Diamond transistor technology

Saturday 11th July 2009
Dr David Moran of Glasgow University: “The motivation at this stage is to examine the intrinsic scalability of device operation with reduced gate length”

In 2007 Dr David Moran at Glasgow University was awarded an EPSRC Advanced Research Fellowship, funding five years of exploration into high power/high frequency diamond FET technology. Such a device, reports Gail Purvis, is highly desirable for specific high performance niche markets, including terahertz imaging, medical applications, and automotive collision detection use.

The original concept was to take his expertise, (as a specialist in short gate length III-V HEMT technology predominantly working with GaAs and InP materials creating compound transistors) over into developing a transistor from diamond. With its large electronic band gap, exceptionally high thermal conductivity, and high tolerance to radiation, diamond is that holy grail electronic material as produced by Element Six.

James Watt Nanofabrication Centre
The work would also fully exploit the on-site James Watt Nanofabrication Centre facilities, especially its recently commissioned Vistek VB6 electronic beam lithography machine. Moran’s aim is to create from several mm square diamond slivers, a stable transistor technology, outperforming other materials in both high power and high frequency.

The diamond square sliver, (right) by contrast to the usual circular substrates that process engineers have to deal with. There is an opinion that had silicon wafers been developed in squares rather than circles, electronics would be quite different today!

The result of  eighteen months work has now resulted in two different runs, and some 50 different devices that have indeed currently yielded the world’s smallest working diamond transistor, with a gate length of just 50nm.

Moran stresses that motivation at this stage is to 'examine the intrinsic scalability of device operation with reduced gate length,' or to see how small a gate length they can achieve while the device still operates as an efficient transistor.

"Speed and power performance is something we are continually working on with these devices and is something that is challenging to maximise given the more complex processing of these reduced device dimensions."

There are, says Moran, three groups that have done considerable diamond field effect transistor research: Ulm University, in Germany and Waseda University and NTT both of Japan. Each have faced the initial challenge of making diamonds conduct.

One approach has been to heavily delta-dope the material with boron to introduce holes as charge carriers. This creates stability, but “unlike HEMT technology, it is very difficult to physically separate the resultant two-dimensional  hole layer from the dopant ions and as such our prized intrinsic mobility goes through the floor,” he says.

"This, however, is by no means the end of the story, as although low field mobility is a nice general metric for material potential, it is at the end of the day carrier saturation velocity beneath the gate that will really decide device speed."

The other method is to hydrogenate the diamond, which leads to the transfer of valence band electrons out from the surface of the diamond into an air-formed adsorbate layer. This "surface transfer dopant" process creates a shallow layer of holes about 10nm below the surface of the diamond.

Short gate length transistor
This is the route Glasgow University have adopted, using the services of Heriot Watt for the hydrogenation which leads to the surface being sensitive to the type of contact metal used and as such makes it straightforward to produce both ohmic and Schottky contacts, dependent upon metal work function layer.

With a short gate length transistor, Moran says, the need is to bring the conduction closer to the gate. In contrast to III-Vs, which suffer inherent "surface Fermi pinning" which makes it difficult to accumulate charge close to the surface, hydrogenated diamond intrinsically offers a shallow conducting layer.

And, he adds, diamond is absolutely the best material for extremely high resolution electron beam lithography work with its minimum backscatter. Work is now in hand to push the record 50nm gate down to a mere 10nm working transistor.

Moran is already deeply involved in the increasingly complex balancing trick, with its trade-offs to produce a transistor, with reduced parasitics, improved ohmic contacts, higher power, higher frequency and a shrinking gate length.

Another aspect that will emerge over the next 42 months work is that hydrogenated diamond transistors rely on air to be conductive. This leads to stability issues which still need to be addressed before a robust diamond transistor technology can be commercialised. Moran makes a passing reference to C60 fullerene passivation as one route that might be explored.

Circuit plans
At present however, what he has achieved beyond the smallest diamond transistor, as both active lecturer and one-man-diamond-nano-transistor-producing-band, is the prospect of some (“still to be trained”) PhD students to help in the work: “It’s not just problems of stability still to address but the developed nano-transistor will also need to be incorporated into circuit plans.”

Moran’s achievement with diamond to date is intriguing and obviously far reaching. It follows on from a project completed in 2008, the micro-machined diamond device initiative with Strathclyde University and the Institute of Photonics developing a ‘toolkit’ for micro and nanoscale manufacturing technology for the next generation of devices on synthetic single crystal diamond for which Element Six is a world leader using CVD techniques.

That initiative led to improved technology for synthesis and processing in production of substrates and epitaxial layers with atomic scale low roughness surfaces. It enabled deposition of thin layers of boron doped diamond at the nanoscale where MESFETs use layers sandwiched between two undoped intrinsic layers to support transistor action, and led to a robust, reproducible dry etching technology suitable for transistor device fabrication.

It also led to Element Six forming Diamond Microwave Devices, now based at Leeds Innovation Centre in the University of Leeds to develop the first high frequency, high power diamond transistors, according to Christopher Ogilvie Thompson, commercial business manager at Element Six and chief executive at DMD.

Technology skills at DMD includes Dr Wolfgang Bosch (CTO Filtronic Integrated Products) Richard Balmer (QinetiQ GaN growth for HFET & MMIC) and Ian Friel, responsible for development of a 'tool-kit' of processing techniques suitable for diamond electronic device fabrication.

Feature courtesy: EM&P

Custom Search

Scotland, Computer News in Scotland, Technology News in Scotland, Computing in Scotland, Web news in Scotland computers, Internet, Communications, advances in communications, communications in Scotland, Energy, Scottish energy, Materials, Biomedicine, Biomedicine in Scotland, articles in Biomedicine, Scottish business, business news in Scotland.

Website : beachshore