© 2011-2017 Evince Technology Limited, Netpark, Sedgefield, County Durham, UK, TS21 3FD

item4
ETLlogoinverseversion2
APPLIED DIAMOND ELECTRONICS

Making A Practical Diamond Device

For too long, people have been trying to make electronic devices in diamond by trying to replicate the techniques used to make silicon devices. It doesn't work.

Diamond has some phenomenal properties. Many of these are due to the fact that the carbon atoms that make diamond what it is, are held together by very tight chemical bonds in a close packed structure. Compounded by the small size of the carbon atom this makes it challenging (but not impossible) to dope diamond in the same way you would with silicon. A good analogy is that with most dopants it's rather like trying to squeeze a tennis ball into a tight packed array of ping-pong balls. The resulting distortion to the regular crystal structure of diamond compromises the performance of the material.

To make practical working devices in diamond, you need to start by tearing up the semiconductor rulebook. Our technology takes diamond, one of the world's best electrical insulators, and turns it at will into a controllable conductor. We do this by exploiting one of diamond's unique properties, which is that it is virtually transparent to free electrons. This phenomenon is widely known and has been used to make switching devices in the past. To make such a device the free electrons were generated from an external source such as a hot filament (just as in a thermionic tube) and therefore needed to operate under a vacuum.

HFDiamondDevice

What these devices did show was that they could handle very high power levels (kilo-amps per cm2) and they were capable of very fast picosecond switching times. As everyone who's ever had to change a hot filament lightbulb will know, the filaments are fragile. So an improvement to the original design was proposed. In the Mark II version, the hot filament was replaced by a cold cathode. Here lots of needle like structures (just like those room ionisers that were so popular in the 1990's) are engineered at the microelectronic scale using semiconductor fabrication techniques. This type of cathode generated electrons by a process known as field emission and had the big advantage that you didn't need to apply large amounts of power to heat a filament.

HFDiamondDevice1

The problem with this Mark II idea was that a vacuum was still required, however all vacuum devices have a fatal flaw which is that the vacuum is never perfect. Occasionally electrons impact and ionise trace gas molecules which are typically 60,000 times heavier than an electron. These ionised gas molecules are attracted back to the cathode points, accelerating in the electric field and when then hit the cathode cause a small amount of damage. Hence slowly but surely the cathodes erode and die. Even so, people are still looking at advanced versions of vacuum devices as a solution to higher speed computation and ultra high frequency (100's to 1000's GHz) needs.

Evince's innovation has been to realise that the superior dielectric properties of diamond can sustain the conditions needed to support field emission within the material itself. Using proprietary surface engineering techniques we have developed a process by which we embed tens of thousands of nano-scale electron emitters into a diamond substrate that act as the electron source necessary to switch diamond from being an insulator to an efficient conductor...

... with no vacuum required!

HFDiamondDevice1a

The result is a new class of devices - unique to diamond - that exhibit the superior high voltage and high frequency switching properties that vacuum electron devices can offer but at microelectronic scale.