produce something with a higher level intelligence.

We've tried neural networks in software, and while they exhibit some basic learning behavior, I've yet to see one that actually has reasoning skills.
0 Votes
+ -
Not convinced of what?
RDrr 29th Jan 2010
The article is about a building block, a replicated neuron... and two of them together replicating a synapse. No neural net, yet.
0 Votes
+ -
This is version 1.0 of the basic building block that can be used to build sentient AI. Who decides what stimuli will provide the rapid tranmission of signals?
0 Votes
+ -
version 1.0 ?
Me_too 31st Jan 2010
I feel it may be version 0.1.A.a.

But it is a nice start.
This is no big deal. Making a biological transistor without the bias of biological survival is the same as an Intel processor.
0 Votes
+ -
What if we look at the smaller picture and instead of seeing this as the leap to artificial intelegence, see this as a way to repair malfunctioning synapse in patients with mental illness.
0 Votes
+ -
I like that suggestion
Kualinar 29th Jan 2010
To use nanotechnology at the subcellular level to help cure neural afflictions and wounds.

It can have applications way broader that mental illness, like mitigating the after effects of a cervical emoragy or a severed spinal chord.
0 Votes
+ -
Contributr
Extreme extrapolation
christopher_jablonski 29th Jan 2010
And, as a physicsworld.com observer noted:

"One can envision a time when a cortical transplant from a dying person might be attached to a prosthetic brain stem of this type which in turn would fit into a cyborg, extending an individual's existence indefinitely."
0 Votes
+ -
This will spark a new a new age in artificial intelligence and robotic cloning leading to new age warfare, space exploration, medical science. The imagination can only ponder what this device will produce in 50 - 100 years from now. The Age of Technology has taken a new turn which will open the next chapter in human evolution. Control robots to fight wars, fly space vehicles, repair broken or sick bodies? Am I going mad muuhaaahahaa...
0 Votes
+ -
Zut alors, a *French* organic transistor?
Heimdall222 29th Jan 2010
Since it's a French organic transistor, by definition it'll automatically sniff its neighbors' backsides and be able to lick its own...ummm, well, you know.
0 Votes
+ -
racist
pcmagik 31st Jan 2010
and offensive
0 Votes
+ -
First the organic transistor, next the organic integrated
circuit!
0 Votes
+ -
Rebooting is all in your head
bobmatch@... 29th Jan 2010
Like they say, the next step in evolution will be man made! Just think, computers in our brains, of course running windows. It?s enough to make one stop and think (or reboot!).
0 Votes
+ -
Or cause a RSOD. Someone did say running Windows.
0 Votes
+ -
RE: An organic transistor that mimics a brain synapse
bbhutchinson@... Updated - 30th Jan 2010
Which means the Government can interface with you and probably track you. What will the I.P. Address be, your social security number? How will the interface work and do you have to take medication to prevent your body from Rejecting this foreign device? Hey, we may now be able to place orders via thought for online/in store purchases...... Many questions to answer as well as many more to ask!
0 Votes
+ -
organic transistor
Peteyboy1 31st Jan 2010
Most of you are missing the point. It's real application is in the medical field. Repairing motor neuron disorders, spinal cord injuries etc.
0 Votes
+ -
spinal cord injuries
Me_too 31st Jan 2010
I can only hope so, but how long after the fact can it be done. I've been waiting for over 13 years but I've been thinking of the other route.
Genetics plus nanobots.
0 Votes
+ -
This is NOT a replicated neuron
DeusExMachina 31st Jan 2010
You state that this device operates in the same manner as biological
neurons. You then go into an explanation that shows both:
that you do not understand how real neurons work, and
that these devices do NOT operate like real neurons.

Neurons work by transducing a low voltage action potential at the
axonal head into a brief, modulated chemical burst. This chemical
secretion crosses the synapse and triggers chemical ligand modulated
ion channels that then open and allow charged particles to enter the
dentritic head. This charges the head, and if that charge is large
enough, causes voltage activated channels to open, further increasing
charge differential, and causing another action potential.

In NO way does this device operate in this fashion. As such, these
comments above claiming this can be used for medical applications,
or those saying this can open the doors for a new brand of AI are
GROSSLY off the mark. While this is an interesting device, it does NOT
operate like neurons, and most certainly could NEVER be used to
repair damaged brain tissue.

FTR, I have a degree in neuroscience and computation neuro.
0 Votes
+ -
Contributr
Where does I state that it does?
christopher_jablonski 31st Jan 2010
Thanks for the comment. However, "To grasp how a NOMFET works requires a quick review of how neuron networks operate." -- does not equal--"..this device operates in the same manner as biological neurons."

NOMFET is an artificial synapse. Neurons use synapses to pass chemical signals to each other, so it's worth pointing out the function of neurons in a simple fashion as this site is for IT pros, and not necessarily neuroscientists like you.

