late 1800's came expanded production of
ferrosilicon. By 1949, the first silicon
pure enough for use in transistors and
other  semiconductor  devices  was
produced.     Beginning  in   1951,
commercial production of high-purity
silicon gave impetus to the rapidly
expanding electronics industry.
Polycrystalline silicon (ultrapure,
99.9%) was a product largely confined to
research until 1956, when its use in
manufacturing    components    for
commercial applications was estimated to
have exceeded that used for research. By
1958, an estimated 20,412 kilograms of
high-purity silicon valued at $16 million
was consumed by the electronics industry
for the manufacture of diodes, rectifiers,
and transistors.
Geology-Resources
The United States has an abundance of
silica deposits for the production of
ferrosilicon and silicon metal. For the
production of these materials, 98% to
99% purity is preferred. However, trace
amounts of aluminum and iron also are
acceptable.  Physically, the material
should not contain fines and should not
crumble easily.
Technology
Silicon metal and ferrosilicon are
produced by the reduction of silica (SiO2)
to silicon (Si) in a submerged arc electric
furnace. A typical charge consists of
silica as beneficiated quartz or quartzite;
coal, coke, or charcoal as a reductant;
wood chips for porosity; and, when
producing ferrosilicon, iron in the form
of steel scrap or iron ore. During the
furnace operation, raw materials are
periodically charged into the top of the
furnace, and the molten metal or alloy is
periodically tapped at the bottom of the
furnace and cast into chills or ingots.
The material is then crushed to specific
size requirements starting at about 20
centimeters down to 200 mesh.
Metallurgical-grade silicon metal is the
starting material for high-purity silicon
consumed by the electronics industry.
First, silicon metal is used to produce an


intermediate product such as silicon
trichlorosilane  (TCS).   A   vapor
deposition process is then used to form a
rod or boule of high-purity polycrystalline
silicon (polysilicon) from the TCS. A
single crystal boule is made from
polysilicon by two methods.     The
Czochralski (CZ) method uses a seed
crystal to grow a rod of single crystal
silicon from a molten crucible of
polysilicon. The Float Zone (FZ) method
uses an induction coil to produce a
molten zone within a rod of polysilicon.
The coil starts at one end of the boule
where a seed crystal starts the crystal
growth. The coil moves slowly down the
boule, allowing a single crystal to be
formed. After being sliced into wafers,
the silicon is used by the electronics
industry to produce integrated circuits.
Recovery of silicon from secondary
sources is not normally practiced. The
only secondary possibility is recovery
from scrap metal. However, any value
of contained silicon would be incidental
to the value of the primary metal. In
1993, the average price for ferrosilicon
was about $0.41 per pound and the
average price for silicon metal was about
$0.66 per pound.    For the future,
recycling of silicon in the form of
ferrosilicon and silicon metal is expected
to be insignificant.
Byproducts and Coproducts
Silicon metal and ferrosilicon furnaces
produce a material that is referred to as
silica fume, silica dust, or microsilica.
Originally, this material was considered
of little or no value.     However,
microsilica is now used as an additive in
a number of different products, including
high-strength concrete.
Economic Factors
Production of silicon metal and silicon
alloys is extremely power intensive,
requiring a power input, for some
operations, of up to 14,000 kilowatt-hours  l
per ton of silicon content.
The location of ferrosilicon and silicon  i
metal smelters is normally determined by  j
balancing  marketing  costs  against


processing costs. Consequently, on a
worldwide basis, smelting plants are near
iron and steel industries or trade routes
serving them, where cheap fossil fuel or
electric energy is available or where a
combination of these factors exists. The
availability of hydroelectric power in
large quantities at competitive prices, the
lack of a sizable steel industry, and the
presence of deep natural and ice-free
harbors contributed to Norway becoming
the largest exporter of ferrosilicon in the
world. The high cost of electric power,
proximity to world trade routes, and the
demands of an established iron and steel
industry have made Germany, Japan, and
the United Kingdom three of the world's
largest importers of silicon metal and
ferrosilicon. During the 1980's, Brazil
established itself as a major producer of
silicon ferroalloys and silicon metal,
joining countries such as Canada, Iceland,
and Norway that have low power costs.
In contrast, high power costs have forced
Japanese ferroalloy producers to cease
production of silicon metal and drastically
curtail silicon ferroalloy production.
By 1960, consumption of high-purity
silicon by the electronics industry was
more than 31,751 kilograms. However,
the market was becoming saturated with
material resulting from increased output
by new plants and the expanded
production from older ones. Prices were
in the range of $150 to $330 per pound
compared with an average price of $330
per pound for metal consumed in 1958.
By the mid-1960's, improvements in
silicon devices and the increased use of
integrated circuits in the computer and
aerospace industries had contributed to an
accelerated consumption of high-purity
silicon.
Demand for metallurgical-grade silicon
alloys and metal is determined by the
level of activity in the steel, ferrous
foundry, aluminum, and chemical
industries and is little affected in the short
term by prices for these materials. As a
result, prices tend to vary widely with
changes in demand and supply.
From the period 1955-69, the price of
silicon alloys and metallurgical-grade
metal remained stable. During this 15-
year period, the domestic producer price


1024






SELICON-1993