Do you know your arsenide from your elbow?

A NEW REPORT claims that the compound semiconductor market is now growing faster than the silicon industry. US consulting firm Kline & Company says in its Global Outlook for Chemicals and Materials in Compound Semiconductors, 2002-2007, that after waiting in the wings for more than two decades while silicon dominated the market, the time has finally come for compound semiconductors based on non-silicon wafers, or wafers that combine silicon with another element such as germanium.

Compound semiconductors go into many types of communications and photonic devices, and growth will be stronger for these devices than for the logic and memory devices that rely on silicon technology, according to preliminary data collected by Kline. In 2001, about $119 billion in integrated circuits were produced, but only $3 billion of this was in the form of compound semiconductors, while silicon accounted for the rest.

Compound semiconductors are more expensive to build; their melting points are lower, which compromises fabrication; and they lack a natural oxide that can serve as a dielectric media. The basic boules from which wafers are cut are much smaller in diameter, and some compounds, such as gallium nitride, are not available at all in bulk boules. So it is not surprising that silicon has been overwhelmingly preferred in semiconductor processing until now.

This imbalance between silicon and compound semiconductors looks set to decrease, however. Compared to conventional silicon-based semiconductors, compounds enable integrated circuits that are faster; can operate at much higher frequencies (300GHz, anyone?); are capable of emitting or detecting visible light and infrared radiation and are both radiation and heat-resistant.

The Kline study focuses on Group IV compounds - mainly silicon-germanium (SiGe) and silicon carbide (SiC) - and combinations of Group III and Group V elements, including gallium arsenide (GaAs), indium phosphide (InP), gallium nitride(GaN).

Four compounds account for most of the current demand for these semiconductors today: Gallium arsenide, the largest at almost three-quarters of the total, followed by gallium nitride at approximately 13%, silicon germanium at 12%, and indium phosphide at 1%. Each material has its niche markets. Gallium nitride emits light in the blue region of the spectrum and is an ideal light source for compact hard disk drives. Indium phosphide is ideal for use in high-speed fibre-optics, while silicon germanium, championed by IBM, is viewed as a power-saving alternative to silicon.

Light-emitting diodes, lasers, UV detectors, solar cells, and many other types of photonic devices depend on compound semiconductors. Compound chips can react to microwaves in real time, converting them to electrical signals. They also emit light and can withstand extreme operating temperatures and radiation (a property put to good use in satellite circuitry). Superior performance in photonics, however, explains most of the current excitement over compound semiconductors. Applications exist today that did were simply not around when gallium arsenide was first developed, especially in digital display appliances and mobile communications. The market has now caught up with this technology, and compound semiconductors are finally poised for strong growth that will exceed that of silicon-based integrated circuits in the next few years.

Andrew Thomas edits III-Vs Review, a monthly magazine focussing on the compound semiconductor industry and claims to know about all this stuff. External links Kline & Company III-Vs Review