Chuck Wickersham, PhD Metallurgical Effects in Sputtering Targets
Chuck Wickersham, PhD
Cabot Performance Materials

Sputtering is an interesting topic, because it involves a very broad spectrum of both materials and requirements. Sputtering was first described 150 years ago by British scientist W. R. Grove, who observed deposition of a thin film while conducting experiments with a vacuum aparatus. It was not until the 1970s that the phenomenon became a widely used process with the development of the "Magnetron Sputtering Process" which enabled reasonable rates of deposition.

In the basic sputtering process, a "target" (which is really the source of the metal to be deposited) is held with a negative charge in a vacuum with a low pressure of argon. A large amount of power is applied (10 - 20 Kw) which generates a plasma at the face of the target. Individual atoms of the target material reach energy levels sufficient to escape, and become deposited on other surfaces in the chamber, most desirably on the item to be coated. The efficiency of the process is improved when a magnetic field is applied to the outside (cold) surface of the target while this is going on, and this development is known as the Magnetron Sputtering Process. In a typical aluminum sputtering process for IC creation on substrates, a deposition rate of about 1 micron/min is achieved.

Target materials must meet many requirements:

-- Targets must be suitable vacuum container components
-- Targets must handle large power (water cooling is used)
-- Magnetic properties are a factor
-- Cyclic Fatigue (1000 - 30000) cycles from vacuum to 1 atm is typical
-- Thermally conductive (Target must not melt during operation)
-- Mechanically strong (15 psia on area of target when under vacuum and hot)
-- Grain Size (has significant effect on deposition)
-- Crystallographic texture
-- Purity (extremely critical in IC applications)

Major applications of sputtering include:
-- Computer Storage Media (magnetic - Co film and optical - Al film)
-- Ink Jet Printer heads - Ta film
-- Flat Panel Displays (particularly front "InSnO" film which is transparent and conductive)
-- Semiconductor Devices (about 60% of total dollar market)

Currently, silicon based ICs use aluminum (Al) film for conducting paths with a Ti junction to SiO2. As greater speeds are required in IC circuits, this will change to copper (Cu) film, with a much more complex Tantalum (Ta) interface (also sputtered) required to be compatible with the Silicon. Conduction paths are typically 1 micron wide, and junction distances (source to drain in a transistor) are about 0.13 micron with thin film used as the connection. Currently the steps of production are to: 1) thermally generate a SiO2 dielectric layer on the Silicon substrate, 2) Sputter Ti and Al, 3) Etch the Al to the circuit pattern, 4) Deposit SiO2. With the new copper film technology the steps become more involved since the substrate and dielectric are etched to the pattern, Ta and Cu are then sputtered, and finally the deposition is planerized back to the pattern.

High purity sputtering targets are made by vacuum melting, then mechanical deformation and heat treatment for properties. They are then bonded to a high strength thermally conductive backplate, inspected (NDT), machined, cleaned, and packaged to Class 100 clean room criteria. The whole process is terribly inefficient. With aluminum being used for ICs, for each 100 g of vacuum cast aluminum, only 2 grams will end up usefully deposited in the IC. About 50% is lost in manufacturing the targets, and another 48% (of the original) is lost either as end-of-life target material or misplaced film during the sputtering process. On a "per Kg" basis the values of the aluminum increases from $1.50 to $500,000.00 due to the complexities of the process and expense of the equipment. There are many opportunities for improvement in this process.

Finally, the furture prospects for sputtering target manufacture are not very bright. It is a mature technology. Growth of the industry has been good but is currently declining. There is currently overcapacity in production. Prices are declining. New demands are for more costly materials. This industry is ripe for consolidation. The opportunity that exists will probably be in a major refinement of the process itself to improve efficiency.