To: maceng2 who wrote (418 ) 9/16/2003 2:36:59 AM From: maceng2 Read Replies (1) | Respond to of 1417 Diamond chips sparkle after breakthroughs in n-doping Steve Bush Two recent developments have bought diamond semiconductor devices closer to reality.electronicsweekly.com Diamond has an extremely high thermal conductivity, can withstand high electric fields, and can be made into a semiconductor - ideal for power devices, one would think. Unfortunately, although it can be p-doped with boron, n-doping is proving to be a problem. Japanese scientists have managed to n-dope using phosphorus, but the most recent development is n-doping using the p-dopant boron. Jacques Chevallier of Centre National de la Recherche Scientifique in France and colleagues from France and Israel exposed boron-doped diamond layers, grown by chemical vapour deposition, to deuterium plasma. For an unknown reason, probably the creation of electron-donating boron-deuterium complexes, the diamond becomes n-type. The process is thought to be reversible with heat driving out the deuterium and reverting the material to p-type. Details have been published in the journal Nature Materials. "If it works like the Nature article, it will be very interesting. You could change from n-type to p-type in one layer and make a transistor," said an industry insider who did not want to be named. Over in Belgium, at the Institute for Materials Research (IMO) of the Limburgs Universitair Centrum, researchers have moved phoshorus doping on to polycrystaline diamond. "With monocrystal diamond, substrates are only 2x2mm, or 5x5mm maximum," said scientist Dr Milos Nesladek. "Polycrystaline can be three inches." Nesladek and collegues have produced poly-diamond p-n junctions with a seven decade forward-reverse current ratio, he said. Voltage drop in the forward direction is around 4.6V from the material's 5.5eV bandgap. Suitable insulators are being developed which could lead to diamond Mosfets. "Calcium fluoride is a possibility," said Nesladek. By bonding oxygen molecules to the diamond surface, a thin insulating layer can be formed. "Nano Fets have been made in Japan using this," said Nesladek. In February, scientists in Germany demonstrated an in-plane gate transistor using the surface properties of diamond, with oxygen and hydrogen (p-type) termination. Another thing on the horizon which could speed diamond semiconductor making is the availability of large synthetic diamonds which avoid the vagaries of natural diamond. Massachusetts-based Apollo Diamond is making "wafers ranging from 3mm square to 10mm square" and anticipates having 25mm square wafers shortly. These wafers, made using chemical vapour deposition, are available from 0.25 to 4mm thick. Thermal conductivity is above 3kW/mK, "almost double that of polycrystalline diamond, and exceeding the best mined diamond", it said.