With the process' discovery, manufacturers could choose the specific surface chemistry for select areas of porous silicon chips, allowing silicon to selectively respond to specific environmental cues. The resulting capability would enable designers to create selective chemical sensors, according to Jillian Buriak, a chemist at Purdue University (West Lafayette, Ind.).

The procedure can be used to integrate light-emitting devices with silicon chips and may also find uses in new types of drug-delivery systems.

"This reaction only works on nanocrystalline silicon — it belongs to this whole field of nanotechnology where properties such as color and reactivity can depend on size," Buriak said. "By cutting a material down to the nanoscale, you can change its properties. Now we can stabilize the surface of a porous silicon wafer for specific applications, since untreated porous silicon disintegrates too quickly to be useful in many environments."

Working with doctoral candidate Michael Stewart, Buriak also demonstrated for the first time specific mechanisms by which nanocrystals of porous silicon — measuring billionths of a meter in diameter — actually work to emit light.

Porous silicon became a hot research topic over a decade ago, when reseachers found that a standard silicon wafer, when deeply etched with hydrofluoric acid, would take on unusual physical properties. Etching produces an array of silicon columns only nanometers in diameter, to which hydrogen atoms — a residue of the process — are attached.

The actual material was accidentally discovered as a by-product at Bell Labs in the 1950s but was avoided until 1990, when Leigh Canham at the Defense Evaluation Research Agency in England discovered that porous silicon could absorb light, as well as emit light when "pumped" by ultraviolet light. That subsequently led to the discovery that porous silicon layers could also emit light when stimulated by an electrical current. The efficient emission of light is something that pure crystals of silicon cannot do, at least at any efficiency level that might lead to practical devices.

The discovery of porous silicon's light-absorbing and -emitting properties prompted researchers to attempt to use light to internally transmit data inside chips, instead of encoding data in electron streams or depending on external gallium arsenide chips to translate electricity into light. However, porous silicon proved to be too unstable to be commercialized, since many normal environmental chemicals — such as oxygen and humidity — corrode the porous silicon, thereby nixing its light-absorption and -emitting capabilities.

"Left untreated, porous silicon is too fragile to hold up in the presence of oxygen and water molecules, which interact with the surface of porous silicon to create a glass-like coating that disrupts its photoluminescent properties," Buriak said.

In 1999 Buriak reported developing a wet-chemistry methodology for functionalizing the entire surface of a porous silicon wafer, but her latest report shows how to employ an inexpensive photolithography step. This latest method allows researchers not only to stabilize the surface, but also to custom-tailor different parts of it to have different responses to specific chemical cues. This capability paves the way for use of the material as sensors.

Carbon-silicon bond

By simply shining moderately intense white light from a tungsten source onto the porous silicon surface for 30 to 60 minutes, the photons cause porous silicon's electrons to jump to a higher energy level, leaving a positively charged hole behind. If the chamber is filled with alkynes — chemical compounds that contain hydrogen and carbon — the holes react with the chemicals to form a carbon-silicon bond that stabilizes the surface.

"This is a very clean, very practical reaction that allows us to stabilize the surface without the need for special equipment," said Buriak.

One of the major benefits of this technique, compared with her previous wet-chemistry approach, is that masks can be used to pattern the stabilizing coat in different ways, on different parts of the chip.

"This method can be exploited to develop new types of sensing devices — for example, if you need a device that responds in a very specific way, you can functionalize the surface of porous silicon to sense and bond only to specific sites," Buriak said.

In the lab, Buriak demonstrated how the porous silicon chips can be patterned into defined areas that prompt specific chemical reactions on different parts of the chip, enabling the integration of light-emitting devices on silicon chips for applications in optoelectronics and bioanalysis. Test chips show specific "readouts" on a chip, changing color in the presence of specific molecules.