Created: A One-Atom Transistor!

Scientists have hit a new low when it comes to size! The newest size of a transistor is just one atomic radius and it is made of phosphorus. A group of physicists from the University of New South Wales and Purdue University have created a transistor out of a single phosphorus atom embedded in a silicon crystal. Moore’s law has been broken, once and for all!

An STM image of the phosphorus atom placed on a Silicon substrate. The surrounding electrodes are the drain and source (see next image). (Courtesy: Arstechnica)

Quantum Mechanics and Choices!

What more, the transistor, instead of relying on the binary electronic states of ‘on’ and ‘off’ can rely on a superposition of quantum states, using so-called qubits. Qubits don’t represent just one of the two positions, but a multitude of all the possibilities, as prescribed by quantum mechanics.

Qubits will help realize the making of quantum computers, and of this, scientists are sure! The computers will be extremely small (for obvious reasons), very fast (information relay over tiny scales and the huge number of qubit states to utilize), energy efficient (no heat dissipation) and be able to solve a huge number of problems within a fraction of a second.

Moore’s Law

Even Moore’s law is happily in trouble! Moore’s law states that every eighteen months, the density of transistors on a chip doubles! Moore’s law has been scaled down to the scale of one atom! It is safe to say that it cannot go down any further.

The colour gradient image of the potential across the neighbourhood of the single phosphorus atom. The G, S and D refer to the gate, source and drain. So-called Field Effect Transistor (FET's) are supposed to regulate the passage of current from the source to the drain, using the voltages applied at the gates (Courtesy: Nature article)

Andreas Heinrich, a physicist at IBM, says the following about this work:

This is at least a 10-year effort to make very tiny electrical wires and combine them with the placement of a phosphorus atom exactly where they want them.

The deposition of the single atom at a precise position was done using a scanning tunneling microscope (STM). The STM was used to ‘cut’ the ‘groove’ into the silicon. Phosphine gas was then used to deposit one atom of phosphorus. It was then covered with a few layers of silicon.

The work appears in an issue in Nature Nanotechnology (link).

Dreams of using a device to relocate just one atom of a substance on a substrate are finally coming true! One of the principal dreamers was Richard Feynman. He would be proud!

Implementing these gated devices as an array of switches to make a working circuit is the present challenge. The Next is already here!

Huge Triumph at Nano Scales: Ultra-Small, Super Fast Single Electron Transistors Created

A team of researchers at the University of Pittsburgh, collaborated with another team from the University of Wisconsin at Madison, to create a transistor, which is just an atomic diameter in size and can be switched on and off by just one or two electrons. These transistors can further be used as solid state devices, such as fast quantum processors (which might replace the current Si processors) and extremely dense memory devices.

Lead researcher Prof. Jeremy Levy of University of Pittsburgh further emphasized that these new materials might be used to create  substances like high temperature super-conductors.

Single Electron Transistor
A Single Electron Transistor. (The red part is the island for housing the electrons)

The Device: A Single Electron Transistor

The device is basically an island, 1.5 nanometers (nm) in diameter, which is made of metal oxide. This island can house zero, one or two electrons only allowing it to be in very specific quantum states. Such exotic materials have never been made. Nanowires, 1 to 1.2 nm thick, carry electrons across the island, thus allowing conduction.

Existent Single Electron Transistors (SET) generally have a size of about a micron, so the miniaturization is a thousand-fold.

The idea has been there for a long time. In order to build an SET, Graphene has also been tried. No real success has been achieved so far, primarily because Graphene doesn’t have a strict off state. This becomes crucial when considering transistors switched on and off by single electrons.

Single Electron Transistor
The SET as FET device (Earlier fabrication)

The wonder

The wonder of the device is its extreme sensitivity to the presence of an electric charge. Further, the oxide base is ferroelectric and can retain electrons even when the device is switched off. If the number of electrons in the island is controlled, the device can act as a memory device, in some state, 0 or 1. Stacking many such islands together can create an ultra-dense solid-state memory device, which can be altered by passage of minute amounts of electric current.

Fabrication in the pure state is a problem currently, but this is usually the case when a new solid state device is made for the first time ever. Today’s improved Atomic Force Microscopy (AFM) Techniques allow engineers to fabricate materials directly on the nano-scale with high precision and this is the current line of attack.