Best,
Chris Jablonski
0 Votes
+ -
Throughout the article
DeusExMachina Updated - 1st Feb 2010
" It exhibits the main behavior of a biological spiking synapse"

No, it doesn't

"capable of responding in a manner similar to the nervous system."

The stated behaviour of NOMFETs is NOT a similar
manner to the response of neurons.

"Basically, we have demonstrated that electric charges flowing
through a mixture of an organic semiconductor and metallic nano-
particles can behave the same way as neurotransmitters through a
synaptic connection in the brain."

I have no doubt that he said that, however, again, they are NOT
behaving in the same way. This just is NOT how neurotransmitters
work. Besides which, there is a LOT more to neural behaviour
than what you incorrectly term "plasticity," and besides which, this
"plasticity" can be modeled without NOMFETs, using standard off the
shelf components.

"To grasp how a NOMFET works requires a quick review of how neuron
networks operate."

This implies a correspondence in function, otherwise there is no need
to understand neuronal behaviour. To wit:

You then go on to discuss varying responses to pulse trains, and refer
to this as "plasticity."
This is NOT what synaptic plasticity means, as shown by your own link
to the wiki article. Plasticity means varying signal strength over time
based upon previous input. It does NOT mean responding to different
pulse encodings in a variable manner. For instance, and example of
the former would be a neuron that fires continuously causing the
neuron it synapses on to decrease its firing in response, over the
course of several hours. An example of the latter would be a neuron
that fires a doublet pulse in response to a fast pulse train, and a
singlet pulse in response to a slow pulse train. The former is
plasticity, the latter is not. The latter is what the NOMFETs do.

"The transistor's performance is comparable to the seven CMOS
transistors (at least) that have been needed until now to build an
electronic synapse and mimic this plasticity."

As I stated above, the only real innovation here, from a practical
standpoint, is the single unit replacing a CMOS circuit. Neat, but not
revolutionary.

This can be summed up in my response to your final statement:

"Neuro-inspired computers produced using this technology eclipse
silicon computers to perform more complex functions comparable to
those of the human brain, such as visual recognition."

First, NO computer has been made with this component, so I have NO
idea where you are pulling that from. They don't do anything of the
kind. Nor would they, even if the DID exist, because, again, silicon
equivalents already exist. One could build the EXACT same
functionality with CMOS transistors, or, more importantly, just model
the behaviour in software on a standard desktop PC.

"NOMFET is an artificial synapse."

For the record, the word synapse means the gap between neurons.
The NOMFET is not really an artificial synapse. Rather it is more an
artificial neuronal junction, modeling both sides of the synaptic cleft.

"Neurons use synapses to pass chemical signals to each other, so it's
worth pointing out the function of neurons in a simple fashion as this
site is for IT pros, and not necessarily neuroscientists like you."

Yes, and these NOMFETs do not, which is why they can not be called
artificial synapses, especially since you directly imply that they mimic
function.
In reading the science daily article you cite in your blog, it is apparent
where the discrepancy comes from, as the author there makes the
same mistakes. Indeed, this is not surprising, seeing as several
sections of that article were simply lifted wholesale, for use in yours.
This is a dangerous practice, especially when the original author has
gotten it wrong.
0 Votes
+ -
Contributr
Good points
christopher_jablonski 1st Feb 2010
DeusExMachina,

Thanks again, I'm going to reread the research paper. But I now see where I'm off and there are lessons learned. Next time I tackle this complex subject matter maybe I could pass it by you or a similar qualified expert for review.

Chris
0 Votes
+ -
Sure
DeusExMachina 2nd Feb 2010
No problem. I just have a bit of a sore spot for all those hard AI articles
that talk about computers soon approaching the capabilities of the
human brain, based on seemingly random parameters, such as total
number of transistors, or calculations per second or whatever. Nothing
could be farther from the truth. We are only marginally closer to
understanding brain function, especially higher order cognitive functions,
and the idea that we are within striking distance to real AI is a laughable
proposition only put forth by computer scientists, never neuroscientists.
So never mind me. I just get a little acerbic on the subject.

But feel free to seek me out if you have questions.
Really, xanax , levitra Brazil and levitra be bright headed for complete you lots of consummate a brief count back just as Viagra as well Vardenafil.

Join the conversation!

Formatting +
BB Codes - Note: HTML is not supported in forums
  • [b] Bold [/b]
  • [i] Italic [/i]
  • [u] Underline [/u]
  • [s] Strikethrough [/s]
  • [q] "Quote" [/q]
  • [ol][*] 1. Ordered List [/ol]
  • [ul][*] · Unordered List [/ul]
  • [pre] Preformat [/pre]
  • [quote] "Blockquote" [/quote]
ie8 fix

The best of ZDNet, delivered

ZDNet Newsletters

Get the best of ZDNet delivered straight to your inbox

Facebook Activity

White Papers, Webcasts, & Resources
ie8 fix