GaAs vs. Si from wikipedia
Comparison with silicon
GaAs advantages
Some electronic properties of gallium arsenide are superior to those of
silicon. It has a higher saturated electron velocity and higher electron
mobility, allowing gallium arsenide transistors to function at frequencies
in excess of 250 GHz. Unlike silicon junctions, GaAs devices are relatively
insensitive to heat owing to their wider bandgap. Also, GaAs devices tend to
have less noise than silicon devices, especially at high frequencies. This
is a result of higher carrier mobilities and lower resistive device
parasitics. These properties recommend GaAs circuitry in mobile phones,
satellite communications, microwave point-to-point links and higher
frequency radar systems. It is used in the manufacture of Gunn diodes for
generation of microwaves.
Another advantage of GaAs is that it has a direct band gap, which means that
it can be used to emit light efficiently. Silicon has an indirect bandgap
and so is very poor at emitting light. Nonetheless, recent advances may make
silicon LEDs and lasers possible.
As a wide direct band gap material with resulting resistance to radiation
damage, GaAs is an excellent material for space electronics and optical
windows in high power applications.
Because of its wide bandgap, pure GaAs is highly resistive. Combined with
the high dielectric constant, this property makes GaAs a very good
electrical substrate and unlike Si provides natural isolation between
devices and circuits. This has made it an ideal material for microwave and
millimeter wave integrated circuits, MMICs, where active and essential
passive components can readily be produced on a single slice of GaAs.
One of the first GaAs microprocessors was developed in the early 1980s by
the RCA corporation and was considered for the Star Wars program of the
United States Department of Defense. Those processors were several times
faster and several orders of magnitude more radiation hard than silicon
counterparts, but they were rather expensive.[8] Other GaAs processors were
implemented by the supercomputer vendors Cray Computer Corporation, Convex,
and Alliant in an attempt to stay ahead of the ever-improving CMOS
microprocessor. Cray eventually built one GaAs-based machine in the early
1990s, the Cray-3, but the effort was not adequately capitalized, and the
company filed for bankruptcy in 1995.
Complex layered structures of gallium arsenide in combination with aluminium
arsenide (AlAs) or the alloy AlxGa1-xAs can be grown using molecular beam
epitaxy (MBE) or using metalorganic vapor phase epitaxy (MOVPE). Because
GaAs and AlAs have almost the same lattice constant, the layers have very
little induced strain, which allows them to be grown almost arbitrarily
thick. This allows for extremely high performance high electron mobility,
HEMT transistors and other quantum well devices.
Silicon advantages
Silicon has three major advantages over GaAs for integrated circuit
manufacture. First, silicon is abundant and cheap to process. Si is highly
abundant in the Earth's crust, in the form of silicate minerals. The economy
of scale available to the silicon industry has also reduced the adoption of
GaAs.
In addition, a Si crystal has an extremely stable structure mechanically and
it can be grown to very large diameter boules and can be processed with
very high yields. It is also a decent thermal conductor, thus enabling very
dense packing of transistors that need to get rid of their heat of operation
, all very desirable for design and manufacturing of very large ICs. Such
good mechanical characteristics also makes it a suitable material for the
rapidly developing field of nanoelectronics.
The second major advantage of Si is the existence of a native oxide (silicon
dioxide), which is used as an insulator in electronic devices. Silicon
dioxide can easily be incorporated onto silicon circuits, and such layers
are adherent to the underlying Si. GaAs does not have a native oxide and
does not easily support a stable adherent insulating layer.
The third, and perhaps most important, advantage of silicon is that it
possesses a much higher hole mobility. This high mobility allows the
fabrication of higher-speed P-channel field effect transistors, which are
required for CMOS logic. Because they lack a fast CMOS structure, GaAs logic
circuits have much higher power consumption, which has made them unable to
compete with silicon logic circuits.
For manufacturing solar cells, silicon has relatively low absorptivity for
the sunlight meaning about 100 micrometers of Si is needed to absorb most
sunlight. Such a layer is relatively robust and easy to handle. In contrast,
the absorptivity of GaAs is so high that a corresponding layer would be
only a few micrometers thick and mechanically unstable.[9]
Silicon is a pure element, avoiding the problems of stoichiometric imbalance
and thermal unmixing of GaAs.
Silicon has a nearly perfect lattice, impurity density is very low and
allows to build very small structures (currently down to 25 nm). GaAs in
contrast has a very high impurity density, which makes it difficult to build
ICs with small structures, so the 500 nm process is a common process for
GaAs.