Boffins build the smallest transistor, controlled by an atom • The Register
"...two small metallic strips containing silver, ...By an electric control pulse, we position a single silver atom into this gap and close the circuit.... The single-atom transistor can drive a tiny current of somewhere between one to eight microAmps."
Not a keen audio device, no. But promising for computing?
What bothers me: 8uA for one electron of charge seems like fabulous transconductance, unless that electron has to be pushed a long way.
The demonstration device is a LOT bigger than "one electron", of course. (Looks an awful lot like Bell Labs' first paper-clip transistor.....)
"...two small metallic strips containing silver, ...By an electric control pulse, we position a single silver atom into this gap and close the circuit.... The single-atom transistor can drive a tiny current of somewhere between one to eight microAmps."
Not a keen audio device, no. But promising for computing?
What bothers me: 8uA for one electron of charge seems like fabulous transconductance, unless that electron has to be pushed a long way.
The demonstration device is a LOT bigger than "one electron", of course. (Looks an awful lot like Bell Labs' first paper-clip transistor.....)
"Moving metal"... well, like discussing the pronoun for a transgender angel. We are a little beyond everyday experience.
We call Silver a metal. But is a single atom of Silver a metal? Properties like electric conductance are properties of the bulk of a metal. Electrons slide from one atom to an adjacent atom. Thermal conductance and strength are also bulk properties. You can't readily split an atom of Silver; you can more easily rip one Silver atom from another.
Which also points-out a misconception in my initial post. I spoke of "transconductance" in the sense of charge transfer theory. But a natural atom of Silver has no charge! But then how do they move it? Is it really a Silver ion?
We call Silver a metal. But is a single atom of Silver a metal? Properties like electric conductance are properties of the bulk of a metal. Electrons slide from one atom to an adjacent atom. Thermal conductance and strength are also bulk properties. You can't readily split an atom of Silver; you can more easily rip one Silver atom from another.
Which also points-out a misconception in my initial post. I spoke of "transconductance" in the sense of charge transfer theory. But a natural atom of Silver has no charge! But then how do they move it? Is it really a Silver ion?
Member
Joined 2006
Dave's right. According to that, it really is a switch with one atom making or breaking the contact. Is it metal? Good question.
" the atomic switch is entirely controlled by an independent third gate electrode, allowing to open and close a metallic contact between the source and drain electrodes by the gate-voltage-controlled relocation of one single silver atom."
Looks like a tiny relay to me.
" the atomic switch is entirely controlled by an independent third gate electrode, allowing to open and close a metallic contact between the source and drain electrodes by the gate-voltage-controlled relocation of one single silver atom."
Looks like a tiny relay to me.
Member
Joined 2006
THANKS!!
That is a very people-readable explanation of what this is.
He is indeed moving one Silver atom to make/break an electric contact, "a relay". The atom is in an electrolyte-- "Silver plating" on an atomic scale.
To my stone-age mind, the atom shift changes two metal points into one metal path. From two metallic-electron clouds to one.
The atom either moves or it does not, quantized. Apparently (after initial fabrication) the shift voltage and conductance are very well defined.
Taking numbers from two sources, I get an estimate of transresistance near 2K Ohms, 0.5 mS transconductance. This is right in the range of what a BJT will do at these currents. (Natural Law is not violated.) However while a BJT (most charge-controlled devices at very low current) will reduce current to half with a 20mV change, the 1-Atom device cuts current dead with a slightly lower change of voltage. So for switching tasks (computing) it is far more sensitive for the purpose.
Speed is unproven. The data only shows 10-second switching. 5 microsecond response is limited by the test instrumentation. The geometric scale promises "picoSecond" response.
The process seems "clean", and potentially CHEAP. While he uses Gold tabs and Silver points, the use is microscopic. The Gold is just a lab-convenient backing. The Silver *could* perhaps be another metal. The electrolyte is not specified but most suspects will be non-noxious and very little is needed.
The picture suggests something a bright child could build with a home microscope. (The instrumentation to prove it works will be daunting at first.)
Attachments
Member
Joined 2006
A silver atom bridging the gap seems to behave like a metal. Whether an isolated silver atom is a metal might down to a debate about the meaning of "is"; the atom is certainly silver, and every chemist knows that silver is a metal.
PS I guess we could argue about my slightly tongue-in-cheek use of the word "piece of metal" in relation to a single metal atom. I think I can argue that it is a piece, in fact the smallest possible piece.
PS I guess we could argue about my slightly tongue-in-cheek use of the word "piece of metal" in relation to a single metal atom. I think I can argue that it is a piece, in fact the smallest possible piece.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.