By James Mitchell Crow of Nature magazine

It might be among the hardest materials known, but place a diamond in a patch of sunlight and it will start to lose atoms, say a team of physicists in Australia. The rate of loss won't significantly trouble tiara wearers or damage diamond rings, but the discovery could prove a boon for researchers working to tap diamond's exceptional optical and electronic properties.

Many of the newest uses of diamonds, from laser light emission to quantum communication and computing, require micro- or nano-features to be built into the surface of the diamond. Physicist Rich Mildren and his team at Macquarie University in Sydney have now shown that beams of ultraviolet (UV) light offer a particularly gentle way to do just that.

Diamonds are usually etched by laser in a process called ablation, which burns atoms from the surface but leaves behind a rough, damaged area more like that of graphite than of diamond. Mildren and his colleagues show that by cutting the pulse power of the laser, a process called desorption takes over, with excited carbon atoms popping off the surface to leave smoothly etched diamond behind.

Mildren and his team discovered the effect by accident while developing diamond-based lasers. "We wanted to show that diamond can operate at wavelengths that other materials cannot, and UV is one such region," says Mildren. The team produced a diamond laser that successfully emitted UV light--but only for about 10 minutes, after which it would always stop working. "It turned out that we were desorbing carbon atoms, making little pits in the diamond surface. It was bad news for laser performance, but good news for this other research direction."

Diamond dilemma

Exactly how the desorption process works is still to be determined, but Mildren has a couple of theories, published this week in Optical Materials Express. The first hint is that the process requires a diamond surface covered with oxygen atoms. The second is that it requires two photons to release one carbon atom. When two photons hit a diamond, they produce an exciton, an excited electron-hole pair, inside the diamond that can diffuse through to the surface, where it could set a carbon atom free. "The energy of the exciton is more than ample for a carbon monoxide molecule to pop off," says Mildren. "It is also possible that the UV is exciting the surface directly, leading to bond breaking and carbon monoxide coming off too."

Although previous research had indicated that diamond can be etched in this way, Mildren's work is the first to probe some of the details of the process, and shows that the relationship between laser pulse power and etch speed is highly linear, which provides control over the etch rate. The process also seems to have no lower threshold, meaning that carbon loss will occur even in ambient sunlight. However, the rate of loss is very slow--even a typical mercury UV lamp in a lab would take around 1,000 years to remove a microgram of diamond.

The slow rate of atom loss limits its application as a technique for etching tiny features into diamond surfaces, says physicist Steven Prawer, who heads the materials research institute at the University of Melbourne in Victoria, Australia, where he studies diamond-based devices. But that's not to say it isn't useful, he adds. "It is more a polishing technique rather than a scribing technique--but you might be able to combine it with ablation to polish an etched surface."

A smooth surface is particularly important for the diamond single-photon emitters that Prawer has been pioneering for quantum communication applications that can transmit information. Photons generated by any defects in the diamond are emitted in all directions, so many of them lost. "To capture the light, we would like to create photons inside a waveguide," says Prawer. "But etching the waveguide often leads to a very rough surface, which causes light scattering--so you lose whatever gains you would have made. A technique to smooth the etched surface would increase the count rate, which is what decides the rate of data transfer."

This article is reproduced with permission from the magazine Nature. The article was first published on July 15, 2011.