If made on a commercial scale, these transistors can make a computer out of quantum processors, having unthinkable speeds, capable of performing calculations in a day, which are estimated to take hundreds of years on today’s supercomputers.


IBM Creates World’s Fastest Transistor Using Graphene

IBM on Tuesday, 12th April, announced that they have made the world’s fastest transistor using graphene and also hinted that they might go into commercial production very soon. This is major news, as graphene might revolutionize the current semi-conductor industry scenario. Graphene may even be good enough to replace silicon, the standard material used in all of today’s semi-conductor devices, in the near future.

What is Graphene? How is it produced?

Graphene is a mono-layer of carbon, with the atoms in hexagonal configuration. Each of the carbon atoms has bonds with three neighboring carbon atoms, maintaining, what is called, a sp2 hybridization. Basically, it is one layer of graphite.

Graphite Structure
The Structure of Graphite

It is produced in the most mundane way you can think of. A pure crystal of graphite is repeatedly stripped off using Scotch Tape, until, about 50 or more repetitions later, graphene is found buried amongst poly- or bi-layered graphite (Pic below). This has to be verified, and can be done so optically, after the extracted graphene is mounted on a Silicon Dioxide (SiO2) substrate of correct thickness. Often, Raman spectroscopy is used for verification.

Graphene under light
How Graphene looks under light and the right thickness of Si layer. (Courtesy: Graphene Industries)

Other methods, which allow graphene to be grown for commercial purposes, are also known. Primary among these is Chemical Vapour Deposition (CVD), in which carbon vapour (obtained from carbon rich substances like acetylene) is deposited on a Ni or Cu substrate.


How did IBM do it?

Graphene grown directly on a SiO2 substrate suffers from the problem of scattering of electrons, resulting in the deterioration of the transport properties and also producing non-uniformity across the SiO2 wafer. The IBM team used a novel substrate called Diamond-like Carbon’ (DLC) on top of the SiO2 layer so as to reduce the scattering. DLC is loosely amorphous (i.e. powdered) diamond. It has all atoms in sp3 configuration (i.e. each atom bonds to four neighbors) and tetrahedral arrangement, but, being a powder, lacks any fracture planes. Thus, having the necessary properties of diamond, it is also flexible and can easily be used as a coating over a substrate.

Graphene couples weakly to the DLC layer and this greatly reduces the scattering, as also the temperature dependence of the material. In fact, the transport properties of DLC-grown graphene remains almost (maybe, exactly) temperature independent right up to 4.3K (which is minus 269 Centigrade).

The IBM team took graphene, made using CVD on a Cu layer, and, after protecting with polymethylmethacrylate (PMMA), dissolved the Cu layer using FeCl3. Then, the PMMA-graphene was transferred to a DLC layer on the SiO2 substrate and the PMMA was got rid of. Raman spectroscopy was used to verify the quality of the graphene layer.

Graphene lends itself readily to fabrication of Field Effect Transistors (FETs). By constructing the gate, drain and source contacts using pure metal and properly calibrating their device, the IBM team achieved input-output characteristics similar to a Si FET. Further, they achieved very high switching speeds up to 26 GHz for a 550 nm long device.


Arrangement of a FET Graphene transistor
Schematic representation of a graphene FET (Courtesy: Nature)

What a graphene transistor actually looks like

How will it score over Si transistors?

Graphene transistors will be ideal in radio frequency (RF) range signals, due to their high switching speeds. Unlike Si transistors, their properties don’t degrade at low temperatures. This means that there will be no unnatural change in transport properties as the temperature is varied.

The problem graphene transistors have is that they have low on-off voltage ratio. However, that is not a very strict condition for RF communication. A more serious problem is that of high contact resistance, which cannot be minimized, unlike MOSFET‘s in case of Si.



IBM reports that transistors with cut-off frequencies (the frequency at which the current gain becomes unity) as high as 155 GHz have been achieved on a 40 nm device using short gate lengths.

A figure of merit is the product of cut-off frequency and gate length. IBM reports the figure of 13 GHz mm for the 550 nm device, which beats the highest value of 9 GHz mm for Si MOSFETS by a long margin.

The future may be here already.